From 843debb889c7a95c7f591acaed185734694b0ff7 Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:30 +0300 Subject: RDMA/core: Commonize RDMA IOCTL declarations location This patch provides one common file (rdma_user_ioctl.h) for all RDMA UAPI IOCTLs. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/Kbuild | 1 + include/uapi/rdma/hfi/hfi1_user.h | 2 +- include/uapi/rdma/ib_user_mad.h | 4 +--- include/uapi/rdma/rdma_user_ioctl.h | 43 +++++++++++++++++++++++++++++++++++++ 4 files changed, 46 insertions(+), 4 deletions(-) create mode 100644 include/uapi/rdma/rdma_user_ioctl.h diff --git a/include/uapi/rdma/Kbuild b/include/uapi/rdma/Kbuild index 82bdf5626859..541311115505 100644 --- a/include/uapi/rdma/Kbuild +++ b/include/uapi/rdma/Kbuild @@ -1,5 +1,6 @@ # UAPI Header export list header-y += ib_user_cm.h +header-y += rdma_user_ioctl.h header-y += ib_user_mad.h header-y += ib_user_sa.h header-y += ib_user_verbs.h diff --git a/include/uapi/rdma/hfi/hfi1_user.h b/include/uapi/rdma/hfi/hfi1_user.h index 587b7360e820..57d3613f2ff8 100644 --- a/include/uapi/rdma/hfi/hfi1_user.h +++ b/include/uapi/rdma/hfi/hfi1_user.h @@ -57,6 +57,7 @@ #define _LINUX__HFI1_USER_H #include +#include /* * This version number is given to the driver by the user code during @@ -132,7 +133,6 @@ * User IOCTLs can not go above 128 if they do then see common.h and change the * base for the snoop ioctl */ -#define IB_IOCTL_MAGIC 0x1b /* See Documentation/ioctl/ioctl-number.txt */ /* * Make the ioctls occupy the last 0xf0-0xff portion of the IB range diff --git a/include/uapi/rdma/ib_user_mad.h b/include/uapi/rdma/ib_user_mad.h index 09f809f323ea..9de7a2b273d7 100644 --- a/include/uapi/rdma/ib_user_mad.h +++ b/include/uapi/rdma/ib_user_mad.h @@ -35,7 +35,7 @@ #define IB_USER_MAD_H #include -#include +#include /* * Increment this value if any changes that break userspace ABI @@ -230,8 +230,6 @@ struct ib_user_mad_reg_req2 { __u8 reserved[3]; }; -#define IB_IOCTL_MAGIC 0x1b - #define IB_USER_MAD_REGISTER_AGENT _IOWR(IB_IOCTL_MAGIC, 1, \ struct ib_user_mad_reg_req) diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h new file mode 100644 index 000000000000..ba1bcdd7178d --- /dev/null +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -0,0 +1,43 @@ +/* + * Copyright (c) 2016 Mellanox Technologies, LTD. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#ifndef RDMA_USER_IOCTL_H +#define RDMA_USER_IOCTL_H + +#include +#include + +/* Documentation/ioctl/ioctl-number.txt */ +#define RDMA_IOCTL_MAGIC 0x1b +#define IB_IOCTL_MAGIC RDMA_IOCTL_MAGIC + +#endif /* RDMA_USER_IOCTL_H */ -- cgit v1.2.3 From 06393bc39e2fdcd9ca281401f25fa3b834bca799 Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:31 +0300 Subject: RDMA/core: Move legacy MAD IOCTL declarations to common file Move legacy MAD IOCTL declarations to rdma_user_ioctl.h file. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/ib_user_mad.h | 10 ---------- include/uapi/rdma/rdma_user_ioctl.h | 11 +++++++++++ 2 files changed, 11 insertions(+), 10 deletions(-) diff --git a/include/uapi/rdma/ib_user_mad.h b/include/uapi/rdma/ib_user_mad.h index 9de7a2b273d7..5c7abd859e0f 100644 --- a/include/uapi/rdma/ib_user_mad.h +++ b/include/uapi/rdma/ib_user_mad.h @@ -230,14 +230,4 @@ struct ib_user_mad_reg_req2 { __u8 reserved[3]; }; -#define IB_USER_MAD_REGISTER_AGENT _IOWR(IB_IOCTL_MAGIC, 1, \ - struct ib_user_mad_reg_req) - -#define IB_USER_MAD_UNREGISTER_AGENT _IOW(IB_IOCTL_MAGIC, 2, __u32) - -#define IB_USER_MAD_ENABLE_PKEY _IO(IB_IOCTL_MAGIC, 3) - -#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(IB_IOCTL_MAGIC, 4, \ - struct ib_user_mad_reg_req2) - #endif /* IB_USER_MAD_H */ diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h index ba1bcdd7178d..820bf3485608 100644 --- a/include/uapi/rdma/rdma_user_ioctl.h +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -35,9 +35,20 @@ #include #include +#include /* Documentation/ioctl/ioctl-number.txt */ #define RDMA_IOCTL_MAGIC 0x1b #define IB_IOCTL_MAGIC RDMA_IOCTL_MAGIC +#define IB_USER_MAD_REGISTER_AGENT _IOWR(IB_IOCTL_MAGIC, 1, \ + struct ib_user_mad_reg_req) + +#define IB_USER_MAD_UNREGISTER_AGENT _IOW(IB_IOCTL_MAGIC, 2, __u32) + +#define IB_USER_MAD_ENABLE_PKEY _IO(IB_IOCTL_MAGIC, 3) + +#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(IB_IOCTL_MAGIC, 4, \ + struct ib_user_mad_reg_req2) + #endif /* RDMA_USER_IOCTL_H */ -- cgit v1.2.3 From 38e8b671bf04af09fc8de74ea857f920285bd211 Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:32 +0300 Subject: RDMA/hfi1: Avoid redeclaration error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Move hfi1 ioctl definitions to a new header which can be included by both the hfi1 and qib drivers to avoid a duplicate enum definition as shown in this build error for qib: CC [M] drivers/infiniband/hw/qib/qib_sysfs.o In file included from ./include/uapi/rdma/rdma_user_ioctl.h:39:0, from include/uapi/rdma/ib_user_mad.h:38, from include/rdma/ib_mad.h:43, from include/rdma/ib_pma.h:38, from drivers/infiniband/hw/qib/qib_mad.h:37, from drivers/infiniband/hw/qib/qib_init.c:49: ./include/uapi/rdma/hfi/hfi1_user.h:370:2: error: redeclaration of enumerator ‘ur_rcvhdrtail’ ur_rcvhdrtail = 0, Move hfi1 structures to separate file to avoid this failure. The actual move of the ioctl definitions comes in a follow on patch. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/hfi/Kbuild | 1 + include/uapi/rdma/hfi/hfi1_ioctl.h | 173 +++++++++++++++++++++++++++++++++++++ include/uapi/rdma/hfi/hfi1_user.h | 120 +------------------------ 3 files changed, 175 insertions(+), 119 deletions(-) create mode 100644 include/uapi/rdma/hfi/hfi1_ioctl.h diff --git a/include/uapi/rdma/hfi/Kbuild b/include/uapi/rdma/hfi/Kbuild index ef23c294fc71..b65b0b3a5f63 100644 --- a/include/uapi/rdma/hfi/Kbuild +++ b/include/uapi/rdma/hfi/Kbuild @@ -1,2 +1,3 @@ # UAPI Header export list header-y += hfi1_user.h +header-y += hfi1_ioctl.h diff --git a/include/uapi/rdma/hfi/hfi1_ioctl.h b/include/uapi/rdma/hfi/hfi1_ioctl.h new file mode 100644 index 000000000000..4791cc8cb09b --- /dev/null +++ b/include/uapi/rdma/hfi/hfi1_ioctl.h @@ -0,0 +1,173 @@ +/* + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * Copyright(c) 2015 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Copyright(c) 2015 Intel Corporation. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +#ifndef _LINUX__HFI1_IOCTL_H +#define _LINUX__HFI1_IOCTL_H +#include + +/* + * This structure is passed to the driver to tell it where + * user code buffers are, sizes, etc. The offsets and sizes of the + * fields must remain unchanged, for binary compatibility. It can + * be extended, if userversion is changed so user code can tell, if needed + */ +struct hfi1_user_info { + /* + * version of user software, to detect compatibility issues. + * Should be set to HFI1_USER_SWVERSION. + */ + __u32 userversion; + __u32 pad; + /* + * If two or more processes wish to share a context, each process + * must set the subcontext_cnt and subcontext_id to the same + * values. The only restriction on the subcontext_id is that + * it be unique for a given node. + */ + __u16 subctxt_cnt; + __u16 subctxt_id; + /* 128bit UUID passed in by PSM. */ + __u8 uuid[16]; +}; + +struct hfi1_ctxt_info { + __u64 runtime_flags; /* chip/drv runtime flags (HFI1_CAP_*) */ + __u32 rcvegr_size; /* size of each eager buffer */ + __u16 num_active; /* number of active units */ + __u16 unit; /* unit (chip) assigned to caller */ + __u16 ctxt; /* ctxt on unit assigned to caller */ + __u16 subctxt; /* subctxt on unit assigned to caller */ + __u16 rcvtids; /* number of Rcv TIDs for this context */ + __u16 credits; /* number of PIO credits for this context */ + __u16 numa_node; /* NUMA node of the assigned device */ + __u16 rec_cpu; /* cpu # for affinity (0xffff if none) */ + __u16 send_ctxt; /* send context in use by this user context */ + __u16 egrtids; /* number of RcvArray entries for Eager Rcvs */ + __u16 rcvhdrq_cnt; /* number of RcvHdrQ entries */ + __u16 rcvhdrq_entsize; /* size (in bytes) for each RcvHdrQ entry */ + __u16 sdma_ring_size; /* number of entries in SDMA request ring */ +}; + +struct hfi1_tid_info { + /* virtual address of first page in transfer */ + __u64 vaddr; + /* pointer to tid array. this array is big enough */ + __u64 tidlist; + /* number of tids programmed by this request */ + __u32 tidcnt; + /* length of transfer buffer programmed by this request */ + __u32 length; +}; + +/* + * This structure is returned by the driver immediately after + * open to get implementation-specific info, and info specific to this + * instance. + * + * This struct must have explicit pad fields where type sizes + * may result in different alignments between 32 and 64 bit + * programs, since the 64 bit * bit kernel requires the user code + * to have matching offsets + */ +struct hfi1_base_info { + /* version of hardware, for feature checking. */ + __u32 hw_version; + /* version of software, for feature checking. */ + __u32 sw_version; + /* Job key */ + __u16 jkey; + __u16 padding1; + /* + * The special QP (queue pair) value that identifies PSM + * protocol packet from standard IB packets. + */ + __u32 bthqp; + /* PIO credit return address, */ + __u64 sc_credits_addr; + /* + * Base address of write-only pio buffers for this process. + * Each buffer has sendpio_credits*64 bytes. + */ + __u64 pio_bufbase_sop; + /* + * Base address of write-only pio buffers for this process. + * Each buffer has sendpio_credits*64 bytes. + */ + __u64 pio_bufbase; + /* address where receive buffer queue is mapped into */ + __u64 rcvhdr_bufbase; + /* base address of Eager receive buffers. */ + __u64 rcvegr_bufbase; + /* base address of SDMA completion ring */ + __u64 sdma_comp_bufbase; + /* + * User register base for init code, not to be used directly by + * protocol or applications. Always maps real chip register space. + * the register addresses are: + * ur_rcvhdrhead, ur_rcvhdrtail, ur_rcvegrhead, ur_rcvegrtail, + * ur_rcvtidflow + */ + __u64 user_regbase; + /* notification events */ + __u64 events_bufbase; + /* status page */ + __u64 status_bufbase; + /* rcvhdrtail update */ + __u64 rcvhdrtail_base; + /* + * shared memory pages for subctxts if ctxt is shared; these cover + * all the processes in the group sharing a single context. + * all have enough space for the num_subcontexts value on this job. + */ + __u64 subctxt_uregbase; + __u64 subctxt_rcvegrbuf; + __u64 subctxt_rcvhdrbuf; +}; +#endif /* _LINIUX__HFI1_IOCTL_H */ diff --git a/include/uapi/rdma/hfi/hfi1_user.h b/include/uapi/rdma/hfi/hfi1_user.h index 57d3613f2ff8..220a020d606d 100644 --- a/include/uapi/rdma/hfi/hfi1_user.h +++ b/include/uapi/rdma/hfi/hfi1_user.h @@ -58,6 +58,7 @@ #include #include +#include /* * This version number is given to the driver by the user code during @@ -211,60 +212,6 @@ struct hfi1_cmd; #define HFI1_POLL_TYPE_ANYRCV 0x0 #define HFI1_POLL_TYPE_URGENT 0x1 -/* - * This structure is passed to the driver to tell it where - * user code buffers are, sizes, etc. The offsets and sizes of the - * fields must remain unchanged, for binary compatibility. It can - * be extended, if userversion is changed so user code can tell, if needed - */ -struct hfi1_user_info { - /* - * version of user software, to detect compatibility issues. - * Should be set to HFI1_USER_SWVERSION. - */ - __u32 userversion; - __u32 pad; - /* - * If two or more processes wish to share a context, each process - * must set the subcontext_cnt and subcontext_id to the same - * values. The only restriction on the subcontext_id is that - * it be unique for a given node. - */ - __u16 subctxt_cnt; - __u16 subctxt_id; - /* 128bit UUID passed in by PSM. */ - __u8 uuid[16]; -}; - -struct hfi1_ctxt_info { - __u64 runtime_flags; /* chip/drv runtime flags (HFI1_CAP_*) */ - __u32 rcvegr_size; /* size of each eager buffer */ - __u16 num_active; /* number of active units */ - __u16 unit; /* unit (chip) assigned to caller */ - __u16 ctxt; /* ctxt on unit assigned to caller */ - __u16 subctxt; /* subctxt on unit assigned to caller */ - __u16 rcvtids; /* number of Rcv TIDs for this context */ - __u16 credits; /* number of PIO credits for this context */ - __u16 numa_node; /* NUMA node of the assigned device */ - __u16 rec_cpu; /* cpu # for affinity (0xffff if none) */ - __u16 send_ctxt; /* send context in use by this user context */ - __u16 egrtids; /* number of RcvArray entries for Eager Rcvs */ - __u16 rcvhdrq_cnt; /* number of RcvHdrQ entries */ - __u16 rcvhdrq_entsize; /* size (in bytes) for each RcvHdrQ entry */ - __u16 sdma_ring_size; /* number of entries in SDMA request ring */ -}; - -struct hfi1_tid_info { - /* virtual address of first page in transfer */ - __u64 vaddr; - /* pointer to tid array. this array is big enough */ - __u64 tidlist; - /* number of tids programmed by this request */ - __u32 tidcnt; - /* length of transfer buffer programmed by this request */ - __u32 length; -}; - enum hfi1_sdma_comp_state { FREE = 0, QUEUED, @@ -289,71 +236,6 @@ struct hfi1_status { char freezemsg[0]; }; -/* - * This structure is returned by the driver immediately after - * open to get implementation-specific info, and info specific to this - * instance. - * - * This struct must have explicit pad fields where type sizes - * may result in different alignments between 32 and 64 bit - * programs, since the 64 bit * bit kernel requires the user code - * to have matching offsets - */ -struct hfi1_base_info { - /* version of hardware, for feature checking. */ - __u32 hw_version; - /* version of software, for feature checking. */ - __u32 sw_version; - /* Job key */ - __u16 jkey; - __u16 padding1; - /* - * The special QP (queue pair) value that identifies PSM - * protocol packet from standard IB packets. - */ - __u32 bthqp; - /* PIO credit return address, */ - __u64 sc_credits_addr; - /* - * Base address of write-only pio buffers for this process. - * Each buffer has sendpio_credits*64 bytes. - */ - __u64 pio_bufbase_sop; - /* - * Base address of write-only pio buffers for this process. - * Each buffer has sendpio_credits*64 bytes. - */ - __u64 pio_bufbase; - /* address where receive buffer queue is mapped into */ - __u64 rcvhdr_bufbase; - /* base address of Eager receive buffers. */ - __u64 rcvegr_bufbase; - /* base address of SDMA completion ring */ - __u64 sdma_comp_bufbase; - /* - * User register base for init code, not to be used directly by - * protocol or applications. Always maps real chip register space. - * the register addresses are: - * ur_rcvhdrhead, ur_rcvhdrtail, ur_rcvegrhead, ur_rcvegrtail, - * ur_rcvtidflow - */ - __u64 user_regbase; - /* notification events */ - __u64 events_bufbase; - /* status page */ - __u64 status_bufbase; - /* rcvhdrtail update */ - __u64 rcvhdrtail_base; - /* - * shared memory pages for subctxts if ctxt is shared; these cover - * all the processes in the group sharing a single context. - * all have enough space for the num_subcontexts value on this job. - */ - __u64 subctxt_uregbase; - __u64 subctxt_rcvegrbuf; - __u64 subctxt_rcvhdrbuf; -}; - enum sdma_req_opcode { EXPECTED = 0, EAGER -- cgit v1.2.3 From 8edec0b55a47d667031d078f29f7a23f0a11122f Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:33 +0300 Subject: RDMA/core: Move HFI1 IOCTL declarations to common file Move HFI1 IOCTL declarations to rdma_user_ioctl.h file. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/hfi/hfi1_user.h | 55 ------------------------------------- include/uapi/rdma/rdma_user_ioctl.h | 54 ++++++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+), 55 deletions(-) diff --git a/include/uapi/rdma/hfi/hfi1_user.h b/include/uapi/rdma/hfi/hfi1_user.h index 220a020d606d..3f4ee93ae5eb 100644 --- a/include/uapi/rdma/hfi/hfi1_user.h +++ b/include/uapi/rdma/hfi/hfi1_user.h @@ -58,7 +58,6 @@ #include #include -#include /* * This version number is given to the driver by the user code during @@ -114,60 +113,6 @@ #define HFI1_RCVHDR_ENTSIZE_16 (1UL << 1) #define HFI1_RCVDHR_ENTSIZE_32 (1UL << 2) -/* User commands. */ -#define HFI1_CMD_ASSIGN_CTXT 1 /* allocate HFI and context */ -#define HFI1_CMD_CTXT_INFO 2 /* find out what resources we got */ -#define HFI1_CMD_USER_INFO 3 /* set up userspace */ -#define HFI1_CMD_TID_UPDATE 4 /* update expected TID entries */ -#define HFI1_CMD_TID_FREE 5 /* free expected TID entries */ -#define HFI1_CMD_CREDIT_UPD 6 /* force an update of PIO credit */ - -#define HFI1_CMD_RECV_CTRL 8 /* control receipt of packets */ -#define HFI1_CMD_POLL_TYPE 9 /* set the kind of polling we want */ -#define HFI1_CMD_ACK_EVENT 10 /* ack & clear user status bits */ -#define HFI1_CMD_SET_PKEY 11 /* set context's pkey */ -#define HFI1_CMD_CTXT_RESET 12 /* reset context's HW send context */ -#define HFI1_CMD_TID_INVAL_READ 13 /* read TID cache invalidations */ -#define HFI1_CMD_GET_VERS 14 /* get the version of the user cdev */ - -/* - * User IOCTLs can not go above 128 if they do then see common.h and change the - * base for the snoop ioctl - */ - -/* - * Make the ioctls occupy the last 0xf0-0xff portion of the IB range - */ -#define __NUM(cmd) (HFI1_CMD_##cmd + 0xe0) - -struct hfi1_cmd; -#define HFI1_IOCTL_ASSIGN_CTXT \ - _IOWR(IB_IOCTL_MAGIC, __NUM(ASSIGN_CTXT), struct hfi1_user_info) -#define HFI1_IOCTL_CTXT_INFO \ - _IOW(IB_IOCTL_MAGIC, __NUM(CTXT_INFO), struct hfi1_ctxt_info) -#define HFI1_IOCTL_USER_INFO \ - _IOW(IB_IOCTL_MAGIC, __NUM(USER_INFO), struct hfi1_base_info) -#define HFI1_IOCTL_TID_UPDATE \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_UPDATE), struct hfi1_tid_info) -#define HFI1_IOCTL_TID_FREE \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_FREE), struct hfi1_tid_info) -#define HFI1_IOCTL_CREDIT_UPD \ - _IO(IB_IOCTL_MAGIC, __NUM(CREDIT_UPD)) -#define HFI1_IOCTL_RECV_CTRL \ - _IOW(IB_IOCTL_MAGIC, __NUM(RECV_CTRL), int) -#define HFI1_IOCTL_POLL_TYPE \ - _IOW(IB_IOCTL_MAGIC, __NUM(POLL_TYPE), int) -#define HFI1_IOCTL_ACK_EVENT \ - _IOW(IB_IOCTL_MAGIC, __NUM(ACK_EVENT), unsigned long) -#define HFI1_IOCTL_SET_PKEY \ - _IOW(IB_IOCTL_MAGIC, __NUM(SET_PKEY), __u16) -#define HFI1_IOCTL_CTXT_RESET \ - _IO(IB_IOCTL_MAGIC, __NUM(CTXT_RESET)) -#define HFI1_IOCTL_TID_INVAL_READ \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_INVAL_READ), struct hfi1_tid_info) -#define HFI1_IOCTL_GET_VERS \ - _IOR(IB_IOCTL_MAGIC, __NUM(GET_VERS), int) - #define _HFI1_EVENT_FROZEN_BIT 0 #define _HFI1_EVENT_LINKDOWN_BIT 1 #define _HFI1_EVENT_LID_CHANGE_BIT 2 diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h index 820bf3485608..e9a69f0b18a2 100644 --- a/include/uapi/rdma/rdma_user_ioctl.h +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -36,6 +36,7 @@ #include #include #include +#include /* Documentation/ioctl/ioctl-number.txt */ #define RDMA_IOCTL_MAGIC 0x1b @@ -51,4 +52,57 @@ #define IB_USER_MAD_REGISTER_AGENT2 _IOWR(IB_IOCTL_MAGIC, 4, \ struct ib_user_mad_reg_req2) +/* User commands. */ +#define HFI1_CMD_ASSIGN_CTXT 1 /* allocate HFI and context */ +#define HFI1_CMD_CTXT_INFO 2 /* find out what resources we got */ +#define HFI1_CMD_USER_INFO 3 /* set up userspace */ +#define HFI1_CMD_TID_UPDATE 4 /* update expected TID entries */ +#define HFI1_CMD_TID_FREE 5 /* free expected TID entries */ +#define HFI1_CMD_CREDIT_UPD 6 /* force an update of PIO credit */ + +#define HFI1_CMD_RECV_CTRL 8 /* control receipt of packets */ +#define HFI1_CMD_POLL_TYPE 9 /* set the kind of polling we want */ +#define HFI1_CMD_ACK_EVENT 10 /* ack & clear user status bits */ +#define HFI1_CMD_SET_PKEY 11 /* set context's pkey */ +#define HFI1_CMD_CTXT_RESET 12 /* reset context's HW send context */ +#define HFI1_CMD_TID_INVAL_READ 13 /* read TID cache invalidations */ +#define HFI1_CMD_GET_VERS 14 /* get the version of the user cdev */ + +/* + * User IOCTLs can not go above 128 if they do then see common.h and change the + * base for the snoop ioctl + */ + +/* + * Make the ioctls occupy the last 0xf0-0xff portion of the IB range + */ +#define __NUM(cmd) (HFI1_CMD_##cmd + 0xe0) + +#define HFI1_IOCTL_ASSIGN_CTXT \ + _IOWR(IB_IOCTL_MAGIC, __NUM(ASSIGN_CTXT), struct hfi1_user_info) +#define HFI1_IOCTL_CTXT_INFO \ + _IOW(IB_IOCTL_MAGIC, __NUM(CTXT_INFO), struct hfi1_ctxt_info) +#define HFI1_IOCTL_USER_INFO \ + _IOW(IB_IOCTL_MAGIC, __NUM(USER_INFO), struct hfi1_base_info) +#define HFI1_IOCTL_TID_UPDATE \ + _IOWR(IB_IOCTL_MAGIC, __NUM(TID_UPDATE), struct hfi1_tid_info) +#define HFI1_IOCTL_TID_FREE \ + _IOWR(IB_IOCTL_MAGIC, __NUM(TID_FREE), struct hfi1_tid_info) +#define HFI1_IOCTL_CREDIT_UPD \ + _IO(IB_IOCTL_MAGIC, __NUM(CREDIT_UPD)) +#define HFI1_IOCTL_RECV_CTRL \ + _IOW(IB_IOCTL_MAGIC, __NUM(RECV_CTRL), int) +#define HFI1_IOCTL_POLL_TYPE \ + _IOW(IB_IOCTL_MAGIC, __NUM(POLL_TYPE), int) +#define HFI1_IOCTL_ACK_EVENT \ + _IOW(IB_IOCTL_MAGIC, __NUM(ACK_EVENT), unsigned long) +#define HFI1_IOCTL_SET_PKEY \ + _IOW(IB_IOCTL_MAGIC, __NUM(SET_PKEY), __u16) +#define HFI1_IOCTL_CTXT_RESET \ + _IO(IB_IOCTL_MAGIC, __NUM(CTXT_RESET)) +#define HFI1_IOCTL_TID_INVAL_READ \ + _IOWR(IB_IOCTL_MAGIC, __NUM(TID_INVAL_READ), struct hfi1_tid_info) +#define HFI1_IOCTL_GET_VERS \ + _IOR(IB_IOCTL_MAGIC, __NUM(GET_VERS), int) + #endif /* RDMA_USER_IOCTL_H */ -- cgit v1.2.3 From 10b31e792e5862ad9e232d7aa2c65e7b7e46bc39 Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:34 +0300 Subject: RDMA/core: Rename RDMA magic number Rename RDMA magic number to better describe IOCTLs. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/rdma_user_ioctl.h | 35 ++++++++++++++++++----------------- 1 file changed, 18 insertions(+), 17 deletions(-) diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h index e9a69f0b18a2..7ecf8cd17f58 100644 --- a/include/uapi/rdma/rdma_user_ioctl.h +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -40,16 +40,17 @@ /* Documentation/ioctl/ioctl-number.txt */ #define RDMA_IOCTL_MAGIC 0x1b +/* Legacy name, for user space application which already use it */ #define IB_IOCTL_MAGIC RDMA_IOCTL_MAGIC -#define IB_USER_MAD_REGISTER_AGENT _IOWR(IB_IOCTL_MAGIC, 1, \ +#define IB_USER_MAD_REGISTER_AGENT _IOWR(RDMA_IOCTL_MAGIC, 1, \ struct ib_user_mad_reg_req) -#define IB_USER_MAD_UNREGISTER_AGENT _IOW(IB_IOCTL_MAGIC, 2, __u32) +#define IB_USER_MAD_UNREGISTER_AGENT _IOW(RDMA_IOCTL_MAGIC, 2, __u32) -#define IB_USER_MAD_ENABLE_PKEY _IO(IB_IOCTL_MAGIC, 3) +#define IB_USER_MAD_ENABLE_PKEY _IO(RDMA_IOCTL_MAGIC, 3) -#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(IB_IOCTL_MAGIC, 4, \ +#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(RDMA_IOCTL_MAGIC, 4, \ struct ib_user_mad_reg_req2) /* User commands. */ @@ -79,30 +80,30 @@ #define __NUM(cmd) (HFI1_CMD_##cmd + 0xe0) #define HFI1_IOCTL_ASSIGN_CTXT \ - _IOWR(IB_IOCTL_MAGIC, __NUM(ASSIGN_CTXT), struct hfi1_user_info) + _IOWR(RDMA_IOCTL_MAGIC, __NUM(ASSIGN_CTXT), struct hfi1_user_info) #define HFI1_IOCTL_CTXT_INFO \ - _IOW(IB_IOCTL_MAGIC, __NUM(CTXT_INFO), struct hfi1_ctxt_info) + _IOW(RDMA_IOCTL_MAGIC, __NUM(CTXT_INFO), struct hfi1_ctxt_info) #define HFI1_IOCTL_USER_INFO \ - _IOW(IB_IOCTL_MAGIC, __NUM(USER_INFO), struct hfi1_base_info) + _IOW(RDMA_IOCTL_MAGIC, __NUM(USER_INFO), struct hfi1_base_info) #define HFI1_IOCTL_TID_UPDATE \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_UPDATE), struct hfi1_tid_info) + _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_UPDATE), struct hfi1_tid_info) #define HFI1_IOCTL_TID_FREE \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_FREE), struct hfi1_tid_info) + _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_FREE), struct hfi1_tid_info) #define HFI1_IOCTL_CREDIT_UPD \ - _IO(IB_IOCTL_MAGIC, __NUM(CREDIT_UPD)) + _IO(RDMA_IOCTL_MAGIC, __NUM(CREDIT_UPD)) #define HFI1_IOCTL_RECV_CTRL \ - _IOW(IB_IOCTL_MAGIC, __NUM(RECV_CTRL), int) + _IOW(RDMA_IOCTL_MAGIC, __NUM(RECV_CTRL), int) #define HFI1_IOCTL_POLL_TYPE \ - _IOW(IB_IOCTL_MAGIC, __NUM(POLL_TYPE), int) + _IOW(RDMA_IOCTL_MAGIC, __NUM(POLL_TYPE), int) #define HFI1_IOCTL_ACK_EVENT \ - _IOW(IB_IOCTL_MAGIC, __NUM(ACK_EVENT), unsigned long) + _IOW(RDMA_IOCTL_MAGIC, __NUM(ACK_EVENT), unsigned long) #define HFI1_IOCTL_SET_PKEY \ - _IOW(IB_IOCTL_MAGIC, __NUM(SET_PKEY), __u16) + _IOW(RDMA_IOCTL_MAGIC, __NUM(SET_PKEY), __u16) #define HFI1_IOCTL_CTXT_RESET \ - _IO(IB_IOCTL_MAGIC, __NUM(CTXT_RESET)) + _IO(RDMA_IOCTL_MAGIC, __NUM(CTXT_RESET)) #define HFI1_IOCTL_TID_INVAL_READ \ - _IOWR(IB_IOCTL_MAGIC, __NUM(TID_INVAL_READ), struct hfi1_tid_info) + _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_INVAL_READ), struct hfi1_tid_info) #define HFI1_IOCTL_GET_VERS \ - _IOR(IB_IOCTL_MAGIC, __NUM(GET_VERS), int) + _IOR(RDMA_IOCTL_MAGIC, __NUM(GET_VERS), int) #endif /* RDMA_USER_IOCTL_H */ -- cgit v1.2.3 From fa83b7936e9799d9cd96a26fbcba22d2baa7ae11 Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Sun, 4 Sep 2016 14:30:35 +0300 Subject: RDMA/core: Unify style of IOCTL commands MAD and HFI1 have different naming convention, this patch simplifies and unifies their defines and names. As part of cleanup, the HFI1 _NUM() macro and command indexes were removed (controversial). This will cause intentional (and arguably unnecessary) breakage to the PSM user space library. Signed-off-by: Matan Barak Signed-off-by: Haggai Eran Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/uapi/rdma/rdma_user_ioctl.h | 98 ++++++++++++++----------------------- 1 file changed, 38 insertions(+), 60 deletions(-) diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h index 7ecf8cd17f58..9388125ad51b 100644 --- a/include/uapi/rdma/rdma_user_ioctl.h +++ b/include/uapi/rdma/rdma_user_ioctl.h @@ -39,71 +39,49 @@ #include /* Documentation/ioctl/ioctl-number.txt */ -#define RDMA_IOCTL_MAGIC 0x1b +#define RDMA_IOCTL_MAGIC 0x1b /* Legacy name, for user space application which already use it */ -#define IB_IOCTL_MAGIC RDMA_IOCTL_MAGIC - -#define IB_USER_MAD_REGISTER_AGENT _IOWR(RDMA_IOCTL_MAGIC, 1, \ - struct ib_user_mad_reg_req) - -#define IB_USER_MAD_UNREGISTER_AGENT _IOW(RDMA_IOCTL_MAGIC, 2, __u32) - -#define IB_USER_MAD_ENABLE_PKEY _IO(RDMA_IOCTL_MAGIC, 3) - -#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(RDMA_IOCTL_MAGIC, 4, \ - struct ib_user_mad_reg_req2) - -/* User commands. */ -#define HFI1_CMD_ASSIGN_CTXT 1 /* allocate HFI and context */ -#define HFI1_CMD_CTXT_INFO 2 /* find out what resources we got */ -#define HFI1_CMD_USER_INFO 3 /* set up userspace */ -#define HFI1_CMD_TID_UPDATE 4 /* update expected TID entries */ -#define HFI1_CMD_TID_FREE 5 /* free expected TID entries */ -#define HFI1_CMD_CREDIT_UPD 6 /* force an update of PIO credit */ - -#define HFI1_CMD_RECV_CTRL 8 /* control receipt of packets */ -#define HFI1_CMD_POLL_TYPE 9 /* set the kind of polling we want */ -#define HFI1_CMD_ACK_EVENT 10 /* ack & clear user status bits */ -#define HFI1_CMD_SET_PKEY 11 /* set context's pkey */ -#define HFI1_CMD_CTXT_RESET 12 /* reset context's HW send context */ -#define HFI1_CMD_TID_INVAL_READ 13 /* read TID cache invalidations */ -#define HFI1_CMD_GET_VERS 14 /* get the version of the user cdev */ +#define IB_IOCTL_MAGIC RDMA_IOCTL_MAGIC /* - * User IOCTLs can not go above 128 if they do then see common.h and change the - * base for the snoop ioctl + * General blocks assignments + * It is closed on purpose do not expose it it user space + * #define MAD_CMD_BASE 0x00 + * #define HFI1_CMD_BAS 0xE0 */ -/* - * Make the ioctls occupy the last 0xf0-0xff portion of the IB range - */ -#define __NUM(cmd) (HFI1_CMD_##cmd + 0xe0) +/* MAD specific section */ +#define IB_USER_MAD_REGISTER_AGENT _IOWR(RDMA_IOCTL_MAGIC, 0x01, struct ib_user_mad_reg_req) +#define IB_USER_MAD_UNREGISTER_AGENT _IOW(RDMA_IOCTL_MAGIC, 0x02, __u32) +#define IB_USER_MAD_ENABLE_PKEY _IO(RDMA_IOCTL_MAGIC, 0x03) +#define IB_USER_MAD_REGISTER_AGENT2 _IOWR(RDMA_IOCTL_MAGIC, 0x04, struct ib_user_mad_reg_req2) -#define HFI1_IOCTL_ASSIGN_CTXT \ - _IOWR(RDMA_IOCTL_MAGIC, __NUM(ASSIGN_CTXT), struct hfi1_user_info) -#define HFI1_IOCTL_CTXT_INFO \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(CTXT_INFO), struct hfi1_ctxt_info) -#define HFI1_IOCTL_USER_INFO \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(USER_INFO), struct hfi1_base_info) -#define HFI1_IOCTL_TID_UPDATE \ - _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_UPDATE), struct hfi1_tid_info) -#define HFI1_IOCTL_TID_FREE \ - _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_FREE), struct hfi1_tid_info) -#define HFI1_IOCTL_CREDIT_UPD \ - _IO(RDMA_IOCTL_MAGIC, __NUM(CREDIT_UPD)) -#define HFI1_IOCTL_RECV_CTRL \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(RECV_CTRL), int) -#define HFI1_IOCTL_POLL_TYPE \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(POLL_TYPE), int) -#define HFI1_IOCTL_ACK_EVENT \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(ACK_EVENT), unsigned long) -#define HFI1_IOCTL_SET_PKEY \ - _IOW(RDMA_IOCTL_MAGIC, __NUM(SET_PKEY), __u16) -#define HFI1_IOCTL_CTXT_RESET \ - _IO(RDMA_IOCTL_MAGIC, __NUM(CTXT_RESET)) -#define HFI1_IOCTL_TID_INVAL_READ \ - _IOWR(RDMA_IOCTL_MAGIC, __NUM(TID_INVAL_READ), struct hfi1_tid_info) -#define HFI1_IOCTL_GET_VERS \ - _IOR(RDMA_IOCTL_MAGIC, __NUM(GET_VERS), int) +/* HFI specific section */ +/* allocate HFI and context */ +#define HFI1_IOCTL_ASSIGN_CTXT _IOWR(RDMA_IOCTL_MAGIC, 0xE1, struct hfi1_user_info) +/* find out what resources we got */ +#define HFI1_IOCTL_CTXT_INFO _IOW(RDMA_IOCTL_MAGIC, 0xE2, struct hfi1_ctxt_info) +/* set up userspace */ +#define HFI1_IOCTL_USER_INFO _IOW(RDMA_IOCTL_MAGIC, 0xE3, struct hfi1_base_info) +/* update expected TID entries */ +#define HFI1_IOCTL_TID_UPDATE _IOWR(RDMA_IOCTL_MAGIC, 0xE4, struct hfi1_tid_info) +/* free expected TID entries */ +#define HFI1_IOCTL_TID_FREE _IOWR(RDMA_IOCTL_MAGIC, 0xE5, struct hfi1_tid_info) +/* force an update of PIO credit */ +#define HFI1_IOCTL_CREDIT_UPD _IO(RDMA_IOCTL_MAGIC, 0xE6) +/* control receipt of packets */ +#define HFI1_IOCTL_RECV_CTRL _IOW(RDMA_IOCTL_MAGIC, 0xE8, int) +/* set the kind of polling we want */ +#define HFI1_IOCTL_POLL_TYPE _IOW(RDMA_IOCTL_MAGIC, 0xE9, int) +/* ack & clear user status bits */ +#define HFI1_IOCTL_ACK_EVENT _IOW(RDMA_IOCTL_MAGIC, 0xEA, unsigned long) +/* set context's pkey */ +#define HFI1_IOCTL_SET_PKEY _IOW(RDMA_IOCTL_MAGIC, 0xEB, __u16) +/* reset context's HW send context */ +#define HFI1_IOCTL_CTXT_RESET _IO(RDMA_IOCTL_MAGIC, 0xEC) +/* read TID cache invalidations */ +#define HFI1_IOCTL_TID_INVAL_READ _IOWR(RDMA_IOCTL_MAGIC, 0xED, struct hfi1_tid_info) +/* get the version of the user cdev */ +#define HFI1_IOCTL_GET_VERS _IOR(RDMA_IOCTL_MAGIC, 0xEE, int) #endif /* RDMA_USER_IOCTL_H */ -- cgit v1.2.3 From 69ae543969abeba48e04dd93277684c8c0895f3b Mon Sep 17 00:00:00 2001 From: Selvin Xavier Date: Mon, 19 Dec 2016 11:28:46 -0800 Subject: RDMA: Adding ethertype ETH_P_IBOE Update the if_ether.h with the ethertype for Infiniband over Ethernet packets. Also, removing the occurances of 0x8915 from infiniband vendor drivers. Signed-off-by: Selvin Xavier Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx4/qp.c | 6 +----- drivers/infiniband/hw/ocrdma/ocrdma_ah.c | 4 ++-- drivers/infiniband/hw/ocrdma/ocrdma_hw.c | 3 ++- drivers/infiniband/hw/ocrdma/ocrdma_sli.h | 5 ----- drivers/infiniband/hw/qedr/qedr_cm.c | 2 +- drivers/infiniband/hw/qedr/qedr_cm.h | 1 - drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h | 1 - drivers/infiniband/hw/usnic/usnic_fwd.h | 3 ++- include/uapi/linux/if_ether.h | 1 + 9 files changed, 9 insertions(+), 17 deletions(-) diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index c068add8838b..7d76f769233c 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -76,10 +76,6 @@ enum { MLX4_IB_LSO_HEADER_SPARE = 128, }; -enum { - MLX4_IB_IBOE_ETHERTYPE = 0x8915 -}; - struct mlx4_ib_sqp { struct mlx4_ib_qp qp; int pkey_index; @@ -2588,7 +2584,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, u16 ether_type; u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13; - ether_type = (!is_udp) ? MLX4_IB_IBOE_ETHERTYPE : + ether_type = (!is_udp) ? ETH_P_IBOE: (ip_version == 4 ? ETH_P_IP : ETH_P_IPV6); mlx->sched_prio = cpu_to_be16(pcp); diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c index 14d33b0f3950..cd66e1e45dd7 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c @@ -59,7 +59,7 @@ static u16 ocrdma_hdr_type_to_proto_num(int devid, u8 hdr_type) { switch (hdr_type) { case OCRDMA_L3_TYPE_IB_GRH: - return (u16)0x8915; + return (u16)ETH_P_IBOE; case OCRDMA_L3_TYPE_IPV4: return (u16)0x0800; case OCRDMA_L3_TYPE_IPV6: @@ -94,7 +94,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, proto_num = ocrdma_hdr_type_to_proto_num(dev->id, ah->hdr_type); if (!proto_num) return -EINVAL; - nxthdr = (proto_num == 0x8915) ? 0x1b : 0x11; + nxthdr = (proto_num == ETH_P_IBOE) ? 0x1b : 0x11; /* VLAN */ if (!vlan_tag || (vlan_tag > 0xFFF)) vlan_tag = dev->pvid; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c index 9a305201545e..aa6967197620 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #include @@ -2984,7 +2985,7 @@ static int ocrdma_parse_dcbxcfg_rsp(struct ocrdma_dev *dev, int ptype, OCRDMA_APP_PARAM_APP_PROTO_MASK; if ( - valid && proto == OCRDMA_APP_PROTO_ROCE && + valid && proto == ETH_P_IBOE && proto_sel == OCRDMA_PROTO_SELECT_L2) { for (slindx = 0; slindx < OCRDMA_MAX_SERVICE_LEVEL_INDEX; slindx++) { diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h index 37df4481bb8f..6ef89c226ad8 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h +++ b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h @@ -1901,7 +1901,6 @@ struct ocrdma_eth_vlan { u8 smac[6]; __be16 eth_type; __be16 vlan_tag; -#define OCRDMA_ROCE_ETH_TYPE 0x8915 __be16 roce_eth_type; } __packed; @@ -2179,10 +2178,6 @@ enum OCRDMA_DCBX_PARAM_TYPE { OCRDMA_PARAMETER_TYPE_PEER = 0x02 }; -enum OCRDMA_DCBX_APP_PROTO { - OCRDMA_APP_PROTO_ROCE = 0x8915 -}; - enum OCRDMA_DCBX_PROTO { OCRDMA_PROTO_SELECT_L2 = 0x00, OCRDMA_PROTO_SELECT_L4 = 0x01 diff --git a/drivers/infiniband/hw/qedr/qedr_cm.c b/drivers/infiniband/hw/qedr/qedr_cm.c index 63890ebb72bd..c6aee0333f30 100644 --- a/drivers/infiniband/hw/qedr/qedr_cm.c +++ b/drivers/infiniband/hw/qedr/qedr_cm.c @@ -293,7 +293,7 @@ static inline int qedr_gsi_build_header(struct qedr_dev *dev, has_udp = (sgid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP); if (!has_udp) { /* RoCE v1 */ - ether_type = ETH_P_ROCE; + ether_type = ETH_P_IBOE; *roce_mode = ROCE_V1; } else if (ipv6_addr_v4mapped((struct in6_addr *)&sgid)) { /* RoCE v2 IPv4 */ diff --git a/drivers/infiniband/hw/qedr/qedr_cm.h b/drivers/infiniband/hw/qedr/qedr_cm.h index 9ba6e15cd93f..78efb1b056d1 100644 --- a/drivers/infiniband/hw/qedr/qedr_cm.h +++ b/drivers/infiniband/hw/qedr/qedr_cm.h @@ -37,7 +37,6 @@ #define QEDR_GSI_MAX_RECV_SGE (1) /* LL2 FW limitation */ -#define ETH_P_ROCE (0x8915) #define QEDR_ROCE_V2_UDP_SPORT (0000) static inline u32 qedr_get_ipv4_from_gid(u8 *gid) diff --git a/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h b/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h index 596e0ed49a8e..bf7d197a9f42 100644 --- a/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h +++ b/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h @@ -34,7 +34,6 @@ #ifndef USNIC_CMN_PKT_HDR_H #define USNIC_CMN_PKT_HDR_H -#define USNIC_ROCE_ETHERTYPE (0x8915) #define USNIC_ROCE_GRH_VER (8) #define USNIC_PROTO_VER (1) #define USNIC_ROCE_GRH_VER_SHIFT (4) diff --git a/drivers/infiniband/hw/usnic/usnic_fwd.h b/drivers/infiniband/hw/usnic/usnic_fwd.h index 3a8add9ddf46..b2ac22be0731 100644 --- a/drivers/infiniband/hw/usnic/usnic_fwd.h +++ b/drivers/infiniband/hw/usnic/usnic_fwd.h @@ -36,6 +36,7 @@ #include #include +#include #include #include @@ -97,7 +98,7 @@ static inline void usnic_fwd_init_usnic_filter(struct filter *filter, uint32_t usnic_id) { filter->type = FILTER_USNIC_ID; - filter->u.usnic.ethtype = USNIC_ROCE_ETHERTYPE; + filter->u.usnic.ethtype = ETH_P_IBOE; filter->u.usnic.flags = FILTER_FIELD_USNIC_ETHTYPE | FILTER_FIELD_USNIC_ID | FILTER_FIELD_USNIC_PROTO; diff --git a/include/uapi/linux/if_ether.h b/include/uapi/linux/if_ether.h index 3e5185e9ef03..5bc9bfd816b7 100644 --- a/include/uapi/linux/if_ether.h +++ b/include/uapi/linux/if_ether.h @@ -93,6 +93,7 @@ #define ETH_P_NCSI 0x88F8 /* NCSI protocol */ #define ETH_P_PRP 0x88FB /* IEC 62439-3 PRP/HSRv0 */ #define ETH_P_FCOE 0x8906 /* Fibre Channel over Ethernet */ +#define ETH_P_IBOE 0x8915 /* Infiniband over Ethernet */ #define ETH_P_TDLS 0x890D /* TDLS */ #define ETH_P_FIP 0x8914 /* FCoE Initialization Protocol */ #define ETH_P_80221 0x8917 /* IEEE 802.21 Media Independent Handover Protocol */ -- cgit v1.2.3 From 8d8f08372091ae3ea92930ebd0d2c33e8bc199c5 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:40 -0800 Subject: IB/rxe: Suppress sparse warnings Avoid that sparse complains about using 0 as a pointer, about missing function declarations and also avoid that sparse complains about endianness. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_net.c | 8 ++++---- drivers/infiniband/sw/rxe/rxe_recv.c | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index 342e78163613..151f639abebf 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -243,8 +243,8 @@ static struct socket *rxe_setup_udp_tunnel(struct net *net, __be16 port, { int err; struct socket *sock; - struct udp_port_cfg udp_cfg = {0}; - struct udp_tunnel_sock_cfg tnl_cfg = {0}; + struct udp_port_cfg udp_cfg = { }; + struct udp_tunnel_sock_cfg tnl_cfg = { }; if (ipv6) { udp_cfg.family = AF_INET6; @@ -658,7 +658,7 @@ struct notifier_block rxe_net_notifier = { .notifier_call = rxe_notify, }; -int rxe_net_ipv4_init(void) +static int rxe_net_ipv4_init(void) { recv_sockets.sk4 = rxe_setup_udp_tunnel(&init_net, htons(ROCE_V2_UDP_DPORT), false); @@ -671,7 +671,7 @@ int rxe_net_ipv4_init(void) return 0; } -int rxe_net_ipv6_init(void) +static int rxe_net_ipv6_init(void) { #if IS_ENABLED(CONFIG_IPV6) diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c index 252b4d637d45..50886031096f 100644 --- a/drivers/infiniband/sw/rxe/rxe_recv.c +++ b/drivers/infiniband/sw/rxe/rxe_recv.c @@ -389,7 +389,7 @@ int rxe_rcv(struct sk_buff *skb) calc_icrc = rxe_icrc_hdr(pkt, skb); calc_icrc = crc32_le(calc_icrc, (u8 *)payload_addr(pkt), payload_size(pkt)); - calc_icrc = cpu_to_be32(~calc_icrc); + calc_icrc = (__force u32)cpu_to_be32(~calc_icrc); if (unlikely(calc_icrc != pack_icrc)) { if (skb->protocol == htons(ETH_P_IPV6)) pr_warn_ratelimited("bad ICRC from %pI6c\n", -- cgit v1.2.3 From 2bec3baded4ff46f51c1810e0dd8dfda55f05451 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:41 -0800 Subject: IB/rxe: Constify the pool name Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_pool.c | 2 +- drivers/infiniband/sw/rxe/rxe_pool.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index d723947a8542..238a6be00d46 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -102,7 +102,7 @@ struct rxe_type_info rxe_type_info[RXE_NUM_TYPES] = { }, }; -static inline char *pool_name(struct rxe_pool *pool) +static inline const char *pool_name(struct rxe_pool *pool) { return rxe_type_info[pool->type].name; } diff --git a/drivers/infiniband/sw/rxe/rxe_pool.h b/drivers/infiniband/sw/rxe/rxe_pool.h index 4d04830adcae..7846ccc58b25 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.h +++ b/drivers/infiniband/sw/rxe/rxe_pool.h @@ -58,7 +58,7 @@ enum rxe_elem_type { }; struct rxe_type_info { - char *name; + const char *name; size_t size; void (*cleanup)(void *obj); enum rxe_pool_flags flags; -- cgit v1.2.3 From c8b82182cb35c256cc0417c1d31a8421d7c87736 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:42 -0800 Subject: IB/rxe: Remove an unused function Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_pool.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index 238a6be00d46..4fe11148b304 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -112,13 +112,6 @@ static inline struct kmem_cache *pool_cache(struct rxe_pool *pool) return rxe_type_info[pool->type].cache; } -static inline enum rxe_elem_type rxe_type(void *arg) -{ - struct rxe_pool_entry *elem = arg; - - return elem->pool->type; -} - int rxe_cache_init(void) { int err; -- cgit v1.2.3 From 175f1244c199f589d23f54fcc3e178fe24d80b7a Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:43 -0800 Subject: IB/rxe: Remove an unused variable and an unused argument The variable 'av' is not used so remove it. Since that change removes the last user of the 'wqe' argument, remove that argument too. Signed-off-by: Bart Van Assche Cc: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_req.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 73d4a97603a1..310cd3350578 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -361,19 +361,14 @@ static inline int check_init_depth(struct rxe_qp *qp, struct rxe_send_wqe *wqe) return -EAGAIN; } -static inline int get_mtu(struct rxe_qp *qp, struct rxe_send_wqe *wqe) +static inline int get_mtu(struct rxe_qp *qp) { struct rxe_dev *rxe = to_rdev(qp->ibqp.device); - struct rxe_port *port; - struct rxe_av *av; if ((qp_type(qp) == IB_QPT_RC) || (qp_type(qp) == IB_QPT_UC)) return qp->mtu; - av = &wqe->av; - port = &rxe->port; - - return port->mtu_cap; + return rxe->port.mtu_cap; } static struct sk_buff *init_req_packet(struct rxe_qp *qp, @@ -679,7 +674,7 @@ next_wqe: goto exit; } - mtu = get_mtu(qp, wqe); + mtu = get_mtu(qp); payload = (mask & RXE_WRITE_OR_SEND) ? wqe->dma.resid : 0; if (payload > mtu) { if (qp_type(qp) == IB_QPT_UD) { -- cgit v1.2.3 From 967335ab906d8bfba09bf179ecfbada927861eb9 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:44 -0800 Subject: IB/rxe: Remove superfluous casts Casting a pointer to 'void *' explicitly is not necessary in C code. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_pool.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index 4fe11148b304..7d1e2862b928 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -454,7 +454,7 @@ void *rxe_pool_get_index(struct rxe_pool *pool, u32 index) out: spin_unlock_irqrestore(&pool->pool_lock, flags); - return node ? (void *)elem : NULL; + return node ? elem : NULL; } void *rxe_pool_get_key(struct rxe_pool *pool, void *key) @@ -490,5 +490,5 @@ void *rxe_pool_get_key(struct rxe_pool *pool, void *key) out: spin_unlock_irqrestore(&pool->pool_lock, flags); - return node ? ((void *)elem) : NULL; + return node ? elem : NULL; } -- cgit v1.2.3 From 046ef24d256bccc3767eb6feef1dbe07338fc5c5 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:45 -0800 Subject: IB/rxe: Enable type checking on SKB_TO_PKT() and PKT_TO_SKB() arguments Let the compiler check the type of the arguments passed to SKB_TO_PKT() and PKT_TO_SKB(). Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_hdr.h | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_hdr.h b/drivers/infiniband/sw/rxe/rxe_hdr.h index d57b5e956ceb..6cb18406f5b8 100644 --- a/drivers/infiniband/sw/rxe/rxe_hdr.h +++ b/drivers/infiniband/sw/rxe/rxe_hdr.h @@ -53,8 +53,16 @@ struct rxe_pkt_info { }; /* Macros should be used only for received skb */ -#define SKB_TO_PKT(skb) ((struct rxe_pkt_info *)(skb)->cb) -#define PKT_TO_SKB(pkt) container_of((void *)(pkt), struct sk_buff, cb) +static inline struct rxe_pkt_info *SKB_TO_PKT(struct sk_buff *skb) +{ + BUILD_BUG_ON(sizeof(struct rxe_pkt_info) > sizeof(skb->cb)); + return (void *)skb->cb; +} + +static inline struct sk_buff *PKT_TO_SKB(struct rxe_pkt_info *pkt) +{ + return container_of((void *)pkt, struct sk_buff, cb); +} /* * IBA header types and methods -- cgit v1.2.3 From 32404fb76408e873d79d0723fda3207b18d235ad Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:46 -0800 Subject: IB/rxe: Let the compiler check the type of the cleanup functions Change the argument type of these functions from void * into struct rxe_pool_entry *. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_cq.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_loc.h | 8 ++++---- drivers/infiniband/sw/rxe/rxe_mcast.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_mr.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_pool.h | 6 ++++-- drivers/infiniband/sw/rxe/rxe_qp.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_verbs.h | 2 +- 7 files changed, 17 insertions(+), 15 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_cq.c b/drivers/infiniband/sw/rxe/rxe_cq.c index e5e6a5e7dee9..49fe42c23f4d 100644 --- a/drivers/infiniband/sw/rxe/rxe_cq.c +++ b/drivers/infiniband/sw/rxe/rxe_cq.c @@ -156,9 +156,9 @@ int rxe_cq_post(struct rxe_cq *cq, struct rxe_cqe *cqe, int solicited) return 0; } -void rxe_cq_cleanup(void *arg) +void rxe_cq_cleanup(struct rxe_pool_entry *arg) { - struct rxe_cq *cq = arg; + struct rxe_cq *cq = container_of(arg, typeof(*cq), pelem); if (cq->queue) rxe_queue_cleanup(cq->queue); diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index efe4c6a35442..da191d7acb6f 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -64,7 +64,7 @@ int rxe_cq_resize_queue(struct rxe_cq *cq, int new_cqe, struct ib_udata *udata); int rxe_cq_post(struct rxe_cq *cq, struct rxe_cqe *cqe, int solicited); -void rxe_cq_cleanup(void *arg); +void rxe_cq_cleanup(struct rxe_pool_entry *arg); /* rxe_mcast.c */ int rxe_mcast_get_grp(struct rxe_dev *rxe, union ib_gid *mgid, @@ -78,7 +78,7 @@ int rxe_mcast_drop_grp_elem(struct rxe_dev *rxe, struct rxe_qp *qp, void rxe_drop_all_mcast_groups(struct rxe_qp *qp); -void rxe_mc_cleanup(void *arg); +void rxe_mc_cleanup(struct rxe_pool_entry *arg); /* rxe_mmap.c */ struct rxe_mmap_info { @@ -137,7 +137,7 @@ int mem_check_range(struct rxe_mem *mem, u64 iova, size_t length); int rxe_mem_map_pages(struct rxe_dev *rxe, struct rxe_mem *mem, u64 *page, int num_pages, u64 iova); -void rxe_mem_cleanup(void *arg); +void rxe_mem_cleanup(struct rxe_pool_entry *arg); int advance_dma_data(struct rxe_dma_info *dma, unsigned int length); @@ -162,7 +162,7 @@ void rxe_qp_error(struct rxe_qp *qp); void rxe_qp_destroy(struct rxe_qp *qp); -void rxe_qp_cleanup(void *arg); +void rxe_qp_cleanup(struct rxe_pool_entry *arg); static inline int qp_num(struct rxe_qp *qp) { diff --git a/drivers/infiniband/sw/rxe/rxe_mcast.c b/drivers/infiniband/sw/rxe/rxe_mcast.c index fa95544ca7e0..e0fb6752f90e 100644 --- a/drivers/infiniband/sw/rxe/rxe_mcast.c +++ b/drivers/infiniband/sw/rxe/rxe_mcast.c @@ -180,9 +180,9 @@ void rxe_drop_all_mcast_groups(struct rxe_qp *qp) } } -void rxe_mc_cleanup(void *arg) +void rxe_mc_cleanup(struct rxe_pool_entry *arg) { - struct rxe_mc_grp *grp = arg; + struct rxe_mc_grp *grp = container_of(arg, typeof(*grp), pelem); struct rxe_dev *rxe = grp->rxe; rxe_drop_key(grp); diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index d0faca294006..8ca3acd327b3 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -89,9 +89,9 @@ static void rxe_mem_init(int access, struct rxe_mem *mem) mem->map_shift = ilog2(RXE_BUF_PER_MAP); } -void rxe_mem_cleanup(void *arg) +void rxe_mem_cleanup(struct rxe_pool_entry *arg) { - struct rxe_mem *mem = arg; + struct rxe_mem *mem = container_of(arg, typeof(*mem), pelem); int i; if (mem->umem) diff --git a/drivers/infiniband/sw/rxe/rxe_pool.h b/drivers/infiniband/sw/rxe/rxe_pool.h index 7846ccc58b25..47df28e43acf 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.h +++ b/drivers/infiniband/sw/rxe/rxe_pool.h @@ -57,10 +57,12 @@ enum rxe_elem_type { RXE_NUM_TYPES, /* keep me last */ }; +struct rxe_pool_entry; + struct rxe_type_info { const char *name; size_t size; - void (*cleanup)(void *obj); + void (*cleanup)(struct rxe_pool_entry *obj); enum rxe_pool_flags flags; u32 max_index; u32 min_index; @@ -91,7 +93,7 @@ struct rxe_pool { spinlock_t pool_lock; /* pool spinlock */ size_t elem_size; struct kref ref_cnt; - void (*cleanup)(void *obj); + void (*cleanup)(struct rxe_pool_entry *obj); enum rxe_pool_state state; enum rxe_pool_flags flags; enum rxe_elem_type type; diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index 486d576e55bc..917147ce4cf9 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -825,9 +825,9 @@ void rxe_qp_destroy(struct rxe_qp *qp) } /* called when the last reference to the qp is dropped */ -void rxe_qp_cleanup(void *arg) +void rxe_qp_cleanup(struct rxe_pool_entry *arg) { - struct rxe_qp *qp = arg; + struct rxe_qp *qp = container_of(arg, typeof(*qp), pelem); rxe_drop_all_mcast_groups(qp); diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index cac1d52a08f0..536974b69ed9 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -475,6 +475,6 @@ static inline struct rxe_mem *to_rmw(struct ib_mw *mw) int rxe_register_device(struct rxe_dev *rxe); int rxe_unregister_device(struct rxe_dev *rxe); -void rxe_mc_cleanup(void *arg); +void rxe_mc_cleanup(struct rxe_pool_entry *arg); #endif /* RXE_VERBS_H */ -- cgit v1.2.3 From 43553b47c3124e7f4a9f115cff9db513b35a5e0a Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:47 -0800 Subject: IB/rxe: Issue warnings once It is strongly recommended to report kernel warnings once instead of every time a condition is hit. Hence change WARN_ON() into WARN_ON_ONCE() / BUILD_BUG_ON() as appropriate. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_comp.c | 2 +- drivers/infiniband/sw/rxe/rxe_mr.c | 6 +++--- drivers/infiniband/sw/rxe/rxe_resp.c | 11 ++++++----- 3 files changed, 10 insertions(+), 9 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index d369f24425f9..e912e5396e8c 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -254,7 +254,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, } break; default: - WARN_ON(1); + WARN_ON_ONCE(1); } /* Check operation validity. */ diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index 8ca3acd327b3..8cf38b253c37 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -123,7 +123,7 @@ static int rxe_mem_alloc(struct rxe_dev *rxe, struct rxe_mem *mem, int num_buf) goto err2; } - WARN_ON(!is_power_of_2(RXE_BUF_PER_MAP)); + BUILD_BUG_ON(!is_power_of_2(RXE_BUF_PER_MAP)); mem->map_shift = ilog2(RXE_BUF_PER_MAP); mem->map_mask = RXE_BUF_PER_MAP - 1; @@ -189,7 +189,7 @@ int rxe_mem_init_user(struct rxe_dev *rxe, struct rxe_pd *pd, u64 start, goto err1; } - WARN_ON(!is_power_of_2(umem->page_size)); + WARN_ON_ONCE(!is_power_of_2(umem->page_size)); mem->page_shift = ilog2(umem->page_size); mem->page_mask = umem->page_size - 1; @@ -375,7 +375,7 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length, return 0; } - WARN_ON(!mem->map); + WARN_ON_ONCE(!mem->map); err = mem_check_range(mem, iova, length); if (err) { diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 3435efff8799..6dbd069689fc 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -307,7 +307,7 @@ static enum resp_states check_op_valid(struct rxe_qp *qp, break; default: - WARN_ON(1); + WARN_ON_ONCE(1); break; } @@ -495,7 +495,7 @@ static enum resp_states check_rkey(struct rxe_qp *qp, } } - WARN_ON(qp->resp.mr); + WARN_ON_ONCE(qp->resp.mr); qp->resp.mr = mem; return RESPST_EXECUTE; @@ -808,9 +808,10 @@ static enum resp_states execute(struct rxe_qp *qp, struct rxe_pkt_info *pkt) err = process_atomic(qp, pkt); if (err) return err; - } else + } else { /* Unreachable */ - WARN_ON(1); + WARN_ON_ONCE(1); + } /* We successfully processed this new request. */ qp->resp.msn++; @@ -1396,7 +1397,7 @@ int rxe_responder(void *arg) goto exit; default: - WARN_ON(1); + WARN_ON_ONCE(1); } } -- cgit v1.2.3 From 642c7cbcaf2ffc1e27f67eda3dc47347ac5aff37 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:48 -0800 Subject: IB/rxe: Add a runtime check in alloc_index() Since index values equal to or above 'range' can trigger memory corruption, complain if index >= range. Signed-off-by: Bart Van Assche Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_pool.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index 7d1e2862b928..75d11ee635ec 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -266,6 +266,7 @@ static u32 alloc_index(struct rxe_pool *pool) if (index >= range) index = find_first_zero_bit(pool->table, range); + WARN_ON_ONCE(index >= range); set_bit(index, pool->table); pool->last = index; return index + pool->min_index; -- cgit v1.2.3 From 723ec9ae2adfe638d4e2bbc44dd3974ac974a6c6 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:49 -0800 Subject: IB/rxe: Introduce functions for queue draining This change makes the code easier to read and avoids that code is duplicated. Signed-off-by: Bart Van Assche Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_comp.c | 63 +++++++++++++----------------------- drivers/infiniband/sw/rxe/rxe_resp.c | 28 ++++++++-------- 2 files changed, 38 insertions(+), 53 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index e912e5396e8c..6769a075501e 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -503,57 +503,40 @@ static inline enum comp_state complete_wqe(struct rxe_qp *qp, return COMPST_GET_WQE; } -int rxe_completer(void *arg) +static void rxe_drain_resp_pkts(struct rxe_qp *qp, bool notify) { - struct rxe_qp *qp = (struct rxe_qp *)arg; - struct rxe_send_wqe *wqe = wqe; - struct sk_buff *skb = NULL; - struct rxe_pkt_info *pkt = NULL; - enum comp_state state; - - rxe_add_ref(qp); - - if (!qp->valid) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; - - while (queue_head(qp->sq.queue)) - advance_consumer(qp->sq.queue); + struct sk_buff *skb; + struct rxe_send_wqe *wqe; - goto exit; + while ((skb = skb_dequeue(&qp->resp_pkts))) { + rxe_drop_ref(qp); + kfree_skb(skb); } - if (qp->req.state == QP_STATE_ERROR) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; - - while ((wqe = queue_head(qp->sq.queue))) { + while ((wqe = queue_head(qp->sq.queue))) { + if (notify) { wqe->status = IB_WC_WR_FLUSH_ERR; do_complete(qp, wqe); + } else { + advance_consumer(qp->sq.queue); } - - goto exit; } +} - if (qp->req.state == QP_STATE_RESET) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; +int rxe_completer(void *arg) +{ + struct rxe_qp *qp = (struct rxe_qp *)arg; + struct rxe_send_wqe *wqe = wqe; + struct sk_buff *skb = NULL; + struct rxe_pkt_info *pkt = NULL; + enum comp_state state; - while (queue_head(qp->sq.queue)) - advance_consumer(qp->sq.queue); + rxe_add_ref(qp); + if (!qp->valid || qp->req.state == QP_STATE_ERROR || + qp->req.state == QP_STATE_RESET) { + rxe_drain_resp_pkts(qp, qp->valid && + qp->req.state == QP_STATE_ERROR); goto exit; } diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 6dbd069689fc..51c134dbc6c8 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -1207,6 +1207,19 @@ static enum resp_states do_class_d1e_error(struct rxe_qp *qp) } } +static void rxe_drain_req_pkts(struct rxe_qp *qp) +{ + struct sk_buff *skb; + + while ((skb = skb_dequeue(&qp->req_pkts))) { + rxe_drop_ref(qp); + kfree_skb(skb); + } + + while (!qp->srq && qp->rq.queue && queue_head(qp->rq.queue)) + advance_consumer(qp->rq.queue); +} + int rxe_responder(void *arg) { struct rxe_qp *qp = (struct rxe_qp *)arg; @@ -1374,21 +1387,10 @@ int rxe_responder(void *arg) goto exit; - case RESPST_RESET: { - struct sk_buff *skb; - - while ((skb = skb_dequeue(&qp->req_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - - while (!qp->srq && qp->rq.queue && - queue_head(qp->rq.queue)) - advance_consumer(qp->rq.queue); - + case RESPST_RESET: + rxe_drain_req_pkts(qp); qp->resp.wqe = NULL; goto exit; - } case RESPST_ERROR: qp->resp.goto_error = 0; -- cgit v1.2.3 From 18d3451c0d7e23d155db37b9ec1a8886c114a5d3 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:50 -0800 Subject: IB/rxe: Generate a completion for all failed work requests Change do_complete() such that an error completion is not only generated if a QP is in the error state but also if a work request failed. Signed-off-by: Bart Van Assche Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_comp.c | 11 ++++++++++- drivers/infiniband/sw/rxe/rxe_loc.h | 1 + drivers/infiniband/sw/rxe/rxe_req.c | 18 +++++++----------- drivers/infiniband/sw/rxe/rxe_resp.c | 4 ++-- 4 files changed, 20 insertions(+), 14 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index 6769a075501e..91317c159b9a 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -412,13 +412,21 @@ static void make_send_cqe(struct rxe_qp *qp, struct rxe_send_wqe *wqe, } } +/* + * IBA Spec. Section 10.7.3.1 SIGNALED COMPLETIONS + * ---------8<---------8<------------- + * ...Note that if a completion error occurs, a Work Completion + * will always be generated, even if the signaling + * indicator requests an Unsignaled Completion. + * ---------8<---------8<------------- + */ static void do_complete(struct rxe_qp *qp, struct rxe_send_wqe *wqe) { struct rxe_cqe cqe; if ((qp->sq_sig_type == IB_SIGNAL_ALL_WR) || (wqe->wr.send_flags & IB_SEND_SIGNALED) || - (qp->req.state == QP_STATE_ERROR)) { + wqe->status != IB_WC_SUCCESS) { make_send_cqe(qp, wqe, &cqe); advance_consumer(qp->sq.queue); rxe_cq_post(qp->scq, &cqe, 0); @@ -709,6 +717,7 @@ int rxe_completer(void *arg) break; case COMPST_ERROR: + WARN_ON_ONCE(wqe->status == IB_WC_SUCCESS); do_complete(qp, wqe); rxe_qp_error(qp); diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index da191d7acb6f..bdec460f1fce 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -225,6 +225,7 @@ extern struct ib_dma_mapping_ops rxe_dma_mapping_ops; void rxe_release(struct kref *kref); +void rxe_drain_req_pkts(struct rxe_qp *qp, bool notify); int rxe_completer(void *arg); int rxe_requester(void *arg); int rxe_responder(void *arg); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 310cd3350578..d62be4828899 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -594,9 +594,14 @@ int rxe_requester(void *arg) rxe_add_ref(qp); next_wqe: - if (unlikely(!qp->valid || qp->req.state == QP_STATE_ERROR)) + if (unlikely(!qp->valid)) goto exit; + if (unlikely(qp->req.state == QP_STATE_ERROR)) { + rxe_drain_req_pkts(qp, true); + goto exit; + } + if (unlikely(qp->req.state == QP_STATE_RESET)) { qp->req.wqe_index = consumer_index(qp->sq.queue); qp->req.opcode = -1; @@ -743,17 +748,8 @@ err: kfree_skb(skb); wqe->status = IB_WC_LOC_PROT_ERR; wqe->state = wqe_state_error; - - /* - * IBA Spec. Section 10.7.3.1 SIGNALED COMPLETIONS - * ---------8<---------8<------------- - * ...Note that if a completion error occurs, a Work Completion - * will always be generated, even if the signaling - * indicator requests an Unsignaled Completion. - * ---------8<---------8<------------- - */ - wqe->wr.send_flags |= IB_SEND_SIGNALED; __rxe_do_task(&qp->comp.task); + exit: rxe_drop_ref(qp); return -EAGAIN; diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 51c134dbc6c8..33defaddc000 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -1207,7 +1207,7 @@ static enum resp_states do_class_d1e_error(struct rxe_qp *qp) } } -static void rxe_drain_req_pkts(struct rxe_qp *qp) +void rxe_drain_req_pkts(struct rxe_qp *qp, bool notify) { struct sk_buff *skb; @@ -1388,7 +1388,7 @@ int rxe_responder(void *arg) goto exit; case RESPST_RESET: - rxe_drain_req_pkts(qp); + rxe_drain_req_pkts(qp, false); qp->resp.wqe = NULL; goto exit; -- cgit v1.2.3 From b3a459961014b14c267544c327db033669493295 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:51 -0800 Subject: IB/rxe: Fix a MR reference leak in check_rkey() Avoid that calling check_rkey() for mem->state == RXE_MEM_STATE_FREE triggers an MR reference leak. Signed-off-by: Bart Van Assche Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_resp.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 33defaddc000..60d78f45aa04 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -418,7 +418,7 @@ static enum resp_states check_length(struct rxe_qp *qp, static enum resp_states check_rkey(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { - struct rxe_mem *mem; + struct rxe_mem *mem = NULL; u64 va; u32 rkey; u32 resid; @@ -459,38 +459,38 @@ static enum resp_states check_rkey(struct rxe_qp *qp, mem = lookup_mem(qp->pd, access, rkey, lookup_remote); if (!mem) { state = RESPST_ERR_RKEY_VIOLATION; - goto err1; + goto err; } if (unlikely(mem->state == RXE_MEM_STATE_FREE)) { state = RESPST_ERR_RKEY_VIOLATION; - goto err1; + goto err; } if (mem_check_range(mem, va, resid)) { state = RESPST_ERR_RKEY_VIOLATION; - goto err2; + goto err; } if (pkt->mask & RXE_WRITE_MASK) { if (resid > mtu) { if (pktlen != mtu || bth_pad(pkt)) { state = RESPST_ERR_LENGTH; - goto err2; + goto err; } resid = mtu; } else { if (pktlen != resid) { state = RESPST_ERR_LENGTH; - goto err2; + goto err; } if ((bth_pad(pkt) != (0x3 & (-resid)))) { /* This case may not be exactly that * but nothing else fits. */ state = RESPST_ERR_LENGTH; - goto err2; + goto err; } } } @@ -500,9 +500,9 @@ static enum resp_states check_rkey(struct rxe_qp *qp, qp->resp.mr = mem; return RESPST_EXECUTE; -err2: - rxe_drop_ref(mem); -err1: +err: + if (mem) + rxe_drop_ref(mem); return state; } -- cgit v1.2.3 From ab17654476a11a1ed7d89f1104e2acdb7ed1c9ed Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:52 -0800 Subject: IB/rxe: Fix reference leaks in memory key invalidation code Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_req.c | 1 + drivers/infiniband/sw/rxe/rxe_resp.c | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index d62be4828899..512e74326b16 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -635,6 +635,7 @@ next_wqe: goto exit; } rmr->state = RXE_MEM_STATE_FREE; + rxe_drop_ref(rmr); wqe->state = wqe_state_done; wqe->status = IB_WC_SUCCESS; } else if (wqe->wr.opcode == IB_WR_REG_MR) { diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 60d78f45aa04..05f374986cea 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -907,6 +907,7 @@ static enum resp_states do_complete(struct rxe_qp *qp, return RESPST_ERROR; } rmr->state = RXE_MEM_STATE_FREE; + rxe_drop_ref(rmr); } wc->qp = &qp->ibqp; -- cgit v1.2.3 From 839f5ac0d806970a102117be03ef05272c50d20e Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:53 -0800 Subject: IB/rxe: Remove a pointless indirection layer Neither rxe->ifc_ops nor any of the function pointers in struct struct rxe_ifc_ops ever change. Hence remove the rxe->ifc_ops indirection mechanism. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Reviewed-by: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe.c | 2 +- drivers/infiniband/sw/rxe/rxe_loc.h | 20 ++++++++++++++-- drivers/infiniband/sw/rxe/rxe_mcast.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_net.c | 43 +++++++++++------------------------ drivers/infiniband/sw/rxe/rxe_req.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_resp.c | 4 ++-- drivers/infiniband/sw/rxe/rxe_verbs.c | 10 ++++---- drivers/infiniband/sw/rxe/rxe_verbs.h | 22 ------------------ 8 files changed, 42 insertions(+), 67 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c index ab6c3c25d7ff..b12dd9b5a89d 100644 --- a/drivers/infiniband/sw/rxe/rxe.c +++ b/drivers/infiniband/sw/rxe/rxe.c @@ -178,7 +178,7 @@ static int rxe_init_ports(struct rxe_dev *rxe) return -ENOMEM; port->pkey_tbl[0] = 0xffff; - port->port_guid = rxe->ifc_ops->port_guid(rxe); + port->port_guid = rxe_port_guid(rxe); spin_lock_init(&port->port_lock); diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index bdec460f1fce..272337e5e948 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -141,6 +141,22 @@ void rxe_mem_cleanup(struct rxe_pool_entry *arg); int advance_dma_data(struct rxe_dma_info *dma, unsigned int length); +/* rxe_net.c */ +int rxe_loopback(struct sk_buff *skb); +int rxe_send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb); +__be64 rxe_port_guid(struct rxe_dev *rxe); +struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av, + int paylen, struct rxe_pkt_info *pkt); +int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb, u32 *crc); +enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num); +const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num); +struct device *rxe_dma_device(struct rxe_dev *rxe); +__be64 rxe_node_guid(struct rxe_dev *rxe); +int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid); +int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid); + /* rxe_qp.c */ int rxe_qp_chk_init(struct rxe_dev *rxe, struct ib_qp_init_attr *init); @@ -257,9 +273,9 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp, if (pkt->mask & RXE_LOOPBACK_MASK) { memcpy(SKB_TO_PKT(skb), pkt, sizeof(*pkt)); - err = rxe->ifc_ops->loopback(skb); + err = rxe_loopback(skb); } else { - err = rxe->ifc_ops->send(rxe, pkt, skb); + err = rxe_send(rxe, pkt, skb); } if (err) { diff --git a/drivers/infiniband/sw/rxe/rxe_mcast.c b/drivers/infiniband/sw/rxe/rxe_mcast.c index e0fb6752f90e..522a7942c56c 100644 --- a/drivers/infiniband/sw/rxe/rxe_mcast.c +++ b/drivers/infiniband/sw/rxe/rxe_mcast.c @@ -61,7 +61,7 @@ int rxe_mcast_get_grp(struct rxe_dev *rxe, union ib_gid *mgid, rxe_add_key(grp, mgid); - err = rxe->ifc_ops->mcast_add(rxe, mgid); + err = rxe_mcast_add(rxe, mgid); if (err) goto err2; @@ -186,5 +186,5 @@ void rxe_mc_cleanup(struct rxe_pool_entry *arg) struct rxe_dev *rxe = grp->rxe; rxe_drop_key(grp); - rxe->ifc_ops->mcast_delete(rxe, &grp->mgid); + rxe_mcast_delete(rxe, &grp->mgid); } diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index 151f639abebf..50144c307eb4 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -102,17 +102,17 @@ static __be64 rxe_mac_to_eui64(struct net_device *ndev) return eui64; } -static __be64 node_guid(struct rxe_dev *rxe) +__be64 rxe_node_guid(struct rxe_dev *rxe) { return rxe_mac_to_eui64(rxe->ndev); } -static __be64 port_guid(struct rxe_dev *rxe) +__be64 rxe_port_guid(struct rxe_dev *rxe) { return rxe_mac_to_eui64(rxe->ndev); } -static struct device *dma_device(struct rxe_dev *rxe) +struct device *rxe_dma_device(struct rxe_dev *rxe) { struct net_device *ndev; @@ -124,7 +124,7 @@ static struct device *dma_device(struct rxe_dev *rxe) return ndev->dev.parent; } -static int mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) +int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) { int err; unsigned char ll_addr[ETH_ALEN]; @@ -135,7 +135,7 @@ static int mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) return err; } -static int mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid) +int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid) { int err; unsigned char ll_addr[ETH_ALEN]; @@ -397,8 +397,8 @@ static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, return 0; } -static int prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb, u32 *crc) +int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb, u32 *crc) { int err = 0; struct rxe_av *av = rxe_get_av(pkt); @@ -424,8 +424,7 @@ static void rxe_skb_tx_dtor(struct sk_buff *skb) rxe_run_task(&qp->req.task, 1); } -static int send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb) +int rxe_send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb) { struct sk_buff *nskb; struct rxe_av *av; @@ -461,7 +460,7 @@ static int send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, return 0; } -static int loopback(struct sk_buff *skb) +int rxe_loopback(struct sk_buff *skb) { return rxe_rcv(skb); } @@ -471,8 +470,8 @@ static inline int addr_same(struct rxe_dev *rxe, struct rxe_av *av) return rxe->port.port_guid == av->grh.dgid.global.interface_id; } -static struct sk_buff *init_packet(struct rxe_dev *rxe, struct rxe_av *av, - int paylen, struct rxe_pkt_info *pkt) +struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av, + int paylen, struct rxe_pkt_info *pkt) { unsigned int hdr_len; struct sk_buff *skb; @@ -511,31 +510,16 @@ static struct sk_buff *init_packet(struct rxe_dev *rxe, struct rxe_av *av, * this is required by rxe_cfg to match rxe devices in * /sys/class/infiniband up with their underlying ethernet devices */ -static char *parent_name(struct rxe_dev *rxe, unsigned int port_num) +const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num) { return rxe->ndev->name; } -static enum rdma_link_layer link_layer(struct rxe_dev *rxe, - unsigned int port_num) +enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num) { return IB_LINK_LAYER_ETHERNET; } -static struct rxe_ifc_ops ifc_ops = { - .node_guid = node_guid, - .port_guid = port_guid, - .dma_device = dma_device, - .mcast_add = mcast_add, - .mcast_delete = mcast_delete, - .prepare = prepare, - .send = send, - .loopback = loopback, - .init_packet = init_packet, - .parent_name = parent_name, - .link_layer = link_layer, -}; - struct rxe_dev *rxe_net_add(struct net_device *ndev) { int err; @@ -545,7 +529,6 @@ struct rxe_dev *rxe_net_add(struct net_device *ndev) if (!rxe) return NULL; - rxe->ifc_ops = &ifc_ops; rxe->ndev = ndev; err = rxe_add(rxe, ndev->mtu); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 512e74326b16..dbfde0dc6ff7 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -404,7 +404,7 @@ static struct sk_buff *init_req_packet(struct rxe_qp *qp, /* init skb */ av = rxe_get_av(pkt); - skb = rxe->ifc_ops->init_packet(rxe, av, paylen, pkt); + skb = rxe_init_packet(rxe, av, paylen, pkt); if (unlikely(!skb)) return NULL; @@ -475,7 +475,7 @@ static int fill_packet(struct rxe_qp *qp, struct rxe_send_wqe *wqe, u32 *p; int err; - err = rxe->ifc_ops->prepare(rxe, pkt, skb, &crc); + err = rxe_prepare(rxe, pkt, skb, &crc); if (err) return err; diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 05f374986cea..7bf20ced2078 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -608,7 +608,7 @@ static struct sk_buff *prepare_ack_packet(struct rxe_qp *qp, pad = (-payload) & 0x3; paylen = rxe_opcode[opcode].length + payload + pad + RXE_ICRC_SIZE; - skb = rxe->ifc_ops->init_packet(rxe, &qp->pri_av, paylen, ack); + skb = rxe_init_packet(rxe, &qp->pri_av, paylen, ack); if (!skb) return NULL; @@ -637,7 +637,7 @@ static struct sk_buff *prepare_ack_packet(struct rxe_qp *qp, if (ack->mask & RXE_ATMACK_MASK) atmack_set_orig(ack, qp->resp.atomic_orig); - err = rxe->ifc_ops->prepare(rxe, ack, skb, &crc); + err = rxe_prepare(rxe, ack, skb, &crc); if (err) { kfree_skb(skb); return NULL; diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c index beb7021ff18a..e4de37fb9aab 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.c +++ b/drivers/infiniband/sw/rxe/rxe_verbs.c @@ -234,7 +234,7 @@ static enum rdma_link_layer rxe_get_link_layer(struct ib_device *dev, { struct rxe_dev *rxe = to_rdev(dev); - return rxe->ifc_ops->link_layer(rxe, port_num); + return rxe_link_layer(rxe, port_num); } static struct ib_ucontext *rxe_alloc_ucontext(struct ib_device *dev, @@ -1209,10 +1209,8 @@ static ssize_t rxe_show_parent(struct device *device, { struct rxe_dev *rxe = container_of(device, struct rxe_dev, ib_dev.dev); - char *name; - name = rxe->ifc_ops->parent_name(rxe, 1); - return snprintf(buf, 16, "%s\n", name); + return snprintf(buf, 16, "%s\n", rxe_parent_name(rxe, 1)); } static DEVICE_ATTR(parent, S_IRUGO, rxe_show_parent, NULL); @@ -1234,9 +1232,9 @@ int rxe_register_device(struct rxe_dev *rxe) dev->node_type = RDMA_NODE_IB_CA; dev->phys_port_cnt = 1; dev->num_comp_vectors = RXE_NUM_COMP_VECTORS; - dev->dma_device = rxe->ifc_ops->dma_device(rxe); + dev->dma_device = rxe_dma_device(rxe); dev->local_dma_lkey = 0; - dev->node_guid = rxe->ifc_ops->node_guid(rxe); + dev->node_guid = rxe_node_guid(rxe); dev->dma_ops = &rxe_dma_mapping_ops; dev->uverbs_abi_ver = RXE_UVERBS_ABI_VERSION; diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index 536974b69ed9..e100c500ae85 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -372,26 +372,6 @@ struct rxe_port { u32 qp_gsi_index; }; -/* callbacks from rdma_rxe to network interface layer */ -struct rxe_ifc_ops { - void (*release)(struct rxe_dev *rxe); - __be64 (*node_guid)(struct rxe_dev *rxe); - __be64 (*port_guid)(struct rxe_dev *rxe); - struct device *(*dma_device)(struct rxe_dev *rxe); - int (*mcast_add)(struct rxe_dev *rxe, union ib_gid *mgid); - int (*mcast_delete)(struct rxe_dev *rxe, union ib_gid *mgid); - int (*prepare)(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb, u32 *crc); - int (*send)(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb); - int (*loopback)(struct sk_buff *skb); - struct sk_buff *(*init_packet)(struct rxe_dev *rxe, struct rxe_av *av, - int paylen, struct rxe_pkt_info *pkt); - char *(*parent_name)(struct rxe_dev *rxe, unsigned int port_num); - enum rdma_link_layer (*link_layer)(struct rxe_dev *rxe, - unsigned int port_num); -}; - struct rxe_dev { struct ib_device ib_dev; struct ib_device_attr attr; @@ -400,8 +380,6 @@ struct rxe_dev { struct kref ref_cnt; struct mutex usdev_lock; - struct rxe_ifc_ops *ifc_ops; - struct net_device *ndev; int xmit_errors; -- cgit v1.2.3 From c5540a0195ec65e2ab9fcc4bde958db2cc66550c Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 10 Jan 2017 11:15:54 -0800 Subject: IB/rxe: Fix an skb leak Additionally, make it easier to detect skb leaks by issuing a warning if a leak occurs. Signed-off-by: Bart Van Assche Reviewed-by: Leon Romanovsky Cc: Andrew Boyer Cc: Moni Shoua Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_comp.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index 91317c159b9a..4cd55d5617f7 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -630,6 +630,7 @@ int rxe_completer(void *arg) if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } goto done; @@ -653,6 +654,7 @@ int rxe_completer(void *arg) qp->qp_timeout_jiffies) mod_timer(&qp->retrans_timer, jiffies + qp->qp_timeout_jiffies); + WARN_ON_ONCE(skb); goto exit; case COMPST_ERROR_RETRY: @@ -665,8 +667,10 @@ int rxe_completer(void *arg) */ /* there is nothing to retry in this case */ - if (!wqe || (wqe->state == wqe_state_posted)) + if (!wqe || (wqe->state == wqe_state_posted)) { + WARN_ON_ONCE(skb); goto exit; + } if (qp->comp.retry_cnt > 0) { if (qp->comp.retry_cnt != 7) @@ -688,8 +692,10 @@ int rxe_completer(void *arg) if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } + WARN_ON_ONCE(skb); goto exit; } else { @@ -709,6 +715,9 @@ int rxe_completer(void *arg) mod_timer(&qp->rnr_nak_timer, jiffies + rnrnak_jiffies(aeth_syn(pkt) & ~AETH_TYPE_MASK)); + rxe_drop_ref(pkt->qp); + kfree_skb(skb); + skb = NULL; goto exit; } else { wqe->status = IB_WC_RNR_RETRY_EXC_ERR; @@ -724,8 +733,10 @@ int rxe_completer(void *arg) if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } + WARN_ON_ONCE(skb); goto exit; } } @@ -734,6 +745,7 @@ exit: /* we come here if we are done with processing and want the task to * exit from the loop calling us */ + WARN_ON_ONCE(skb); rxe_drop_ref(qp); return -EAGAIN; @@ -741,6 +753,7 @@ done: /* we come here if we have processed a packet we want the task to call * us again to see if there is anything else to do */ + WARN_ON_ONCE(skb); rxe_drop_ref(qp); return 0; } -- cgit v1.2.3 From 6554c9f7f7499ee8229e0bdf0e873d9b7a93405d Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 16 Dec 2016 17:03:16 -0800 Subject: RDMA/nes: use designated initializers Prepare to mark sensitive kernel structures for randomization by making sure they're using designated initializers. These were identified during allyesconfig builds of x86, arm, and arm64, with most initializer fixes extracted from grsecurity. Signed-off-by: Kees Cook Signed-off-by: Doug Ledford --- drivers/infiniband/hw/nes/nes_cm.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index 8e703479e7ce..fb983df7c157 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -135,17 +135,17 @@ static void record_ird_ord(struct nes_cm_node *, u16, u16); /* instance of function pointers for client API */ /* set address of this instance to cm_core->cm_ops at cm_core alloc */ static const struct nes_cm_ops nes_cm_api = { - mini_cm_accelerated, - mini_cm_listen, - mini_cm_del_listen, - mini_cm_connect, - mini_cm_close, - mini_cm_accept, - mini_cm_reject, - mini_cm_recv_pkt, - mini_cm_dealloc_core, - mini_cm_get, - mini_cm_set + .accelerated = mini_cm_accelerated, + .listen = mini_cm_listen, + .stop_listener = mini_cm_del_listen, + .connect = mini_cm_connect, + .close = mini_cm_close, + .accept = mini_cm_accept, + .reject = mini_cm_reject, + .recv_pkt = mini_cm_recv_pkt, + .destroy_cm_core = mini_cm_dealloc_core, + .get = mini_cm_get, + .set = mini_cm_set }; static struct nes_cm_core *g_cm_core; -- cgit v1.2.3 From 7f6856b789ff13ccfd317c936e4b161c0bbd88a3 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 16 Dec 2016 17:05:42 -0800 Subject: RDMA/i40iw: use designated initializers Prepare to mark sensitive kernel structures for randomization by making sure they're using designated initializers. These were identified during allyesconfig builds of x86, arm, and arm64, with most initializer fixes extracted from grsecurity. Signed-off-by: Kees Cook Acked-by: Shiraz Saleem Signed-off-by: Doug Ledford --- drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 128 +++++++++++++++---------------- drivers/infiniband/hw/i40iw/i40iw_uk.c | 34 ++++---- 2 files changed, 80 insertions(+), 82 deletions(-) diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c index 98923a8cf86d..dced4f48cbdd 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -4851,46 +4851,46 @@ void i40iw_vsi_stats_free(struct i40iw_sc_vsi *vsi) } static struct i40iw_cqp_ops iw_cqp_ops = { - i40iw_sc_cqp_init, - i40iw_sc_cqp_create, - i40iw_sc_cqp_post_sq, - i40iw_sc_cqp_get_next_send_wqe, - i40iw_sc_cqp_destroy, - i40iw_sc_poll_for_cqp_op_done + .cqp_init = i40iw_sc_cqp_init, + .cqp_create = i40iw_sc_cqp_create, + .cqp_post_sq = i40iw_sc_cqp_post_sq, + .cqp_get_next_send_wqe = i40iw_sc_cqp_get_next_send_wqe, + .cqp_destroy = i40iw_sc_cqp_destroy, + .poll_for_cqp_op_done = i40iw_sc_poll_for_cqp_op_done }; static struct i40iw_ccq_ops iw_ccq_ops = { - i40iw_sc_ccq_init, - i40iw_sc_ccq_create, - i40iw_sc_ccq_destroy, - i40iw_sc_ccq_create_done, - i40iw_sc_ccq_get_cqe_info, - i40iw_sc_ccq_arm + .ccq_init = i40iw_sc_ccq_init, + .ccq_create = i40iw_sc_ccq_create, + .ccq_destroy = i40iw_sc_ccq_destroy, + .ccq_create_done = i40iw_sc_ccq_create_done, + .ccq_get_cqe_info = i40iw_sc_ccq_get_cqe_info, + .ccq_arm = i40iw_sc_ccq_arm }; static struct i40iw_ceq_ops iw_ceq_ops = { - i40iw_sc_ceq_init, - i40iw_sc_ceq_create, - i40iw_sc_cceq_create_done, - i40iw_sc_cceq_destroy_done, - i40iw_sc_cceq_create, - i40iw_sc_ceq_destroy, - i40iw_sc_process_ceq + .ceq_init = i40iw_sc_ceq_init, + .ceq_create = i40iw_sc_ceq_create, + .cceq_create_done = i40iw_sc_cceq_create_done, + .cceq_destroy_done = i40iw_sc_cceq_destroy_done, + .cceq_create = i40iw_sc_cceq_create, + .ceq_destroy = i40iw_sc_ceq_destroy, + .process_ceq = i40iw_sc_process_ceq }; static struct i40iw_aeq_ops iw_aeq_ops = { - i40iw_sc_aeq_init, - i40iw_sc_aeq_create, - i40iw_sc_aeq_destroy, - i40iw_sc_get_next_aeqe, - i40iw_sc_repost_aeq_entries, - i40iw_sc_aeq_create_done, - i40iw_sc_aeq_destroy_done + .aeq_init = i40iw_sc_aeq_init, + .aeq_create = i40iw_sc_aeq_create, + .aeq_destroy = i40iw_sc_aeq_destroy, + .get_next_aeqe = i40iw_sc_get_next_aeqe, + .repost_aeq_entries = i40iw_sc_repost_aeq_entries, + .aeq_create_done = i40iw_sc_aeq_create_done, + .aeq_destroy_done = i40iw_sc_aeq_destroy_done }; /* iwarp pd ops */ static struct i40iw_pd_ops iw_pd_ops = { - i40iw_sc_pd_init, + .pd_init = i40iw_sc_pd_init, }; static struct i40iw_priv_qp_ops iw_priv_qp_ops = { @@ -4909,53 +4909,51 @@ static struct i40iw_priv_qp_ops iw_priv_qp_ops = { }; static struct i40iw_priv_cq_ops iw_priv_cq_ops = { - i40iw_sc_cq_init, - i40iw_sc_cq_create, - i40iw_sc_cq_destroy, - i40iw_sc_cq_modify, + .cq_init = i40iw_sc_cq_init, + .cq_create = i40iw_sc_cq_create, + .cq_destroy = i40iw_sc_cq_destroy, + .cq_modify = i40iw_sc_cq_modify, }; static struct i40iw_mr_ops iw_mr_ops = { - i40iw_sc_alloc_stag, - i40iw_sc_mr_reg_non_shared, - i40iw_sc_mr_reg_shared, - i40iw_sc_dealloc_stag, - i40iw_sc_query_stag, - i40iw_sc_mw_alloc + .alloc_stag = i40iw_sc_alloc_stag, + .mr_reg_non_shared = i40iw_sc_mr_reg_non_shared, + .mr_reg_shared = i40iw_sc_mr_reg_shared, + .dealloc_stag = i40iw_sc_dealloc_stag, + .query_stag = i40iw_sc_query_stag, + .mw_alloc = i40iw_sc_mw_alloc }; static struct i40iw_cqp_misc_ops iw_cqp_misc_ops = { - i40iw_sc_manage_push_page, - i40iw_sc_manage_hmc_pm_func_table, - i40iw_sc_set_hmc_resource_profile, - i40iw_sc_commit_fpm_values, - i40iw_sc_query_fpm_values, - i40iw_sc_static_hmc_pages_allocated, - i40iw_sc_add_arp_cache_entry, - i40iw_sc_del_arp_cache_entry, - i40iw_sc_query_arp_cache_entry, - i40iw_sc_manage_apbvt_entry, - i40iw_sc_manage_qhash_table_entry, - i40iw_sc_alloc_local_mac_ipaddr_entry, - i40iw_sc_add_local_mac_ipaddr_entry, - i40iw_sc_del_local_mac_ipaddr_entry, - i40iw_sc_cqp_nop, - i40iw_sc_commit_fpm_values_done, - i40iw_sc_query_fpm_values_done, - i40iw_sc_manage_hmc_pm_func_table_done, - i40iw_sc_suspend_qp, - i40iw_sc_resume_qp + .manage_push_page = i40iw_sc_manage_push_page, + .manage_hmc_pm_func_table = i40iw_sc_manage_hmc_pm_func_table, + .set_hmc_resource_profile = i40iw_sc_set_hmc_resource_profile, + .commit_fpm_values = i40iw_sc_commit_fpm_values, + .query_fpm_values = i40iw_sc_query_fpm_values, + .static_hmc_pages_allocated = i40iw_sc_static_hmc_pages_allocated, + .add_arp_cache_entry = i40iw_sc_add_arp_cache_entry, + .del_arp_cache_entry = i40iw_sc_del_arp_cache_entry, + .query_arp_cache_entry = i40iw_sc_query_arp_cache_entry, + .manage_apbvt_entry = i40iw_sc_manage_apbvt_entry, + .manage_qhash_table_entry = i40iw_sc_manage_qhash_table_entry, + .alloc_local_mac_ipaddr_table_entry = i40iw_sc_alloc_local_mac_ipaddr_entry, + .add_local_mac_ipaddr_entry = i40iw_sc_add_local_mac_ipaddr_entry, + .del_local_mac_ipaddr_entry = i40iw_sc_del_local_mac_ipaddr_entry, + .cqp_nop = i40iw_sc_cqp_nop, + .commit_fpm_values_done = i40iw_sc_commit_fpm_values_done, + .query_fpm_values_done = i40iw_sc_query_fpm_values_done, + .manage_hmc_pm_func_table_done = i40iw_sc_manage_hmc_pm_func_table_done, + .update_suspend_qp = i40iw_sc_suspend_qp, + .update_resume_qp = i40iw_sc_resume_qp }; static struct i40iw_hmc_ops iw_hmc_ops = { - i40iw_sc_init_iw_hmc, - i40iw_sc_parse_fpm_query_buf, - i40iw_sc_configure_iw_fpm, - i40iw_sc_parse_fpm_commit_buf, - i40iw_sc_create_hmc_obj, - i40iw_sc_del_hmc_obj, - NULL, - NULL + .init_iw_hmc = i40iw_sc_init_iw_hmc, + .parse_fpm_query_buf = i40iw_sc_parse_fpm_query_buf, + .configure_iw_fpm = i40iw_sc_configure_iw_fpm, + .parse_fpm_commit_buf = i40iw_sc_parse_fpm_commit_buf, + .create_hmc_object = i40iw_sc_create_hmc_obj, + .del_hmc_object = i40iw_sc_del_hmc_obj }; /** diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c index 2800f796271c..b0d3a0e8a9b5 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_uk.c +++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c @@ -913,29 +913,29 @@ enum i40iw_status_code i40iw_get_wqe_shift(u32 wqdepth, u32 sge, u32 inline_data } static struct i40iw_qp_uk_ops iw_qp_uk_ops = { - i40iw_qp_post_wr, - i40iw_qp_ring_push_db, - i40iw_rdma_write, - i40iw_rdma_read, - i40iw_send, - i40iw_inline_rdma_write, - i40iw_inline_send, - i40iw_stag_local_invalidate, - i40iw_mw_bind, - i40iw_post_receive, - i40iw_nop + .iw_qp_post_wr = i40iw_qp_post_wr, + .iw_qp_ring_push_db = i40iw_qp_ring_push_db, + .iw_rdma_write = i40iw_rdma_write, + .iw_rdma_read = i40iw_rdma_read, + .iw_send = i40iw_send, + .iw_inline_rdma_write = i40iw_inline_rdma_write, + .iw_inline_send = i40iw_inline_send, + .iw_stag_local_invalidate = i40iw_stag_local_invalidate, + .iw_mw_bind = i40iw_mw_bind, + .iw_post_receive = i40iw_post_receive, + .iw_post_nop = i40iw_nop }; static struct i40iw_cq_ops iw_cq_ops = { - i40iw_cq_request_notification, - i40iw_cq_poll_completion, - i40iw_cq_post_entries, - i40iw_clean_cq + .iw_cq_request_notification = i40iw_cq_request_notification, + .iw_cq_poll_completion = i40iw_cq_poll_completion, + .iw_cq_post_entries = i40iw_cq_post_entries, + .iw_cq_clean = i40iw_clean_cq }; static struct i40iw_device_uk_ops iw_device_uk_ops = { - i40iw_cq_uk_init, - i40iw_qp_uk_init, + .iwarp_cq_uk_init = i40iw_cq_uk_init, + .iwarp_qp_uk_init = i40iw_qp_uk_init, }; /** -- cgit v1.2.3 From 850d08721aa10de0cce98fc695bd1ca775dea92e Mon Sep 17 00:00:00 2001 From: Cao jin Date: Mon, 19 Dec 2016 14:46:35 +0800 Subject: RDMA/qib: drop qib_pci_link_reset() In AER recovery, pci_error_handlers.link_reset() is never called, drop it now. Signed-off-by: Cao jin Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qib/qib_pcie.c | 8 -------- 1 file changed, 8 deletions(-) diff --git a/drivers/infiniband/hw/qib/qib_pcie.c b/drivers/infiniband/hw/qib/qib_pcie.c index 6abe1c621aa4..c379b8342a09 100644 --- a/drivers/infiniband/hw/qib/qib_pcie.c +++ b/drivers/infiniband/hw/qib/qib_pcie.c @@ -682,13 +682,6 @@ qib_pci_slot_reset(struct pci_dev *pdev) return PCI_ERS_RESULT_CAN_RECOVER; } -static pci_ers_result_t -qib_pci_link_reset(struct pci_dev *pdev) -{ - qib_devinfo(pdev, "QIB link_reset function called, ignored\n"); - return PCI_ERS_RESULT_CAN_RECOVER; -} - static void qib_pci_resume(struct pci_dev *pdev) { @@ -707,7 +700,6 @@ qib_pci_resume(struct pci_dev *pdev) const struct pci_error_handlers qib_pci_err_handler = { .error_detected = qib_pci_error_detected, .mmio_enabled = qib_pci_mmio_enabled, - .link_reset = qib_pci_link_reset, .slot_reset = qib_pci_slot_reset, .resume = qib_pci_resume, }; -- cgit v1.2.3 From e8f4eb3bfaf4ddb8765cf11c24cadfb1f20e714a Mon Sep 17 00:00:00 2001 From: Cao jin Date: Mon, 19 Dec 2016 14:46:36 +0800 Subject: RDMA/hfi1: drop pci_link_reset() In AER recovery, pci_error_handlers.link_reset() is never called, drop it now. Signed-off-by: Cao jin Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/pcie.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c index 4ac8f330c5cb..ebd941fc8a92 100644 --- a/drivers/infiniband/hw/hfi1/pcie.c +++ b/drivers/infiniband/hw/hfi1/pcie.c @@ -598,15 +598,6 @@ pci_slot_reset(struct pci_dev *pdev) return PCI_ERS_RESULT_CAN_RECOVER; } -static pci_ers_result_t -pci_link_reset(struct pci_dev *pdev) -{ - struct hfi1_devdata *dd = pci_get_drvdata(pdev); - - dd_dev_info(dd, "HFI1 link_reset function called, ignored\n"); - return PCI_ERS_RESULT_CAN_RECOVER; -} - static void pci_resume(struct pci_dev *pdev) { @@ -625,7 +616,6 @@ pci_resume(struct pci_dev *pdev) const struct pci_error_handlers hfi1_pci_err_handler = { .error_detected = pci_error_detected, .mmio_enabled = pci_mmio_enabled, - .link_reset = pci_link_reset, .slot_reset = pci_slot_reset, .resume = pci_resume, }; -- cgit v1.2.3 From bb75f33cf0f8f332dd69e9c1966a8eb0769ee38b Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Tue, 20 Dec 2016 22:02:12 +0800 Subject: RDMA/qib: use rb_entry() To make the code clearer, use rb_entry() instead of container_of() to deal with rbtree. Signed-off-by: Geliang Tang Reviewed-by: Leon Romanovsky Reviewed-by: Dennis Dalessandro Acked-by: Mike Marciniszyn Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qib/qib_user_sdma.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/qib/qib_user_sdma.c b/drivers/infiniband/hw/qib/qib_user_sdma.c index 3e0677c51276..926f3c8eba69 100644 --- a/drivers/infiniband/hw/qib/qib_user_sdma.c +++ b/drivers/infiniband/hw/qib/qib_user_sdma.c @@ -144,8 +144,8 @@ qib_user_sdma_rb_search(struct rb_root *root, pid_t pid) struct rb_node *node = root->rb_node; while (node) { - sdma_rb_node = container_of(node, - struct qib_user_sdma_rb_node, node); + sdma_rb_node = rb_entry(node, struct qib_user_sdma_rb_node, + node); if (pid < sdma_rb_node->pid) node = node->rb_left; else if (pid > sdma_rb_node->pid) @@ -164,7 +164,7 @@ qib_user_sdma_rb_insert(struct rb_root *root, struct qib_user_sdma_rb_node *new) struct qib_user_sdma_rb_node *got; while (*node) { - got = container_of(*node, struct qib_user_sdma_rb_node, node); + got = rb_entry(*node, struct qib_user_sdma_rb_node, node); parent = *node; if (new->pid < got->pid) node = &((*node)->rb_left); -- cgit v1.2.3 From df15856132bc837b512caa36d2227d2350cf64d8 Mon Sep 17 00:00:00 2001 From: "Amrani, Ram" Date: Thu, 22 Dec 2016 14:52:24 +0200 Subject: RDMA/qedr: restructure functions that create/destroy QPs Simplify function and sub-function flow of QP creation and destruction. This also serves as a preparation for SRQ and iWARP support. Signed-off-by: Ram Amrani Reviewed-by: Michal Kalderon Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qedr/verbs.c | 549 ++++++++++++++++--------------------- 1 file changed, 236 insertions(+), 313 deletions(-) diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c index 57c8de208077..bf035a29ab5b 100644 --- a/drivers/infiniband/hw/qedr/verbs.c +++ b/drivers/infiniband/hw/qedr/verbs.c @@ -1076,30 +1076,6 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, return 0; } -static void qedr_cleanup_user_sq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - qedr_free_pbl(dev, &qp->usq.pbl_info, qp->usq.pbl_tbl); - ib_umem_release(qp->usq.umem); -} - -static void qedr_cleanup_user_rq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - qedr_free_pbl(dev, &qp->urq.pbl_info, qp->urq.pbl_tbl); - ib_umem_release(qp->urq.umem); -} - -static void qedr_cleanup_kernel_sq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); - kfree(qp->wqe_wr_id); -} - -static void qedr_cleanup_kernel_rq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); - kfree(qp->rqe_wr_id); -} - static int qedr_check_qp_attrs(struct ib_pd *ibpd, struct qedr_dev *dev, struct ib_qp_init_attr *attrs) { @@ -1188,15 +1164,13 @@ static int qedr_copy_qp_uresp(struct qedr_dev *dev, return rc; } -static void qedr_set_qp_init_params(struct qedr_dev *dev, - struct qedr_qp *qp, - struct qedr_pd *pd, - struct ib_qp_init_attr *attrs) +static void qedr_set_common_qp_params(struct qedr_dev *dev, + struct qedr_qp *qp, + struct qedr_pd *pd, + struct ib_qp_init_attr *attrs) { - qp->pd = pd; - spin_lock_init(&qp->q_lock); - + qp->pd = pd; qp->qp_type = attrs->qp_type; qp->max_inline_data = attrs->cap.max_inline_data; qp->sq.max_sges = attrs->cap.max_send_sge; @@ -1205,7 +1179,11 @@ static void qedr_set_qp_init_params(struct qedr_dev *dev, qp->sq_cq = get_qedr_cq(attrs->send_cq); qp->rq_cq = get_qedr_cq(attrs->recv_cq); qp->dev = dev; + qp->rq.max_sges = attrs->cap.max_recv_sge; + DP_DEBUG(dev, QEDR_MSG_QP, + "RQ params:\trq_max_sges = %d, rq_cq_id = %d\n", + qp->rq.max_sges, qp->rq_cq->icid); DP_DEBUG(dev, QEDR_MSG_QP, "QP params:\tpd = %d, qp_type = %d, max_inline_data = %d, state = %d, signaled = %d, use_srq=%d\n", pd->pd_id, qp->qp_type, qp->max_inline_data, @@ -1213,95 +1191,149 @@ static void qedr_set_qp_init_params(struct qedr_dev *dev, DP_DEBUG(dev, QEDR_MSG_QP, "SQ params:\tsq_max_sges = %d, sq_cq_id = %d\n", qp->sq.max_sges, qp->sq_cq->icid); - qp->rq.max_sges = attrs->cap.max_recv_sge; - DP_DEBUG(dev, QEDR_MSG_QP, - "RQ params:\trq_max_sges = %d, rq_cq_id = %d\n", - qp->rq.max_sges, qp->rq_cq->icid); } -static inline void -qedr_init_qp_user_params(struct qed_rdma_create_qp_in_params *params, - struct qedr_create_qp_ureq *ureq) -{ - /* QP handle to be written in CQE */ - params->qp_handle_lo = ureq->qp_handle_lo; - params->qp_handle_hi = ureq->qp_handle_hi; -} - -static inline void -qedr_init_qp_kernel_doorbell_sq(struct qedr_dev *dev, struct qedr_qp *qp) +static void qedr_set_roce_db_info(struct qedr_dev *dev, struct qedr_qp *qp) { qp->sq.db = dev->db_addr + DB_ADDR_SHIFT(DQ_PWM_OFFSET_XCM_RDMA_SQ_PROD); qp->sq.db_data.data.icid = qp->icid + 1; + qp->rq.db = dev->db_addr + + DB_ADDR_SHIFT(DQ_PWM_OFFSET_TCM_ROCE_RQ_PROD); + qp->rq.db_data.data.icid = qp->icid; } static inline void -qedr_init_qp_kernel_doorbell_rq(struct qedr_dev *dev, struct qedr_qp *qp) +qedr_init_common_qp_in_params(struct qedr_dev *dev, + struct qedr_pd *pd, + struct qedr_qp *qp, + struct ib_qp_init_attr *attrs, + bool fmr_and_reserved_lkey, + struct qed_rdma_create_qp_in_params *params) { - qp->rq.db = dev->db_addr + - DB_ADDR_SHIFT(DQ_PWM_OFFSET_TCM_ROCE_RQ_PROD); - qp->rq.db_data.data.icid = qp->icid; + /* QP handle to be written in an async event */ + params->qp_handle_async_lo = lower_32_bits((uintptr_t) qp); + params->qp_handle_async_hi = upper_32_bits((uintptr_t) qp); + + params->signal_all = (attrs->sq_sig_type == IB_SIGNAL_ALL_WR); + params->fmr_and_reserved_lkey = fmr_and_reserved_lkey; + params->pd = pd->pd_id; + params->dpi = pd->uctx ? pd->uctx->dpi : dev->dpi; + params->sq_cq_id = get_qedr_cq(attrs->send_cq)->icid; + params->stats_queue = 0; + params->rq_cq_id = get_qedr_cq(attrs->recv_cq)->icid; + params->srq_id = 0; + params->use_srq = false; } -static inline int -qedr_init_qp_kernel_params_rq(struct qedr_dev *dev, - struct qedr_qp *qp, struct ib_qp_init_attr *attrs) +static inline void qedr_qp_user_print(struct qedr_dev *dev, struct qedr_qp *qp) { - /* Allocate driver internal RQ array */ - qp->rqe_wr_id = kcalloc(qp->rq.max_wr, sizeof(*qp->rqe_wr_id), - GFP_KERNEL); - if (!qp->rqe_wr_id) - return -ENOMEM; + DP_DEBUG(dev, QEDR_MSG_QP, "create qp: successfully created user QP. " + "qp=%p. " + "sq_addr=0x%llx, " + "sq_len=%zd, " + "rq_addr=0x%llx, " + "rq_len=%zd" + "\n", + qp, + qp->usq.buf_addr, + qp->usq.buf_len, qp->urq.buf_addr, qp->urq.buf_len); +} - DP_DEBUG(dev, QEDR_MSG_QP, "RQ max_wr set to %d.\n", qp->rq.max_wr); +static void qedr_cleanup_user(struct qedr_dev *dev, struct qedr_qp *qp) +{ + if (qp->usq.umem) + ib_umem_release(qp->usq.umem); + qp->usq.umem = NULL; - return 0; + if (qp->urq.umem) + ib_umem_release(qp->urq.umem); + qp->urq.umem = NULL; } -static inline int -qedr_init_qp_kernel_params_sq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct qed_rdma_create_qp_in_params *params) +static int qedr_create_user_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct ib_pd *ibpd, + struct ib_udata *udata, + struct ib_qp_init_attr *attrs) { - u32 temp_max_wr; + struct qed_rdma_create_qp_in_params in_params; + struct qed_rdma_create_qp_out_params out_params; + struct qedr_pd *pd = get_qedr_pd(ibpd); + struct ib_ucontext *ib_ctx = NULL; + struct qedr_ucontext *ctx = NULL; + struct qedr_create_qp_ureq ureq; + int rc = -EINVAL; - /* Allocate driver internal SQ array */ - temp_max_wr = attrs->cap.max_send_wr * dev->wq_multiplier; - temp_max_wr = min_t(u32, temp_max_wr, dev->attr.max_sqe); + ib_ctx = ibpd->uobject->context; + ctx = get_qedr_ucontext(ib_ctx); - /* temp_max_wr < attr->max_sqe < u16 so the casting is safe */ - qp->sq.max_wr = (u16)temp_max_wr; - qp->wqe_wr_id = kcalloc(qp->sq.max_wr, sizeof(*qp->wqe_wr_id), - GFP_KERNEL); - if (!qp->wqe_wr_id) - return -ENOMEM; + memset(&ureq, 0, sizeof(ureq)); + rc = ib_copy_from_udata(&ureq, udata, sizeof(ureq)); + if (rc) { + DP_ERR(dev, "Problem copying data from user space\n"); + return rc; + } - DP_DEBUG(dev, QEDR_MSG_QP, "SQ max_wr set to %d.\n", qp->sq.max_wr); + /* SQ - read access only (0), dma sync not required (0) */ + rc = qedr_init_user_queue(ib_ctx, dev, &qp->usq, ureq.sq_addr, + ureq.sq_len, 0, 0); + if (rc) + return rc; - /* QP handle to be written in CQE */ - params->qp_handle_lo = lower_32_bits((uintptr_t)qp); - params->qp_handle_hi = upper_32_bits((uintptr_t)qp); + /* RQ - read access only (0), dma sync not required (0) */ + rc = qedr_init_user_queue(ib_ctx, dev, &qp->urq, ureq.rq_addr, + ureq.rq_len, 0, 0); + + if (rc) + return rc; + + memset(&in_params, 0, sizeof(in_params)); + qedr_init_common_qp_in_params(dev, pd, qp, attrs, false, &in_params); + in_params.qp_handle_lo = ureq.qp_handle_lo; + in_params.qp_handle_hi = ureq.qp_handle_hi; + in_params.sq_num_pages = qp->usq.pbl_info.num_pbes; + in_params.sq_pbl_ptr = qp->usq.pbl_tbl->pa; + in_params.rq_num_pages = qp->urq.pbl_info.num_pbes; + in_params.rq_pbl_ptr = qp->urq.pbl_tbl->pa; + + qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, + &in_params, &out_params); + + if (!qp->qed_qp) { + rc = -ENOMEM; + goto err1; + } + + qp->qp_id = out_params.qp_id; + qp->icid = out_params.icid; + + rc = qedr_copy_qp_uresp(dev, qp, udata); + if (rc) + goto err; + + qedr_qp_user_print(dev, qp); return 0; +err: + rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); + if (rc) + DP_ERR(dev, "create qp: fatal fault. rc=%d", rc); + +err1: + qedr_cleanup_user(dev, qp); + return rc; } -static inline int qedr_init_qp_kernel_sq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs) +static int +qedr_roce_create_kernel_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct qed_rdma_create_qp_in_params *in_params, + u32 n_sq_elems, u32 n_rq_elems) { - u32 n_sq_elems, n_sq_entries; + struct qed_rdma_create_qp_out_params out_params; int rc; - /* A single work request may take up to QEDR_MAX_SQ_WQE_SIZE elements in - * the ring. The ring should allow at least a single WR, even if the - * user requested none, due to allocation issues. - */ - n_sq_entries = attrs->cap.max_send_wr; - n_sq_entries = min_t(u32, n_sq_entries, dev->attr.max_sqe); - n_sq_entries = max_t(u32, n_sq_entries, 1); - n_sq_elems = n_sq_entries * QEDR_MAX_SQE_ELEMENTS_PER_SQE; rc = dev->ops->common->chain_alloc(dev->cdev, QED_CHAIN_USE_TO_PRODUCE, QED_CHAIN_MODE_PBL, @@ -1309,31 +1341,13 @@ static inline int qedr_init_qp_kernel_sq(struct qedr_dev *dev, n_sq_elems, QEDR_SQE_ELEMENT_SIZE, &qp->sq.pbl); - if (rc) { - DP_ERR(dev, "failed to allocate QP %p SQ\n", qp); - return rc; - } - DP_DEBUG(dev, QEDR_MSG_SQ, - "SQ Pbl base addr = %llx max_send_wr=%d max_wr=%d capacity=%d, rc=%d\n", - qed_chain_get_pbl_phys(&qp->sq.pbl), attrs->cap.max_send_wr, - n_sq_entries, qed_chain_get_capacity(&qp->sq.pbl), rc); - return 0; -} + if (rc) + return rc; -static inline int qedr_init_qp_kernel_rq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs) -{ - u32 n_rq_elems, n_rq_entries; - int rc; + in_params->sq_num_pages = qed_chain_get_page_cnt(&qp->sq.pbl); + in_params->sq_pbl_ptr = qed_chain_get_pbl_phys(&qp->sq.pbl); - /* A single work request may take up to QEDR_MAX_RQ_WQE_SIZE elements in - * the ring. There ring should allow at least a single WR, even if the - * user requested none, due to allocation issues. - */ - n_rq_entries = max_t(u32, attrs->cap.max_recv_wr, 1); - n_rq_elems = n_rq_entries * QEDR_MAX_RQE_ELEMENTS_PER_RQE; rc = dev->ops->common->chain_alloc(dev->cdev, QED_CHAIN_USE_TO_CONSUME_PRODUCE, QED_CHAIN_MODE_PBL, @@ -1341,136 +1355,102 @@ static inline int qedr_init_qp_kernel_rq(struct qedr_dev *dev, n_rq_elems, QEDR_RQE_ELEMENT_SIZE, &qp->rq.pbl); + if (rc) + return rc; - if (rc) { - DP_ERR(dev, "failed to allocate memory for QP %p RQ\n", qp); - return -ENOMEM; - } - - DP_DEBUG(dev, QEDR_MSG_RQ, - "RQ Pbl base addr = %llx max_recv_wr=%d max_wr=%d capacity=%d, rc=%d\n", - qed_chain_get_pbl_phys(&qp->rq.pbl), attrs->cap.max_recv_wr, - n_rq_entries, qed_chain_get_capacity(&qp->rq.pbl), rc); + in_params->rq_num_pages = qed_chain_get_page_cnt(&qp->rq.pbl); + in_params->rq_pbl_ptr = qed_chain_get_pbl_phys(&qp->rq.pbl); - /* n_rq_entries < u16 so the casting is safe */ - qp->rq.max_wr = (u16)n_rq_entries; + qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, + in_params, &out_params); - return 0; -} + if (!qp->qed_qp) + return -EINVAL; -static inline void -qedr_init_qp_in_params_sq(struct qedr_dev *dev, - struct qedr_pd *pd, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct ib_udata *udata, - struct qed_rdma_create_qp_in_params *params) -{ - /* QP handle to be written in an async event */ - params->qp_handle_async_lo = lower_32_bits((uintptr_t)qp); - params->qp_handle_async_hi = upper_32_bits((uintptr_t)qp); + qp->qp_id = out_params.qp_id; + qp->icid = out_params.icid; - params->signal_all = (attrs->sq_sig_type == IB_SIGNAL_ALL_WR); - params->fmr_and_reserved_lkey = !udata; - params->pd = pd->pd_id; - params->dpi = pd->uctx ? pd->uctx->dpi : dev->dpi; - params->sq_cq_id = get_qedr_cq(attrs->send_cq)->icid; - params->max_sq_sges = 0; - params->stats_queue = 0; + qedr_set_roce_db_info(dev, qp); - if (udata) { - params->sq_num_pages = qp->usq.pbl_info.num_pbes; - params->sq_pbl_ptr = qp->usq.pbl_tbl->pa; - } else { - params->sq_num_pages = qed_chain_get_page_cnt(&qp->sq.pbl); - params->sq_pbl_ptr = qed_chain_get_pbl_phys(&qp->sq.pbl); - } + return 0; } -static inline void -qedr_init_qp_in_params_rq(struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct ib_udata *udata, - struct qed_rdma_create_qp_in_params *params) +static void qedr_cleanup_kernel(struct qedr_dev *dev, struct qedr_qp *qp) { - params->rq_cq_id = get_qedr_cq(attrs->recv_cq)->icid; - params->srq_id = 0; - params->use_srq = false; + dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); + kfree(qp->wqe_wr_id); - if (udata) { - params->rq_num_pages = qp->urq.pbl_info.num_pbes; - params->rq_pbl_ptr = qp->urq.pbl_tbl->pa; - } else { - params->rq_num_pages = qed_chain_get_page_cnt(&qp->rq.pbl); - params->rq_pbl_ptr = qed_chain_get_pbl_phys(&qp->rq.pbl); - } + dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); + kfree(qp->rqe_wr_id); } -static inline void qedr_qp_user_print(struct qedr_dev *dev, struct qedr_qp *qp) +static int qedr_create_kernel_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct ib_pd *ibpd, + struct ib_qp_init_attr *attrs) { - DP_DEBUG(dev, QEDR_MSG_QP, - "create qp: successfully created user QP. qp=%p, sq_addr=0x%llx, sq_len=%zd, rq_addr=0x%llx, rq_len=%zd\n", - qp, qp->usq.buf_addr, qp->usq.buf_len, qp->urq.buf_addr, - qp->urq.buf_len); -} + struct qed_rdma_create_qp_in_params in_params; + struct qedr_pd *pd = get_qedr_pd(ibpd); + int rc = -EINVAL; + u32 n_rq_elems; + u32 n_sq_elems; + u32 n_sq_entries; -static inline int qedr_init_user_qp(struct ib_ucontext *ib_ctx, - struct qedr_dev *dev, - struct qedr_qp *qp, - struct qedr_create_qp_ureq *ureq) -{ - int rc; + memset(&in_params, 0, sizeof(in_params)); - /* SQ - read access only (0), dma sync not required (0) */ - rc = qedr_init_user_queue(ib_ctx, dev, &qp->usq, ureq->sq_addr, - ureq->sq_len, 0, 0); - if (rc) - return rc; + /* A single work request may take up to QEDR_MAX_SQ_WQE_SIZE elements in + * the ring. The ring should allow at least a single WR, even if the + * user requested none, due to allocation issues. + * We should add an extra WR since the prod and cons indices of + * wqe_wr_id are managed in such a way that the WQ is considered full + * when (prod+1)%max_wr==cons. We currently don't do that because we + * double the number of entries due an iSER issue that pushes far more + * WRs than indicated. If we decline its ib_post_send() then we get + * error prints in the dmesg we'd like to avoid. + */ + qp->sq.max_wr = min_t(u32, attrs->cap.max_send_wr * dev->wq_multiplier, + dev->attr.max_sqe); - /* RQ - read access only (0), dma sync not required (0) */ - rc = qedr_init_user_queue(ib_ctx, dev, &qp->urq, ureq->rq_addr, - ureq->rq_len, 0, 0); + qp->wqe_wr_id = kzalloc(qp->sq.max_wr * sizeof(*qp->wqe_wr_id), + GFP_KERNEL); + if (!qp->wqe_wr_id) { + DP_ERR(dev, "create qp: failed SQ shadow memory allocation\n"); + return -ENOMEM; + } - if (rc) - qedr_cleanup_user_sq(dev, qp); - return rc; -} + /* QP handle to be written in CQE */ + in_params.qp_handle_lo = lower_32_bits((uintptr_t) qp); + in_params.qp_handle_hi = upper_32_bits((uintptr_t) qp); -static inline int -qedr_init_kernel_qp(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct qed_rdma_create_qp_in_params *params) -{ - int rc; + /* A single work request may take up to QEDR_MAX_RQ_WQE_SIZE elements in + * the ring. There ring should allow at least a single WR, even if the + * user requested none, due to allocation issues. + */ + qp->rq.max_wr = (u16) max_t(u32, attrs->cap.max_recv_wr, 1); - rc = qedr_init_qp_kernel_sq(dev, qp, attrs); - if (rc) { - DP_ERR(dev, "failed to init kernel QP %p SQ\n", qp); - return rc; + /* Allocate driver internal RQ array */ + qp->rqe_wr_id = kzalloc(qp->rq.max_wr * sizeof(*qp->rqe_wr_id), + GFP_KERNEL); + if (!qp->rqe_wr_id) { + DP_ERR(dev, + "create qp: failed RQ shadow memory allocation\n"); + kfree(qp->wqe_wr_id); + return -ENOMEM; } - rc = qedr_init_qp_kernel_params_sq(dev, qp, attrs, params); - if (rc) { - dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); - DP_ERR(dev, "failed to init kernel QP %p SQ params\n", qp); - return rc; - } + qedr_init_common_qp_in_params(dev, pd, qp, attrs, true, &in_params); - rc = qedr_init_qp_kernel_rq(dev, qp, attrs); - if (rc) { - qedr_cleanup_kernel_sq(dev, qp); - DP_ERR(dev, "failed to init kernel QP %p RQ\n", qp); - return rc; - } + n_sq_entries = attrs->cap.max_send_wr; + n_sq_entries = min_t(u32, n_sq_entries, dev->attr.max_sqe); + n_sq_entries = max_t(u32, n_sq_entries, 1); + n_sq_elems = n_sq_entries * QEDR_MAX_SQE_ELEMENTS_PER_SQE; - rc = qedr_init_qp_kernel_params_rq(dev, qp, attrs); - if (rc) { - DP_ERR(dev, "failed to init kernel QP %p RQ params\n", qp); - qedr_cleanup_kernel_sq(dev, qp); - dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); - return rc; - } + n_rq_elems = qp->rq.max_wr * QEDR_MAX_RQE_ELEMENTS_PER_RQE; + + rc = qedr_roce_create_kernel_qp(dev, qp, &in_params, + n_sq_elems, n_rq_elems); + if (rc) + qedr_cleanup_kernel(dev, qp); return rc; } @@ -1480,12 +1460,7 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd, struct ib_udata *udata) { struct qedr_dev *dev = get_qedr_dev(ibpd->device); - struct qed_rdma_create_qp_out_params out_params; - struct qed_rdma_create_qp_in_params in_params; struct qedr_pd *pd = get_qedr_pd(ibpd); - struct ib_ucontext *ib_ctx = NULL; - struct qedr_ucontext *ctx = NULL; - struct qedr_create_qp_ureq ureq; struct qedr_qp *qp; struct ib_qp *ibqp; int rc = 0; @@ -1500,101 +1475,42 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd, if (attrs->srq) return ERR_PTR(-EINVAL); - qp = kzalloc(sizeof(*qp), GFP_KERNEL); - if (!qp) - return ERR_PTR(-ENOMEM); - DP_DEBUG(dev, QEDR_MSG_QP, - "create qp: sq_cq=%p, sq_icid=%d, rq_cq=%p, rq_icid=%d\n", + "create qp: called from %s, event_handler=%p, eepd=%p sq_cq=%p, sq_icid=%d, rq_cq=%p, rq_icid=%d\n", + udata ? "user library" : "kernel", attrs->event_handler, pd, get_qedr_cq(attrs->send_cq), get_qedr_cq(attrs->send_cq)->icid, get_qedr_cq(attrs->recv_cq), get_qedr_cq(attrs->recv_cq)->icid); - qedr_set_qp_init_params(dev, qp, pd, attrs); + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) { + DP_ERR(dev, "create qp: failed allocating memory\n"); + return ERR_PTR(-ENOMEM); + } + + qedr_set_common_qp_params(dev, qp, pd, attrs); if (attrs->qp_type == IB_QPT_GSI) { - if (udata) { - DP_ERR(dev, - "create qp: unexpected udata when creating GSI QP\n"); - goto err0; - } ibqp = qedr_create_gsi_qp(dev, attrs, qp); if (IS_ERR(ibqp)) kfree(qp); return ibqp; } - memset(&in_params, 0, sizeof(in_params)); - - if (udata) { - if (!(udata && ibpd->uobject && ibpd->uobject->context)) - goto err0; - - ib_ctx = ibpd->uobject->context; - ctx = get_qedr_ucontext(ib_ctx); - - memset(&ureq, 0, sizeof(ureq)); - if (ib_copy_from_udata(&ureq, udata, sizeof(ureq))) { - DP_ERR(dev, - "create qp: problem copying data from user space\n"); - goto err0; - } - - rc = qedr_init_user_qp(ib_ctx, dev, qp, &ureq); - if (rc) - goto err0; - - qedr_init_qp_user_params(&in_params, &ureq); - } else { - rc = qedr_init_kernel_qp(dev, qp, attrs, &in_params); - if (rc) - goto err0; - } - - qedr_init_qp_in_params_sq(dev, pd, qp, attrs, udata, &in_params); - qedr_init_qp_in_params_rq(qp, attrs, udata, &in_params); - - qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, - &in_params, &out_params); + if (udata) + rc = qedr_create_user_qp(dev, qp, ibpd, udata, attrs); + else + rc = qedr_create_kernel_qp(dev, qp, ibpd, attrs); - if (!qp->qed_qp) - goto err1; + if (rc) + goto err; - qp->qp_id = out_params.qp_id; - qp->icid = out_params.icid; qp->ibqp.qp_num = qp->qp_id; - if (udata) { - rc = qedr_copy_qp_uresp(dev, qp, udata); - if (rc) - goto err2; - - qedr_qp_user_print(dev, qp); - } else { - qedr_init_qp_kernel_doorbell_sq(dev, qp); - qedr_init_qp_kernel_doorbell_rq(dev, qp); - } - - DP_DEBUG(dev, QEDR_MSG_QP, "created %s space QP %p\n", - udata ? "user" : "kernel", qp); - return &qp->ibqp; -err2: - rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); - if (rc) - DP_ERR(dev, "create qp: fatal fault. rc=%d", rc); -err1: - if (udata) { - qedr_cleanup_user_sq(dev, qp); - qedr_cleanup_user_rq(dev, qp); - } else { - qedr_cleanup_kernel_sq(dev, qp); - qedr_cleanup_kernel_rq(dev, qp); - } - -err0: +err: kfree(qp); return ERR_PTR(-EFAULT); @@ -2067,6 +1983,24 @@ err: return rc; } +int qedr_free_qp_resources(struct qedr_dev *dev, struct qedr_qp *qp) +{ + int rc = 0; + + if (qp->qp_type != IB_QPT_GSI) { + rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); + if (rc) + return rc; + } + + if (qp->ibqp.uobject && qp->ibqp.uobject->context) + qedr_cleanup_user(dev, qp); + else + qedr_cleanup_kernel(dev, qp); + + return 0; +} + int qedr_destroy_qp(struct ib_qp *ibqp) { struct qedr_qp *qp = get_qedr_qp(ibqp); @@ -2089,21 +2023,10 @@ int qedr_destroy_qp(struct ib_qp *ibqp) qedr_modify_qp(ibqp, &attr, attr_mask, NULL); } - if (qp->qp_type != IB_QPT_GSI) { - rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); - if (rc) - return rc; - } else { + if (qp->qp_type == IB_QPT_GSI) qedr_destroy_gsi_qp(dev); - } - if (ibqp->uobject && ibqp->uobject->context) { - qedr_cleanup_user_sq(dev, qp); - qedr_cleanup_user_rq(dev, qp); - } else { - qedr_cleanup_kernel_sq(dev, qp); - qedr_cleanup_kernel_rq(dev, qp); - } + qedr_free_qp_resources(dev, qp); kfree(qp); -- cgit v1.2.3 From 55efcfcd7776165b294f8b5cd6e05ca00ec89b7c Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Thu, 22 Dec 2016 18:07:52 -0700 Subject: RDMA/core: Fix incorrect structure packing for booleans The RDMA core uses ib_pack() to convert from unpacked CPU structs to on-the-wire bitpacked structs. This process requires that 1 bit fields are declared as u8 in the unpacked struct, otherwise the packing process does not read the value properly and the packed result is wired to 0. Several places wrongly used int. Crucially this means the kernel has never, set reversible correctly in the path record request. It has always asked for irreversible paths even if the ULP requests otherwise. When the kernel is used with a SM that supports this feature, it completely breaks communication management if reversible paths are not properly requested. The only reason this ever worked is because opensm ignores the reversible bit. Cc: stable@vger.kernel.org Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Jason Gunthorpe Signed-off-by: Doug Ledford --- include/rdma/ib_sa.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h index 5ee7aab95eb8..fd0e53219f93 100644 --- a/include/rdma/ib_sa.h +++ b/include/rdma/ib_sa.h @@ -153,12 +153,12 @@ struct ib_sa_path_rec { union ib_gid sgid; __be16 dlid; __be16 slid; - int raw_traffic; + u8 raw_traffic; /* reserved */ __be32 flow_label; u8 hop_limit; u8 traffic_class; - int reversible; + u8 reversible; u8 numb_path; __be16 pkey; __be16 qos_class; @@ -220,7 +220,7 @@ struct ib_sa_mcmember_rec { u8 hop_limit; u8 scope; u8 join_state; - int proxy_join; + u8 proxy_join; }; /* Service Record Component Mask Sec 15.2.5.14 Ver 1.1 */ -- cgit v1.2.3 From b462b06eb6b4b8006fc384ac9e8fe6b3d35b0c05 Mon Sep 17 00:00:00 2001 From: ssh10 Date: Sat, 24 Dec 2016 07:14:35 +0530 Subject: RDMA/cxgb4: Use AF_INET for sin_family field Elsewhere the sin_family field holds a value with a name of the form AF_..., so it seems reasonable to do so here as well. Also the values of PF_INET and AF_INET are the same. The semantic patch that makes this change is as follows: // @@ struct sockaddr_in sip; @@ ( sip.sin_family == - PF_INET + AF_INET | sip.sin_family != - PF_INET + AF_INET | sip.sin_family = - PF_INET + AF_INET ) // Signed-off-by: Shyam Saini Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index f1510cc76d2d..9749c52f8729 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -2516,18 +2516,18 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) struct sockaddr_in *sin = (struct sockaddr_in *) &child_ep->com.local_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = local_port; sin->sin_addr.s_addr = *(__be32 *)local_ip; sin = (struct sockaddr_in *)&child_ep->com.local_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = ((struct sockaddr_in *) &parent_ep->com.local_addr)->sin_port; sin->sin_addr.s_addr = *(__be32 *)local_ip; sin = (struct sockaddr_in *)&child_ep->com.remote_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = peer_port; sin->sin_addr.s_addr = *(__be32 *)peer_ip; } else { -- cgit v1.2.3 From db287ec5cb58e765e19c9de0e26411fc64c65de3 Mon Sep 17 00:00:00 2001 From: ssh10 Date: Sat, 24 Dec 2016 21:50:06 +0530 Subject: RDMA/ocrdma: Replace BUG() with BUG_ON() Replace BUG() with BUG_ON() using coccinelle Signed-off-by: Shyam Saini Signed-off-by: Doug Ledford --- drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 15 +++++---------- 1 file changed, 5 insertions(+), 10 deletions(-) diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 6af44f8db3d5..e06ad7250963 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -1170,8 +1170,7 @@ int ocrdma_destroy_cq(struct ib_cq *ibcq) dev->cq_tbl[cq->id] = NULL; indx = ocrdma_get_eq_table_index(dev, cq->eqn); - if (indx == -EINVAL) - BUG(); + BUG_ON(indx == -EINVAL); eq = &dev->eq_tbl[indx]; irq = ocrdma_get_irq(dev, eq); @@ -1741,8 +1740,7 @@ static void ocrdma_discard_cqes(struct ocrdma_qp *qp, struct ocrdma_cq *cq) wqe_idx = (le32_to_cpu(cqe->rq.buftag_qpn) >> OCRDMA_CQE_BUFTAG_SHIFT) & qp->srq->rq.max_wqe_idx; - if (wqe_idx < 1) - BUG(); + BUG_ON(wqe_idx < 1); spin_lock_irqsave(&qp->srq->q_lock, flags); ocrdma_hwq_inc_tail(&qp->srq->rq); ocrdma_srq_toggle_bit(qp->srq, wqe_idx - 1); @@ -2388,15 +2386,13 @@ static int ocrdma_srq_get_idx(struct ocrdma_srq *srq) if (srq->idx_bit_fields[row]) { indx = ffs(srq->idx_bit_fields[row]); indx = (row * 32) + (indx - 1); - if (indx >= srq->rq.max_cnt) - BUG(); + BUG_ON(indx >= srq->rq.max_cnt); ocrdma_srq_toggle_bit(srq, indx); break; } } - if (row == srq->bit_fields_len) - BUG(); + BUG_ON(row == srq->bit_fields_len); return indx + 1; /* Use from index 1 */ } @@ -2754,8 +2750,7 @@ static void ocrdma_update_free_srq_cqe(struct ib_wc *ibwc, srq = get_ocrdma_srq(qp->ibqp.srq); wqe_idx = (le32_to_cpu(cqe->rq.buftag_qpn) >> OCRDMA_CQE_BUFTAG_SHIFT) & srq->rq.max_wqe_idx; - if (wqe_idx < 1) - BUG(); + BUG_ON(wqe_idx < 1); ibwc->wr_id = srq->rqe_wr_id_tbl[wqe_idx]; spin_lock_irqsave(&srq->q_lock, flags); -- cgit v1.2.3 From 29da686dff757968fbf5a9bb5246070ddf602664 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:20 +0200 Subject: IB/ipoib: When given an invalid UD MTU, give debug msg In datagram mode, the IB UD (Unreliable Datagram) transport is used so the MTU of the interface is equal to the IB L2 MTU minus the IPoIB encapsulation header. Any request to change the MTU value above the maximum range will change the MTU to the max allowed, but will not show any warning message. An ipoib_warn is issued in such cases, letting the user know that even though the value is legal, it can't be currently applied. Signed-off-by: Feras Daoud Signed-off-by: Noa Osherovich Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 3ce0765a05ab..b2a75d85bc23 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -229,6 +229,10 @@ static int ipoib_change_mtu(struct net_device *dev, int new_mtu) priv->admin_mtu = new_mtu; + if (priv->mcast_mtu < priv->admin_mtu) + ipoib_dbg(priv, "MTU must be smaller than the underlying " + "link layer MTU - 4 (%u)\n", priv->mcast_mtu); + dev->mtu = min(priv->mcast_mtu, priv->admin_mtu); return 0; -- cgit v1.2.3 From 80b5b35aba62232521b31440f0a3cf6caa033849 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:21 +0200 Subject: IB/ipoib: Set device connection mode only when needed When changing the connection mode, the ipoib_set_mode function did not check if the previous connection mode equals to the new one. This commit adds the required check and return 0 if the new mode equals to the previous one. Fixes: 839fcaba355a ("IPoIB: Connected mode experimental support") Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Alex Vesker Reviewed-by: Yuval Shaia Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index b2a75d85bc23..cbd06a882a60 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -474,6 +474,13 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) { struct ipoib_dev_priv *priv = netdev_priv(dev); + if ((test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags) && + !strcmp(buf, "connected\n")) || + (!test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags) && + !strcmp(buf, "datagram\n"))) { + return 0; + } + /* flush paths if we switch modes so that connections are restarted */ if (IPOIB_CM_SUPPORTED(dev->dev_addr) && !strcmp(buf, "connected\n")) { set_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags); -- cgit v1.2.3 From 1c3098cdb05207e740715857df7b0998e372f527 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:22 +0200 Subject: IB/ipoib: Fix deadlock over vlan_mutex This patch fixes Deadlock while executing ipoib_vlan_delete. The function takes the vlan_rwsem semaphore and calls unregister_netdevice. The later function calls ipoib_mcast_stop_thread that cause workqueue flush. When the queue has one of the ipoib_ib_dev_flush_xxx events, a deadlock occur because these events also tries to catch the same vlan_rwsem semaphore. To fix, unregister_netdevice should be called after releasing the semaphore. Fixes: cbbe1efa4972 ("IPoIB: Fix deadlock between ipoib_open() and child interface create") Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Alex Vesker Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_vlan.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index fd811115af49..96821838c352 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -196,7 +196,6 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) list_for_each_entry_safe(priv, tpriv, &ppriv->child_intfs, list) { if (priv->pkey == pkey && priv->child_type == IPOIB_LEGACY_CHILD) { - unregister_netdevice(priv->dev); list_del(&priv->list); dev = priv->dev; break; @@ -204,6 +203,11 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) } up_write(&ppriv->vlan_rwsem); + if (dev) { + ipoib_dbg(ppriv, "delete child vlan %s\n", dev->name); + unregister_netdevice(dev); + } + rtnl_unlock(); if (dev) { -- cgit v1.2.3 From 0a0007f28304cb9fc87809c86abb80ec71317f20 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:23 +0200 Subject: IB/ipoib: Fix deadlock between rmmod and set_mode When calling set_mode from sys/fs, the call flow locks the sys/fs lock first and then tries to lock rtnl_lock (when calling ipoib_set_mod). On the other hand, the rmmod call flow takes the rtnl_lock first (when calling unregister_netdev) and then tries to take the sys/fs lock. Deadlock a->b, b->a. The problem starts when ipoib_set_mod frees it's rtnl_lck and tries to get it after that. set_mod: [] ? check_preempt_curr+0x6d/0x90 [] __mutex_lock_slowpath+0x13e/0x180 [] ? __rtnl_unlock+0x15/0x20 [] mutex_lock+0x2b/0x50 [] rtnl_lock+0x15/0x20 [] ipoib_set_mode+0x97/0x160 [ib_ipoib] [] set_mode+0x3b/0x80 [ib_ipoib] [] dev_attr_store+0x20/0x30 [] sysfs_write_file+0xe5/0x170 [] vfs_write+0xb8/0x1a0 [] sys_write+0x51/0x90 [] system_call_fastpath+0x16/0x1b rmmod: [] ? put_dec+0x10c/0x110 [] ? number+0x2ee/0x320 [] schedule_timeout+0x215/0x2e0 [] ? vsnprintf+0x484/0x5f0 [] ? string+0x40/0x100 [] wait_for_common+0x123/0x180 [] ? default_wake_function+0x0/0x20 [] ? ifind_fast+0x5e/0xb0 [] wait_for_completion+0x1d/0x20 [] sysfs_addrm_finish+0x228/0x270 [] sysfs_remove_dir+0xa3/0xf0 [] kobject_del+0x16/0x40 [] device_del+0x184/0x1e0 [] netdev_unregister_kobject+0xab/0xc0 [] rollback_registered+0xae/0x130 [] unregister_netdevice+0x22/0x70 [] unregister_netdev+0x1e/0x30 [] ipoib_remove_one+0xe0/0x120 [ib_ipoib] [] ib_unregister_device+0x4f/0x100 [ib_core] [] mlx4_ib_remove+0x41/0x180 [mlx4_ib] [] mlx4_remove_device+0x71/0x90 [mlx4_core] Fixes: 862096a8bbf8 ("IB/ipoib: Add more rtnl_link_ops callbacks") Cc: # v3.6+ Cc: Or Gerlitz Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 12 +++++++----- drivers/infiniband/ulp/ipoib/ipoib_main.c | 6 ++---- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 096c4f6fbd65..1c7a9a16efc7 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1507,12 +1507,14 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, ret = ipoib_set_mode(dev, buf); - rtnl_unlock(); - - if (!ret) - return count; + /* The assumption is that the function ipoib_set_mode returned + * with the rtnl held by it, if not the value -EBUSY returned, + * then no need to rtnl_unlock + */ + if (ret != -EBUSY) + rtnl_unlock(); - return ret; + return (!ret || ret == -EBUSY) ? count : ret; } static DEVICE_ATTR(mode, S_IWUSR | S_IRUGO, show_mode, set_mode); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index cbd06a882a60..045f844d7f4a 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -492,8 +492,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; ipoib_flush_paths(dev); - rtnl_lock(); - return 0; + return (!rtnl_trylock()) ? -EBUSY : 0; } if (!strcmp(buf, "datagram\n")) { @@ -502,8 +501,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) dev_set_mtu(dev, min(priv->mcast_mtu, dev->mtu)); rtnl_unlock(); ipoib_flush_paths(dev); - rtnl_lock(); - return 0; + return (!rtnl_trylock()) ? -EBUSY : 0; } return -EINVAL; -- cgit v1.2.3 From 89a3987ab7a923c047c6dec008e60ad6f41fac22 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:24 +0200 Subject: IB/ipoib: rtnl_unlock can not come after free_netdev The ipoib_vlan_add function calls rtnl_unlock after free_netdev, rtnl_unlock not only releases the lock, but also calls netdev_run_todo. The latter function browses the net_todo_list array and completes the unregistration of all its net_device instances. If we call free_netdev before rtnl_unlock, then netdev_run_todo call over the freed device causes panic. To fix, move rtnl_unlock call before free_netdev call. Fixes: 9baa0b036410 ("IB/ipoib: Add rtnl_link_ops support") Cc: Or Gerlitz Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Yuval Shaia Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_vlan.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index 96821838c352..d9dab4a109f7 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -168,11 +168,11 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) out: up_write(&ppriv->vlan_rwsem); + rtnl_unlock(); + if (result) free_netdev(priv->dev); - rtnl_unlock(); - return result; } -- cgit v1.2.3 From d32b9a81d7c9bf111536b547a60b50c8dd7fccd1 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:25 +0200 Subject: IB/ipoib: Add detailed error message to dev_queue_xmit call Add a detailed return code to dev_queue_xmit function when calling to requeue packet via __skb_dequeue. Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Yuval Shaia Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 7 ++++--- drivers/infiniband/ulp/ipoib/ipoib_main.c | 8 +++++--- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 6 ++++-- 3 files changed, 13 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 1c7a9a16efc7..a720d2ddddc1 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1015,9 +1015,10 @@ static int ipoib_cm_rep_handler(struct ib_cm_id *cm_id, struct ib_cm_event *even while ((skb = __skb_dequeue(&skqueue))) { skb->dev = p->dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed " - "to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s:dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); } ret = ib_send_cm_rtu(cm_id, NULL, 0); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 045f844d7f4a..1086858f1cbf 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -843,10 +843,12 @@ static void path_rec_completion(int status, ipoib_put_ah(old_ah); while ((skb = __skb_dequeue(&skqueue))) { + int ret; skb->dev = dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed " - "to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s: dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); } } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index fddff403d5d2..7c6c67bbdab3 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -314,9 +314,11 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, netif_tx_unlock_bh(dev); skb->dev = dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s:dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); netif_tx_lock_bh(dev); } netif_tx_unlock_bh(dev); -- cgit v1.2.3 From 13ee429a0271b997f57b22ee67d40fc601f0b220 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:26 +0200 Subject: IB/ipoib: Use debug prints instead of warnings in RNR WC status If a receive request has not been posted to the work queue, the incoming message is rejected and the peer will receive a receiver-not-ready (RNR) error. In IPoIB, IB_WC_RNR_RETRY_EXC_ERR error is part of the life cycle therefore ipoib_cm_handle_tx_wc function will print to debug instead of warnings. Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Signed-off-by: Leon Romanovsky Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index a720d2ddddc1..c433e72994f5 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -820,9 +820,12 @@ void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc) wc->status != IB_WC_WR_FLUSH_ERR) { struct ipoib_neigh *neigh; - ipoib_dbg(priv, "failed cm send event " - "(status=%d, wrid=%d vend_err %x)\n", - wc->status, wr_id, wc->vendor_err); + if (wc->status != IB_WC_RNR_RETRY_EXC_ERR) + ipoib_warn(priv, "failed cm send event (status=%d, wrid=%d vend_err %x)\n", + wc->status, wr_id, wc->vendor_err); + else + ipoib_dbg(priv, "failed cm send event (status=%d, wrid=%d vend_err %x)\n", + wc->status, wr_id, wc->vendor_err); spin_lock_irqsave(&priv->lock, flags); neigh = tx->neigh; -- cgit v1.2.3 From c586071d1dc8227a7182179b8e50ee92cc43f6d2 Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:27 +0200 Subject: IB/ipoib: Replace list_del of the neigh->list with list_del_init In order to resolve a situation where a few process delete the same list element in sequence and cause panic, list_del is replaced with list_del_init. In this case if the first process that calls list_del releases the lock before acquiring it again, other processes who can acquire the lock will call list_del_init. Fixes: b63b70d87741 ("IPoIB: Use a private hash table for path lookup") Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Alex Vesker Signed-off-by: Leon Romanovsky Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 1086858f1cbf..ee32a2a87e2d 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1297,7 +1297,7 @@ static void __ipoib_reap_neigh(struct ipoib_dev_priv *priv) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from path/mc list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } else { np = &neigh->hnext; @@ -1461,7 +1461,7 @@ void ipoib_neigh_free(struct ipoib_neigh *neigh) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from parent list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); return; } else { @@ -1546,7 +1546,7 @@ void ipoib_del_neighs_by_gid(struct net_device *dev, u8 *gid) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from parent list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } else { np = &neigh->hnext; @@ -1588,7 +1588,7 @@ static void ipoib_flush_neighs(struct ipoib_dev_priv *priv) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from path/mc list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } } -- cgit v1.2.3 From 27d41d29c7f093f6f77843624fbb080c1b4a8b9c Mon Sep 17 00:00:00 2001 From: Feras Daoud Date: Wed, 28 Dec 2016 14:47:28 +0200 Subject: IB/ipoib: Change list_del to list_del_init in the tx object Since ipoib_cm_tx_start function and ipoib_cm_tx_reap function belong to different work queues, they can run in parallel. In this case if ipoib_cm_tx_reap calls list_del and release the lock, ipoib_cm_tx_start may acquire it and call list_del_init on the already deleted object. Changing list_del to list_del_init in ipoib_cm_tx_reap fixes the problem. Fixes: 839fcaba355a ("IPoIB: Connected mode experimental support") Signed-off-by: Feras Daoud Signed-off-by: Erez Shitrit Reviewed-by: Alex Vesker Signed-off-by: Leon Romanovsky Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index c433e72994f5..53d69d7d9ad4 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1392,7 +1392,7 @@ static void ipoib_cm_tx_reap(struct work_struct *work) while (!list_empty(&priv->cm.reap_list)) { p = list_entry(priv->cm.reap_list.next, typeof(*p), list); - list_del(&p->list); + list_del_init(&p->list); spin_unlock_irqrestore(&priv->lock, flags); netif_tx_unlock_bh(dev); ipoib_cm_tx_destroy(p); -- cgit v1.2.3 From aaaca121c7cf9217ab81d9db0a04835d52aabebe Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Mon, 2 Jan 2017 13:17:36 +0100 Subject: RDMA/core: add port state cache We need a port state cache in ib_core, later we will use in rdma_cm. Signed-off-by: Jack Wang Reviewed-by: Michael Wang Acked-by: Sean Hefty Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 10 +++++++++- include/rdma/ib_verbs.h | 1 + 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index ae04826e82fc..87fd7c39fbc0 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -1105,6 +1105,8 @@ static void ib_cache_update(struct ib_device *device, } device->cache.lmc_cache[port - rdma_start_port(device)] = tprops->lmc; + device->cache.port_state_cache[port - rdma_start_port(device)] = + tprops->state; write_unlock_irq(&device->cache.lock); @@ -1164,7 +1166,11 @@ int ib_cache_setup_one(struct ib_device *device) (rdma_end_port(device) - rdma_start_port(device) + 1), GFP_KERNEL); - if (!device->cache.pkey_cache || + device->cache.port_state_cache = kmalloc(sizeof *device->cache.port_state_cache * + (rdma_end_port(device) - + rdma_start_port(device) + 1), + GFP_KERNEL); + if (!device->cache.pkey_cache || !device->cache.port_state_cache || !device->cache.lmc_cache) { err = -ENOMEM; goto free; @@ -1190,6 +1196,7 @@ err: free: kfree(device->cache.pkey_cache); kfree(device->cache.lmc_cache); + kfree(device->cache.port_state_cache); return err; } @@ -1211,6 +1218,7 @@ void ib_cache_release_one(struct ib_device *device) gid_table_release_one(device); kfree(device->cache.pkey_cache); kfree(device->cache.lmc_cache); + kfree(device->cache.port_state_cache); } void ib_cache_cleanup_one(struct ib_device *device) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 958a24d8fae7..fafa988e0e9a 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1781,6 +1781,7 @@ struct ib_cache { struct ib_pkey_cache **pkey_cache; struct ib_gid_table **gid_cache; u8 *lmc_cache; + enum ib_port_state *port_state_cache; }; struct ib_dma_mapping_ops { -- cgit v1.2.3 From 9e2c3f1c7f3edc6bdd83dab601418ac6c3b6df73 Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Mon, 2 Jan 2017 13:19:24 +0100 Subject: RDMA/core: export ib_get_cached_port_state Export function for rdma_cm, patch for rdma_cm to follow. Signed-off-by: Jack Wang Reviewed-by: Michael Wang Acked-by: Sean Hefty Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 18 ++++++++++++++++++ include/rdma/ib_cache.h | 13 +++++++++++++ 2 files changed, 31 insertions(+) diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 87fd7c39fbc0..f91886bab9d1 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -1021,6 +1021,24 @@ int ib_get_cached_lmc(struct ib_device *device, } EXPORT_SYMBOL(ib_get_cached_lmc); +int ib_get_cached_port_state(struct ib_device *device, + u8 port_num, + enum ib_port_state *port_state) +{ + unsigned long flags; + int ret = 0; + + if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + return -EINVAL; + + read_lock_irqsave(&device->cache.lock, flags); + *port_state = device->cache.port_state_cache[port_num - rdma_start_port(device)]; + read_unlock_irqrestore(&device->cache.lock, flags); + + return ret; +} +EXPORT_SYMBOL(ib_get_cached_port_state); + static void ib_cache_update(struct ib_device *device, u8 port) { diff --git a/include/rdma/ib_cache.h b/include/rdma/ib_cache.h index e30f19bd4a41..385ec88ee9e5 100644 --- a/include/rdma/ib_cache.h +++ b/include/rdma/ib_cache.h @@ -165,4 +165,17 @@ int ib_get_cached_lmc(struct ib_device *device, u8 port_num, u8 *lmc); +/** + * ib_get_cached_port_state - Returns a cached port state table entry + * @device: The device to query. + * @port_num: The port number of the device to query. + * @port_state: port_state for the specified port for that device. + * + * ib_get_cached_port_state() fetches the specified port_state table entry stored in + * the local software cache. + */ +int ib_get_cached_port_state(struct ib_device *device, + u8 port_num, + enum ib_port_state *port_active); + #endif /* _IB_CACHE_H */ -- cgit v1.2.3 From 93b1f29de71f3bd3b88ec8a2c2874c3847cd38bc Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Mon, 2 Jan 2017 13:20:44 +0100 Subject: RDMA/cma: resolve to first active ib port When we try to resolve a dest addr, if we don't give src addr, cma core will try to resolve to our source ib device automatically. The current logic only checks if a given port has the same subnet_prefix as our dest, which is not enough if we use default well known subnet_prefix on our active port, as it will be the same as the subnet_prefix on inactive ports and we might match against an inactive port by accident. To resolve this, we should also check if port is active before we resolve it as a suitable src address for a given dest. Signed-off-by: Jack Wang Reviewed-by: Michael Wang Acked-by: Sean Hefty Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index e7dcfac877ca..40e2b726aa02 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -709,6 +709,7 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) union ib_gid gid, sgid, *dgid; u16 pkey, index; u8 p; + enum ib_port_state port_state; int i; cma_dev = NULL; @@ -724,6 +725,8 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) if (ib_find_cached_pkey(cur_dev->device, p, pkey, &index)) continue; + if (ib_get_cached_port_state(cur_dev->device, p, &port_state)) + continue; for (i = 0; !ib_get_cached_gid(cur_dev->device, p, i, &gid, NULL); i++) { @@ -735,7 +738,8 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) } if (!cma_dev && (gid.global.subnet_prefix == - dgid->global.subnet_prefix)) { + dgid->global.subnet_prefix) && + port_state == IB_PORT_ACTIVE) { cma_dev = cur_dev; sgid = gid; id_priv->id.port_num = p; -- cgit v1.2.3 From 102c5ce082f557405a0f71ad5a8920f9df50cd9e Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Mon, 2 Jan 2017 13:22:05 +0100 Subject: RDMA/cma: use cached port state when bind loopback Signed-off-by: Jack Wang Reviewed-by: Michael Wang Acked-by: Sean Hefty Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 40e2b726aa02..a14a18cfc8f3 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -2656,8 +2656,8 @@ static void cma_set_loopback(struct sockaddr *addr) static int cma_bind_loopback(struct rdma_id_private *id_priv) { struct cma_device *cma_dev, *cur_dev; - struct ib_port_attr port_attr; union ib_gid gid; + enum ib_port_state port_state; u16 pkey; int ret; u8 p; @@ -2673,8 +2673,8 @@ static int cma_bind_loopback(struct rdma_id_private *id_priv) cma_dev = cur_dev; for (p = 1; p <= cur_dev->device->phys_port_cnt; ++p) { - if (!ib_query_port(cur_dev->device, p, &port_attr) && - port_attr.state == IB_PORT_ACTIVE) { + if (!ib_get_cached_port_state(cur_dev->device, p, &port_state) && + port_state == IB_PORT_ACTIVE) { cma_dev = cur_dev; goto port_found; } -- cgit v1.2.3 From 6c6e51a617df3ef6acdbdb550d61c70899e6033e Mon Sep 17 00:00:00 2001 From: Yuval Shaia Date: Wed, 4 Jan 2017 22:17:14 +0200 Subject: IB/core: Fix typo in comment Signed-off-by: Yuval Shaia Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/cq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/core/cq.c b/drivers/infiniband/core/cq.c index a754fc727de5..838147ef727f 100644 --- a/drivers/infiniband/core/cq.c +++ b/drivers/infiniband/core/cq.c @@ -120,7 +120,7 @@ static void ib_cq_completion_workqueue(struct ib_cq *cq, void *private) * * This is the proper interface to allocate a CQ for in-kernel users. A * CQ allocated with this interface will automatically be polled from the - * specified context. The ULP needs must use wr->wr_cqe instead of wr->wr_id + * specified context. The ULP must use wr->wr_cqe instead of wr->wr_id * to use this CQ abstraction. */ struct ib_cq *ib_alloc_cq(struct ib_device *dev, void *private, -- cgit v1.2.3 From 1dd70ea36077f320dadad120d1a611b9ce706ea5 Mon Sep 17 00:00:00 2001 From: Yuval Shaia Date: Tue, 17 Jan 2017 10:34:00 +0200 Subject: IB/vmw_pvrdma: Remove unused qp_type Remove the unused qp_type parameter from function's args Signed-off-by: Yuval Shaia Reviewed-by: Leon Romanovsky Acked-by: Adit Ranadive Signed-off-by: Doug Ledford --- drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c index c8c01e558125..dbbfd35e7da7 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c @@ -151,7 +151,7 @@ static int pvrdma_set_rq_size(struct pvrdma_dev *dev, } static int pvrdma_set_sq_size(struct pvrdma_dev *dev, struct ib_qp_cap *req_cap, - enum ib_qp_type type, struct pvrdma_qp *qp) + struct pvrdma_qp *qp) { if (req_cap->max_send_wr > dev->dsr->caps.max_qp_wr || req_cap->max_send_sge > dev->dsr->caps.max_sge) { @@ -276,8 +276,7 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, qp->is_kernel = true; ret = pvrdma_set_sq_size(to_vdev(pd->device), - &init_attr->cap, - init_attr->qp_type, qp); + &init_attr->cap, qp); if (ret) goto err_qp; -- cgit v1.2.3 From f57e8ca50e23b9e69b2d9f057ca07f9817daab09 Mon Sep 17 00:00:00 2001 From: Yuval Shaia Date: Thu, 19 Jan 2017 16:48:31 +0200 Subject: IB/mad: Add port_num to error message Print the invalid port number to ease troubleshooting. Signed-off-by: Yuval Shaia Reviewed-by: Leon Romanovsky Reviewed-by: Johannes Thumshirn Reviewed-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/core/mad.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index a009f7132c73..57f231f1c721 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -316,7 +316,9 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, /* Validate device and port */ port_priv = ib_get_mad_port(device, port_num); if (!port_priv) { - dev_notice(&device->dev, "ib_register_mad_agent: Invalid port\n"); + dev_notice(&device->dev, + "ib_register_mad_agent: Invalid port %d\n", + port_num); ret = ERR_PTR(-ENODEV); goto error1; } -- cgit v1.2.3 From bab572f1d42dc79d366095fe939f012c59f603f4 Mon Sep 17 00:00:00 2001 From: Ganesh Goudar Date: Fri, 20 Jan 2017 12:32:01 +0530 Subject: iw_cxgb4: Guard against null cm_id in dump_ep/qp Endpoints that are aborting can have already dereferenced the cm_id and set ep->com.cm_id to NULL. So guard against that in dump_ep() and dump_qp(). Also create a common function for setting up ip address pointers since the same logic is needed in several places. Reviewed-by: Steve Wise Signed-off-by: Ganesh Goudar Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/device.c | 133 ++++++++++++++++++++++------------- 1 file changed, 84 insertions(+), 49 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c index 516b0ae6dc3f..c4b344e1f7e7 100644 --- a/drivers/infiniband/hw/cxgb4/device.c +++ b/drivers/infiniband/hw/cxgb4/device.c @@ -214,6 +214,52 @@ static const struct file_operations wr_log_debugfs_fops = { .write = wr_log_clear, }; +static struct sockaddr_in zero_sin = { + .sin_family = AF_INET, +}; + +static struct sockaddr_in6 zero_sin6 = { + .sin6_family = AF_INET6, +}; + +static void set_ep_sin_addrs(struct c4iw_ep *ep, + struct sockaddr_in **lsin, + struct sockaddr_in **rsin, + struct sockaddr_in **m_lsin, + struct sockaddr_in **m_rsin) +{ + struct iw_cm_id *id = ep->com.cm_id; + + *lsin = (struct sockaddr_in *)&ep->com.local_addr; + *rsin = (struct sockaddr_in *)&ep->com.remote_addr; + if (id) { + *m_lsin = (struct sockaddr_in *)&id->m_local_addr; + *m_rsin = (struct sockaddr_in *)&id->m_remote_addr; + } else { + *m_lsin = &zero_sin; + *m_rsin = &zero_sin; + } +} + +static void set_ep_sin6_addrs(struct c4iw_ep *ep, + struct sockaddr_in6 **lsin6, + struct sockaddr_in6 **rsin6, + struct sockaddr_in6 **m_lsin6, + struct sockaddr_in6 **m_rsin6) +{ + struct iw_cm_id *id = ep->com.cm_id; + + *lsin6 = (struct sockaddr_in6 *)&ep->com.local_addr; + *rsin6 = (struct sockaddr_in6 *)&ep->com.remote_addr; + if (id) { + *m_lsin6 = (struct sockaddr_in6 *)&id->m_local_addr; + *m_rsin6 = (struct sockaddr_in6 *)&id->m_remote_addr; + } else { + *m_lsin6 = &zero_sin6; + *m_rsin6 = &zero_sin6; + } +} + static int dump_qp(int id, void *p, void *data) { struct c4iw_qp *qp = p; @@ -229,16 +275,15 @@ static int dump_qp(int id, void *p, void *data) return 1; if (qp->ep) { - if (qp->ep->com.local_addr.ss_family == AF_INET) { - struct sockaddr_in *lsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->local_addr; - struct sockaddr_in *rsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->remote_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->m_local_addr; - struct sockaddr_in *mapped_rsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->m_remote_addr; + struct c4iw_ep *ep = qp->ep; + + if (ep->com.local_addr.ss_family == AF_INET) { + struct sockaddr_in *lsin; + struct sockaddr_in *rsin; + struct sockaddr_in *m_lsin; + struct sockaddr_in *m_rsin; + set_ep_sin_addrs(ep, &lsin, &rsin, &m_lsin, &m_rsin); cc = snprintf(qpd->buf + qpd->pos, space, "rc qp sq id %u rq id %u state %u " "onchip %u ep tid %u state %u " @@ -246,23 +291,19 @@ static int dump_qp(int id, void *p, void *data) qp->wq.sq.qid, qp->wq.rq.qid, (int)qp->attr.state, qp->wq.sq.flags & T4_SQ_ONCHIP, - qp->ep->hwtid, (int)qp->ep->com.state, + ep->hwtid, (int)ep->com.state, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port), + ntohs(m_lsin->sin_port), &rsin->sin_addr, ntohs(rsin->sin_port), - ntohs(mapped_rsin->sin_port)); + ntohs(m_rsin->sin_port)); } else { - struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) - &qp->ep->com.cm_id->local_addr; - struct sockaddr_in6 *rsin6 = (struct sockaddr_in6 *) - &qp->ep->com.cm_id->remote_addr; - struct sockaddr_in6 *mapped_lsin6 = - (struct sockaddr_in6 *) - &qp->ep->com.cm_id->m_local_addr; - struct sockaddr_in6 *mapped_rsin6 = - (struct sockaddr_in6 *) - &qp->ep->com.cm_id->m_remote_addr; + struct sockaddr_in6 *lsin6; + struct sockaddr_in6 *rsin6; + struct sockaddr_in6 *m_lsin6; + struct sockaddr_in6 *m_rsin6; + set_ep_sin6_addrs(ep, &lsin6, &rsin6, &m_lsin6, + &m_rsin6); cc = snprintf(qpd->buf + qpd->pos, space, "rc qp sq id %u rq id %u state %u " "onchip %u ep tid %u state %u " @@ -270,13 +311,13 @@ static int dump_qp(int id, void *p, void *data) qp->wq.sq.qid, qp->wq.rq.qid, (int)qp->attr.state, qp->wq.sq.flags & T4_SQ_ONCHIP, - qp->ep->hwtid, (int)qp->ep->com.state, + ep->hwtid, (int)ep->com.state, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port), + ntohs(m_lsin6->sin6_port), &rsin6->sin6_addr, ntohs(rsin6->sin6_port), - ntohs(mapped_rsin6->sin6_port)); + ntohs(m_rsin6->sin6_port)); } } else cc = snprintf(qpd->buf + qpd->pos, space, @@ -533,15 +574,12 @@ static int dump_ep(int id, void *p, void *data) return 1; if (ep->com.local_addr.ss_family == AF_INET) { - struct sockaddr_in *lsin = (struct sockaddr_in *) - &ep->com.cm_id->local_addr; - struct sockaddr_in *rsin = (struct sockaddr_in *) - &ep->com.cm_id->remote_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) - &ep->com.cm_id->m_local_addr; - struct sockaddr_in *mapped_rsin = (struct sockaddr_in *) - &ep->com.cm_id->m_remote_addr; + struct sockaddr_in *lsin; + struct sockaddr_in *rsin; + struct sockaddr_in *m_lsin; + struct sockaddr_in *m_rsin; + set_ep_sin_addrs(ep, &lsin, &rsin, &m_lsin, &m_rsin); cc = snprintf(epd->buf + epd->pos, space, "ep %p cm_id %p qp %p state %d flags 0x%lx " "history 0x%lx hwtid %d atid %d " @@ -553,19 +591,16 @@ static int dump_ep(int id, void *p, void *data) ep->stats.connect_neg_adv, ep->stats.abort_neg_adv, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port), + ntohs(m_lsin->sin_port), &rsin->sin_addr, ntohs(rsin->sin_port), - ntohs(mapped_rsin->sin_port)); + ntohs(m_rsin->sin_port)); } else { - struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->local_addr; - struct sockaddr_in6 *rsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->remote_addr; - struct sockaddr_in6 *mapped_lsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->m_local_addr; - struct sockaddr_in6 *mapped_rsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->m_remote_addr; + struct sockaddr_in6 *lsin6; + struct sockaddr_in6 *rsin6; + struct sockaddr_in6 *m_lsin6; + struct sockaddr_in6 *m_rsin6; + set_ep_sin6_addrs(ep, &lsin6, &rsin6, &m_lsin6, &m_rsin6); cc = snprintf(epd->buf + epd->pos, space, "ep %p cm_id %p qp %p state %d flags 0x%lx " "history 0x%lx hwtid %d atid %d " @@ -577,9 +612,9 @@ static int dump_ep(int id, void *p, void *data) ep->stats.connect_neg_adv, ep->stats.abort_neg_adv, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port), + ntohs(m_lsin6->sin6_port), &rsin6->sin6_addr, ntohs(rsin6->sin6_port), - ntohs(mapped_rsin6->sin6_port)); + ntohs(m_rsin6->sin6_port)); } if (cc < space) epd->pos += cc; @@ -600,7 +635,7 @@ static int dump_listen_ep(int id, void *p, void *data) if (ep->com.local_addr.ss_family == AF_INET) { struct sockaddr_in *lsin = (struct sockaddr_in *) &ep->com.cm_id->local_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) + struct sockaddr_in *m_lsin = (struct sockaddr_in *) &ep->com.cm_id->m_local_addr; cc = snprintf(epd->buf + epd->pos, space, @@ -609,11 +644,11 @@ static int dump_listen_ep(int id, void *p, void *data) ep, ep->com.cm_id, (int)ep->com.state, ep->com.flags, ep->stid, ep->backlog, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port)); + ntohs(m_lsin->sin_port)); } else { struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) &ep->com.cm_id->local_addr; - struct sockaddr_in6 *mapped_lsin6 = (struct sockaddr_in6 *) + struct sockaddr_in6 *m_lsin6 = (struct sockaddr_in6 *) &ep->com.cm_id->m_local_addr; cc = snprintf(epd->buf + epd->pos, space, @@ -622,7 +657,7 @@ static int dump_listen_ep(int id, void *p, void *data) ep, ep->com.cm_id, (int)ep->com.state, ep->com.flags, ep->stid, ep->backlog, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port)); + ntohs(m_lsin6->sin6_port)); } if (cc < space) epd->pos += cc; -- cgit v1.2.3 From 506f71d1811aaf2d6a6c0bd132ceba29f5c14f3e Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Fri, 30 Dec 2016 14:38:05 +0000 Subject: IB/isert: fix spelling mistake: "teminating" -> "terminating" Trivial fix to spelling mistake in isert_warn message Signed-off-by: Colin Ian King Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/isert/ib_isert.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index 314e95516068..91cbe86b25c8 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -728,7 +728,7 @@ isert_disconnected_handler(struct rdma_cm_id *cma_id, iscsit_cause_connection_reinstatement(isert_conn->conn, 0); break; default: - isert_warn("conn %p teminating in state %d\n", + isert_warn("conn %p terminating in state %d\n", isert_conn, isert_conn->state); } mutex_unlock(&isert_conn->mutex); -- cgit v1.2.3 From f7534f45dcbc1a2ecb486b019db0443188e957d6 Mon Sep 17 00:00:00 2001 From: Zhu Yanjun Date: Thu, 5 Jan 2017 03:56:08 -0500 Subject: IB/ipoib: Remove unnecessary returned value check In the function ipoib_set_dev_features, the returned value is always 0. As such, it is not necessary to check the returned value. This is not a bug. It is a trivial problem. Reviewed-by: Guanglei Li Signed-off-by: Zhu Yanjun Reviewed-by: Yuval Shaia Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib.h | 2 +- drivers/infiniband/ulp/ipoib/ipoib_main.c | 8 ++------ drivers/infiniband/ulp/ipoib/ipoib_vlan.c | 4 +--- 3 files changed, 4 insertions(+), 10 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index da12717a3eb7..f5680642960a 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -593,7 +593,7 @@ void ipoib_pkey_open(struct ipoib_dev_priv *priv); void ipoib_drain_cq(struct net_device *dev); void ipoib_set_ethtool_ops(struct net_device *dev); -int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca); +void ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca); #define IPOIB_FLAGS_RC 0x80 #define IPOIB_FLAGS_UC 0x40 diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index ee32a2a87e2d..c8af6377a288 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1995,7 +1995,7 @@ int ipoib_add_pkey_attr(struct net_device *dev) return device_create_file(&dev->dev, &dev_attr_pkey); } -int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca) +void ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca) { priv->hca_caps = hca->attrs.device_cap_flags; @@ -2007,8 +2007,6 @@ int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca) priv->dev->features |= priv->dev->hw_features; } - - return 0; } static struct net_device *ipoib_add_port(const char *format, @@ -2048,9 +2046,7 @@ static struct net_device *ipoib_add_port(const char *format, goto device_init_failed; } - result = ipoib_set_dev_features(priv, hca); - if (result) - goto device_init_failed; + ipoib_set_dev_features(priv, hca); /* * Set the full membership bit, so that we join the right diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index d9dab4a109f7..deedb6fc1b05 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -61,9 +61,7 @@ int __ipoib_vlan_add(struct ipoib_dev_priv *ppriv, struct ipoib_dev_priv *priv, priv->parent = ppriv->dev; set_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags); - result = ipoib_set_dev_features(priv, ppriv->ca); - if (result) - goto err; + ipoib_set_dev_features(priv, ppriv->ca); priv->pkey = pkey; -- cgit v1.2.3 From 498683c6a7eec5e6ca4f8371d92061618f0827bc Mon Sep 17 00:00:00 2001 From: Moni Shoua Date: Thu, 5 Jan 2017 16:16:35 +0200 Subject: IB/cma: Add debug messages to error flows Print debug messages to the kernel log to add more information about RDMA_CM events that indicate an error. Signed-off-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/cm.c | 2 ++ drivers/infiniband/core/cma.c | 25 +++++++++++++++++++++++-- 2 files changed, 25 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index cf1edfa1cbac..6535f09dc575 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -3409,6 +3409,8 @@ static void cm_process_send_error(struct ib_mad_send_buf *msg, if (msg != cm_id_priv->msg || state != cm_id_priv->id.state) goto discard; + pr_debug_ratelimited("CM: failed sending MAD in state %d. (%s)\n", + state, ib_wc_status_msg(wc_status)); switch (state) { case IB_CM_REQ_SENT: case IB_CM_MRA_REQ_RCVD: diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index a14a18cfc8f3..37e464468539 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1693,6 +1693,7 @@ static int cma_rep_recv(struct rdma_id_private *id_priv) return 0; reject: + pr_debug_ratelimited("RDMA CM: CONNECT_ERROR: failed to handle reply. status %d\n", ret); cma_modify_qp_err(id_priv); ib_send_cm_rej(id_priv->cm_id.ib, IB_CM_REJ_CONSUMER_DEFINED, NULL, 0, NULL, 0); @@ -1764,6 +1765,8 @@ static int cma_ib_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) /* ignore event */ goto out; case IB_CM_REJ_RECEIVED: + pr_debug_ratelimited("RDMA CM: REJECTED: %s\n", rdma_reject_msg(&id_priv->id, + ib_event->param.rej_rcvd.reason)); cma_modify_qp_err(id_priv); event.status = ib_event->param.rej_rcvd.reason; event.event = RDMA_CM_EVENT_REJECTED; @@ -2289,6 +2292,8 @@ static void cma_query_handler(int status, struct ib_sa_path_rec *path_rec, work->new_state = RDMA_CM_ADDR_RESOLVED; work->event.event = RDMA_CM_EVENT_ROUTE_ERROR; work->event.status = status; + pr_debug_ratelimited("RDMA CM: ROUTE_ERROR: failed to query path. status %d\n", + status); } queue_work(cma_wq, &work->work); @@ -2724,8 +2729,14 @@ static void addr_handler(int status, struct sockaddr *src_addr, goto out; memcpy(cma_src_addr(id_priv), src_addr, rdma_addr_size(src_addr)); - if (!status && !id_priv->cma_dev) + if (!status && !id_priv->cma_dev) { status = cma_acquire_dev(id_priv, NULL); + if (status) + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to acquire device. status %d\n", + status); + } else { + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to resolve IP. status %d\n", status); + } if (status) { if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_RESOLVED, @@ -3311,10 +3322,13 @@ static int cma_sidr_rep_handler(struct ib_cm_id *cm_id, if (rep->status != IB_SIDR_SUCCESS) { event.event = RDMA_CM_EVENT_UNREACHABLE; event.status = ib_event->param.sidr_rep_rcvd.status; + pr_debug_ratelimited("RDMA CM: UNREACHABLE: bad SIDR reply. status %d\n", + event.status); break; } ret = cma_set_qkey(id_priv, rep->qkey); if (ret) { + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to set qkey. status %d\n", ret); event.event = RDMA_CM_EVENT_ADDR_ERROR; event.status = ret; break; @@ -3763,10 +3777,17 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) if (!status) status = cma_set_qkey(id_priv, be32_to_cpu(multicast->rec.qkey)); + else + pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to join multicast. status %d\n", + status); mutex_lock(&id_priv->qp_mutex); - if (!status && id_priv->id.qp) + if (!status && id_priv->id.qp) { status = ib_attach_mcast(id_priv->id.qp, &multicast->rec.mgid, be16_to_cpu(multicast->rec.mlid)); + if (!status) + pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to attach QP. status %d\n", + status); + } mutex_unlock(&id_priv->qp_mutex); memset(&event, 0, sizeof event); -- cgit v1.2.3 From 19b752a19dcec188f404a070f4d4c33607ea493b Mon Sep 17 00:00:00 2001 From: Moni Shoua Date: Thu, 5 Jan 2017 16:16:36 +0200 Subject: IB/cma: Allow port reuse for rdma_id When allocating a port number for binding to a rdma_id, assuming the allocation is not for a specific port, the rule is to allow only ports that were not in use before by any other rdma_id. This condition is too strong to achieve the goal of a unique 5 tuple rdma_id. Instead, we can compare current rdma_id with other rdma_id for difference in one of destination port, source address and destination address to allow port reuse. Signed-off-by: Moni Shoua Signed-off-by: Leon Romanovsky Acked-by: Sean Hefty Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 67 ++++++++++++++++++++++++++++++++++++++----- 1 file changed, 60 insertions(+), 7 deletions(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 37e464468539..f9133c7606a8 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -2847,20 +2847,26 @@ int rdma_resolve_addr(struct rdma_cm_id *id, struct sockaddr *src_addr, int ret; id_priv = container_of(id, struct rdma_id_private, id); + memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr)); if (id_priv->state == RDMA_CM_IDLE) { ret = cma_bind_addr(id, src_addr, dst_addr); - if (ret) + if (ret) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return ret; + } } - if (cma_family(id_priv) != dst_addr->sa_family) + if (cma_family(id_priv) != dst_addr->sa_family) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return -EINVAL; + } - if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY)) + if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY)) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return -EINVAL; + } atomic_inc(&id_priv->refcount); - memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr)); if (cma_any_addr(dst_addr)) { ret = cma_resolve_loopback(id_priv); } else { @@ -2976,6 +2982,43 @@ err: return ret == -ENOSPC ? -EADDRNOTAVAIL : ret; } +static int cma_port_is_unique(struct rdma_bind_list *bind_list, + struct rdma_id_private *id_priv) +{ + struct rdma_id_private *cur_id; + struct sockaddr *daddr = cma_dst_addr(id_priv); + struct sockaddr *saddr = cma_src_addr(id_priv); + __be16 dport = cma_port(daddr); + + hlist_for_each_entry(cur_id, &bind_list->owners, node) { + struct sockaddr *cur_daddr = cma_dst_addr(cur_id); + struct sockaddr *cur_saddr = cma_src_addr(cur_id); + __be16 cur_dport = cma_port(cur_daddr); + + if (id_priv == cur_id) + continue; + + /* different dest port -> unique */ + if (!cma_any_port(cur_daddr) && + (dport != cur_dport)) + continue; + + /* different src address -> unique */ + if (!cma_any_addr(saddr) && + !cma_any_addr(cur_saddr) && + cma_addr_cmp(saddr, cur_saddr)) + continue; + + /* different dst address -> unique */ + if (!cma_any_addr(cur_daddr) && + cma_addr_cmp(daddr, cur_daddr)) + continue; + + return -EADDRNOTAVAIL; + } + return 0; +} + static int cma_alloc_any_port(enum rdma_port_space ps, struct rdma_id_private *id_priv) { @@ -2988,9 +3031,19 @@ static int cma_alloc_any_port(enum rdma_port_space ps, remaining = (high - low) + 1; rover = prandom_u32() % remaining + low; retry: - if (last_used_port != rover && - !cma_ps_find(net, ps, (unsigned short)rover)) { - int ret = cma_alloc_port(ps, id_priv, rover); + if (last_used_port != rover) { + struct rdma_bind_list *bind_list; + int ret; + + bind_list = cma_ps_find(net, ps, (unsigned short)rover); + + if (!bind_list) { + ret = cma_alloc_port(ps, id_priv, rover); + } else { + ret = cma_port_is_unique(bind_list, id_priv); + if (!ret) + cma_bind_port(bind_list, id_priv); + } /* * Remember previously used port number in order to avoid * re-using same port immediately after it is closed. -- cgit v1.2.3 From 820cd30ac2ad28fcfeb28f382aa21a29e921678a Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 9 Jan 2017 23:12:16 +0300 Subject: i40iw: fix some indenting in i40iw_sc_vsi_init() The debug printk was indented more than it should have been and we can remove an unnecessary line break. Signed-off-by: Dan Carpenter Acked-by: Shiraz Saleem Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c index dced4f48cbdd..ed84584a97b7 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -4498,9 +4498,9 @@ void i40iw_sc_vsi_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_init_info *inf i40iw_fill_qos_list(info->params->qs_handle_list); for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) { - vsi->qos[i].qs_handle = - info->params->qs_handle_list[i]; - i40iw_debug(vsi->dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, vsi->qos[i].qs_handle); + vsi->qos[i].qs_handle = info->params->qs_handle_list[i]; + i40iw_debug(vsi->dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, + vsi->qos[i].qs_handle); spin_lock_init(&vsi->qos[i].lock); INIT_LIST_HEAD(&vsi->qos[i].qplist); } -- cgit v1.2.3 From dfc0e5550664a727a59921db7d9e7a41c21d03bb Mon Sep 17 00:00:00 2001 From: Zhu Yanjun Date: Thu, 12 Jan 2017 02:39:01 -0500 Subject: IB/ipoib: function interface change The ipoib_ib_dev_down/ipoib_ib_dev_stop return zero unconditionally and the callers never check the returned values, change the return type to void and remove the redundant return values. Reviewed-by: Shan Hai Signed-off-by: Zhu Yanjun Reviewed-by: Yuval Shaia Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib.h | 4 ++-- drivers/infiniband/ulp/ipoib/ipoib_ib.c | 8 ++------ 2 files changed, 4 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index f5680642960a..6dac79454708 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -501,8 +501,8 @@ void ipoib_ib_dev_cleanup(struct net_device *dev); int ipoib_ib_dev_open(struct net_device *dev); int ipoib_ib_dev_up(struct net_device *dev); -int ipoib_ib_dev_down(struct net_device *dev); -int ipoib_ib_dev_stop(struct net_device *dev); +void ipoib_ib_dev_down(struct net_device *dev); +void ipoib_ib_dev_stop(struct net_device *dev); void ipoib_pkey_dev_check_presence(struct net_device *dev); int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 5038f9d2d753..6f443f725fca 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -771,7 +771,7 @@ int ipoib_ib_dev_up(struct net_device *dev) return ipoib_mcast_start_thread(dev); } -int ipoib_ib_dev_down(struct net_device *dev) +void ipoib_ib_dev_down(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); @@ -784,8 +784,6 @@ int ipoib_ib_dev_down(struct net_device *dev) ipoib_mcast_dev_flush(dev); ipoib_flush_paths(dev); - - return 0; } static int recvs_pending(struct net_device *dev) @@ -840,7 +838,7 @@ void ipoib_drain_cq(struct net_device *dev) local_bh_enable(); } -int ipoib_ib_dev_stop(struct net_device *dev) +void ipoib_ib_dev_stop(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); struct ib_qp_attr qp_attr; @@ -913,8 +911,6 @@ timeout: ipoib_flush_ah(dev); ib_req_notify_cq(priv->recv_cq, IB_CQ_NEXT_COMP); - - return 0; } int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port) -- cgit v1.2.3 From 21d6454a392d552c7e845f39884f7cf86f9426b9 Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Tue, 17 Jan 2017 10:11:12 +0100 Subject: RDMA/core: create struct ib_port_cache As Jason suggested, we have 4 elements for per port arrays, it's better to have a separate structure to represent them. It simplifies code a bit, ~ 30 lines of code less :) Signed-off-by: Jack Wang Reviewed-by: Michael Wang Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 134 +++++++++++++++------------------------- include/rdma/ib_verbs.h | 12 ++-- 2 files changed, 59 insertions(+), 87 deletions(-) diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index f91886bab9d1..2e52021aa999 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -314,14 +314,13 @@ static void make_default_gid(struct net_device *dev, union ib_gid *gid) int ib_cache_gid_add(struct ib_device *ib_dev, u8 port, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; int ret = 0; struct net_device *idev; int empty; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (!memcmp(gid, &zgid, sizeof(*gid))) return -EINVAL; @@ -369,11 +368,10 @@ out_unlock: int ib_cache_gid_del(struct ib_device *ib_dev, u8 port, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; mutex_lock(&table->lock); write_lock_irq(&table->rwlock); @@ -399,12 +397,11 @@ out_unlock: int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, struct net_device *ndev) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; bool deleted = false; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; mutex_lock(&table->lock); write_lock_irq(&table->rwlock); @@ -428,10 +425,9 @@ int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, static int __ib_cache_gid_get(struct ib_device *ib_dev, u8 port, int index, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (index < 0 || index >= table->sz) return -EINVAL; @@ -455,14 +451,13 @@ static int _ib_cache_gid_table_find(struct ib_device *ib_dev, unsigned long mask, u8 *port, u16 *index) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; u8 p; int local_index; unsigned long flags; for (p = 0; p < ib_dev->phys_port_cnt; p++) { - table = ports_table[p]; + table = ib_dev->cache.ports[p].gid; read_lock_irqsave(&table->rwlock, flags); local_index = find_gid(table, gid, val, false, mask, NULL); if (local_index >= 0) { @@ -503,7 +498,6 @@ int ib_find_cached_gid_by_port(struct ib_device *ib_dev, u16 *index) { int local_index; - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; unsigned long mask = GID_ATTR_FIND_MASK_GID | GID_ATTR_FIND_MASK_GID_TYPE; @@ -514,7 +508,7 @@ int ib_find_cached_gid_by_port(struct ib_device *ib_dev, port > rdma_end_port(ib_dev)) return -ENOENT; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (ndev) mask |= GID_ATTR_FIND_MASK_NETDEV; @@ -562,21 +556,18 @@ static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev, void *context, u16 *index) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; unsigned int i; unsigned long flags; bool found = false; - if (!ports_table) - return -EOPNOTSUPP; if (port < rdma_start_port(ib_dev) || port > rdma_end_port(ib_dev) || !rdma_protocol_roce(ib_dev, port)) return -EPROTONOSUPPORT; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; read_lock_irqsave(&table->rwlock, flags); for (i = 0; i < table->sz; i++) { @@ -668,14 +659,13 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port, unsigned long gid_type_mask, enum ib_cache_gid_default_mode mode) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; union ib_gid gid; struct ib_gid_attr gid_attr; struct ib_gid_attr zattr_type = zattr; struct ib_gid_table *table; unsigned int gid_type; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; make_default_gid(ndev, &gid); memset(&gid_attr, 0, sizeof(gid_attr)); @@ -766,71 +756,64 @@ static int gid_table_reserve_default(struct ib_device *ib_dev, u8 port, static int _gid_table_setup_one(struct ib_device *ib_dev) { u8 port; - struct ib_gid_table **table; + struct ib_gid_table *table; int err = 0; - table = kcalloc(ib_dev->phys_port_cnt, sizeof(*table), GFP_KERNEL); - if (!table) - return -ENOMEM; - for (port = 0; port < ib_dev->phys_port_cnt; port++) { u8 rdma_port = port + rdma_start_port(ib_dev); - table[port] = + table = alloc_gid_table( ib_dev->port_immutable[rdma_port].gid_tbl_len); - if (!table[port]) { + if (!table) { err = -ENOMEM; goto rollback_table_setup; } err = gid_table_reserve_default(ib_dev, port + rdma_start_port(ib_dev), - table[port]); + table); if (err) goto rollback_table_setup; + ib_dev->cache.ports[port].gid = table; } - ib_dev->cache.gid_cache = table; return 0; rollback_table_setup: for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; + cleanup_gid_table_port(ib_dev, port + rdma_start_port(ib_dev), - table[port]); - release_gid_table(table[port]); + table); + release_gid_table(table); } - kfree(table); return err; } static void gid_table_release_one(struct ib_device *ib_dev) { - struct ib_gid_table **table = ib_dev->cache.gid_cache; + struct ib_gid_table *table; u8 port; - if (!table) - return; - - for (port = 0; port < ib_dev->phys_port_cnt; port++) - release_gid_table(table[port]); - - kfree(table); - ib_dev->cache.gid_cache = NULL; + for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; + release_gid_table(table); + ib_dev->cache.ports[port].gid = NULL; + } } static void gid_table_cleanup_one(struct ib_device *ib_dev) { - struct ib_gid_table **table = ib_dev->cache.gid_cache; + struct ib_gid_table *table; u8 port; - if (!table) - return; - - for (port = 0; port < ib_dev->phys_port_cnt; port++) + for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; cleanup_gid_table_port(ib_dev, port + rdma_start_port(ib_dev), - table[port]); + table); + } } static int gid_table_setup_one(struct ib_device *ib_dev) @@ -860,12 +843,12 @@ int ib_get_cached_gid(struct ib_device *device, { int res; unsigned long flags; - struct ib_gid_table **ports_table = device->cache.gid_cache; - struct ib_gid_table *table = ports_table[port_num - rdma_start_port(device)]; + struct ib_gid_table *table; if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) return -EINVAL; + table = device->cache.ports[port_num - rdma_start_port(device)].gid; read_lock_irqsave(&table->rwlock, flags); res = __ib_cache_gid_get(device, port_num, index, gid, gid_attr); read_unlock_irqrestore(&table->rwlock, flags); @@ -917,7 +900,7 @@ int ib_get_cached_pkey(struct ib_device *device, read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; if (index < 0 || index >= cache->table_len) ret = -EINVAL; @@ -946,7 +929,7 @@ int ib_find_cached_pkey(struct ib_device *device, read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; *index = -1; @@ -986,7 +969,7 @@ int ib_find_exact_cached_pkey(struct ib_device *device, read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; *index = -1; @@ -1014,7 +997,7 @@ int ib_get_cached_lmc(struct ib_device *device, return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - *lmc = device->cache.lmc_cache[port_num - rdma_start_port(device)]; + *lmc = device->cache.ports[port_num - rdma_start_port(device)].lmc; read_unlock_irqrestore(&device->cache.lock, flags); return ret; @@ -1032,7 +1015,8 @@ int ib_get_cached_port_state(struct ib_device *device, return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - *port_state = device->cache.port_state_cache[port_num - rdma_start_port(device)]; + *port_state = device->cache.ports[port_num + - rdma_start_port(device)].port_state; read_unlock_irqrestore(&device->cache.lock, flags); return ret; @@ -1051,14 +1035,13 @@ static void ib_cache_update(struct ib_device *device, int i; int ret; struct ib_gid_table *table; - struct ib_gid_table **ports_table = device->cache.gid_cache; bool use_roce_gid_table = rdma_cap_roce_gid_table(device, port); if (port < rdma_start_port(device) || port > rdma_end_port(device)) return; - table = ports_table[port - rdma_start_port(device)]; + table = device->cache.ports[port - rdma_start_port(device)].gid; tprops = kmalloc(sizeof *tprops, GFP_KERNEL); if (!tprops) @@ -1110,9 +1093,10 @@ static void ib_cache_update(struct ib_device *device, write_lock_irq(&device->cache.lock); - old_pkey_cache = device->cache.pkey_cache[port - rdma_start_port(device)]; + old_pkey_cache = device->cache.ports[port - + rdma_start_port(device)].pkey; - device->cache.pkey_cache[port - rdma_start_port(device)] = pkey_cache; + device->cache.ports[port - rdma_start_port(device)].pkey = pkey_cache; if (!use_roce_gid_table) { write_lock(&table->rwlock); for (i = 0; i < gid_cache->table_len; i++) { @@ -1122,8 +1106,8 @@ static void ib_cache_update(struct ib_device *device, write_unlock(&table->rwlock); } - device->cache.lmc_cache[port - rdma_start_port(device)] = tprops->lmc; - device->cache.port_state_cache[port - rdma_start_port(device)] = + device->cache.ports[port - rdma_start_port(device)].lmc = tprops->lmc; + device->cache.ports[port - rdma_start_port(device)].port_state = tprops->state; write_unlock_irq(&device->cache.lock); @@ -1177,26 +1161,17 @@ int ib_cache_setup_one(struct ib_device *device) rwlock_init(&device->cache.lock); - device->cache.pkey_cache = - kzalloc(sizeof *device->cache.pkey_cache * + device->cache.ports = + kzalloc(sizeof(*device->cache.ports) * (rdma_end_port(device) - rdma_start_port(device) + 1), GFP_KERNEL); - device->cache.lmc_cache = kmalloc(sizeof *device->cache.lmc_cache * - (rdma_end_port(device) - - rdma_start_port(device) + 1), - GFP_KERNEL); - device->cache.port_state_cache = kmalloc(sizeof *device->cache.port_state_cache * - (rdma_end_port(device) - - rdma_start_port(device) + 1), - GFP_KERNEL); - if (!device->cache.pkey_cache || !device->cache.port_state_cache || - !device->cache.lmc_cache) { + if (!device->cache.ports) { err = -ENOMEM; - goto free; + goto out; } err = gid_table_setup_one(device); if (err) - goto free; + goto out; for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p) ib_cache_update(device, p + rdma_start_port(device)); @@ -1211,10 +1186,7 @@ int ib_cache_setup_one(struct ib_device *device) err: gid_table_cleanup_one(device); -free: - kfree(device->cache.pkey_cache); - kfree(device->cache.lmc_cache); - kfree(device->cache.port_state_cache); +out: return err; } @@ -1228,15 +1200,11 @@ void ib_cache_release_one(struct ib_device *device) * all the device's resources when the cache could no * longer be accessed. */ - if (device->cache.pkey_cache) - for (p = 0; - p <= rdma_end_port(device) - rdma_start_port(device); ++p) - kfree(device->cache.pkey_cache[p]); + for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p) + kfree(device->cache.ports[p].pkey); gid_table_release_one(device); - kfree(device->cache.pkey_cache); - kfree(device->cache.lmc_cache); - kfree(device->cache.port_state_cache); + kfree(device->cache.ports); } void ib_cache_cleanup_one(struct ib_device *device) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index fafa988e0e9a..e55afec6bb84 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1775,13 +1775,17 @@ enum ib_mad_result { #define IB_DEVICE_NAME_MAX 64 +struct ib_port_cache { + struct ib_pkey_cache *pkey; + struct ib_gid_table *gid; + u8 lmc; + enum ib_port_state port_state; +}; + struct ib_cache { rwlock_t lock; struct ib_event_handler event_handler; - struct ib_pkey_cache **pkey_cache; - struct ib_gid_table **gid_cache; - u8 *lmc_cache; - enum ib_port_state *port_state_cache; + struct ib_port_cache *ports; }; struct ib_dma_mapping_ops { -- cgit v1.2.3 From 564649b4ea862588be41da4ae2b96b41208b57be Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Wed, 18 Jan 2017 07:57:58 +0200 Subject: IB/qib: Remove empty function Commit f06267104dd9 ("RDMA: Update workqueue usage") removed content of qib_qsfp_deinit(...) and left it empty. This patch deletes all leftovers of that function. Signed-off-by: Leon Romanovsky Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qib/qib_iba7322.c | 1 - drivers/infiniband/hw/qib/qib_qsfp.c | 10 ---------- drivers/infiniband/hw/qib/qib_qsfp.h | 1 - 3 files changed, 12 deletions(-) diff --git a/drivers/infiniband/hw/qib/qib_iba7322.c b/drivers/infiniband/hw/qib/qib_iba7322.c index c4a3616062f1..9cc97bd42775 100644 --- a/drivers/infiniband/hw/qib/qib_iba7322.c +++ b/drivers/infiniband/hw/qib/qib_iba7322.c @@ -2893,7 +2893,6 @@ static void qib_setup_7322_cleanup(struct qib_devdata *dd) dd->cspec->gpio_mask &= ~mask; qib_write_kreg(dd, kr_gpio_mask, dd->cspec->gpio_mask); spin_unlock_irqrestore(&dd->cspec->gpio_lock, flags); - qib_qsfp_deinit(&dd->pport[i].cpspec->qsfp_data); } } } diff --git a/drivers/infiniband/hw/qib/qib_qsfp.c b/drivers/infiniband/hw/qib/qib_qsfp.c index 4c7c3c84a741..295d40a83bb6 100644 --- a/drivers/infiniband/hw/qib/qib_qsfp.c +++ b/drivers/infiniband/hw/qib/qib_qsfp.c @@ -485,16 +485,6 @@ void qib_qsfp_init(struct qib_qsfp_data *qd, dd->f_gpio_mod(dd, mask, mask, mask); } -void qib_qsfp_deinit(struct qib_qsfp_data *qd) -{ - /* - * There is nothing to do here for now. our work is scheduled - * with queue_work(), and flush_workqueue() from remove_one - * will block until all work setup with queue_work() - * completes. - */ -} - int qib_qsfp_dump(struct qib_pportdata *ppd, char *buf, int len) { struct qib_qsfp_cache cd; diff --git a/drivers/infiniband/hw/qib/qib_qsfp.h b/drivers/infiniband/hw/qib/qib_qsfp.h index 91908f533a2b..ad8dbd6ac0cf 100644 --- a/drivers/infiniband/hw/qib/qib_qsfp.h +++ b/drivers/infiniband/hw/qib/qib_qsfp.h @@ -186,4 +186,3 @@ extern int qib_refresh_qsfp_cache(struct qib_pportdata *ppd, extern int qib_qsfp_mod_present(struct qib_pportdata *ppd); extern void qib_qsfp_init(struct qib_qsfp_data *qd, void (*fevent)(struct work_struct *)); -extern void qib_qsfp_deinit(struct qib_qsfp_data *qd); -- cgit v1.2.3 From 3f9fade5e77da67f618ac766ce1a6ded1679fb1a Mon Sep 17 00:00:00 2001 From: Shiraz Saleem Date: Wed, 18 Jan 2017 11:48:29 -0600 Subject: i40iw: Set maj_err and min_err in i40iw_sc_cqp_create Set maj_err and min_err in i40iw_sc_cqp_create so that it returns correct values for all return cases. This also addresses an uninitialized variable warning for maj_err and min_err in i40iw_create_cqp. Reported-by: Dan Carpenter Reported-by: Yuval Shaia Signed-off-by: Shiraz Saleem Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c index ed84584a97b7..f82483b3d1e7 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -450,6 +450,9 @@ static enum i40iw_status_code i40iw_sc_cqp_create(struct i40iw_sc_cqp *cqp, u32 cnt = 0, p1, p2, val = 0, err_code; enum i40iw_status_code ret_code; + *maj_err = 0; + *min_err = 0; + ret_code = i40iw_allocate_dma_mem(cqp->dev->hw, &cqp->sdbuf, 128, -- cgit v1.2.3 From 5c37077fd025c0fa3aa3ab2e6b607d653a4fc604 Mon Sep 17 00:00:00 2001 From: Zhu Yanjun Date: Wed, 18 Jan 2017 23:16:06 -0500 Subject: IB/ipoib: Remove the unnecessary error check The function ipoib_mcast_start_thread/ipoib_ib_dev_up always return zero. As such, in the function ipoib_open, err_stop will never be reached. So remove this err_stop and change the return type of the function ipoib_mcast_start_thread/ipoib_ib_dev_up to void. Signed-off-by: Zhu Yanjun Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib.h | 4 ++-- drivers/infiniband/ulp/ipoib/ipoib_ib.c | 6 +++--- drivers/infiniband/ulp/ipoib/ipoib_main.c | 6 +----- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 4 +--- 4 files changed, 7 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index 6dac79454708..bed233bf45c3 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -500,7 +500,7 @@ void ipoib_pkey_event(struct work_struct *work); void ipoib_ib_dev_cleanup(struct net_device *dev); int ipoib_ib_dev_open(struct net_device *dev); -int ipoib_ib_dev_up(struct net_device *dev); +void ipoib_ib_dev_up(struct net_device *dev); void ipoib_ib_dev_down(struct net_device *dev); void ipoib_ib_dev_stop(struct net_device *dev); void ipoib_pkey_dev_check_presence(struct net_device *dev); @@ -513,7 +513,7 @@ void ipoib_mcast_carrier_on_task(struct work_struct *work); void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb); void ipoib_mcast_restart_task(struct work_struct *work); -int ipoib_mcast_start_thread(struct net_device *dev); +void ipoib_mcast_start_thread(struct net_device *dev); int ipoib_mcast_stop_thread(struct net_device *dev); void ipoib_mcast_dev_down(struct net_device *dev); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 6f443f725fca..12c4f84a6639 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -755,7 +755,7 @@ void ipoib_pkey_dev_check_presence(struct net_device *dev) set_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); } -int ipoib_ib_dev_up(struct net_device *dev) +void ipoib_ib_dev_up(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); @@ -763,12 +763,12 @@ int ipoib_ib_dev_up(struct net_device *dev) if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) { ipoib_dbg(priv, "PKEY is not assigned.\n"); - return 0; + return; } set_bit(IPOIB_FLAG_OPER_UP, &priv->flags); - return ipoib_mcast_start_thread(dev); + ipoib_mcast_start_thread(dev); } void ipoib_ib_dev_down(struct net_device *dev) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index c8af6377a288..e3bfa8a99ad2 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -126,8 +126,7 @@ int ipoib_open(struct net_device *dev) goto err_disable; } - if (ipoib_ib_dev_up(dev)) - goto err_stop; + ipoib_ib_dev_up(dev); if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) { struct ipoib_dev_priv *cpriv; @@ -150,9 +149,6 @@ int ipoib_open(struct net_device *dev) return 0; -err_stop: - ipoib_ib_dev_stop(dev); - err_disable: clear_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 7c6c67bbdab3..69e146cdc306 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -676,7 +676,7 @@ out: spin_unlock_irq(&priv->lock); } -int ipoib_mcast_start_thread(struct net_device *dev) +void ipoib_mcast_start_thread(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); unsigned long flags; @@ -686,8 +686,6 @@ int ipoib_mcast_start_thread(struct net_device *dev) spin_lock_irqsave(&priv->lock, flags); __ipoib_mcast_schedule_join_thread(priv, NULL, 0); spin_unlock_irqrestore(&priv->lock, flags); - - return 0; } int ipoib_mcast_stop_thread(struct net_device *dev) -- cgit v1.2.3 From a3dd3a48a5f6ce433e06863d311dfdd57a26bf67 Mon Sep 17 00:00:00 2001 From: Christophe Jaillet Date: Fri, 27 Jan 2017 14:15:35 +0100 Subject: IB/cma: Fix reversed test This test looks reverted. We should log an error message only if 'ib_attach_mcast()' fails. Signed-off-by: Christophe JAILLET Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index f9133c7606a8..7cb01a9463da 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -3837,7 +3837,7 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) if (!status && id_priv->id.qp) { status = ib_attach_mcast(id_priv->id.qp, &multicast->rec.mgid, be16_to_cpu(multicast->rec.mlid)); - if (!status) + if (status) pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to attach QP. status %d\n", status); } -- cgit v1.2.3 From 2bce1a6d2209c8c776a9598741f5aa1991689dcb Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 9 Dec 2016 11:00:33 -0800 Subject: IB/srpt: Accept GUIDs as port names Port and ACL information must be configured before an initiator logs in. Make it possible to configure this information before a subnet prefix has been assigned to a port by not only accepting GIDs as target port and initiator port names but by also accepting port GUIDs. Add a 'priv' member to struct se_wwn to allow target drivers to associate their own data with struct se_wwn. Reported-by: Doug Ledford References: http://www.spinics.net/lists/linux-rdma/msg39505.html Signed-off-by: Bart Van Assche Cc: Christoph Hellwig Cc: Nicholas Bellinger Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srpt/ib_srpt.c | 139 ++++++++++++++++++++-------------- drivers/infiniband/ulp/srpt/ib_srpt.h | 18 +++-- drivers/target/target_core_tpg.c | 1 + include/target/target_core_base.h | 1 + 4 files changed, 98 insertions(+), 61 deletions(-) diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index d21ba9d857c3..bc5a2d86ae7e 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -500,6 +500,7 @@ static int srpt_refresh_port(struct srpt_port *sport) struct ib_mad_reg_req reg_req; struct ib_port_modify port_modify; struct ib_port_attr port_attr; + __be16 *guid; int ret; memset(&port_modify, 0, sizeof(port_modify)); @@ -522,10 +523,17 @@ static int srpt_refresh_port(struct srpt_port *sport) if (ret) goto err_query_port; + sport->port_guid_wwn.priv = sport; + guid = (__be16 *)&sport->gid.global.interface_id; snprintf(sport->port_guid, sizeof(sport->port_guid), - "0x%016llx%016llx", - be64_to_cpu(sport->gid.global.subnet_prefix), - be64_to_cpu(sport->gid.global.interface_id)); + "%04x:%04x:%04x:%04x", + be16_to_cpu(guid[0]), be16_to_cpu(guid[1]), + be16_to_cpu(guid[2]), be16_to_cpu(guid[3])); + sport->port_gid_wwn.priv = sport; + snprintf(sport->port_gid, sizeof(sport->port_gid), + "0x%016llx%016llx", + be64_to_cpu(sport->gid.global.subnet_prefix), + be64_to_cpu(sport->gid.global.interface_id)); if (!sport->mad_agent) { memset(®_req, 0, sizeof(reg_req)); @@ -1838,6 +1846,7 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id, struct srp_login_rej *rej; struct ib_cm_rep_param *rep_param; struct srpt_rdma_ch *ch, *tmp_ch; + __be16 *guid; u32 it_iu_len; int i, ret = 0; @@ -1983,26 +1992,30 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id, goto destroy_ib; } - /* - * Use the initator port identifier as the session name, when - * checking against se_node_acl->initiatorname[] this can be - * with or without preceeding '0x'. - */ + guid = (__be16 *)¶m->primary_path->sgid.global.interface_id; + snprintf(ch->ini_guid, sizeof(ch->ini_guid), "%04x:%04x:%04x:%04x", + be16_to_cpu(guid[0]), be16_to_cpu(guid[1]), + be16_to_cpu(guid[2]), be16_to_cpu(guid[3])); snprintf(ch->sess_name, sizeof(ch->sess_name), "0x%016llx%016llx", be64_to_cpu(*(__be64 *)ch->i_port_id), be64_to_cpu(*(__be64 *)(ch->i_port_id + 8))); pr_debug("registering session %s\n", ch->sess_name); - ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0, + if (sport->port_guid_tpg.se_tpg_wwn) + ch->sess = target_alloc_session(&sport->port_guid_tpg, 0, 0, + TARGET_PROT_NORMAL, + ch->ini_guid, ch, NULL); + if (sport->port_gid_tpg.se_tpg_wwn && IS_ERR_OR_NULL(ch->sess)) + ch->sess = target_alloc_session(&sport->port_gid_tpg, 0, 0, TARGET_PROT_NORMAL, ch->sess_name, ch, NULL); /* Retry without leading "0x" */ - if (IS_ERR(ch->sess)) - ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0, + if (sport->port_gid_tpg.se_tpg_wwn && IS_ERR_OR_NULL(ch->sess)) + ch->sess = target_alloc_session(&sport->port_gid_tpg, 0, 0, TARGET_PROT_NORMAL, ch->sess_name + 2, ch, NULL); - if (IS_ERR(ch->sess)) { + if (IS_ERR_OR_NULL(ch->sess)) { pr_info("Rejected login because no ACL has been configured yet for initiator %s.\n", ch->sess_name); rej->reason = cpu_to_be32((PTR_ERR(ch->sess) == -ENOMEM) ? @@ -2420,7 +2433,7 @@ static int srpt_release_sdev(struct srpt_device *sdev) return 0; } -static struct srpt_port *__srpt_lookup_port(const char *name) +static struct se_wwn *__srpt_lookup_wwn(const char *name) { struct ib_device *dev; struct srpt_device *sdev; @@ -2435,23 +2448,25 @@ static struct srpt_port *__srpt_lookup_port(const char *name) for (i = 0; i < dev->phys_port_cnt; i++) { sport = &sdev->port[i]; - if (!strcmp(sport->port_guid, name)) - return sport; + if (strcmp(sport->port_guid, name) == 0) + return &sport->port_guid_wwn; + if (strcmp(sport->port_gid, name) == 0) + return &sport->port_gid_wwn; } } return NULL; } -static struct srpt_port *srpt_lookup_port(const char *name) +static struct se_wwn *srpt_lookup_wwn(const char *name) { - struct srpt_port *sport; + struct se_wwn *wwn; spin_lock(&srpt_dev_lock); - sport = __srpt_lookup_port(name); + wwn = __srpt_lookup_wwn(name); spin_unlock(&srpt_dev_lock); - return sport; + return wwn; } /** @@ -2643,11 +2658,19 @@ static char *srpt_get_fabric_name(void) return "srpt"; } +static struct srpt_port *srpt_tpg_to_sport(struct se_portal_group *tpg) +{ + return tpg->se_tpg_wwn->priv; +} + static char *srpt_get_fabric_wwn(struct se_portal_group *tpg) { - struct srpt_port *sport = container_of(tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(tpg); - return sport->port_guid; + WARN_ON_ONCE(tpg != &sport->port_guid_tpg && + tpg != &sport->port_gid_tpg); + return tpg == &sport->port_guid_tpg ? sport->port_guid : + sport->port_gid; } static u16 srpt_get_tag(struct se_portal_group *tpg) @@ -2737,6 +2760,19 @@ static int srpt_get_tcm_cmd_state(struct se_cmd *se_cmd) return srpt_get_cmd_state(ioctx); } +static int srpt_parse_guid(u64 *guid, const char *name) +{ + u16 w[4]; + int ret = -EINVAL; + + if (sscanf(name, "%hx:%hx:%hx:%hx", &w[0], &w[1], &w[2], &w[3]) != 4) + goto out; + *guid = get_unaligned_be64(w); + ret = 0; +out: + return ret; +} + /** * srpt_parse_i_port_id() - Parse an initiator port ID. * @name: ASCII representation of a 128-bit initiator port ID. @@ -2772,20 +2808,23 @@ out: */ static int srpt_init_nodeacl(struct se_node_acl *se_nacl, const char *name) { + u64 guid; u8 i_port_id[16]; + int ret; - if (srpt_parse_i_port_id(i_port_id, name) < 0) { + ret = srpt_parse_guid(&guid, name); + if (ret < 0) + ret = srpt_parse_i_port_id(i_port_id, name); + if (ret < 0) pr_err("invalid initiator port ID %s\n", name); - return -EINVAL; - } - return 0; + return ret; } static ssize_t srpt_tpg_attrib_srp_max_rdma_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_max_rdma_size); } @@ -2794,7 +2833,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rdma_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2822,7 +2861,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rsp_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_max_rsp_size); } @@ -2831,7 +2870,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rsp_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2859,7 +2898,7 @@ static ssize_t srpt_tpg_attrib_srp_sq_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_sq_size); } @@ -2868,7 +2907,7 @@ static ssize_t srpt_tpg_attrib_srp_sq_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2906,7 +2945,7 @@ static struct configfs_attribute *srpt_tpg_attrib_attrs[] = { static ssize_t srpt_tpg_enable_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return snprintf(page, PAGE_SIZE, "%d\n", (sport->enabled) ? 1: 0); } @@ -2915,7 +2954,7 @@ static ssize_t srpt_tpg_enable_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); struct srpt_device *sdev = sport->sdev; struct srpt_rdma_ch *ch; unsigned long tmp; @@ -2967,15 +3006,19 @@ static struct se_portal_group *srpt_make_tpg(struct se_wwn *wwn, struct config_group *group, const char *name) { - struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); + struct srpt_port *sport = wwn->priv; + static struct se_portal_group *tpg; int res; - /* Initialize sport->port_wwn and sport->port_tpg_1 */ - res = core_tpg_register(&sport->port_wwn, &sport->port_tpg_1, SCSI_PROTOCOL_SRP); + WARN_ON_ONCE(wwn != &sport->port_guid_wwn && + wwn != &sport->port_gid_wwn); + tpg = wwn == &sport->port_guid_wwn ? &sport->port_guid_tpg : + &sport->port_gid_tpg; + res = core_tpg_register(wwn, tpg, SCSI_PROTOCOL_SRP); if (res) return ERR_PTR(res); - return &sport->port_tpg_1; + return tpg; } /** @@ -2984,11 +3027,10 @@ static struct se_portal_group *srpt_make_tpg(struct se_wwn *wwn, */ static void srpt_drop_tpg(struct se_portal_group *tpg) { - struct srpt_port *sport = container_of(tpg, - struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(tpg); sport->enabled = false; - core_tpg_deregister(&sport->port_tpg_1); + core_tpg_deregister(tpg); } /** @@ -2999,19 +3041,7 @@ static struct se_wwn *srpt_make_tport(struct target_fabric_configfs *tf, struct config_group *group, const char *name) { - struct srpt_port *sport; - int ret; - - sport = srpt_lookup_port(name); - pr_debug("make_tport(%s)\n", name); - ret = -EINVAL; - if (!sport) - goto err; - - return &sport->port_wwn; - -err: - return ERR_PTR(ret); + return srpt_lookup_wwn(name) ? : ERR_PTR(-EINVAL); } /** @@ -3020,9 +3050,6 @@ err: */ static void srpt_drop_tport(struct se_wwn *wwn) { - struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); - - pr_debug("drop_tport(%s\n", config_item_name(&sport->port_wwn.wwn_group.cg_item)); } static ssize_t srpt_wwn_version_show(struct config_item *item, char *buf) diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.h b/drivers/infiniband/ulp/srpt/ib_srpt.h index 581878782854..cc1183851af5 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.h +++ b/drivers/infiniband/ulp/srpt/ib_srpt.h @@ -258,6 +258,7 @@ enum rdma_ch_state { * against concurrent modification by the cm_id spinlock. * @sess: Session information associated with this SRP channel. * @sess_name: Session name. + * @ini_guid: Initiator port GUID. * @release_work: Allows scheduling of srpt_release_channel(). * @release_done: Enables waiting for srpt_release_channel() completion. */ @@ -284,6 +285,7 @@ struct srpt_rdma_ch { struct list_head cmd_wait_list; struct se_session *sess; u8 sess_name[36]; + u8 ini_guid[24]; struct work_struct release_work; struct completion *release_done; }; @@ -306,28 +308,34 @@ struct srpt_port_attrib { * @mad_agent: per-port management datagram processing information. * @enabled: Whether or not this target port is enabled. * @port_guid: ASCII representation of Port GUID + * @port_gid: ASCII representation of Port GID * @port: one-based port number. * @sm_lid: cached value of the port's sm_lid. * @lid: cached value of the port's lid. * @gid: cached value of the port's gid. * @port_acl_lock spinlock for port_acl_list: * @work: work structure for refreshing the aforementioned cached values. - * @port_tpg_1 Target portal group = 1 data. - * @port_wwn: Target core WWN data. + * @port_guid_tpg: TPG associated with target port GUID. + * @port_guid_wwn: WWN associated with target port GUID. + * @port_gid_tpg: TPG associated with target port GID. + * @port_gid_wwn: WWN associated with target port GID. * @port_acl_list: Head of the list with all node ACLs for this port. */ struct srpt_port { struct srpt_device *sdev; struct ib_mad_agent *mad_agent; bool enabled; - u8 port_guid[64]; + u8 port_guid[24]; + u8 port_gid[64]; u8 port; u16 sm_lid; u16 lid; union ib_gid gid; struct work_struct work; - struct se_portal_group port_tpg_1; - struct se_wwn port_wwn; + struct se_portal_group port_guid_tpg; + struct se_wwn port_guid_wwn; + struct se_portal_group port_gid_tpg; + struct se_wwn port_gid_wwn; struct srpt_port_attrib port_attrib; }; diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c index d99752c6cd60..4a8b180c478b 100644 --- a/drivers/target/target_core_tpg.c +++ b/drivers/target/target_core_tpg.c @@ -448,6 +448,7 @@ static void core_tpg_lun_ref_release(struct percpu_ref *ref) complete(&lun->lun_ref_comp); } +/* Does not change se_wwn->priv. */ int core_tpg_register( struct se_wwn *se_wwn, struct se_portal_group *se_tpg, diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 29e6858bb164..541b52bc811c 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -909,6 +909,7 @@ static inline struct se_portal_group *param_to_tpg(struct config_item *item) struct se_wwn { struct target_fabric_configfs *wwn_tf; + void *priv; struct config_group wwn_group; struct config_group fabric_stat_group; }; -- cgit v1.2.3 From 24dc831b77eca9361cf835be59fa69ea0e471afc Mon Sep 17 00:00:00 2001 From: Yuval Shaia Date: Wed, 25 Jan 2017 18:41:37 +0200 Subject: IB/core: Add inline function to validate port Signed-off-by: Yuval Shaia Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 18 ++++++++---------- drivers/infiniband/core/cma.c | 6 ++---- drivers/infiniband/core/device.c | 4 ++-- drivers/infiniband/core/verbs.c | 3 +-- include/rdma/ib_verbs.h | 7 +++++++ 5 files changed, 20 insertions(+), 18 deletions(-) diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 2e52021aa999..b1371eb9f46c 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -504,8 +504,7 @@ int ib_find_cached_gid_by_port(struct ib_device *ib_dev, struct ib_gid_attr val = {.ndev = ndev, .gid_type = gid_type}; unsigned long flags; - if (port < rdma_start_port(ib_dev) || - port > rdma_end_port(ib_dev)) + if (!rdma_is_port_valid(ib_dev, port)) return -ENOENT; table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; @@ -562,8 +561,7 @@ static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev, bool found = false; - if (port < rdma_start_port(ib_dev) || - port > rdma_end_port(ib_dev) || + if (!rdma_is_port_valid(ib_dev, port) || !rdma_protocol_roce(ib_dev, port)) return -EPROTONOSUPPORT; @@ -845,7 +843,7 @@ int ib_get_cached_gid(struct ib_device *device, unsigned long flags; struct ib_gid_table *table; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; table = device->cache.ports[port_num - rdma_start_port(device)].gid; @@ -895,7 +893,7 @@ int ib_get_cached_pkey(struct ib_device *device, unsigned long flags; int ret = 0; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); @@ -924,7 +922,7 @@ int ib_find_cached_pkey(struct ib_device *device, int ret = -ENOENT; int partial_ix = -1; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); @@ -964,7 +962,7 @@ int ib_find_exact_cached_pkey(struct ib_device *device, int i; int ret = -ENOENT; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); @@ -993,7 +991,7 @@ int ib_get_cached_lmc(struct ib_device *device, unsigned long flags; int ret = 0; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); @@ -1038,7 +1036,7 @@ static void ib_cache_update(struct ib_device *device, bool use_roce_gid_table = rdma_cap_roce_gid_table(device, port); - if (port < rdma_start_port(device) || port > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port)) return; table = device->cache.ports[port - rdma_start_port(device)].gid; diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 7cb01a9463da..9c93e2fa969b 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -269,8 +269,7 @@ struct cma_device *cma_enum_devices_by_ibdev(cma_device_filter filter, int cma_get_default_gid_type(struct cma_device *cma_dev, unsigned int port) { - if (port < rdma_start_port(cma_dev->device) || - port > rdma_end_port(cma_dev->device)) + if (!rdma_is_port_valid(cma_dev->device, port)) return -EINVAL; return cma_dev->default_gid_type[port - rdma_start_port(cma_dev->device)]; @@ -282,8 +281,7 @@ int cma_set_default_gid_type(struct cma_device *cma_dev, { unsigned long supported_gids; - if (port < rdma_start_port(cma_dev->device) || - port > rdma_end_port(cma_dev->device)) + if (!rdma_is_port_valid(cma_dev->device, port)) return -EINVAL; supported_gids = roce_gid_type_mask_support(cma_dev->device, port); diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 571974cd3919..f2e48655a906 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -659,7 +659,7 @@ int ib_query_port(struct ib_device *device, union ib_gid gid; int err; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; memset(port_attr, 0, sizeof(*port_attr)); @@ -825,7 +825,7 @@ int ib_modify_port(struct ib_device *device, if (!device->modify_port) return -ENOSYS; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; return device->modify_port(device, port_num, port_modify_mask, diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 71580cc28c9e..9b77fbc86903 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1205,8 +1205,7 @@ int ib_resolve_eth_dmac(struct ib_device *device, { int ret = 0; - if (ah_attr->port_num < rdma_start_port(device) || - ah_attr->port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, ah_attr->port_num)) return -EINVAL; if (!rdma_cap_eth_ah(device, ah_attr->port_num)) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index e55afec6bb84..b1ac9735fbbe 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -2280,6 +2280,13 @@ static inline u8 rdma_end_port(const struct ib_device *device) return rdma_cap_ib_switch(device) ? 0 : device->phys_port_cnt; } +static inline int rdma_is_port_valid(const struct ib_device *device, + unsigned int port) +{ + return (port >= rdma_start_port(device) && + port <= rdma_end_port(device)); +} + static inline bool rdma_protocol_ib(const struct ib_device *device, u8 port_num) { return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_IB; -- cgit v1.2.3 From 1ac5a404797523cedaf424a3aaa3cf8f9548dff8 Mon Sep 17 00:00:00 2001 From: Selvin Xavier Date: Fri, 10 Feb 2017 03:19:33 -0800 Subject: RDMA/bnxt_re: Add bnxt_re RoCE driver This patch introduces the RoCE driver for the Broadcom NetXtreme-E 10/25/40/50G RoCE HCAs. The RoCE driver is a two part driver that relies on the parent bnxt_en NIC driver to operate. The changes needed in the bnxt_en driver have already been incorporated via Dave Miller's net tree into the mainline kernel. The vendor official git repository for this driver is available on github as: https://github.com/Broadcom/linux-rdma-nxt/ Signed-off-by: Eddie Wai Signed-off-by: Devesh Sharma Signed-off-by: Somnath Kotur Signed-off-by: Sriharsha Basavapatna Signed-off-by: Selvin Xavier Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/bnxt_re/bnxt_re.h | 146 ++ drivers/infiniband/hw/bnxt_re/ib_verbs.c | 3202 ++++++++++++++++++++++++++++ drivers/infiniband/hw/bnxt_re/ib_verbs.h | 197 ++ drivers/infiniband/hw/bnxt_re/main.c | 1315 ++++++++++++ drivers/infiniband/hw/bnxt_re/qplib_fp.c | 2167 +++++++++++++++++++ drivers/infiniband/hw/bnxt_re/qplib_fp.h | 439 ++++ drivers/infiniband/hw/bnxt_re/qplib_rcfw.c | 694 ++++++ drivers/infiniband/hw/bnxt_re/qplib_rcfw.h | 231 ++ drivers/infiniband/hw/bnxt_re/qplib_res.c | 825 +++++++ drivers/infiniband/hw/bnxt_re/qplib_res.h | 223 ++ drivers/infiniband/hw/bnxt_re/qplib_sp.c | 838 ++++++++ drivers/infiniband/hw/bnxt_re/qplib_sp.h | 160 ++ drivers/infiniband/hw/bnxt_re/roce_hsi.h | 2821 ++++++++++++++++++++++++ include/uapi/rdma/bnxt_re-abi.h | 89 + 14 files changed, 13347 insertions(+) create mode 100644 drivers/infiniband/hw/bnxt_re/bnxt_re.h create mode 100644 drivers/infiniband/hw/bnxt_re/ib_verbs.c create mode 100644 drivers/infiniband/hw/bnxt_re/ib_verbs.h create mode 100644 drivers/infiniband/hw/bnxt_re/main.c create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_fp.c create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_fp.h create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_rcfw.c create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_rcfw.h create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_res.c create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_res.h create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_sp.c create mode 100644 drivers/infiniband/hw/bnxt_re/qplib_sp.h create mode 100644 drivers/infiniband/hw/bnxt_re/roce_hsi.h create mode 100644 include/uapi/rdma/bnxt_re-abi.h diff --git a/drivers/infiniband/hw/bnxt_re/bnxt_re.h b/drivers/infiniband/hw/bnxt_re/bnxt_re.h new file mode 100644 index 000000000000..ebf7be8d4139 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/bnxt_re.h @@ -0,0 +1,146 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators (header) + * + */ + +#ifndef __BNXT_RE_H__ +#define __BNXT_RE_H__ +#define ROCE_DRV_MODULE_NAME "bnxt_re" +#define ROCE_DRV_MODULE_VERSION "1.0.0" + +#define BNXT_RE_DESC "Broadcom NetXtreme-C/E RoCE Driver" + +#define BNXT_RE_PAGE_SIZE_4K BIT(12) +#define BNXT_RE_PAGE_SIZE_8K BIT(13) +#define BNXT_RE_PAGE_SIZE_64K BIT(16) +#define BNXT_RE_PAGE_SIZE_2M BIT(21) +#define BNXT_RE_PAGE_SIZE_8M BIT(23) +#define BNXT_RE_PAGE_SIZE_1G BIT(30) + +#define BNXT_RE_MAX_QPC_COUNT (64 * 1024) +#define BNXT_RE_MAX_MRW_COUNT (64 * 1024) +#define BNXT_RE_MAX_SRQC_COUNT (64 * 1024) +#define BNXT_RE_MAX_CQ_COUNT (64 * 1024) + +struct bnxt_re_work { + struct work_struct work; + unsigned long event; + struct bnxt_re_dev *rdev; + struct net_device *vlan_dev; +}; + +struct bnxt_re_sqp_entries { + struct bnxt_qplib_sge sge; + u64 wrid; + /* For storing the actual qp1 cqe */ + struct bnxt_qplib_cqe cqe; + struct bnxt_re_qp *qp1_qp; +}; + +#define BNXT_RE_MIN_MSIX 2 +#define BNXT_RE_MAX_MSIX 16 +#define BNXT_RE_AEQ_IDX 0 +#define BNXT_RE_NQ_IDX 1 + +struct bnxt_re_dev { + struct ib_device ibdev; + struct list_head list; + unsigned long flags; +#define BNXT_RE_FLAG_NETDEV_REGISTERED 0 +#define BNXT_RE_FLAG_IBDEV_REGISTERED 1 +#define BNXT_RE_FLAG_GOT_MSIX 2 +#define BNXT_RE_FLAG_RCFW_CHANNEL_EN 8 +#define BNXT_RE_FLAG_QOS_WORK_REG 16 + struct net_device *netdev; + unsigned int version, major, minor; + struct bnxt_en_dev *en_dev; + struct bnxt_msix_entry msix_entries[BNXT_RE_MAX_MSIX]; + int num_msix; + + int id; + + struct delayed_work worker; + u8 cur_prio_map; + + /* FP Notification Queue (CQ & SRQ) */ + struct tasklet_struct nq_task; + + /* RCFW Channel */ + struct bnxt_qplib_rcfw rcfw; + + /* NQ */ + struct bnxt_qplib_nq nq; + + /* Device Resources */ + struct bnxt_qplib_dev_attr dev_attr; + struct bnxt_qplib_ctx qplib_ctx; + struct bnxt_qplib_res qplib_res; + struct bnxt_qplib_dpi dpi_privileged; + + atomic_t qp_count; + struct mutex qp_lock; /* protect qp list */ + struct list_head qp_list; + + atomic_t cq_count; + atomic_t srq_count; + atomic_t mr_count; + atomic_t mw_count; + /* Max of 2 lossless traffic class supported per port */ + u16 cosq[2]; + + /* QP for for handling QP1 packets */ + u32 sqp_id; + struct bnxt_re_qp *qp1_sqp; + struct bnxt_re_ah *sqp_ah; + struct bnxt_re_sqp_entries sqp_tbl[1024]; +}; + +#define to_bnxt_re_dev(ptr, member) \ + container_of((ptr), struct bnxt_re_dev, member) + +#define BNXT_RE_ROCE_V1_PACKET 0 +#define BNXT_RE_ROCEV2_IPV4_PACKET 2 +#define BNXT_RE_ROCEV2_IPV6_PACKET 3 + +static inline struct device *rdev_to_dev(struct bnxt_re_dev *rdev) +{ + if (rdev) + return &rdev->ibdev.dev; + return NULL; +} + +#endif diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.c b/drivers/infiniband/hw/bnxt_re/ib_verbs.c new file mode 100644 index 000000000000..33af2e3de399 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.c @@ -0,0 +1,3202 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: IB Verbs interpreter + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include "bnxt_ulp.h" + +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_fp.h" +#include "qplib_rcfw.h" + +#include "bnxt_re.h" +#include "ib_verbs.h" +#include + +static int bnxt_re_build_sgl(struct ib_sge *ib_sg_list, + struct bnxt_qplib_sge *sg_list, int num) +{ + int i, total = 0; + + for (i = 0; i < num; i++) { + sg_list[i].addr = ib_sg_list[i].addr; + sg_list[i].lkey = ib_sg_list[i].lkey; + sg_list[i].size = ib_sg_list[i].length; + total += sg_list[i].size; + } + return total; +} + +/* Device */ +struct net_device *bnxt_re_get_netdev(struct ib_device *ibdev, u8 port_num) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct net_device *netdev = NULL; + + rcu_read_lock(); + if (rdev) + netdev = rdev->netdev; + if (netdev) + dev_hold(netdev); + + rcu_read_unlock(); + return netdev; +} + +int bnxt_re_query_device(struct ib_device *ibdev, + struct ib_device_attr *ib_attr, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + + memset(ib_attr, 0, sizeof(*ib_attr)); + + ib_attr->fw_ver = (u64)(unsigned long)(dev_attr->fw_ver); + bnxt_qplib_get_guid(rdev->netdev->dev_addr, + (u8 *)&ib_attr->sys_image_guid); + ib_attr->max_mr_size = ~0ull; + ib_attr->page_size_cap = BNXT_RE_PAGE_SIZE_4K | BNXT_RE_PAGE_SIZE_8K | + BNXT_RE_PAGE_SIZE_64K | BNXT_RE_PAGE_SIZE_2M | + BNXT_RE_PAGE_SIZE_8M | BNXT_RE_PAGE_SIZE_1G; + + ib_attr->vendor_id = rdev->en_dev->pdev->vendor; + ib_attr->vendor_part_id = rdev->en_dev->pdev->device; + ib_attr->hw_ver = rdev->en_dev->pdev->subsystem_device; + ib_attr->max_qp = dev_attr->max_qp; + ib_attr->max_qp_wr = dev_attr->max_qp_wqes; + ib_attr->device_cap_flags = + IB_DEVICE_CURR_QP_STATE_MOD + | IB_DEVICE_RC_RNR_NAK_GEN + | IB_DEVICE_SHUTDOWN_PORT + | IB_DEVICE_SYS_IMAGE_GUID + | IB_DEVICE_LOCAL_DMA_LKEY + | IB_DEVICE_RESIZE_MAX_WR + | IB_DEVICE_PORT_ACTIVE_EVENT + | IB_DEVICE_N_NOTIFY_CQ + | IB_DEVICE_MEM_WINDOW + | IB_DEVICE_MEM_WINDOW_TYPE_2B + | IB_DEVICE_MEM_MGT_EXTENSIONS; + ib_attr->max_sge = dev_attr->max_qp_sges; + ib_attr->max_sge_rd = dev_attr->max_qp_sges; + ib_attr->max_cq = dev_attr->max_cq; + ib_attr->max_cqe = dev_attr->max_cq_wqes; + ib_attr->max_mr = dev_attr->max_mr; + ib_attr->max_pd = dev_attr->max_pd; + ib_attr->max_qp_rd_atom = dev_attr->max_qp_rd_atom; + ib_attr->max_qp_init_rd_atom = dev_attr->max_qp_rd_atom; + ib_attr->atomic_cap = IB_ATOMIC_HCA; + ib_attr->masked_atomic_cap = IB_ATOMIC_HCA; + + ib_attr->max_ee_rd_atom = 0; + ib_attr->max_res_rd_atom = 0; + ib_attr->max_ee_init_rd_atom = 0; + ib_attr->max_ee = 0; + ib_attr->max_rdd = 0; + ib_attr->max_mw = dev_attr->max_mw; + ib_attr->max_raw_ipv6_qp = 0; + ib_attr->max_raw_ethy_qp = dev_attr->max_raw_ethy_qp; + ib_attr->max_mcast_grp = 0; + ib_attr->max_mcast_qp_attach = 0; + ib_attr->max_total_mcast_qp_attach = 0; + ib_attr->max_ah = dev_attr->max_ah; + + ib_attr->max_fmr = dev_attr->max_fmr; + ib_attr->max_map_per_fmr = 1; /* ? */ + + ib_attr->max_srq = dev_attr->max_srq; + ib_attr->max_srq_wr = dev_attr->max_srq_wqes; + ib_attr->max_srq_sge = dev_attr->max_srq_sges; + + ib_attr->max_fast_reg_page_list_len = MAX_PBL_LVL_1_PGS; + + ib_attr->max_pkeys = 1; + ib_attr->local_ca_ack_delay = 0; + return 0; +} + +int bnxt_re_modify_device(struct ib_device *ibdev, + int device_modify_mask, + struct ib_device_modify *device_modify) +{ + switch (device_modify_mask) { + case IB_DEVICE_MODIFY_SYS_IMAGE_GUID: + /* Modify the GUID requires the modification of the GID table */ + /* GUID should be made as READ-ONLY */ + break; + case IB_DEVICE_MODIFY_NODE_DESC: + /* Node Desc should be made as READ-ONLY */ + break; + default: + break; + } + return 0; +} + +static void __to_ib_speed_width(struct net_device *netdev, u8 *speed, u8 *width) +{ + struct ethtool_link_ksettings lksettings; + u32 espeed; + + if (netdev->ethtool_ops && netdev->ethtool_ops->get_link_ksettings) { + memset(&lksettings, 0, sizeof(lksettings)); + rtnl_lock(); + netdev->ethtool_ops->get_link_ksettings(netdev, &lksettings); + rtnl_unlock(); + espeed = lksettings.base.speed; + } else { + espeed = SPEED_UNKNOWN; + } + switch (espeed) { + case SPEED_1000: + *speed = IB_SPEED_SDR; + *width = IB_WIDTH_1X; + break; + case SPEED_10000: + *speed = IB_SPEED_QDR; + *width = IB_WIDTH_1X; + break; + case SPEED_20000: + *speed = IB_SPEED_DDR; + *width = IB_WIDTH_4X; + break; + case SPEED_25000: + *speed = IB_SPEED_EDR; + *width = IB_WIDTH_1X; + break; + case SPEED_40000: + *speed = IB_SPEED_QDR; + *width = IB_WIDTH_4X; + break; + case SPEED_50000: + break; + default: + *speed = IB_SPEED_SDR; + *width = IB_WIDTH_1X; + break; + } +} + +/* Port */ +int bnxt_re_query_port(struct ib_device *ibdev, u8 port_num, + struct ib_port_attr *port_attr) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + + memset(port_attr, 0, sizeof(*port_attr)); + + if (netif_running(rdev->netdev) && netif_carrier_ok(rdev->netdev)) { + port_attr->state = IB_PORT_ACTIVE; + port_attr->phys_state = 5; + } else { + port_attr->state = IB_PORT_DOWN; + port_attr->phys_state = 3; + } + port_attr->max_mtu = IB_MTU_4096; + port_attr->active_mtu = iboe_get_mtu(rdev->netdev->mtu); + port_attr->gid_tbl_len = dev_attr->max_sgid; + port_attr->port_cap_flags = IB_PORT_CM_SUP | IB_PORT_REINIT_SUP | + IB_PORT_DEVICE_MGMT_SUP | + IB_PORT_VENDOR_CLASS_SUP | + IB_PORT_IP_BASED_GIDS; + + /* Max MSG size set to 2G for now */ + port_attr->max_msg_sz = 0x80000000; + port_attr->bad_pkey_cntr = 0; + port_attr->qkey_viol_cntr = 0; + port_attr->pkey_tbl_len = dev_attr->max_pkey; + port_attr->lid = 0; + port_attr->sm_lid = 0; + port_attr->lmc = 0; + port_attr->max_vl_num = 4; + port_attr->sm_sl = 0; + port_attr->subnet_timeout = 0; + port_attr->init_type_reply = 0; + /* call the underlying netdev's ethtool hooks to query speed settings + * for which we acquire rtnl_lock _only_ if it's registered with + * IB stack to avoid race in the NETDEV_UNREG path + */ + if (test_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) + __to_ib_speed_width(rdev->netdev, &port_attr->active_speed, + &port_attr->active_width); + return 0; +} + +int bnxt_re_modify_port(struct ib_device *ibdev, u8 port_num, + int port_modify_mask, + struct ib_port_modify *port_modify) +{ + switch (port_modify_mask) { + case IB_PORT_SHUTDOWN: + break; + case IB_PORT_INIT_TYPE: + break; + case IB_PORT_RESET_QKEY_CNTR: + break; + default: + break; + } + return 0; +} + +int bnxt_re_get_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr port_attr; + + if (bnxt_re_query_port(ibdev, port_num, &port_attr)) + return -EINVAL; + + immutable->pkey_tbl_len = port_attr.pkey_tbl_len; + immutable->gid_tbl_len = port_attr.gid_tbl_len; + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + immutable->core_cap_flags |= RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + return 0; +} + +int bnxt_re_query_pkey(struct ib_device *ibdev, u8 port_num, + u16 index, u16 *pkey) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + + /* Ignore port_num */ + + memset(pkey, 0, sizeof(*pkey)); + return bnxt_qplib_get_pkey(&rdev->qplib_res, + &rdev->qplib_res.pkey_tbl, index, pkey); +} + +int bnxt_re_query_gid(struct ib_device *ibdev, u8 port_num, + int index, union ib_gid *gid) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + int rc = 0; + + /* Ignore port_num */ + memset(gid, 0, sizeof(*gid)); + rc = bnxt_qplib_get_sgid(&rdev->qplib_res, + &rdev->qplib_res.sgid_tbl, index, + (struct bnxt_qplib_gid *)gid); + return rc; +} + +int bnxt_re_del_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, void **context) +{ + int rc = 0; + struct bnxt_re_gid_ctx *ctx, **ctx_tbl; + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_sgid_tbl *sgid_tbl = &rdev->qplib_res.sgid_tbl; + + /* Delete the entry from the hardware */ + ctx = *context; + if (!ctx) + return -EINVAL; + + if (sgid_tbl && sgid_tbl->active) { + if (ctx->idx >= sgid_tbl->max) + return -EINVAL; + ctx->refcnt--; + if (!ctx->refcnt) { + rc = bnxt_qplib_del_sgid + (sgid_tbl, + &sgid_tbl->tbl[ctx->idx], true); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to remove GID: %#x", rc); + ctx_tbl = sgid_tbl->ctx; + ctx_tbl[ctx->idx] = NULL; + kfree(ctx); + } + } else { + return -EINVAL; + } + return rc; +} + +int bnxt_re_add_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, const union ib_gid *gid, + const struct ib_gid_attr *attr, void **context) +{ + int rc; + u32 tbl_idx = 0; + u16 vlan_id = 0xFFFF; + struct bnxt_re_gid_ctx *ctx, **ctx_tbl; + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_sgid_tbl *sgid_tbl = &rdev->qplib_res.sgid_tbl; + + if ((attr->ndev) && is_vlan_dev(attr->ndev)) + vlan_id = vlan_dev_vlan_id(attr->ndev); + + rc = bnxt_qplib_add_sgid(sgid_tbl, (struct bnxt_qplib_gid *)gid, + rdev->qplib_res.netdev->dev_addr, + vlan_id, true, &tbl_idx); + if (rc == -EALREADY) { + ctx_tbl = sgid_tbl->ctx; + ctx_tbl[tbl_idx]->refcnt++; + *context = ctx_tbl[tbl_idx]; + return 0; + } + + if (rc < 0) { + dev_err(rdev_to_dev(rdev), "Failed to add GID: %#x", rc); + return rc; + } + + ctx = kmalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + ctx_tbl = sgid_tbl->ctx; + ctx->idx = tbl_idx; + ctx->refcnt = 1; + ctx_tbl[tbl_idx] = ctx; + + return rc; +} + +enum rdma_link_layer bnxt_re_get_link_layer(struct ib_device *ibdev, + u8 port_num) +{ + return IB_LINK_LAYER_ETHERNET; +} + +/* Protection Domains */ +int bnxt_re_dealloc_pd(struct ib_pd *ib_pd) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + int rc; + + if (ib_pd->uobject && pd->dpi.dbr) { + struct ib_ucontext *ib_uctx = ib_pd->uobject->context; + struct bnxt_re_ucontext *ucntx; + + /* Free DPI only if this is the first PD allocated by the + * application and mark the context dpi as NULL + */ + ucntx = container_of(ib_uctx, struct bnxt_re_ucontext, ib_uctx); + + rc = bnxt_qplib_dealloc_dpi(&rdev->qplib_res, + &rdev->qplib_res.dpi_tbl, + &pd->dpi); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to deallocate HW DPI"); + /* Don't fail, continue*/ + ucntx->dpi = NULL; + } + + rc = bnxt_qplib_dealloc_pd(&rdev->qplib_res, + &rdev->qplib_res.pd_tbl, + &pd->qplib_pd); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to deallocate HW PD"); + return rc; + } + + kfree(pd); + return 0; +} + +struct ib_pd *bnxt_re_alloc_pd(struct ib_device *ibdev, + struct ib_ucontext *ucontext, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_re_ucontext *ucntx = container_of(ucontext, + struct bnxt_re_ucontext, + ib_uctx); + struct bnxt_re_pd *pd; + int rc; + + pd = kzalloc(sizeof(*pd), GFP_KERNEL); + if (!pd) + return ERR_PTR(-ENOMEM); + + pd->rdev = rdev; + if (bnxt_qplib_alloc_pd(&rdev->qplib_res.pd_tbl, &pd->qplib_pd)) { + dev_err(rdev_to_dev(rdev), "Failed to allocate HW PD"); + rc = -ENOMEM; + goto fail; + } + + if (udata) { + struct bnxt_re_pd_resp resp; + + if (!ucntx->dpi) { + /* Allocate DPI in alloc_pd to avoid failing of + * ibv_devinfo and family of application when DPIs + * are depleted. + */ + if (bnxt_qplib_alloc_dpi(&rdev->qplib_res.dpi_tbl, + &pd->dpi, ucntx)) { + rc = -ENOMEM; + goto dbfail; + } + ucntx->dpi = &pd->dpi; + } + + resp.pdid = pd->qplib_pd.id; + /* Still allow mapping this DBR to the new user PD. */ + resp.dpi = ucntx->dpi->dpi; + resp.dbr = (u64)ucntx->dpi->umdbr; + + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to copy user response\n"); + goto dbfail; + } + } + + return &pd->ib_pd; +dbfail: + (void)bnxt_qplib_dealloc_pd(&rdev->qplib_res, &rdev->qplib_res.pd_tbl, + &pd->qplib_pd); +fail: + kfree(pd); + return ERR_PTR(rc); +} + +/* Address Handles */ +int bnxt_re_destroy_ah(struct ib_ah *ib_ah) +{ + struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); + struct bnxt_re_dev *rdev = ah->rdev; + int rc; + + rc = bnxt_qplib_destroy_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW AH"); + return rc; + } + kfree(ah); + return 0; +} + +struct ib_ah *bnxt_re_create_ah(struct ib_pd *ib_pd, + struct ib_ah_attr *ah_attr, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_ah *ah; + int rc; + u16 vlan_tag; + u8 nw_type; + + struct ib_gid_attr sgid_attr; + + if (!(ah_attr->ah_flags & IB_AH_GRH)) { + dev_err(rdev_to_dev(rdev), "Failed to alloc AH: GRH not set"); + return ERR_PTR(-EINVAL); + } + ah = kzalloc(sizeof(*ah), GFP_ATOMIC); + if (!ah) + return ERR_PTR(-ENOMEM); + + ah->rdev = rdev; + ah->qplib_ah.pd = &pd->qplib_pd; + + /* Supply the configuration for the HW */ + memcpy(ah->qplib_ah.dgid.data, ah_attr->grh.dgid.raw, + sizeof(union ib_gid)); + /* + * If RoCE V2 is enabled, stack will have two entries for + * each GID entry. Avoiding this duplicte entry in HW. Dividing + * the GID index by 2 for RoCE V2 + */ + ah->qplib_ah.sgid_index = ah_attr->grh.sgid_index / 2; + ah->qplib_ah.host_sgid_index = ah_attr->grh.sgid_index; + ah->qplib_ah.traffic_class = ah_attr->grh.traffic_class; + ah->qplib_ah.flow_label = ah_attr->grh.flow_label; + ah->qplib_ah.hop_limit = ah_attr->grh.hop_limit; + ah->qplib_ah.sl = ah_attr->sl; + if (ib_pd->uobject && + !rdma_is_multicast_addr((struct in6_addr *) + ah_attr->grh.dgid.raw) && + !rdma_link_local_addr((struct in6_addr *) + ah_attr->grh.dgid.raw)) { + union ib_gid sgid; + + rc = ib_get_cached_gid(&rdev->ibdev, 1, + ah_attr->grh.sgid_index, &sgid, + &sgid_attr); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to query gid at index %d", + ah_attr->grh.sgid_index); + goto fail; + } + if (sgid_attr.ndev) { + if (is_vlan_dev(sgid_attr.ndev)) + vlan_tag = vlan_dev_vlan_id(sgid_attr.ndev); + dev_put(sgid_attr.ndev); + } + /* Get network header type for this GID */ + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V2IPV4; + break; + case RDMA_NETWORK_IPV6: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V2IPV6; + break; + default: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V1; + break; + } + rc = rdma_addr_find_l2_eth_by_grh(&sgid, &ah_attr->grh.dgid, + ah_attr->dmac, &vlan_tag, + &sgid_attr.ndev->ifindex, + NULL); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to get dmac\n"); + goto fail; + } + } + + memcpy(ah->qplib_ah.dmac, ah_attr->dmac, ETH_ALEN); + rc = bnxt_qplib_create_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to allocate HW AH"); + goto fail; + } + + /* Write AVID to shared page. */ + if (ib_pd->uobject) { + struct ib_ucontext *ib_uctx = ib_pd->uobject->context; + struct bnxt_re_ucontext *uctx; + unsigned long flag; + u32 *wrptr; + + uctx = container_of(ib_uctx, struct bnxt_re_ucontext, ib_uctx); + spin_lock_irqsave(&uctx->sh_lock, flag); + wrptr = (u32 *)(uctx->shpg + BNXT_RE_AVID_OFFT); + *wrptr = ah->qplib_ah.id; + wmb(); /* make sure cache is updated. */ + spin_unlock_irqrestore(&uctx->sh_lock, flag); + } + + return &ah->ib_ah; + +fail: + kfree(ah); + return ERR_PTR(rc); +} + +int bnxt_re_modify_ah(struct ib_ah *ib_ah, struct ib_ah_attr *ah_attr) +{ + return 0; +} + +int bnxt_re_query_ah(struct ib_ah *ib_ah, struct ib_ah_attr *ah_attr) +{ + struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); + + memcpy(ah_attr->grh.dgid.raw, ah->qplib_ah.dgid.data, + sizeof(union ib_gid)); + ah_attr->grh.sgid_index = ah->qplib_ah.host_sgid_index; + ah_attr->grh.traffic_class = ah->qplib_ah.traffic_class; + ah_attr->sl = ah->qplib_ah.sl; + memcpy(ah_attr->dmac, ah->qplib_ah.dmac, ETH_ALEN); + ah_attr->ah_flags = IB_AH_GRH; + ah_attr->port_num = 1; + ah_attr->static_rate = 0; + return 0; +} + +/* Queue Pairs */ +int bnxt_re_destroy_qp(struct ib_qp *ib_qp) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + int rc; + + rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW QP"); + return rc; + } + if (ib_qp->qp_type == IB_QPT_GSI && rdev->qp1_sqp) { + rc = bnxt_qplib_destroy_ah(&rdev->qplib_res, + &rdev->sqp_ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to destroy HW AH for shadow QP"); + return rc; + } + + rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, + &rdev->qp1_sqp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to destroy Shadow QP"); + return rc; + } + mutex_lock(&rdev->qp_lock); + list_del(&rdev->qp1_sqp->list); + atomic_dec(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + + kfree(rdev->sqp_ah); + kfree(rdev->qp1_sqp); + } + + if (qp->rumem && !IS_ERR(qp->rumem)) + ib_umem_release(qp->rumem); + if (qp->sumem && !IS_ERR(qp->sumem)) + ib_umem_release(qp->sumem); + + mutex_lock(&rdev->qp_lock); + list_del(&qp->list); + atomic_dec(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + kfree(qp); + return 0; +} + +static u8 __from_ib_qp_type(enum ib_qp_type type) +{ + switch (type) { + case IB_QPT_GSI: + return CMDQ_CREATE_QP1_TYPE_GSI; + case IB_QPT_RC: + return CMDQ_CREATE_QP_TYPE_RC; + case IB_QPT_UD: + return CMDQ_CREATE_QP_TYPE_UD; + default: + return IB_QPT_MAX; + } +} + +static int bnxt_re_init_user_qp(struct bnxt_re_dev *rdev, struct bnxt_re_pd *pd, + struct bnxt_re_qp *qp, struct ib_udata *udata) +{ + struct bnxt_re_qp_req ureq; + struct bnxt_qplib_qp *qplib_qp = &qp->qplib_qp; + struct ib_umem *umem; + int bytes = 0; + struct ib_ucontext *context = pd->ib_pd.uobject->context; + struct bnxt_re_ucontext *cntx = container_of(context, + struct bnxt_re_ucontext, + ib_uctx); + if (ib_copy_from_udata(&ureq, udata, sizeof(ureq))) + return -EFAULT; + + bytes = (qplib_qp->sq.max_wqe * BNXT_QPLIB_MAX_SQE_ENTRY_SIZE); + /* Consider mapping PSN search memory only for RC QPs. */ + if (qplib_qp->type == CMDQ_CREATE_QP_TYPE_RC) + bytes += (qplib_qp->sq.max_wqe * sizeof(struct sq_psn_search)); + bytes = PAGE_ALIGN(bytes); + umem = ib_umem_get(context, ureq.qpsva, bytes, + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(umem)) + return PTR_ERR(umem); + + qp->sumem = umem; + qplib_qp->sq.sglist = umem->sg_head.sgl; + qplib_qp->sq.nmap = umem->nmap; + qplib_qp->qp_handle = ureq.qp_handle; + + if (!qp->qplib_qp.srq) { + bytes = (qplib_qp->rq.max_wqe * BNXT_QPLIB_MAX_RQE_ENTRY_SIZE); + bytes = PAGE_ALIGN(bytes); + umem = ib_umem_get(context, ureq.qprva, bytes, + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(umem)) + goto rqfail; + qp->rumem = umem; + qplib_qp->rq.sglist = umem->sg_head.sgl; + qplib_qp->rq.nmap = umem->nmap; + } + + qplib_qp->dpi = cntx->dpi; + return 0; +rqfail: + ib_umem_release(qp->sumem); + qp->sumem = NULL; + qplib_qp->sq.sglist = NULL; + qplib_qp->sq.nmap = 0; + + return PTR_ERR(umem); +} + +static struct bnxt_re_ah *bnxt_re_create_shadow_qp_ah + (struct bnxt_re_pd *pd, + struct bnxt_qplib_res *qp1_res, + struct bnxt_qplib_qp *qp1_qp) +{ + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_ah *ah; + union ib_gid sgid; + int rc; + + ah = kzalloc(sizeof(*ah), GFP_KERNEL); + if (!ah) + return NULL; + + memset(ah, 0, sizeof(*ah)); + ah->rdev = rdev; + ah->qplib_ah.pd = &pd->qplib_pd; + + rc = bnxt_re_query_gid(&rdev->ibdev, 1, 0, &sgid); + if (rc) + goto fail; + + /* supply the dgid data same as sgid */ + memcpy(ah->qplib_ah.dgid.data, &sgid.raw, + sizeof(union ib_gid)); + ah->qplib_ah.sgid_index = 0; + + ah->qplib_ah.traffic_class = 0; + ah->qplib_ah.flow_label = 0; + ah->qplib_ah.hop_limit = 1; + ah->qplib_ah.sl = 0; + /* Have DMAC same as SMAC */ + ether_addr_copy(ah->qplib_ah.dmac, rdev->netdev->dev_addr); + + rc = bnxt_qplib_create_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate HW AH for Shadow QP"); + goto fail; + } + + return ah; + +fail: + kfree(ah); + return NULL; +} + +static struct bnxt_re_qp *bnxt_re_create_shadow_qp + (struct bnxt_re_pd *pd, + struct bnxt_qplib_res *qp1_res, + struct bnxt_qplib_qp *qp1_qp) +{ + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_qp *qp; + int rc; + + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) + return NULL; + + memset(qp, 0, sizeof(*qp)); + qp->rdev = rdev; + + /* Initialize the shadow QP structure from the QP1 values */ + ether_addr_copy(qp->qplib_qp.smac, rdev->netdev->dev_addr); + + qp->qplib_qp.pd = &pd->qplib_pd; + qp->qplib_qp.qp_handle = (u64)(unsigned long)(&qp->qplib_qp); + qp->qplib_qp.type = IB_QPT_UD; + + qp->qplib_qp.max_inline_data = 0; + qp->qplib_qp.sig_type = true; + + /* Shadow QP SQ depth should be same as QP1 RQ depth */ + qp->qplib_qp.sq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.sq.max_sge = 2; + + qp->qplib_qp.scq = qp1_qp->scq; + qp->qplib_qp.rcq = qp1_qp->rcq; + + qp->qplib_qp.rq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.rq.max_sge = qp1_qp->rq.max_sge; + + qp->qplib_qp.mtu = qp1_qp->mtu; + + qp->qplib_qp.sq_hdr_buf_size = 0; + qp->qplib_qp.rq_hdr_buf_size = BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6; + qp->qplib_qp.dpi = &rdev->dpi_privileged; + + rc = bnxt_qplib_create_qp(qp1_res, &qp->qplib_qp); + if (rc) + goto fail; + + rdev->sqp_id = qp->qplib_qp.id; + + spin_lock_init(&qp->sq_lock); + INIT_LIST_HEAD(&qp->list); + mutex_lock(&rdev->qp_lock); + list_add_tail(&qp->list, &rdev->qp_list); + atomic_inc(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + return qp; +fail: + kfree(qp); + return NULL; +} + +struct ib_qp *bnxt_re_create_qp(struct ib_pd *ib_pd, + struct ib_qp_init_attr *qp_init_attr, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_re_qp *qp; + struct bnxt_re_cq *cq; + int rc, entries; + + if ((qp_init_attr->cap.max_send_wr > dev_attr->max_qp_wqes) || + (qp_init_attr->cap.max_recv_wr > dev_attr->max_qp_wqes) || + (qp_init_attr->cap.max_send_sge > dev_attr->max_qp_sges) || + (qp_init_attr->cap.max_recv_sge > dev_attr->max_qp_sges) || + (qp_init_attr->cap.max_inline_data > dev_attr->max_inline_data)) + return ERR_PTR(-EINVAL); + + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) + return ERR_PTR(-ENOMEM); + + qp->rdev = rdev; + ether_addr_copy(qp->qplib_qp.smac, rdev->netdev->dev_addr); + qp->qplib_qp.pd = &pd->qplib_pd; + qp->qplib_qp.qp_handle = (u64)(unsigned long)(&qp->qplib_qp); + qp->qplib_qp.type = __from_ib_qp_type(qp_init_attr->qp_type); + if (qp->qplib_qp.type == IB_QPT_MAX) { + dev_err(rdev_to_dev(rdev), "QP type 0x%x not supported", + qp->qplib_qp.type); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.max_inline_data = qp_init_attr->cap.max_inline_data; + qp->qplib_qp.sig_type = ((qp_init_attr->sq_sig_type == + IB_SIGNAL_ALL_WR) ? true : false); + + entries = roundup_pow_of_two(qp_init_attr->cap.max_send_wr + 1); + qp->qplib_qp.sq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + + qp->qplib_qp.sq.max_sge = qp_init_attr->cap.max_send_sge; + if (qp->qplib_qp.sq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.sq.max_sge = dev_attr->max_qp_sges; + + if (qp_init_attr->send_cq) { + cq = container_of(qp_init_attr->send_cq, struct bnxt_re_cq, + ib_cq); + if (!cq) { + dev_err(rdev_to_dev(rdev), "Send CQ not found"); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.scq = &cq->qplib_cq; + } + + if (qp_init_attr->recv_cq) { + cq = container_of(qp_init_attr->recv_cq, struct bnxt_re_cq, + ib_cq); + if (!cq) { + dev_err(rdev_to_dev(rdev), "Receive CQ not found"); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.rcq = &cq->qplib_cq; + } + + if (qp_init_attr->srq) { + dev_err(rdev_to_dev(rdev), "SRQ not supported"); + rc = -ENOTSUPP; + goto fail; + } else { + /* Allocate 1 more than what's provided so posting max doesn't + * mean empty + */ + entries = roundup_pow_of_two(qp_init_attr->cap.max_recv_wr + 1); + qp->qplib_qp.rq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + + qp->qplib_qp.rq.max_sge = qp_init_attr->cap.max_recv_sge; + if (qp->qplib_qp.rq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + } + + qp->qplib_qp.mtu = ib_mtu_enum_to_int(iboe_get_mtu(rdev->netdev->mtu)); + + if (qp_init_attr->qp_type == IB_QPT_GSI) { + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + if (qp->qplib_qp.rq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + qp->qplib_qp.sq.max_sge++; + if (qp->qplib_qp.sq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.sq.max_sge = dev_attr->max_qp_sges; + + qp->qplib_qp.rq_hdr_buf_size = + BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2; + + qp->qplib_qp.sq_hdr_buf_size = + BNXT_QPLIB_MAX_QP1_SQ_HDR_SIZE_V2; + qp->qplib_qp.dpi = &rdev->dpi_privileged; + rc = bnxt_qplib_create_qp1(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW QP1"); + goto fail; + } + /* Create a shadow QP to handle the QP1 traffic */ + rdev->qp1_sqp = bnxt_re_create_shadow_qp(pd, &rdev->qplib_res, + &qp->qplib_qp); + if (!rdev->qp1_sqp) { + rc = -EINVAL; + dev_err(rdev_to_dev(rdev), + "Failed to create Shadow QP for QP1"); + goto qp_destroy; + } + rdev->sqp_ah = bnxt_re_create_shadow_qp_ah(pd, &rdev->qplib_res, + &qp->qplib_qp); + if (!rdev->sqp_ah) { + bnxt_qplib_destroy_qp(&rdev->qplib_res, + &rdev->qp1_sqp->qplib_qp); + rc = -EINVAL; + dev_err(rdev_to_dev(rdev), + "Failed to create AH entry for ShadowQP"); + goto qp_destroy; + } + + } else { + qp->qplib_qp.max_rd_atomic = dev_attr->max_qp_rd_atom; + qp->qplib_qp.max_dest_rd_atomic = dev_attr->max_qp_init_rd_atom; + if (udata) { + rc = bnxt_re_init_user_qp(rdev, pd, qp, udata); + if (rc) + goto fail; + } else { + qp->qplib_qp.dpi = &rdev->dpi_privileged; + } + + rc = bnxt_qplib_create_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW QP"); + goto fail; + } + } + + qp->ib_qp.qp_num = qp->qplib_qp.id; + spin_lock_init(&qp->sq_lock); + + if (udata) { + struct bnxt_re_qp_resp resp; + + resp.qpid = qp->ib_qp.qp_num; + resp.rsvd = 0; + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy QP udata"); + goto qp_destroy; + } + } + INIT_LIST_HEAD(&qp->list); + mutex_lock(&rdev->qp_lock); + list_add_tail(&qp->list, &rdev->qp_list); + atomic_inc(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + + return &qp->ib_qp; +qp_destroy: + bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp); +fail: + kfree(qp); + return ERR_PTR(rc); +} + +static u8 __from_ib_qp_state(enum ib_qp_state state) +{ + switch (state) { + case IB_QPS_RESET: + return CMDQ_MODIFY_QP_NEW_STATE_RESET; + case IB_QPS_INIT: + return CMDQ_MODIFY_QP_NEW_STATE_INIT; + case IB_QPS_RTR: + return CMDQ_MODIFY_QP_NEW_STATE_RTR; + case IB_QPS_RTS: + return CMDQ_MODIFY_QP_NEW_STATE_RTS; + case IB_QPS_SQD: + return CMDQ_MODIFY_QP_NEW_STATE_SQD; + case IB_QPS_SQE: + return CMDQ_MODIFY_QP_NEW_STATE_SQE; + case IB_QPS_ERR: + default: + return CMDQ_MODIFY_QP_NEW_STATE_ERR; + } +} + +static enum ib_qp_state __to_ib_qp_state(u8 state) +{ + switch (state) { + case CMDQ_MODIFY_QP_NEW_STATE_RESET: + return IB_QPS_RESET; + case CMDQ_MODIFY_QP_NEW_STATE_INIT: + return IB_QPS_INIT; + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + return IB_QPS_RTR; + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + return IB_QPS_RTS; + case CMDQ_MODIFY_QP_NEW_STATE_SQD: + return IB_QPS_SQD; + case CMDQ_MODIFY_QP_NEW_STATE_SQE: + return IB_QPS_SQE; + case CMDQ_MODIFY_QP_NEW_STATE_ERR: + default: + return IB_QPS_ERR; + } +} + +static u32 __from_ib_mtu(enum ib_mtu mtu) +{ + switch (mtu) { + case IB_MTU_256: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_256; + case IB_MTU_512: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_512; + case IB_MTU_1024: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_1024; + case IB_MTU_2048: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + case IB_MTU_4096: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_4096; + default: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + } +} + +static enum ib_mtu __to_ib_mtu(u32 mtu) +{ + switch (mtu & CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK) { + case CMDQ_MODIFY_QP_PATH_MTU_MTU_256: + return IB_MTU_256; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_512: + return IB_MTU_512; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_1024: + return IB_MTU_1024; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_2048: + return IB_MTU_2048; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_4096: + return IB_MTU_4096; + default: + return IB_MTU_2048; + } +} + +static int __from_ib_access_flags(int iflags) +{ + int qflags = 0; + + if (iflags & IB_ACCESS_LOCAL_WRITE) + qflags |= BNXT_QPLIB_ACCESS_LOCAL_WRITE; + if (iflags & IB_ACCESS_REMOTE_READ) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_READ; + if (iflags & IB_ACCESS_REMOTE_WRITE) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_WRITE; + if (iflags & IB_ACCESS_REMOTE_ATOMIC) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_ATOMIC; + if (iflags & IB_ACCESS_MW_BIND) + qflags |= BNXT_QPLIB_ACCESS_MW_BIND; + if (iflags & IB_ZERO_BASED) + qflags |= BNXT_QPLIB_ACCESS_ZERO_BASED; + if (iflags & IB_ACCESS_ON_DEMAND) + qflags |= BNXT_QPLIB_ACCESS_ON_DEMAND; + return qflags; +}; + +static enum ib_access_flags __to_ib_access_flags(int qflags) +{ + enum ib_access_flags iflags = 0; + + if (qflags & BNXT_QPLIB_ACCESS_LOCAL_WRITE) + iflags |= IB_ACCESS_LOCAL_WRITE; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_WRITE) + iflags |= IB_ACCESS_REMOTE_WRITE; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_READ) + iflags |= IB_ACCESS_REMOTE_READ; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_ATOMIC) + iflags |= IB_ACCESS_REMOTE_ATOMIC; + if (qflags & BNXT_QPLIB_ACCESS_MW_BIND) + iflags |= IB_ACCESS_MW_BIND; + if (qflags & BNXT_QPLIB_ACCESS_ZERO_BASED) + iflags |= IB_ZERO_BASED; + if (qflags & BNXT_QPLIB_ACCESS_ON_DEMAND) + iflags |= IB_ACCESS_ON_DEMAND; + return iflags; +}; + +static int bnxt_re_modify_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp1_qp, + int qp_attr_mask) +{ + struct bnxt_re_qp *qp = rdev->qp1_sqp; + int rc = 0; + + if (qp_attr_mask & IB_QP_STATE) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_STATE; + qp->qplib_qp.state = qp1_qp->qplib_qp.state; + } + if (qp_attr_mask & IB_QP_PKEY_INDEX) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PKEY; + qp->qplib_qp.pkey_index = qp1_qp->qplib_qp.pkey_index; + } + + if (qp_attr_mask & IB_QP_QKEY) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_QKEY; + /* Using a Random QKEY */ + qp->qplib_qp.qkey = 0x81818181; + } + if (qp_attr_mask & IB_QP_SQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN; + qp->qplib_qp.sq.psn = qp1_qp->qplib_qp.sq.psn; + } + + rc = bnxt_qplib_modify_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to modify Shadow QP for QP1"); + return rc; +} + +int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_udata *udata) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + enum ib_qp_state curr_qp_state, new_qp_state; + int rc, entries; + int status; + union ib_gid sgid; + struct ib_gid_attr sgid_attr; + u8 nw_type; + + qp->qplib_qp.modify_flags = 0; + if (qp_attr_mask & IB_QP_STATE) { + curr_qp_state = __to_ib_qp_state(qp->qplib_qp.cur_qp_state); + new_qp_state = qp_attr->qp_state; + if (!ib_modify_qp_is_ok(curr_qp_state, new_qp_state, + ib_qp->qp_type, qp_attr_mask, + IB_LINK_LAYER_ETHERNET)) { + dev_err(rdev_to_dev(rdev), + "Invalid attribute mask: %#x specified ", + qp_attr_mask); + dev_err(rdev_to_dev(rdev), + "for qpn: %#x type: %#x", + ib_qp->qp_num, ib_qp->qp_type); + dev_err(rdev_to_dev(rdev), + "curr_qp_state=0x%x, new_qp_state=0x%x\n", + curr_qp_state, new_qp_state); + return -EINVAL; + } + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_STATE; + qp->qplib_qp.state = __from_ib_qp_state(qp_attr->qp_state); + } + if (qp_attr_mask & IB_QP_EN_SQD_ASYNC_NOTIFY) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_EN_SQD_ASYNC_NOTIFY; + qp->qplib_qp.en_sqd_async_notify = true; + } + if (qp_attr_mask & IB_QP_ACCESS_FLAGS) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS; + qp->qplib_qp.access = + __from_ib_access_flags(qp_attr->qp_access_flags); + /* LOCAL_WRITE access must be set to allow RC receive */ + qp->qplib_qp.access |= BNXT_QPLIB_ACCESS_LOCAL_WRITE; + } + if (qp_attr_mask & IB_QP_PKEY_INDEX) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PKEY; + qp->qplib_qp.pkey_index = qp_attr->pkey_index; + } + if (qp_attr_mask & IB_QP_QKEY) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_QKEY; + qp->qplib_qp.qkey = qp_attr->qkey; + } + if (qp_attr_mask & IB_QP_AV) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_DGID | + CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL | + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX | + CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT | + CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC | + CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID; + memcpy(qp->qplib_qp.ah.dgid.data, qp_attr->ah_attr.grh.dgid.raw, + sizeof(qp->qplib_qp.ah.dgid.data)); + qp->qplib_qp.ah.flow_label = qp_attr->ah_attr.grh.flow_label; + /* If RoCE V2 is enabled, stack will have two entries for + * each GID entry. Avoiding this duplicte entry in HW. Dividing + * the GID index by 2 for RoCE V2 + */ + qp->qplib_qp.ah.sgid_index = + qp_attr->ah_attr.grh.sgid_index / 2; + qp->qplib_qp.ah.host_sgid_index = + qp_attr->ah_attr.grh.sgid_index; + qp->qplib_qp.ah.hop_limit = qp_attr->ah_attr.grh.hop_limit; + qp->qplib_qp.ah.traffic_class = + qp_attr->ah_attr.grh.traffic_class; + qp->qplib_qp.ah.sl = qp_attr->ah_attr.sl; + ether_addr_copy(qp->qplib_qp.ah.dmac, qp_attr->ah_attr.dmac); + + status = ib_get_cached_gid(&rdev->ibdev, 1, + qp_attr->ah_attr.grh.sgid_index, + &sgid, &sgid_attr); + if (!status && sgid_attr.ndev) { + memcpy(qp->qplib_qp.smac, sgid_attr.ndev->dev_addr, + ETH_ALEN); + dev_put(sgid_attr.ndev); + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, + &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV4; + break; + case RDMA_NETWORK_IPV6: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV6; + break; + default: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV1; + break; + } + } + } + + if (qp_attr_mask & IB_QP_PATH_MTU) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->qplib_qp.path_mtu = __from_ib_mtu(qp_attr->path_mtu); + } else if (qp_attr->qp_state == IB_QPS_RTR) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->qplib_qp.path_mtu = + __from_ib_mtu(iboe_get_mtu(rdev->netdev->mtu)); + } + + if (qp_attr_mask & IB_QP_TIMEOUT) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT; + qp->qplib_qp.timeout = qp_attr->timeout; + } + if (qp_attr_mask & IB_QP_RETRY_CNT) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT; + qp->qplib_qp.retry_cnt = qp_attr->retry_cnt; + } + if (qp_attr_mask & IB_QP_RNR_RETRY) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY; + qp->qplib_qp.rnr_retry = qp_attr->rnr_retry; + } + if (qp_attr_mask & IB_QP_MIN_RNR_TIMER) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER; + qp->qplib_qp.min_rnr_timer = qp_attr->min_rnr_timer; + } + if (qp_attr_mask & IB_QP_RQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN; + qp->qplib_qp.rq.psn = qp_attr->rq_psn; + } + if (qp_attr_mask & IB_QP_MAX_QP_RD_ATOMIC) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC; + qp->qplib_qp.max_rd_atomic = qp_attr->max_rd_atomic; + } + if (qp_attr_mask & IB_QP_SQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN; + qp->qplib_qp.sq.psn = qp_attr->sq_psn; + } + if (qp_attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC; + qp->qplib_qp.max_dest_rd_atomic = qp_attr->max_dest_rd_atomic; + } + if (qp_attr_mask & IB_QP_CAP) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SIZE | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SIZE | + CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SGE | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SGE | + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_INLINE_DATA; + if ((qp_attr->cap.max_send_wr >= dev_attr->max_qp_wqes) || + (qp_attr->cap.max_recv_wr >= dev_attr->max_qp_wqes) || + (qp_attr->cap.max_send_sge >= dev_attr->max_qp_sges) || + (qp_attr->cap.max_recv_sge >= dev_attr->max_qp_sges) || + (qp_attr->cap.max_inline_data >= + dev_attr->max_inline_data)) { + dev_err(rdev_to_dev(rdev), + "Create QP failed - max exceeded"); + return -EINVAL; + } + entries = roundup_pow_of_two(qp_attr->cap.max_send_wr); + qp->qplib_qp.sq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + qp->qplib_qp.sq.max_sge = qp_attr->cap.max_send_sge; + if (qp->qplib_qp.rq.max_wqe) { + entries = roundup_pow_of_two(qp_attr->cap.max_recv_wr); + qp->qplib_qp.rq.max_wqe = + min_t(u32, entries, dev_attr->max_qp_wqes + 1); + qp->qplib_qp.rq.max_sge = qp_attr->cap.max_recv_sge; + } else { + /* SRQ was used prior, just ignore the RQ caps */ + } + } + if (qp_attr_mask & IB_QP_DEST_QPN) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID; + qp->qplib_qp.dest_qpn = qp_attr->dest_qp_num; + } + rc = bnxt_qplib_modify_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to modify HW QP"); + return rc; + } + if (ib_qp->qp_type == IB_QPT_GSI && rdev->qp1_sqp) + rc = bnxt_re_modify_shadow_qp(rdev, qp, qp_attr_mask); + return rc; +} + +int bnxt_re_query_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_qplib_qp qplib_qp; + int rc; + + memset(&qplib_qp, 0, sizeof(struct bnxt_qplib_qp)); + qplib_qp.id = qp->qplib_qp.id; + qplib_qp.ah.host_sgid_index = qp->qplib_qp.ah.host_sgid_index; + + rc = bnxt_qplib_query_qp(&rdev->qplib_res, &qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to query HW QP"); + return rc; + } + qp_attr->qp_state = __to_ib_qp_state(qplib_qp.state); + qp_attr->en_sqd_async_notify = qplib_qp.en_sqd_async_notify ? 1 : 0; + qp_attr->qp_access_flags = __to_ib_access_flags(qplib_qp.access); + qp_attr->pkey_index = qplib_qp.pkey_index; + qp_attr->qkey = qplib_qp.qkey; + memcpy(qp_attr->ah_attr.grh.dgid.raw, qplib_qp.ah.dgid.data, + sizeof(qplib_qp.ah.dgid.data)); + qp_attr->ah_attr.grh.flow_label = qplib_qp.ah.flow_label; + qp_attr->ah_attr.grh.sgid_index = qplib_qp.ah.host_sgid_index; + qp_attr->ah_attr.grh.hop_limit = qplib_qp.ah.hop_limit; + qp_attr->ah_attr.grh.traffic_class = qplib_qp.ah.traffic_class; + qp_attr->ah_attr.sl = qplib_qp.ah.sl; + ether_addr_copy(qp_attr->ah_attr.dmac, qplib_qp.ah.dmac); + qp_attr->path_mtu = __to_ib_mtu(qplib_qp.path_mtu); + qp_attr->timeout = qplib_qp.timeout; + qp_attr->retry_cnt = qplib_qp.retry_cnt; + qp_attr->rnr_retry = qplib_qp.rnr_retry; + qp_attr->min_rnr_timer = qplib_qp.min_rnr_timer; + qp_attr->rq_psn = qplib_qp.rq.psn; + qp_attr->max_rd_atomic = qplib_qp.max_rd_atomic; + qp_attr->sq_psn = qplib_qp.sq.psn; + qp_attr->max_dest_rd_atomic = qplib_qp.max_dest_rd_atomic; + qp_init_attr->sq_sig_type = qplib_qp.sig_type ? IB_SIGNAL_ALL_WR : + IB_SIGNAL_REQ_WR; + qp_attr->dest_qp_num = qplib_qp.dest_qpn; + + qp_attr->cap.max_send_wr = qp->qplib_qp.sq.max_wqe; + qp_attr->cap.max_send_sge = qp->qplib_qp.sq.max_sge; + qp_attr->cap.max_recv_wr = qp->qplib_qp.rq.max_wqe; + qp_attr->cap.max_recv_sge = qp->qplib_qp.rq.max_sge; + qp_attr->cap.max_inline_data = qp->qplib_qp.max_inline_data; + qp_init_attr->cap = qp_attr->cap; + + return 0; +} + +/* Routine for sending QP1 packets for RoCE V1 an V2 + */ +static int bnxt_re_build_qp1_send_v2(struct bnxt_re_qp *qp, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe, + int payload_size) +{ + struct ib_device *ibdev = &qp->rdev->ibdev; + struct bnxt_re_ah *ah = container_of(ud_wr(wr)->ah, struct bnxt_re_ah, + ib_ah); + struct bnxt_qplib_ah *qplib_ah = &ah->qplib_ah; + struct bnxt_qplib_sge sge; + union ib_gid sgid; + u8 nw_type; + u16 ether_type; + struct ib_gid_attr sgid_attr; + union ib_gid dgid; + bool is_eth = false; + bool is_vlan = false; + bool is_grh = false; + bool is_udp = false; + u8 ip_version = 0; + u16 vlan_id = 0xFFFF; + void *buf; + int i, rc = 0, size; + + memset(&qp->qp1_hdr, 0, sizeof(qp->qp1_hdr)); + + rc = ib_get_cached_gid(ibdev, 1, + qplib_ah->host_sgid_index, &sgid, + &sgid_attr); + if (rc) { + dev_err(rdev_to_dev(qp->rdev), + "Failed to query gid at index %d", + qplib_ah->host_sgid_index); + return rc; + } + if (sgid_attr.ndev) { + if (is_vlan_dev(sgid_attr.ndev)) + vlan_id = vlan_dev_vlan_id(sgid_attr.ndev); + dev_put(sgid_attr.ndev); + } + /* Get network header type for this GID */ + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + nw_type = BNXT_RE_ROCEV2_IPV4_PACKET; + break; + case RDMA_NETWORK_IPV6: + nw_type = BNXT_RE_ROCEV2_IPV6_PACKET; + break; + default: + nw_type = BNXT_RE_ROCE_V1_PACKET; + break; + } + memcpy(&dgid.raw, &qplib_ah->dgid, 16); + is_udp = sgid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP; + if (is_udp) { + if (ipv6_addr_v4mapped((struct in6_addr *)&sgid)) { + ip_version = 4; + ether_type = ETH_P_IP; + } else { + ip_version = 6; + ether_type = ETH_P_IPV6; + } + is_grh = false; + } else { + ether_type = ETH_P_IBOE; + is_grh = true; + } + + is_eth = true; + is_vlan = (vlan_id && (vlan_id < 0x1000)) ? true : false; + + ib_ud_header_init(payload_size, !is_eth, is_eth, is_vlan, is_grh, + ip_version, is_udp, 0, &qp->qp1_hdr); + + /* ETH */ + ether_addr_copy(qp->qp1_hdr.eth.dmac_h, ah->qplib_ah.dmac); + ether_addr_copy(qp->qp1_hdr.eth.smac_h, qp->qplib_qp.smac); + + /* For vlan, check the sgid for vlan existence */ + + if (!is_vlan) { + qp->qp1_hdr.eth.type = cpu_to_be16(ether_type); + } else { + qp->qp1_hdr.vlan.type = cpu_to_be16(ether_type); + qp->qp1_hdr.vlan.tag = cpu_to_be16(vlan_id); + } + + if (is_grh || (ip_version == 6)) { + memcpy(qp->qp1_hdr.grh.source_gid.raw, sgid.raw, sizeof(sgid)); + memcpy(qp->qp1_hdr.grh.destination_gid.raw, qplib_ah->dgid.data, + sizeof(sgid)); + qp->qp1_hdr.grh.hop_limit = qplib_ah->hop_limit; + } + + if (ip_version == 4) { + qp->qp1_hdr.ip4.tos = 0; + qp->qp1_hdr.ip4.id = 0; + qp->qp1_hdr.ip4.frag_off = htons(IP_DF); + qp->qp1_hdr.ip4.ttl = qplib_ah->hop_limit; + + memcpy(&qp->qp1_hdr.ip4.saddr, sgid.raw + 12, 4); + memcpy(&qp->qp1_hdr.ip4.daddr, qplib_ah->dgid.data + 12, 4); + qp->qp1_hdr.ip4.check = ib_ud_ip4_csum(&qp->qp1_hdr); + } + + if (is_udp) { + qp->qp1_hdr.udp.dport = htons(ROCE_V2_UDP_DPORT); + qp->qp1_hdr.udp.sport = htons(0x8CD1); + qp->qp1_hdr.udp.csum = 0; + } + + /* BTH */ + if (wr->opcode == IB_WR_SEND_WITH_IMM) { + qp->qp1_hdr.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; + qp->qp1_hdr.immediate_present = 1; + } else { + qp->qp1_hdr.bth.opcode = IB_OPCODE_UD_SEND_ONLY; + } + if (wr->send_flags & IB_SEND_SOLICITED) + qp->qp1_hdr.bth.solicited_event = 1; + /* pad_count */ + qp->qp1_hdr.bth.pad_count = (4 - payload_size) & 3; + + /* P_key for QP1 is for all members */ + qp->qp1_hdr.bth.pkey = cpu_to_be16(0xFFFF); + qp->qp1_hdr.bth.destination_qpn = IB_QP1; + qp->qp1_hdr.bth.ack_req = 0; + qp->send_psn++; + qp->send_psn &= BTH_PSN_MASK; + qp->qp1_hdr.bth.psn = cpu_to_be32(qp->send_psn); + /* DETH */ + /* Use the priviledged Q_Key for QP1 */ + qp->qp1_hdr.deth.qkey = cpu_to_be32(IB_QP1_QKEY); + qp->qp1_hdr.deth.source_qpn = IB_QP1; + + /* Pack the QP1 to the transmit buffer */ + buf = bnxt_qplib_get_qp1_sq_buf(&qp->qplib_qp, &sge); + if (buf) { + size = ib_ud_header_pack(&qp->qp1_hdr, buf); + for (i = wqe->num_sge; i; i--) { + wqe->sg_list[i].addr = wqe->sg_list[i - 1].addr; + wqe->sg_list[i].lkey = wqe->sg_list[i - 1].lkey; + wqe->sg_list[i].size = wqe->sg_list[i - 1].size; + } + + /* + * Max Header buf size for IPV6 RoCE V2 is 86, + * which is same as the QP1 SQ header buffer. + * Header buf size for IPV4 RoCE V2 can be 66. + * ETH(14) + VLAN(4)+ IP(20) + UDP (8) + BTH(20). + * Subtract 20 bytes from QP1 SQ header buf size + */ + if (is_udp && ip_version == 4) + sge.size -= 20; + /* + * Max Header buf size for RoCE V1 is 78. + * ETH(14) + VLAN(4) + GRH(40) + BTH(20). + * Subtract 8 bytes from QP1 SQ header buf size + */ + if (!is_udp) + sge.size -= 8; + + /* Subtract 4 bytes for non vlan packets */ + if (!is_vlan) + sge.size -= 4; + + wqe->sg_list[0].addr = sge.addr; + wqe->sg_list[0].lkey = sge.lkey; + wqe->sg_list[0].size = sge.size; + wqe->num_sge++; + + } else { + dev_err(rdev_to_dev(qp->rdev), "QP1 buffer is empty!"); + rc = -ENOMEM; + } + return rc; +} + +/* For the MAD layer, it only provides the recv SGE the size of + * ib_grh + MAD datagram. No Ethernet headers, Ethertype, BTH, DETH, + * nor RoCE iCRC. The Cu+ solution must provide buffer for the entire + * receive packet (334 bytes) with no VLAN and then copy the GRH + * and the MAD datagram out to the provided SGE. + */ +static int bnxt_re_build_qp1_shadow_qp_recv(struct bnxt_re_qp *qp, + struct ib_recv_wr *wr, + struct bnxt_qplib_swqe *wqe, + int payload_size) +{ + struct bnxt_qplib_sge ref, sge; + u32 rq_prod_index; + struct bnxt_re_sqp_entries *sqp_entry; + + rq_prod_index = bnxt_qplib_get_rq_prod_index(&qp->qplib_qp); + + if (!bnxt_qplib_get_qp1_rq_buf(&qp->qplib_qp, &sge)) + return -ENOMEM; + + /* Create 1 SGE to receive the entire + * ethernet packet + */ + /* Save the reference from ULP */ + ref.addr = wqe->sg_list[0].addr; + ref.lkey = wqe->sg_list[0].lkey; + ref.size = wqe->sg_list[0].size; + + sqp_entry = &qp->rdev->sqp_tbl[rq_prod_index]; + + /* SGE 1 */ + wqe->sg_list[0].addr = sge.addr; + wqe->sg_list[0].lkey = sge.lkey; + wqe->sg_list[0].size = BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2; + sge.size -= wqe->sg_list[0].size; + + sqp_entry->sge.addr = ref.addr; + sqp_entry->sge.lkey = ref.lkey; + sqp_entry->sge.size = ref.size; + /* Store the wrid for reporting completion */ + sqp_entry->wrid = wqe->wr_id; + /* change the wqe->wrid to table index */ + wqe->wr_id = rq_prod_index; + return 0; +} + +static int is_ud_qp(struct bnxt_re_qp *qp) +{ + return qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_UD; +} + +static int bnxt_re_build_send_wqe(struct bnxt_re_qp *qp, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_re_ah *ah = NULL; + + if (is_ud_qp(qp)) { + ah = container_of(ud_wr(wr)->ah, struct bnxt_re_ah, ib_ah); + wqe->send.q_key = ud_wr(wr)->remote_qkey; + wqe->send.dst_qp = ud_wr(wr)->remote_qpn; + wqe->send.avid = ah->qplib_ah.id; + } + switch (wr->opcode) { + case IB_WR_SEND: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND; + break; + case IB_WR_SEND_WITH_IMM: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM; + wqe->send.imm_data = wr->ex.imm_data; + break; + case IB_WR_SEND_WITH_INV: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV; + wqe->send.inv_key = wr->ex.invalidate_rkey; + break; + default: + return -EINVAL; + } + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + if (wr->send_flags & IB_SEND_INLINE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_INLINE; + + return 0; +} + +static int bnxt_re_build_rdma_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + switch (wr->opcode) { + case IB_WR_RDMA_WRITE: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE; + break; + case IB_WR_RDMA_WRITE_WITH_IMM: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM; + wqe->rdma.imm_data = wr->ex.imm_data; + break; + case IB_WR_RDMA_READ: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_READ; + wqe->rdma.inv_key = wr->ex.invalidate_rkey; + break; + default: + return -EINVAL; + } + wqe->rdma.remote_va = rdma_wr(wr)->remote_addr; + wqe->rdma.r_key = rdma_wr(wr)->rkey; + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + if (wr->send_flags & IB_SEND_INLINE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_INLINE; + + return 0; +} + +static int bnxt_re_build_atomic_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + switch (wr->opcode) { + case IB_WR_ATOMIC_CMP_AND_SWP: + wqe->type = BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP; + wqe->atomic.swap_data = atomic_wr(wr)->swap; + break; + case IB_WR_ATOMIC_FETCH_AND_ADD: + wqe->type = BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD; + wqe->atomic.cmp_data = atomic_wr(wr)->compare_add; + break; + default: + return -EINVAL; + } + wqe->atomic.remote_va = atomic_wr(wr)->remote_addr; + wqe->atomic.r_key = atomic_wr(wr)->rkey; + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + return 0; +} + +static int bnxt_re_build_inv_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + wqe->type = BNXT_QPLIB_SWQE_TYPE_LOCAL_INV; + wqe->local_inv.inv_l_key = wr->ex.invalidate_rkey; + + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + + return 0; +} + +static int bnxt_re_build_reg_wqe(struct ib_reg_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_re_mr *mr = container_of(wr->mr, struct bnxt_re_mr, ib_mr); + struct bnxt_qplib_frpl *qplib_frpl = &mr->qplib_frpl; + int access = wr->access; + + wqe->frmr.pbl_ptr = (__le64 *)qplib_frpl->hwq.pbl_ptr[0]; + wqe->frmr.pbl_dma_ptr = qplib_frpl->hwq.pbl_dma_ptr[0]; + wqe->frmr.page_list = mr->pages; + wqe->frmr.page_list_len = mr->npages; + wqe->frmr.levels = qplib_frpl->hwq.level + 1; + wqe->type = BNXT_QPLIB_SWQE_TYPE_REG_MR; + + if (wr->wr.send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->wr.send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + + if (access & IB_ACCESS_LOCAL_WRITE) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE; + if (access & IB_ACCESS_REMOTE_READ) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_READ; + if (access & IB_ACCESS_REMOTE_WRITE) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_WRITE; + if (access & IB_ACCESS_REMOTE_ATOMIC) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_ATOMIC; + if (access & IB_ACCESS_MW_BIND) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_WINDOW_BIND; + + wqe->frmr.l_key = wr->key; + wqe->frmr.length = wr->mr->length; + wqe->frmr.pbl_pg_sz_log = (wr->mr->page_size >> PAGE_SHIFT_4K) - 1; + wqe->frmr.va = wr->mr->iova; + return 0; +} + +static int bnxt_re_copy_inline_data(struct bnxt_re_dev *rdev, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + /* Copy the inline data to the data field */ + u8 *in_data; + u32 i, sge_len; + void *sge_addr; + + in_data = wqe->inline_data; + for (i = 0; i < wr->num_sge; i++) { + sge_addr = (void *)(unsigned long) + wr->sg_list[i].addr; + sge_len = wr->sg_list[i].length; + + if ((sge_len + wqe->inline_len) > + BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH) { + dev_err(rdev_to_dev(rdev), + "Inline data size requested > supported value"); + return -EINVAL; + } + sge_len = wr->sg_list[i].length; + + memcpy(in_data, sge_addr, sge_len); + in_data += wr->sg_list[i].length; + wqe->inline_len += wr->sg_list[i].length; + } + return wqe->inline_len; +} + +static int bnxt_re_copy_wr_payload(struct bnxt_re_dev *rdev, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + int payload_sz = 0; + + if (wr->send_flags & IB_SEND_INLINE) + payload_sz = bnxt_re_copy_inline_data(rdev, wr, wqe); + else + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe->sg_list, + wqe->num_sge); + + return payload_sz; +} + +static int bnxt_re_post_send_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp, + struct ib_send_wr *wr) +{ + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + unsigned long flags; + + spin_lock_irqsave(&qp->sq_lock, flags); + memset(&wqe, 0, sizeof(wqe)); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.sq.max_sge) { + dev_err(rdev_to_dev(rdev), + "Limit exceeded for Send SGEs"); + rc = -EINVAL; + goto bad; + } + + payload_sz = bnxt_re_copy_wr_payload(qp->rdev, wr, &wqe); + if (payload_sz < 0) { + rc = -EINVAL; + goto bad; + } + wqe.wr_id = wr->wr_id; + + wqe.type = BNXT_QPLIB_SWQE_TYPE_SEND; + + rc = bnxt_re_build_send_wqe(qp, wr, &wqe); + if (!rc) + rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); +bad: + if (rc) { + dev_err(rdev_to_dev(rdev), + "Post send failed opcode = %#x rc = %d", + wr->opcode, rc); + break; + } + wr = wr->next; + } + bnxt_qplib_post_send_db(&qp->qplib_qp); + spin_unlock_irqrestore(&qp->sq_lock, flags); + return rc; +} + +int bnxt_re_post_send(struct ib_qp *ib_qp, struct ib_send_wr *wr, + struct ib_send_wr **bad_wr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + unsigned long flags; + + spin_lock_irqsave(&qp->sq_lock, flags); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.sq.max_sge) { + dev_err(rdev_to_dev(qp->rdev), + "Limit exceeded for Send SGEs"); + rc = -EINVAL; + goto bad; + } + + payload_sz = bnxt_re_copy_wr_payload(qp->rdev, wr, &wqe); + if (payload_sz < 0) { + rc = -EINVAL; + goto bad; + } + wqe.wr_id = wr->wr_id; + + switch (wr->opcode) { + case IB_WR_SEND: + case IB_WR_SEND_WITH_IMM: + if (ib_qp->qp_type == IB_QPT_GSI) { + rc = bnxt_re_build_qp1_send_v2(qp, wr, &wqe, + payload_sz); + if (rc) + goto bad; + wqe.rawqp1.lflags |= + SQ_SEND_RAWETH_QP1_LFLAGS_ROCE_CRC; + } + switch (wr->send_flags) { + case IB_SEND_IP_CSUM: + wqe.rawqp1.lflags |= + SQ_SEND_RAWETH_QP1_LFLAGS_IP_CHKSUM; + break; + default: + break; + } + /* Fall thru to build the wqe */ + case IB_WR_SEND_WITH_INV: + rc = bnxt_re_build_send_wqe(qp, wr, &wqe); + break; + case IB_WR_RDMA_WRITE: + case IB_WR_RDMA_WRITE_WITH_IMM: + case IB_WR_RDMA_READ: + rc = bnxt_re_build_rdma_wqe(wr, &wqe); + break; + case IB_WR_ATOMIC_CMP_AND_SWP: + case IB_WR_ATOMIC_FETCH_AND_ADD: + rc = bnxt_re_build_atomic_wqe(wr, &wqe); + break; + case IB_WR_RDMA_READ_WITH_INV: + dev_err(rdev_to_dev(qp->rdev), + "RDMA Read with Invalidate is not supported"); + rc = -EINVAL; + goto bad; + case IB_WR_LOCAL_INV: + rc = bnxt_re_build_inv_wqe(wr, &wqe); + break; + case IB_WR_REG_MR: + rc = bnxt_re_build_reg_wqe(reg_wr(wr), &wqe); + break; + default: + /* Unsupported WRs */ + dev_err(rdev_to_dev(qp->rdev), + "WR (%#x) is not supported", wr->opcode); + rc = -EINVAL; + goto bad; + } + if (!rc) + rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); +bad: + if (rc) { + dev_err(rdev_to_dev(qp->rdev), + "post_send failed op:%#x qps = %#x rc = %d\n", + wr->opcode, qp->qplib_qp.state, rc); + *bad_wr = wr; + break; + } + wr = wr->next; + } + bnxt_qplib_post_send_db(&qp->qplib_qp); + spin_unlock_irqrestore(&qp->sq_lock, flags); + + return rc; +} + +static int bnxt_re_post_recv_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp, + struct ib_recv_wr *wr) +{ + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + + memset(&wqe, 0, sizeof(wqe)); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.rq.max_sge) { + dev_err(rdev_to_dev(rdev), + "Limit exceeded for Receive SGEs"); + rc = -EINVAL; + break; + } + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe.sg_list, + wr->num_sge); + wqe.wr_id = wr->wr_id; + wqe.type = BNXT_QPLIB_SWQE_TYPE_RECV; + + rc = bnxt_qplib_post_recv(&qp->qplib_qp, &wqe); + if (rc) + break; + + wr = wr->next; + } + if (!rc) + bnxt_qplib_post_recv_db(&qp->qplib_qp); + return rc; +} + +int bnxt_re_post_recv(struct ib_qp *ib_qp, struct ib_recv_wr *wr, + struct ib_recv_wr **bad_wr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.rq.max_sge) { + dev_err(rdev_to_dev(qp->rdev), + "Limit exceeded for Receive SGEs"); + rc = -EINVAL; + *bad_wr = wr; + break; + } + + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe.sg_list, + wr->num_sge); + wqe.wr_id = wr->wr_id; + wqe.type = BNXT_QPLIB_SWQE_TYPE_RECV; + + if (ib_qp->qp_type == IB_QPT_GSI) + rc = bnxt_re_build_qp1_shadow_qp_recv(qp, wr, &wqe, + payload_sz); + if (!rc) + rc = bnxt_qplib_post_recv(&qp->qplib_qp, &wqe); + if (rc) { + *bad_wr = wr; + break; + } + wr = wr->next; + } + bnxt_qplib_post_recv_db(&qp->qplib_qp); + return rc; +} + +/* Completion Queues */ +int bnxt_re_destroy_cq(struct ib_cq *ib_cq) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + struct bnxt_re_dev *rdev = cq->rdev; + int rc; + + rc = bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW CQ"); + return rc; + } + if (cq->umem && !IS_ERR(cq->umem)) + ib_umem_release(cq->umem); + + if (cq) { + kfree(cq->cql); + kfree(cq); + } + atomic_dec(&rdev->cq_count); + rdev->nq.budget--; + return 0; +} + +struct ib_cq *bnxt_re_create_cq(struct ib_device *ibdev, + const struct ib_cq_init_attr *attr, + struct ib_ucontext *context, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_re_cq *cq = NULL; + int rc, entries; + int cqe = attr->cqe; + + /* Validate CQ fields */ + if (cqe < 1 || cqe > dev_attr->max_cq_wqes) { + dev_err(rdev_to_dev(rdev), "Failed to create CQ -max exceeded"); + return ERR_PTR(-EINVAL); + } + cq = kzalloc(sizeof(*cq), GFP_KERNEL); + if (!cq) + return ERR_PTR(-ENOMEM); + + cq->rdev = rdev; + cq->qplib_cq.cq_handle = (u64)(unsigned long)(&cq->qplib_cq); + + entries = roundup_pow_of_two(cqe + 1); + if (entries > dev_attr->max_cq_wqes + 1) + entries = dev_attr->max_cq_wqes + 1; + + if (context) { + struct bnxt_re_cq_req req; + struct bnxt_re_ucontext *uctx = container_of + (context, + struct bnxt_re_ucontext, + ib_uctx); + if (ib_copy_from_udata(&req, udata, sizeof(req))) { + rc = -EFAULT; + goto fail; + } + + cq->umem = ib_umem_get(context, req.cq_va, + entries * sizeof(struct cq_base), + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(cq->umem)) { + rc = PTR_ERR(cq->umem); + goto fail; + } + cq->qplib_cq.sghead = cq->umem->sg_head.sgl; + cq->qplib_cq.nmap = cq->umem->nmap; + cq->qplib_cq.dpi = uctx->dpi; + } else { + cq->max_cql = min_t(u32, entries, MAX_CQL_PER_POLL); + cq->cql = kcalloc(cq->max_cql, sizeof(struct bnxt_qplib_cqe), + GFP_KERNEL); + if (!cq->cql) { + rc = -ENOMEM; + goto fail; + } + + cq->qplib_cq.dpi = &rdev->dpi_privileged; + cq->qplib_cq.sghead = NULL; + cq->qplib_cq.nmap = 0; + } + cq->qplib_cq.max_wqe = entries; + cq->qplib_cq.cnq_hw_ring_id = rdev->nq.ring_id; + + rc = bnxt_qplib_create_cq(&rdev->qplib_res, &cq->qplib_cq); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW CQ"); + goto fail; + } + + cq->ib_cq.cqe = entries; + cq->cq_period = cq->qplib_cq.period; + rdev->nq.budget++; + + atomic_inc(&rdev->cq_count); + + if (context) { + struct bnxt_re_cq_resp resp; + + resp.cqid = cq->qplib_cq.id; + resp.tail = cq->qplib_cq.hwq.cons; + resp.phase = cq->qplib_cq.period; + resp.rsvd = 0; + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy CQ udata"); + bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); + goto c2fail; + } + } + + return &cq->ib_cq; + +c2fail: + if (context) + ib_umem_release(cq->umem); +fail: + kfree(cq->cql); + kfree(cq); + return ERR_PTR(rc); +} + +static u8 __req_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_REQ_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_REQ_STATUS_BAD_RESPONSE_ERR: + return IB_WC_BAD_RESP_ERR; + case CQ_REQ_STATUS_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_REQ_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_REQ_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_REQ_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_REQ_STATUS_REMOTE_INVALID_REQUEST_ERR: + return IB_WC_REM_INV_REQ_ERR; + case CQ_REQ_STATUS_REMOTE_ACCESS_ERR: + return IB_WC_REM_ACCESS_ERR; + case CQ_REQ_STATUS_REMOTE_OPERATION_ERR: + return IB_WC_REM_OP_ERR; + case CQ_REQ_STATUS_RNR_NAK_RETRY_CNT_ERR: + return IB_WC_RNR_RETRY_EXC_ERR; + case CQ_REQ_STATUS_TRANSPORT_RETRY_CNT_ERR: + return IB_WC_RETRY_EXC_ERR; + case CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } + return 0; +} + +static u8 __rawqp1_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_RES_RAWETH_QP1_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_ACCESS_ERROR: + return IB_WC_LOC_ACCESS_ERR; + case CQ_RES_RAWETH_QP1_STATUS_HW_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_RES_RAWETH_QP1_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_RES_RAWETH_QP1_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + case CQ_RES_RAWETH_QP1_STATUS_HW_FLUSH_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } +} + +static u8 __rc_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_RES_RC_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_RES_RC_STATUS_LOCAL_ACCESS_ERROR: + return IB_WC_LOC_ACCESS_ERR; + case CQ_RES_RC_STATUS_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_RES_RC_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_RES_RC_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_RES_RC_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_RES_RC_STATUS_REMOTE_INVALID_REQUEST_ERR: + return IB_WC_REM_INV_REQ_ERR; + case CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + case CQ_RES_RC_STATUS_HW_FLUSH_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } +} + +static void bnxt_re_process_req_wc(struct ib_wc *wc, struct bnxt_qplib_cqe *cqe) +{ + switch (cqe->type) { + case BNXT_QPLIB_SWQE_TYPE_SEND: + wc->opcode = IB_WC_SEND; + break; + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM: + wc->opcode = IB_WC_SEND; + wc->wc_flags |= IB_WC_WITH_IMM; + break; + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV: + wc->opcode = IB_WC_SEND; + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE: + wc->opcode = IB_WC_RDMA_WRITE; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM: + wc->opcode = IB_WC_RDMA_WRITE; + wc->wc_flags |= IB_WC_WITH_IMM; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_READ: + wc->opcode = IB_WC_RDMA_READ; + break; + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP: + wc->opcode = IB_WC_COMP_SWAP; + break; + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD: + wc->opcode = IB_WC_FETCH_ADD; + break; + case BNXT_QPLIB_SWQE_TYPE_LOCAL_INV: + wc->opcode = IB_WC_LOCAL_INV; + break; + case BNXT_QPLIB_SWQE_TYPE_REG_MR: + wc->opcode = IB_WC_REG_MR; + break; + default: + wc->opcode = IB_WC_SEND; + break; + } + + wc->status = __req_to_ib_wc_status(cqe->status); +} + +static int bnxt_re_check_packet_type(u16 raweth_qp1_flags, + u16 raweth_qp1_flags2) +{ + bool is_udp = false, is_ipv6 = false, is_ipv4 = false; + + /* raweth_qp1_flags Bit 9-6 indicates itype */ + if ((raweth_qp1_flags & CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE) + != CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE) + return -1; + + if (raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_CS_CALC && + raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_L4_CS_CALC) { + is_udp = true; + /* raweth_qp1_flags2 Bit 8 indicates ip_type. 0-v4 1 - v6 */ + (raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_TYPE) ? + (is_ipv6 = true) : (is_ipv4 = true); + return ((is_ipv6) ? + BNXT_RE_ROCEV2_IPV6_PACKET : + BNXT_RE_ROCEV2_IPV4_PACKET); + } else { + return BNXT_RE_ROCE_V1_PACKET; + } +} + +static int bnxt_re_to_ib_nw_type(int nw_type) +{ + u8 nw_hdr_type = 0xFF; + + switch (nw_type) { + case BNXT_RE_ROCE_V1_PACKET: + nw_hdr_type = RDMA_NETWORK_ROCE_V1; + break; + case BNXT_RE_ROCEV2_IPV4_PACKET: + nw_hdr_type = RDMA_NETWORK_IPV4; + break; + case BNXT_RE_ROCEV2_IPV6_PACKET: + nw_hdr_type = RDMA_NETWORK_IPV6; + break; + } + return nw_hdr_type; +} + +static bool bnxt_re_is_loopback_packet(struct bnxt_re_dev *rdev, + void *rq_hdr_buf) +{ + u8 *tmp_buf = NULL; + struct ethhdr *eth_hdr; + u16 eth_type; + bool rc = false; + + tmp_buf = (u8 *)rq_hdr_buf; + /* + * If dest mac is not same as I/F mac, this could be a + * loopback address or multicast address, check whether + * it is a loopback packet + */ + if (!ether_addr_equal(tmp_buf, rdev->netdev->dev_addr)) { + tmp_buf += 4; + /* Check the ether type */ + eth_hdr = (struct ethhdr *)tmp_buf; + eth_type = ntohs(eth_hdr->h_proto); + switch (eth_type) { + case ETH_P_IBOE: + rc = true; + break; + case ETH_P_IP: + case ETH_P_IPV6: { + u32 len; + struct udphdr *udp_hdr; + + len = (eth_type == ETH_P_IP ? sizeof(struct iphdr) : + sizeof(struct ipv6hdr)); + tmp_buf += sizeof(struct ethhdr) + len; + udp_hdr = (struct udphdr *)tmp_buf; + if (ntohs(udp_hdr->dest) == + ROCE_V2_UDP_DPORT) + rc = true; + break; + } + default: + break; + } + } + + return rc; +} + +static int bnxt_re_process_raw_qp_pkt_rx(struct bnxt_re_qp *qp1_qp, + struct bnxt_qplib_cqe *cqe) +{ + struct bnxt_re_dev *rdev = qp1_qp->rdev; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + struct bnxt_re_qp *qp = rdev->qp1_sqp; + struct ib_send_wr *swr; + struct ib_ud_wr udwr; + struct ib_recv_wr rwr; + int pkt_type = 0; + u32 tbl_idx; + void *rq_hdr_buf; + dma_addr_t rq_hdr_buf_map; + dma_addr_t shrq_hdr_buf_map; + u32 offset = 0; + u32 skip_bytes = 0; + struct ib_sge s_sge[2]; + struct ib_sge r_sge[2]; + int rc; + + memset(&udwr, 0, sizeof(udwr)); + memset(&rwr, 0, sizeof(rwr)); + memset(&s_sge, 0, sizeof(s_sge)); + memset(&r_sge, 0, sizeof(r_sge)); + + swr = &udwr.wr; + tbl_idx = cqe->wr_id; + + rq_hdr_buf = qp1_qp->qplib_qp.rq_hdr_buf + + (tbl_idx * qp1_qp->qplib_qp.rq_hdr_buf_size); + rq_hdr_buf_map = bnxt_qplib_get_qp_buf_from_index(&qp1_qp->qplib_qp, + tbl_idx); + + /* Shadow QP header buffer */ + shrq_hdr_buf_map = bnxt_qplib_get_qp_buf_from_index(&qp->qplib_qp, + tbl_idx); + sqp_entry = &rdev->sqp_tbl[tbl_idx]; + + /* Store this cqe */ + memcpy(&sqp_entry->cqe, cqe, sizeof(struct bnxt_qplib_cqe)); + sqp_entry->qp1_qp = qp1_qp; + + /* Find packet type from the cqe */ + + pkt_type = bnxt_re_check_packet_type(cqe->raweth_qp1_flags, + cqe->raweth_qp1_flags2); + if (pkt_type < 0) { + dev_err(rdev_to_dev(rdev), "Invalid packet\n"); + return -EINVAL; + } + + /* Adjust the offset for the user buffer and post in the rq */ + + if (pkt_type == BNXT_RE_ROCEV2_IPV4_PACKET) + offset = 20; + + /* + * QP1 loopback packet has 4 bytes of internal header before + * ether header. Skip these four bytes. + */ + if (bnxt_re_is_loopback_packet(rdev, rq_hdr_buf)) + skip_bytes = 4; + + /* First send SGE . Skip the ether header*/ + s_sge[0].addr = rq_hdr_buf_map + BNXT_QPLIB_MAX_QP1_RQ_ETH_HDR_SIZE + + skip_bytes; + s_sge[0].lkey = 0xFFFFFFFF; + s_sge[0].length = offset ? BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV4 : + BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6; + + /* Second Send SGE */ + s_sge[1].addr = s_sge[0].addr + s_sge[0].length + + BNXT_QPLIB_MAX_QP1_RQ_BDETH_HDR_SIZE; + if (pkt_type != BNXT_RE_ROCE_V1_PACKET) + s_sge[1].addr += 8; + s_sge[1].lkey = 0xFFFFFFFF; + s_sge[1].length = 256; + + /* First recv SGE */ + + r_sge[0].addr = shrq_hdr_buf_map; + r_sge[0].lkey = 0xFFFFFFFF; + r_sge[0].length = 40; + + r_sge[1].addr = sqp_entry->sge.addr + offset; + r_sge[1].lkey = sqp_entry->sge.lkey; + r_sge[1].length = BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6 + 256 - offset; + + /* Create receive work request */ + rwr.num_sge = 2; + rwr.sg_list = r_sge; + rwr.wr_id = tbl_idx; + rwr.next = NULL; + + rc = bnxt_re_post_recv_shadow_qp(rdev, qp, &rwr); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to post Rx buffers to shadow QP"); + return -ENOMEM; + } + + swr->num_sge = 2; + swr->sg_list = s_sge; + swr->wr_id = tbl_idx; + swr->opcode = IB_WR_SEND; + swr->next = NULL; + + udwr.ah = &rdev->sqp_ah->ib_ah; + udwr.remote_qpn = rdev->qp1_sqp->qplib_qp.id; + udwr.remote_qkey = rdev->qp1_sqp->qplib_qp.qkey; + + /* post data received in the send queue */ + rc = bnxt_re_post_send_shadow_qp(rdev, qp, swr); + + return 0; +} + +static void bnxt_re_process_res_rawqp1_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rawqp1_to_ib_wc_status(cqe->status); + wc->wc_flags |= IB_WC_GRH; +} + +static void bnxt_re_process_res_rc_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rc_to_ib_wc_status(cqe->status); + + if (cqe->flags & CQ_RES_RC_FLAGS_IMM) + wc->wc_flags |= IB_WC_WITH_IMM; + if (cqe->flags & CQ_RES_RC_FLAGS_INV) + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + if ((cqe->flags & (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) == + (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) + wc->opcode = IB_WC_RECV_RDMA_WITH_IMM; +} + +static void bnxt_re_process_res_shadow_qp_wc(struct bnxt_re_qp *qp, + struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + u32 tbl_idx; + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_re_qp *qp1_qp = NULL; + struct bnxt_qplib_cqe *orig_cqe = NULL; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + int nw_type; + + tbl_idx = cqe->wr_id; + + sqp_entry = &rdev->sqp_tbl[tbl_idx]; + qp1_qp = sqp_entry->qp1_qp; + orig_cqe = &sqp_entry->cqe; + + wc->wr_id = sqp_entry->wrid; + wc->byte_len = orig_cqe->length; + wc->qp = &qp1_qp->ib_qp; + + wc->ex.imm_data = orig_cqe->immdata; + wc->src_qp = orig_cqe->src_qp; + memcpy(wc->smac, orig_cqe->smac, ETH_ALEN); + wc->port_num = 1; + wc->vendor_err = orig_cqe->status; + + wc->opcode = IB_WC_RECV; + wc->status = __rawqp1_to_ib_wc_status(orig_cqe->status); + wc->wc_flags |= IB_WC_GRH; + + nw_type = bnxt_re_check_packet_type(orig_cqe->raweth_qp1_flags, + orig_cqe->raweth_qp1_flags2); + if (nw_type >= 0) { + wc->network_hdr_type = bnxt_re_to_ib_nw_type(nw_type); + wc->wc_flags |= IB_WC_WITH_NETWORK_HDR_TYPE; + } +} + +static void bnxt_re_process_res_ud_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rc_to_ib_wc_status(cqe->status); + + if (cqe->flags & CQ_RES_RC_FLAGS_IMM) + wc->wc_flags |= IB_WC_WITH_IMM; + if (cqe->flags & CQ_RES_RC_FLAGS_INV) + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + if ((cqe->flags & (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) == + (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) + wc->opcode = IB_WC_RECV_RDMA_WITH_IMM; +} + +int bnxt_re_poll_cq(struct ib_cq *ib_cq, int num_entries, struct ib_wc *wc) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + struct bnxt_re_qp *qp; + struct bnxt_qplib_cqe *cqe; + int i, ncqe, budget; + u32 tbl_idx; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + unsigned long flags; + + spin_lock_irqsave(&cq->cq_lock, flags); + budget = min_t(u32, num_entries, cq->max_cql); + if (!cq->cql) { + dev_err(rdev_to_dev(cq->rdev), "POLL CQ : no CQL to use"); + goto exit; + } + cqe = &cq->cql[0]; + while (budget) { + ncqe = bnxt_qplib_poll_cq(&cq->qplib_cq, cqe, budget); + if (!ncqe) + break; + + for (i = 0; i < ncqe; i++, cqe++) { + /* Transcribe each qplib_wqe back to ib_wc */ + memset(wc, 0, sizeof(*wc)); + + wc->wr_id = cqe->wr_id; + wc->byte_len = cqe->length; + qp = container_of + ((struct bnxt_qplib_qp *) + (unsigned long)(cqe->qp_handle), + struct bnxt_re_qp, qplib_qp); + if (!qp) { + dev_err(rdev_to_dev(cq->rdev), + "POLL CQ : bad QP handle"); + continue; + } + wc->qp = &qp->ib_qp; + wc->ex.imm_data = cqe->immdata; + wc->src_qp = cqe->src_qp; + memcpy(wc->smac, cqe->smac, ETH_ALEN); + wc->port_num = 1; + wc->vendor_err = cqe->status; + + switch (cqe->opcode) { + case CQ_BASE_CQE_TYPE_REQ: + if (qp->qplib_qp.id == + qp->rdev->qp1_sqp->qplib_qp.id) { + /* Handle this completion with + * the stored completion + */ + memset(wc, 0, sizeof(*wc)); + continue; + } + bnxt_re_process_req_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + if (!cqe->status) { + int rc = 0; + + rc = bnxt_re_process_raw_qp_pkt_rx + (qp, cqe); + if (!rc) { + memset(wc, 0, sizeof(*wc)); + continue; + } + cqe->status = -1; + } + /* Errors need not be looped back. + * But change the wr_id to the one + * stored in the table + */ + tbl_idx = cqe->wr_id; + sqp_entry = &cq->rdev->sqp_tbl[tbl_idx]; + wc->wr_id = sqp_entry->wrid; + bnxt_re_process_res_rawqp1_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_RC: + bnxt_re_process_res_rc_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_UD: + if (qp->qplib_qp.id == + qp->rdev->qp1_sqp->qplib_qp.id) { + /* Handle this completion with + * the stored completion + */ + if (cqe->status) { + continue; + } else { + bnxt_re_process_res_shadow_qp_wc + (qp, wc, cqe); + break; + } + } + bnxt_re_process_res_ud_wc(wc, cqe); + break; + default: + dev_err(rdev_to_dev(cq->rdev), + "POLL CQ : type 0x%x not handled", + cqe->opcode); + continue; + } + wc++; + budget--; + } + } +exit: + spin_unlock_irqrestore(&cq->cq_lock, flags); + return num_entries - budget; +} + +int bnxt_re_req_notify_cq(struct ib_cq *ib_cq, + enum ib_cq_notify_flags ib_cqn_flags) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + int type = 0; + + /* Trigger on the very next completion */ + if (ib_cqn_flags & IB_CQ_NEXT_COMP) + type = DBR_DBR_TYPE_CQ_ARMALL; + /* Trigger on the next solicited completion */ + else if (ib_cqn_flags & IB_CQ_SOLICITED) + type = DBR_DBR_TYPE_CQ_ARMSE; + + bnxt_qplib_req_notify_cq(&cq->qplib_cq, type); + + return 0; +} + +/* Memory Regions */ +struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr; + u64 pbl = 0; + int rc; + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + + /* Allocate and register 0 as the address */ + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) + goto fail; + + mr->qplib_mr.hwq.level = PBL_LVL_MAX; + mr->qplib_mr.total_size = -1; /* Infinte length */ + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, &pbl, 0, false); + if (rc) + goto fail_mr; + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + if (mr_access_flags & (IB_ACCESS_REMOTE_WRITE | IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_ATOMIC)) + mr->ib_mr.rkey = mr->ib_mr.lkey; + atomic_inc(&rdev->mr_count); + + return &mr->ib_mr; + +fail_mr: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +fail: + kfree(mr); + return ERR_PTR(rc); +} + +int bnxt_re_dereg_mr(struct ib_mr *ib_mr) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + struct bnxt_re_dev *rdev = mr->rdev; + int rc = 0; + + if (mr->npages && mr->pages) { + rc = bnxt_qplib_free_fast_reg_page_list(&rdev->qplib_res, + &mr->qplib_frpl); + kfree(mr->pages); + mr->npages = 0; + mr->pages = NULL; + } + rc = bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); + + if (!IS_ERR(mr->ib_umem) && mr->ib_umem) + ib_umem_release(mr->ib_umem); + + kfree(mr); + atomic_dec(&rdev->mr_count); + return rc; +} + +static int bnxt_re_set_page(struct ib_mr *ib_mr, u64 addr) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + + if (unlikely(mr->npages == mr->qplib_frpl.max_pg_ptrs)) + return -ENOMEM; + + mr->pages[mr->npages++] = addr; + return 0; +} + +int bnxt_re_map_mr_sg(struct ib_mr *ib_mr, struct scatterlist *sg, int sg_nents, + unsigned int *sg_offset) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + + mr->npages = 0; + return ib_sg_to_pages(ib_mr, sg, sg_nents, sg_offset, bnxt_re_set_page); +} + +struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type type, + u32 max_num_sg) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr = NULL; + int rc; + + if (type != IB_MR_TYPE_MEM_REG) { + dev_dbg(rdev_to_dev(rdev), "MR type 0x%x not supported", type); + return ERR_PTR(-EINVAL); + } + if (max_num_sg > MAX_PBL_LVL_1_PGS) + return ERR_PTR(-EINVAL); + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = BNXT_QPLIB_FR_PMR; + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) + goto fail; + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + mr->ib_mr.rkey = mr->ib_mr.lkey; + + mr->pages = kcalloc(max_num_sg, sizeof(u64), GFP_KERNEL); + if (!mr->pages) { + rc = -ENOMEM; + goto fail; + } + rc = bnxt_qplib_alloc_fast_reg_page_list(&rdev->qplib_res, + &mr->qplib_frpl, max_num_sg); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate HW FR page list"); + goto fail_mr; + } + + atomic_inc(&rdev->mr_count); + return &mr->ib_mr; + +fail_mr: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +fail: + kfree(mr->pages); + kfree(mr); + return ERR_PTR(rc); +} + +/* Fast Memory Regions */ +struct ib_fmr *bnxt_re_alloc_fmr(struct ib_pd *ib_pd, int mr_access_flags, + struct ib_fmr_attr *fmr_attr) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_fmr *fmr; + int rc; + + if (fmr_attr->max_pages > MAX_PBL_LVL_2_PGS || + fmr_attr->max_maps > rdev->dev_attr.max_map_per_fmr) { + dev_err(rdev_to_dev(rdev), "Allocate FMR exceeded Max limit"); + return ERR_PTR(-ENOMEM); + } + fmr = kzalloc(sizeof(*fmr), GFP_KERNEL); + if (!fmr) + return ERR_PTR(-ENOMEM); + + fmr->rdev = rdev; + fmr->qplib_fmr.pd = &pd->qplib_pd; + fmr->qplib_fmr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &fmr->qplib_fmr); + if (rc) + goto fail; + + fmr->qplib_fmr.flags = __from_ib_access_flags(mr_access_flags); + fmr->ib_fmr.lkey = fmr->qplib_fmr.lkey; + fmr->ib_fmr.rkey = fmr->ib_fmr.lkey; + + atomic_inc(&rdev->mr_count); + return &fmr->ib_fmr; +fail: + kfree(fmr); + return ERR_PTR(rc); +} + +int bnxt_re_map_phys_fmr(struct ib_fmr *ib_fmr, u64 *page_list, int list_len, + u64 iova) +{ + struct bnxt_re_fmr *fmr = container_of(ib_fmr, struct bnxt_re_fmr, + ib_fmr); + struct bnxt_re_dev *rdev = fmr->rdev; + int rc; + + fmr->qplib_fmr.va = iova; + fmr->qplib_fmr.total_size = list_len * PAGE_SIZE; + + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &fmr->qplib_fmr, page_list, + list_len, true); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to map FMR for lkey = 0x%x!", + fmr->ib_fmr.lkey); + return rc; +} + +int bnxt_re_unmap_fmr(struct list_head *fmr_list) +{ + struct bnxt_re_dev *rdev; + struct bnxt_re_fmr *fmr; + struct ib_fmr *ib_fmr; + int rc = 0; + + /* Validate each FMRs inside the fmr_list */ + list_for_each_entry(ib_fmr, fmr_list, list) { + fmr = container_of(ib_fmr, struct bnxt_re_fmr, ib_fmr); + rdev = fmr->rdev; + + if (rdev) { + rc = bnxt_qplib_dereg_mrw(&rdev->qplib_res, + &fmr->qplib_fmr, true); + if (rc) + break; + } + } + return rc; +} + +int bnxt_re_dealloc_fmr(struct ib_fmr *ib_fmr) +{ + struct bnxt_re_fmr *fmr = container_of(ib_fmr, struct bnxt_re_fmr, + ib_fmr); + struct bnxt_re_dev *rdev = fmr->rdev; + int rc; + + rc = bnxt_qplib_free_mrw(&rdev->qplib_res, &fmr->qplib_fmr); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to free FMR"); + + kfree(fmr); + atomic_dec(&rdev->mr_count); + return rc; +} + +/* uverbs */ +struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *ib_pd, u64 start, u64 length, + u64 virt_addr, int mr_access_flags, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr; + struct ib_umem *umem; + u64 *pbl_tbl, *pbl_tbl_orig; + int i, umem_pgs, pages, page_shift, rc; + struct scatterlist *sg; + int entry; + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_MR; + + umem = ib_umem_get(ib_pd->uobject->context, start, length, + mr_access_flags, 0); + if (IS_ERR(umem)) { + dev_err(rdev_to_dev(rdev), "Failed to get umem"); + rc = -EFAULT; + goto free_mr; + } + mr->ib_umem = umem; + + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to allocate MR"); + goto release_umem; + } + /* The fixed portion of the rkey is the same as the lkey */ + mr->ib_mr.rkey = mr->qplib_mr.rkey; + + mr->qplib_mr.va = virt_addr; + umem_pgs = ib_umem_page_count(umem); + if (!umem_pgs) { + dev_err(rdev_to_dev(rdev), "umem is invalid!"); + rc = -EINVAL; + goto free_mrw; + } + mr->qplib_mr.total_size = length; + + pbl_tbl = kcalloc(umem_pgs, sizeof(u64 *), GFP_KERNEL); + if (!pbl_tbl) { + rc = -EINVAL; + goto free_mrw; + } + pbl_tbl_orig = pbl_tbl; + + page_shift = ilog2(umem->page_size); + if (umem->hugetlb) { + dev_err(rdev_to_dev(rdev), "umem hugetlb not supported!"); + rc = -EFAULT; + goto fail; + } + if (umem->page_size != PAGE_SIZE) { + dev_err(rdev_to_dev(rdev), "umem page size unsupported!"); + rc = -EFAULT; + goto fail; + } + /* Map umem buf ptrs to the PBL */ + for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { + pages = sg_dma_len(sg) >> page_shift; + for (i = 0; i < pages; i++, pbl_tbl++) + *pbl_tbl = sg_dma_address(sg) + (i << page_shift); + } + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, pbl_tbl_orig, + umem_pgs, false); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to register user MR"); + goto fail; + } + + kfree(pbl_tbl_orig); + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + mr->ib_mr.rkey = mr->qplib_mr.lkey; + atomic_inc(&rdev->mr_count); + + return &mr->ib_mr; +fail: + kfree(pbl_tbl_orig); +free_mrw: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +release_umem: + ib_umem_release(umem); +free_mr: + kfree(mr); + return ERR_PTR(rc); +} + +struct ib_ucontext *bnxt_re_alloc_ucontext(struct ib_device *ibdev, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_re_uctx_resp resp; + struct bnxt_re_ucontext *uctx; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + int rc; + + dev_dbg(rdev_to_dev(rdev), "ABI version requested %d", + ibdev->uverbs_abi_ver); + + if (ibdev->uverbs_abi_ver != BNXT_RE_ABI_VERSION) { + dev_dbg(rdev_to_dev(rdev), " is different from the device %d ", + BNXT_RE_ABI_VERSION); + return ERR_PTR(-EPERM); + } + + uctx = kzalloc(sizeof(*uctx), GFP_KERNEL); + if (!uctx) + return ERR_PTR(-ENOMEM); + + uctx->rdev = rdev; + + uctx->shpg = (void *)__get_free_page(GFP_KERNEL); + if (!uctx->shpg) { + rc = -ENOMEM; + goto fail; + } + spin_lock_init(&uctx->sh_lock); + + resp.dev_id = rdev->en_dev->pdev->devfn; /*Temp, Use idr_alloc instead*/ + resp.max_qp = rdev->qplib_ctx.qpc_count; + resp.pg_size = PAGE_SIZE; + resp.cqe_sz = sizeof(struct cq_base); + resp.max_cqd = dev_attr->max_cq_wqes; + resp.rsvd = 0; + + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy user context"); + rc = -EFAULT; + goto cfail; + } + + return &uctx->ib_uctx; +cfail: + free_page((unsigned long)uctx->shpg); + uctx->shpg = NULL; +fail: + kfree(uctx); + return ERR_PTR(rc); +} + +int bnxt_re_dealloc_ucontext(struct ib_ucontext *ib_uctx) +{ + struct bnxt_re_ucontext *uctx = container_of(ib_uctx, + struct bnxt_re_ucontext, + ib_uctx); + if (uctx->shpg) + free_page((unsigned long)uctx->shpg); + kfree(uctx); + return 0; +} + +/* Helper function to mmap the virtual memory from user app */ +int bnxt_re_mmap(struct ib_ucontext *ib_uctx, struct vm_area_struct *vma) +{ + struct bnxt_re_ucontext *uctx = container_of(ib_uctx, + struct bnxt_re_ucontext, + ib_uctx); + struct bnxt_re_dev *rdev = uctx->rdev; + u64 pfn; + + if (vma->vm_end - vma->vm_start != PAGE_SIZE) + return -EINVAL; + + if (vma->vm_pgoff) { + vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); + if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, + PAGE_SIZE, vma->vm_page_prot)) { + dev_err(rdev_to_dev(rdev), "Failed to map DPI"); + return -EAGAIN; + } + } else { + pfn = virt_to_phys(uctx->shpg) >> PAGE_SHIFT; + if (remap_pfn_range(vma, vma->vm_start, + pfn, PAGE_SIZE, vma->vm_page_prot)) { + dev_err(rdev_to_dev(rdev), + "Failed to map shared page"); + return -EAGAIN; + } + } + + return 0; +} diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.h b/drivers/infiniband/hw/bnxt_re/ib_verbs.h new file mode 100644 index 000000000000..b4084c252f06 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.h @@ -0,0 +1,197 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: IB Verbs interpreter (header) + */ + +#ifndef __BNXT_RE_IB_VERBS_H__ +#define __BNXT_RE_IB_VERBS_H__ + +struct bnxt_re_gid_ctx { + u32 idx; + u32 refcnt; +}; + +struct bnxt_re_pd { + struct bnxt_re_dev *rdev; + struct ib_pd ib_pd; + struct bnxt_qplib_pd qplib_pd; + struct bnxt_qplib_dpi dpi; +}; + +struct bnxt_re_ah { + struct bnxt_re_dev *rdev; + struct ib_ah ib_ah; + struct bnxt_qplib_ah qplib_ah; +}; + +struct bnxt_re_qp { + struct list_head list; + struct bnxt_re_dev *rdev; + struct ib_qp ib_qp; + spinlock_t sq_lock; /* protect sq */ + struct bnxt_qplib_qp qplib_qp; + struct ib_umem *sumem; + struct ib_umem *rumem; + /* QP1 */ + u32 send_psn; + struct ib_ud_header qp1_hdr; +}; + +struct bnxt_re_cq { + struct bnxt_re_dev *rdev; + spinlock_t cq_lock; /* protect cq */ + u16 cq_count; + u16 cq_period; + struct ib_cq ib_cq; + struct bnxt_qplib_cq qplib_cq; + struct bnxt_qplib_cqe *cql; +#define MAX_CQL_PER_POLL 1024 + u32 max_cql; + struct ib_umem *umem; +}; + +struct bnxt_re_mr { + struct bnxt_re_dev *rdev; + struct ib_mr ib_mr; + struct ib_umem *ib_umem; + struct bnxt_qplib_mrw qplib_mr; + u32 npages; + u64 *pages; + struct bnxt_qplib_frpl qplib_frpl; +}; + +struct bnxt_re_frpl { + struct bnxt_re_dev *rdev; + struct bnxt_qplib_frpl qplib_frpl; + u64 *page_list; +}; + +struct bnxt_re_fmr { + struct bnxt_re_dev *rdev; + struct ib_fmr ib_fmr; + struct bnxt_qplib_mrw qplib_fmr; +}; + +struct bnxt_re_mw { + struct bnxt_re_dev *rdev; + struct ib_mw ib_mw; + struct bnxt_qplib_mrw qplib_mw; +}; + +struct bnxt_re_ucontext { + struct bnxt_re_dev *rdev; + struct ib_ucontext ib_uctx; + struct bnxt_qplib_dpi *dpi; + void *shpg; + spinlock_t sh_lock; /* protect shpg */ +}; + +struct net_device *bnxt_re_get_netdev(struct ib_device *ibdev, u8 port_num); + +int bnxt_re_query_device(struct ib_device *ibdev, + struct ib_device_attr *ib_attr, + struct ib_udata *udata); +int bnxt_re_modify_device(struct ib_device *ibdev, + int device_modify_mask, + struct ib_device_modify *device_modify); +int bnxt_re_query_port(struct ib_device *ibdev, u8 port_num, + struct ib_port_attr *port_attr); +int bnxt_re_modify_port(struct ib_device *ibdev, u8 port_num, + int port_modify_mask, + struct ib_port_modify *port_modify); +int bnxt_re_get_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable); +int bnxt_re_query_pkey(struct ib_device *ibdev, u8 port_num, + u16 index, u16 *pkey); +int bnxt_re_del_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, void **context); +int bnxt_re_add_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, const union ib_gid *gid, + const struct ib_gid_attr *attr, void **context); +int bnxt_re_query_gid(struct ib_device *ibdev, u8 port_num, + int index, union ib_gid *gid); +enum rdma_link_layer bnxt_re_get_link_layer(struct ib_device *ibdev, + u8 port_num); +struct ib_pd *bnxt_re_alloc_pd(struct ib_device *ibdev, + struct ib_ucontext *context, + struct ib_udata *udata); +int bnxt_re_dealloc_pd(struct ib_pd *pd); +struct ib_ah *bnxt_re_create_ah(struct ib_pd *pd, + struct ib_ah_attr *ah_attr, + struct ib_udata *udata); +int bnxt_re_modify_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr); +int bnxt_re_query_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr); +int bnxt_re_destroy_ah(struct ib_ah *ah); +struct ib_qp *bnxt_re_create_qp(struct ib_pd *pd, + struct ib_qp_init_attr *qp_init_attr, + struct ib_udata *udata); +int bnxt_re_modify_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_udata *udata); +int bnxt_re_query_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); +int bnxt_re_destroy_qp(struct ib_qp *qp); +int bnxt_re_post_send(struct ib_qp *qp, struct ib_send_wr *send_wr, + struct ib_send_wr **bad_send_wr); +int bnxt_re_post_recv(struct ib_qp *qp, struct ib_recv_wr *recv_wr, + struct ib_recv_wr **bad_recv_wr); +struct ib_cq *bnxt_re_create_cq(struct ib_device *ibdev, + const struct ib_cq_init_attr *attr, + struct ib_ucontext *context, + struct ib_udata *udata); +int bnxt_re_destroy_cq(struct ib_cq *cq); +int bnxt_re_poll_cq(struct ib_cq *cq, int num_entries, struct ib_wc *wc); +int bnxt_re_req_notify_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags); +struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *pd, int mr_access_flags); + +int bnxt_re_map_mr_sg(struct ib_mr *ib_mr, struct scatterlist *sg, int sg_nents, + unsigned int *sg_offset); +struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type mr_type, + u32 max_num_sg); +int bnxt_re_dereg_mr(struct ib_mr *mr); +struct ib_fmr *bnxt_re_alloc_fmr(struct ib_pd *pd, int mr_access_flags, + struct ib_fmr_attr *fmr_attr); +int bnxt_re_map_phys_fmr(struct ib_fmr *fmr, u64 *page_list, int list_len, + u64 iova); +int bnxt_re_unmap_fmr(struct list_head *fmr_list); +int bnxt_re_dealloc_fmr(struct ib_fmr *fmr); +struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, + u64 virt_addr, int mr_access_flags, + struct ib_udata *udata); +struct ib_ucontext *bnxt_re_alloc_ucontext(struct ib_device *ibdev, + struct ib_udata *udata); +int bnxt_re_dealloc_ucontext(struct ib_ucontext *context); +int bnxt_re_mmap(struct ib_ucontext *context, struct vm_area_struct *vma); +#endif /* __BNXT_RE_IB_VERBS_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/main.c b/drivers/infiniband/hw/bnxt_re/main.c new file mode 100644 index 000000000000..6b9f1178050f --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/main.c @@ -0,0 +1,1315 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Main component of the bnxt_re driver + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include "bnxt_ulp.h" +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_fp.h" +#include "qplib_rcfw.h" +#include "bnxt_re.h" +#include "ib_verbs.h" +#include +#include "bnxt.h" +static char version[] = + BNXT_RE_DESC " v" ROCE_DRV_MODULE_VERSION "\n"; + +MODULE_AUTHOR("Eddie Wai "); +MODULE_DESCRIPTION(BNXT_RE_DESC " Driver"); +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_VERSION(ROCE_DRV_MODULE_VERSION); + +/* globals */ +static struct list_head bnxt_re_dev_list = LIST_HEAD_INIT(bnxt_re_dev_list); +/* Mutex to protect the list of bnxt_re devices added */ +static DEFINE_MUTEX(bnxt_re_dev_lock); +static struct workqueue_struct *bnxt_re_wq; + +/* for handling bnxt_en callbacks later */ +static void bnxt_re_stop(void *p) +{ +} + +static void bnxt_re_start(void *p) +{ +} + +static void bnxt_re_sriov_config(void *p, int num_vfs) +{ +} + +static struct bnxt_ulp_ops bnxt_re_ulp_ops = { + .ulp_async_notifier = NULL, + .ulp_stop = bnxt_re_stop, + .ulp_start = bnxt_re_start, + .ulp_sriov_config = bnxt_re_sriov_config +}; + +/* RoCE -> Net driver */ + +/* Driver registration routines used to let the networking driver (bnxt_en) + * to know that the RoCE driver is now installed + */ +static int bnxt_re_unregister_netdev(struct bnxt_re_dev *rdev, bool lock_wait) +{ + struct bnxt_en_dev *en_dev; + int rc; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + /* Acquire rtnl lock if it is not invokded from netdev event */ + if (lock_wait) + rtnl_lock(); + + rc = en_dev->en_ops->bnxt_unregister_device(rdev->en_dev, + BNXT_ROCE_ULP); + if (lock_wait) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_register_netdev(struct bnxt_re_dev *rdev) +{ + struct bnxt_en_dev *en_dev; + int rc = 0; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + rtnl_lock(); + rc = en_dev->en_ops->bnxt_register_device(en_dev, BNXT_ROCE_ULP, + &bnxt_re_ulp_ops, rdev); + rtnl_unlock(); + return rc; +} + +static int bnxt_re_free_msix(struct bnxt_re_dev *rdev, bool lock_wait) +{ + struct bnxt_en_dev *en_dev; + int rc; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + if (lock_wait) + rtnl_lock(); + + rc = en_dev->en_ops->bnxt_free_msix(rdev->en_dev, BNXT_ROCE_ULP); + + if (lock_wait) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_request_msix(struct bnxt_re_dev *rdev) +{ + int rc = 0, num_msix_want = BNXT_RE_MIN_MSIX, num_msix_got; + struct bnxt_en_dev *en_dev; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + rtnl_lock(); + num_msix_got = en_dev->en_ops->bnxt_request_msix(en_dev, BNXT_ROCE_ULP, + rdev->msix_entries, + num_msix_want); + if (num_msix_got < BNXT_RE_MIN_MSIX) { + rc = -EINVAL; + goto done; + } + if (num_msix_got != num_msix_want) { + dev_warn(rdev_to_dev(rdev), + "Requested %d MSI-X vectors, got %d\n", + num_msix_want, num_msix_got); + } + rdev->num_msix = num_msix_got; +done: + rtnl_unlock(); + return rc; +} + +static void bnxt_re_init_hwrm_hdr(struct bnxt_re_dev *rdev, struct input *hdr, + u16 opcd, u16 crid, u16 trid) +{ + hdr->req_type = cpu_to_le16(opcd); + hdr->cmpl_ring = cpu_to_le16(crid); + hdr->target_id = cpu_to_le16(trid); +} + +static void bnxt_re_fill_fw_msg(struct bnxt_fw_msg *fw_msg, void *msg, + int msg_len, void *resp, int resp_max_len, + int timeout) +{ + fw_msg->msg = msg; + fw_msg->msg_len = msg_len; + fw_msg->resp = resp; + fw_msg->resp_max_len = resp_max_len; + fw_msg->timeout = timeout; +} + +static int bnxt_re_net_ring_free(struct bnxt_re_dev *rdev, u16 fw_ring_id, + bool lock_wait) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_ring_free_input req = {0}; + struct hwrm_ring_free_output resp; + struct bnxt_fw_msg fw_msg; + bool do_unlock = false; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + if (lock_wait) { + rtnl_lock(); + do_unlock = true; + } + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_RING_FREE, -1, -1); + req.ring_type = RING_ALLOC_REQ_RING_TYPE_CMPL; + req.ring_id = cpu_to_le16(fw_ring_id); + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to free HW ring:%d :%#x", req.ring_id, rc); + if (do_unlock) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_ring_alloc(struct bnxt_re_dev *rdev, dma_addr_t *dma_arr, + int pages, int type, u32 ring_mask, + u32 map_index, u16 *fw_ring_id) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_ring_alloc_input req = {0}; + struct hwrm_ring_alloc_output resp; + struct bnxt_fw_msg fw_msg; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + rtnl_lock(); + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_RING_ALLOC, -1, -1); + req.enables = 0; + req.page_tbl_addr = cpu_to_le64(dma_arr[0]); + if (pages > 1) { + /* Page size is in log2 units */ + req.page_size = BNXT_PAGE_SHIFT; + req.page_tbl_depth = 1; + } + req.fbo = 0; + /* Association of ring index with doorbell index and MSIX number */ + req.logical_id = cpu_to_le16(map_index); + req.length = cpu_to_le32(ring_mask + 1); + req.ring_type = RING_ALLOC_REQ_RING_TYPE_CMPL; + req.int_mode = RING_ALLOC_REQ_INT_MODE_MSIX; + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (!rc) + *fw_ring_id = le16_to_cpu(resp.ring_id); + + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_stats_ctx_free(struct bnxt_re_dev *rdev, + u32 fw_stats_ctx_id, bool lock_wait) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_stat_ctx_free_input req = {0}; + struct bnxt_fw_msg fw_msg; + bool do_unlock = false; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + if (lock_wait) { + rtnl_lock(); + do_unlock = true; + } + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_STAT_CTX_FREE, -1, -1); + req.stat_ctx_id = cpu_to_le32(fw_stats_ctx_id); + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&req, + sizeof(req), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to free HW stats context %#x", rc); + + if (do_unlock) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_stats_ctx_alloc(struct bnxt_re_dev *rdev, + dma_addr_t dma_map, + u32 *fw_stats_ctx_id) +{ + struct hwrm_stat_ctx_alloc_output resp = {0}; + struct hwrm_stat_ctx_alloc_input req = {0}; + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct bnxt_fw_msg fw_msg; + int rc = -EINVAL; + + *fw_stats_ctx_id = INVALID_STATS_CTX_ID; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + rtnl_lock(); + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_STAT_CTX_ALLOC, -1, -1); + req.update_period_ms = cpu_to_le32(1000); + req.stats_dma_addr = cpu_to_le64(dma_map); + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (!rc) + *fw_stats_ctx_id = le32_to_cpu(resp.stat_ctx_id); + + rtnl_unlock(); + return rc; +} + +/* Device */ + +static bool is_bnxt_re_dev(struct net_device *netdev) +{ + struct ethtool_drvinfo drvinfo; + + if (netdev->ethtool_ops && netdev->ethtool_ops->get_drvinfo) { + memset(&drvinfo, 0, sizeof(drvinfo)); + netdev->ethtool_ops->get_drvinfo(netdev, &drvinfo); + + if (strcmp(drvinfo.driver, "bnxt_en")) + return false; + return true; + } + return false; +} + +static struct bnxt_re_dev *bnxt_re_from_netdev(struct net_device *netdev) +{ + struct bnxt_re_dev *rdev; + + rcu_read_lock(); + list_for_each_entry_rcu(rdev, &bnxt_re_dev_list, list) { + if (rdev->netdev == netdev) { + rcu_read_unlock(); + return rdev; + } + } + rcu_read_unlock(); + return NULL; +} + +static void bnxt_re_dev_unprobe(struct net_device *netdev, + struct bnxt_en_dev *en_dev) +{ + dev_put(netdev); + module_put(en_dev->pdev->driver->driver.owner); +} + +static struct bnxt_en_dev *bnxt_re_dev_probe(struct net_device *netdev) +{ + struct bnxt *bp = netdev_priv(netdev); + struct bnxt_en_dev *en_dev; + struct pci_dev *pdev; + + /* Call bnxt_en's RoCE probe via indirect API */ + if (!bp->ulp_probe) + return ERR_PTR(-EINVAL); + + en_dev = bp->ulp_probe(netdev); + if (IS_ERR(en_dev)) + return en_dev; + + pdev = en_dev->pdev; + if (!pdev) + return ERR_PTR(-EINVAL); + + if (!(en_dev->flags & BNXT_EN_FLAG_ROCE_CAP)) { + dev_dbg(&pdev->dev, + "%s: probe error: RoCE is not supported on this device", + ROCE_DRV_MODULE_NAME); + return ERR_PTR(-ENODEV); + } + + /* Bump net device reference count */ + if (!try_module_get(pdev->driver->driver.owner)) + return ERR_PTR(-ENODEV); + + dev_hold(netdev); + + return en_dev; +} + +static void bnxt_re_unregister_ib(struct bnxt_re_dev *rdev) +{ + ib_unregister_device(&rdev->ibdev); +} + +static int bnxt_re_register_ib(struct bnxt_re_dev *rdev) +{ + struct ib_device *ibdev = &rdev->ibdev; + + /* ib device init */ + ibdev->owner = THIS_MODULE; + ibdev->node_type = RDMA_NODE_IB_CA; + strlcpy(ibdev->name, "bnxt_re%d", IB_DEVICE_NAME_MAX); + strlcpy(ibdev->node_desc, BNXT_RE_DESC " HCA", + strlen(BNXT_RE_DESC) + 5); + ibdev->phys_port_cnt = 1; + + bnxt_qplib_get_guid(rdev->netdev->dev_addr, (u8 *)&ibdev->node_guid); + + ibdev->num_comp_vectors = 1; + ibdev->dma_device = &rdev->en_dev->pdev->dev; + ibdev->local_dma_lkey = BNXT_QPLIB_RSVD_LKEY; + + /* User space */ + ibdev->uverbs_abi_ver = BNXT_RE_ABI_VERSION; + ibdev->uverbs_cmd_mask = + (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | + (1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) | + (1ull << IB_USER_VERBS_CMD_QUERY_PORT) | + (1ull << IB_USER_VERBS_CMD_ALLOC_PD) | + (1ull << IB_USER_VERBS_CMD_DEALLOC_PD) | + (1ull << IB_USER_VERBS_CMD_REG_MR) | + (1ull << IB_USER_VERBS_CMD_REREG_MR) | + (1ull << IB_USER_VERBS_CMD_DEREG_MR) | + (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) | + (1ull << IB_USER_VERBS_CMD_CREATE_CQ) | + (1ull << IB_USER_VERBS_CMD_RESIZE_CQ) | + (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) | + (1ull << IB_USER_VERBS_CMD_CREATE_QP) | + (1ull << IB_USER_VERBS_CMD_MODIFY_QP) | + (1ull << IB_USER_VERBS_CMD_QUERY_QP) | + (1ull << IB_USER_VERBS_CMD_DESTROY_QP) | + (1ull << IB_USER_VERBS_CMD_CREATE_SRQ) | + (1ull << IB_USER_VERBS_CMD_MODIFY_SRQ) | + (1ull << IB_USER_VERBS_CMD_QUERY_SRQ) | + (1ull << IB_USER_VERBS_CMD_DESTROY_SRQ) | + (1ull << IB_USER_VERBS_CMD_CREATE_AH) | + (1ull << IB_USER_VERBS_CMD_MODIFY_AH) | + (1ull << IB_USER_VERBS_CMD_QUERY_AH) | + (1ull << IB_USER_VERBS_CMD_DESTROY_AH); + /* POLL_CQ and REQ_NOTIFY_CQ is directly handled in libbnxt_re */ + + /* Kernel verbs */ + ibdev->query_device = bnxt_re_query_device; + ibdev->modify_device = bnxt_re_modify_device; + + ibdev->query_port = bnxt_re_query_port; + ibdev->modify_port = bnxt_re_modify_port; + ibdev->get_port_immutable = bnxt_re_get_port_immutable; + ibdev->query_pkey = bnxt_re_query_pkey; + ibdev->query_gid = bnxt_re_query_gid; + ibdev->get_netdev = bnxt_re_get_netdev; + ibdev->add_gid = bnxt_re_add_gid; + ibdev->del_gid = bnxt_re_del_gid; + ibdev->get_link_layer = bnxt_re_get_link_layer; + + ibdev->alloc_pd = bnxt_re_alloc_pd; + ibdev->dealloc_pd = bnxt_re_dealloc_pd; + + ibdev->create_ah = bnxt_re_create_ah; + ibdev->modify_ah = bnxt_re_modify_ah; + ibdev->query_ah = bnxt_re_query_ah; + ibdev->destroy_ah = bnxt_re_destroy_ah; + + ibdev->create_qp = bnxt_re_create_qp; + ibdev->modify_qp = bnxt_re_modify_qp; + ibdev->query_qp = bnxt_re_query_qp; + ibdev->destroy_qp = bnxt_re_destroy_qp; + + ibdev->post_send = bnxt_re_post_send; + ibdev->post_recv = bnxt_re_post_recv; + + ibdev->create_cq = bnxt_re_create_cq; + ibdev->destroy_cq = bnxt_re_destroy_cq; + ibdev->poll_cq = bnxt_re_poll_cq; + ibdev->req_notify_cq = bnxt_re_req_notify_cq; + + ibdev->get_dma_mr = bnxt_re_get_dma_mr; + ibdev->dereg_mr = bnxt_re_dereg_mr; + ibdev->alloc_mr = bnxt_re_alloc_mr; + ibdev->map_mr_sg = bnxt_re_map_mr_sg; + ibdev->alloc_fmr = bnxt_re_alloc_fmr; + ibdev->map_phys_fmr = bnxt_re_map_phys_fmr; + ibdev->unmap_fmr = bnxt_re_unmap_fmr; + ibdev->dealloc_fmr = bnxt_re_dealloc_fmr; + + ibdev->reg_user_mr = bnxt_re_reg_user_mr; + ibdev->alloc_ucontext = bnxt_re_alloc_ucontext; + ibdev->dealloc_ucontext = bnxt_re_dealloc_ucontext; + ibdev->mmap = bnxt_re_mmap; + + return ib_register_device(ibdev, NULL); +} + +static ssize_t show_rev(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "0x%x\n", rdev->en_dev->pdev->vendor); +} + +static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "%s\n", rdev->dev_attr.fw_ver); +} + +static ssize_t show_hca(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "%s\n", rdev->ibdev.node_desc); +} + +static DEVICE_ATTR(hw_rev, 0444, show_rev, NULL); +static DEVICE_ATTR(fw_rev, 0444, show_fw_ver, NULL); +static DEVICE_ATTR(hca_type, 0444, show_hca, NULL); + +static struct device_attribute *bnxt_re_attributes[] = { + &dev_attr_hw_rev, + &dev_attr_fw_rev, + &dev_attr_hca_type +}; + +static void bnxt_re_dev_remove(struct bnxt_re_dev *rdev) +{ + dev_put(rdev->netdev); + rdev->netdev = NULL; + + mutex_lock(&bnxt_re_dev_lock); + list_del_rcu(&rdev->list); + mutex_unlock(&bnxt_re_dev_lock); + + synchronize_rcu(); + flush_workqueue(bnxt_re_wq); + + ib_dealloc_device(&rdev->ibdev); + /* rdev is gone */ +} + +static struct bnxt_re_dev *bnxt_re_dev_add(struct net_device *netdev, + struct bnxt_en_dev *en_dev) +{ + struct bnxt_re_dev *rdev; + + /* Allocate bnxt_re_dev instance here */ + rdev = (struct bnxt_re_dev *)ib_alloc_device(sizeof(*rdev)); + if (!rdev) { + dev_err(NULL, "%s: bnxt_re_dev allocation failure!", + ROCE_DRV_MODULE_NAME); + return NULL; + } + /* Default values */ + rdev->netdev = netdev; + dev_hold(rdev->netdev); + rdev->en_dev = en_dev; + rdev->id = rdev->en_dev->pdev->devfn; + INIT_LIST_HEAD(&rdev->qp_list); + mutex_init(&rdev->qp_lock); + atomic_set(&rdev->qp_count, 0); + atomic_set(&rdev->cq_count, 0); + atomic_set(&rdev->srq_count, 0); + atomic_set(&rdev->mr_count, 0); + atomic_set(&rdev->mw_count, 0); + rdev->cosq[0] = 0xFFFF; + rdev->cosq[1] = 0xFFFF; + + mutex_lock(&bnxt_re_dev_lock); + list_add_tail_rcu(&rdev->list, &bnxt_re_dev_list); + mutex_unlock(&bnxt_re_dev_lock); + return rdev; +} + +static int bnxt_re_aeq_handler(struct bnxt_qplib_rcfw *rcfw, + struct creq_func_event *aeqe) +{ + switch (aeqe->event) { + case CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TQM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCS_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCC_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TIM_ERROR: + break; + default: + return -EINVAL; + } + return 0; +} + +static int bnxt_re_cqn_handler(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *handle) +{ + struct bnxt_re_cq *cq = container_of(handle, struct bnxt_re_cq, + qplib_cq); + + if (!cq) { + dev_err(NULL, "%s: CQ is NULL, CQN not handled", + ROCE_DRV_MODULE_NAME); + return -EINVAL; + } + if (cq->ib_cq.comp_handler) { + /* Lock comp_handler? */ + (*cq->ib_cq.comp_handler)(&cq->ib_cq, cq->ib_cq.cq_context); + } + + return 0; +} + +static void bnxt_re_cleanup_res(struct bnxt_re_dev *rdev) +{ + if (rdev->nq.hwq.max_elements) + bnxt_qplib_disable_nq(&rdev->nq); + + if (rdev->qplib_res.rcfw) + bnxt_qplib_cleanup_res(&rdev->qplib_res); +} + +static int bnxt_re_init_res(struct bnxt_re_dev *rdev) +{ + int rc = 0; + + bnxt_qplib_init_res(&rdev->qplib_res); + + if (rdev->msix_entries[BNXT_RE_NQ_IDX].vector <= 0) + return -EINVAL; + + rc = bnxt_qplib_enable_nq(rdev->en_dev->pdev, &rdev->nq, + rdev->msix_entries[BNXT_RE_NQ_IDX].vector, + rdev->msix_entries[BNXT_RE_NQ_IDX].db_offset, + &bnxt_re_cqn_handler, + NULL); + + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to enable NQ: %#x", rc); + + return rc; +} + +static void bnxt_re_free_res(struct bnxt_re_dev *rdev, bool lock_wait) +{ + if (rdev->nq.hwq.max_elements) { + bnxt_re_net_ring_free(rdev, rdev->nq.ring_id, lock_wait); + bnxt_qplib_free_nq(&rdev->nq); + } + if (rdev->qplib_res.dpi_tbl.max) { + bnxt_qplib_dealloc_dpi(&rdev->qplib_res, + &rdev->qplib_res.dpi_tbl, + &rdev->dpi_privileged); + } + if (rdev->qplib_res.rcfw) { + bnxt_qplib_free_res(&rdev->qplib_res); + rdev->qplib_res.rcfw = NULL; + } +} + +static int bnxt_re_alloc_res(struct bnxt_re_dev *rdev) +{ + int rc = 0; + + /* Configure and allocate resources for qplib */ + rdev->qplib_res.rcfw = &rdev->rcfw; + rc = bnxt_qplib_get_dev_attr(&rdev->rcfw, &rdev->dev_attr); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_res(&rdev->qplib_res, rdev->en_dev->pdev, + rdev->netdev, &rdev->dev_attr); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_dpi(&rdev->qplib_res.dpi_tbl, + &rdev->dpi_privileged, + rdev); + if (rc) + goto fail; + + rdev->nq.hwq.max_elements = BNXT_RE_MAX_CQ_COUNT + + BNXT_RE_MAX_SRQC_COUNT + 2; + rc = bnxt_qplib_alloc_nq(rdev->en_dev->pdev, &rdev->nq); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate NQ memory: %#x", rc); + goto fail; + } + rc = bnxt_re_net_ring_alloc + (rdev, rdev->nq.hwq.pbl[PBL_LVL_0].pg_map_arr, + rdev->nq.hwq.pbl[rdev->nq.hwq.level].pg_count, + HWRM_RING_ALLOC_CMPL, BNXT_QPLIB_NQE_MAX_CNT - 1, + rdev->msix_entries[BNXT_RE_NQ_IDX].ring_idx, + &rdev->nq.ring_id); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate NQ ring: %#x", rc); + goto free_nq; + } + return 0; +free_nq: + bnxt_qplib_free_nq(&rdev->nq); +fail: + rdev->qplib_res.rcfw = NULL; + return rc; +} + +static void bnxt_re_dispatch_event(struct ib_device *ibdev, struct ib_qp *qp, + u8 port_num, enum ib_event_type event) +{ + struct ib_event ib_event; + + ib_event.device = ibdev; + if (qp) + ib_event.element.qp = qp; + else + ib_event.element.port_num = port_num; + ib_event.event = event; + ib_dispatch_event(&ib_event); +} + +#define HWRM_QUEUE_PRI2COS_QCFG_INPUT_FLAGS_IVLAN 0x02 +static int bnxt_re_query_hwrm_pri2cos(struct bnxt_re_dev *rdev, u8 dir, + u64 *cid_map) +{ + struct hwrm_queue_pri2cos_qcfg_input req = {0}; + struct bnxt *bp = netdev_priv(rdev->netdev); + struct hwrm_queue_pri2cos_qcfg_output resp; + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct bnxt_fw_msg fw_msg; + u32 flags = 0; + u8 *qcfgmap, *tmp_map; + int rc = 0, i; + + if (!cid_map) + return -EINVAL; + + memset(&fw_msg, 0, sizeof(fw_msg)); + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, + HWRM_QUEUE_PRI2COS_QCFG, -1, -1); + flags |= (dir & 0x01); + flags |= HWRM_QUEUE_PRI2COS_QCFG_INPUT_FLAGS_IVLAN; + req.flags = cpu_to_le32(flags); + req.port_id = bp->pf.port_id; + + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + return rc; + + if (resp.queue_cfg_info) { + dev_warn(rdev_to_dev(rdev), + "Asymmetric cos queue configuration detected"); + dev_warn(rdev_to_dev(rdev), + " on device, QoS may not be fully functional\n"); + } + qcfgmap = &resp.pri0_cos_queue_id; + tmp_map = (u8 *)cid_map; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + tmp_map[i] = qcfgmap[i]; + + return rc; +} + +static bool bnxt_re_is_qp1_or_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp) +{ + return (qp->ib_qp.qp_type == IB_QPT_GSI) || (qp == rdev->qp1_sqp); +} + +static void bnxt_re_dev_stop(struct bnxt_re_dev *rdev) +{ + int mask = IB_QP_STATE; + struct ib_qp_attr qp_attr; + struct bnxt_re_qp *qp; + + qp_attr.qp_state = IB_QPS_ERR; + mutex_lock(&rdev->qp_lock); + list_for_each_entry(qp, &rdev->qp_list, list) { + /* Modify the state of all QPs except QP1/Shadow QP */ + if (!bnxt_re_is_qp1_or_shadow_qp(rdev, qp)) { + if (qp->qplib_qp.state != + CMDQ_MODIFY_QP_NEW_STATE_RESET && + qp->qplib_qp.state != + CMDQ_MODIFY_QP_NEW_STATE_ERR) { + bnxt_re_dispatch_event(&rdev->ibdev, &qp->ib_qp, + 1, IB_EVENT_QP_FATAL); + bnxt_re_modify_qp(&qp->ib_qp, &qp_attr, mask, + NULL); + } + } + } + mutex_unlock(&rdev->qp_lock); +} + +static u32 bnxt_re_get_priority_mask(struct bnxt_re_dev *rdev) +{ + u32 prio_map = 0, tmp_map = 0; + struct net_device *netdev; + struct dcb_app app; + + netdev = rdev->netdev; + + memset(&app, 0, sizeof(app)); + app.selector = IEEE_8021QAZ_APP_SEL_ETHERTYPE; + app.protocol = ETH_P_IBOE; + tmp_map = dcb_ieee_getapp_mask(netdev, &app); + prio_map = tmp_map; + + app.selector = IEEE_8021QAZ_APP_SEL_DGRAM; + app.protocol = ROCE_V2_UDP_DPORT; + tmp_map = dcb_ieee_getapp_mask(netdev, &app); + prio_map |= tmp_map; + + if (!prio_map) + prio_map = -EFAULT; + return prio_map; +} + +static void bnxt_re_parse_cid_map(u8 prio_map, u8 *cid_map, u16 *cosq) +{ + u16 prio; + u8 id; + + for (prio = 0, id = 0; prio < 8; prio++) { + if (prio_map & (1 << prio)) { + cosq[id] = cid_map[prio]; + id++; + if (id == 2) /* Max 2 tcs supported */ + break; + } + } +} + +static int bnxt_re_setup_qos(struct bnxt_re_dev *rdev) +{ + u8 prio_map = 0; + u64 cid_map; + int rc; + + /* Get priority for roce */ + rc = bnxt_re_get_priority_mask(rdev); + if (rc < 0) + return rc; + prio_map = (u8)rc; + + if (prio_map == rdev->cur_prio_map) + return 0; + rdev->cur_prio_map = prio_map; + /* Get cosq id for this priority */ + rc = bnxt_re_query_hwrm_pri2cos(rdev, 0, &cid_map); + if (rc) { + dev_warn(rdev_to_dev(rdev), "no cos for p_mask %x\n", prio_map); + return rc; + } + /* Parse CoS IDs for app priority */ + bnxt_re_parse_cid_map(prio_map, (u8 *)&cid_map, rdev->cosq); + + /* Config BONO. */ + rc = bnxt_qplib_map_tc2cos(&rdev->qplib_res, rdev->cosq); + if (rc) { + dev_warn(rdev_to_dev(rdev), "no tc for cos{%x, %x}\n", + rdev->cosq[0], rdev->cosq[1]); + return rc; + } + + return 0; +} + +static void bnxt_re_ib_unreg(struct bnxt_re_dev *rdev, bool lock_wait) +{ + int i, rc; + + if (test_and_clear_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) { + for (i = 0; i < ARRAY_SIZE(bnxt_re_attributes); i++) + device_remove_file(&rdev->ibdev.dev, + bnxt_re_attributes[i]); + /* Cleanup ib dev */ + bnxt_re_unregister_ib(rdev); + } + if (test_and_clear_bit(BNXT_RE_FLAG_QOS_WORK_REG, &rdev->flags)) + cancel_delayed_work(&rdev->worker); + + bnxt_re_cleanup_res(rdev); + bnxt_re_free_res(rdev, lock_wait); + + if (test_and_clear_bit(BNXT_RE_FLAG_RCFW_CHANNEL_EN, &rdev->flags)) { + rc = bnxt_qplib_deinit_rcfw(&rdev->rcfw); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to deinitialize RCFW: %#x", rc); + bnxt_re_net_stats_ctx_free(rdev, rdev->qplib_ctx.stats.fw_id, + lock_wait); + bnxt_qplib_free_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx); + bnxt_qplib_disable_rcfw_channel(&rdev->rcfw); + bnxt_re_net_ring_free(rdev, rdev->rcfw.creq_ring_id, lock_wait); + bnxt_qplib_free_rcfw_channel(&rdev->rcfw); + } + if (test_and_clear_bit(BNXT_RE_FLAG_GOT_MSIX, &rdev->flags)) { + rc = bnxt_re_free_msix(rdev, lock_wait); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to free MSI-X vectors: %#x", rc); + } + if (test_and_clear_bit(BNXT_RE_FLAG_NETDEV_REGISTERED, &rdev->flags)) { + rc = bnxt_re_unregister_netdev(rdev, lock_wait); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to unregister with netdev: %#x", rc); + } +} + +static void bnxt_re_set_resource_limits(struct bnxt_re_dev *rdev) +{ + u32 i; + + rdev->qplib_ctx.qpc_count = BNXT_RE_MAX_QPC_COUNT; + rdev->qplib_ctx.mrw_count = BNXT_RE_MAX_MRW_COUNT; + rdev->qplib_ctx.srqc_count = BNXT_RE_MAX_SRQC_COUNT; + rdev->qplib_ctx.cq_count = BNXT_RE_MAX_CQ_COUNT; + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) + rdev->qplib_ctx.tqm_count[i] = + rdev->dev_attr.tqm_alloc_reqs[i]; +} + +/* worker thread for polling periodic events. Now used for QoS programming*/ +static void bnxt_re_worker(struct work_struct *work) +{ + struct bnxt_re_dev *rdev = container_of(work, struct bnxt_re_dev, + worker.work); + + bnxt_re_setup_qos(rdev); + schedule_delayed_work(&rdev->worker, msecs_to_jiffies(30000)); +} + +static int bnxt_re_ib_reg(struct bnxt_re_dev *rdev) +{ + int i, j, rc; + + /* Registered a new RoCE device instance to netdev */ + rc = bnxt_re_register_netdev(rdev); + if (rc) { + pr_err("Failed to register with netedev: %#x\n", rc); + return -EINVAL; + } + set_bit(BNXT_RE_FLAG_NETDEV_REGISTERED, &rdev->flags); + + rc = bnxt_re_request_msix(rdev); + if (rc) { + pr_err("Failed to get MSI-X vectors: %#x\n", rc); + rc = -EINVAL; + goto fail; + } + set_bit(BNXT_RE_FLAG_GOT_MSIX, &rdev->flags); + + /* Establish RCFW Communication Channel to initialize the context + * memory for the function and all child VFs + */ + rc = bnxt_qplib_alloc_rcfw_channel(rdev->en_dev->pdev, &rdev->rcfw); + if (rc) + goto fail; + + rc = bnxt_re_net_ring_alloc + (rdev, rdev->rcfw.creq.pbl[PBL_LVL_0].pg_map_arr, + rdev->rcfw.creq.pbl[rdev->rcfw.creq.level].pg_count, + HWRM_RING_ALLOC_CMPL, BNXT_QPLIB_CREQE_MAX_CNT - 1, + rdev->msix_entries[BNXT_RE_AEQ_IDX].ring_idx, + &rdev->rcfw.creq_ring_id); + if (rc) { + pr_err("Failed to allocate CREQ: %#x\n", rc); + goto free_rcfw; + } + rc = bnxt_qplib_enable_rcfw_channel + (rdev->en_dev->pdev, &rdev->rcfw, + rdev->msix_entries[BNXT_RE_AEQ_IDX].vector, + rdev->msix_entries[BNXT_RE_AEQ_IDX].db_offset, + 0, &bnxt_re_aeq_handler); + if (rc) { + pr_err("Failed to enable RCFW channel: %#x\n", rc); + goto free_ring; + } + + rc = bnxt_qplib_get_dev_attr(&rdev->rcfw, &rdev->dev_attr); + if (rc) + goto disable_rcfw; + bnxt_re_set_resource_limits(rdev); + + rc = bnxt_qplib_alloc_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx, 0); + if (rc) { + pr_err("Failed to allocate QPLIB context: %#x\n", rc); + goto disable_rcfw; + } + rc = bnxt_re_net_stats_ctx_alloc(rdev, + rdev->qplib_ctx.stats.dma_map, + &rdev->qplib_ctx.stats.fw_id); + if (rc) { + pr_err("Failed to allocate stats context: %#x\n", rc); + goto free_ctx; + } + + rc = bnxt_qplib_init_rcfw(&rdev->rcfw, &rdev->qplib_ctx, 0); + if (rc) { + pr_err("Failed to initialize RCFW: %#x\n", rc); + goto free_sctx; + } + set_bit(BNXT_RE_FLAG_RCFW_CHANNEL_EN, &rdev->flags); + + /* Resources based on the 'new' device caps */ + rc = bnxt_re_alloc_res(rdev); + if (rc) { + pr_err("Failed to allocate resources: %#x\n", rc); + goto fail; + } + rc = bnxt_re_init_res(rdev); + if (rc) { + pr_err("Failed to initialize resources: %#x\n", rc); + goto fail; + } + + rc = bnxt_re_setup_qos(rdev); + if (rc) + pr_info("RoCE priority not yet configured\n"); + + INIT_DELAYED_WORK(&rdev->worker, bnxt_re_worker); + set_bit(BNXT_RE_FLAG_QOS_WORK_REG, &rdev->flags); + schedule_delayed_work(&rdev->worker, msecs_to_jiffies(30000)); + + /* Register ib dev */ + rc = bnxt_re_register_ib(rdev); + if (rc) { + pr_err("Failed to register with IB: %#x\n", rc); + goto fail; + } + dev_info(rdev_to_dev(rdev), "Device registered successfully"); + for (i = 0; i < ARRAY_SIZE(bnxt_re_attributes); i++) { + rc = device_create_file(&rdev->ibdev.dev, + bnxt_re_attributes[i]); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to create IB sysfs: %#x", rc); + /* Must clean up all created device files */ + for (j = 0; j < i; j++) + device_remove_file(&rdev->ibdev.dev, + bnxt_re_attributes[j]); + bnxt_re_unregister_ib(rdev); + goto fail; + } + } + set_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags); + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, IB_EVENT_PORT_ACTIVE); + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, IB_EVENT_GID_CHANGE); + + return 0; +free_sctx: + bnxt_re_net_stats_ctx_free(rdev, rdev->qplib_ctx.stats.fw_id, true); +free_ctx: + bnxt_qplib_free_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx); +disable_rcfw: + bnxt_qplib_disable_rcfw_channel(&rdev->rcfw); +free_ring: + bnxt_re_net_ring_free(rdev, rdev->rcfw.creq_ring_id, true); +free_rcfw: + bnxt_qplib_free_rcfw_channel(&rdev->rcfw); +fail: + bnxt_re_ib_unreg(rdev, true); + return rc; +} + +static void bnxt_re_dev_unreg(struct bnxt_re_dev *rdev) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct net_device *netdev = rdev->netdev; + + bnxt_re_dev_remove(rdev); + + if (netdev) + bnxt_re_dev_unprobe(netdev, en_dev); +} + +static int bnxt_re_dev_reg(struct bnxt_re_dev **rdev, struct net_device *netdev) +{ + struct bnxt_en_dev *en_dev; + int rc = 0; + + if (!is_bnxt_re_dev(netdev)) + return -ENODEV; + + en_dev = bnxt_re_dev_probe(netdev); + if (IS_ERR(en_dev)) { + if (en_dev != ERR_PTR(-ENODEV)) + pr_err("%s: Failed to probe\n", ROCE_DRV_MODULE_NAME); + rc = PTR_ERR(en_dev); + goto exit; + } + *rdev = bnxt_re_dev_add(netdev, en_dev); + if (!*rdev) { + rc = -ENOMEM; + bnxt_re_dev_unprobe(netdev, en_dev); + goto exit; + } +exit: + return rc; +} + +static void bnxt_re_remove_one(struct bnxt_re_dev *rdev) +{ + pci_dev_put(rdev->en_dev->pdev); +} + +/* Handle all deferred netevents tasks */ +static void bnxt_re_task(struct work_struct *work) +{ + struct bnxt_re_work *re_work; + struct bnxt_re_dev *rdev; + int rc = 0; + + re_work = container_of(work, struct bnxt_re_work, work); + rdev = re_work->rdev; + + if (re_work->event != NETDEV_REGISTER && + !test_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) + return; + + switch (re_work->event) { + case NETDEV_REGISTER: + rc = bnxt_re_ib_reg(rdev); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to register with IB: %#x", rc); + break; + case NETDEV_UP: + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, + IB_EVENT_PORT_ACTIVE); + break; + case NETDEV_DOWN: + bnxt_re_dev_stop(rdev); + break; + case NETDEV_CHANGE: + if (!netif_carrier_ok(rdev->netdev)) + bnxt_re_dev_stop(rdev); + else if (netif_carrier_ok(rdev->netdev)) + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, + IB_EVENT_PORT_ACTIVE); + break; + default: + break; + } + kfree(re_work); +} + +static void bnxt_re_init_one(struct bnxt_re_dev *rdev) +{ + pci_dev_get(rdev->en_dev->pdev); +} + +/* + * "Notifier chain callback can be invoked for the same chain from + * different CPUs at the same time". + * + * For cases when the netdev is already present, our call to the + * register_netdevice_notifier() will actually get the rtnl_lock() + * before sending NETDEV_REGISTER and (if up) NETDEV_UP + * events. + * + * But for cases when the netdev is not already present, the notifier + * chain is subjected to be invoked from different CPUs simultaneously. + * + * This is protected by the netdev_mutex. + */ +static int bnxt_re_netdev_event(struct notifier_block *notifier, + unsigned long event, void *ptr) +{ + struct net_device *real_dev, *netdev = netdev_notifier_info_to_dev(ptr); + struct bnxt_re_work *re_work; + struct bnxt_re_dev *rdev; + int rc = 0; + bool sch_work = false; + + real_dev = rdma_vlan_dev_real_dev(netdev); + if (!real_dev) + real_dev = netdev; + + rdev = bnxt_re_from_netdev(real_dev); + if (!rdev && event != NETDEV_REGISTER) + goto exit; + if (real_dev != netdev) + goto exit; + + switch (event) { + case NETDEV_REGISTER: + if (rdev) + break; + rc = bnxt_re_dev_reg(&rdev, real_dev); + if (rc == -ENODEV) + break; + if (rc) { + pr_err("Failed to register with the device %s: %#x\n", + real_dev->name, rc); + break; + } + bnxt_re_init_one(rdev); + sch_work = true; + break; + + case NETDEV_UNREGISTER: + bnxt_re_ib_unreg(rdev, false); + bnxt_re_remove_one(rdev); + bnxt_re_dev_unreg(rdev); + break; + + default: + sch_work = true; + break; + } + if (sch_work) { + /* Allocate for the deferred task */ + re_work = kzalloc(sizeof(*re_work), GFP_ATOMIC); + if (re_work) { + re_work->rdev = rdev; + re_work->event = event; + re_work->vlan_dev = (real_dev == netdev ? + NULL : netdev); + INIT_WORK(&re_work->work, bnxt_re_task); + queue_work(bnxt_re_wq, &re_work->work); + } + } + +exit: + return NOTIFY_DONE; +} + +static struct notifier_block bnxt_re_netdev_notifier = { + .notifier_call = bnxt_re_netdev_event +}; + +static int __init bnxt_re_mod_init(void) +{ + int rc = 0; + + pr_info("%s: %s", ROCE_DRV_MODULE_NAME, version); + + bnxt_re_wq = create_singlethread_workqueue("bnxt_re"); + if (!bnxt_re_wq) + return -ENOMEM; + + INIT_LIST_HEAD(&bnxt_re_dev_list); + + rc = register_netdevice_notifier(&bnxt_re_netdev_notifier); + if (rc) { + pr_err("%s: Cannot register to netdevice_notifier", + ROCE_DRV_MODULE_NAME); + goto err_netdev; + } + return 0; + +err_netdev: + destroy_workqueue(bnxt_re_wq); + + return rc; +} + +static void __exit bnxt_re_mod_exit(void) +{ + unregister_netdevice_notifier(&bnxt_re_netdev_notifier); + if (bnxt_re_wq) + destroy_workqueue(bnxt_re_wq); +} + +module_init(bnxt_re_mod_init); +module_exit(bnxt_re_mod_exit); diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.c b/drivers/infiniband/hw/bnxt_re/qplib_fp.c new file mode 100644 index 000000000000..43d08b5e9085 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.c @@ -0,0 +1,2167 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Fast Path Operators + */ + +#include +#include +#include +#include +#include +#include + +#include "roce_hsi.h" + +#include "qplib_res.h" +#include "qplib_rcfw.h" +#include "qplib_sp.h" +#include "qplib_fp.h" + +static void bnxt_qplib_arm_cq_enable(struct bnxt_qplib_cq *cq); + +static void bnxt_qplib_free_qp_hdr_buf(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_q *sq = &qp->sq; + + if (qp->rq_hdr_buf) + dma_free_coherent(&res->pdev->dev, + rq->hwq.max_elements * qp->rq_hdr_buf_size, + qp->rq_hdr_buf, qp->rq_hdr_buf_map); + if (qp->sq_hdr_buf) + dma_free_coherent(&res->pdev->dev, + sq->hwq.max_elements * qp->sq_hdr_buf_size, + qp->sq_hdr_buf, qp->sq_hdr_buf_map); + qp->rq_hdr_buf = NULL; + qp->sq_hdr_buf = NULL; + qp->rq_hdr_buf_map = 0; + qp->sq_hdr_buf_map = 0; + qp->sq_hdr_buf_size = 0; + qp->rq_hdr_buf_size = 0; +} + +static int bnxt_qplib_alloc_qp_hdr_buf(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_q *sq = &qp->rq; + int rc = 0; + + if (qp->sq_hdr_buf_size && sq->hwq.max_elements) { + qp->sq_hdr_buf = dma_alloc_coherent(&res->pdev->dev, + sq->hwq.max_elements * + qp->sq_hdr_buf_size, + &qp->sq_hdr_buf_map, GFP_KERNEL); + if (!qp->sq_hdr_buf) { + rc = -ENOMEM; + dev_err(&res->pdev->dev, + "QPLIB: Failed to create sq_hdr_buf"); + goto fail; + } + } + + if (qp->rq_hdr_buf_size && rq->hwq.max_elements) { + qp->rq_hdr_buf = dma_alloc_coherent(&res->pdev->dev, + rq->hwq.max_elements * + qp->rq_hdr_buf_size, + &qp->rq_hdr_buf_map, + GFP_KERNEL); + if (!qp->rq_hdr_buf) { + rc = -ENOMEM; + dev_err(&res->pdev->dev, + "QPLIB: Failed to create rq_hdr_buf"); + goto fail; + } + } + return 0; + +fail: + bnxt_qplib_free_qp_hdr_buf(res, qp); + return rc; +} + +static void bnxt_qplib_service_nq(unsigned long data) +{ + struct bnxt_qplib_nq *nq = (struct bnxt_qplib_nq *)data; + struct bnxt_qplib_hwq *hwq = &nq->hwq; + struct nq_base *nqe, **nq_ptr; + int num_cqne_processed = 0; + u32 sw_cons, raw_cons; + u16 type; + int budget = nq->budget; + u64 q_handle; + + /* Service the NQ until empty */ + raw_cons = hwq->cons; + while (budget--) { + sw_cons = HWQ_CMP(raw_cons, hwq); + nq_ptr = (struct nq_base **)hwq->pbl_ptr; + nqe = &nq_ptr[NQE_PG(sw_cons)][NQE_IDX(sw_cons)]; + if (!NQE_CMP_VALID(nqe, raw_cons, hwq->max_elements)) + break; + + type = le16_to_cpu(nqe->info10_type) & NQ_BASE_TYPE_MASK; + switch (type) { + case NQ_BASE_TYPE_CQ_NOTIFICATION: + { + struct nq_cn *nqcne = (struct nq_cn *)nqe; + + q_handle = le32_to_cpu(nqcne->cq_handle_low); + q_handle |= (u64)le32_to_cpu(nqcne->cq_handle_high) + << 32; + bnxt_qplib_arm_cq_enable((struct bnxt_qplib_cq *) + ((unsigned long)q_handle)); + if (!nq->cqn_handler(nq, (struct bnxt_qplib_cq *) + ((unsigned long)q_handle))) + num_cqne_processed++; + else + dev_warn(&nq->pdev->dev, + "QPLIB: cqn - type 0x%x not handled", + type); + break; + } + case NQ_BASE_TYPE_DBQ_EVENT: + break; + default: + dev_warn(&nq->pdev->dev, + "QPLIB: nqe with type = 0x%x not handled", + type); + break; + } + raw_cons++; + } + if (hwq->cons != raw_cons) { + hwq->cons = raw_cons; + NQ_DB_REARM(nq->bar_reg_iomem, hwq->cons, hwq->max_elements); + } +} + +static irqreturn_t bnxt_qplib_nq_irq(int irq, void *dev_instance) +{ + struct bnxt_qplib_nq *nq = dev_instance; + struct bnxt_qplib_hwq *hwq = &nq->hwq; + struct nq_base **nq_ptr; + u32 sw_cons; + + /* Prefetch the NQ element */ + sw_cons = HWQ_CMP(hwq->cons, hwq); + nq_ptr = (struct nq_base **)nq->hwq.pbl_ptr; + prefetch(&nq_ptr[NQE_PG(sw_cons)][NQE_IDX(sw_cons)]); + + /* Fan out to CPU affinitized kthreads? */ + tasklet_schedule(&nq->worker); + + return IRQ_HANDLED; +} + +void bnxt_qplib_disable_nq(struct bnxt_qplib_nq *nq) +{ + /* Make sure the HW is stopped! */ + synchronize_irq(nq->vector); + tasklet_disable(&nq->worker); + tasklet_kill(&nq->worker); + + if (nq->requested) { + free_irq(nq->vector, nq); + nq->requested = false; + } + if (nq->bar_reg_iomem) + iounmap(nq->bar_reg_iomem); + nq->bar_reg_iomem = NULL; + + nq->cqn_handler = NULL; + nq->srqn_handler = NULL; + nq->vector = 0; +} + +int bnxt_qplib_enable_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq, + int msix_vector, int bar_reg_offset, + int (*cqn_handler)(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *), + int (*srqn_handler)(struct bnxt_qplib_nq *nq, + void *, u8 event)) +{ + resource_size_t nq_base; + int rc; + + nq->pdev = pdev; + nq->vector = msix_vector; + + nq->cqn_handler = cqn_handler; + + nq->srqn_handler = srqn_handler; + + tasklet_init(&nq->worker, bnxt_qplib_service_nq, (unsigned long)nq); + + nq->requested = false; + rc = request_irq(nq->vector, bnxt_qplib_nq_irq, 0, "bnxt_qplib_nq", nq); + if (rc) { + dev_err(&nq->pdev->dev, + "Failed to request IRQ for NQ: %#x", rc); + bnxt_qplib_disable_nq(nq); + goto fail; + } + nq->requested = true; + nq->bar_reg = NQ_CONS_PCI_BAR_REGION; + nq->bar_reg_off = bar_reg_offset; + nq_base = pci_resource_start(pdev, nq->bar_reg); + if (!nq_base) { + rc = -ENOMEM; + goto fail; + } + nq->bar_reg_iomem = ioremap_nocache(nq_base + nq->bar_reg_off, 4); + if (!nq->bar_reg_iomem) { + rc = -ENOMEM; + goto fail; + } + NQ_DB_REARM(nq->bar_reg_iomem, nq->hwq.cons, nq->hwq.max_elements); + + return 0; +fail: + bnxt_qplib_disable_nq(nq); + return rc; +} + +void bnxt_qplib_free_nq(struct bnxt_qplib_nq *nq) +{ + if (nq->hwq.max_elements) + bnxt_qplib_free_hwq(nq->pdev, &nq->hwq); +} + +int bnxt_qplib_alloc_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq) +{ + nq->pdev = pdev; + if (!nq->hwq.max_elements || + nq->hwq.max_elements > BNXT_QPLIB_NQE_MAX_CNT) + nq->hwq.max_elements = BNXT_QPLIB_NQE_MAX_CNT; + + if (bnxt_qplib_alloc_init_hwq(nq->pdev, &nq->hwq, NULL, 0, + &nq->hwq.max_elements, + BNXT_QPLIB_MAX_NQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_L2_CMPL)) + return -ENOMEM; + + nq->budget = 8; + return 0; +} + +/* QP */ +int bnxt_qplib_create_qp1(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_qp1 req; + struct creq_create_qp1_resp *resp; + struct bnxt_qplib_pbl *pbl; + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_q *rq = &qp->rq; + int rc; + u16 cmd_flags = 0; + u32 qp_flags = 0; + + RCFW_CMD_PREP(req, CREATE_QP1, cmd_flags); + + /* General */ + req.type = qp->type; + req.dpi = cpu_to_le32(qp->dpi->dpi); + req.qp_handle = cpu_to_le64(qp->qp_handle); + + /* SQ */ + sq->hwq.max_elements = sq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &sq->hwq, NULL, 0, + &sq->hwq.max_elements, + BNXT_QPLIB_MAX_SQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + sq->swq = kcalloc(sq->hwq.max_elements, sizeof(*sq->swq), GFP_KERNEL); + if (!sq->swq) { + rc = -ENOMEM; + goto fail_sq; + } + pbl = &sq->hwq.pbl[PBL_LVL_0]; + req.sq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.sq_pg_size_sq_lvl = + ((sq->hwq.level & CMDQ_CREATE_QP1_SQ_LVL_MASK) + << CMDQ_CREATE_QP1_SQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K); + + if (qp->scq) + req.scq_cid = cpu_to_le32(qp->scq->id); + + qp_flags |= CMDQ_CREATE_QP1_QP_FLAGS_RESERVED_LKEY_ENABLE; + + /* RQ */ + if (rq->max_wqe) { + rq->hwq.max_elements = qp->rq.max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &rq->hwq, NULL, 0, + &rq->hwq.max_elements, + BNXT_QPLIB_MAX_RQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto fail_sq; + + rq->swq = kcalloc(rq->hwq.max_elements, sizeof(*rq->swq), + GFP_KERNEL); + if (!rq->swq) { + rc = -ENOMEM; + goto fail_rq; + } + pbl = &rq->hwq.pbl[PBL_LVL_0]; + req.rq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.rq_pg_size_rq_lvl = + ((rq->hwq.level & CMDQ_CREATE_QP1_RQ_LVL_MASK) << + CMDQ_CREATE_QP1_RQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K); + if (qp->rcq) + req.rcq_cid = cpu_to_le32(qp->rcq->id); + } + + /* Header buffer - allow hdr_buf pass in */ + rc = bnxt_qplib_alloc_qp_hdr_buf(res, qp); + if (rc) { + rc = -ENOMEM; + goto fail; + } + req.qp_flags = cpu_to_le32(qp_flags); + req.sq_size = cpu_to_le32(sq->hwq.max_elements); + req.rq_size = cpu_to_le32(rq->hwq.max_elements); + + req.sq_fwo_sq_sge = + cpu_to_le16((sq->max_sge & CMDQ_CREATE_QP1_SQ_SGE_MASK) << + CMDQ_CREATE_QP1_SQ_SGE_SFT); + req.rq_fwo_rq_sge = + cpu_to_le16((rq->max_sge & CMDQ_CREATE_QP1_RQ_SGE_MASK) << + CMDQ_CREATE_QP1_RQ_SGE_SFT); + + req.pd_id = cpu_to_le32(qp->pd->id); + + resp = (struct creq_create_qp1_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&res->pdev->dev, "QPLIB: FP: CREATE_QP1 send failed"); + rc = -EINVAL; + goto fail; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_QP1 timed out"); + rc = -ETIMEDOUT; + goto fail; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_QP1 failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + rc = -EINVAL; + goto fail; + } + qp->id = le32_to_cpu(resp->xid); + qp->cur_qp_state = CMDQ_MODIFY_QP_NEW_STATE_RESET; + sq->flush_in_progress = false; + rq->flush_in_progress = false; + + return 0; + +fail: + bnxt_qplib_free_qp_hdr_buf(res, qp); +fail_rq: + bnxt_qplib_free_hwq(res->pdev, &rq->hwq); + kfree(rq->swq); +fail_sq: + bnxt_qplib_free_hwq(res->pdev, &sq->hwq); + kfree(sq->swq); +exit: + return rc; +} + +int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct sq_send *hw_sq_send_hdr, **hw_sq_send_ptr; + struct cmdq_create_qp req; + struct creq_create_qp_resp *resp; + struct bnxt_qplib_pbl *pbl; + struct sq_psn_search **psn_search_ptr; + unsigned long int psn_search, poff = 0; + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_hwq *xrrq; + int i, rc, req_size, psn_sz; + u16 cmd_flags = 0, max_ssge; + u32 sw_prod, qp_flags = 0; + + RCFW_CMD_PREP(req, CREATE_QP, cmd_flags); + + /* General */ + req.type = qp->type; + req.dpi = cpu_to_le32(qp->dpi->dpi); + req.qp_handle = cpu_to_le64(qp->qp_handle); + + /* SQ */ + psn_sz = (qp->type == CMDQ_CREATE_QP_TYPE_RC) ? + sizeof(struct sq_psn_search) : 0; + sq->hwq.max_elements = sq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &sq->hwq, sq->sglist, + sq->nmap, &sq->hwq.max_elements, + BNXT_QPLIB_MAX_SQE_ENTRY_SIZE, + psn_sz, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + sq->swq = kcalloc(sq->hwq.max_elements, sizeof(*sq->swq), GFP_KERNEL); + if (!sq->swq) { + rc = -ENOMEM; + goto fail_sq; + } + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + if (psn_sz) { + psn_search_ptr = (struct sq_psn_search **) + &hw_sq_send_ptr[get_sqe_pg + (sq->hwq.max_elements)]; + psn_search = (unsigned long int) + &hw_sq_send_ptr[get_sqe_pg(sq->hwq.max_elements)] + [get_sqe_idx(sq->hwq.max_elements)]; + if (psn_search & ~PAGE_MASK) { + /* If the psn_search does not start on a page boundary, + * then calculate the offset + */ + poff = (psn_search & ~PAGE_MASK) / + BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE; + } + for (i = 0; i < sq->hwq.max_elements; i++) + sq->swq[i].psn_search = + &psn_search_ptr[get_psne_pg(i + poff)] + [get_psne_idx(i + poff)]; + } + pbl = &sq->hwq.pbl[PBL_LVL_0]; + req.sq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.sq_pg_size_sq_lvl = + ((sq->hwq.level & CMDQ_CREATE_QP_SQ_LVL_MASK) + << CMDQ_CREATE_QP_SQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K); + + /* initialize all SQ WQEs to LOCAL_INVALID (sq prep for hw fetch) */ + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + for (sw_prod = 0; sw_prod < sq->hwq.max_elements; sw_prod++) { + hw_sq_send_hdr = &hw_sq_send_ptr[get_sqe_pg(sw_prod)] + [get_sqe_idx(sw_prod)]; + hw_sq_send_hdr->wqe_type = SQ_BASE_WQE_TYPE_LOCAL_INVALID; + } + + if (qp->scq) + req.scq_cid = cpu_to_le32(qp->scq->id); + + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_RESERVED_LKEY_ENABLE; + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_FR_PMR_ENABLED; + if (qp->sig_type) + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_FORCE_COMPLETION; + + /* RQ */ + if (rq->max_wqe) { + rq->hwq.max_elements = rq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &rq->hwq, rq->sglist, + rq->nmap, &rq->hwq.max_elements, + BNXT_QPLIB_MAX_RQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto fail_sq; + + rq->swq = kcalloc(rq->hwq.max_elements, sizeof(*rq->swq), + GFP_KERNEL); + if (!rq->swq) { + rc = -ENOMEM; + goto fail_rq; + } + pbl = &rq->hwq.pbl[PBL_LVL_0]; + req.rq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.rq_pg_size_rq_lvl = + ((rq->hwq.level & CMDQ_CREATE_QP_RQ_LVL_MASK) << + CMDQ_CREATE_QP_RQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K); + } + + if (qp->rcq) + req.rcq_cid = cpu_to_le32(qp->rcq->id); + req.qp_flags = cpu_to_le32(qp_flags); + req.sq_size = cpu_to_le32(sq->hwq.max_elements); + req.rq_size = cpu_to_le32(rq->hwq.max_elements); + qp->sq_hdr_buf = NULL; + qp->rq_hdr_buf = NULL; + + rc = bnxt_qplib_alloc_qp_hdr_buf(res, qp); + if (rc) + goto fail_rq; + + /* CTRL-22434: Irrespective of the requested SGE count on the SQ + * always create the QP with max send sges possible if the requested + * inline size is greater than 0. + */ + max_ssge = qp->max_inline_data ? 6 : sq->max_sge; + req.sq_fwo_sq_sge = cpu_to_le16( + ((max_ssge & CMDQ_CREATE_QP_SQ_SGE_MASK) + << CMDQ_CREATE_QP_SQ_SGE_SFT) | 0); + req.rq_fwo_rq_sge = cpu_to_le16( + ((rq->max_sge & CMDQ_CREATE_QP_RQ_SGE_MASK) + << CMDQ_CREATE_QP_RQ_SGE_SFT) | 0); + /* ORRQ and IRRQ */ + if (psn_sz) { + xrrq = &qp->orrq; + xrrq->max_elements = + ORD_LIMIT_TO_ORRQ_SLOTS(qp->max_rd_atomic); + req_size = xrrq->max_elements * + BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE + PAGE_SIZE - 1; + req_size &= ~(PAGE_SIZE - 1); + rc = bnxt_qplib_alloc_init_hwq(res->pdev, xrrq, NULL, 0, + &xrrq->max_elements, + BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE, + 0, req_size, HWQ_TYPE_CTX); + if (rc) + goto fail_buf_free; + pbl = &xrrq->pbl[PBL_LVL_0]; + req.orrq_addr = cpu_to_le64(pbl->pg_map_arr[0]); + + xrrq = &qp->irrq; + xrrq->max_elements = IRD_LIMIT_TO_IRRQ_SLOTS( + qp->max_dest_rd_atomic); + req_size = xrrq->max_elements * + BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE + PAGE_SIZE - 1; + req_size &= ~(PAGE_SIZE - 1); + + rc = bnxt_qplib_alloc_init_hwq(res->pdev, xrrq, NULL, 0, + &xrrq->max_elements, + BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE, + 0, req_size, HWQ_TYPE_CTX); + if (rc) + goto fail_orrq; + + pbl = &xrrq->pbl[PBL_LVL_0]; + req.irrq_addr = cpu_to_le64(pbl->pg_map_arr[0]); + } + req.pd_id = cpu_to_le32(qp->pd->id); + + resp = (struct creq_create_qp_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_QP send failed"); + rc = -EINVAL; + goto fail; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_QP timed out"); + rc = -ETIMEDOUT; + goto fail; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_QP failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + rc = -EINVAL; + goto fail; + } + qp->id = le32_to_cpu(resp->xid); + qp->cur_qp_state = CMDQ_MODIFY_QP_NEW_STATE_RESET; + sq->flush_in_progress = false; + rq->flush_in_progress = false; + + return 0; + +fail: + if (qp->irrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->irrq); +fail_orrq: + if (qp->orrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->orrq); +fail_buf_free: + bnxt_qplib_free_qp_hdr_buf(res, qp); +fail_rq: + bnxt_qplib_free_hwq(res->pdev, &rq->hwq); + kfree(rq->swq); +fail_sq: + bnxt_qplib_free_hwq(res->pdev, &sq->hwq); + kfree(sq->swq); +exit: + return rc; +} + +static void __modify_flags_from_init_state(struct bnxt_qplib_qp *qp) +{ + switch (qp->state) { + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + /* INIT->RTR, configure the path_mtu to the default + * 2048 if not being requested + */ + if (!(qp->modify_flags & + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU)) { + qp->modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->path_mtu = + CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + } + qp->modify_flags &= + ~CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID; + /* Bono FW require the max_dest_rd_atomic to be >= 1 */ + if (qp->max_dest_rd_atomic < 1) + qp->max_dest_rd_atomic = 1; + qp->modify_flags &= ~CMDQ_MODIFY_QP_MODIFY_MASK_SRC_MAC; + /* Bono FW 20.6.5 requires SGID_INDEX configuration */ + if (!(qp->modify_flags & + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX)) { + qp->modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX; + qp->ah.sgid_index = 0; + } + break; + default: + break; + } +} + +static void __modify_flags_from_rtr_state(struct bnxt_qplib_qp *qp) +{ + switch (qp->state) { + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + /* Bono FW requires the max_rd_atomic to be >= 1 */ + if (qp->max_rd_atomic < 1) + qp->max_rd_atomic = 1; + /* Bono FW does not allow PKEY_INDEX, + * DGID, FLOW_LABEL, SGID_INDEX, HOP_LIMIT, + * TRAFFIC_CLASS, DEST_MAC, PATH_MTU, RQ_PSN, + * MIN_RNR_TIMER, MAX_DEST_RD_ATOMIC, DEST_QP_ID + * modification + */ + qp->modify_flags &= + ~(CMDQ_MODIFY_QP_MODIFY_MASK_PKEY | + CMDQ_MODIFY_QP_MODIFY_MASK_DGID | + CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL | + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX | + CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT | + CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC | + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN | + CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER | + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID); + break; + default: + break; + } +} + +static void __filter_modify_flags(struct bnxt_qplib_qp *qp) +{ + switch (qp->cur_qp_state) { + case CMDQ_MODIFY_QP_NEW_STATE_RESET: + break; + case CMDQ_MODIFY_QP_NEW_STATE_INIT: + __modify_flags_from_init_state(qp); + break; + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + __modify_flags_from_rtr_state(qp); + break; + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + break; + case CMDQ_MODIFY_QP_NEW_STATE_SQD: + break; + case CMDQ_MODIFY_QP_NEW_STATE_SQE: + break; + case CMDQ_MODIFY_QP_NEW_STATE_ERR: + break; + default: + break; + } +} + +int bnxt_qplib_modify_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_modify_qp req; + struct creq_modify_qp_resp *resp; + u16 cmd_flags = 0, pkey; + u32 temp32[4]; + u32 bmask; + + RCFW_CMD_PREP(req, MODIFY_QP, cmd_flags); + + /* Filter out the qp_attr_mask based on the state->new transition */ + __filter_modify_flags(qp); + bmask = qp->modify_flags; + req.modify_mask = cpu_to_le32(qp->modify_flags); + req.qp_cid = cpu_to_le32(qp->id); + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_STATE) { + req.network_type_en_sqd_async_notify_new_state = + (qp->state & CMDQ_MODIFY_QP_NEW_STATE_MASK) | + (qp->en_sqd_async_notify ? + CMDQ_MODIFY_QP_EN_SQD_ASYNC_NOTIFY : 0); + } + req.network_type_en_sqd_async_notify_new_state |= qp->nw_type; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS) + req.access = qp->access; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_PKEY) { + if (!bnxt_qplib_get_pkey(res, &res->pkey_tbl, + qp->pkey_index, &pkey)) + req.pkey = cpu_to_le16(pkey); + } + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_QKEY) + req.qkey = cpu_to_le32(qp->qkey); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DGID) { + memcpy(temp32, qp->ah.dgid.data, sizeof(struct bnxt_qplib_gid)); + req.dgid[0] = cpu_to_le32(temp32[0]); + req.dgid[1] = cpu_to_le32(temp32[1]); + req.dgid[2] = cpu_to_le32(temp32[2]); + req.dgid[3] = cpu_to_le32(temp32[3]); + } + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL) + req.flow_label = cpu_to_le32(qp->ah.flow_label); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX) + req.sgid_index = cpu_to_le16(res->sgid_tbl.hw_id + [qp->ah.sgid_index]); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT) + req.hop_limit = qp->ah.hop_limit; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS) + req.traffic_class = qp->ah.traffic_class; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC) + memcpy(req.dest_mac, qp->ah.dmac, 6); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU) + req.path_mtu = qp->path_mtu; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT) + req.timeout = qp->timeout; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT) + req.retry_cnt = qp->retry_cnt; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY) + req.rnr_retry = qp->rnr_retry; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER) + req.min_rnr_timer = qp->min_rnr_timer; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN) + req.rq_psn = cpu_to_le32(qp->rq.psn); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN) + req.sq_psn = cpu_to_le32(qp->sq.psn); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC) + req.max_rd_atomic = + ORD_LIMIT_TO_ORRQ_SLOTS(qp->max_rd_atomic); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC) + req.max_dest_rd_atomic = + IRD_LIMIT_TO_IRRQ_SLOTS(qp->max_dest_rd_atomic); + + req.sq_size = cpu_to_le32(qp->sq.hwq.max_elements); + req.rq_size = cpu_to_le32(qp->rq.hwq.max_elements); + req.sq_sge = cpu_to_le16(qp->sq.max_sge); + req.rq_sge = cpu_to_le16(qp->rq.max_sge); + req.max_inline_data = cpu_to_le32(qp->max_inline_data); + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID) + req.dest_qp_id = cpu_to_le32(qp->dest_qpn); + + req.vlan_pcp_vlan_dei_vlan_id = cpu_to_le16(qp->vlan_id); + + resp = (struct creq_modify_qp_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: MODIFY_QP send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: MODIFY_QP timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: MODIFY_QP failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + qp->cur_qp_state = qp->state; + return 0; +} + +int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_query_qp req; + struct creq_query_qp_resp *resp; + struct creq_query_qp_resp_sb *sb; + u16 cmd_flags = 0; + u32 temp32[4]; + int i; + + RCFW_CMD_PREP(req, QUERY_QP, cmd_flags); + + req.qp_cid = cpu_to_le32(qp->id); + req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS; + resp = (struct creq_query_qp_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void **)&sb, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: QUERY_QP send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: QUERY_QP timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: QUERY_QP failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + /* Extract the context from the side buffer */ + qp->state = sb->en_sqd_async_notify_state & + CREQ_QUERY_QP_RESP_SB_STATE_MASK; + qp->en_sqd_async_notify = sb->en_sqd_async_notify_state & + CREQ_QUERY_QP_RESP_SB_EN_SQD_ASYNC_NOTIFY ? + true : false; + qp->access = sb->access; + qp->pkey_index = le16_to_cpu(sb->pkey); + qp->qkey = le32_to_cpu(sb->qkey); + + temp32[0] = le32_to_cpu(sb->dgid[0]); + temp32[1] = le32_to_cpu(sb->dgid[1]); + temp32[2] = le32_to_cpu(sb->dgid[2]); + temp32[3] = le32_to_cpu(sb->dgid[3]); + memcpy(qp->ah.dgid.data, temp32, sizeof(qp->ah.dgid.data)); + + qp->ah.flow_label = le32_to_cpu(sb->flow_label); + + qp->ah.sgid_index = 0; + for (i = 0; i < res->sgid_tbl.max; i++) { + if (res->sgid_tbl.hw_id[i] == le16_to_cpu(sb->sgid_index)) { + qp->ah.sgid_index = i; + break; + } + } + if (i == res->sgid_tbl.max) + dev_warn(&res->pdev->dev, "QPLIB: SGID not found??"); + + qp->ah.hop_limit = sb->hop_limit; + qp->ah.traffic_class = sb->traffic_class; + memcpy(qp->ah.dmac, sb->dest_mac, 6); + qp->ah.vlan_id = (le16_to_cpu(sb->path_mtu_dest_vlan_id) & + CREQ_QUERY_QP_RESP_SB_VLAN_ID_MASK) >> + CREQ_QUERY_QP_RESP_SB_VLAN_ID_SFT; + qp->path_mtu = (le16_to_cpu(sb->path_mtu_dest_vlan_id) & + CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK) >> + CREQ_QUERY_QP_RESP_SB_PATH_MTU_SFT; + qp->timeout = sb->timeout; + qp->retry_cnt = sb->retry_cnt; + qp->rnr_retry = sb->rnr_retry; + qp->min_rnr_timer = sb->min_rnr_timer; + qp->rq.psn = le32_to_cpu(sb->rq_psn); + qp->max_rd_atomic = ORRQ_SLOTS_TO_ORD_LIMIT(sb->max_rd_atomic); + qp->sq.psn = le32_to_cpu(sb->sq_psn); + qp->max_dest_rd_atomic = + IRRQ_SLOTS_TO_IRD_LIMIT(sb->max_dest_rd_atomic); + qp->sq.max_wqe = qp->sq.hwq.max_elements; + qp->rq.max_wqe = qp->rq.hwq.max_elements; + qp->sq.max_sge = le16_to_cpu(sb->sq_sge); + qp->rq.max_sge = le16_to_cpu(sb->rq_sge); + qp->max_inline_data = le32_to_cpu(sb->max_inline_data); + qp->dest_qpn = le32_to_cpu(sb->dest_qp_id); + memcpy(qp->smac, sb->src_mac, 6); + qp->vlan_id = le16_to_cpu(sb->vlan_pcp_vlan_dei_vlan_id); + return 0; +} + +static void __clean_cq(struct bnxt_qplib_cq *cq, u64 qp) +{ + struct bnxt_qplib_hwq *cq_hwq = &cq->hwq; + struct cq_base *hw_cqe, **hw_cqe_ptr; + int i; + + for (i = 0; i < cq_hwq->max_elements; i++) { + hw_cqe_ptr = (struct cq_base **)cq_hwq->pbl_ptr; + hw_cqe = &hw_cqe_ptr[CQE_PG(i)][CQE_IDX(i)]; + if (!CQE_CMP_VALID(hw_cqe, i, cq_hwq->max_elements)) + continue; + switch (hw_cqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK) { + case CQ_BASE_CQE_TYPE_REQ: + case CQ_BASE_CQE_TYPE_TERMINAL: + { + struct cq_req *cqe = (struct cq_req *)hw_cqe; + + if (qp == le64_to_cpu(cqe->qp_handle)) + cqe->qp_handle = 0; + break; + } + case CQ_BASE_CQE_TYPE_RES_RC: + case CQ_BASE_CQE_TYPE_RES_UD: + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + { + struct cq_res_rc *cqe = (struct cq_res_rc *)hw_cqe; + + if (qp == le64_to_cpu(cqe->qp_handle)) + cqe->qp_handle = 0; + break; + } + default: + break; + } + } +} + +int bnxt_qplib_destroy_qp(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_qp req; + struct creq_destroy_qp_resp *resp; + unsigned long flags; + u16 cmd_flags = 0; + + RCFW_CMD_PREP(req, DESTROY_QP, cmd_flags); + + req.qp_cid = cpu_to_le32(qp->id); + resp = (struct creq_destroy_qp_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_QP send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_QP timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_QP failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + + /* Must walk the associated CQs to nullified the QP ptr */ + spin_lock_irqsave(&qp->scq->hwq.lock, flags); + + __clean_cq(qp->scq, (u64)(unsigned long)qp); + + if (qp->rcq && qp->rcq != qp->scq) { + spin_lock(&qp->rcq->hwq.lock); + __clean_cq(qp->rcq, (u64)(unsigned long)qp); + spin_unlock(&qp->rcq->hwq.lock); + } + + spin_unlock_irqrestore(&qp->scq->hwq.lock, flags); + + bnxt_qplib_free_qp_hdr_buf(res, qp); + bnxt_qplib_free_hwq(res->pdev, &qp->sq.hwq); + kfree(qp->sq.swq); + + bnxt_qplib_free_hwq(res->pdev, &qp->rq.hwq); + kfree(qp->rq.swq); + + if (qp->irrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->irrq); + if (qp->orrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->orrq); + + return 0; +} + +void *bnxt_qplib_get_qp1_sq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge) +{ + struct bnxt_qplib_q *sq = &qp->sq; + u32 sw_prod; + + memset(sge, 0, sizeof(*sge)); + + if (qp->sq_hdr_buf) { + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + sge->addr = (dma_addr_t)(qp->sq_hdr_buf_map + + sw_prod * qp->sq_hdr_buf_size); + sge->lkey = 0xFFFFFFFF; + sge->size = qp->sq_hdr_buf_size; + return qp->sq_hdr_buf + sw_prod * sge->size; + } + return NULL; +} + +u32 bnxt_qplib_get_rq_prod_index(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + + return HWQ_CMP(rq->hwq.prod, &rq->hwq); +} + +dma_addr_t bnxt_qplib_get_qp_buf_from_index(struct bnxt_qplib_qp *qp, u32 index) +{ + return (qp->rq_hdr_buf_map + index * qp->rq_hdr_buf_size); +} + +void *bnxt_qplib_get_qp1_rq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge) +{ + struct bnxt_qplib_q *rq = &qp->rq; + u32 sw_prod; + + memset(sge, 0, sizeof(*sge)); + + if (qp->rq_hdr_buf) { + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + sge->addr = (dma_addr_t)(qp->rq_hdr_buf_map + + sw_prod * qp->rq_hdr_buf_size); + sge->lkey = 0xFFFFFFFF; + sge->size = qp->rq_hdr_buf_size; + return qp->rq_hdr_buf + sw_prod * sge->size; + } + return NULL; +} + +void bnxt_qplib_post_send_db(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *sq = &qp->sq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_prod; + + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + + db_msg.index = cpu_to_le32((sw_prod << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((qp->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_SQ); + /* Flush all the WQE writes to HW */ + wmb(); + __iowrite64_copy(qp->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_post_send(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_swq *swq; + struct sq_send *hw_sq_send_hdr, **hw_sq_send_ptr; + struct sq_sge *hw_sge; + u32 sw_prod; + u8 wqe_size16; + int i, rc = 0, data_len = 0, pkt_num = 0; + __le32 temp32; + + if (qp->state != CMDQ_MODIFY_QP_NEW_STATE_RTS) { + rc = -EINVAL; + goto done; + } + if (HWQ_CMP((sq->hwq.prod + 1), &sq->hwq) == + HWQ_CMP(sq->hwq.cons, &sq->hwq)) { + rc = -ENOMEM; + goto done; + } + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + swq = &sq->swq[sw_prod]; + swq->wr_id = wqe->wr_id; + swq->type = wqe->type; + swq->flags = wqe->flags; + if (qp->sig_type) + swq->flags |= SQ_SEND_FLAGS_SIGNAL_COMP; + swq->start_psn = sq->psn & BTH_PSN_MASK; + + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + hw_sq_send_hdr = &hw_sq_send_ptr[get_sqe_pg(sw_prod)] + [get_sqe_idx(sw_prod)]; + + memset(hw_sq_send_hdr, 0, BNXT_QPLIB_MAX_SQE_ENTRY_SIZE); + + if (wqe->flags & BNXT_QPLIB_SWQE_FLAGS_INLINE) { + /* Copy the inline data */ + if (wqe->inline_len > BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH) { + dev_warn(&sq->hwq.pdev->dev, + "QPLIB: Inline data length > 96 detected"); + data_len = BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH; + } else { + data_len = wqe->inline_len; + } + memcpy(hw_sq_send_hdr->data, wqe->inline_data, data_len); + wqe_size16 = (data_len + 15) >> 4; + } else { + for (i = 0, hw_sge = (struct sq_sge *)hw_sq_send_hdr->data; + i < wqe->num_sge; i++, hw_sge++) { + hw_sge->va_or_pa = cpu_to_le64(wqe->sg_list[i].addr); + hw_sge->l_key = cpu_to_le32(wqe->sg_list[i].lkey); + hw_sge->size = cpu_to_le32(wqe->sg_list[i].size); + data_len += wqe->sg_list[i].size; + } + /* Each SGE entry = 1 WQE size16 */ + wqe_size16 = wqe->num_sge; + } + + /* Specifics */ + switch (wqe->type) { + case BNXT_QPLIB_SWQE_TYPE_SEND: + if (qp->type == CMDQ_CREATE_QP1_TYPE_GSI) { + /* Assemble info for Raw Ethertype QPs */ + struct sq_send_raweth_qp1 *sqe = + (struct sq_send_raweth_qp1 *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->cfa_action = cpu_to_le16(wqe->rawqp1.cfa_action); + sqe->lflags = cpu_to_le16(wqe->rawqp1.lflags); + sqe->length = cpu_to_le32(data_len); + sqe->cfa_meta = cpu_to_le32((wqe->rawqp1.cfa_meta & + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_MASK) << + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_SFT); + + break; + } + /* else, just fall thru */ + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM: + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV: + { + struct sq_send *sqe = (struct sq_send *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->inv_key_or_imm_data = cpu_to_le32( + wqe->send.inv_key); + if (qp->type == CMDQ_CREATE_QP_TYPE_UD) { + sqe->q_key = cpu_to_le32(wqe->send.q_key); + sqe->dst_qp = cpu_to_le32( + wqe->send.dst_qp & SQ_SEND_DST_QP_MASK); + sqe->length = cpu_to_le32(data_len); + sqe->avid = cpu_to_le32(wqe->send.avid & + SQ_SEND_AVID_MASK); + sq->psn = (sq->psn + 1) & BTH_PSN_MASK; + } else { + sqe->length = cpu_to_le32(data_len); + sqe->dst_qp = 0; + sqe->avid = 0; + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + } + break; + } + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE: + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM: + case BNXT_QPLIB_SWQE_TYPE_RDMA_READ: + { + struct sq_rdma *sqe = (struct sq_rdma *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->imm_data = cpu_to_le32(wqe->rdma.inv_key); + sqe->length = cpu_to_le32((u32)data_len); + sqe->remote_va = cpu_to_le64(wqe->rdma.remote_va); + sqe->remote_key = cpu_to_le32(wqe->rdma.r_key); + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + break; + } + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP: + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD: + { + struct sq_atomic *sqe = (struct sq_atomic *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->remote_key = cpu_to_le32(wqe->atomic.r_key); + sqe->remote_va = cpu_to_le64(wqe->atomic.remote_va); + sqe->swap_data = cpu_to_le64(wqe->atomic.swap_data); + sqe->cmp_data = cpu_to_le64(wqe->atomic.cmp_data); + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + break; + } + case BNXT_QPLIB_SWQE_TYPE_LOCAL_INV: + { + struct sq_localinvalidate *sqe = + (struct sq_localinvalidate *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->inv_l_key = cpu_to_le32(wqe->local_inv.inv_l_key); + + break; + } + case BNXT_QPLIB_SWQE_TYPE_FAST_REG_MR: + { + struct sq_fr_pmr *sqe = (struct sq_fr_pmr *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->access_cntl = wqe->frmr.access_cntl | + SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE; + sqe->zero_based_page_size_log = + (wqe->frmr.pg_sz_log & SQ_FR_PMR_PAGE_SIZE_LOG_MASK) << + SQ_FR_PMR_PAGE_SIZE_LOG_SFT | + (wqe->frmr.zero_based ? SQ_FR_PMR_ZERO_BASED : 0); + sqe->l_key = cpu_to_le32(wqe->frmr.l_key); + temp32 = cpu_to_le32(wqe->frmr.length); + memcpy(sqe->length, &temp32, sizeof(wqe->frmr.length)); + sqe->numlevels_pbl_page_size_log = + ((wqe->frmr.pbl_pg_sz_log << + SQ_FR_PMR_PBL_PAGE_SIZE_LOG_SFT) & + SQ_FR_PMR_PBL_PAGE_SIZE_LOG_MASK) | + ((wqe->frmr.levels << SQ_FR_PMR_NUMLEVELS_SFT) & + SQ_FR_PMR_NUMLEVELS_MASK); + + for (i = 0; i < wqe->frmr.page_list_len; i++) + wqe->frmr.pbl_ptr[i] = cpu_to_le64( + wqe->frmr.page_list[i] | + PTU_PTE_VALID); + sqe->pblptr = cpu_to_le64(wqe->frmr.pbl_dma_ptr); + sqe->va = cpu_to_le64(wqe->frmr.va); + + break; + } + case BNXT_QPLIB_SWQE_TYPE_BIND_MW: + { + struct sq_bind *sqe = (struct sq_bind *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->access_cntl = wqe->bind.access_cntl; + sqe->mw_type_zero_based = wqe->bind.mw_type | + (wqe->bind.zero_based ? SQ_BIND_ZERO_BASED : 0); + sqe->parent_l_key = cpu_to_le32(wqe->bind.parent_l_key); + sqe->l_key = cpu_to_le32(wqe->bind.r_key); + sqe->va = cpu_to_le64(wqe->bind.va); + temp32 = cpu_to_le32(wqe->bind.length); + memcpy(&sqe->length, &temp32, sizeof(wqe->bind.length)); + break; + } + default: + /* Bad wqe, return error */ + rc = -EINVAL; + goto done; + } + swq->next_psn = sq->psn & BTH_PSN_MASK; + if (swq->psn_search) { + swq->psn_search->opcode_start_psn = cpu_to_le32( + ((swq->start_psn << SQ_PSN_SEARCH_START_PSN_SFT) & + SQ_PSN_SEARCH_START_PSN_MASK) | + ((wqe->type << SQ_PSN_SEARCH_OPCODE_SFT) & + SQ_PSN_SEARCH_OPCODE_MASK)); + swq->psn_search->flags_next_psn = cpu_to_le32( + ((swq->next_psn << SQ_PSN_SEARCH_NEXT_PSN_SFT) & + SQ_PSN_SEARCH_NEXT_PSN_MASK)); + } + + sq->hwq.prod++; +done: + return rc; +} + +void bnxt_qplib_post_recv_db(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_prod; + + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + db_msg.index = cpu_to_le32((sw_prod << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((qp->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_RQ); + + /* Flush the writes to HW Rx WQE before the ringing Rx DB */ + wmb(); + __iowrite64_copy(qp->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_post_recv(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct rq_wqe *rqe, **rqe_ptr; + struct sq_sge *hw_sge; + u32 sw_prod; + int i, rc = 0; + + if (qp->state == CMDQ_MODIFY_QP_NEW_STATE_ERR) { + dev_err(&rq->hwq.pdev->dev, + "QPLIB: FP: QP (0x%x) is in the 0x%x state", + qp->id, qp->state); + rc = -EINVAL; + goto done; + } + if (HWQ_CMP((rq->hwq.prod + 1), &rq->hwq) == + HWQ_CMP(rq->hwq.cons, &rq->hwq)) { + dev_err(&rq->hwq.pdev->dev, + "QPLIB: FP: QP (0x%x) RQ is full!", qp->id); + rc = -EINVAL; + goto done; + } + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + rq->swq[sw_prod].wr_id = wqe->wr_id; + + rqe_ptr = (struct rq_wqe **)rq->hwq.pbl_ptr; + rqe = &rqe_ptr[RQE_PG(sw_prod)][RQE_IDX(sw_prod)]; + + memset(rqe, 0, BNXT_QPLIB_MAX_RQE_ENTRY_SIZE); + + /* Calculate wqe_size16 and data_len */ + for (i = 0, hw_sge = (struct sq_sge *)rqe->data; + i < wqe->num_sge; i++, hw_sge++) { + hw_sge->va_or_pa = cpu_to_le64(wqe->sg_list[i].addr); + hw_sge->l_key = cpu_to_le32(wqe->sg_list[i].lkey); + hw_sge->size = cpu_to_le32(wqe->sg_list[i].size); + } + rqe->wqe_type = wqe->type; + rqe->flags = wqe->flags; + rqe->wqe_size = wqe->num_sge + + ((offsetof(typeof(*rqe), data) + 15) >> 4); + + /* Supply the rqe->wr_id index to the wr_id_tbl for now */ + rqe->wr_id[0] = cpu_to_le32(sw_prod); + + rq->hwq.prod++; +done: + return rc; +} + +/* CQ */ + +/* Spinlock must be held */ +static void bnxt_qplib_arm_cq_enable(struct bnxt_qplib_cq *cq) +{ + struct dbr_dbr db_msg = { 0 }; + + db_msg.type_xid = + cpu_to_le32(((cq->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_CQ_ARMENA); + /* Flush memory writes before enabling the CQ */ + wmb(); + __iowrite64_copy(cq->dbr_base, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +static void bnxt_qplib_arm_cq(struct bnxt_qplib_cq *cq, u32 arm_type) +{ + struct bnxt_qplib_hwq *cq_hwq = &cq->hwq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_cons; + + /* Ring DB */ + sw_cons = HWQ_CMP(cq_hwq->cons, cq_hwq); + db_msg.index = cpu_to_le32((sw_cons << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((cq->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + arm_type); + /* flush memory writes before arming the CQ */ + wmb(); + __iowrite64_copy(cq->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_cq req; + struct creq_create_cq_resp *resp; + struct bnxt_qplib_pbl *pbl; + u16 cmd_flags = 0; + int rc; + + cq->hwq.max_elements = cq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &cq->hwq, cq->sghead, + cq->nmap, &cq->hwq.max_elements, + BNXT_QPLIB_MAX_CQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + RCFW_CMD_PREP(req, CREATE_CQ, cmd_flags); + + if (!cq->dpi) { + dev_err(&rcfw->pdev->dev, + "QPLIB: FP: CREATE_CQ failed due to NULL DPI"); + return -EINVAL; + } + req.dpi = cpu_to_le32(cq->dpi->dpi); + req.cq_handle = cpu_to_le64(cq->cq_handle); + + req.cq_size = cpu_to_le32(cq->hwq.max_elements); + pbl = &cq->hwq.pbl[PBL_LVL_0]; + req.pg_size_lvl = cpu_to_le32( + ((cq->hwq.level & CMDQ_CREATE_CQ_LVL_MASK) << + CMDQ_CREATE_CQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? CMDQ_CREATE_CQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? CMDQ_CREATE_CQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? CMDQ_CREATE_CQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? CMDQ_CREATE_CQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? CMDQ_CREATE_CQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? CMDQ_CREATE_CQ_PG_SIZE_PG_1G : + CMDQ_CREATE_CQ_PG_SIZE_PG_4K)); + + req.pbl = cpu_to_le64(pbl->pg_map_arr[0]); + + req.cq_fco_cnq_id = cpu_to_le32( + (cq->cnq_hw_ring_id & CMDQ_CREATE_CQ_CNQ_ID_MASK) << + CMDQ_CREATE_CQ_CNQ_ID_SFT); + + resp = (struct creq_create_cq_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_CQ send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_CQ timed out"); + rc = -ETIMEDOUT; + goto fail; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: CREATE_CQ failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + rc = -EINVAL; + goto fail; + } + cq->id = le32_to_cpu(resp->xid); + cq->dbr_base = res->dpi_tbl.dbr_bar_reg_iomem; + cq->period = BNXT_QPLIB_QUEUE_START_PERIOD; + init_waitqueue_head(&cq->waitq); + + bnxt_qplib_arm_cq_enable(cq); + return 0; + +fail: + bnxt_qplib_free_hwq(res->pdev, &cq->hwq); +exit: + return rc; +} + +int bnxt_qplib_destroy_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_cq req; + struct creq_destroy_cq_resp *resp; + u16 cmd_flags = 0; + + RCFW_CMD_PREP(req, DESTROY_CQ, cmd_flags); + + req.cq_cid = cpu_to_le32(cq->id); + resp = (struct creq_destroy_cq_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_CQ send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_CQ timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: FP: DESTROY_CQ failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + bnxt_qplib_free_hwq(res->pdev, &cq->hwq); + return 0; +} + +static int __flush_sq(struct bnxt_qplib_q *sq, struct bnxt_qplib_qp *qp, + struct bnxt_qplib_cqe **pcqe, int *budget) +{ + u32 sw_prod, sw_cons; + struct bnxt_qplib_cqe *cqe; + int rc = 0; + + /* Now complete all outstanding SQEs with FLUSHED_ERR */ + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_cons == sw_prod) { + sq->flush_in_progress = false; + break; + } + memset(cqe, 0, sizeof(*cqe)); + cqe->status = CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR; + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->wr_id = sq->swq[sw_cons].wr_id; + cqe->src_qp = qp->id; + cqe->type = sq->swq[sw_cons].type; + cqe++; + (*budget)--; + sq->hwq.cons++; + } + *pcqe = cqe; + if (!(*budget) && HWQ_CMP(sq->hwq.cons, &sq->hwq) != sw_prod) + /* Out of budget */ + rc = -EAGAIN; + + return rc; +} + +static int __flush_rq(struct bnxt_qplib_q *rq, struct bnxt_qplib_qp *qp, + int opcode, struct bnxt_qplib_cqe **pcqe, int *budget) +{ + struct bnxt_qplib_cqe *cqe; + u32 sw_prod, sw_cons; + int rc = 0; + + /* Flush the rest of the RQ */ + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(rq->hwq.cons, &rq->hwq); + if (sw_cons == sw_prod) + break; + memset(cqe, 0, sizeof(*cqe)); + cqe->status = + CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR; + cqe->opcode = opcode; + cqe->qp_handle = (unsigned long)qp; + cqe->wr_id = rq->swq[sw_cons].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + } + *pcqe = cqe; + if (!*budget && HWQ_CMP(rq->hwq.cons, &rq->hwq) != sw_prod) + /* Out of budget */ + rc = -EAGAIN; + + return rc; +} + +static int bnxt_qplib_cq_process_req(struct bnxt_qplib_cq *cq, + struct cq_req *hwcqe, + struct bnxt_qplib_cqe **pcqe, int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *sq; + struct bnxt_qplib_cqe *cqe; + u32 sw_cons, cqe_cons; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: Process Req qp is NULL"); + return -EINVAL; + } + sq = &qp->sq; + + cqe_cons = HWQ_CMP(le16_to_cpu(hwcqe->sq_cons_idx), &sq->hwq); + if (cqe_cons > sq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process req reported "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: sq_cons_idx 0x%x which exceeded max 0x%x", + cqe_cons, sq->hwq.max_elements); + return -EINVAL; + } + /* If we were in the middle of flushing the SQ, continue */ + if (sq->flush_in_progress) + goto flush; + + /* Require to walk the sq's swq to fabricate CQEs for all previously + * signaled SWQEs due to CQE aggregation from the current sq cons + * to the cqe_cons + */ + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_cons == cqe_cons) + break; + memset(cqe, 0, sizeof(*cqe)); + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->src_qp = qp->id; + cqe->wr_id = sq->swq[sw_cons].wr_id; + cqe->type = sq->swq[sw_cons].type; + + /* For the last CQE, check for status. For errors, regardless + * of the request being signaled or not, it must complete with + * the hwcqe error status + */ + if (HWQ_CMP((sw_cons + 1), &sq->hwq) == cqe_cons && + hwcqe->status != CQ_REQ_STATUS_OK) { + cqe->status = hwcqe->status; + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Processed Req "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id[%d] = 0x%llx with status 0x%x", + sw_cons, cqe->wr_id, cqe->status); + cqe++; + (*budget)--; + sq->flush_in_progress = true; + /* Must block new posting of SQ and RQ */ + qp->state = CMDQ_MODIFY_QP_NEW_STATE_ERR; + } else { + if (sq->swq[sw_cons].flags & + SQ_SEND_FLAGS_SIGNAL_COMP) { + cqe->status = CQ_REQ_STATUS_OK; + cqe++; + (*budget)--; + } + } + sq->hwq.cons++; + } + *pcqe = cqe; + if (!*budget && HWQ_CMP(sq->hwq.cons, &sq->hwq) != cqe_cons) { + /* Out of budget */ + rc = -EAGAIN; + goto done; + } + if (!sq->flush_in_progress) + goto done; +flush: + /* Require to walk the sq's swq to fabricate CQEs for all + * previously posted SWQEs due to the error CQE received + */ + rc = __flush_sq(sq, qp, pcqe, budget); + if (!rc) + sq->flush_in_progress = false; +done: + return rc; +} + +static int bnxt_qplib_cq_process_res_rc(struct bnxt_qplib_cq *cq, + struct cq_res_rc *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: process_cq RC qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->length = le32_to_cpu(hwcqe->length); + cqe->invrkey = le32_to_cpu(hwcqe->imm_data_or_inv_r_key); + cqe->mr_handle = le64_to_cpu(hwcqe->mr_handle); + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->status = hwcqe->status; + cqe->qp_handle = (u64)(unsigned long)qp; + + wr_id_idx = le32_to_cpu(hwcqe->srq_or_rq_wr_id) & + CQ_RES_RC_SRQ_OR_RQ_WR_ID_MASK; + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process RC "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id idx 0x%x exceeded RQ max 0x%x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_RC, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +static int bnxt_qplib_cq_process_res_ud(struct bnxt_qplib_cq *cq, + struct cq_res_ud *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: process_cq UD qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->length = le32_to_cpu(hwcqe->length); + cqe->invrkey = le32_to_cpu(hwcqe->imm_data); + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->status = hwcqe->status; + cqe->qp_handle = (u64)(unsigned long)qp; + memcpy(cqe->smac, hwcqe->src_mac, 6); + wr_id_idx = le32_to_cpu(hwcqe->src_qp_high_srq_or_rq_wr_id) + & CQ_RES_UD_SRQ_OR_RQ_WR_ID_MASK; + cqe->src_qp = le16_to_cpu(hwcqe->src_qp_low) | + ((le32_to_cpu( + hwcqe->src_qp_high_srq_or_rq_wr_id) & + CQ_RES_UD_SRC_QP_HIGH_MASK) >> 8); + + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process UD "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id idx %#x exceeded RQ max %#x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_UD, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +static int bnxt_qplib_cq_process_res_raweth_qp1(struct bnxt_qplib_cq *cq, + struct cq_res_raweth_qp1 *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cq Raw/QP1 qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->qp_handle = (u64)(unsigned long)qp; + + wr_id_idx = + le32_to_cpu(hwcqe->raweth_qp1_payload_offset_srq_or_rq_wr_id) + & CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_MASK; + cqe->src_qp = qp->id; + if (qp->id == 1 && !cqe->length) { + /* Add workaround for the length misdetection */ + cqe->length = 296; + } else { + cqe->length = le16_to_cpu(hwcqe->length); + } + cqe->pkey_index = qp->pkey_index; + memcpy(cqe->smac, qp->smac, 6); + + cqe->raweth_qp1_flags = le16_to_cpu(hwcqe->raweth_qp1_flags); + cqe->raweth_qp1_flags2 = le32_to_cpu(hwcqe->raweth_qp1_flags2); + + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process Raw/QP1 RQ wr_id "); + dev_err(&cq->hwq.pdev->dev, "QPLIB: ix 0x%x exceeded RQ max 0x%x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_RAWETH_QP1, pcqe, + budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +static int bnxt_qplib_cq_process_terminal(struct bnxt_qplib_cq *cq, + struct cq_terminal *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *sq, *rq; + struct bnxt_qplib_cqe *cqe; + u32 sw_cons = 0, cqe_cons; + int rc = 0; + u8 opcode = 0; + + /* Check the Status */ + if (hwcqe->status != CQ_TERMINAL_STATUS_OK) + dev_warn(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process Terminal Error status = 0x%x", + hwcqe->status); + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process terminal qp is NULL"); + return -EINVAL; + } + /* Must block new posting of SQ and RQ */ + qp->state = CMDQ_MODIFY_QP_NEW_STATE_ERR; + + sq = &qp->sq; + rq = &qp->rq; + + cqe_cons = le16_to_cpu(hwcqe->sq_cons_idx); + if (cqe_cons == 0xFFFF) + goto do_rq; + + if (cqe_cons > sq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process terminal reported "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: sq_cons_idx 0x%x which exceeded max 0x%x", + cqe_cons, sq->hwq.max_elements); + goto do_rq; + } + /* If we were in the middle of flushing, continue */ + if (sq->flush_in_progress) + goto flush_sq; + + /* Terminal CQE can also include aggregated successful CQEs prior. + * So we must complete all CQEs from the current sq's cons to the + * cq_cons with status OK + */ + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_cons == cqe_cons) + break; + if (sq->swq[sw_cons].flags & SQ_SEND_FLAGS_SIGNAL_COMP) { + memset(cqe, 0, sizeof(*cqe)); + cqe->status = CQ_REQ_STATUS_OK; + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->src_qp = qp->id; + cqe->wr_id = sq->swq[sw_cons].wr_id; + cqe->type = sq->swq[sw_cons].type; + cqe++; + (*budget)--; + } + sq->hwq.cons++; + } + *pcqe = cqe; + if (!(*budget) && sw_cons != cqe_cons) { + /* Out of budget */ + rc = -EAGAIN; + goto sq_done; + } + sq->flush_in_progress = true; +flush_sq: + rc = __flush_sq(sq, qp, pcqe, budget); + if (!rc) + sq->flush_in_progress = false; +sq_done: + if (rc) + return rc; +do_rq: + cqe_cons = le16_to_cpu(hwcqe->rq_cons_idx); + if (cqe_cons == 0xFFFF) { + goto done; + } else if (cqe_cons > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Processed terminal "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: reported rq_cons_idx 0x%x exceeds max 0x%x", + cqe_cons, rq->hwq.max_elements); + goto done; + } + /* Terminal CQE requires all posted RQEs to complete with FLUSHED_ERR + * from the current rq->cons to the rq->prod regardless what the + * rq->cons the terminal CQE indicates + */ + rq->flush_in_progress = true; + switch (qp->type) { + case CMDQ_CREATE_QP1_TYPE_GSI: + opcode = CQ_BASE_CQE_TYPE_RES_RAWETH_QP1; + break; + case CMDQ_CREATE_QP_TYPE_RC: + opcode = CQ_BASE_CQE_TYPE_RES_RC; + break; + case CMDQ_CREATE_QP_TYPE_UD: + opcode = CQ_BASE_CQE_TYPE_RES_UD; + break; + } + + rc = __flush_rq(rq, qp, opcode, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; +done: + return rc; +} + +static int bnxt_qplib_cq_process_cutoff(struct bnxt_qplib_cq *cq, + struct cq_cutoff *hwcqe) +{ + /* Check the Status */ + if (hwcqe->status != CQ_CUTOFF_STATUS_OK) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process Cutoff Error status = 0x%x", + hwcqe->status); + return -EINVAL; + } + clear_bit(CQ_FLAGS_RESIZE_IN_PROG, &cq->flags); + wake_up_interruptible(&cq->waitq); + + return 0; +} + +int bnxt_qplib_poll_cq(struct bnxt_qplib_cq *cq, struct bnxt_qplib_cqe *cqe, + int num_cqes) +{ + struct cq_base *hw_cqe, **hw_cqe_ptr; + unsigned long flags; + u32 sw_cons, raw_cons; + int budget, rc = 0; + + spin_lock_irqsave(&cq->hwq.lock, flags); + raw_cons = cq->hwq.cons; + budget = num_cqes; + + while (budget) { + sw_cons = HWQ_CMP(raw_cons, &cq->hwq); + hw_cqe_ptr = (struct cq_base **)cq->hwq.pbl_ptr; + hw_cqe = &hw_cqe_ptr[CQE_PG(sw_cons)][CQE_IDX(sw_cons)]; + + /* Check for Valid bit */ + if (!CQE_CMP_VALID(hw_cqe, raw_cons, cq->hwq.max_elements)) + break; + + /* From the device's respective CQE format to qplib_wc*/ + switch (hw_cqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK) { + case CQ_BASE_CQE_TYPE_REQ: + rc = bnxt_qplib_cq_process_req(cq, + (struct cq_req *)hw_cqe, + &cqe, &budget); + break; + case CQ_BASE_CQE_TYPE_RES_RC: + rc = bnxt_qplib_cq_process_res_rc(cq, + (struct cq_res_rc *) + hw_cqe, &cqe, + &budget); + break; + case CQ_BASE_CQE_TYPE_RES_UD: + rc = bnxt_qplib_cq_process_res_ud + (cq, (struct cq_res_ud *)hw_cqe, &cqe, + &budget); + break; + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + rc = bnxt_qplib_cq_process_res_raweth_qp1 + (cq, (struct cq_res_raweth_qp1 *) + hw_cqe, &cqe, &budget); + break; + case CQ_BASE_CQE_TYPE_TERMINAL: + rc = bnxt_qplib_cq_process_terminal + (cq, (struct cq_terminal *)hw_cqe, + &cqe, &budget); + break; + case CQ_BASE_CQE_TYPE_CUT_OFF: + bnxt_qplib_cq_process_cutoff + (cq, (struct cq_cutoff *)hw_cqe); + /* Done processing this CQ */ + goto exit; + default: + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cq unknown type 0x%lx", + hw_cqe->cqe_type_toggle & + CQ_BASE_CQE_TYPE_MASK); + rc = -EINVAL; + break; + } + if (rc < 0) { + if (rc == -EAGAIN) + break; + /* Error while processing the CQE, just skip to the + * next one + */ + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cqe error rc = 0x%x", rc); + } + raw_cons++; + } + if (cq->hwq.cons != raw_cons) { + cq->hwq.cons = raw_cons; + bnxt_qplib_arm_cq(cq, DBR_DBR_TYPE_CQ); + } +exit: + spin_unlock_irqrestore(&cq->hwq.lock, flags); + return num_cqes - budget; +} + +void bnxt_qplib_req_notify_cq(struct bnxt_qplib_cq *cq, u32 arm_type) +{ + unsigned long flags; + + spin_lock_irqsave(&cq->hwq.lock, flags); + if (arm_type) + bnxt_qplib_arm_cq(cq, arm_type); + + spin_unlock_irqrestore(&cq->hwq.lock, flags); +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.h b/drivers/infiniband/hw/bnxt_re/qplib_fp.h new file mode 100644 index 000000000000..f0150f8da1e3 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.h @@ -0,0 +1,439 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Fast Path Operators (header) + */ + +#ifndef __BNXT_QPLIB_FP_H__ +#define __BNXT_QPLIB_FP_H__ + +struct bnxt_qplib_sge { + u64 addr; + u32 lkey; + u32 size; +}; + +#define BNXT_QPLIB_MAX_SQE_ENTRY_SIZE sizeof(struct sq_send) + +#define SQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_SQE_ENTRY_SIZE) +#define SQE_MAX_IDX_PER_PG (SQE_CNT_PER_PG - 1) + +static inline u32 get_sqe_pg(u32 val) +{ + return ((val & ~SQE_MAX_IDX_PER_PG) / SQE_CNT_PER_PG); +} + +static inline u32 get_sqe_idx(u32 val) +{ + return (val & SQE_MAX_IDX_PER_PG); +} + +#define BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE sizeof(struct sq_psn_search) + +#define PSNE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE) +#define PSNE_MAX_IDX_PER_PG (PSNE_CNT_PER_PG - 1) + +static inline u32 get_psne_pg(u32 val) +{ + return ((val & ~PSNE_MAX_IDX_PER_PG) / PSNE_CNT_PER_PG); +} + +static inline u32 get_psne_idx(u32 val) +{ + return (val & PSNE_MAX_IDX_PER_PG); +} + +#define BNXT_QPLIB_QP_MAX_SGL 6 + +struct bnxt_qplib_swq { + u64 wr_id; + u8 type; + u8 flags; + u32 start_psn; + u32 next_psn; + struct sq_psn_search *psn_search; +}; + +struct bnxt_qplib_swqe { + /* General */ + u64 wr_id; + u8 reqs_type; + u8 type; +#define BNXT_QPLIB_SWQE_TYPE_SEND 0 +#define BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM 1 +#define BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV 2 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE 4 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM 5 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_READ 6 +#define BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP 8 +#define BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD 11 +#define BNXT_QPLIB_SWQE_TYPE_LOCAL_INV 12 +#define BNXT_QPLIB_SWQE_TYPE_FAST_REG_MR 13 +#define BNXT_QPLIB_SWQE_TYPE_REG_MR 13 +#define BNXT_QPLIB_SWQE_TYPE_BIND_MW 14 +#define BNXT_QPLIB_SWQE_TYPE_RECV 128 +#define BNXT_QPLIB_SWQE_TYPE_RECV_RDMA_IMM 129 + u8 flags; +#define BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP BIT(0) +#define BNXT_QPLIB_SWQE_FLAGS_RD_ATOMIC_FENCE BIT(1) +#define BNXT_QPLIB_SWQE_FLAGS_UC_FENCE BIT(2) +#define BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT BIT(3) +#define BNXT_QPLIB_SWQE_FLAGS_INLINE BIT(4) + struct bnxt_qplib_sge sg_list[BNXT_QPLIB_QP_MAX_SGL]; + int num_sge; + /* Max inline data is 96 bytes */ + u32 inline_len; +#define BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH 96 + u8 inline_data[BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH]; + + union { + /* Send, with imm, inval key */ + struct { + union { + __be32 imm_data; + u32 inv_key; + }; + u32 q_key; + u32 dst_qp; + u16 avid; + } send; + + /* Send Raw Ethernet and QP1 */ + struct { + u16 lflags; + u16 cfa_action; + u32 cfa_meta; + } rawqp1; + + /* RDMA write, with imm, read */ + struct { + union { + __be32 imm_data; + u32 inv_key; + }; + u64 remote_va; + u32 r_key; + } rdma; + + /* Atomic cmp/swap, fetch/add */ + struct { + u64 remote_va; + u32 r_key; + u64 swap_data; + u64 cmp_data; + } atomic; + + /* Local Invalidate */ + struct { + u32 inv_l_key; + } local_inv; + + /* FR-PMR */ + struct { + u8 access_cntl; + u8 pg_sz_log; + bool zero_based; + u32 l_key; + u32 length; + u8 pbl_pg_sz_log; +#define BNXT_QPLIB_SWQE_PAGE_SIZE_4K 0 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_8K 1 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_64K 4 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_256K 6 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_1M 8 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_2M 9 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_4M 10 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_1G 18 + u8 levels; +#define PAGE_SHIFT_4K 12 + __le64 *pbl_ptr; + dma_addr_t pbl_dma_ptr; + u64 *page_list; + u16 page_list_len; + u64 va; + } frmr; + + /* Bind */ + struct { + u8 access_cntl; +#define BNXT_QPLIB_BIND_SWQE_ACCESS_LOCAL_WRITE BIT(0) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_READ BIT(1) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_WRITE BIT(2) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_ATOMIC BIT(3) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_WINDOW_BIND BIT(4) + bool zero_based; + u8 mw_type; + u32 parent_l_key; + u32 r_key; + u64 va; + u32 length; + } bind; + }; +}; + +#define BNXT_QPLIB_MAX_RQE_ENTRY_SIZE sizeof(struct rq_wqe) + +#define RQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_RQE_ENTRY_SIZE) +#define RQE_MAX_IDX_PER_PG (RQE_CNT_PER_PG - 1) +#define RQE_PG(x) (((x) & ~RQE_MAX_IDX_PER_PG) / RQE_CNT_PER_PG) +#define RQE_IDX(x) ((x) & RQE_MAX_IDX_PER_PG) + +struct bnxt_qplib_q { + struct bnxt_qplib_hwq hwq; + struct bnxt_qplib_swq *swq; + struct scatterlist *sglist; + u32 nmap; + u32 max_wqe; + u16 max_sge; + u32 psn; + bool flush_in_progress; +}; + +struct bnxt_qplib_qp { + struct bnxt_qplib_pd *pd; + struct bnxt_qplib_dpi *dpi; + u64 qp_handle; + u32 id; + u8 type; + u8 sig_type; + u32 modify_flags; + u8 state; + u8 cur_qp_state; + u32 max_inline_data; + u32 mtu; + u8 path_mtu; + bool en_sqd_async_notify; + u16 pkey_index; + u32 qkey; + u32 dest_qp_id; + u8 access; + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u32 min_rnr_timer; + u32 max_rd_atomic; + u32 max_dest_rd_atomic; + u32 dest_qpn; + u8 smac[6]; + u16 vlan_id; + u8 nw_type; + struct bnxt_qplib_ah ah; + +#define BTH_PSN_MASK ((1 << 24) - 1) + /* SQ */ + struct bnxt_qplib_q sq; + /* RQ */ + struct bnxt_qplib_q rq; + /* SRQ */ + struct bnxt_qplib_srq *srq; + /* CQ */ + struct bnxt_qplib_cq *scq; + struct bnxt_qplib_cq *rcq; + /* IRRQ and ORRQ */ + struct bnxt_qplib_hwq irrq; + struct bnxt_qplib_hwq orrq; + /* Header buffer for QP1 */ + int sq_hdr_buf_size; + int rq_hdr_buf_size; +/* + * Buffer space for ETH(14), IP or GRH(40), UDP header(8) + * and ib_bth + ib_deth (20). + * Max required is 82 when RoCE V2 is enabled + */ +#define BNXT_QPLIB_MAX_QP1_SQ_HDR_SIZE_V2 86 + /* Ethernet header = 14 */ + /* ib_grh = 40 (provided by MAD) */ + /* ib_bth + ib_deth = 20 */ + /* MAD = 256 (provided by MAD) */ + /* iCRC = 4 */ +#define BNXT_QPLIB_MAX_QP1_RQ_ETH_HDR_SIZE 14 +#define BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2 512 +#define BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV4 20 +#define BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6 40 +#define BNXT_QPLIB_MAX_QP1_RQ_BDETH_HDR_SIZE 20 + void *sq_hdr_buf; + dma_addr_t sq_hdr_buf_map; + void *rq_hdr_buf; + dma_addr_t rq_hdr_buf_map; +}; + +#define BNXT_QPLIB_MAX_CQE_ENTRY_SIZE sizeof(struct cq_base) + +#define CQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_CQE_ENTRY_SIZE) +#define CQE_MAX_IDX_PER_PG (CQE_CNT_PER_PG - 1) +#define CQE_PG(x) (((x) & ~CQE_MAX_IDX_PER_PG) / CQE_CNT_PER_PG) +#define CQE_IDX(x) ((x) & CQE_MAX_IDX_PER_PG) + +#define ROCE_CQE_CMP_V 0 +#define CQE_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!((hdr)->cqe_type_toggle & CQ_BASE_TOGGLE) == \ + !((raw_cons) & (cp_bit))) + +struct bnxt_qplib_cqe { + u8 status; + u8 type; + u8 opcode; + u32 length; + u64 wr_id; + union { + __be32 immdata; + u32 invrkey; + }; + u64 qp_handle; + u64 mr_handle; + u16 flags; + u8 smac[6]; + u32 src_qp; + u16 raweth_qp1_flags; + u16 raweth_qp1_errors; + u16 raweth_qp1_cfa_code; + u32 raweth_qp1_flags2; + u32 raweth_qp1_metadata; + u8 raweth_qp1_payload_offset; + u16 pkey_index; +}; + +#define BNXT_QPLIB_QUEUE_START_PERIOD 0x01 +struct bnxt_qplib_cq { + struct bnxt_qplib_dpi *dpi; + void __iomem *dbr_base; + u32 max_wqe; + u32 id; + u16 count; + u16 period; + struct bnxt_qplib_hwq hwq; + u32 cnq_hw_ring_id; + bool resize_in_progress; + struct scatterlist *sghead; + u32 nmap; + u64 cq_handle; + +#define CQ_RESIZE_WAIT_TIME_MS 500 + unsigned long flags; +#define CQ_FLAGS_RESIZE_IN_PROG 1 + wait_queue_head_t waitq; +}; + +#define BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE sizeof(struct xrrq_irrq) +#define BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE sizeof(struct xrrq_orrq) +#define IRD_LIMIT_TO_IRRQ_SLOTS(x) (2 * (x) + 2) +#define IRRQ_SLOTS_TO_IRD_LIMIT(s) (((s) >> 1) - 1) +#define ORD_LIMIT_TO_ORRQ_SLOTS(x) ((x) + 1) +#define ORRQ_SLOTS_TO_ORD_LIMIT(s) ((s) - 1) + +#define BNXT_QPLIB_MAX_NQE_ENTRY_SIZE sizeof(struct nq_base) + +#define NQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_NQE_ENTRY_SIZE) +#define NQE_MAX_IDX_PER_PG (NQE_CNT_PER_PG - 1) +#define NQE_PG(x) (((x) & ~NQE_MAX_IDX_PER_PG) / NQE_CNT_PER_PG) +#define NQE_IDX(x) ((x) & NQE_MAX_IDX_PER_PG) + +#define NQE_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!(le32_to_cpu((hdr)->info63_v[0]) & NQ_BASE_V) == \ + !((raw_cons) & (cp_bit))) + +#define BNXT_QPLIB_NQE_MAX_CNT (128 * 1024) + +#define NQ_CONS_PCI_BAR_REGION 2 +#define NQ_DB_KEY_CP (0x2 << CMPL_DOORBELL_KEY_SFT) +#define NQ_DB_IDX_VALID CMPL_DOORBELL_IDX_VALID +#define NQ_DB_IRQ_DIS CMPL_DOORBELL_MASK +#define NQ_DB_CP_FLAGS_REARM (NQ_DB_KEY_CP | \ + NQ_DB_IDX_VALID) +#define NQ_DB_CP_FLAGS (NQ_DB_KEY_CP | \ + NQ_DB_IDX_VALID | \ + NQ_DB_IRQ_DIS) +#define NQ_DB_REARM(db, raw_cons, cp_bit) \ + writel(NQ_DB_CP_FLAGS_REARM | ((raw_cons) & ((cp_bit) - 1)), db) +#define NQ_DB(db, raw_cons, cp_bit) \ + writel(NQ_DB_CP_FLAGS | ((raw_cons) & ((cp_bit) - 1)), db) + +struct bnxt_qplib_nq { + struct pci_dev *pdev; + + int vector; + int budget; + bool requested; + struct tasklet_struct worker; + struct bnxt_qplib_hwq hwq; + + u16 bar_reg; + u16 bar_reg_off; + u16 ring_id; + void __iomem *bar_reg_iomem; + + int (*cqn_handler) + (struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *cq); + int (*srqn_handler) + (struct bnxt_qplib_nq *nq, + void *srq, + u8 event); +}; + +void bnxt_qplib_disable_nq(struct bnxt_qplib_nq *nq); +int bnxt_qplib_enable_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq, + int msix_vector, int bar_reg_offset, + int (*cqn_handler)(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *cq), + int (*srqn_handler)(struct bnxt_qplib_nq *nq, + void *srq, + u8 event)); +int bnxt_qplib_create_qp1(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_modify_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_destroy_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +void *bnxt_qplib_get_qp1_sq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge); +void *bnxt_qplib_get_qp1_rq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge); +u32 bnxt_qplib_get_rq_prod_index(struct bnxt_qplib_qp *qp); +dma_addr_t bnxt_qplib_get_qp_buf_from_index(struct bnxt_qplib_qp *qp, + u32 index); +void bnxt_qplib_post_send_db(struct bnxt_qplib_qp *qp); +int bnxt_qplib_post_send(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe); +void bnxt_qplib_post_recv_db(struct bnxt_qplib_qp *qp); +int bnxt_qplib_post_recv(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe); +int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq); +int bnxt_qplib_destroy_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq); +int bnxt_qplib_poll_cq(struct bnxt_qplib_cq *cq, struct bnxt_qplib_cqe *cqe, + int num); +void bnxt_qplib_req_notify_cq(struct bnxt_qplib_cq *cq, u32 arm_type); +void bnxt_qplib_free_nq(struct bnxt_qplib_nq *nq); +int bnxt_qplib_alloc_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq); +#endif /* __BNXT_QPLIB_FP_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c new file mode 100644 index 000000000000..23fb7260662b --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c @@ -0,0 +1,694 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RDMA Controller HW interface + */ +#include +#include +#include +#include +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_rcfw.h" +static void bnxt_qplib_service_creq(unsigned long data); + +/* Hardware communication channel */ +int bnxt_qplib_rcfw_wait_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie) +{ + u16 cbit; + int rc; + + cookie &= RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (!test_bit(cbit, rcfw->cmdq_bitmap)) + dev_warn(&rcfw->pdev->dev, + "QPLIB: CMD bit %d for cookie 0x%x is not set?", + cbit, cookie); + + rc = wait_event_timeout(rcfw->waitq, + !test_bit(cbit, rcfw->cmdq_bitmap), + msecs_to_jiffies(RCFW_CMD_WAIT_TIME_MS)); + if (!rc) { + dev_warn(&rcfw->pdev->dev, + "QPLIB: Bono Error: timeout %d msec, msg {0x%x}\n", + RCFW_CMD_WAIT_TIME_MS, cookie); + } + + return rc; +}; + +int bnxt_qplib_rcfw_block_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie) +{ + u32 count = -1; + u16 cbit; + + cookie &= RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (!test_bit(cbit, rcfw->cmdq_bitmap)) + goto done; + do { + bnxt_qplib_service_creq((unsigned long)rcfw); + } while (test_bit(cbit, rcfw->cmdq_bitmap) && --count); +done: + return count; +}; + +void *bnxt_qplib_rcfw_send_message(struct bnxt_qplib_rcfw *rcfw, + struct cmdq_base *req, void **crsbe, + u8 is_block) +{ + struct bnxt_qplib_crsq *crsq = &rcfw->crsq; + struct bnxt_qplib_cmdqe *cmdqe, **cmdq_ptr; + struct bnxt_qplib_hwq *cmdq = &rcfw->cmdq; + struct bnxt_qplib_hwq *crsb = &rcfw->crsb; + struct bnxt_qplib_crsqe *crsqe = NULL; + struct bnxt_qplib_crsbe **crsb_ptr; + u32 sw_prod, cmdq_prod; + u8 retry_cnt = 0xFF; + dma_addr_t dma_addr; + unsigned long flags; + u32 size, opcode; + u16 cookie, cbit; + int pg, idx; + u8 *preq; + +retry: + opcode = req->opcode; + if (!test_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags) && + (opcode != CMDQ_BASE_OPCODE_QUERY_FUNC && + opcode != CMDQ_BASE_OPCODE_INITIALIZE_FW)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW not initialized, reject opcode 0x%x", + opcode); + return NULL; + } + + if (test_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags) && + opcode == CMDQ_BASE_OPCODE_INITIALIZE_FW) { + dev_err(&rcfw->pdev->dev, "QPLIB: RCFW already initialized!"); + return NULL; + } + + /* Cmdq are in 16-byte units, each request can consume 1 or more + * cmdqe + */ + spin_lock_irqsave(&cmdq->lock, flags); + if (req->cmd_size > cmdq->max_elements - + ((HWQ_CMP(cmdq->prod, cmdq) - HWQ_CMP(cmdq->cons, cmdq)) & + (cmdq->max_elements - 1))) { + dev_err(&rcfw->pdev->dev, "QPLIB: RCFW: CMDQ is full!"); + spin_unlock_irqrestore(&cmdq->lock, flags); + + if (!retry_cnt--) + return NULL; + goto retry; + } + + retry_cnt = 0xFF; + + cookie = atomic_inc_return(&rcfw->seq_num) & RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (is_block) + cookie |= RCFW_CMD_IS_BLOCKING; + req->cookie = cpu_to_le16(cookie); + if (test_and_set_bit(cbit, rcfw->cmdq_bitmap)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW MAX outstanding cmd reached!"); + atomic_dec(&rcfw->seq_num); + spin_unlock_irqrestore(&cmdq->lock, flags); + + if (!retry_cnt--) + return NULL; + goto retry; + } + /* Reserve a resp buffer slot if requested */ + if (req->resp_size && crsbe) { + spin_lock(&crsb->lock); + sw_prod = HWQ_CMP(crsb->prod, crsb); + crsb_ptr = (struct bnxt_qplib_crsbe **)crsb->pbl_ptr; + *crsbe = (void *)&crsb_ptr[get_crsb_pg(sw_prod)] + [get_crsb_idx(sw_prod)]; + bnxt_qplib_crsb_dma_next(crsb->pbl_dma_ptr, sw_prod, &dma_addr); + req->resp_addr = cpu_to_le64(dma_addr); + crsb->prod++; + spin_unlock(&crsb->lock); + + req->resp_size = (sizeof(struct bnxt_qplib_crsbe) + + BNXT_QPLIB_CMDQE_UNITS - 1) / + BNXT_QPLIB_CMDQE_UNITS; + } + cmdq_ptr = (struct bnxt_qplib_cmdqe **)cmdq->pbl_ptr; + preq = (u8 *)req; + size = req->cmd_size * BNXT_QPLIB_CMDQE_UNITS; + do { + pg = 0; + idx = 0; + + /* Locate the next cmdq slot */ + sw_prod = HWQ_CMP(cmdq->prod, cmdq); + cmdqe = &cmdq_ptr[get_cmdq_pg(sw_prod)][get_cmdq_idx(sw_prod)]; + if (!cmdqe) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW request failed with no cmdqe!"); + goto done; + } + /* Copy a segment of the req cmd to the cmdq */ + memset(cmdqe, 0, sizeof(*cmdqe)); + memcpy(cmdqe, preq, min_t(u32, size, sizeof(*cmdqe))); + preq += min_t(u32, size, sizeof(*cmdqe)); + size -= min_t(u32, size, sizeof(*cmdqe)); + cmdq->prod++; + } while (size > 0); + + cmdq_prod = cmdq->prod; + if (rcfw->flags & FIRMWARE_FIRST_FLAG) { + /* The very first doorbell write is required to set this flag + * which prompts the FW to reset its internal pointers + */ + cmdq_prod |= FIRMWARE_FIRST_FLAG; + rcfw->flags &= ~FIRMWARE_FIRST_FLAG; + } + sw_prod = HWQ_CMP(crsq->prod, crsq); + crsqe = &crsq->crsq[sw_prod]; + memset(crsqe, 0, sizeof(*crsqe)); + crsq->prod++; + crsqe->req_size = req->cmd_size; + + /* ring CMDQ DB */ + writel(cmdq_prod, rcfw->cmdq_bar_reg_iomem + + rcfw->cmdq_bar_reg_prod_off); + writel(RCFW_CMDQ_TRIG_VAL, rcfw->cmdq_bar_reg_iomem + + rcfw->cmdq_bar_reg_trig_off); +done: + spin_unlock_irqrestore(&cmdq->lock, flags); + /* Return the CREQ response pointer */ + return crsqe ? &crsqe->qp_event : NULL; +} + +/* Completions */ +static int bnxt_qplib_process_func_event(struct bnxt_qplib_rcfw *rcfw, + struct creq_func_event *func_event) +{ + switch (func_event->event) { + case CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TQM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCS_ERROR: + /* SRQ ctx error, call srq_handler?? + * But there's no SRQ handle! + */ + break; + case CREQ_FUNC_EVENT_EVENT_CFCC_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TIM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_VF_COMM_REQUEST: + break; + case CREQ_FUNC_EVENT_EVENT_RESOURCE_EXHAUSTED: + break; + default: + return -EINVAL; + } + return 0; +} + +static int bnxt_qplib_process_qp_event(struct bnxt_qplib_rcfw *rcfw, + struct creq_qp_event *qp_event) +{ + struct bnxt_qplib_crsq *crsq = &rcfw->crsq; + struct bnxt_qplib_hwq *cmdq = &rcfw->cmdq; + struct bnxt_qplib_crsqe *crsqe; + u16 cbit, cookie, blocked = 0; + unsigned long flags; + u32 sw_cons; + + switch (qp_event->event) { + case CREQ_QP_EVENT_EVENT_QP_ERROR_NOTIFICATION: + dev_dbg(&rcfw->pdev->dev, + "QPLIB: Received QP error notification"); + break; + default: + /* Command Response */ + spin_lock_irqsave(&cmdq->lock, flags); + sw_cons = HWQ_CMP(crsq->cons, crsq); + crsqe = &crsq->crsq[sw_cons]; + crsq->cons++; + memcpy(&crsqe->qp_event, qp_event, sizeof(crsqe->qp_event)); + + cookie = le16_to_cpu(crsqe->qp_event.cookie); + blocked = cookie & RCFW_CMD_IS_BLOCKING; + cookie &= RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (!test_and_clear_bit(cbit, rcfw->cmdq_bitmap)) + dev_warn(&rcfw->pdev->dev, + "QPLIB: CMD bit %d was not requested", cbit); + + cmdq->cons += crsqe->req_size; + spin_unlock_irqrestore(&cmdq->lock, flags); + if (!blocked) + wake_up(&rcfw->waitq); + break; + } + return 0; +} + +/* SP - CREQ Completion handlers */ +static void bnxt_qplib_service_creq(unsigned long data) +{ + struct bnxt_qplib_rcfw *rcfw = (struct bnxt_qplib_rcfw *)data; + struct bnxt_qplib_hwq *creq = &rcfw->creq; + struct creq_base *creqe, **creq_ptr; + u32 sw_cons, raw_cons; + unsigned long flags; + u32 type; + + /* Service the CREQ until empty */ + spin_lock_irqsave(&creq->lock, flags); + raw_cons = creq->cons; + while (1) { + sw_cons = HWQ_CMP(raw_cons, creq); + creq_ptr = (struct creq_base **)creq->pbl_ptr; + creqe = &creq_ptr[get_creq_pg(sw_cons)][get_creq_idx(sw_cons)]; + if (!CREQ_CMP_VALID(creqe, raw_cons, creq->max_elements)) + break; + + type = creqe->type & CREQ_BASE_TYPE_MASK; + switch (type) { + case CREQ_BASE_TYPE_QP_EVENT: + if (!bnxt_qplib_process_qp_event + (rcfw, (struct creq_qp_event *)creqe)) + rcfw->creq_qp_event_processed++; + else { + dev_warn(&rcfw->pdev->dev, "QPLIB: crsqe with"); + dev_warn(&rcfw->pdev->dev, + "QPLIB: type = 0x%x not handled", + type); + } + break; + case CREQ_BASE_TYPE_FUNC_EVENT: + if (!bnxt_qplib_process_func_event + (rcfw, (struct creq_func_event *)creqe)) + rcfw->creq_func_event_processed++; + else + dev_warn + (&rcfw->pdev->dev, "QPLIB:aeqe:%#x Not handled", + type); + break; + default: + dev_warn(&rcfw->pdev->dev, "QPLIB: creqe with "); + dev_warn(&rcfw->pdev->dev, + "QPLIB: op_event = 0x%x not handled", type); + break; + } + raw_cons++; + } + if (creq->cons != raw_cons) { + creq->cons = raw_cons; + CREQ_DB_REARM(rcfw->creq_bar_reg_iomem, raw_cons, + creq->max_elements); + } + spin_unlock_irqrestore(&creq->lock, flags); +} + +static irqreturn_t bnxt_qplib_creq_irq(int irq, void *dev_instance) +{ + struct bnxt_qplib_rcfw *rcfw = dev_instance; + struct bnxt_qplib_hwq *creq = &rcfw->creq; + struct creq_base **creq_ptr; + u32 sw_cons; + + /* Prefetch the CREQ element */ + sw_cons = HWQ_CMP(creq->cons, creq); + creq_ptr = (struct creq_base **)rcfw->creq.pbl_ptr; + prefetch(&creq_ptr[get_creq_pg(sw_cons)][get_creq_idx(sw_cons)]); + + tasklet_schedule(&rcfw->worker); + + return IRQ_HANDLED; +} + +/* RCFW */ +int bnxt_qplib_deinit_rcfw(struct bnxt_qplib_rcfw *rcfw) +{ + struct creq_deinitialize_fw_resp *resp; + struct cmdq_deinitialize_fw req; + u16 cmd_flags = 0; + + RCFW_CMD_PREP(req, DEINITIALIZE_FW, cmd_flags); + resp = (struct creq_deinitialize_fw_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) + return -EINVAL; + + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) + return -ETIMEDOUT; + + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) + return -EFAULT; + + clear_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags); + return 0; +} + +static int __get_pbl_pg_idx(struct bnxt_qplib_pbl *pbl) +{ + return (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_1G : + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K); +} + +int bnxt_qplib_init_rcfw(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_ctx *ctx, int is_virtfn) +{ + struct creq_initialize_fw_resp *resp; + struct cmdq_initialize_fw req; + u16 cmd_flags = 0, level; + + RCFW_CMD_PREP(req, INITIALIZE_FW, cmd_flags); + + /* + * VFs need not setup the HW context area, PF + * shall setup this area for VF. Skipping the + * HW programming + */ + if (is_virtfn) + goto skip_ctx_setup; + + level = ctx->qpc_tbl.level; + req.qpc_pg_size_qpc_lvl = (level << CMDQ_INITIALIZE_FW_QPC_LVL_SFT) | + __get_pbl_pg_idx(&ctx->qpc_tbl.pbl[level]); + level = ctx->mrw_tbl.level; + req.mrw_pg_size_mrw_lvl = (level << CMDQ_INITIALIZE_FW_MRW_LVL_SFT) | + __get_pbl_pg_idx(&ctx->mrw_tbl.pbl[level]); + level = ctx->srqc_tbl.level; + req.srq_pg_size_srq_lvl = (level << CMDQ_INITIALIZE_FW_SRQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->srqc_tbl.pbl[level]); + level = ctx->cq_tbl.level; + req.cq_pg_size_cq_lvl = (level << CMDQ_INITIALIZE_FW_CQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->cq_tbl.pbl[level]); + level = ctx->srqc_tbl.level; + req.srq_pg_size_srq_lvl = (level << CMDQ_INITIALIZE_FW_SRQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->srqc_tbl.pbl[level]); + level = ctx->cq_tbl.level; + req.cq_pg_size_cq_lvl = (level << CMDQ_INITIALIZE_FW_CQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->cq_tbl.pbl[level]); + level = ctx->tim_tbl.level; + req.tim_pg_size_tim_lvl = (level << CMDQ_INITIALIZE_FW_TIM_LVL_SFT) | + __get_pbl_pg_idx(&ctx->tim_tbl.pbl[level]); + level = ctx->tqm_pde_level; + req.tqm_pg_size_tqm_lvl = (level << CMDQ_INITIALIZE_FW_TQM_LVL_SFT) | + __get_pbl_pg_idx(&ctx->tqm_pde.pbl[level]); + + req.qpc_page_dir = + cpu_to_le64(ctx->qpc_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.mrw_page_dir = + cpu_to_le64(ctx->mrw_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.srq_page_dir = + cpu_to_le64(ctx->srqc_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.cq_page_dir = + cpu_to_le64(ctx->cq_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.tim_page_dir = + cpu_to_le64(ctx->tim_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.tqm_page_dir = + cpu_to_le64(ctx->tqm_pde.pbl[PBL_LVL_0].pg_map_arr[0]); + + req.number_of_qp = cpu_to_le32(ctx->qpc_tbl.max_elements); + req.number_of_mrw = cpu_to_le32(ctx->mrw_tbl.max_elements); + req.number_of_srq = cpu_to_le32(ctx->srqc_tbl.max_elements); + req.number_of_cq = cpu_to_le32(ctx->cq_tbl.max_elements); + + req.max_qp_per_vf = cpu_to_le32(ctx->vf_res.max_qp_per_vf); + req.max_mrw_per_vf = cpu_to_le32(ctx->vf_res.max_mrw_per_vf); + req.max_srq_per_vf = cpu_to_le32(ctx->vf_res.max_srq_per_vf); + req.max_cq_per_vf = cpu_to_le32(ctx->vf_res.max_cq_per_vf); + req.max_gid_per_vf = cpu_to_le32(ctx->vf_res.max_gid_per_vf); + +skip_ctx_setup: + req.stat_ctx_id = cpu_to_le32(ctx->stats.fw_id); + resp = (struct creq_initialize_fw_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW: INITIALIZE_FW send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW: INITIALIZE_FW timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW: INITIALIZE_FW failed"); + return -EINVAL; + } + set_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags); + return 0; +} + +void bnxt_qplib_free_rcfw_channel(struct bnxt_qplib_rcfw *rcfw) +{ + bnxt_qplib_free_hwq(rcfw->pdev, &rcfw->crsb); + kfree(rcfw->crsq.crsq); + bnxt_qplib_free_hwq(rcfw->pdev, &rcfw->cmdq); + bnxt_qplib_free_hwq(rcfw->pdev, &rcfw->creq); + + rcfw->pdev = NULL; +} + +int bnxt_qplib_alloc_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw) +{ + rcfw->pdev = pdev; + rcfw->creq.max_elements = BNXT_QPLIB_CREQE_MAX_CNT; + if (bnxt_qplib_alloc_init_hwq(rcfw->pdev, &rcfw->creq, NULL, 0, + &rcfw->creq.max_elements, + BNXT_QPLIB_CREQE_UNITS, 0, PAGE_SIZE, + HWQ_TYPE_L2_CMPL)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: HW channel CREQ allocation failed"); + goto fail; + } + rcfw->cmdq.max_elements = BNXT_QPLIB_CMDQE_MAX_CNT; + if (bnxt_qplib_alloc_init_hwq(rcfw->pdev, &rcfw->cmdq, NULL, 0, + &rcfw->cmdq.max_elements, + BNXT_QPLIB_CMDQE_UNITS, 0, PAGE_SIZE, + HWQ_TYPE_CTX)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: HW channel CMDQ allocation failed"); + goto fail; + } + + rcfw->crsq.max_elements = rcfw->cmdq.max_elements; + rcfw->crsq.crsq = kcalloc(rcfw->crsq.max_elements, + sizeof(*rcfw->crsq.crsq), GFP_KERNEL); + if (!rcfw->crsq.crsq) + goto fail; + + rcfw->crsb.max_elements = BNXT_QPLIB_CRSBE_MAX_CNT; + if (bnxt_qplib_alloc_init_hwq(rcfw->pdev, &rcfw->crsb, NULL, 0, + &rcfw->crsb.max_elements, + BNXT_QPLIB_CRSBE_UNITS, 0, PAGE_SIZE, + HWQ_TYPE_CTX)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: HW channel CRSB allocation failed"); + goto fail; + } + return 0; + +fail: + bnxt_qplib_free_rcfw_channel(rcfw); + return -ENOMEM; +} + +void bnxt_qplib_disable_rcfw_channel(struct bnxt_qplib_rcfw *rcfw) +{ + unsigned long indx; + + /* Make sure the HW channel is stopped! */ + synchronize_irq(rcfw->vector); + tasklet_disable(&rcfw->worker); + tasklet_kill(&rcfw->worker); + + if (rcfw->requested) { + free_irq(rcfw->vector, rcfw); + rcfw->requested = false; + } + if (rcfw->cmdq_bar_reg_iomem) + iounmap(rcfw->cmdq_bar_reg_iomem); + rcfw->cmdq_bar_reg_iomem = NULL; + + if (rcfw->creq_bar_reg_iomem) + iounmap(rcfw->creq_bar_reg_iomem); + rcfw->creq_bar_reg_iomem = NULL; + + indx = find_first_bit(rcfw->cmdq_bitmap, rcfw->bmap_size); + if (indx != rcfw->bmap_size) + dev_err(&rcfw->pdev->dev, + "QPLIB: disabling RCFW with pending cmd-bit %lx", indx); + kfree(rcfw->cmdq_bitmap); + rcfw->bmap_size = 0; + + rcfw->aeq_handler = NULL; + rcfw->vector = 0; +} + +int bnxt_qplib_enable_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw, + int msix_vector, + int cp_bar_reg_off, int virt_fn, + int (*aeq_handler)(struct bnxt_qplib_rcfw *, + struct creq_func_event *)) +{ + resource_size_t res_base; + struct cmdq_init init; + u16 bmap_size; + int rc; + + /* General */ + atomic_set(&rcfw->seq_num, 0); + rcfw->flags = FIRMWARE_FIRST_FLAG; + bmap_size = BITS_TO_LONGS(RCFW_MAX_OUTSTANDING_CMD * + sizeof(unsigned long)); + rcfw->cmdq_bitmap = kzalloc(bmap_size, GFP_KERNEL); + if (!rcfw->cmdq_bitmap) + return -ENOMEM; + rcfw->bmap_size = bmap_size; + + /* CMDQ */ + rcfw->cmdq_bar_reg = RCFW_COMM_PCI_BAR_REGION; + res_base = pci_resource_start(pdev, rcfw->cmdq_bar_reg); + if (!res_base) + return -ENOMEM; + + rcfw->cmdq_bar_reg_iomem = ioremap_nocache(res_base + + RCFW_COMM_BASE_OFFSET, + RCFW_COMM_SIZE); + if (!rcfw->cmdq_bar_reg_iomem) { + dev_err(&rcfw->pdev->dev, + "QPLIB: CMDQ BAR region %d mapping failed", + rcfw->cmdq_bar_reg); + return -ENOMEM; + } + + rcfw->cmdq_bar_reg_prod_off = virt_fn ? RCFW_VF_COMM_PROD_OFFSET : + RCFW_PF_COMM_PROD_OFFSET; + + rcfw->cmdq_bar_reg_trig_off = RCFW_COMM_TRIG_OFFSET; + + /* CRSQ */ + rcfw->crsq.prod = 0; + rcfw->crsq.cons = 0; + + /* CREQ */ + rcfw->creq_bar_reg = RCFW_COMM_CONS_PCI_BAR_REGION; + res_base = pci_resource_start(pdev, rcfw->creq_bar_reg); + if (!res_base) + dev_err(&rcfw->pdev->dev, + "QPLIB: CREQ BAR region %d resc start is 0!", + rcfw->creq_bar_reg); + rcfw->creq_bar_reg_iomem = ioremap_nocache(res_base + cp_bar_reg_off, + 4); + if (!rcfw->creq_bar_reg_iomem) { + dev_err(&rcfw->pdev->dev, + "QPLIB: CREQ BAR region %d mapping failed", + rcfw->creq_bar_reg); + return -ENOMEM; + } + rcfw->creq_qp_event_processed = 0; + rcfw->creq_func_event_processed = 0; + + rcfw->vector = msix_vector; + if (aeq_handler) + rcfw->aeq_handler = aeq_handler; + + tasklet_init(&rcfw->worker, bnxt_qplib_service_creq, + (unsigned long)rcfw); + + rcfw->requested = false; + rc = request_irq(rcfw->vector, bnxt_qplib_creq_irq, 0, + "bnxt_qplib_creq", rcfw); + if (rc) { + dev_err(&rcfw->pdev->dev, + "QPLIB: Failed to request IRQ for CREQ rc = 0x%x", rc); + bnxt_qplib_disable_rcfw_channel(rcfw); + return rc; + } + rcfw->requested = true; + + init_waitqueue_head(&rcfw->waitq); + + CREQ_DB_REARM(rcfw->creq_bar_reg_iomem, 0, rcfw->creq.max_elements); + + init.cmdq_pbl = cpu_to_le64(rcfw->cmdq.pbl[PBL_LVL_0].pg_map_arr[0]); + init.cmdq_size_cmdq_lvl = cpu_to_le16( + ((BNXT_QPLIB_CMDQE_MAX_CNT << CMDQ_INIT_CMDQ_SIZE_SFT) & + CMDQ_INIT_CMDQ_SIZE_MASK) | + ((rcfw->cmdq.level << CMDQ_INIT_CMDQ_LVL_SFT) & + CMDQ_INIT_CMDQ_LVL_MASK)); + init.creq_ring_id = cpu_to_le16(rcfw->creq_ring_id); + + /* Write to the Bono mailbox register */ + __iowrite32_copy(rcfw->cmdq_bar_reg_iomem, &init, sizeof(init) / 4); + return 0; +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h new file mode 100644 index 000000000000..d3567d75bf58 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h @@ -0,0 +1,231 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RDMA Controller HW interface (header) + */ + +#ifndef __BNXT_QPLIB_RCFW_H__ +#define __BNXT_QPLIB_RCFW_H__ + +#define RCFW_CMDQ_TRIG_VAL 1 +#define RCFW_COMM_PCI_BAR_REGION 0 +#define RCFW_COMM_CONS_PCI_BAR_REGION 2 +#define RCFW_COMM_BASE_OFFSET 0x600 +#define RCFW_PF_COMM_PROD_OFFSET 0xc +#define RCFW_VF_COMM_PROD_OFFSET 0xc +#define RCFW_COMM_TRIG_OFFSET 0x100 +#define RCFW_COMM_SIZE 0x104 + +#define RCFW_DBR_PCI_BAR_REGION 2 + +#define RCFW_CMD_PREP(req, CMD, cmd_flags) \ + do { \ + memset(&(req), 0, sizeof((req))); \ + (req).opcode = CMDQ_BASE_OPCODE_##CMD; \ + (req).cmd_size = (sizeof((req)) + \ + BNXT_QPLIB_CMDQE_UNITS - 1) / \ + BNXT_QPLIB_CMDQE_UNITS; \ + (req).flags = cpu_to_le16(cmd_flags); \ + } while (0) + +#define RCFW_CMD_WAIT_TIME_MS 20000 /* 20 Seconds timeout */ + +/* CMDQ elements */ +#define BNXT_QPLIB_CMDQE_MAX_CNT 256 +#define BNXT_QPLIB_CMDQE_UNITS sizeof(struct bnxt_qplib_cmdqe) +#define BNXT_QPLIB_CMDQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_CMDQE_UNITS) + +#define MAX_CMDQ_IDX (BNXT_QPLIB_CMDQE_MAX_CNT - 1) +#define MAX_CMDQ_IDX_PER_PG (BNXT_QPLIB_CMDQE_CNT_PER_PG - 1) + +#define RCFW_MAX_OUTSTANDING_CMD BNXT_QPLIB_CMDQE_MAX_CNT +#define RCFW_MAX_COOKIE_VALUE 0x7FFF +#define RCFW_CMD_IS_BLOCKING 0x8000 + +/* Cmdq contains a fix number of a 16-Byte slots */ +struct bnxt_qplib_cmdqe { + u8 data[16]; +}; + +static inline u32 get_cmdq_pg(u32 val) +{ + return (val & ~MAX_CMDQ_IDX_PER_PG) / BNXT_QPLIB_CMDQE_CNT_PER_PG; +} + +static inline u32 get_cmdq_idx(u32 val) +{ + return val & MAX_CMDQ_IDX_PER_PG; +} + +/* Crsq buf is 1024-Byte */ +struct bnxt_qplib_crsbe { + u8 data[1024]; +}; + +/* CRSQ SB */ +#define BNXT_QPLIB_CRSBE_MAX_CNT 4 +#define BNXT_QPLIB_CRSBE_UNITS sizeof(struct bnxt_qplib_crsbe) +#define BNXT_QPLIB_CRSBE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_CRSBE_UNITS) + +#define MAX_CRSB_IDX (BNXT_QPLIB_CRSBE_MAX_CNT - 1) +#define MAX_CRSB_IDX_PER_PG (BNXT_QPLIB_CRSBE_CNT_PER_PG - 1) + +static inline u32 get_crsb_pg(u32 val) +{ + return (val & ~MAX_CRSB_IDX_PER_PG) / BNXT_QPLIB_CRSBE_CNT_PER_PG; +} + +static inline u32 get_crsb_idx(u32 val) +{ + return val & MAX_CRSB_IDX_PER_PG; +} + +static inline void bnxt_qplib_crsb_dma_next(dma_addr_t *pg_map_arr, + u32 prod, dma_addr_t *dma_addr) +{ + *dma_addr = pg_map_arr[(prod) / BNXT_QPLIB_CRSBE_CNT_PER_PG]; + *dma_addr += ((prod) % BNXT_QPLIB_CRSBE_CNT_PER_PG) * + BNXT_QPLIB_CRSBE_UNITS; +} + +/* CREQ */ +/* Allocate 1 per QP for async error notification for now */ +#define BNXT_QPLIB_CREQE_MAX_CNT (64 * 1024) +#define BNXT_QPLIB_CREQE_UNITS 16 /* 16-Bytes per prod unit */ +#define BNXT_QPLIB_CREQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_CREQE_UNITS) + +#define MAX_CREQ_IDX (BNXT_QPLIB_CREQE_MAX_CNT - 1) +#define MAX_CREQ_IDX_PER_PG (BNXT_QPLIB_CREQE_CNT_PER_PG - 1) + +static inline u32 get_creq_pg(u32 val) +{ + return (val & ~MAX_CREQ_IDX_PER_PG) / BNXT_QPLIB_CREQE_CNT_PER_PG; +} + +static inline u32 get_creq_idx(u32 val) +{ + return val & MAX_CREQ_IDX_PER_PG; +} + +#define BNXT_QPLIB_CREQE_PER_PG (PAGE_SIZE / sizeof(struct creq_base)) + +#define CREQ_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!((hdr)->v & CREQ_BASE_V) == \ + !((raw_cons) & (cp_bit))) + +#define CREQ_DB_KEY_CP (0x2 << CMPL_DOORBELL_KEY_SFT) +#define CREQ_DB_IDX_VALID CMPL_DOORBELL_IDX_VALID +#define CREQ_DB_IRQ_DIS CMPL_DOORBELL_MASK +#define CREQ_DB_CP_FLAGS_REARM (CREQ_DB_KEY_CP | \ + CREQ_DB_IDX_VALID) +#define CREQ_DB_CP_FLAGS (CREQ_DB_KEY_CP | \ + CREQ_DB_IDX_VALID | \ + CREQ_DB_IRQ_DIS) +#define CREQ_DB_REARM(db, raw_cons, cp_bit) \ + writel(CREQ_DB_CP_FLAGS_REARM | ((raw_cons) & ((cp_bit) - 1)), db) +#define CREQ_DB(db, raw_cons, cp_bit) \ + writel(CREQ_DB_CP_FLAGS | ((raw_cons) & ((cp_bit) - 1)), db) + +/* HWQ */ +struct bnxt_qplib_crsqe { + struct creq_qp_event qp_event; + u32 req_size; +}; + +struct bnxt_qplib_crsq { + struct bnxt_qplib_crsqe *crsq; + u32 prod; + u32 cons; + u32 max_elements; +}; + +/* RCFW Communication Channels */ +struct bnxt_qplib_rcfw { + struct pci_dev *pdev; + int vector; + struct tasklet_struct worker; + bool requested; + unsigned long *cmdq_bitmap; + u32 bmap_size; + unsigned long flags; +#define FIRMWARE_INITIALIZED_FLAG 1 +#define FIRMWARE_FIRST_FLAG BIT(31) + wait_queue_head_t waitq; + int (*aeq_handler)(struct bnxt_qplib_rcfw *, + struct creq_func_event *); + atomic_t seq_num; + + /* Bar region info */ + void __iomem *cmdq_bar_reg_iomem; + u16 cmdq_bar_reg; + u16 cmdq_bar_reg_prod_off; + u16 cmdq_bar_reg_trig_off; + u16 creq_ring_id; + u16 creq_bar_reg; + void __iomem *creq_bar_reg_iomem; + + /* Cmd-Resp and Async Event notification queue */ + struct bnxt_qplib_hwq creq; + u64 creq_qp_event_processed; + u64 creq_func_event_processed; + + /* Actual Cmd and Resp Queues */ + struct bnxt_qplib_hwq cmdq; + struct bnxt_qplib_crsq crsq; + struct bnxt_qplib_hwq crsb; +}; + +void bnxt_qplib_free_rcfw_channel(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_alloc_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw); +void bnxt_qplib_disable_rcfw_channel(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_enable_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw, + int msix_vector, + int cp_bar_reg_off, int virt_fn, + int (*aeq_handler) + (struct bnxt_qplib_rcfw *, + struct creq_func_event *)); + +int bnxt_qplib_rcfw_block_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie); +int bnxt_qplib_rcfw_wait_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie); +void *bnxt_qplib_rcfw_send_message(struct bnxt_qplib_rcfw *rcfw, + struct cmdq_base *req, void **crsbe, + u8 is_block); + +int bnxt_qplib_deinit_rcfw(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_init_rcfw(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_ctx *ctx, int is_virtfn); +#endif /* __BNXT_QPLIB_RCFW_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.c b/drivers/infiniband/hw/bnxt_re/qplib_res.c new file mode 100644 index 000000000000..62447b3badec --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.c @@ -0,0 +1,825 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: QPLib resource manager + */ + +#include +#include +#include +#include +#include +#include +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_rcfw.h" + +static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats); +static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats); + +/* PBL */ +static void __free_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl, + bool is_umem) +{ + int i; + + if (!is_umem) { + for (i = 0; i < pbl->pg_count; i++) { + if (pbl->pg_arr[i]) + dma_free_coherent(&pdev->dev, pbl->pg_size, + (void *)((unsigned long) + pbl->pg_arr[i] & + PAGE_MASK), + pbl->pg_map_arr[i]); + else + dev_warn(&pdev->dev, + "QPLIB: PBL free pg_arr[%d] empty?!", + i); + pbl->pg_arr[i] = NULL; + } + } + kfree(pbl->pg_arr); + pbl->pg_arr = NULL; + kfree(pbl->pg_map_arr); + pbl->pg_map_arr = NULL; + pbl->pg_count = 0; + pbl->pg_size = 0; +} + +static int __alloc_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl, + struct scatterlist *sghead, u32 pages, u32 pg_size) +{ + struct scatterlist *sg; + bool is_umem = false; + int i; + + /* page ptr arrays */ + pbl->pg_arr = kcalloc(pages, sizeof(void *), GFP_KERNEL); + if (!pbl->pg_arr) + return -ENOMEM; + + pbl->pg_map_arr = kcalloc(pages, sizeof(dma_addr_t), GFP_KERNEL); + if (!pbl->pg_map_arr) { + kfree(pbl->pg_arr); + pbl->pg_arr = NULL; + return -ENOMEM; + } + pbl->pg_count = 0; + pbl->pg_size = pg_size; + + if (!sghead) { + for (i = 0; i < pages; i++) { + pbl->pg_arr[i] = dma_alloc_coherent(&pdev->dev, + pbl->pg_size, + &pbl->pg_map_arr[i], + GFP_KERNEL); + if (!pbl->pg_arr[i]) + goto fail; + memset(pbl->pg_arr[i], 0, pbl->pg_size); + pbl->pg_count++; + } + } else { + i = 0; + is_umem = true; + for_each_sg(sghead, sg, pages, i) { + pbl->pg_map_arr[i] = sg_dma_address(sg); + pbl->pg_arr[i] = sg_virt(sg); + if (!pbl->pg_arr[i]) + goto fail; + + pbl->pg_count++; + } + } + + return 0; + +fail: + __free_pbl(pdev, pbl, is_umem); + return -ENOMEM; +} + +/* HWQ */ +void bnxt_qplib_free_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq) +{ + int i; + + if (!hwq->max_elements) + return; + if (hwq->level >= PBL_LVL_MAX) + return; + + for (i = 0; i < hwq->level + 1; i++) { + if (i == hwq->level) + __free_pbl(pdev, &hwq->pbl[i], hwq->is_user); + else + __free_pbl(pdev, &hwq->pbl[i], false); + } + + hwq->level = PBL_LVL_MAX; + hwq->max_elements = 0; + hwq->element_size = 0; + hwq->prod = 0; + hwq->cons = 0; + hwq->cp_bit = 0; +} + +/* All HWQs are power of 2 in size */ +int bnxt_qplib_alloc_init_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq, + struct scatterlist *sghead, int nmap, + u32 *elements, u32 element_size, u32 aux, + u32 pg_size, enum bnxt_qplib_hwq_type hwq_type) +{ + u32 pages, slots, size, aux_pages = 0, aux_size = 0; + dma_addr_t *src_phys_ptr, **dst_virt_ptr; + int i, rc; + + hwq->level = PBL_LVL_MAX; + + slots = roundup_pow_of_two(*elements); + if (aux) { + aux_size = roundup_pow_of_two(aux); + aux_pages = (slots * aux_size) / pg_size; + if ((slots * aux_size) % pg_size) + aux_pages++; + } + size = roundup_pow_of_two(element_size); + + if (!sghead) { + hwq->is_user = false; + pages = (slots * size) / pg_size + aux_pages; + if ((slots * size) % pg_size) + pages++; + if (!pages) + return -EINVAL; + } else { + hwq->is_user = true; + pages = nmap; + } + + /* Alloc the 1st memory block; can be a PDL/PTL/PBL */ + if (sghead && (pages == MAX_PBL_LVL_0_PGS)) + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_0], sghead, + pages, pg_size); + else + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_0], NULL, 1, pg_size); + if (rc) + goto fail; + + hwq->level = PBL_LVL_0; + + if (pages > MAX_PBL_LVL_0_PGS) { + if (pages > MAX_PBL_LVL_1_PGS) { + /* 2 levels of indirection */ + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_1], NULL, + MAX_PBL_LVL_1_PGS_FOR_LVL_2, pg_size); + if (rc) + goto fail; + /* Fill in lvl0 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_0].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_1].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_1].pg_count; i++) + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | PTU_PDE_VALID; + hwq->level = PBL_LVL_1; + + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_2], sghead, + pages, pg_size); + if (rc) + goto fail; + + /* Fill in lvl1 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_1].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_2].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_2].pg_count; i++) { + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | PTU_PTE_VALID; + } + if (hwq_type == HWQ_TYPE_QUEUE) { + /* Find the last pg of the size */ + i = hwq->pbl[PBL_LVL_2].pg_count; + dst_virt_ptr[PTR_PG(i - 1)][PTR_IDX(i - 1)] |= + PTU_PTE_LAST; + if (i > 1) + dst_virt_ptr[PTR_PG(i - 2)] + [PTR_IDX(i - 2)] |= + PTU_PTE_NEXT_TO_LAST; + } + hwq->level = PBL_LVL_2; + } else { + u32 flag = hwq_type == HWQ_TYPE_L2_CMPL ? 0 : + PTU_PTE_VALID; + + /* 1 level of indirection */ + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_1], sghead, + pages, pg_size); + if (rc) + goto fail; + /* Fill in lvl0 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_0].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_1].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_1].pg_count; i++) { + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | flag; + } + if (hwq_type == HWQ_TYPE_QUEUE) { + /* Find the last pg of the size */ + i = hwq->pbl[PBL_LVL_1].pg_count; + dst_virt_ptr[PTR_PG(i - 1)][PTR_IDX(i - 1)] |= + PTU_PTE_LAST; + if (i > 1) + dst_virt_ptr[PTR_PG(i - 2)] + [PTR_IDX(i - 2)] |= + PTU_PTE_NEXT_TO_LAST; + } + hwq->level = PBL_LVL_1; + } + } + hwq->pdev = pdev; + spin_lock_init(&hwq->lock); + hwq->prod = 0; + hwq->cons = 0; + *elements = hwq->max_elements = slots; + hwq->element_size = size; + + /* For direct access to the elements */ + hwq->pbl_ptr = hwq->pbl[hwq->level].pg_arr; + hwq->pbl_dma_ptr = hwq->pbl[hwq->level].pg_map_arr; + + return 0; + +fail: + bnxt_qplib_free_hwq(pdev, hwq); + return -ENOMEM; +} + +/* Context Tables */ +void bnxt_qplib_free_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx) +{ + int i; + + bnxt_qplib_free_hwq(pdev, &ctx->qpc_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->mrw_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->srqc_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->cq_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->tim_tbl); + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) + bnxt_qplib_free_hwq(pdev, &ctx->tqm_tbl[i]); + bnxt_qplib_free_hwq(pdev, &ctx->tqm_pde); + bnxt_qplib_free_stats_ctx(pdev, &ctx->stats); +} + +/* + * Routine: bnxt_qplib_alloc_ctx + * Description: + * Context tables are memories which are used by the chip fw. + * The 6 tables defined are: + * QPC ctx - holds QP states + * MRW ctx - holds memory region and window + * SRQ ctx - holds shared RQ states + * CQ ctx - holds completion queue states + * TQM ctx - holds Tx Queue Manager context + * TIM ctx - holds timer context + * Depending on the size of the tbl requested, either a 1 Page Buffer List + * or a 1-to-2-stage indirection Page Directory List + 1 PBL is used + * instead. + * Table might be employed as follows: + * For 0 < ctx size <= 1 PAGE, 0 level of ind is used + * For 1 PAGE < ctx size <= 512 entries size, 1 level of ind is used + * For 512 < ctx size <= MAX, 2 levels of ind is used + * Returns: + * 0 if success, else -ERRORS + */ +int bnxt_qplib_alloc_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx, + bool virt_fn) +{ + int i, j, k, rc = 0; + int fnz_idx = -1; + __le64 **pbl_ptr; + + if (virt_fn) + goto stats_alloc; + + /* QPC Tables */ + ctx->qpc_tbl.max_elements = ctx->qpc_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->qpc_tbl, NULL, 0, + &ctx->qpc_tbl.max_elements, + BNXT_QPLIB_MAX_QP_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* MRW Tables */ + ctx->mrw_tbl.max_elements = ctx->mrw_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->mrw_tbl, NULL, 0, + &ctx->mrw_tbl.max_elements, + BNXT_QPLIB_MAX_MRW_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* SRQ Tables */ + ctx->srqc_tbl.max_elements = ctx->srqc_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->srqc_tbl, NULL, 0, + &ctx->srqc_tbl.max_elements, + BNXT_QPLIB_MAX_SRQ_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* CQ Tables */ + ctx->cq_tbl.max_elements = ctx->cq_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->cq_tbl, NULL, 0, + &ctx->cq_tbl.max_elements, + BNXT_QPLIB_MAX_CQ_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* TQM Buffer */ + ctx->tqm_pde.max_elements = 512; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tqm_pde, NULL, 0, + &ctx->tqm_pde.max_elements, sizeof(u64), + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) { + if (!ctx->tqm_count[i]) + continue; + ctx->tqm_tbl[i].max_elements = ctx->qpc_count * + ctx->tqm_count[i]; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tqm_tbl[i], NULL, 0, + &ctx->tqm_tbl[i].max_elements, 1, + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + } + pbl_ptr = (__le64 **)ctx->tqm_pde.pbl_ptr; + for (i = 0, j = 0; i < MAX_TQM_ALLOC_REQ; + i++, j += MAX_TQM_ALLOC_BLK_SIZE) { + if (!ctx->tqm_tbl[i].max_elements) + continue; + if (fnz_idx == -1) + fnz_idx = i; + switch (ctx->tqm_tbl[i].level) { + case PBL_LVL_2: + for (k = 0; k < ctx->tqm_tbl[i].pbl[PBL_LVL_1].pg_count; + k++) + pbl_ptr[PTR_PG(j + k)][PTR_IDX(j + k)] = + cpu_to_le64( + ctx->tqm_tbl[i].pbl[PBL_LVL_1].pg_map_arr[k] + | PTU_PTE_VALID); + break; + case PBL_LVL_1: + case PBL_LVL_0: + default: + pbl_ptr[PTR_PG(j)][PTR_IDX(j)] = cpu_to_le64( + ctx->tqm_tbl[i].pbl[PBL_LVL_0].pg_map_arr[0] | + PTU_PTE_VALID); + break; + } + } + if (fnz_idx == -1) + fnz_idx = 0; + ctx->tqm_pde_level = ctx->tqm_tbl[fnz_idx].level == PBL_LVL_2 ? + PBL_LVL_2 : ctx->tqm_tbl[fnz_idx].level + 1; + + /* TIM Buffer */ + ctx->tim_tbl.max_elements = ctx->qpc_count * 16; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tim_tbl, NULL, 0, + &ctx->tim_tbl.max_elements, 1, + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + +stats_alloc: + /* Stats */ + rc = bnxt_qplib_alloc_stats_ctx(pdev, &ctx->stats); + if (rc) + goto fail; + + return 0; + +fail: + bnxt_qplib_free_ctx(pdev, ctx); + return rc; +} + +/* GUID */ +void bnxt_qplib_get_guid(u8 *dev_addr, u8 *guid) +{ + u8 mac[ETH_ALEN]; + + /* MAC-48 to EUI-64 mapping */ + memcpy(mac, dev_addr, ETH_ALEN); + guid[0] = mac[0] ^ 2; + guid[1] = mac[1]; + guid[2] = mac[2]; + guid[3] = 0xff; + guid[4] = 0xfe; + guid[5] = mac[3]; + guid[6] = mac[4]; + guid[7] = mac[5]; +} + +static void bnxt_qplib_free_sgid_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl) +{ + kfree(sgid_tbl->tbl); + kfree(sgid_tbl->hw_id); + kfree(sgid_tbl->ctx); + sgid_tbl->tbl = NULL; + sgid_tbl->hw_id = NULL; + sgid_tbl->ctx = NULL; + sgid_tbl->max = 0; + sgid_tbl->active = 0; +} + +static int bnxt_qplib_alloc_sgid_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl, + u16 max) +{ + sgid_tbl->tbl = kcalloc(max, sizeof(struct bnxt_qplib_gid), GFP_KERNEL); + if (!sgid_tbl->tbl) + return -ENOMEM; + + sgid_tbl->hw_id = kcalloc(max, sizeof(u16), GFP_KERNEL); + if (!sgid_tbl->hw_id) + goto out_free1; + + sgid_tbl->ctx = kcalloc(max, sizeof(void *), GFP_KERNEL); + if (!sgid_tbl->ctx) + goto out_free2; + + sgid_tbl->max = max; + return 0; +out_free2: + kfree(sgid_tbl->hw_id); + sgid_tbl->hw_id = NULL; +out_free1: + kfree(sgid_tbl->tbl); + sgid_tbl->tbl = NULL; + return -ENOMEM; +}; + +static void bnxt_qplib_cleanup_sgid_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl) +{ + int i; + + for (i = 0; i < sgid_tbl->max; i++) { + if (memcmp(&sgid_tbl->tbl[i], &bnxt_qplib_gid_zero, + sizeof(bnxt_qplib_gid_zero))) + bnxt_qplib_del_sgid(sgid_tbl, &sgid_tbl->tbl[i], true); + } + memset(sgid_tbl->tbl, 0, sizeof(struct bnxt_qplib_gid) * sgid_tbl->max); + memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); + sgid_tbl->active = 0; +} + +static void bnxt_qplib_init_sgid_tbl(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct net_device *netdev) +{ + memset(sgid_tbl->tbl, 0, sizeof(struct bnxt_qplib_gid) * sgid_tbl->max); + memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); +} + +static void bnxt_qplib_free_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + if (!pkey_tbl->tbl) + dev_dbg(&res->pdev->dev, "QPLIB: PKEY tbl not present"); + else + kfree(pkey_tbl->tbl); + + pkey_tbl->tbl = NULL; + pkey_tbl->max = 0; + pkey_tbl->active = 0; +} + +static int bnxt_qplib_alloc_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, + u16 max) +{ + pkey_tbl->tbl = kcalloc(max, sizeof(u16), GFP_KERNEL); + if (!pkey_tbl->tbl) + return -ENOMEM; + + pkey_tbl->max = max; + return 0; +}; + +/* PDs */ +int bnxt_qplib_alloc_pd(struct bnxt_qplib_pd_tbl *pdt, struct bnxt_qplib_pd *pd) +{ + u32 bit_num; + + bit_num = find_first_bit(pdt->tbl, pdt->max); + if (bit_num == pdt->max) + return -ENOMEM; + + /* Found unused PD */ + clear_bit(bit_num, pdt->tbl); + pd->id = bit_num; + return 0; +} + +int bnxt_qplib_dealloc_pd(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pdt, + struct bnxt_qplib_pd *pd) +{ + if (test_and_set_bit(pd->id, pdt->tbl)) { + dev_warn(&res->pdev->dev, "Freeing an unused PD? pdn = %d", + pd->id); + return -EINVAL; + } + pd->id = 0; + return 0; +} + +static void bnxt_qplib_free_pd_tbl(struct bnxt_qplib_pd_tbl *pdt) +{ + kfree(pdt->tbl); + pdt->tbl = NULL; + pdt->max = 0; +} + +static int bnxt_qplib_alloc_pd_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pdt, + u32 max) +{ + u32 bytes; + + bytes = max >> 3; + if (!bytes) + bytes = 1; + pdt->tbl = kmalloc(bytes, GFP_KERNEL); + if (!pdt->tbl) + return -ENOMEM; + + pdt->max = max; + memset((u8 *)pdt->tbl, 0xFF, bytes); + + return 0; +} + +/* DPIs */ +int bnxt_qplib_alloc_dpi(struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi, + void *app) +{ + u32 bit_num; + + bit_num = find_first_bit(dpit->tbl, dpit->max); + if (bit_num == dpit->max) + return -ENOMEM; + + /* Found unused DPI */ + clear_bit(bit_num, dpit->tbl); + dpit->app_tbl[bit_num] = app; + + dpi->dpi = bit_num; + dpi->dbr = dpit->dbr_bar_reg_iomem + (bit_num * PAGE_SIZE); + dpi->umdbr = dpit->unmapped_dbr + (bit_num * PAGE_SIZE); + + return 0; +} + +int bnxt_qplib_dealloc_dpi(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi) +{ + if (dpi->dpi >= dpit->max) { + dev_warn(&res->pdev->dev, "Invalid DPI? dpi = %d", dpi->dpi); + return -EINVAL; + } + if (test_and_set_bit(dpi->dpi, dpit->tbl)) { + dev_warn(&res->pdev->dev, "Freeing an unused DPI? dpi = %d", + dpi->dpi); + return -EINVAL; + } + if (dpit->app_tbl) + dpit->app_tbl[dpi->dpi] = NULL; + memset(dpi, 0, sizeof(*dpi)); + + return 0; +} + +static void bnxt_qplib_free_dpi_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit) +{ + kfree(dpit->tbl); + kfree(dpit->app_tbl); + if (dpit->dbr_bar_reg_iomem) + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + memset(dpit, 0, sizeof(*dpit)); +} + +static int bnxt_qplib_alloc_dpi_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit, + u32 dbr_offset) +{ + u32 dbr_bar_reg = RCFW_DBR_PCI_BAR_REGION; + resource_size_t bar_reg_base; + u32 dbr_len, bytes; + + if (dpit->dbr_bar_reg_iomem) { + dev_err(&res->pdev->dev, + "QPLIB: DBR BAR region %d already mapped", dbr_bar_reg); + return -EALREADY; + } + + bar_reg_base = pci_resource_start(res->pdev, dbr_bar_reg); + if (!bar_reg_base) { + dev_err(&res->pdev->dev, + "QPLIB: BAR region %d resc start failed", dbr_bar_reg); + return -ENOMEM; + } + + dbr_len = pci_resource_len(res->pdev, dbr_bar_reg) - dbr_offset; + if (!dbr_len || ((dbr_len & (PAGE_SIZE - 1)) != 0)) { + dev_err(&res->pdev->dev, "QPLIB: Invalid DBR length %d", + dbr_len); + return -ENOMEM; + } + + dpit->dbr_bar_reg_iomem = ioremap_nocache(bar_reg_base + dbr_offset, + dbr_len); + if (!dpit->dbr_bar_reg_iomem) { + dev_err(&res->pdev->dev, + "QPLIB: FP: DBR BAR region %d mapping failed", + dbr_bar_reg); + return -ENOMEM; + } + + dpit->unmapped_dbr = bar_reg_base + dbr_offset; + dpit->max = dbr_len / PAGE_SIZE; + + dpit->app_tbl = kcalloc(dpit->max, sizeof(void *), GFP_KERNEL); + if (!dpit->app_tbl) { + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + dev_err(&res->pdev->dev, + "QPLIB: DPI app tbl allocation failed"); + return -ENOMEM; + } + + bytes = dpit->max >> 3; + if (!bytes) + bytes = 1; + + dpit->tbl = kmalloc(bytes, GFP_KERNEL); + if (!dpit->tbl) { + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + kfree(dpit->app_tbl); + dpit->app_tbl = NULL; + dev_err(&res->pdev->dev, + "QPLIB: DPI tbl allocation failed for size = %d", + bytes); + return -ENOMEM; + } + + memset((u8 *)dpit->tbl, 0xFF, bytes); + + return 0; +} + +/* PKEYs */ +static void bnxt_qplib_cleanup_pkey_tbl(struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + memset(pkey_tbl->tbl, 0, sizeof(u16) * pkey_tbl->max); + pkey_tbl->active = 0; +} + +static void bnxt_qplib_init_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + u16 pkey = 0xFFFF; + + memset(pkey_tbl->tbl, 0, sizeof(u16) * pkey_tbl->max); + + /* pkey default = 0xFFFF */ + bnxt_qplib_add_pkey(res, pkey_tbl, &pkey, false); +} + +/* Stats */ +static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats) +{ + if (stats->dma) { + dma_free_coherent(&pdev->dev, stats->size, + stats->dma, stats->dma_map); + } + memset(stats, 0, sizeof(*stats)); + stats->fw_id = -1; +} + +static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats) +{ + memset(stats, 0, sizeof(*stats)); + stats->fw_id = -1; + stats->size = sizeof(struct ctx_hw_stats); + stats->dma = dma_alloc_coherent(&pdev->dev, stats->size, + &stats->dma_map, GFP_KERNEL); + if (!stats->dma) { + dev_err(&pdev->dev, "QPLIB: Stats DMA allocation failed"); + return -ENOMEM; + } + return 0; +} + +void bnxt_qplib_cleanup_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_cleanup_pkey_tbl(&res->pkey_tbl); + bnxt_qplib_cleanup_sgid_tbl(res, &res->sgid_tbl); +} + +int bnxt_qplib_init_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_init_sgid_tbl(&res->sgid_tbl, res->netdev); + bnxt_qplib_init_pkey_tbl(res, &res->pkey_tbl); + + return 0; +} + +void bnxt_qplib_free_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_free_pkey_tbl(res, &res->pkey_tbl); + bnxt_qplib_free_sgid_tbl(res, &res->sgid_tbl); + bnxt_qplib_free_pd_tbl(&res->pd_tbl); + bnxt_qplib_free_dpi_tbl(res, &res->dpi_tbl); + + res->netdev = NULL; + res->pdev = NULL; +} + +int bnxt_qplib_alloc_res(struct bnxt_qplib_res *res, struct pci_dev *pdev, + struct net_device *netdev, + struct bnxt_qplib_dev_attr *dev_attr) +{ + int rc = 0; + + res->pdev = pdev; + res->netdev = netdev; + + rc = bnxt_qplib_alloc_sgid_tbl(res, &res->sgid_tbl, dev_attr->max_sgid); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_pkey_tbl(res, &res->pkey_tbl, dev_attr->max_pkey); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_pd_tbl(res, &res->pd_tbl, dev_attr->max_pd); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_dpi_tbl(res, &res->dpi_tbl, dev_attr->l2_db_size); + if (rc) + goto fail; + + return 0; +fail: + bnxt_qplib_free_res(res); + return rc; +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.h b/drivers/infiniband/hw/bnxt_re/qplib_res.h new file mode 100644 index 000000000000..6277d802ca4b --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.h @@ -0,0 +1,223 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: QPLib resource manager (header) + */ + +#ifndef __BNXT_QPLIB_RES_H__ +#define __BNXT_QPLIB_RES_H__ + +extern const struct bnxt_qplib_gid bnxt_qplib_gid_zero; + +#define PTR_CNT_PER_PG (PAGE_SIZE / sizeof(void *)) +#define PTR_MAX_IDX_PER_PG (PTR_CNT_PER_PG - 1) +#define PTR_PG(x) (((x) & ~PTR_MAX_IDX_PER_PG) / PTR_CNT_PER_PG) +#define PTR_IDX(x) ((x) & PTR_MAX_IDX_PER_PG) + +#define HWQ_CMP(idx, hwq) ((idx) & ((hwq)->max_elements - 1)) + +enum bnxt_qplib_hwq_type { + HWQ_TYPE_CTX, + HWQ_TYPE_QUEUE, + HWQ_TYPE_L2_CMPL +}; + +#define MAX_PBL_LVL_0_PGS 1 +#define MAX_PBL_LVL_1_PGS 512 +#define MAX_PBL_LVL_1_PGS_SHIFT 9 +#define MAX_PBL_LVL_1_PGS_FOR_LVL_2 256 +#define MAX_PBL_LVL_2_PGS (256 * 512) + +enum bnxt_qplib_pbl_lvl { + PBL_LVL_0, + PBL_LVL_1, + PBL_LVL_2, + PBL_LVL_MAX +}; + +#define ROCE_PG_SIZE_4K (4 * 1024) +#define ROCE_PG_SIZE_8K (8 * 1024) +#define ROCE_PG_SIZE_64K (64 * 1024) +#define ROCE_PG_SIZE_2M (2 * 1024 * 1024) +#define ROCE_PG_SIZE_8M (8 * 1024 * 1024) +#define ROCE_PG_SIZE_1G (1024 * 1024 * 1024) + +struct bnxt_qplib_pbl { + u32 pg_count; + u32 pg_size; + void **pg_arr; + dma_addr_t *pg_map_arr; +}; + +struct bnxt_qplib_hwq { + struct pci_dev *pdev; + /* lock to protect qplib_hwq */ + spinlock_t lock; + struct bnxt_qplib_pbl pbl[PBL_LVL_MAX]; + enum bnxt_qplib_pbl_lvl level; /* 0, 1, or 2 */ + /* ptr for easy access to the PBL entries */ + void **pbl_ptr; + /* ptr for easy access to the dma_addr */ + dma_addr_t *pbl_dma_ptr; + u32 max_elements; + u16 element_size; /* Size of each entry */ + + u32 prod; /* raw */ + u32 cons; /* raw */ + u8 cp_bit; + u8 is_user; +}; + +/* Tables */ +struct bnxt_qplib_pd_tbl { + unsigned long *tbl; + u32 max; +}; + +struct bnxt_qplib_sgid_tbl { + struct bnxt_qplib_gid *tbl; + u16 *hw_id; + u16 max; + u16 active; + void *ctx; +}; + +struct bnxt_qplib_pkey_tbl { + u16 *tbl; + u16 max; + u16 active; +}; + +struct bnxt_qplib_dpi { + u32 dpi; + void __iomem *dbr; + u64 umdbr; +}; + +struct bnxt_qplib_dpi_tbl { + void **app_tbl; + unsigned long *tbl; + u16 max; + void __iomem *dbr_bar_reg_iomem; + u64 unmapped_dbr; +}; + +struct bnxt_qplib_stats { + dma_addr_t dma_map; + void *dma; + u32 size; + u32 fw_id; +}; + +struct bnxt_qplib_vf_res { + u32 max_qp_per_vf; + u32 max_mrw_per_vf; + u32 max_srq_per_vf; + u32 max_cq_per_vf; + u32 max_gid_per_vf; +}; + +#define BNXT_QPLIB_MAX_QP_CTX_ENTRY_SIZE 448 +#define BNXT_QPLIB_MAX_SRQ_CTX_ENTRY_SIZE 64 +#define BNXT_QPLIB_MAX_CQ_CTX_ENTRY_SIZE 64 +#define BNXT_QPLIB_MAX_MRW_CTX_ENTRY_SIZE 128 + +struct bnxt_qplib_ctx { + u32 qpc_count; + struct bnxt_qplib_hwq qpc_tbl; + u32 mrw_count; + struct bnxt_qplib_hwq mrw_tbl; + u32 srqc_count; + struct bnxt_qplib_hwq srqc_tbl; + u32 cq_count; + struct bnxt_qplib_hwq cq_tbl; + struct bnxt_qplib_hwq tim_tbl; +#define MAX_TQM_ALLOC_REQ 32 +#define MAX_TQM_ALLOC_BLK_SIZE 8 + u8 tqm_count[MAX_TQM_ALLOC_REQ]; + struct bnxt_qplib_hwq tqm_pde; + u32 tqm_pde_level; + struct bnxt_qplib_hwq tqm_tbl[MAX_TQM_ALLOC_REQ]; + struct bnxt_qplib_stats stats; + struct bnxt_qplib_vf_res vf_res; +}; + +struct bnxt_qplib_res { + struct pci_dev *pdev; + struct net_device *netdev; + + struct bnxt_qplib_rcfw *rcfw; + + struct bnxt_qplib_pd_tbl pd_tbl; + struct bnxt_qplib_sgid_tbl sgid_tbl; + struct bnxt_qplib_pkey_tbl pkey_tbl; + struct bnxt_qplib_dpi_tbl dpi_tbl; +}; + +#define to_bnxt_qplib(ptr, type, member) \ + container_of(ptr, type, member) + +struct bnxt_qplib_pd; +struct bnxt_qplib_dev_attr; + +void bnxt_qplib_free_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq); +int bnxt_qplib_alloc_init_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq, + struct scatterlist *sl, int nmap, u32 *elements, + u32 elements_per_page, u32 aux, u32 pg_size, + enum bnxt_qplib_hwq_type hwq_type); +void bnxt_qplib_get_guid(u8 *dev_addr, u8 *guid); +int bnxt_qplib_alloc_pd(struct bnxt_qplib_pd_tbl *pd_tbl, + struct bnxt_qplib_pd *pd); +int bnxt_qplib_dealloc_pd(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pd_tbl, + struct bnxt_qplib_pd *pd); +int bnxt_qplib_alloc_dpi(struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi, + void *app); +int bnxt_qplib_dealloc_dpi(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpi_tbl, + struct bnxt_qplib_dpi *dpi); +void bnxt_qplib_cleanup_res(struct bnxt_qplib_res *res); +int bnxt_qplib_init_res(struct bnxt_qplib_res *res); +void bnxt_qplib_free_res(struct bnxt_qplib_res *res); +int bnxt_qplib_alloc_res(struct bnxt_qplib_res *res, struct pci_dev *pdev, + struct net_device *netdev, + struct bnxt_qplib_dev_attr *dev_attr); +void bnxt_qplib_free_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx); +int bnxt_qplib_alloc_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx, + bool virt_fn); +#endif /* __BNXT_QPLIB_RES_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.c b/drivers/infiniband/hw/bnxt_re/qplib_sp.c new file mode 100644 index 000000000000..7b31eccedf11 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.c @@ -0,0 +1,838 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators + */ + +#include +#include +#include +#include + +#include "roce_hsi.h" + +#include "qplib_res.h" +#include "qplib_rcfw.h" +#include "qplib_sp.h" + +const struct bnxt_qplib_gid bnxt_qplib_gid_zero = {{ 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0 } }; + +/* Device */ +int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_dev_attr *attr) +{ + struct cmdq_query_func req; + struct creq_query_func_resp *resp; + struct creq_query_func_resp_sb *sb; + u16 cmd_flags = 0; + u32 temp; + u8 *tqm_alloc; + int i; + + RCFW_CMD_PREP(req, QUERY_FUNC, cmd_flags); + + req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS; + resp = (struct creq_query_func_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void **)&sb, + 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: QUERY_FUNC send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: SP: QUERY_FUNC timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: QUERY_FUNC failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + /* Extract the context from the side buffer */ + attr->max_qp = le32_to_cpu(sb->max_qp); + attr->max_qp_rd_atom = + sb->max_qp_rd_atom > BNXT_QPLIB_MAX_OUT_RD_ATOM ? + BNXT_QPLIB_MAX_OUT_RD_ATOM : sb->max_qp_rd_atom; + attr->max_qp_init_rd_atom = + sb->max_qp_init_rd_atom > BNXT_QPLIB_MAX_OUT_RD_ATOM ? + BNXT_QPLIB_MAX_OUT_RD_ATOM : sb->max_qp_init_rd_atom; + attr->max_qp_wqes = le16_to_cpu(sb->max_qp_wr); + attr->max_qp_sges = sb->max_sge; + attr->max_cq = le32_to_cpu(sb->max_cq); + attr->max_cq_wqes = le32_to_cpu(sb->max_cqe); + attr->max_cq_sges = attr->max_qp_sges; + attr->max_mr = le32_to_cpu(sb->max_mr); + attr->max_mw = le32_to_cpu(sb->max_mw); + + attr->max_mr_size = le64_to_cpu(sb->max_mr_size); + attr->max_pd = 64 * 1024; + attr->max_raw_ethy_qp = le32_to_cpu(sb->max_raw_eth_qp); + attr->max_ah = le32_to_cpu(sb->max_ah); + + attr->max_fmr = le32_to_cpu(sb->max_fmr); + attr->max_map_per_fmr = sb->max_map_per_fmr; + + attr->max_srq = le16_to_cpu(sb->max_srq); + attr->max_srq_wqes = le32_to_cpu(sb->max_srq_wr) - 1; + attr->max_srq_sges = sb->max_srq_sge; + /* Bono only reports 1 PKEY for now, but it can support > 1 */ + attr->max_pkey = le32_to_cpu(sb->max_pkeys); + + attr->max_inline_data = le32_to_cpu(sb->max_inline_data); + attr->l2_db_size = (sb->l2_db_space_size + 1) * PAGE_SIZE; + attr->max_sgid = le32_to_cpu(sb->max_gid); + + strlcpy(attr->fw_ver, "20.6.28.0", sizeof(attr->fw_ver)); + + for (i = 0; i < MAX_TQM_ALLOC_REQ / 4; i++) { + temp = le32_to_cpu(sb->tqm_alloc_reqs[i]); + tqm_alloc = (u8 *)&temp; + attr->tqm_alloc_reqs[i * 4] = *tqm_alloc; + attr->tqm_alloc_reqs[i * 4 + 1] = *(++tqm_alloc); + attr->tqm_alloc_reqs[i * 4 + 2] = *(++tqm_alloc); + attr->tqm_alloc_reqs[i * 4 + 3] = *(++tqm_alloc); + } + return 0; +} + +/* SGID */ +int bnxt_qplib_get_sgid(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl, int index, + struct bnxt_qplib_gid *gid) +{ + if (index > sgid_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: Index %d exceeded SGID table max (%d)", + index, sgid_tbl->max); + return -EINVAL; + } + memcpy(gid, &sgid_tbl->tbl[index], sizeof(*gid)); + return 0; +} + +int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct bnxt_qplib_gid *gid, bool update) +{ + struct bnxt_qplib_res *res = to_bnxt_qplib(sgid_tbl, + struct bnxt_qplib_res, + sgid_tbl); + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + int index; + + if (!sgid_tbl) { + dev_err(&res->pdev->dev, "QPLIB: SGID table not allocated"); + return -EINVAL; + } + /* Do we need a sgid_lock here? */ + if (!sgid_tbl->active) { + dev_err(&res->pdev->dev, + "QPLIB: SGID table has no active entries"); + return -ENOMEM; + } + for (index = 0; index < sgid_tbl->max; index++) { + if (!memcmp(&sgid_tbl->tbl[index], gid, sizeof(*gid))) + break; + } + if (index == sgid_tbl->max) { + dev_warn(&res->pdev->dev, "GID not found in the SGID table"); + return 0; + } + /* Remove GID from the SGID table */ + if (update) { + struct cmdq_delete_gid req; + struct creq_delete_gid_resp *resp; + u16 cmd_flags = 0; + + RCFW_CMD_PREP(req, DELETE_GID, cmd_flags); + if (sgid_tbl->hw_id[index] == 0xFFFF) { + dev_err(&res->pdev->dev, + "QPLIB: GID entry contains an invalid HW id"); + return -EINVAL; + } + req.gid_index = cpu_to_le16(sgid_tbl->hw_id[index]); + resp = (struct creq_delete_gid_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, NULL, + 0); + if (!resp) { + dev_err(&res->pdev->dev, + "QPLIB: SP: DELETE_GID send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, + le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&res->pdev->dev, + "QPLIB: SP: DELETE_GID timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&res->pdev->dev, + "QPLIB: SP: DELETE_GID failed "); + dev_err(&res->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + } + memcpy(&sgid_tbl->tbl[index], &bnxt_qplib_gid_zero, + sizeof(bnxt_qplib_gid_zero)); + sgid_tbl->active--; + dev_dbg(&res->pdev->dev, + "QPLIB: SGID deleted hw_id[0x%x] = 0x%x active = 0x%x", + index, sgid_tbl->hw_id[index], sgid_tbl->active); + sgid_tbl->hw_id[index] = (u16)-1; + + /* unlock */ + return 0; +} + +int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct bnxt_qplib_gid *gid, u8 *smac, u16 vlan_id, + bool update, u32 *index) +{ + struct bnxt_qplib_res *res = to_bnxt_qplib(sgid_tbl, + struct bnxt_qplib_res, + sgid_tbl); + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + int i, free_idx, rc = 0; + + if (!sgid_tbl) { + dev_err(&res->pdev->dev, "QPLIB: SGID table not allocated"); + return -EINVAL; + } + /* Do we need a sgid_lock here? */ + if (sgid_tbl->active == sgid_tbl->max) { + dev_err(&res->pdev->dev, "QPLIB: SGID table is full"); + return -ENOMEM; + } + free_idx = sgid_tbl->max; + for (i = 0; i < sgid_tbl->max; i++) { + if (!memcmp(&sgid_tbl->tbl[i], gid, sizeof(*gid))) { + dev_dbg(&res->pdev->dev, + "QPLIB: SGID entry already exist in entry %d!", + i); + *index = i; + return -EALREADY; + } else if (!memcmp(&sgid_tbl->tbl[i], &bnxt_qplib_gid_zero, + sizeof(bnxt_qplib_gid_zero)) && + free_idx == sgid_tbl->max) { + free_idx = i; + } + } + if (free_idx == sgid_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: SGID table is FULL but count is not MAX??"); + return -ENOMEM; + } + if (update) { + struct cmdq_add_gid req; + struct creq_add_gid_resp *resp; + u16 cmd_flags = 0; + u32 temp32[4]; + u16 temp16[3]; + + RCFW_CMD_PREP(req, ADD_GID, cmd_flags); + + memcpy(temp32, gid->data, sizeof(struct bnxt_qplib_gid)); + req.gid[0] = cpu_to_be32(temp32[3]); + req.gid[1] = cpu_to_be32(temp32[2]); + req.gid[2] = cpu_to_be32(temp32[1]); + req.gid[3] = cpu_to_be32(temp32[0]); + if (vlan_id != 0xFFFF) + req.vlan = cpu_to_le16((vlan_id & + CMDQ_ADD_GID_VLAN_VLAN_ID_MASK) | + CMDQ_ADD_GID_VLAN_TPID_TPID_8100 | + CMDQ_ADD_GID_VLAN_VLAN_EN); + + /* MAC in network format */ + memcpy(temp16, smac, 6); + req.src_mac[0] = cpu_to_be16(temp16[0]); + req.src_mac[1] = cpu_to_be16(temp16[1]); + req.src_mac[2] = cpu_to_be16(temp16[2]); + + resp = (struct creq_add_gid_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&res->pdev->dev, + "QPLIB: SP: ADD_GID send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, + le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&res->pdev->dev, + "QPIB: SP: ADD_GID timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&res->pdev->dev, "QPLIB: SP: ADD_GID failed "); + dev_err(&res->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + sgid_tbl->hw_id[free_idx] = le32_to_cpu(resp->xid); + } + /* Add GID to the sgid_tbl */ + memcpy(&sgid_tbl->tbl[free_idx], gid, sizeof(*gid)); + sgid_tbl->active++; + dev_dbg(&res->pdev->dev, + "QPLIB: SGID added hw_id[0x%x] = 0x%x active = 0x%x", + free_idx, sgid_tbl->hw_id[free_idx], sgid_tbl->active); + + *index = free_idx; + /* unlock */ + return rc; +} + +/* pkeys */ +int bnxt_qplib_get_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 index, + u16 *pkey) +{ + if (index == 0xFFFF) { + *pkey = 0xFFFF; + return 0; + } + if (index > pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: Index %d exceeded PKEY table max (%d)", + index, pkey_tbl->max); + return -EINVAL; + } + memcpy(pkey, &pkey_tbl->tbl[index], sizeof(*pkey)); + return 0; +} + +int bnxt_qplib_del_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update) +{ + int i, rc = 0; + + if (!pkey_tbl) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table not allocated"); + return -EINVAL; + } + + /* Do we need a pkey_lock here? */ + if (!pkey_tbl->active) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY table has no active entries"); + return -ENOMEM; + } + for (i = 0; i < pkey_tbl->max; i++) { + if (!memcmp(&pkey_tbl->tbl[i], pkey, sizeof(*pkey))) + break; + } + if (i == pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY 0x%04x not found in the pkey table", + *pkey); + return -ENOMEM; + } + memset(&pkey_tbl->tbl[i], 0, sizeof(*pkey)); + pkey_tbl->active--; + + /* unlock */ + return rc; +} + +int bnxt_qplib_add_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update) +{ + int i, free_idx, rc = 0; + + if (!pkey_tbl) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table not allocated"); + return -EINVAL; + } + + /* Do we need a pkey_lock here? */ + if (pkey_tbl->active == pkey_tbl->max) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table is full"); + return -ENOMEM; + } + free_idx = pkey_tbl->max; + for (i = 0; i < pkey_tbl->max; i++) { + if (!memcmp(&pkey_tbl->tbl[i], pkey, sizeof(*pkey))) + return -EALREADY; + else if (!pkey_tbl->tbl[i] && free_idx == pkey_tbl->max) + free_idx = i; + } + if (free_idx == pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY table is FULL but count is not MAX??"); + return -ENOMEM; + } + /* Add PKEY to the pkey_tbl */ + memcpy(&pkey_tbl->tbl[free_idx], pkey, sizeof(*pkey)); + pkey_tbl->active++; + + /* unlock */ + return rc; +} + +/* AH */ +int bnxt_qplib_create_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_ah req; + struct creq_create_ah_resp *resp; + u16 cmd_flags = 0; + u32 temp32[4]; + u16 temp16[3]; + + RCFW_CMD_PREP(req, CREATE_AH, cmd_flags); + + memcpy(temp32, ah->dgid.data, sizeof(struct bnxt_qplib_gid)); + req.dgid[0] = cpu_to_le32(temp32[0]); + req.dgid[1] = cpu_to_le32(temp32[1]); + req.dgid[2] = cpu_to_le32(temp32[2]); + req.dgid[3] = cpu_to_le32(temp32[3]); + + req.type = ah->nw_type; + req.hop_limit = ah->hop_limit; + req.sgid_index = cpu_to_le16(res->sgid_tbl.hw_id[ah->sgid_index]); + req.dest_vlan_id_flow_label = cpu_to_le32((ah->flow_label & + CMDQ_CREATE_AH_FLOW_LABEL_MASK) | + CMDQ_CREATE_AH_DEST_VLAN_ID_MASK); + req.pd_id = cpu_to_le32(ah->pd->id); + req.traffic_class = ah->traffic_class; + + /* MAC in network format */ + memcpy(temp16, ah->dmac, 6); + req.dest_mac[0] = cpu_to_le16(temp16[0]); + req.dest_mac[1] = cpu_to_le16(temp16[1]); + req.dest_mac[2] = cpu_to_le16(temp16[2]); + + resp = (struct creq_create_ah_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 1); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: CREATE_AH send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_block_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: SP: CREATE_AH timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: CREATE_AH failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + ah->id = le32_to_cpu(resp->xid); + return 0; +} + +int bnxt_qplib_destroy_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_ah req; + struct creq_destroy_ah_resp *resp; + u16 cmd_flags = 0; + + /* Clean up the AH table in the device */ + RCFW_CMD_PREP(req, DESTROY_AH, cmd_flags); + + req.ah_cid = cpu_to_le32(ah->id); + + resp = (struct creq_destroy_ah_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 1); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: DESTROY_AH send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_block_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: SP: DESTROY_AH timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: DESTROY_AH failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + return 0; +} + +/* MRW */ +int bnxt_qplib_free_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_deallocate_key req; + struct creq_deallocate_key_resp *resp; + u16 cmd_flags = 0; + + if (mrw->lkey == 0xFFFFFFFF) { + dev_info(&res->pdev->dev, + "QPLIB: SP: Free a reserved lkey MRW"); + return 0; + } + + RCFW_CMD_PREP(req, DEALLOCATE_KEY, cmd_flags); + + req.mrw_flags = mrw->type; + + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B)) + req.key = cpu_to_le32(mrw->rkey); + else + req.key = cpu_to_le32(mrw->lkey); + + resp = (struct creq_deallocate_key_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&res->pdev->dev, "QPLIB: SP: FREE_MR send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&res->pdev->dev, "QPLIB: SP: FREE_MR timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&res->pdev->dev, "QPLIB: SP: FREE_MR failed "); + dev_err(&res->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + /* Free the qplib's MRW memory */ + if (mrw->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mrw->hwq); + + return 0; +} + +int bnxt_qplib_alloc_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_allocate_mrw req; + struct creq_allocate_mrw_resp *resp; + u16 cmd_flags = 0; + unsigned long tmp; + + RCFW_CMD_PREP(req, ALLOCATE_MRW, cmd_flags); + + req.pd_id = cpu_to_le32(mrw->pd->id); + req.mrw_flags = mrw->type; + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR && + mrw->flags & BNXT_QPLIB_FR_PMR) || + mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A || + mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B) + req.access = CMDQ_ALLOCATE_MRW_ACCESS_CONSUMER_OWNED_KEY; + tmp = (unsigned long)mrw; + req.mrw_handle = cpu_to_le64(tmp); + + resp = (struct creq_allocate_mrw_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, 0); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: ALLOC_MRW send failed"); + return -EINVAL; + } + if (!bnxt_qplib_rcfw_wait_for_resp(rcfw, le16_to_cpu(req.cookie))) { + /* Cmd timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: SP: ALLOC_MRW timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: ALLOC_MRW failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B)) + mrw->rkey = le32_to_cpu(resp->xid); + else + mrw->lkey = le32_to_cpu(resp->xid); + return 0; +} + +int bnxt_qplib_dereg_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw, + bool block) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_deregister_mr req; + struct creq_deregister_mr_resp *resp; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DEREGISTER_MR, cmd_flags); + + req.lkey = cpu_to_le32(mrw->lkey); + resp = (struct creq_deregister_mr_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, block); + if (!resp) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: DEREG_MR send failed"); + return -EINVAL; + } + if (block) + rc = bnxt_qplib_rcfw_block_for_resp(rcfw, + le16_to_cpu(req.cookie)); + else + rc = bnxt_qplib_rcfw_wait_for_resp(rcfw, + le16_to_cpu(req.cookie)); + if (!rc) { + /* Cmd timed out */ + dev_err(&res->pdev->dev, "QPLIB: SP: DEREG_MR timed out"); + return -ETIMEDOUT; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&rcfw->pdev->dev, "QPLIB: SP: DEREG_MR failed "); + dev_err(&rcfw->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + + /* Free the qplib's MR memory */ + if (mrw->hwq.max_elements) { + mrw->va = 0; + mrw->total_size = 0; + bnxt_qplib_free_hwq(res->pdev, &mrw->hwq); + } + + return 0; +} + +int bnxt_qplib_reg_mr(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr, + u64 *pbl_tbl, int num_pbls, bool block) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_register_mr req; + struct creq_register_mr_resp *resp; + u16 cmd_flags = 0, level; + int pg_ptrs, pages, i, rc; + dma_addr_t **pbl_ptr; + u32 pg_size; + + if (num_pbls) { + pg_ptrs = roundup_pow_of_two(num_pbls); + pages = pg_ptrs >> MAX_PBL_LVL_1_PGS_SHIFT; + if (!pages) + pages++; + + if (pages > MAX_PBL_LVL_1_PGS) { + dev_err(&res->pdev->dev, "QPLIB: SP: Reg MR pages "); + dev_err(&res->pdev->dev, + "requested (0x%x) exceeded max (0x%x)", + pages, MAX_PBL_LVL_1_PGS); + return -ENOMEM; + } + /* Free the hwq if it already exist, must be a rereg */ + if (mr->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mr->hwq); + + mr->hwq.max_elements = pages; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &mr->hwq, NULL, 0, + &mr->hwq.max_elements, + PAGE_SIZE, 0, PAGE_SIZE, + HWQ_TYPE_CTX); + if (rc) { + dev_err(&res->pdev->dev, + "SP: Reg MR memory allocation failed"); + return -ENOMEM; + } + /* Write to the hwq */ + pbl_ptr = (dma_addr_t **)mr->hwq.pbl_ptr; + for (i = 0; i < num_pbls; i++) + pbl_ptr[PTR_PG(i)][PTR_IDX(i)] = + (pbl_tbl[i] & PAGE_MASK) | PTU_PTE_VALID; + } + + RCFW_CMD_PREP(req, REGISTER_MR, cmd_flags); + + /* Configure the request */ + if (mr->hwq.level == PBL_LVL_MAX) { + level = 0; + req.pbl = 0; + pg_size = PAGE_SIZE; + } else { + level = mr->hwq.level + 1; + req.pbl = cpu_to_le64(mr->hwq.pbl[PBL_LVL_0].pg_map_arr[0]); + pg_size = mr->hwq.pbl[PBL_LVL_0].pg_size; + } + req.log2_pg_size_lvl = (level << CMDQ_REGISTER_MR_LVL_SFT) | + ((ilog2(pg_size) << + CMDQ_REGISTER_MR_LOG2_PG_SIZE_SFT) & + CMDQ_REGISTER_MR_LOG2_PG_SIZE_MASK); + req.access = (mr->flags & 0xFFFF); + req.va = cpu_to_le64(mr->va); + req.key = cpu_to_le32(mr->lkey); + req.mr_size = cpu_to_le64(mr->total_size); + + resp = (struct creq_register_mr_resp *) + bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + NULL, block); + if (!resp) { + dev_err(&res->pdev->dev, "SP: REG_MR send failed"); + rc = -EINVAL; + goto fail; + } + if (block) + rc = bnxt_qplib_rcfw_block_for_resp(rcfw, + le16_to_cpu(req.cookie)); + else + rc = bnxt_qplib_rcfw_wait_for_resp(rcfw, + le16_to_cpu(req.cookie)); + if (!rc) { + /* Cmd timed out */ + dev_err(&res->pdev->dev, "SP: REG_MR timed out"); + rc = -ETIMEDOUT; + goto fail; + } + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&res->pdev->dev, "QPLIB: SP: REG_MR failed "); + dev_err(&res->pdev->dev, + "QPLIB: SP: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + rc = -EINVAL; + goto fail; + } + return 0; + +fail: + if (mr->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mr->hwq); + return rc; +} + +int bnxt_qplib_alloc_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl, + int max_pg_ptrs) +{ + int pg_ptrs, pages, rc; + + /* Re-calculate the max to fit the HWQ allocation model */ + pg_ptrs = roundup_pow_of_two(max_pg_ptrs); + pages = pg_ptrs >> MAX_PBL_LVL_1_PGS_SHIFT; + if (!pages) + pages++; + + if (pages > MAX_PBL_LVL_1_PGS) + return -ENOMEM; + + frpl->hwq.max_elements = pages; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &frpl->hwq, NULL, 0, + &frpl->hwq.max_elements, PAGE_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (!rc) + frpl->max_pg_ptrs = pg_ptrs; + + return rc; +} + +int bnxt_qplib_free_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl) +{ + bnxt_qplib_free_hwq(res->pdev, &frpl->hwq); + return 0; +} + +int bnxt_qplib_map_tc2cos(struct bnxt_qplib_res *res, u16 *cids) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_map_tc_to_cos req; + struct creq_map_tc_to_cos_resp *resp; + u16 cmd_flags = 0; + int tleft; + + RCFW_CMD_PREP(req, MAP_TC_TO_COS, cmd_flags); + req.cos0 = cpu_to_le16(cids[0]); + req.cos1 = cpu_to_le16(cids[1]); + + resp = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, NULL, 0); + if (!resp) { + dev_err(&res->pdev->dev, "QPLIB: SP: MAP_TC2COS send failed"); + return -EINVAL; + } + + tleft = bnxt_qplib_rcfw_block_for_resp(rcfw, le16_to_cpu(req.cookie)); + if (!tleft) { + dev_err(&res->pdev->dev, "QPLIB: SP: MAP_TC2COS timed out"); + return -ETIMEDOUT; + } + + if (resp->status || + le16_to_cpu(resp->cookie) != le16_to_cpu(req.cookie)) { + dev_err(&res->pdev->dev, "QPLIB: SP: MAP_TC2COS failed "); + dev_err(&res->pdev->dev, + "QPLIB: with status 0x%x cmdq 0x%x resp 0x%x", + resp->status, le16_to_cpu(req.cookie), + le16_to_cpu(resp->cookie)); + return -EINVAL; + } + + return 0; +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.h b/drivers/infiniband/hw/bnxt_re/qplib_sp.h new file mode 100644 index 000000000000..1442a617e968 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.h @@ -0,0 +1,160 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators (header) + * + */ + +#ifndef __BNXT_QPLIB_SP_H__ +#define __BNXT_QPLIB_SP_H__ + +struct bnxt_qplib_dev_attr { + char fw_ver[32]; + u16 max_sgid; + u16 max_mrw; + u32 max_qp; +#define BNXT_QPLIB_MAX_OUT_RD_ATOM 126 + u32 max_qp_rd_atom; + u32 max_qp_init_rd_atom; + u32 max_qp_wqes; + u32 max_qp_sges; + u32 max_cq; + u32 max_cq_wqes; + u32 max_cq_sges; + u32 max_mr; + u64 max_mr_size; + u32 max_pd; + u32 max_mw; + u32 max_raw_ethy_qp; + u32 max_ah; + u32 max_fmr; + u32 max_map_per_fmr; + u32 max_srq; + u32 max_srq_wqes; + u32 max_srq_sges; + u32 max_pkey; + u32 max_inline_data; + u32 l2_db_size; + u8 tqm_alloc_reqs[MAX_TQM_ALLOC_REQ]; +}; + +struct bnxt_qplib_pd { + u32 id; +}; + +struct bnxt_qplib_gid { + u8 data[16]; +}; + +struct bnxt_qplib_ah { + struct bnxt_qplib_gid dgid; + struct bnxt_qplib_pd *pd; + u32 id; + u8 sgid_index; + /* For Query AH if the hw table and SW table are differnt */ + u8 host_sgid_index; + u8 traffic_class; + u32 flow_label; + u8 hop_limit; + u8 sl; + u8 dmac[6]; + u16 vlan_id; + u8 nw_type; +}; + +struct bnxt_qplib_mrw { + struct bnxt_qplib_pd *pd; + int type; + u32 flags; +#define BNXT_QPLIB_FR_PMR 0x80000000 + u32 lkey; + u32 rkey; +#define BNXT_QPLIB_RSVD_LKEY 0xFFFFFFFF + u64 va; + u64 total_size; + u32 npages; + u64 mr_handle; + struct bnxt_qplib_hwq hwq; +}; + +struct bnxt_qplib_frpl { + int max_pg_ptrs; + struct bnxt_qplib_hwq hwq; +}; + +#define BNXT_QPLIB_ACCESS_LOCAL_WRITE BIT(0) +#define BNXT_QPLIB_ACCESS_REMOTE_READ BIT(1) +#define BNXT_QPLIB_ACCESS_REMOTE_WRITE BIT(2) +#define BNXT_QPLIB_ACCESS_REMOTE_ATOMIC BIT(3) +#define BNXT_QPLIB_ACCESS_MW_BIND BIT(4) +#define BNXT_QPLIB_ACCESS_ZERO_BASED BIT(5) +#define BNXT_QPLIB_ACCESS_ON_DEMAND BIT(6) + +int bnxt_qplib_get_sgid(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl, int index, + struct bnxt_qplib_gid *gid); +int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct bnxt_qplib_gid *gid, bool update); +int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct bnxt_qplib_gid *gid, u8 *mac, u16 vlan_id, + bool update, u32 *index); +int bnxt_qplib_get_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 index, + u16 *pkey); +int bnxt_qplib_del_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update); +int bnxt_qplib_add_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update); +int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_dev_attr *attr); +int bnxt_qplib_create_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah); +int bnxt_qplib_destroy_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah); +int bnxt_qplib_alloc_mrw(struct bnxt_qplib_res *res, + struct bnxt_qplib_mrw *mrw); +int bnxt_qplib_dereg_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw, + bool block); +int bnxt_qplib_reg_mr(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr, + u64 *pbl_tbl, int num_pbls, bool block); +int bnxt_qplib_free_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr); +int bnxt_qplib_alloc_fast_reg_mr(struct bnxt_qplib_res *res, + struct bnxt_qplib_mrw *mr, int max); +int bnxt_qplib_alloc_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl, int max); +int bnxt_qplib_free_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl); +int bnxt_qplib_map_tc2cos(struct bnxt_qplib_res *res, u16 *cids); +#endif /* __BNXT_QPLIB_SP_H__*/ diff --git a/drivers/infiniband/hw/bnxt_re/roce_hsi.h b/drivers/infiniband/hw/bnxt_re/roce_hsi.h new file mode 100644 index 000000000000..fc23477ac52f --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/roce_hsi.h @@ -0,0 +1,2821 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RoCE HSI File - Autogenerated + */ + +#ifndef __BNXT_RE_HSI_H__ +#define __BNXT_RE_HSI_H__ + +/* include bnxt_hsi.h from bnxt_en driver */ +#include "bnxt_hsi.h" + +/* CMP Door Bell Format (4 bytes) */ +struct cmpl_doorbell { + __le32 key_mask_valid_idx; + #define CMPL_DOORBELL_IDX_MASK 0xffffffUL + #define CMPL_DOORBELL_IDX_SFT 0 + #define CMPL_DOORBELL_RESERVED_MASK 0x3000000UL + #define CMPL_DOORBELL_RESERVED_SFT 24 + #define CMPL_DOORBELL_IDX_VALID 0x4000000UL + #define CMPL_DOORBELL_MASK 0x8000000UL + #define CMPL_DOORBELL_KEY_MASK 0xf0000000UL + #define CMPL_DOORBELL_KEY_SFT 28 + #define CMPL_DOORBELL_KEY_CMPL (0x2UL << 28) +}; + +/* Status Door Bell Format (4 bytes) */ +struct status_doorbell { + __le32 key_idx; + #define STATUS_DOORBELL_IDX_MASK 0xffffffUL + #define STATUS_DOORBELL_IDX_SFT 0 + #define STATUS_DOORBELL_RESERVED_MASK 0xf000000UL + #define STATUS_DOORBELL_RESERVED_SFT 24 + #define STATUS_DOORBELL_KEY_MASK 0xf0000000UL + #define STATUS_DOORBELL_KEY_SFT 28 + #define STATUS_DOORBELL_KEY_STAT (0x3UL << 28) +}; + +/* RoCE Host Structures */ + +/* Doorbell Structures */ +/* 64b Doorbell Format (8 bytes) */ +struct dbr_dbr { + __le32 index; + #define DBR_DBR_INDEX_MASK 0xfffffUL + #define DBR_DBR_INDEX_SFT 0 + #define DBR_DBR_RESERVED12_MASK 0xfff00000UL + #define DBR_DBR_RESERVED12_SFT 20 + __le32 type_xid; + #define DBR_DBR_XID_MASK 0xfffffUL + #define DBR_DBR_XID_SFT 0 + #define DBR_DBR_RESERVED8_MASK 0xff00000UL + #define DBR_DBR_RESERVED8_SFT 20 + #define DBR_DBR_TYPE_MASK 0xf0000000UL + #define DBR_DBR_TYPE_SFT 28 + #define DBR_DBR_TYPE_SQ (0x0UL << 28) + #define DBR_DBR_TYPE_RQ (0x1UL << 28) + #define DBR_DBR_TYPE_SRQ (0x2UL << 28) + #define DBR_DBR_TYPE_SRQ_ARM (0x3UL << 28) + #define DBR_DBR_TYPE_CQ (0x4UL << 28) + #define DBR_DBR_TYPE_CQ_ARMSE (0x5UL << 28) + #define DBR_DBR_TYPE_CQ_ARMALL (0x6UL << 28) + #define DBR_DBR_TYPE_CQ_ARMENA (0x7UL << 28) + #define DBR_DBR_TYPE_SRQ_ARMENA (0x8UL << 28) + #define DBR_DBR_TYPE_CQ_CUTOFF_ACK (0x9UL << 28) + #define DBR_DBR_TYPE_NULL (0xfUL << 28) +}; + +/* 32b Doorbell Format (4 bytes) */ +struct dbr_dbr32 { + __le32 type_abs_incr_xid; + #define DBR_DBR32_XID_MASK 0xfffffUL + #define DBR_DBR32_XID_SFT 0 + #define DBR_DBR32_RESERVED4_MASK 0xf00000UL + #define DBR_DBR32_RESERVED4_SFT 20 + #define DBR_DBR32_INCR_MASK 0xf000000UL + #define DBR_DBR32_INCR_SFT 24 + #define DBR_DBR32_ABS 0x10000000UL + #define DBR_DBR32_TYPE_MASK 0xe0000000UL + #define DBR_DBR32_TYPE_SFT 29 + #define DBR_DBR32_TYPE_SQ (0x0UL << 29) +}; + +/* SQ WQE Structures */ +/* Base SQ WQE (8 bytes) */ +struct sq_base { + u8 wqe_type; + #define SQ_BASE_WQE_TYPE_SEND 0x0UL + #define SQ_BASE_WQE_TYPE_SEND_W_IMMEAD 0x1UL + #define SQ_BASE_WQE_TYPE_SEND_W_INVALID 0x2UL + #define SQ_BASE_WQE_TYPE_WRITE_WQE 0x4UL + #define SQ_BASE_WQE_TYPE_WRITE_W_IMMEAD 0x5UL + #define SQ_BASE_WQE_TYPE_READ_WQE 0x6UL + #define SQ_BASE_WQE_TYPE_ATOMIC_CS 0x8UL + #define SQ_BASE_WQE_TYPE_ATOMIC_FA 0xbUL + #define SQ_BASE_WQE_TYPE_LOCAL_INVALID 0xcUL + #define SQ_BASE_WQE_TYPE_FR_PMR 0xdUL + #define SQ_BASE_WQE_TYPE_BIND 0xeUL + u8 unused_0[7]; +}; + +/* WQE SGE (16 bytes) */ +struct sq_sge { + __le64 va_or_pa; + __le32 l_key; + __le32 size; +}; + +/* PSN Search Structure (8 bytes) */ +struct sq_psn_search { + __le32 opcode_start_psn; + #define SQ_PSN_SEARCH_START_PSN_MASK 0xffffffUL + #define SQ_PSN_SEARCH_START_PSN_SFT 0 + #define SQ_PSN_SEARCH_OPCODE_MASK 0xff000000UL + #define SQ_PSN_SEARCH_OPCODE_SFT 24 + __le32 flags_next_psn; + #define SQ_PSN_SEARCH_NEXT_PSN_MASK 0xffffffUL + #define SQ_PSN_SEARCH_NEXT_PSN_SFT 0 + #define SQ_PSN_SEARCH_FLAGS_MASK 0xff000000UL + #define SQ_PSN_SEARCH_FLAGS_SFT 24 +}; + +/* Send SQ WQE (40 bytes) */ +struct sq_send { + u8 wqe_type; + #define SQ_SEND_WQE_TYPE_SEND 0x0UL + #define SQ_SEND_WQE_TYPE_SEND_W_IMMEAD 0x1UL + #define SQ_SEND_WQE_TYPE_SEND_W_INVALID 0x2UL + u8 flags; + #define SQ_SEND_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_SEND_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_SEND_FLAGS_UC_FENCE 0x4UL + #define SQ_SEND_FLAGS_SE 0x8UL + #define SQ_SEND_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8_1; + __le32 inv_key_or_imm_data; + __le32 length; + __le32 q_key; + __le32 dst_qp; + #define SQ_SEND_DST_QP_MASK 0xffffffUL + #define SQ_SEND_DST_QP_SFT 0 + #define SQ_SEND_RESERVED8_2_MASK 0xff000000UL + #define SQ_SEND_RESERVED8_2_SFT 24 + __le32 avid; + #define SQ_SEND_AVID_MASK 0xfffffUL + #define SQ_SEND_AVID_SFT 0 + #define SQ_SEND_RESERVED_AVID_MASK 0xfff00000UL + #define SQ_SEND_RESERVED_AVID_SFT 20 + __le64 reserved64; + __le32 data[24]; +}; + +/* Send Raw Ethernet and QP1 SQ WQE (40 bytes) */ +struct sq_send_raweth_qp1 { + u8 wqe_type; + #define SQ_SEND_RAWETH_QP1_WQE_TYPE_SEND 0x0UL + u8 flags; + #define SQ_SEND_RAWETH_QP1_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_SEND_RAWETH_QP1_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_SEND_RAWETH_QP1_FLAGS_UC_FENCE 0x4UL + #define SQ_SEND_RAWETH_QP1_FLAGS_SE 0x8UL + #define SQ_SEND_RAWETH_QP1_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8; + __le16 lflags; + #define SQ_SEND_RAWETH_QP1_LFLAGS_TCP_UDP_CHKSUM 0x1UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_IP_CHKSUM 0x2UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_NOCRC 0x4UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_STAMP 0x8UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_T_IP_CHKSUM 0x10UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_1 0x20UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_2 0x40UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_3 0x80UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_ROCE_CRC 0x100UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_FCOE_CRC 0x200UL + __le16 cfa_action; + __le32 length; + __le32 reserved32_1; + __le32 cfa_meta; + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_MASK 0xfffUL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_SFT 0 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_DE 0x1000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_PRI_MASK 0xe000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_PRI_SFT 13 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_MASK 0x70000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_SFT 16 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID88A8 (0x0UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID8100 (0x1UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9100 (0x2UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9200 (0x3UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9300 (0x4UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPIDCFG (0x5UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_LAST \ + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPIDCFG + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_RESERVED_MASK 0xff80000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_RESERVED_SFT 19 + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_MASK 0xf0000000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_SFT 28 + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_NONE (0x0UL << 28) + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_VLAN_TAG (0x1UL << 28) + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_LAST \ + SQ_SEND_RAWETH_QP1_CFA_META_KEY_VLAN_TAG + __le32 reserved32_2; + __le64 reserved64; + __le32 data[24]; +}; + +/* RDMA SQ WQE (40 bytes) */ +struct sq_rdma { + u8 wqe_type; + #define SQ_RDMA_WQE_TYPE_WRITE_WQE 0x4UL + #define SQ_RDMA_WQE_TYPE_WRITE_W_IMMEAD 0x5UL + #define SQ_RDMA_WQE_TYPE_READ_WQE 0x6UL + u8 flags; + #define SQ_RDMA_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_RDMA_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_RDMA_FLAGS_UC_FENCE 0x4UL + #define SQ_RDMA_FLAGS_SE 0x8UL + #define SQ_RDMA_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8; + __le32 imm_data; + __le32 length; + __le32 reserved32_1; + __le64 remote_va; + __le32 remote_key; + __le32 reserved32_2; + __le32 data[24]; +}; + +/* Atomic SQ WQE (40 bytes) */ +struct sq_atomic { + u8 wqe_type; + #define SQ_ATOMIC_WQE_TYPE_ATOMIC_CS 0x8UL + #define SQ_ATOMIC_WQE_TYPE_ATOMIC_FA 0xbUL + u8 flags; + #define SQ_ATOMIC_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_ATOMIC_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_ATOMIC_FLAGS_UC_FENCE 0x4UL + #define SQ_ATOMIC_FLAGS_SE 0x8UL + #define SQ_ATOMIC_FLAGS_INLINE 0x10UL + __le16 reserved16; + __le32 remote_key; + __le64 remote_va; + __le64 swap_data; + __le64 cmp_data; + __le32 data[24]; +}; + +/* Local Invalidate SQ WQE (40 bytes) */ +struct sq_localinvalidate { + u8 wqe_type; + #define SQ_LOCALINVALIDATE_WQE_TYPE_LOCAL_INVALID 0xcUL + u8 flags; + #define SQ_LOCALINVALIDATE_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_LOCALINVALIDATE_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_LOCALINVALIDATE_FLAGS_UC_FENCE 0x4UL + #define SQ_LOCALINVALIDATE_FLAGS_SE 0x8UL + #define SQ_LOCALINVALIDATE_FLAGS_INLINE 0x10UL + __le16 reserved16; + __le32 inv_l_key; + __le64 reserved64; + __le32 reserved128[4]; + __le32 data[24]; +}; + +/* FR-PMR SQ WQE (40 bytes) */ +struct sq_fr_pmr { + u8 wqe_type; + #define SQ_FR_PMR_WQE_TYPE_FR_PMR 0xdUL + u8 flags; + #define SQ_FR_PMR_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_FR_PMR_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_FR_PMR_FLAGS_UC_FENCE 0x4UL + #define SQ_FR_PMR_FLAGS_SE 0x8UL + #define SQ_FR_PMR_FLAGS_INLINE 0x10UL + u8 access_cntl; + #define SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE 0x1UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_READ 0x2UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_WRITE 0x4UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_ATOMIC 0x8UL + #define SQ_FR_PMR_ACCESS_CNTL_WINDOW_BIND 0x10UL + u8 zero_based_page_size_log; + #define SQ_FR_PMR_PAGE_SIZE_LOG_MASK 0x1fUL + #define SQ_FR_PMR_PAGE_SIZE_LOG_SFT 0 + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_4K 0x0UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_8K 0x1UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_64K 0x4UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_256K 0x6UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_1M 0x8UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_2M 0x9UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_4M 0xaUL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_1G 0x12UL + #define SQ_FR_PMR_ZERO_BASED 0x20UL + #define SQ_FR_PMR_RESERVED2_MASK 0xc0UL + #define SQ_FR_PMR_RESERVED2_SFT 6 + __le32 l_key; + u8 length[5]; + u8 reserved8_1; + u8 reserved8_2; + u8 numlevels_pbl_page_size_log; + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_MASK 0x1fUL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_SFT 0 + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_4K 0x0UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_8K 0x1UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_64K 0x4UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_256K 0x6UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_1M 0x8UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_2M 0x9UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_4M 0xaUL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_1G 0x12UL + #define SQ_FR_PMR_RESERVED1 0x20UL + #define SQ_FR_PMR_NUMLEVELS_MASK 0xc0UL + #define SQ_FR_PMR_NUMLEVELS_SFT 6 + #define SQ_FR_PMR_NUMLEVELS_PHYSICAL (0x0UL << 6) + #define SQ_FR_PMR_NUMLEVELS_LAYER1 (0x1UL << 6) + #define SQ_FR_PMR_NUMLEVELS_LAYER2 (0x2UL << 6) + __le64 pblptr; + __le64 va; + __le32 data[24]; +}; + +/* Bind SQ WQE (40 bytes) */ +struct sq_bind { + u8 wqe_type; + #define SQ_BIND_WQE_TYPE_BIND 0xeUL + u8 flags; + #define SQ_BIND_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_BIND_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_BIND_FLAGS_UC_FENCE 0x4UL + #define SQ_BIND_FLAGS_SE 0x8UL + #define SQ_BIND_FLAGS_INLINE 0x10UL + u8 access_cntl; + #define SQ_BIND_ACCESS_CNTL_LOCAL_WRITE 0x1UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_READ 0x2UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_WRITE 0x4UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_ATOMIC 0x8UL + #define SQ_BIND_ACCESS_CNTL_WINDOW_BIND 0x10UL + u8 reserved8_1; + u8 mw_type_zero_based; + #define SQ_BIND_ZERO_BASED 0x1UL + #define SQ_BIND_MW_TYPE 0x2UL + #define SQ_BIND_MW_TYPE_TYPE1 (0x0UL << 1) + #define SQ_BIND_MW_TYPE_TYPE2 (0x1UL << 1) + #define SQ_BIND_RESERVED6_MASK 0xfcUL + #define SQ_BIND_RESERVED6_SFT 2 + u8 reserved8_2; + __le16 reserved16; + __le32 parent_l_key; + __le32 l_key; + __le64 va; + u8 length[5]; + u8 data_reserved24[99]; + #define SQ_BIND_RESERVED24_MASK 0xffffff00UL + #define SQ_BIND_RESERVED24_SFT 8 + #define SQ_BIND_DATA_MASK 0xffffffffUL + #define SQ_BIND_DATA_SFT 0 +}; + +/* RQ/SRQ WQE Structures */ +/* RQ/SRQ WQE (40 bytes) */ +struct rq_wqe { + u8 wqe_type; + #define RQ_WQE_WQE_TYPE_RCV 0x80UL + u8 flags; + u8 wqe_size; + u8 reserved8; + __le32 reserved32; + __le32 wr_id[2]; + #define RQ_WQE_WR_ID_MASK 0xfffffUL + #define RQ_WQE_WR_ID_SFT 0 + #define RQ_WQE_RESERVED44_MASK 0xfff00000UL + #define RQ_WQE_RESERVED44_SFT 20 + __le32 reserved128[4]; + __le32 data[24]; +}; + +/* CQ CQE Structures */ +/* Base CQE (32 bytes) */ +struct cq_base { + __le64 reserved64_1; + __le64 reserved64_2; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_BASE_TOGGLE 0x1UL + #define CQ_BASE_CQE_TYPE_MASK 0x1eUL + #define CQ_BASE_CQE_TYPE_SFT 1 + #define CQ_BASE_CQE_TYPE_REQ (0x0UL << 1) + #define CQ_BASE_CQE_TYPE_RES_RC (0x1UL << 1) + #define CQ_BASE_CQE_TYPE_RES_UD (0x2UL << 1) + #define CQ_BASE_CQE_TYPE_RES_RAWETH_QP1 (0x3UL << 1) + #define CQ_BASE_CQE_TYPE_TERMINAL (0xeUL << 1) + #define CQ_BASE_CQE_TYPE_CUT_OFF (0xfUL << 1) + #define CQ_BASE_RESERVED3_MASK 0xe0UL + #define CQ_BASE_RESERVED3_SFT 5 + u8 status; + __le16 reserved16; + __le32 reserved32; +}; + +/* Requester CQ CQE (32 bytes) */ +struct cq_req { + __le64 qp_handle; + __le16 sq_cons_idx; + __le16 reserved16_1; + __le32 reserved32_2; + __le64 reserved64; + u8 cqe_type_toggle; + #define CQ_REQ_TOGGLE 0x1UL + #define CQ_REQ_CQE_TYPE_MASK 0x1eUL + #define CQ_REQ_CQE_TYPE_SFT 1 + #define CQ_REQ_CQE_TYPE_REQ (0x0UL << 1) + #define CQ_REQ_RESERVED3_MASK 0xe0UL + #define CQ_REQ_RESERVED3_SFT 5 + u8 status; + #define CQ_REQ_STATUS_OK 0x0UL + #define CQ_REQ_STATUS_BAD_RESPONSE_ERR 0x1UL + #define CQ_REQ_STATUS_LOCAL_LENGTH_ERR 0x2UL + #define CQ_REQ_STATUS_LOCAL_QP_OPERATION_ERR 0x3UL + #define CQ_REQ_STATUS_LOCAL_PROTECTION_ERR 0x4UL + #define CQ_REQ_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_REQ_STATUS_REMOTE_INVALID_REQUEST_ERR 0x6UL + #define CQ_REQ_STATUS_REMOTE_ACCESS_ERR 0x7UL + #define CQ_REQ_STATUS_REMOTE_OPERATION_ERR 0x8UL + #define CQ_REQ_STATUS_RNR_NAK_RETRY_CNT_ERR 0x9UL + #define CQ_REQ_STATUS_TRANSPORT_RETRY_CNT_ERR 0xaUL + #define CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR 0xbUL + __le16 reserved16_2; + __le32 reserved32_1; +}; + +/* Responder RC CQE (32 bytes) */ +struct cq_res_rc { + __le32 length; + __le32 imm_data_or_inv_r_key; + __le64 qp_handle; + __le64 mr_handle; + u8 cqe_type_toggle; + #define CQ_RES_RC_TOGGLE 0x1UL + #define CQ_RES_RC_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_RC_CQE_TYPE_SFT 1 + #define CQ_RES_RC_CQE_TYPE_RES_RC (0x1UL << 1) + #define CQ_RES_RC_RESERVED3_MASK 0xe0UL + #define CQ_RES_RC_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_RC_STATUS_OK 0x0UL + #define CQ_RES_RC_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_RC_STATUS_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_RC_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_RC_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_RC_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_RC_STATUS_REMOTE_INVALID_REQUEST_ERR 0x6UL + #define CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_RC_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_RC_FLAGS_SRQ 0x1UL + #define CQ_RES_RC_FLAGS_SRQ_RQ (0x0UL << 0) + #define CQ_RES_RC_FLAGS_SRQ_SRQ (0x1UL << 0) + #define CQ_RES_RC_FLAGS_SRQ_LAST CQ_RES_RC_FLAGS_SRQ_SRQ + #define CQ_RES_RC_FLAGS_IMM 0x2UL + #define CQ_RES_RC_FLAGS_INV 0x4UL + #define CQ_RES_RC_FLAGS_RDMA 0x8UL + #define CQ_RES_RC_FLAGS_RDMA_SEND (0x0UL << 3) + #define CQ_RES_RC_FLAGS_RDMA_RDMA_WRITE (0x1UL << 3) + #define CQ_RES_RC_FLAGS_RDMA_LAST CQ_RES_RC_FLAGS_RDMA_RDMA_WRITE + __le32 srq_or_rq_wr_id; + #define CQ_RES_RC_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_RC_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_RC_RESERVED12_MASK 0xfff00000UL + #define CQ_RES_RC_RESERVED12_SFT 20 +}; + +/* Responder UD CQE (32 bytes) */ +struct cq_res_ud { + __le32 length; + #define CQ_RES_UD_LENGTH_MASK 0x3fffUL + #define CQ_RES_UD_LENGTH_SFT 0 + #define CQ_RES_UD_RESERVED18_MASK 0xffffc000UL + #define CQ_RES_UD_RESERVED18_SFT 14 + __le32 imm_data; + __le64 qp_handle; + __le16 src_mac[3]; + __le16 src_qp_low; + u8 cqe_type_toggle; + #define CQ_RES_UD_TOGGLE 0x1UL + #define CQ_RES_UD_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_UD_CQE_TYPE_SFT 1 + #define CQ_RES_UD_CQE_TYPE_RES_UD (0x2UL << 1) + #define CQ_RES_UD_RESERVED3_MASK 0xe0UL + #define CQ_RES_UD_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_UD_STATUS_OK 0x0UL + #define CQ_RES_UD_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_UD_STATUS_HW_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_UD_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_UD_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_UD_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_UD_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_UD_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_UD_FLAGS_SRQ 0x1UL + #define CQ_RES_UD_FLAGS_SRQ_RQ (0x0UL << 0) + #define CQ_RES_UD_FLAGS_SRQ_SRQ (0x1UL << 0) + #define CQ_RES_UD_FLAGS_SRQ_LAST CQ_RES_UD_FLAGS_SRQ_SRQ + #define CQ_RES_UD_FLAGS_IMM 0x2UL + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_MASK 0xcUL + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_SFT 2 + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V1 (0x0UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV4 (0x2UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV6 (0x3UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_LAST \ + CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV6 + __le32 src_qp_high_srq_or_rq_wr_id; + #define CQ_RES_UD_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_UD_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_UD_RESERVED4_MASK 0xf00000UL + #define CQ_RES_UD_RESERVED4_SFT 20 + #define CQ_RES_UD_SRC_QP_HIGH_MASK 0xff000000UL + #define CQ_RES_UD_SRC_QP_HIGH_SFT 24 +}; + +/* Responder RawEth and QP1 CQE (32 bytes) */ +struct cq_res_raweth_qp1 { + __le16 length; + #define CQ_RES_RAWETH_QP1_LENGTH_MASK 0x3fffUL + #define CQ_RES_RAWETH_QP1_LENGTH_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED2_MASK 0xc000UL + #define CQ_RES_RAWETH_QP1_RESERVED2_SFT 14 + __le16 raweth_qp1_flags; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ERROR 0x1UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_RESERVED5_1_MASK 0x3eUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_RESERVED5_1_SFT 1 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_MASK 0x3c0UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_SFT 6 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_NOT_KNOWN (0x0UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_IP (0x1UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_TCP (0x2UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_UDP (0x3UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_FCOE (0x4UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE (0x5UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ICMP (0x7UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_WO_TIMESTAMP \ + (0x8UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_W_TIMESTAMP \ + (0x9UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_W_TIMESTAMP + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_MASK 0x3ffUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED6_MASK 0xfc00UL + #define CQ_RES_RAWETH_QP1_RESERVED6_SFT 10 + __le16 raweth_qp1_errors; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_RESERVED4_MASK 0xfUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_RESERVED4_SFT 0 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_IP_CS_ERROR 0x10UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_L4_CS_ERROR 0x20UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_IP_CS_ERROR 0x40UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_L4_CS_ERROR 0x80UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_CRC_ERROR 0x100UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_MASK 0xe00UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_SFT 9 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_NO_ERROR \ + (0x0UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_VERSION \ + (0x1UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_HDR_LEN \ + (0x2UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_TUNNEL_TOTAL_ERROR \ + (0x3UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_IP_TOTAL_ERROR \ + (0x4UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_UDP_TOTAL_ERROR \ + (0x5UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_TTL \ + (0x6UL << 9) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_TTL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_MASK 0xf000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_SFT 12 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_NO_ERROR \ + (0x0UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_VERSION \ + (0x1UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_HDR_LEN \ + (0x2UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_TTL \ + (0x3UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_IP_TOTAL_ERROR \ + (0x4UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_UDP_TOTAL_ERROR \ + (0x5UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_HDR_LEN \ + (0x6UL << 12) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_HDR_LEN_TOO_SMALL\ + (0x7UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_OPT_LEN \ + (0x8UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_OPT_LEN + __le16 raweth_qp1_cfa_code; + __le64 qp_handle; + __le32 raweth_qp1_flags2; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_CS_CALC 0x1UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_L4_CS_CALC 0x2UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_T_IP_CS_CALC 0x4UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_T_L4_CS_CALC 0x8UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_MASK 0xf0UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_SFT 4 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_NONE \ + (0x0UL << 4) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_VLAN \ + (0x1UL << 4) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_LAST\ + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_VLAN + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_TYPE 0x100UL + __le32 raweth_qp1_metadata; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_VID_MASK 0xfffUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_VID_SFT 0 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_DE 0x1000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_PRI_MASK 0xe000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_PRI_SFT 13 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_TPID_MASK 0xffff0000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_TPID_SFT 16 + u8 cqe_type_toggle; + #define CQ_RES_RAWETH_QP1_TOGGLE 0x1UL + #define CQ_RES_RAWETH_QP1_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_RAWETH_QP1_CQE_TYPE_SFT 1 + #define CQ_RES_RAWETH_QP1_CQE_TYPE_RES_RAWETH_QP1 (0x3UL << 1) + #define CQ_RES_RAWETH_QP1_RESERVED3_MASK 0xe0UL + #define CQ_RES_RAWETH_QP1_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_RAWETH_QP1_STATUS_OK 0x0UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_RAWETH_QP1_STATUS_HW_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_RAWETH_QP1_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_RAWETH_QP1_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_RAWETH_QP1_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ 0x1UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_RQ 0x0UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_SRQ 0x1UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_LAST \ + CQ_RES_RAWETH_QP1_FLAGS_SRQ_SRQ + __le32 raweth_qp1_payload_offset_srq_or_rq_wr_id; + #define CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED4_MASK 0xf00000UL + #define CQ_RES_RAWETH_QP1_RESERVED4_SFT 20 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_PAYLOAD_OFFSET_MASK 0xff000000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_PAYLOAD_OFFSET_SFT 24 +}; + +/* Terminal CQE (32 bytes) */ +struct cq_terminal { + __le64 qp_handle; + __le16 sq_cons_idx; + __le16 rq_cons_idx; + __le32 reserved32_1; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_TERMINAL_TOGGLE 0x1UL + #define CQ_TERMINAL_CQE_TYPE_MASK 0x1eUL + #define CQ_TERMINAL_CQE_TYPE_SFT 1 + #define CQ_TERMINAL_CQE_TYPE_TERMINAL (0xeUL << 1) + #define CQ_TERMINAL_RESERVED3_MASK 0xe0UL + #define CQ_TERMINAL_RESERVED3_SFT 5 + u8 status; + #define CQ_TERMINAL_STATUS_OK 0x0UL + __le16 reserved16; + __le32 reserved32_2; +}; + +/* Cutoff CQE (32 bytes) */ +struct cq_cutoff { + __le64 reserved64_1; + __le64 reserved64_2; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_CUTOFF_TOGGLE 0x1UL + #define CQ_CUTOFF_CQE_TYPE_MASK 0x1eUL + #define CQ_CUTOFF_CQE_TYPE_SFT 1 + #define CQ_CUTOFF_CQE_TYPE_CUT_OFF (0xfUL << 1) + #define CQ_CUTOFF_RESERVED3_MASK 0xe0UL + #define CQ_CUTOFF_RESERVED3_SFT 5 + u8 status; + #define CQ_CUTOFF_STATUS_OK 0x0UL + __le16 reserved16; + __le32 reserved32; +}; + +/* Notification Queue (NQ) Structures */ +/* Base NQ Record (16 bytes) */ +struct nq_base { + __le16 info10_type; + #define NQ_BASE_TYPE_MASK 0x3fUL + #define NQ_BASE_TYPE_SFT 0 + #define NQ_BASE_TYPE_CQ_NOTIFICATION 0x30UL + #define NQ_BASE_TYPE_SRQ_EVENT 0x32UL + #define NQ_BASE_TYPE_DBQ_EVENT 0x34UL + #define NQ_BASE_TYPE_QP_EVENT 0x38UL + #define NQ_BASE_TYPE_FUNC_EVENT 0x3aUL + #define NQ_BASE_INFO10_MASK 0xffc0UL + #define NQ_BASE_INFO10_SFT 6 + __le16 info16; + __le32 info32; + __le32 info63_v[2]; + #define NQ_BASE_V 0x1UL + #define NQ_BASE_INFO63_MASK 0xfffffffeUL + #define NQ_BASE_INFO63_SFT 1 +}; + +/* Completion Queue Notification (16 bytes) */ +struct nq_cn { + __le16 type; + #define NQ_CN_TYPE_MASK 0x3fUL + #define NQ_CN_TYPE_SFT 0 + #define NQ_CN_TYPE_CQ_NOTIFICATION 0x30UL + #define NQ_CN_RESERVED9_MASK 0xffc0UL + #define NQ_CN_RESERVED9_SFT 6 + __le16 reserved16; + __le32 cq_handle_low; + __le32 v; + #define NQ_CN_V 0x1UL + #define NQ_CN_RESERVED31_MASK 0xfffffffeUL + #define NQ_CN_RESERVED31_SFT 1 + __le32 cq_handle_high; +}; + +/* SRQ Event Notification (16 bytes) */ +struct nq_srq_event { + u8 type; + #define NQ_SRQ_EVENT_TYPE_MASK 0x3fUL + #define NQ_SRQ_EVENT_TYPE_SFT 0 + #define NQ_SRQ_EVENT_TYPE_SRQ_EVENT 0x32UL + #define NQ_SRQ_EVENT_RESERVED1_MASK 0xc0UL + #define NQ_SRQ_EVENT_RESERVED1_SFT 6 + u8 event; + #define NQ_SRQ_EVENT_EVENT_SRQ_THRESHOLD_EVENT 0x1UL + __le16 reserved16; + __le32 srq_handle_low; + __le32 v; + #define NQ_SRQ_EVENT_V 0x1UL + #define NQ_SRQ_EVENT_RESERVED31_MASK 0xfffffffeUL + #define NQ_SRQ_EVENT_RESERVED31_SFT 1 + __le32 srq_handle_high; +}; + +/* DBQ Async Event Notification (16 bytes) */ +struct nq_dbq_event { + u8 type; + #define NQ_DBQ_EVENT_TYPE_MASK 0x3fUL + #define NQ_DBQ_EVENT_TYPE_SFT 0 + #define NQ_DBQ_EVENT_TYPE_DBQ_EVENT 0x34UL + #define NQ_DBQ_EVENT_RESERVED1_MASK 0xc0UL + #define NQ_DBQ_EVENT_RESERVED1_SFT 6 + u8 event; + #define NQ_DBQ_EVENT_EVENT_DBQ_THRESHOLD_EVENT 0x1UL + __le16 db_pfid; + #define NQ_DBQ_EVENT_DB_PFID_MASK 0xfUL + #define NQ_DBQ_EVENT_DB_PFID_SFT 0 + #define NQ_DBQ_EVENT_RESERVED12_MASK 0xfff0UL + #define NQ_DBQ_EVENT_RESERVED12_SFT 4 + __le32 db_dpi; + #define NQ_DBQ_EVENT_DB_DPI_MASK 0xfffffUL + #define NQ_DBQ_EVENT_DB_DPI_SFT 0 + #define NQ_DBQ_EVENT_RESERVED12_2_MASK 0xfff00000UL + #define NQ_DBQ_EVENT_RESERVED12_2_SFT 20 + __le32 v; + #define NQ_DBQ_EVENT_V 0x1UL + #define NQ_DBQ_EVENT_RESERVED32_MASK 0xfffffffeUL + #define NQ_DBQ_EVENT_RESERVED32_SFT 1 + __le32 db_type_db_xid; + #define NQ_DBQ_EVENT_DB_XID_MASK 0xfffffUL + #define NQ_DBQ_EVENT_DB_XID_SFT 0 + #define NQ_DBQ_EVENT_RESERVED8_MASK 0xff00000UL + #define NQ_DBQ_EVENT_RESERVED8_SFT 20 + #define NQ_DBQ_EVENT_DB_TYPE_MASK 0xf0000000UL + #define NQ_DBQ_EVENT_DB_TYPE_SFT 28 +}; + +/* Read Request/Response Queue Structures */ +/* Input Read Request Queue (IRRQ) Message (32 bytes) */ +struct xrrq_irrq { + __le16 credits_type; + #define XRRQ_IRRQ_TYPE 0x1UL + #define XRRQ_IRRQ_TYPE_READ_REQ 0x0UL + #define XRRQ_IRRQ_TYPE_ATOMIC_REQ 0x1UL + #define XRRQ_IRRQ_RESERVED10_MASK 0x7feUL + #define XRRQ_IRRQ_RESERVED10_SFT 1 + #define XRRQ_IRRQ_CREDITS_MASK 0xf800UL + #define XRRQ_IRRQ_CREDITS_SFT 11 + __le16 reserved16; + __le32 reserved32; + __le32 psn; + #define XRRQ_IRRQ_PSN_MASK 0xffffffUL + #define XRRQ_IRRQ_PSN_SFT 0 + #define XRRQ_IRRQ_RESERVED8_1_MASK 0xff000000UL + #define XRRQ_IRRQ_RESERVED8_1_SFT 24 + __le32 msn; + #define XRRQ_IRRQ_MSN_MASK 0xffffffUL + #define XRRQ_IRRQ_MSN_SFT 0 + #define XRRQ_IRRQ_RESERVED8_2_MASK 0xff000000UL + #define XRRQ_IRRQ_RESERVED8_2_SFT 24 + __le64 va_or_atomic_result; + __le32 rdma_r_key; + __le32 length; +}; + +/* Output Read Request Queue (ORRQ) Message (32 bytes) */ +struct xrrq_orrq { + __le16 num_sges_type; + #define XRRQ_ORRQ_TYPE 0x1UL + #define XRRQ_ORRQ_TYPE_READ_REQ 0x0UL + #define XRRQ_ORRQ_TYPE_ATOMIC_REQ 0x1UL + #define XRRQ_ORRQ_RESERVED10_MASK 0x7feUL + #define XRRQ_ORRQ_RESERVED10_SFT 1 + #define XRRQ_ORRQ_NUM_SGES_MASK 0xf800UL + #define XRRQ_ORRQ_NUM_SGES_SFT 11 + __le16 reserved16; + __le32 length; + __le32 psn; + #define XRRQ_ORRQ_PSN_MASK 0xffffffUL + #define XRRQ_ORRQ_PSN_SFT 0 + #define XRRQ_ORRQ_RESERVED8_1_MASK 0xff000000UL + #define XRRQ_ORRQ_RESERVED8_1_SFT 24 + __le32 end_psn; + #define XRRQ_ORRQ_END_PSN_MASK 0xffffffUL + #define XRRQ_ORRQ_END_PSN_SFT 0 + #define XRRQ_ORRQ_RESERVED8_2_MASK 0xff000000UL + #define XRRQ_ORRQ_RESERVED8_2_SFT 24 + __le64 first_sge_phy_or_sing_sge_va; + __le32 single_sge_l_key; + __le32 single_sge_size; +}; + +/* Page Buffer List Memory Structures (PBL) */ +/* Page Table Entry (PTE) (8 bytes) */ +struct ptu_pte { + __le32 page_next_to_last_last_valid[2]; + #define PTU_PTE_VALID 0x1UL + #define PTU_PTE_LAST 0x2UL + #define PTU_PTE_NEXT_TO_LAST 0x4UL + #define PTU_PTE_PAGE_MASK 0xfffff000UL + #define PTU_PTE_PAGE_SFT 12 +}; + +/* Page Directory Entry (PDE) (8 bytes) */ +struct ptu_pde { + __le32 page_valid[2]; + #define PTU_PDE_VALID 0x1UL + #define PTU_PDE_PAGE_MASK 0xfffff000UL + #define PTU_PDE_PAGE_SFT 12 +}; + +/* RoCE Fastpath Host Structures */ +/* Command Queue (CMDQ) Interface */ +/* Init CMDQ (16 bytes) */ +struct cmdq_init { + __le64 cmdq_pbl; + __le16 cmdq_size_cmdq_lvl; + #define CMDQ_INIT_CMDQ_LVL_MASK 0x3UL + #define CMDQ_INIT_CMDQ_LVL_SFT 0 + #define CMDQ_INIT_CMDQ_SIZE_MASK 0xfffcUL + #define CMDQ_INIT_CMDQ_SIZE_SFT 2 + __le16 creq_ring_id; + __le32 prod_idx; +}; + +/* Update CMDQ producer index (16 bytes) */ +struct cmdq_update { + __le64 reserved64; + __le32 reserved32; + __le32 prod_idx; +}; + +/* CMDQ common header structure (16 bytes) */ +struct cmdq_base { + u8 opcode; + #define CMDQ_BASE_OPCODE_CREATE_QP 0x1UL + #define CMDQ_BASE_OPCODE_DESTROY_QP 0x2UL + #define CMDQ_BASE_OPCODE_MODIFY_QP 0x3UL + #define CMDQ_BASE_OPCODE_QUERY_QP 0x4UL + #define CMDQ_BASE_OPCODE_CREATE_SRQ 0x5UL + #define CMDQ_BASE_OPCODE_DESTROY_SRQ 0x6UL + #define CMDQ_BASE_OPCODE_QUERY_SRQ 0x8UL + #define CMDQ_BASE_OPCODE_CREATE_CQ 0x9UL + #define CMDQ_BASE_OPCODE_DESTROY_CQ 0xaUL + #define CMDQ_BASE_OPCODE_RESIZE_CQ 0xcUL + #define CMDQ_BASE_OPCODE_ALLOCATE_MRW 0xdUL + #define CMDQ_BASE_OPCODE_DEALLOCATE_KEY 0xeUL + #define CMDQ_BASE_OPCODE_REGISTER_MR 0xfUL + #define CMDQ_BASE_OPCODE_DEREGISTER_MR 0x10UL + #define CMDQ_BASE_OPCODE_ADD_GID 0x11UL + #define CMDQ_BASE_OPCODE_DELETE_GID 0x12UL + #define CMDQ_BASE_OPCODE_MODIFY_GID 0x17UL + #define CMDQ_BASE_OPCODE_QUERY_GID 0x18UL + #define CMDQ_BASE_OPCODE_CREATE_QP1 0x13UL + #define CMDQ_BASE_OPCODE_DESTROY_QP1 0x14UL + #define CMDQ_BASE_OPCODE_CREATE_AH 0x15UL + #define CMDQ_BASE_OPCODE_DESTROY_AH 0x16UL + #define CMDQ_BASE_OPCODE_INITIALIZE_FW 0x80UL + #define CMDQ_BASE_OPCODE_DEINITIALIZE_FW 0x81UL + #define CMDQ_BASE_OPCODE_STOP_FUNC 0x82UL + #define CMDQ_BASE_OPCODE_QUERY_FUNC 0x83UL + #define CMDQ_BASE_OPCODE_SET_FUNC_RESOURCES 0x84UL + #define CMDQ_BASE_OPCODE_READ_CONTEXT 0x85UL + #define CMDQ_BASE_OPCODE_VF_BACKCHANNEL_REQUEST 0x86UL + #define CMDQ_BASE_OPCODE_READ_VF_MEMORY 0x87UL + #define CMDQ_BASE_OPCODE_COMPLETE_VF_REQUEST 0x88UL + #define CMDQ_BASE_OPCODE_EXTEND_CONTEXT_ARRRAY 0x89UL + #define CMDQ_BASE_OPCODE_MAP_TC_TO_COS 0x8aUL + #define CMDQ_BASE_OPCODE_QUERY_VERSION 0x8bUL + #define CMDQ_BASE_OPCODE_MODIFY_CC 0x8cUL + #define CMDQ_BASE_OPCODE_QUERY_CC 0x8dUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Create QP command (96 bytes) */ +struct cmdq_create_qp { + u8 opcode; + #define CMDQ_CREATE_QP_OPCODE_CREATE_QP 0x1UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 qp_handle; + __le32 qp_flags; + #define CMDQ_CREATE_QP_QP_FLAGS_SRQ_USED 0x1UL + #define CMDQ_CREATE_QP_QP_FLAGS_FORCE_COMPLETION 0x2UL + #define CMDQ_CREATE_QP_QP_FLAGS_RESERVED_LKEY_ENABLE 0x4UL + #define CMDQ_CREATE_QP_QP_FLAGS_FR_PMR_ENABLED 0x8UL + u8 type; + #define CMDQ_CREATE_QP_TYPE_RC 0x2UL + #define CMDQ_CREATE_QP_TYPE_UD 0x4UL + #define CMDQ_CREATE_QP_TYPE_RAW_ETHERTYPE 0x6UL + u8 sq_pg_size_sq_lvl; + #define CMDQ_CREATE_QP_SQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP_SQ_LVL_SFT 0 + #define CMDQ_CREATE_QP_SQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP_SQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP_SQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP_SQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP_SQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 rq_pg_size_rq_lvl; + #define CMDQ_CREATE_QP_RQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP_RQ_LVL_SFT 0 + #define CMDQ_CREATE_QP_RQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP_RQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP_RQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP_RQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP_RQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 unused_0; + __le32 dpi; + __le32 sq_size; + __le32 rq_size; + __le16 sq_fwo_sq_sge; + #define CMDQ_CREATE_QP_SQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP_SQ_SGE_SFT 0 + #define CMDQ_CREATE_QP_SQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP_SQ_FWO_SFT 4 + __le16 rq_fwo_rq_sge; + #define CMDQ_CREATE_QP_RQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP_RQ_SGE_SFT 0 + #define CMDQ_CREATE_QP_RQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP_RQ_FWO_SFT 4 + __le32 scq_cid; + __le32 rcq_cid; + __le32 srq_cid; + __le32 pd_id; + __le64 sq_pbl; + __le64 rq_pbl; + __le64 irrq_addr; + __le64 orrq_addr; +}; + +/* Destroy QP command (24 bytes) */ +struct cmdq_destroy_qp { + u8 opcode; + #define CMDQ_DESTROY_QP_OPCODE_DESTROY_QP 0x2UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp_cid; + __le32 unused_0; +}; + +/* Modify QP command (112 bytes) */ +struct cmdq_modify_qp { + u8 opcode; + #define CMDQ_MODIFY_QP_OPCODE_MODIFY_QP 0x3UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 modify_mask; + #define CMDQ_MODIFY_QP_MODIFY_MASK_STATE 0x1UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_EN_SQD_ASYNC_NOTIFY 0x2UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS 0x4UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_PKEY 0x8UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_QKEY 0x10UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DGID 0x20UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL 0x40UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX 0x80UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT 0x100UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS 0x200UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC 0x400UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU 0x1000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT 0x2000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT 0x4000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY 0x8000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN 0x10000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC 0x20000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER 0x40000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN 0x80000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC 0x100000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SIZE 0x200000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SIZE 0x400000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SGE 0x800000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SGE 0x1000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_INLINE_DATA 0x2000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID 0x4000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SRC_MAC 0x8000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID 0x10000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_ENABLE_CC 0x20000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TOS_ECN 0x40000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TOS_DSCP 0x80000000UL + __le32 qp_cid; + u8 network_type_en_sqd_async_notify_new_state; + #define CMDQ_MODIFY_QP_NEW_STATE_MASK 0xfUL + #define CMDQ_MODIFY_QP_NEW_STATE_SFT 0 + #define CMDQ_MODIFY_QP_NEW_STATE_RESET 0x0UL + #define CMDQ_MODIFY_QP_NEW_STATE_INIT 0x1UL + #define CMDQ_MODIFY_QP_NEW_STATE_RTR 0x2UL + #define CMDQ_MODIFY_QP_NEW_STATE_RTS 0x3UL + #define CMDQ_MODIFY_QP_NEW_STATE_SQD 0x4UL + #define CMDQ_MODIFY_QP_NEW_STATE_SQE 0x5UL + #define CMDQ_MODIFY_QP_NEW_STATE_ERR 0x6UL + #define CMDQ_MODIFY_QP_EN_SQD_ASYNC_NOTIFY 0x10UL + #define CMDQ_MODIFY_QP_NETWORK_TYPE_MASK 0xc0UL + #define CMDQ_MODIFY_QP_NETWORK_TYPE_SFT 6 + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV1 (0x0UL << 6) + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV4 (0x2UL << 6) + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV6 (0x3UL << 6) + u8 access; + #define CMDQ_MODIFY_QP_ACCESS_LOCAL_WRITE 0x1UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_WRITE 0x2UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_READ 0x4UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_ATOMIC 0x8UL + __le16 pkey; + __le32 qkey; + __le32 dgid[4]; + __le32 flow_label; + __le16 sgid_index; + u8 hop_limit; + u8 traffic_class; + __le16 dest_mac[3]; + u8 tos_dscp_tos_ecn; + #define CMDQ_MODIFY_QP_TOS_ECN_MASK 0x3UL + #define CMDQ_MODIFY_QP_TOS_ECN_SFT 0 + #define CMDQ_MODIFY_QP_TOS_DSCP_MASK 0xfcUL + #define CMDQ_MODIFY_QP_TOS_DSCP_SFT 2 + u8 path_mtu; + #define CMDQ_MODIFY_QP_PATH_MTU_MASK 0xf0UL + #define CMDQ_MODIFY_QP_PATH_MTU_SFT 4 + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_256 (0x0UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_512 (0x1UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_1024 (0x2UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_2048 (0x3UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_4096 (0x4UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_8192 (0x5UL << 4) + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u8 min_rnr_timer; + __le32 rq_psn; + __le32 sq_psn; + u8 max_rd_atomic; + u8 max_dest_rd_atomic; + __le16 enable_cc; + #define CMDQ_MODIFY_QP_ENABLE_CC 0x1UL + __le32 sq_size; + __le32 rq_size; + __le16 sq_sge; + __le16 rq_sge; + __le32 max_inline_data; + __le32 dest_qp_id; + __le32 unused_3; + __le16 src_mac[3]; + __le16 vlan_pcp_vlan_dei_vlan_id; + #define CMDQ_MODIFY_QP_VLAN_ID_MASK 0xfffUL + #define CMDQ_MODIFY_QP_VLAN_ID_SFT 0 + #define CMDQ_MODIFY_QP_VLAN_DEI 0x1000UL + #define CMDQ_MODIFY_QP_VLAN_PCP_MASK 0xe000UL + #define CMDQ_MODIFY_QP_VLAN_PCP_SFT 13 +}; + +/* Query QP command (24 bytes) */ +struct cmdq_query_qp { + u8 opcode; + #define CMDQ_QUERY_QP_OPCODE_QUERY_QP 0x4UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp_cid; + __le32 unused_0; +}; + +/* Create SRQ command (48 bytes) */ +struct cmdq_create_srq { + u8 opcode; + #define CMDQ_CREATE_SRQ_OPCODE_CREATE_SRQ 0x5UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 srq_handle; + __le16 pg_size_lvl; + #define CMDQ_CREATE_SRQ_LVL_MASK 0x3UL + #define CMDQ_CREATE_SRQ_LVL_SFT 0 + #define CMDQ_CREATE_SRQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_SRQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_SRQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_SRQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_CREATE_SRQ_PG_SIZE_SFT 2 + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_1G (0x5UL << 2) + __le16 eventq_id; + #define CMDQ_CREATE_SRQ_EVENTQ_ID_MASK 0xfffUL + #define CMDQ_CREATE_SRQ_EVENTQ_ID_SFT 0 + __le16 srq_size; + __le16 srq_fwo; + __le32 dpi; + __le32 pd_id; + __le64 pbl; +}; + +/* Destroy SRQ command (24 bytes) */ +struct cmdq_destroy_srq { + u8 opcode; + #define CMDQ_DESTROY_SRQ_OPCODE_DESTROY_SRQ 0x6UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 srq_cid; + __le32 unused_0; +}; + +/* Query SRQ command (24 bytes) */ +struct cmdq_query_srq { + u8 opcode; + #define CMDQ_QUERY_SRQ_OPCODE_QUERY_SRQ 0x8UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 srq_cid; + __le32 unused_0; +}; + +/* Create CQ command (48 bytes) */ +struct cmdq_create_cq { + u8 opcode; + #define CMDQ_CREATE_CQ_OPCODE_CREATE_CQ 0x9UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 cq_handle; + __le32 pg_size_lvl; + #define CMDQ_CREATE_CQ_LVL_MASK 0x3UL + #define CMDQ_CREATE_CQ_LVL_SFT 0 + #define CMDQ_CREATE_CQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_CQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_CQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_CQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_CREATE_CQ_PG_SIZE_SFT 2 + #define CMDQ_CREATE_CQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_1G (0x5UL << 2) + __le32 cq_fco_cnq_id; + #define CMDQ_CREATE_CQ_CNQ_ID_MASK 0xfffUL + #define CMDQ_CREATE_CQ_CNQ_ID_SFT 0 + #define CMDQ_CREATE_CQ_CQ_FCO_MASK 0xfffff000UL + #define CMDQ_CREATE_CQ_CQ_FCO_SFT 12 + __le32 dpi; + __le32 cq_size; + __le64 pbl; +}; + +/* Destroy CQ command (24 bytes) */ +struct cmdq_destroy_cq { + u8 opcode; + #define CMDQ_DESTROY_CQ_OPCODE_DESTROY_CQ 0xaUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 cq_cid; + __le32 unused_0; +}; + +/* Resize CQ command (40 bytes) */ +struct cmdq_resize_cq { + u8 opcode; + #define CMDQ_RESIZE_CQ_OPCODE_RESIZE_CQ 0xcUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 cq_cid; + __le32 new_cq_size_pg_size_lvl; + #define CMDQ_RESIZE_CQ_LVL_MASK 0x3UL + #define CMDQ_RESIZE_CQ_LVL_SFT 0 + #define CMDQ_RESIZE_CQ_LVL_LVL_0 0x0UL + #define CMDQ_RESIZE_CQ_LVL_LVL_1 0x1UL + #define CMDQ_RESIZE_CQ_LVL_LVL_2 0x2UL + #define CMDQ_RESIZE_CQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_RESIZE_CQ_PG_SIZE_SFT 2 + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_1G (0x5UL << 2) + #define CMDQ_RESIZE_CQ_NEW_CQ_SIZE_MASK 0x1fffe0UL + #define CMDQ_RESIZE_CQ_NEW_CQ_SIZE_SFT 5 + __le64 new_pbl; + __le32 new_cq_fco; + __le32 unused_2; +}; + +/* Allocate MRW command (32 bytes) */ +struct cmdq_allocate_mrw { + u8 opcode; + #define CMDQ_ALLOCATE_MRW_OPCODE_ALLOCATE_MRW 0xdUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 mrw_handle; + u8 mrw_flags; + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MASK 0xfUL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_SFT 0 + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MR 0x0UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR 0x1UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1 0x2UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A 0x3UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B 0x4UL + u8 access; + #define CMDQ_ALLOCATE_MRW_ACCESS_RESERVED_MASK 0x1fUL + #define CMDQ_ALLOCATE_MRW_ACCESS_RESERVED_SFT 0 + #define CMDQ_ALLOCATE_MRW_ACCESS_CONSUMER_OWNED_KEY 0x20UL + __le16 unused_1; + __le32 pd_id; +}; + +/* De-allocate key command (24 bytes) */ +struct cmdq_deallocate_key { + u8 opcode; + #define CMDQ_DEALLOCATE_KEY_OPCODE_DEALLOCATE_KEY 0xeUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 mrw_flags; + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MASK 0xfUL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_SFT 0 + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MR 0x0UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_PMR 0x1UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE1 0x2UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE2A 0x3UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE2B 0x4UL + u8 unused_1[3]; + __le32 key; +}; + +/* Register MR command (48 bytes) */ +struct cmdq_register_mr { + u8 opcode; + #define CMDQ_REGISTER_MR_OPCODE_REGISTER_MR 0xfUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 log2_pg_size_lvl; + #define CMDQ_REGISTER_MR_LVL_MASK 0x3UL + #define CMDQ_REGISTER_MR_LVL_SFT 0 + #define CMDQ_REGISTER_MR_LVL_LVL_0 0x0UL + #define CMDQ_REGISTER_MR_LVL_LVL_1 0x1UL + #define CMDQ_REGISTER_MR_LVL_LVL_2 0x2UL + #define CMDQ_REGISTER_MR_LOG2_PG_SIZE_MASK 0x7cUL + #define CMDQ_REGISTER_MR_LOG2_PG_SIZE_SFT 2 + u8 access; + #define CMDQ_REGISTER_MR_ACCESS_LOCAL_WRITE 0x1UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_READ 0x2UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_WRITE 0x4UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_ATOMIC 0x8UL + #define CMDQ_REGISTER_MR_ACCESS_MW_BIND 0x10UL + #define CMDQ_REGISTER_MR_ACCESS_ZERO_BASED 0x20UL + __le16 unused_1; + __le32 key; + __le64 pbl; + __le64 va; + __le64 mr_size; +}; + +/* Deregister MR command (24 bytes) */ +struct cmdq_deregister_mr { + u8 opcode; + #define CMDQ_DEREGISTER_MR_OPCODE_DEREGISTER_MR 0x10UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 lkey; + __le32 unused_0; +}; + +/* Add GID command (48 bytes) */ +struct cmdq_add_gid { + u8 opcode; + #define CMDQ_ADD_GID_OPCODE_ADD_GID 0x11UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __be32 gid[4]; + __be16 src_mac[3]; + __le16 vlan; + #define CMDQ_ADD_GID_VLAN_VLAN_ID_MASK 0xfffUL + #define CMDQ_ADD_GID_VLAN_VLAN_ID_SFT 0 + #define CMDQ_ADD_GID_VLAN_TPID_MASK 0x7000UL + #define CMDQ_ADD_GID_VLAN_TPID_SFT 12 + #define CMDQ_ADD_GID_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_LAST CMDQ_ADD_GID_VLAN_TPID_TPID_CFG3 + #define CMDQ_ADD_GID_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 stats_ctx; + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_ID_MASK 0x7fffUL + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_ID_SFT 0 + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_VALID 0x8000UL + __le32 unused_0; +}; + +/* Delete GID command (24 bytes) */ +struct cmdq_delete_gid { + u8 opcode; + #define CMDQ_DELETE_GID_OPCODE_DELETE_GID 0x12UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 gid_index; + __le16 unused_0; + __le32 unused_1; +}; + +/* Modify GID command (48 bytes) */ +struct cmdq_modify_gid { + u8 opcode; + #define CMDQ_MODIFY_GID_OPCODE_MODIFY_GID 0x17UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 gid[4]; + __le16 src_mac[3]; + __le16 vlan; + #define CMDQ_MODIFY_GID_VLAN_VLAN_ID_MASK 0xfffUL + #define CMDQ_MODIFY_GID_VLAN_VLAN_ID_SFT 0 + #define CMDQ_MODIFY_GID_VLAN_TPID_MASK 0x7000UL + #define CMDQ_MODIFY_GID_VLAN_TPID_SFT 12 + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_LAST \ + CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG3 + #define CMDQ_MODIFY_GID_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 gid_index; + __le16 stats_ctx; + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_ID_MASK 0x7fffUL + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_ID_SFT 0 + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_VALID 0x8000UL + __le16 unused_0; +}; + +/* Query GID command (24 bytes) */ +struct cmdq_query_gid { + u8 opcode; + #define CMDQ_QUERY_GID_OPCODE_QUERY_GID 0x18UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 gid_index; + __le16 unused_0; + __le32 unused_1; +}; + +/* Create QP1 command (80 bytes) */ +struct cmdq_create_qp1 { + u8 opcode; + #define CMDQ_CREATE_QP1_OPCODE_CREATE_QP1 0x13UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 qp_handle; + __le32 qp_flags; + #define CMDQ_CREATE_QP1_QP_FLAGS_SRQ_USED 0x1UL + #define CMDQ_CREATE_QP1_QP_FLAGS_FORCE_COMPLETION 0x2UL + #define CMDQ_CREATE_QP1_QP_FLAGS_RESERVED_LKEY_ENABLE 0x4UL + u8 type; + #define CMDQ_CREATE_QP1_TYPE_GSI 0x1UL + u8 sq_pg_size_sq_lvl; + #define CMDQ_CREATE_QP1_SQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP1_SQ_LVL_SFT 0 + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 rq_pg_size_rq_lvl; + #define CMDQ_CREATE_QP1_RQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP1_RQ_LVL_SFT 0 + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 unused_0; + __le32 dpi; + __le32 sq_size; + __le32 rq_size; + __le16 sq_fwo_sq_sge; + #define CMDQ_CREATE_QP1_SQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP1_SQ_SGE_SFT 0 + #define CMDQ_CREATE_QP1_SQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP1_SQ_FWO_SFT 4 + __le16 rq_fwo_rq_sge; + #define CMDQ_CREATE_QP1_RQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP1_RQ_SGE_SFT 0 + #define CMDQ_CREATE_QP1_RQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP1_RQ_FWO_SFT 4 + __le32 scq_cid; + __le32 rcq_cid; + __le32 srq_cid; + __le32 pd_id; + __le64 sq_pbl; + __le64 rq_pbl; +}; + +/* Destroy QP1 command (24 bytes) */ +struct cmdq_destroy_qp1 { + u8 opcode; + #define CMDQ_DESTROY_QP1_OPCODE_DESTROY_QP1 0x14UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp1_cid; + __le32 unused_0; +}; + +/* Create AH command (64 bytes) */ +struct cmdq_create_ah { + u8 opcode; + #define CMDQ_CREATE_AH_OPCODE_CREATE_AH 0x15UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 ah_handle; + __le32 dgid[4]; + u8 type; + #define CMDQ_CREATE_AH_TYPE_V1 0x0UL + #define CMDQ_CREATE_AH_TYPE_V2IPV4 0x2UL + #define CMDQ_CREATE_AH_TYPE_V2IPV6 0x3UL + u8 hop_limit; + __le16 sgid_index; + __le32 dest_vlan_id_flow_label; + #define CMDQ_CREATE_AH_FLOW_LABEL_MASK 0xfffffUL + #define CMDQ_CREATE_AH_FLOW_LABEL_SFT 0 + #define CMDQ_CREATE_AH_DEST_VLAN_ID_MASK 0xfff00000UL + #define CMDQ_CREATE_AH_DEST_VLAN_ID_SFT 20 + __le32 pd_id; + __le32 unused_0; + __le16 dest_mac[3]; + u8 traffic_class; + u8 unused_1; +}; + +/* Destroy AH command (24 bytes) */ +struct cmdq_destroy_ah { + u8 opcode; + #define CMDQ_DESTROY_AH_OPCODE_DESTROY_AH 0x16UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 ah_cid; + __le32 unused_0; +}; + +/* Initialize Firmware command (112 bytes) */ +struct cmdq_initialize_fw { + u8 opcode; + #define CMDQ_INITIALIZE_FW_OPCODE_INITIALIZE_FW 0x80UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 qpc_pg_size_qpc_lvl; + #define CMDQ_INITIALIZE_FW_QPC_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_QPC_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_1G (0x5UL << 4) + u8 mrw_pg_size_mrw_lvl; + #define CMDQ_INITIALIZE_FW_MRW_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_MRW_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_1G (0x5UL << 4) + u8 srq_pg_size_srq_lvl; + #define CMDQ_INITIALIZE_FW_SRQ_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 cq_pg_size_cq_lvl; + #define CMDQ_INITIALIZE_FW_CQ_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_CQ_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 tqm_pg_size_tqm_lvl; + #define CMDQ_INITIALIZE_FW_TQM_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_TQM_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_1G (0x5UL << 4) + u8 tim_pg_size_tim_lvl; + #define CMDQ_INITIALIZE_FW_TIM_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_TIM_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_1G (0x5UL << 4) + __le16 reserved16; + __le64 qpc_page_dir; + __le64 mrw_page_dir; + __le64 srq_page_dir; + __le64 cq_page_dir; + __le64 tqm_page_dir; + __le64 tim_page_dir; + __le32 number_of_qp; + __le32 number_of_mrw; + __le32 number_of_srq; + __le32 number_of_cq; + __le32 max_qp_per_vf; + __le32 max_mrw_per_vf; + __le32 max_srq_per_vf; + __le32 max_cq_per_vf; + __le32 max_gid_per_vf; + __le32 stat_ctx_id; +}; + +/* De-initialize Firmware command (16 bytes) */ +struct cmdq_deinitialize_fw { + u8 opcode; + #define CMDQ_DEINITIALIZE_FW_OPCODE_DEINITIALIZE_FW 0x81UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Stop function command (16 bytes) */ +struct cmdq_stop_func { + u8 opcode; + #define CMDQ_STOP_FUNC_OPCODE_STOP_FUNC 0x82UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Query function command (16 bytes) */ +struct cmdq_query_func { + u8 opcode; + #define CMDQ_QUERY_FUNC_OPCODE_QUERY_FUNC 0x83UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Set function resources command (16 bytes) */ +struct cmdq_set_func_resources { + u8 opcode; + #define CMDQ_SET_FUNC_RESOURCES_OPCODE_SET_FUNC_RESOURCES 0x84UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Read hardware resource context command (24 bytes) */ +struct cmdq_read_context { + u8 opcode; + #define CMDQ_READ_CONTEXT_OPCODE_READ_CONTEXT 0x85UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 type_xid; + #define CMDQ_READ_CONTEXT_XID_MASK 0xffffffUL + #define CMDQ_READ_CONTEXT_XID_SFT 0 + #define CMDQ_READ_CONTEXT_TYPE_MASK 0xff000000UL + #define CMDQ_READ_CONTEXT_TYPE_SFT 24 + #define CMDQ_READ_CONTEXT_TYPE_QPC (0x0UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_CQ (0x1UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_MRW (0x2UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_SRQ (0x3UL << 24) + __le32 unused_0; +}; + +/* Map TC to COS. Can only be issued from a PF (24 bytes) */ +struct cmdq_map_tc_to_cos { + u8 opcode; + #define CMDQ_MAP_TC_TO_COS_OPCODE_MAP_TC_TO_COS 0x8aUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 cos0; + #define CMDQ_MAP_TC_TO_COS_COS0_NO_CHANGE 0xffffUL + __le16 cos1; + #define CMDQ_MAP_TC_TO_COS_COS1_DISABLE 0x8000UL + #define CMDQ_MAP_TC_TO_COS_COS1_NO_CHANGE 0xffffUL + __le32 unused_0; +}; + +/* Query version command (16 bytes) */ +struct cmdq_query_version { + u8 opcode; + #define CMDQ_QUERY_VERSION_OPCODE_QUERY_VERSION 0x8bUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Command-Response Event Queue (CREQ) Structures */ +/* Base CREQ Record (16 bytes) */ +struct creq_base { + u8 type; + #define CREQ_BASE_TYPE_MASK 0x3fUL + #define CREQ_BASE_TYPE_SFT 0 + #define CREQ_BASE_TYPE_QP_EVENT 0x38UL + #define CREQ_BASE_TYPE_FUNC_EVENT 0x3aUL + #define CREQ_BASE_RESERVED2_MASK 0xc0UL + #define CREQ_BASE_RESERVED2_SFT 6 + u8 reserved56[7]; + u8 v; + #define CREQ_BASE_V 0x1UL + #define CREQ_BASE_RESERVED7_MASK 0xfeUL + #define CREQ_BASE_RESERVED7_SFT 1 + u8 event; + __le16 reserved48[3]; +}; + +/* RoCE Function Async Event Notification (16 bytes) */ +struct creq_func_event { + u8 type; + #define CREQ_FUNC_EVENT_TYPE_MASK 0x3fUL + #define CREQ_FUNC_EVENT_TYPE_SFT 0 + #define CREQ_FUNC_EVENT_TYPE_FUNC_EVENT 0x3aUL + #define CREQ_FUNC_EVENT_RESERVED2_MASK 0xc0UL + #define CREQ_FUNC_EVENT_RESERVED2_SFT 6 + u8 reserved56[7]; + u8 v; + #define CREQ_FUNC_EVENT_V 0x1UL + #define CREQ_FUNC_EVENT_RESERVED7_MASK 0xfeUL + #define CREQ_FUNC_EVENT_RESERVED7_SFT 1 + u8 event; + #define CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR 0x1UL + #define CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR 0x2UL + #define CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR 0x3UL + #define CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR 0x4UL + #define CREQ_FUNC_EVENT_EVENT_CQ_ERROR 0x5UL + #define CREQ_FUNC_EVENT_EVENT_TQM_ERROR 0x6UL + #define CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR 0x7UL + #define CREQ_FUNC_EVENT_EVENT_CFCS_ERROR 0x8UL + #define CREQ_FUNC_EVENT_EVENT_CFCC_ERROR 0x9UL + #define CREQ_FUNC_EVENT_EVENT_CFCM_ERROR 0xaUL + #define CREQ_FUNC_EVENT_EVENT_TIM_ERROR 0xbUL + #define CREQ_FUNC_EVENT_EVENT_VF_COMM_REQUEST 0x80UL + #define CREQ_FUNC_EVENT_EVENT_RESOURCE_EXHAUSTED 0x81UL + __le16 reserved48[3]; +}; + +/* RoCE Slowpath Command Completion (16 bytes) */ +struct creq_qp_event { + u8 type; + #define CREQ_QP_EVENT_TYPE_MASK 0x3fUL + #define CREQ_QP_EVENT_TYPE_SFT 0 + #define CREQ_QP_EVENT_TYPE_QP_EVENT 0x38UL + #define CREQ_QP_EVENT_RESERVED2_MASK 0xc0UL + #define CREQ_QP_EVENT_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_QP_EVENT_V 0x1UL + #define CREQ_QP_EVENT_RESERVED7_MASK 0xfeUL + #define CREQ_QP_EVENT_RESERVED7_SFT 1 + u8 event; + #define CREQ_QP_EVENT_EVENT_CREATE_QP 0x1UL + #define CREQ_QP_EVENT_EVENT_DESTROY_QP 0x2UL + #define CREQ_QP_EVENT_EVENT_MODIFY_QP 0x3UL + #define CREQ_QP_EVENT_EVENT_QUERY_QP 0x4UL + #define CREQ_QP_EVENT_EVENT_CREATE_SRQ 0x5UL + #define CREQ_QP_EVENT_EVENT_DESTROY_SRQ 0x6UL + #define CREQ_QP_EVENT_EVENT_QUERY_SRQ 0x8UL + #define CREQ_QP_EVENT_EVENT_CREATE_CQ 0x9UL + #define CREQ_QP_EVENT_EVENT_DESTROY_CQ 0xaUL + #define CREQ_QP_EVENT_EVENT_RESIZE_CQ 0xcUL + #define CREQ_QP_EVENT_EVENT_ALLOCATE_MRW 0xdUL + #define CREQ_QP_EVENT_EVENT_DEALLOCATE_KEY 0xeUL + #define CREQ_QP_EVENT_EVENT_REGISTER_MR 0xfUL + #define CREQ_QP_EVENT_EVENT_DEREGISTER_MR 0x10UL + #define CREQ_QP_EVENT_EVENT_ADD_GID 0x11UL + #define CREQ_QP_EVENT_EVENT_DELETE_GID 0x12UL + #define CREQ_QP_EVENT_EVENT_MODIFY_GID 0x17UL + #define CREQ_QP_EVENT_EVENT_QUERY_GID 0x18UL + #define CREQ_QP_EVENT_EVENT_CREATE_QP1 0x13UL + #define CREQ_QP_EVENT_EVENT_DESTROY_QP1 0x14UL + #define CREQ_QP_EVENT_EVENT_CREATE_AH 0x15UL + #define CREQ_QP_EVENT_EVENT_DESTROY_AH 0x16UL + #define CREQ_QP_EVENT_EVENT_INITIALIZE_FW 0x80UL + #define CREQ_QP_EVENT_EVENT_DEINITIALIZE_FW 0x81UL + #define CREQ_QP_EVENT_EVENT_STOP_FUNC 0x82UL + #define CREQ_QP_EVENT_EVENT_QUERY_FUNC 0x83UL + #define CREQ_QP_EVENT_EVENT_SET_FUNC_RESOURCES 0x84UL + #define CREQ_QP_EVENT_EVENT_MAP_TC_TO_COS 0x8aUL + #define CREQ_QP_EVENT_EVENT_QUERY_VERSION 0x8bUL + #define CREQ_QP_EVENT_EVENT_MODIFY_CC 0x8cUL + #define CREQ_QP_EVENT_EVENT_QUERY_CC 0x8dUL + #define CREQ_QP_EVENT_EVENT_QP_ERROR_NOTIFICATION 0xc0UL + __le16 reserved48[3]; +}; + +/* Create QP command response (16 bytes) */ +struct creq_create_qp_resp { + u8 type; + #define CREQ_CREATE_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_QP_RESP_TYPE_SFT 0 + #define CREQ_CREATE_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_QP_RESP_V 0x1UL + #define CREQ_CREATE_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_QP_RESP_EVENT_CREATE_QP 0x1UL + __le16 reserved48[3]; +}; + +/* Destroy QP command response (16 bytes) */ +struct creq_destroy_qp_resp { + u8 type; + #define CREQ_DESTROY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_QP_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_QP_RESP_V 0x1UL + #define CREQ_DESTROY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_QP_RESP_EVENT_DESTROY_QP 0x2UL + __le16 reserved48[3]; +}; + +/* Modify QP command response (16 bytes) */ +struct creq_modify_qp_resp { + u8 type; + #define CREQ_MODIFY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_QP_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_MODIFY_QP_RESP_V 0x1UL + #define CREQ_MODIFY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_QP_RESP_EVENT_MODIFY_QP 0x3UL + __le16 reserved48[3]; +}; + +/* Query QP command response (16 bytes) */ +struct creq_query_qp_resp { + u8 type; + #define CREQ_QUERY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_QP_RESP_TYPE_SFT 0 + #define CREQ_QUERY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_QP_RESP_V 0x1UL + #define CREQ_QUERY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_QP_RESP_EVENT_QUERY_QP 0x4UL + __le16 reserved48[3]; +}; + +/* Query QP command response side buffer structure (104 bytes) */ +struct creq_query_qp_resp_sb { + u8 opcode; + #define CREQ_QUERY_QP_RESP_SB_OPCODE_QUERY_QP 0x4UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 xid; + u8 en_sqd_async_notify_state; + #define CREQ_QUERY_QP_RESP_SB_STATE_MASK 0xfUL + #define CREQ_QUERY_QP_RESP_SB_STATE_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_STATE_RESET 0x0UL + #define CREQ_QUERY_QP_RESP_SB_STATE_INIT 0x1UL + #define CREQ_QUERY_QP_RESP_SB_STATE_RTR 0x2UL + #define CREQ_QUERY_QP_RESP_SB_STATE_RTS 0x3UL + #define CREQ_QUERY_QP_RESP_SB_STATE_SQD 0x4UL + #define CREQ_QUERY_QP_RESP_SB_STATE_SQE 0x5UL + #define CREQ_QUERY_QP_RESP_SB_STATE_ERR 0x6UL + #define CREQ_QUERY_QP_RESP_SB_EN_SQD_ASYNC_NOTIFY 0x10UL + u8 access; + #define CREQ_QUERY_QP_RESP_SB_ACCESS_LOCAL_WRITE 0x1UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_WRITE 0x2UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_READ 0x4UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_ATOMIC 0x8UL + __le16 pkey; + __le32 qkey; + __le32 reserved32; + __le32 dgid[4]; + __le32 flow_label; + __le16 sgid_index; + u8 hop_limit; + u8 traffic_class; + __le16 dest_mac[3]; + __le16 path_mtu_dest_vlan_id; + #define CREQ_QUERY_QP_RESP_SB_DEST_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_QP_RESP_SB_DEST_VLAN_ID_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK 0xf000UL + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_SFT 12 + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_256 (0x0UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_512 (0x1UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_1024 (0x2UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_2048 (0x3UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_4096 (0x4UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_8192 (0x5UL << 12) + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u8 min_rnr_timer; + __le32 rq_psn; + __le32 sq_psn; + u8 max_rd_atomic; + u8 max_dest_rd_atomic; + u8 tos_dscp_tos_ecn; + #define CREQ_QUERY_QP_RESP_SB_TOS_ECN_MASK 0x3UL + #define CREQ_QUERY_QP_RESP_SB_TOS_ECN_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_TOS_DSCP_MASK 0xfcUL + #define CREQ_QUERY_QP_RESP_SB_TOS_DSCP_SFT 2 + u8 enable_cc; + #define CREQ_QUERY_QP_RESP_SB_ENABLE_CC 0x1UL + #define CREQ_QUERY_QP_RESP_SB_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_QP_RESP_SB_RESERVED7_SFT 1 + __le32 sq_size; + __le32 rq_size; + __le16 sq_sge; + __le16 rq_sge; + __le32 max_inline_data; + __le32 dest_qp_id; + __le32 unused_1; + __le16 src_mac[3]; + __le16 vlan_pcp_vlan_dei_vlan_id; + #define CREQ_QUERY_QP_RESP_SB_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_QP_RESP_SB_VLAN_ID_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_VLAN_DEI 0x1000UL + #define CREQ_QUERY_QP_RESP_SB_VLAN_PCP_MASK 0xe000UL + #define CREQ_QUERY_QP_RESP_SB_VLAN_PCP_SFT 13 +}; + +/* Create SRQ command response (16 bytes) */ +struct creq_create_srq_resp { + u8 type; + #define CREQ_CREATE_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_SRQ_RESP_TYPE_SFT 0 + #define CREQ_CREATE_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_SRQ_RESP_V 0x1UL + #define CREQ_CREATE_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_SRQ_RESP_EVENT_CREATE_SRQ 0x5UL + __le16 reserved48[3]; +}; + +/* Destroy SRQ command response (16 bytes) */ +struct creq_destroy_srq_resp { + u8 type; + #define CREQ_DESTROY_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_SRQ_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_SRQ_RESP_V 0x1UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_SRQ_RESP_EVENT_DESTROY_SRQ 0x6UL + __le16 enable_for_arm[3]; + #define CREQ_DESTROY_SRQ_RESP_ENABLE_FOR_ARM_MASK 0x30000UL + #define CREQ_DESTROY_SRQ_RESP_ENABLE_FOR_ARM_SFT 16 + #define CREQ_DESTROY_SRQ_RESP_RESERVED46_MASK 0xfffc0000UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED46_SFT 18 +}; + +/* Query SRQ command response (16 bytes) */ +struct creq_query_srq_resp { + u8 type; + #define CREQ_QUERY_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_SRQ_RESP_TYPE_SFT 0 + #define CREQ_QUERY_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_SRQ_RESP_V 0x1UL + #define CREQ_QUERY_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_SRQ_RESP_EVENT_QUERY_SRQ 0x8UL + __le16 reserved48[3]; +}; + +/* Query SRQ command response side buffer structure (24 bytes) */ +struct creq_query_srq_resp_sb { + u8 opcode; + #define CREQ_QUERY_SRQ_RESP_SB_OPCODE_QUERY_SRQ 0x8UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 xid; + __le16 srq_limit; + __le16 reserved16; + __le32 data[4]; +}; + +/* Create CQ command Response (16 bytes) */ +struct creq_create_cq_resp { + u8 type; + #define CREQ_CREATE_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_CQ_RESP_TYPE_SFT 0 + #define CREQ_CREATE_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_CQ_RESP_V 0x1UL + #define CREQ_CREATE_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_CQ_RESP_EVENT_CREATE_CQ 0x9UL + __le16 reserved48[3]; +}; + +/* Destroy CQ command response (16 bytes) */ +struct creq_destroy_cq_resp { + u8 type; + #define CREQ_DESTROY_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_CQ_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_CQ_RESP_V 0x1UL + #define CREQ_DESTROY_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_CQ_RESP_EVENT_DESTROY_CQ 0xaUL + __le16 cq_arm_lvl; + #define CREQ_DESTROY_CQ_RESP_CQ_ARM_LVL_MASK 0x3UL + #define CREQ_DESTROY_CQ_RESP_CQ_ARM_LVL_SFT 0 + #define CREQ_DESTROY_CQ_RESP_RESERVED14_MASK 0xfffcUL + #define CREQ_DESTROY_CQ_RESP_RESERVED14_SFT 2 + __le16 total_cnq_events; + __le16 reserved16; +}; + +/* Resize CQ command response (16 bytes) */ +struct creq_resize_cq_resp { + u8 type; + #define CREQ_RESIZE_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_RESIZE_CQ_RESP_TYPE_SFT 0 + #define CREQ_RESIZE_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_RESIZE_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_RESIZE_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_RESIZE_CQ_RESP_V 0x1UL + #define CREQ_RESIZE_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_RESIZE_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_RESIZE_CQ_RESP_EVENT_RESIZE_CQ 0xcUL + __le16 reserved48[3]; +}; + +/* Allocate MRW command response (16 bytes) */ +struct creq_allocate_mrw_resp { + u8 type; + #define CREQ_ALLOCATE_MRW_RESP_TYPE_MASK 0x3fUL + #define CREQ_ALLOCATE_MRW_RESP_TYPE_SFT 0 + #define CREQ_ALLOCATE_MRW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_ALLOCATE_MRW_RESP_V 0x1UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_ALLOCATE_MRW_RESP_EVENT_ALLOCATE_MRW 0xdUL + __le16 reserved48[3]; +}; + +/* De-allocate key command response (16 bytes) */ +struct creq_deallocate_key_resp { + u8 type; + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_SFT 0 + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DEALLOCATE_KEY_RESP_V 0x1UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEALLOCATE_KEY_RESP_EVENT_DEALLOCATE_KEY 0xeUL + __le16 reserved16; + __le32 bound_window_info; +}; + +/* Register MR command response (16 bytes) */ +struct creq_register_mr_resp { + u8 type; + #define CREQ_REGISTER_MR_RESP_TYPE_MASK 0x3fUL + #define CREQ_REGISTER_MR_RESP_TYPE_SFT 0 + #define CREQ_REGISTER_MR_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_REGISTER_MR_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_REGISTER_MR_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_REGISTER_MR_RESP_V 0x1UL + #define CREQ_REGISTER_MR_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_REGISTER_MR_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_REGISTER_MR_RESP_EVENT_REGISTER_MR 0xfUL + __le16 reserved48[3]; +}; + +/* Deregister MR command response (16 bytes) */ +struct creq_deregister_mr_resp { + u8 type; + #define CREQ_DEREGISTER_MR_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEREGISTER_MR_RESP_TYPE_SFT 0 + #define CREQ_DEREGISTER_MR_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DEREGISTER_MR_RESP_V 0x1UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEREGISTER_MR_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEREGISTER_MR_RESP_EVENT_DEREGISTER_MR 0x10UL + __le16 reserved16; + __le32 bound_windows; +}; + +/* Add GID command response (16 bytes) */ +struct creq_add_gid_resp { + u8 type; + #define CREQ_ADD_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_ADD_GID_RESP_TYPE_SFT 0 + #define CREQ_ADD_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_ADD_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_ADD_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_ADD_GID_RESP_V 0x1UL + #define CREQ_ADD_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_ADD_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_ADD_GID_RESP_EVENT_ADD_GID 0x11UL + __le16 reserved48[3]; +}; + +/* Delete GID command response (16 bytes) */ +struct creq_delete_gid_resp { + u8 type; + #define CREQ_DELETE_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_DELETE_GID_RESP_TYPE_SFT 0 + #define CREQ_DELETE_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DELETE_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DELETE_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DELETE_GID_RESP_V 0x1UL + #define CREQ_DELETE_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DELETE_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DELETE_GID_RESP_EVENT_DELETE_GID 0x12UL + __le16 reserved48[3]; +}; + +/* Modify GID command response (16 bytes) */ +struct creq_modify_gid_resp { + u8 type; + #define CREQ_MODIFY_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_GID_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_MODIFY_GID_RESP_V 0x1UL + #define CREQ_MODIFY_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_GID_RESP_EVENT_ADD_GID 0x11UL + __le16 reserved48[3]; +}; + +/* Query GID command response (16 bytes) */ +struct creq_query_gid_resp { + u8 type; + #define CREQ_QUERY_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_GID_RESP_TYPE_SFT 0 + #define CREQ_QUERY_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_GID_RESP_V 0x1UL + #define CREQ_QUERY_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_GID_RESP_EVENT_QUERY_GID 0x18UL + __le16 reserved48[3]; +}; + +/* Query GID command response side buffer structure (40 bytes) */ +struct creq_query_gid_resp_sb { + u8 opcode; + #define CREQ_QUERY_GID_RESP_SB_OPCODE_QUERY_GID 0x18UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 gid[4]; + __le16 src_mac[3]; + __le16 vlan; + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_ID_SFT 0 + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_MASK 0x7000UL + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_SFT 12 + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_LAST \ + CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG3 + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 gid_index; + __le32 unused_0; +}; + +/* Create QP1 command response (16 bytes) */ +struct creq_create_qp1_resp { + u8 type; + #define CREQ_CREATE_QP1_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_QP1_RESP_TYPE_SFT 0 + #define CREQ_CREATE_QP1_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_QP1_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_QP1_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_QP1_RESP_V 0x1UL + #define CREQ_CREATE_QP1_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_QP1_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_QP1_RESP_EVENT_CREATE_QP1 0x13UL + __le16 reserved48[3]; +}; + +/* Destroy QP1 command response (16 bytes) */ +struct creq_destroy_qp1_resp { + u8 type; + #define CREQ_DESTROY_QP1_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_QP1_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_QP1_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_QP1_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_QP1_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_QP1_RESP_V 0x1UL + #define CREQ_DESTROY_QP1_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_QP1_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_QP1_RESP_EVENT_DESTROY_QP1 0x14UL + __le16 reserved48[3]; +}; + +/* Create AH command response (16 bytes) */ +struct creq_create_ah_resp { + u8 type; + #define CREQ_CREATE_AH_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_AH_RESP_TYPE_SFT 0 + #define CREQ_CREATE_AH_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_AH_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_AH_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_AH_RESP_V 0x1UL + #define CREQ_CREATE_AH_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_AH_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_AH_RESP_EVENT_CREATE_AH 0x15UL + __le16 reserved48[3]; +}; + +/* Destroy AH command response (16 bytes) */ +struct creq_destroy_ah_resp { + u8 type; + #define CREQ_DESTROY_AH_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_AH_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_AH_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_AH_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_AH_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_AH_RESP_V 0x1UL + #define CREQ_DESTROY_AH_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_AH_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_AH_RESP_EVENT_DESTROY_AH 0x16UL + __le16 reserved48[3]; +}; + +/* Initialize Firmware command response (16 bytes) */ +struct creq_initialize_fw_resp { + u8 type; + #define CREQ_INITIALIZE_FW_RESP_TYPE_MASK 0x3fUL + #define CREQ_INITIALIZE_FW_RESP_TYPE_SFT 0 + #define CREQ_INITIALIZE_FW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_INITIALIZE_FW_RESP_V 0x1UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_INITIALIZE_FW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_INITIALIZE_FW_RESP_EVENT_INITIALIZE_FW 0x80UL + __le16 reserved48[3]; +}; + +/* De-initialize Firmware command response (16 bytes) */ +struct creq_deinitialize_fw_resp { + u8 type; + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_SFT 0 + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_DEINITIALIZE_FW_RESP_V 0x1UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEINITIALIZE_FW_RESP_EVENT_DEINITIALIZE_FW 0x81UL + __le16 reserved48[3]; +}; + +/* Stop function command response (16 bytes) */ +struct creq_stop_func_resp { + u8 type; + #define CREQ_STOP_FUNC_RESP_TYPE_MASK 0x3fUL + #define CREQ_STOP_FUNC_RESP_TYPE_SFT 0 + #define CREQ_STOP_FUNC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_STOP_FUNC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_STOP_FUNC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_STOP_FUNC_RESP_V 0x1UL + #define CREQ_STOP_FUNC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_STOP_FUNC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_STOP_FUNC_RESP_EVENT_STOP_FUNC 0x82UL + __le16 reserved48[3]; +}; + +/* Query function command response (16 bytes) */ +struct creq_query_func_resp { + u8 type; + #define CREQ_QUERY_FUNC_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_FUNC_RESP_TYPE_SFT 0 + #define CREQ_QUERY_FUNC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_FUNC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_FUNC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_FUNC_RESP_V 0x1UL + #define CREQ_QUERY_FUNC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_FUNC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_FUNC_RESP_EVENT_QUERY_FUNC 0x83UL + __le16 reserved48[3]; +}; + +/* Query function command response side buffer structure (88 bytes) */ +struct creq_query_func_resp_sb { + u8 opcode; + #define CREQ_QUERY_FUNC_RESP_SB_OPCODE_QUERY_FUNC 0x83UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le64 max_mr_size; + __le32 max_qp; + __le16 max_qp_wr; + __le16 dev_cap_flags; + #define CREQ_QUERY_FUNC_RESP_SB_DEV_CAP_FLAGS_RESIZE_QP 0x1UL + __le32 max_cq; + __le32 max_cqe; + __le32 max_pd; + u8 max_sge; + u8 max_srq_sge; + u8 max_qp_rd_atom; + u8 max_qp_init_rd_atom; + __le32 max_mr; + __le32 max_mw; + __le32 max_raw_eth_qp; + __le32 max_ah; + __le32 max_fmr; + __le32 max_srq_wr; + __le32 max_pkeys; + __le32 max_inline_data; + u8 max_map_per_fmr; + u8 l2_db_space_size; + __le16 max_srq; + __le32 max_gid; + __le32 tqm_alloc_reqs[8]; +}; + +/* Set resources command response (16 bytes) */ +struct creq_set_func_resources_resp { + u8 type; + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_MASK 0x3fUL + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_SFT 0 + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_SET_FUNC_RESOURCES_RESP_V 0x1UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_SET_FUNC_RESOURCES_RESP_EVENT_SET_FUNC_RESOURCES 0x84UL + __le16 reserved48[3]; +}; + +/* Map TC to COS response (16 bytes) */ +struct creq_map_tc_to_cos_resp { + u8 type; + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_MASK 0x3fUL + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_SFT 0 + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_MAP_TC_TO_COS_RESP_V 0x1UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MAP_TC_TO_COS_RESP_EVENT_MAP_TC_TO_COS 0x8aUL + __le16 reserved48[3]; +}; + +/* Query version response (16 bytes) */ +struct creq_query_version_resp { + u8 type; + #define CREQ_QUERY_VERSION_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_VERSION_RESP_TYPE_SFT 0 + #define CREQ_QUERY_VERSION_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_VERSION_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_VERSION_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + u8 fw_maj; + u8 fw_minor; + u8 fw_bld; + u8 fw_rsvd; + u8 v; + #define CREQ_QUERY_VERSION_RESP_V 0x1UL + #define CREQ_QUERY_VERSION_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_VERSION_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_VERSION_RESP_EVENT_QUERY_VERSION 0x8bUL + __le16 reserved16; + u8 intf_maj; + u8 intf_minor; + u8 intf_bld; + u8 intf_rsvd; +}; + +/* Modify congestion control command response (16 bytes) */ +struct creq_modify_cc_resp { + u8 type; + #define CREQ_MODIFY_CC_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_CC_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_CC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_CC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_CC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_MODIFY_CC_RESP_V 0x1UL + #define CREQ_MODIFY_CC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_CC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_CC_RESP_EVENT_MODIFY_CC 0x8cUL + __le16 reserved48[3]; +}; + +/* Query congestion control command response (16 bytes) */ +struct creq_query_cc_resp { + u8 type; + #define CREQ_QUERY_CC_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_CC_RESP_TYPE_SFT 0 + #define CREQ_QUERY_CC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_CC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_CC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_CC_RESP_V 0x1UL + #define CREQ_QUERY_CC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_CC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_CC_RESP_EVENT_QUERY_CC 0x8dUL + __le16 reserved48[3]; +}; + +/* Query congestion control command response side buffer structure (32 bytes) */ +struct creq_query_cc_resp_sb { + u8 opcode; + #define CREQ_QUERY_CC_RESP_SB_OPCODE_QUERY_CC 0x8dUL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + u8 enable_cc; + #define CREQ_QUERY_CC_RESP_SB_ENABLE_CC 0x1UL + u8 g; + #define CREQ_QUERY_CC_RESP_SB_G_MASK 0x7UL + #define CREQ_QUERY_CC_RESP_SB_G_SFT 0 + u8 num_phases_per_state; + __le16 init_cr; + u8 unused_2; + __le16 unused_3; + u8 unused_4; + __le16 init_tr; + u8 tos_dscp_tos_ecn; + #define CREQ_QUERY_CC_RESP_SB_TOS_ECN_MASK 0x3UL + #define CREQ_QUERY_CC_RESP_SB_TOS_ECN_SFT 0 + #define CREQ_QUERY_CC_RESP_SB_TOS_DSCP_MASK 0xfcUL + #define CREQ_QUERY_CC_RESP_SB_TOS_DSCP_SFT 2 + __le64 reserved64; + __le64 reserved64_1; +}; + +/* QP error notification event (16 bytes) */ +struct creq_qp_error_notification { + u8 type; + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_MASK 0x3fUL + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_SFT 0 + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_QP_EVENT 0x38UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED2_MASK 0xc0UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED2_SFT 6 + u8 status; + u8 req_slow_path_state; + u8 req_err_state_reason; + __le32 xid; + u8 v; + #define CREQ_QP_ERROR_NOTIFICATION_V 0x1UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED7_MASK 0xfeUL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED7_SFT 1 + u8 event; + #define CREQ_QP_ERROR_NOTIFICATION_EVENT_QP_ERROR_NOTIFICATION 0xc0UL + u8 res_slow_path_state; + u8 res_err_state_reason; + __le16 sq_cons_idx; + __le16 rq_cons_idx; +}; + +/* RoCE Slowpath HSI Specification 1.6.0 */ +#define ROCE_SP_HSI_VERSION_MAJOR 1 +#define ROCE_SP_HSI_VERSION_MINOR 6 +#define ROCE_SP_HSI_VERSION_UPDATE 0 + +#define ROCE_SP_HSI_VERSION_STR "1.6.0" +/* + * Following is the signature for ROCE_SP_HSI message field that indicates not + * applicable (All F's). Need to cast it the size of the field if needed. + */ +#define ROCE_SP_HSI_NA_SIGNATURE ((__le32)(-1)) +#endif /* __BNXT_RE_HSI_H__ */ diff --git a/include/uapi/rdma/bnxt_re-abi.h b/include/uapi/rdma/bnxt_re-abi.h new file mode 100644 index 000000000000..e2c8a3f0ccec --- /dev/null +++ b/include/uapi/rdma/bnxt_re-abi.h @@ -0,0 +1,89 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Uverbs ABI header file + */ + +#ifndef __BNXT_RE_UVERBS_ABI_H__ +#define __BNXT_RE_UVERBS_ABI_H__ + +#define BNXT_RE_ABI_VERSION 1 + +struct bnxt_re_uctx_resp { + __u32 dev_id; + __u32 max_qp; + __u32 pg_size; + __u32 cqe_sz; + __u32 max_cqd; + __u32 rsvd; +}; + +struct bnxt_re_pd_resp { + __u32 pdid; + __u32 dpi; + __u64 dbr; +}; + +struct bnxt_re_cq_req { + __u64 cq_va; + __u64 cq_handle; +}; + +struct bnxt_re_cq_resp { + __u32 cqid; + __u32 tail; + __u32 phase; + __u32 rsvd; +}; + +struct bnxt_re_qp_req { + __u64 qpsva; + __u64 qprva; + __u64 qp_handle; +}; + +struct bnxt_re_qp_resp { + __u32 qpid; + __u32 rsvd; +}; + +enum bnxt_re_shpg_offt { + BNXT_RE_BEG_RESV_OFFT = 0x00, + BNXT_RE_AVID_OFFT = 0x10, + BNXT_RE_AVID_SIZE = 0x04, + BNXT_RE_END_RESV_OFFT = 0xFF0 +}; + +#endif /* __BNXT_RE_UVERBS_ABI_H__*/ -- cgit v1.2.3 From 592e8b3226a2b6d116589908d96c45fa13302ad7 Mon Sep 17 00:00:00 2001 From: Selvin Xavier Date: Fri, 10 Feb 2017 03:19:53 -0800 Subject: RDMA/bnxt_re: Add bnxt_re driver build support Makefile and Kconfig changes for enabling bnxt_re compilation Signed-off-by: Devesh Sharma Signed-off-by: Somnath Kotur Signed-off-by: Sriharsha Basavapatna Signed-off-by: Selvin Xavier Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- MAINTAINERS | 11 +++++++++++ drivers/infiniband/Kconfig | 2 ++ drivers/infiniband/hw/Makefile | 1 + drivers/infiniband/hw/bnxt_re/Kconfig | 9 +++++++++ drivers/infiniband/hw/bnxt_re/Makefile | 6 ++++++ 5 files changed, 29 insertions(+) create mode 100644 drivers/infiniband/hw/bnxt_re/Kconfig create mode 100644 drivers/infiniband/hw/bnxt_re/Makefile diff --git a/MAINTAINERS b/MAINTAINERS index 5f0420a0da5b..468d2e8e2d4d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2808,6 +2808,17 @@ L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm64/boot/dts/broadcom/vulcan* +BROADCOM NETXTREME-E ROCE DRIVER +M: Selvin Xavier +M: Devesh Sharma +M: Somnath Kotur +M: Sriharsha Basavapatna +L: linux-rdma@vger.kernel.org +W: http://www.broadcom.com +S: Supported +F: drivers/infiniband/hw/bnxt_re/ +F: include/uapi/rdma/bnxt_re-abi.h + BROCADE BFA FC SCSI DRIVER M: Anil Gurumurthy M: Sudarsana Kalluru diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 670917387eda..66f86027ed47 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -92,4 +92,6 @@ source "drivers/infiniband/hw/hfi1/Kconfig" source "drivers/infiniband/hw/qedr/Kconfig" +source "drivers/infiniband/hw/bnxt_re/Kconfig" + endif # INFINIBAND diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile index ed553de2ca12..34c93abf0fe0 100644 --- a/drivers/infiniband/hw/Makefile +++ b/drivers/infiniband/hw/Makefile @@ -12,3 +12,4 @@ obj-$(CONFIG_INFINIBAND_USNIC) += usnic/ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ obj-$(CONFIG_INFINIBAND_HNS) += hns/ obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ +obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ diff --git a/drivers/infiniband/hw/bnxt_re/Kconfig b/drivers/infiniband/hw/bnxt_re/Kconfig new file mode 100644 index 000000000000..cd0175e32584 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/Kconfig @@ -0,0 +1,9 @@ +config INFINIBAND_BNXT_RE + tristate "Broadcom Netxtreme HCA support" + depends on ETHERNET && NETDEVICES && PCI && INET + select NET_VENDOR_BROADCOM + select BNXT + ---help--- + This driver supports Broadcom NetXtreme-E 10/25/40/50 gigabit + RoCE HCAs. To compile this driver as a module, choose M here: + the module will be called bnxt_re. diff --git a/drivers/infiniband/hw/bnxt_re/Makefile b/drivers/infiniband/hw/bnxt_re/Makefile new file mode 100644 index 000000000000..036f84efbc73 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/Makefile @@ -0,0 +1,6 @@ + +ccflags-y := -Idrivers/net/ethernet/broadcom/bnxt +obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re.o +bnxt_re-y := main.o ib_verbs.o \ + qplib_res.o qplib_rcfw.o \ + qplib_sp.o qplib_fp.o -- cgit v1.2.3 From 2b0841766a898aba84630fb723989a77a9d3b4e6 Mon Sep 17 00:00:00 2001 From: Erez Shitrit Date: Wed, 1 Feb 2017 19:10:05 +0200 Subject: IB/IPoIB: Add destination address when re-queue packet When sending packet to destination that was not resolved yet via path query, the driver keeps the skb and tries to re-send it again when the path is resolved. But when re-sending via dev_queue_xmit the kernel doesn't call to dev_hard_header, so IPoIB needs to keep 20 bytes in the skb and to put the destination address inside them. In that way the dev_start_xmit will have the correct destination, and the driver won't take the destination from the skb->data, while nothing exists there, which causes to packet be be dropped. The test flow is: 1. Run the SM on remote node, 2. Restart the driver. 4. Ping some destination, 3. Observe that first ICMP request will be dropped. Fixes: fc791b633515 ("IB/ipoib: move back IB LL address into the hard header") Cc: # v4.8+ Signed-off-by: Erez Shitrit Signed-off-by: Noa Osherovich Signed-off-by: Leon Romanovsky Tested-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 30 +++++++++++++++++------------- 1 file changed, 17 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index e3bfa8a99ad2..259c59f67394 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -721,6 +721,14 @@ int ipoib_check_sm_sendonly_fullmember_support(struct ipoib_dev_priv *priv) return ret; } +static void push_pseudo_header(struct sk_buff *skb, const char *daddr) +{ + struct ipoib_pseudo_header *phdr; + + phdr = (struct ipoib_pseudo_header *)skb_push(skb, sizeof(*phdr)); + memcpy(phdr->hwaddr, daddr, INFINIBAND_ALEN); +} + void ipoib_flush_paths(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); @@ -947,8 +955,7 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr, } if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, neigh->daddr); __skb_queue_tail(&neigh->queue, skb); } else { ipoib_warn(priv, "queue length limit %d. Packet drop.\n", @@ -966,10 +973,12 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr, if (!path->query && path_rec_start(dev, path)) goto err_path; - if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) + if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { + push_pseudo_header(skb, neigh->daddr); __skb_queue_tail(&neigh->queue, skb); - else + } else { goto err_drop; + } } spin_unlock_irqrestore(&priv->lock, flags); @@ -1005,8 +1014,7 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, } if (path) { if (skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); } else { ++dev->stats.tx_dropped; @@ -1038,8 +1046,7 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, return; } else if ((path->query || !path_rec_start(dev, path)) && skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); } else { ++dev->stats.tx_dropped; @@ -1120,8 +1127,7 @@ send_using_neigh: } if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, sizeof(*phdr)); + push_pseudo_header(skb, phdr->hwaddr); spin_lock_irqsave(&priv->lock, flags); __skb_queue_tail(&neigh->queue, skb); spin_unlock_irqrestore(&priv->lock, flags); @@ -1153,7 +1159,6 @@ static int ipoib_hard_header(struct sk_buff *skb, unsigned short type, const void *daddr, const void *saddr, unsigned len) { - struct ipoib_pseudo_header *phdr; struct ipoib_header *header; header = (struct ipoib_header *) skb_push(skb, sizeof *header); @@ -1166,8 +1171,7 @@ static int ipoib_hard_header(struct sk_buff *skb, * destination address into skb hard header so we can figure out where * to send the packet later. */ - phdr = (struct ipoib_pseudo_header *) skb_push(skb, sizeof(*phdr)); - memcpy(phdr->hwaddr, daddr, INFINIBAND_ALEN); + push_pseudo_header(skb, daddr); return IPOIB_HARD_LEN; } -- cgit v1.2.3 From 590396084039b6504097b371cda6474dfcb83648 Mon Sep 17 00:00:00 2001 From: Parav Pandit Date: Thu, 2 Feb 2017 07:14:08 +0200 Subject: IB/core: Remove pointer casting from void to net_device This patch avoids unnecessary type casting from void to net_device. CC: Yuval Shaia Signed-off-by: Parav Pandit Signed-off-by: Leon Romanovsky Reviewed-by: Yuval Shaia Signed-off-by: Doug Ledford --- drivers/infiniband/core/roce_gid_mgmt.c | 28 ++++++++++------------------ 1 file changed, 10 insertions(+), 18 deletions(-) diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c index 0621f4455732..db958d3207ef 100644 --- a/drivers/infiniband/core/roce_gid_mgmt.c +++ b/drivers/infiniband/core/roce_gid_mgmt.c @@ -144,7 +144,6 @@ static enum bonding_slave_state is_eth_active_slave_of_bonding_rcu(struct net_de static int is_eth_port_of_netdev(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; struct net_device *real_dev; int res; @@ -152,11 +151,11 @@ static int is_eth_port_of_netdev(struct ib_device *ib_dev, u8 port, return 0; rcu_read_lock(); - real_dev = rdma_vlan_dev_real_dev(event_ndev); + real_dev = rdma_vlan_dev_real_dev(cookie); if (!real_dev) - real_dev = event_ndev; + real_dev = cookie; - res = ((rdma_is_upper_dev_rcu(rdma_ndev, event_ndev) && + res = ((rdma_is_upper_dev_rcu(rdma_ndev, cookie) && (is_eth_active_slave_of_bonding_rcu(rdma_ndev, real_dev) & REQUIRED_BOND_STATES)) || real_dev == rdma_ndev); @@ -192,17 +191,16 @@ static int pass_all_filter(struct ib_device *ib_dev, u8 port, static int upper_device_filter(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; int res; if (!rdma_ndev) return 0; - if (rdma_ndev == event_ndev) + if (rdma_ndev == cookie) return 1; rcu_read_lock(); - res = rdma_is_upper_dev_rcu(rdma_ndev, event_ndev); + res = rdma_is_upper_dev_rcu(rdma_ndev, cookie); rcu_read_unlock(); return res; @@ -379,18 +377,14 @@ static void _add_netdev_ips(struct ib_device *ib_dev, u8 port, static void add_netdev_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - enum_netdev_default_gids(ib_dev, port, event_ndev, rdma_ndev); - _add_netdev_ips(ib_dev, port, event_ndev); + enum_netdev_default_gids(ib_dev, port, cookie, rdma_ndev); + _add_netdev_ips(ib_dev, port, cookie); } static void del_netdev_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - ib_cache_gid_del_all_netdev_gids(ib_dev, port, event_ndev); + ib_cache_gid_del_all_netdev_gids(ib_dev, port, cookie); } static void enum_all_gids_of_dev_cb(struct ib_device *ib_dev, @@ -460,7 +454,7 @@ static void handle_netdev_upper(struct ib_device *ib_dev, u8 port, u8 port, struct net_device *ndev)) { - struct net_device *ndev = (struct net_device *)cookie; + struct net_device *ndev = cookie; struct upper_list *upper_iter; struct upper_list *upper_temp; LIST_HEAD(upper_list); @@ -519,9 +513,7 @@ static void del_netdev_default_ips_join(struct ib_device *ib_dev, u8 port, static void del_netdev_default_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - bond_delete_netdev_default_gids(ib_dev, port, event_ndev, rdma_ndev); + bond_delete_netdev_default_gids(ib_dev, port, cookie, rdma_ndev); } /* The following functions operate on all IB devices. netdevice_event and -- cgit v1.2.3 From 89052d784bc977c2a0b92393f6bd57140952c206 Mon Sep 17 00:00:00 2001 From: Majd Dibbiny Date: Tue, 14 Feb 2017 07:21:52 +0200 Subject: IB/cma: Add default RoCE TOS to CMA configfs Add new entry to the RDMA-CM configfs that allows users to select default TOS for RDMA-CM QPs. This is useful for users that want to control the TOS for legacy applications without changing their code. Application that sets the TOS explicitly using the rdma_set_option API will continue to work as expected, meaning overriding the configfs value. CC: Dan Carpenter Signed-off-by: Majd Dibbiny Reviewed-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- Documentation/ABI/testing/configfs-rdma_cm | 8 +++++ drivers/infiniband/core/cma.c | 54 +++++++++++++++++++++++++++--- drivers/infiniband/core/cma_configfs.c | 42 +++++++++++++++++++++++ drivers/infiniband/core/core_priv.h | 3 ++ 4 files changed, 102 insertions(+), 5 deletions(-) diff --git a/Documentation/ABI/testing/configfs-rdma_cm b/Documentation/ABI/testing/configfs-rdma_cm index 5c389aaf5291..74f9506f42e7 100644 --- a/Documentation/ABI/testing/configfs-rdma_cm +++ b/Documentation/ABI/testing/configfs-rdma_cm @@ -20,3 +20,11 @@ Description: RDMA-CM based connections from HCA at port will be initiated with this RoCE type as default. The possible RoCE types are either "IB/RoCE v1" or "RoCE v2". This parameter has RW access. + +What: /config/rdma_cm//ports//default_roce_tos +Date: February 7, 2017 +KernelVersion: 4.11.0 +Description: RDMA-CM QPs from HCA at port + will be created with this TOS as default. + This can be overridden by using the rdma_set_option API. + The possible RoCE TOS values are 0-255. diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 9c93e2fa969b..f98ec19a851a 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -198,6 +198,7 @@ struct cma_device { atomic_t refcount; struct list_head id_list; enum ib_gid_type *default_gid_type; + u8 *default_roce_tos; }; struct rdma_bind_list { @@ -295,6 +296,25 @@ int cma_set_default_gid_type(struct cma_device *cma_dev, return 0; } +int cma_get_default_roce_tos(struct cma_device *cma_dev, unsigned int port) +{ + if (!rdma_is_port_valid(cma_dev->device, port)) + return -EINVAL; + + return cma_dev->default_roce_tos[port - rdma_start_port(cma_dev->device)]; +} + +int cma_set_default_roce_tos(struct cma_device *cma_dev, unsigned int port, + u8 default_roce_tos) +{ + if (!rdma_is_port_valid(cma_dev->device, port)) + return -EINVAL; + + cma_dev->default_roce_tos[port - rdma_start_port(cma_dev->device)] = + default_roce_tos; + + return 0; +} struct ib_device *cma_get_ib_dev(struct cma_device *cma_dev) { return cma_dev->device; @@ -341,6 +361,7 @@ struct rdma_id_private { u32 options; u8 srq; u8 tos; + bool tos_set; u8 reuseaddr; u8 afonly; enum ib_gid_type gid_type; @@ -780,6 +801,7 @@ struct rdma_cm_id *rdma_create_id(struct net *net, id_priv->id.event_handler = event_handler; id_priv->id.ps = ps; id_priv->id.qp_type = qp_type; + id_priv->tos_set = false; spin_lock_init(&id_priv->lock); mutex_init(&id_priv->qp_mutex); init_completion(&id_priv->comp); @@ -2271,6 +2293,7 @@ void rdma_set_service_type(struct rdma_cm_id *id, int tos) id_priv = container_of(id, struct rdma_id_private, id); id_priv->tos = (u8) tos; + id_priv->tos_set = true; } EXPORT_SYMBOL(rdma_set_service_type); @@ -2507,6 +2530,9 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) struct cma_work *work; int ret; struct net_device *ndev = NULL; + u8 default_roce_tos = id_priv->cma_dev->default_roce_tos[id_priv->id.port_num - + rdma_start_port(id_priv->cma_dev->device)]; + u8 tos = id_priv->tos_set ? id_priv->tos : default_roce_tos; work = kzalloc(sizeof *work, GFP_KERNEL); @@ -2580,7 +2606,8 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) route->path_rec->reversible = 1; route->path_rec->pkey = cpu_to_be16(0xffff); route->path_rec->mtu_selector = IB_SA_EQ; - route->path_rec->sl = iboe_tos_to_sl(ndev, id_priv->tos); + route->path_rec->sl = iboe_tos_to_sl(ndev, tos); + route->path_rec->traffic_class = tos; route->path_rec->mtu = iboe_get_mtu(ndev->mtu); route->path_rec->rate_selector = IB_SA_EQ; route->path_rec->rate = iboe_get_rate(ndev); @@ -4304,15 +4331,21 @@ static void cma_add_one(struct ib_device *device) cma_dev->default_gid_type = kcalloc(device->phys_port_cnt, sizeof(*cma_dev->default_gid_type), GFP_KERNEL); - if (!cma_dev->default_gid_type) { - kfree(cma_dev); - return; - } + if (!cma_dev->default_gid_type) + goto free_cma_dev; + + cma_dev->default_roce_tos = kcalloc(device->phys_port_cnt, + sizeof(*cma_dev->default_roce_tos), + GFP_KERNEL); + if (!cma_dev->default_roce_tos) + goto free_gid_type; + for (i = rdma_start_port(device); i <= rdma_end_port(device); i++) { supported_gids = roce_gid_type_mask_support(device, i); WARN_ON(!supported_gids); cma_dev->default_gid_type[i - rdma_start_port(device)] = find_first_bit(&supported_gids, BITS_PER_LONG); + cma_dev->default_roce_tos[i - rdma_start_port(device)] = 0; } init_completion(&cma_dev->comp); @@ -4325,6 +4358,16 @@ static void cma_add_one(struct ib_device *device) list_for_each_entry(id_priv, &listen_any_list, list) cma_listen_on_dev(id_priv, cma_dev); mutex_unlock(&lock); + + return; + +free_gid_type: + kfree(cma_dev->default_gid_type); + +free_cma_dev: + kfree(cma_dev); + + return; } static int cma_remove_id_dev(struct rdma_id_private *id_priv) @@ -4393,6 +4436,7 @@ static void cma_remove_one(struct ib_device *device, void *client_data) mutex_unlock(&lock); cma_process_remove(cma_dev); + kfree(cma_dev->default_roce_tos); kfree(cma_dev->default_gid_type); kfree(cma_dev); } diff --git a/drivers/infiniband/core/cma_configfs.c b/drivers/infiniband/core/cma_configfs.c index 41573df1d9fc..54076a3e8007 100644 --- a/drivers/infiniband/core/cma_configfs.c +++ b/drivers/infiniband/core/cma_configfs.c @@ -139,8 +139,50 @@ static ssize_t default_roce_mode_store(struct config_item *item, CONFIGFS_ATTR(, default_roce_mode); +static ssize_t default_roce_tos_show(struct config_item *item, char *buf) +{ + struct cma_device *cma_dev; + struct cma_dev_port_group *group; + ssize_t ret; + u8 tos; + + ret = cma_configfs_params_get(item, &cma_dev, &group); + if (ret) + return ret; + + tos = cma_get_default_roce_tos(cma_dev, group->port_num); + cma_configfs_params_put(cma_dev); + + return sprintf(buf, "%u\n", tos); +} + +static ssize_t default_roce_tos_store(struct config_item *item, + const char *buf, size_t count) +{ + struct cma_device *cma_dev; + struct cma_dev_port_group *group; + ssize_t ret; + u8 tos; + + ret = kstrtou8(buf, 0, &tos); + if (ret) + return ret; + + ret = cma_configfs_params_get(item, &cma_dev, &group); + if (ret) + return ret; + + ret = cma_set_default_roce_tos(cma_dev, group->port_num, tos); + cma_configfs_params_put(cma_dev); + + return ret ? ret : strnlen(buf, count); +} + +CONFIGFS_ATTR(, default_roce_tos); + static struct configfs_attribute *cma_configfs_attributes[] = { &attr_default_roce_mode, + &attr_default_roce_tos, NULL, }; diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index d29372624f3a..912ab4cd6eae 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -62,6 +62,9 @@ int cma_get_default_gid_type(struct cma_device *cma_dev, int cma_set_default_gid_type(struct cma_device *cma_dev, unsigned int port, enum ib_gid_type default_gid_type); +int cma_get_default_roce_tos(struct cma_device *cma_dev, unsigned int port); +int cma_set_default_roce_tos(struct cma_device *a_dev, unsigned int port, + u8 default_roce_tos); struct ib_device *cma_get_ib_dev(struct cma_device *cma_dev); int ib_device_register_sysfs(struct ib_device *device, -- cgit v1.2.3 From 6df6b4a9ce43deb57d3e23b0e79a529464b5cf37 Mon Sep 17 00:00:00 2001 From: Moni Shoua Date: Tue, 14 Feb 2017 07:24:08 +0200 Subject: IB/cma: Destination and source addr families must match The destination address in a listening rdma_id does not have an address family. Since address family in both sides of a connection must be the same in rdma_bind_addr() we set the address family of the destination to the address family of the source. This patch serves the logic in cma_port_is_unique() which requires to know if destination address that is associated with a rdma_id is any address (cma_zero_addr() and cma_loopback_addr()). This can happen when port reuse is checked for a port number that is being listened to. Fixes: 19b752a19dce ("IB/cma: Allow port reuse for rdma_id") Signed-off-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index f98ec19a851a..ba60e4b5114c 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -3297,6 +3297,7 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr) { struct rdma_id_private *id_priv; int ret; + struct sockaddr *daddr; if (addr->sa_family != AF_INET && addr->sa_family != AF_INET6 && addr->sa_family != AF_IB) @@ -3336,6 +3337,9 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr) if (ret) goto err2; + daddr = cma_dst_addr(id_priv); + daddr->sa_family = addr->sa_family; + return 0; err2: if (id_priv->cma_dev) -- cgit v1.2.3 From 192539f4ce36e5b80b9eb1f4ccd759a92eed2ff9 Mon Sep 17 00:00:00 2001 From: Ganesh Goudar Date: Tue, 31 Jan 2017 12:00:09 +0530 Subject: iw_cxgb4: clean up send_connect() Clean up send_connect() and make use of t6 specific active open request struct. Acked-by: Steve Wise Signed-off-by: Bharat Teja Signed-off-by: Ganesh Goudar Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 56 ++++++++++++++++++++++++++-------------- 1 file changed, 36 insertions(+), 20 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index a33f37998e63..03a1b0e64fc3 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -692,6 +692,10 @@ static int send_connect(struct c4iw_ep *ep) int ret; enum chip_type adapter_type = ep->com.dev->rdev.lldi.adapter_type; u32 isn = (prandom_u32() & ~7UL) - 1; + struct net_device *netdev; + u64 params; + + netdev = ep->com.dev->rdev.lldi.ports[0]; switch (CHELSIO_CHIP_VERSION(adapter_type)) { case CHELSIO_T4: @@ -768,6 +772,8 @@ static int send_connect(struct c4iw_ep *ep) opt2 |= T5_ISS_F; } + params = cxgb4_select_ntuple(netdev, ep->l2t); + if (ep->com.remote_addr.ss_family == AF_INET6) cxgb4_clip_get(ep->com.dev->rdev.lldi.ports[0], (const u32 *)&la6->sin6_addr.s6_addr, 1); @@ -809,18 +815,22 @@ static int send_connect(struct c4iw_ep *ep) req->opt0 = cpu_to_be64(opt0); if (is_t4(ep->com.dev->rdev.lldi.adapter_type)) { - req->params = cpu_to_be32(cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t)); + req->params = cpu_to_be32(params); req->opt2 = cpu_to_be32(opt2); } else { - t5req->params = cpu_to_be64(FILTER_TUPLE_V( - cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t))); - t5req->rsvd = cpu_to_be32(isn); - PDBG("%s snd_isn %u\n", __func__, t5req->rsvd); - t5req->opt2 = cpu_to_be32(opt2); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + t5req->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t5req->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, t5req->rsvd); + t5req->opt2 = cpu_to_be32(opt2); + } else { + t6req->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t6req->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, t6req->rsvd); + t6req->opt2 = cpu_to_be32(opt2); + } } } else { switch (CHELSIO_CHIP_VERSION(adapter_type)) { @@ -859,18 +869,24 @@ static int send_connect(struct c4iw_ep *ep) req6->opt0 = cpu_to_be64(opt0); if (is_t4(ep->com.dev->rdev.lldi.adapter_type)) { - req6->params = cpu_to_be32(cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t)); + req6->params = cpu_to_be32(cxgb4_select_ntuple(netdev, + ep->l2t)); req6->opt2 = cpu_to_be32(opt2); } else { - t5req6->params = cpu_to_be64(FILTER_TUPLE_V( - cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t))); - t5req6->rsvd = cpu_to_be32(isn); - PDBG("%s snd_isn %u\n", __func__, t5req6->rsvd); - t5req6->opt2 = cpu_to_be32(opt2); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + t5req6->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t5req6->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, t5req6->rsvd); + t5req6->opt2 = cpu_to_be32(opt2); + } else { + t6req6->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t6req6->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, t6req6->rsvd); + t6req6->opt2 = cpu_to_be32(opt2); + } + } } -- cgit v1.2.3 From 32f8e839edebe3310dd077439ce0e9a96e0c2352 Mon Sep 17 00:00:00 2001 From: Max Gurtovoy Date: Thu, 2 Feb 2017 01:55:18 +0200 Subject: IB/iser: Protect completion context active_qps update As iser connections can share completion contexts, we need to protect the active_qps update each time we set it. Signed-off-by: Max Gurtovoy Acked-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/iser/iser_verbs.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 6a9d1cb548ee..30b622f2ab73 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -597,7 +597,9 @@ static void iser_free_ib_conn_res(struct iser_conn *iser_conn, iser_conn, ib_conn->cma_id, ib_conn->qp); if (ib_conn->qp != NULL) { + mutex_lock(&ig.connlist_mutex); ib_conn->comp->active_qps--; + mutex_unlock(&ig.connlist_mutex); rdma_destroy_qp(ib_conn->cma_id); ib_conn->qp = NULL; } -- cgit v1.2.3 From 8d8a47338089d96093fa5fb8a686d9faefce04a1 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Tue, 7 Feb 2017 15:15:10 +0000 Subject: IB/rxe: use setup_timer to simplify the code Use setup_timer function instead of initializing timer with the function and data fields. Signed-off-by: Wei Yongjun Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rxe/rxe_qp.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index 5749ff0ea9ba..f98a19e61a3d 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -273,13 +273,8 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, rxe_init_task(rxe, &qp->comp.task, qp, rxe_completer, "comp"); - init_timer(&qp->rnr_nak_timer); - qp->rnr_nak_timer.function = rnr_nak_timer; - qp->rnr_nak_timer.data = (unsigned long)qp; - - init_timer(&qp->retrans_timer); - qp->retrans_timer.function = retransmit_timer; - qp->retrans_timer.data = (unsigned long)qp; + setup_timer(&qp->rnr_nak_timer, rnr_nak_timer, (unsigned long)qp); + setup_timer(&qp->retrans_timer, retransmit_timer, (unsigned long)qp); qp->qp_timeout_jiffies = 0; /* Can't be set for UD/UC in modify_qp */ return 0; -- cgit v1.2.3 From 4fcf1de5a79bfe8ee4c638bd3902d20cd59376de Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:25:56 -0800 Subject: IB/hfi1: Correct defered count after processing qp_wait_list The qp_wait_list processing leaves the defered ack count at its prior value. This can result in a premature send of an ack. Fixed by unconditionally reseting the defered ack count in hfi1_send_rc_ack(). Fixes: Commit 7c091e5c0685 ("staging/rdma/hfi1: add ACK coalescing logic") Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/rc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 809b26eb6d3c..1dd999e9349c 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -853,6 +853,10 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, struct ib_header hdr; struct ib_other_headers *ohdr; unsigned long flags; + struct hfi1_qp_priv *priv = qp->priv; + + /* clear the defer count */ + priv->r_adefered = 0; /* Don't send ACK or NAK if a RDMA read or atomic is pending. */ if (qp->s_flags & RVT_S_RESP_PENDING) -- cgit v1.2.3 From a82a7fcd1f23497a97df237829622dfdfacf24fe Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:26:02 -0800 Subject: IB/hfi1: Process qp wait list in IRQ thread periodically In the event that the IRQ thread is extremely busy, the processing of an rcd wait list can be delayed by quite a bit until the IRQ thread completes its work. The QP reset reference count wait can then appear to be stuck, thus causing up a QP destroy to emit the hung task diagnostic. Fix by processing the qp wait list periodically from the thread. The interval is a multiple (currently 4) of the MAX_PKT_RECV. Also, reduce some of the excessive inlining. The guidelines are per packet is ok inline, otherwise the choice is based on likelyhood of execution. Reviewed-by: Sebastian Sanchez Reviewed-by: Dennis Dalessandro Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/driver.c | 121 ++++++++++++++++++++---------------- 1 file changed, 67 insertions(+), 54 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c index 4fbaee68012b..29db67371775 100644 --- a/drivers/infiniband/hw/hfi1/driver.c +++ b/drivers/infiniband/hw/hfi1/driver.c @@ -100,6 +100,11 @@ MODULE_VERSION(HFI1_DRIVER_VERSION); * MAX_PKT_RCV is the max # if packets processed per receive interrupt. */ #define MAX_PKT_RECV 64 +/* + * MAX_PKT_THREAD_RCV is the max # of packets processed before + * the qp_wait_list queue is flushed. + */ +#define MAX_PKT_RECV_THREAD (MAX_PKT_RECV * 4) #define EGR_HEAD_UPDATE_THRESHOLD 16 struct hfi1_ib_stats hfi1_stats; @@ -259,7 +264,7 @@ static inline void *get_egrbuf(const struct hfi1_ctxtdata *rcd, u64 rhf, * allowed size ranges for the respective type and, optionally, * return the proper encoding. */ -inline int hfi1_rcvbuf_validate(u32 size, u8 type, u16 *encoded) +int hfi1_rcvbuf_validate(u32 size, u8 type, u16 *encoded) { if (unlikely(!PAGE_ALIGNED(size))) return 0; @@ -654,24 +659,68 @@ next: } } -static inline int skip_rcv_packet(struct hfi1_packet *packet, int thread) +static void process_rcv_qp_work(struct hfi1_ctxtdata *rcd) +{ + struct rvt_qp *qp, *nqp; + + /* + * Iterate over all QPs waiting to respond. + * The list won't change since the IRQ is only run on one CPU. + */ + list_for_each_entry_safe(qp, nqp, &rcd->qp_wait_list, rspwait) { + list_del_init(&qp->rspwait); + if (qp->r_flags & RVT_R_RSP_NAK) { + qp->r_flags &= ~RVT_R_RSP_NAK; + hfi1_send_rc_ack(rcd, qp, 0); + } + if (qp->r_flags & RVT_R_RSP_SEND) { + unsigned long flags; + + qp->r_flags &= ~RVT_R_RSP_SEND; + spin_lock_irqsave(&qp->s_lock, flags); + if (ib_rvt_state_ops[qp->state] & + RVT_PROCESS_OR_FLUSH_SEND) + hfi1_schedule_send(qp); + spin_unlock_irqrestore(&qp->s_lock, flags); + } + rvt_put_qp(qp); + } +} + +static noinline int max_packet_exceeded(struct hfi1_packet *packet, int thread) +{ + if (thread) { + if ((packet->numpkt & (MAX_PKT_RECV_THREAD - 1)) == 0) + /* allow defered processing */ + process_rcv_qp_work(packet->rcd); + cond_resched(); + return RCV_PKT_OK; + } else { + this_cpu_inc(*packet->rcd->dd->rcv_limit); + return RCV_PKT_LIMIT; + } +} + +static inline int check_max_packet(struct hfi1_packet *packet, int thread) { int ret = RCV_PKT_OK; + if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) + ret = max_packet_exceeded(packet, thread); + return ret; +} + +static noinline int skip_rcv_packet(struct hfi1_packet *packet, int thread) +{ + int ret; + /* Set up for the next packet */ packet->rhqoff += packet->rsize; if (packet->rhqoff >= packet->maxcnt) packet->rhqoff = 0; packet->numpkt++; - if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) { - if (thread) { - cond_resched(); - } else { - ret = RCV_PKT_LIMIT; - this_cpu_inc(*packet->rcd->dd->rcv_limit); - } - } + ret = check_max_packet(packet, thread); packet->rhf_addr = (__le32 *)packet->rcd->rcvhdrq + packet->rhqoff + packet->rcd->dd->rhf_offset; @@ -682,7 +731,7 @@ static inline int skip_rcv_packet(struct hfi1_packet *packet, int thread) static inline int process_rcv_packet(struct hfi1_packet *packet, int thread) { - int ret = RCV_PKT_OK; + int ret; packet->hdr = hfi1_get_msgheader(packet->rcd->dd, packet->rhf_addr); @@ -723,14 +772,7 @@ static inline int process_rcv_packet(struct hfi1_packet *packet, int thread) if (packet->rhqoff >= packet->maxcnt) packet->rhqoff = 0; - if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) { - if (thread) { - cond_resched(); - } else { - ret = RCV_PKT_LIMIT; - this_cpu_inc(*packet->rcd->dd->rcv_limit); - } - } + ret = check_max_packet(packet, thread); packet->rhf_addr = (__le32 *)packet->rcd->rcvhdrq + packet->rhqoff + packet->rcd->dd->rhf_offset; @@ -767,38 +809,6 @@ static inline void finish_packet(struct hfi1_packet *packet) packet->etail, rcv_intr_dynamic, packet->numpkt); } -static inline void process_rcv_qp_work(struct hfi1_packet *packet) -{ - struct hfi1_ctxtdata *rcd; - struct rvt_qp *qp, *nqp; - - rcd = packet->rcd; - rcd->head = packet->rhqoff; - - /* - * Iterate over all QPs waiting to respond. - * The list won't change since the IRQ is only run on one CPU. - */ - list_for_each_entry_safe(qp, nqp, &rcd->qp_wait_list, rspwait) { - list_del_init(&qp->rspwait); - if (qp->r_flags & RVT_R_RSP_NAK) { - qp->r_flags &= ~RVT_R_RSP_NAK; - hfi1_send_rc_ack(rcd, qp, 0); - } - if (qp->r_flags & RVT_R_RSP_SEND) { - unsigned long flags; - - qp->r_flags &= ~RVT_R_RSP_SEND; - spin_lock_irqsave(&qp->s_lock, flags); - if (ib_rvt_state_ops[qp->state] & - RVT_PROCESS_OR_FLUSH_SEND) - hfi1_schedule_send(qp); - spin_unlock_irqrestore(&qp->s_lock, flags); - } - rvt_put_qp(qp); - } -} - /* * Handle receive interrupts when using the no dma rtail option. */ @@ -826,7 +836,8 @@ int handle_receive_interrupt_nodma_rtail(struct hfi1_ctxtdata *rcd, int thread) last = RCV_PKT_DONE; process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: finish_packet(&packet); return last; @@ -854,7 +865,8 @@ int handle_receive_interrupt_dma_rtail(struct hfi1_ctxtdata *rcd, int thread) last = RCV_PKT_DONE; process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: finish_packet(&packet); return last; @@ -1024,7 +1036,8 @@ int handle_receive_interrupt(struct hfi1_ctxtdata *rcd, int thread) process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: /* -- cgit v1.2.3 From eb04ff09d834c1a531d02ee0b3e4b503253df609 Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:26:08 -0800 Subject: IB/hfi1: Ensure read of producer s_head is correct The read of s_head in the hfi1_make_rc_req() and qib_make_rc_req() lack the necesary barrier instuctions. Correct other ACCESS_ONCE() warnings in the same file. Reviewed-by: Ashutosh Dixit Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/rc.c | 7 ++++--- drivers/infiniband/hw/qib/qib_rc.c | 7 ++++--- 2 files changed, 8 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 1dd999e9349c..6446179843e9 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -414,7 +414,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) goto bail; /* We are in the error state, flush the work request. */ smp_read_barrier_depends(); /* see post_one_send() */ - if (qp->s_last == ACCESS_ONCE(qp->s_head)) + if (qp->s_last == READ_ONCE(qp->s_head)) goto bail; /* If DMAs are in progress, we can't flush immediately. */ if (iowait_sdma_pending(&priv->s_iowait)) { @@ -457,7 +457,8 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) newreq = 0; if (qp->s_cur == qp->s_tail) { /* Check if send work queue is empty. */ - if (qp->s_tail == qp->s_head) { + smp_read_barrier_depends(); /* see post_one_send() */ + if (qp->s_tail == READ_ONCE(qp->s_head)) { clear_ahg(qp); goto bail; } @@ -1590,7 +1591,7 @@ static void rc_rcv_resp(struct hfi1_ibport *ibp, /* Ignore invalid responses. */ smp_read_barrier_depends(); /* see post_one_send */ - if (cmp_psn(psn, ACCESS_ONCE(qp->s_next_psn)) >= 0) + if (cmp_psn(psn, READ_ONCE(qp->s_next_psn)) >= 0) goto ack_done; /* Ignore duplicate responses. */ diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 031433cb7206..696bcd0d087c 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -257,7 +257,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) goto bail; /* We are in the error state, flush the work request. */ smp_read_barrier_depends(); /* see post_one_send() */ - if (qp->s_last == ACCESS_ONCE(qp->s_head)) + if (qp->s_last == READ_ONCE(qp->s_head)) goto bail; /* If DMAs are in progress, we can't flush immediately. */ if (atomic_read(&priv->s_dma_busy)) { @@ -303,7 +303,8 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) newreq = 0; if (qp->s_cur == qp->s_tail) { /* Check if send work queue is empty. */ - if (qp->s_tail == qp->s_head) + smp_read_barrier_depends(); /* see post_one_send() */ + if (qp->s_tail == READ_ONCE(qp->s_head)) goto bail; /* * If a fence is requested, wait for previous @@ -1390,7 +1391,7 @@ static void qib_rc_rcv_resp(struct qib_ibport *ibp, /* Ignore invalid responses. */ smp_read_barrier_depends(); /* see post_one_send */ - if (qib_cmp24(psn, ACCESS_ONCE(qp->s_next_psn)) >= 0) + if (qib_cmp24(psn, READ_ONCE(qp->s_next_psn)) >= 0) goto ack_done; /* Ignore duplicate responses. */ -- cgit v1.2.3 From 39e2afa8d042a53d855137d4c5a689a6f5492b39 Mon Sep 17 00:00:00 2001 From: Easwar Hariharan Date: Wed, 8 Feb 2017 05:26:14 -0800 Subject: IB/hfi1: Use static CTLE with Preset 6 for integrated HFIs After extended testing, it was found that the previous PCIe Gen 3 recipe, which used adaptive CTLE with Preset 4, could cause an NMI/Surprise Link Down in about 1 in 100 to 1 in 1000 power cycles on some platforms. New EV data combined with extensive empirical data indicates that the new recipe should use static CTLE with Preset 6 for all integrated silicon SKUs. Fixes: c3f8de0b334c ("IB/hfi1: Add static PCIe Gen3 CTLE tuning") Reviewed-by: Dennis Dalessandro Signed-off-by: Easwar Hariharan Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/pcie.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c index ebd941fc8a92..0829fce06172 100644 --- a/drivers/infiniband/hw/hfi1/pcie.c +++ b/drivers/infiniband/hw/hfi1/pcie.c @@ -663,12 +663,12 @@ MODULE_PARM_DESC(pcie_retry, "Driver will try this many times to reach requested #define UNSET_PSET 255 #define DEFAULT_DISCRETE_PSET 2 /* discrete HFI */ -#define DEFAULT_MCP_PSET 4 /* MCP HFI */ +#define DEFAULT_MCP_PSET 6 /* MCP HFI */ static uint pcie_pset = UNSET_PSET; module_param(pcie_pset, uint, S_IRUGO); MODULE_PARM_DESC(pcie_pset, "PCIe Eq Pset value to use, range is 0-10"); -static uint pcie_ctle = 1; /* discrete on, integrated off */ +static uint pcie_ctle = 3; /* discrete on, integrated on */ module_param(pcie_ctle, uint, S_IRUGO); MODULE_PARM_DESC(pcie_ctle, "PCIe static CTLE mode, bit 0 - discrete on/off, bit 1 - integrated on/off"); -- cgit v1.2.3 From a8715b97d63718fc5c4daebc465407c259aea265 Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:26:20 -0800 Subject: IB/hfi1: Correct error calldown locking The resource specific wait locking missed correcting the lock for the notify_error_qp() calldown. The code is fixed to correctly use the iowait lock field to protect the head that is protected by that lock. Fixes: Commit 4e045572e2c2 ("IB/hfi1: Add unique txwait_lock for txreq events") Reviewed-by: Sebastian Sanchez Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/qp.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index d752d6768a49..c2a166d473b2 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -961,17 +961,20 @@ int get_pmtu_from_attr(struct rvt_dev_info *rdi, struct rvt_qp *qp, void notify_error_qp(struct rvt_qp *qp) { - struct hfi1_ibdev *dev = to_idev(qp->ibqp.device); struct hfi1_qp_priv *priv = qp->priv; + seqlock_t *lock = priv->s_iowait.lock; - write_seqlock(&dev->iowait_lock); - if (!list_empty(&priv->s_iowait.list) && !(qp->s_flags & RVT_S_BUSY)) { - qp->s_flags &= ~RVT_S_ANY_WAIT_IO; - list_del_init(&priv->s_iowait.list); - priv->s_iowait.lock = NULL; - rvt_put_qp(qp); + if (lock) { + write_seqlock(lock); + if (!list_empty(&priv->s_iowait.list) && + !(qp->s_flags & RVT_S_BUSY)) { + qp->s_flags &= ~RVT_S_ANY_WAIT_IO; + list_del_init(&priv->s_iowait.list); + priv->s_iowait.lock = NULL; + rvt_put_qp(qp); + } + write_sequnlock(lock); } - write_sequnlock(&dev->iowait_lock); if (!(qp->s_flags & RVT_S_BUSY)) { qp->s_hdrwords = 0; -- cgit v1.2.3 From f3e862cb6894389a35d0beb10f73d62eb3317beb Mon Sep 17 00:00:00 2001 From: Sebastian Sanchez Date: Wed, 8 Feb 2017 05:26:25 -0800 Subject: IB/hfi1: Access hfi1_ibport through rcd pointer Receive code paths use the QP's device and port number to access the struct hfi1_ibport. When an instance of struct hfi1_ctxtdata is present, it can be used to access struct hfi1_ibport through a pointer. This makes struct hfi1_ibport lookup time faster as an array doesn't have to be indexed and access fields in other cache-lines. Reviewed-by: Mike Marciniszyn Signed-off-by: Sebastian Sanchez Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/driver.c | 4 ++-- drivers/infiniband/hw/hfi1/hfi.h | 5 +++++ drivers/infiniband/hw/hfi1/rc.c | 10 +++++----- drivers/infiniband/hw/hfi1/uc.c | 2 +- drivers/infiniband/hw/hfi1/ud.c | 2 +- drivers/infiniband/hw/hfi1/verbs.c | 4 ++-- 6 files changed, 16 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c index 29db67371775..3881c951f6af 100644 --- a/drivers/infiniband/hw/hfi1/driver.c +++ b/drivers/infiniband/hw/hfi1/driver.c @@ -284,7 +284,7 @@ static void rcv_hdrerr(struct hfi1_ctxtdata *rcd, struct hfi1_pportdata *ppd, struct ib_header *rhdr = packet->hdr; u32 rte = rhf_rcv_type_err(packet->rhf); int lnh = be16_to_cpu(rhdr->lrh[0]) & 3; - struct hfi1_ibport *ibp = &ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct hfi1_devdata *dd = ppd->dd; struct rvt_dev_info *rdi = &dd->verbs_dev.rdi; @@ -599,7 +599,7 @@ static void __prescan_rxq(struct hfi1_packet *packet) while (1) { struct hfi1_devdata *dd = rcd->dd; - struct hfi1_ibport *ibp = &rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(rcd); __le32 *rhf_addr = (__le32 *)rcd->rcvhdrq + mdata.ps_head + dd->rhf_offset; struct rvt_qp *qp; diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 751a0fb29fa5..1d0873914a6f 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1584,6 +1584,11 @@ static inline struct hfi1_ibport *to_iport(struct ib_device *ibdev, u8 port) return &dd->pport[pidx].ibport_data; } +static inline struct hfi1_ibport *rcd_to_iport(struct hfi1_ctxtdata *rcd) +{ + return &rcd->ppd->ibport_data; +} + void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt, bool do_cnp); static inline bool process_ecn(struct rvt_qp *qp, struct hfi1_packet *pkt, diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 6446179843e9..abea4b7c92f4 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -841,7 +841,7 @@ bail_no_tx: void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, int is_fecn) { - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); u64 pbc, pbc_flags = 0; u16 lrh0; @@ -1326,7 +1326,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, if (aeth >> 29) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); - ibp = to_iport(qp->ibqp.device, qp->port_num); + ibp = rcd_to_iport(rcd); /* * The MSN might be for a later WQE than the PSN indicates so @@ -1791,7 +1791,7 @@ static noinline int rc_rcv_error(struct ib_other_headers *ohdr, void *data, struct rvt_qp *qp, u32 opcode, u32 psn, int diff, struct hfi1_ctxtdata *rcd) { - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct rvt_ack_entry *e; unsigned long flags; u8 i, prev; @@ -2100,7 +2100,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) void *data = packet->ebuf; u32 tlen = packet->tlen; struct rvt_qp *qp = packet->qp; - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct ib_other_headers *ohdr = packet->ohdr; u32 bth0, opcode; u32 hdrsize = packet->hlen; @@ -2552,7 +2552,7 @@ void hfi1_rc_hdrerr( { int has_grh = rcv_flags & HFI1_HAS_GRH; struct ib_other_headers *ohdr; - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); int diff; u32 opcode; u32 psn, bth0; diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c index b141a78ae38b..74b7b2be458c 100644 --- a/drivers/infiniband/hw/hfi1/uc.c +++ b/drivers/infiniband/hw/hfi1/uc.c @@ -296,7 +296,7 @@ bail_no_tx: */ void hfi1_uc_rcv(struct hfi1_packet *packet) { - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct ib_header *hdr = packet->hdr; u32 rcv_flags = packet->rcv_flags; void *data = packet->ebuf; diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c index c071955c0272..6d81d79720b0 100644 --- a/drivers/infiniband/hw/hfi1/ud.c +++ b/drivers/infiniband/hw/hfi1/ud.c @@ -672,7 +672,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) u32 src_qp; u16 dlid, pkey; int mgmt_pkey_idx = -1; - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); struct ib_header *hdr = packet->hdr; u32 rcv_flags = packet->rcv_flags; diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index 95ed4d6da510..b937a23efc75 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -576,7 +576,7 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) struct ib_header *hdr = packet->hdr; u32 tlen = packet->tlen; struct hfi1_pportdata *ppd = rcd->ppd; - struct hfi1_ibport *ibp = &ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct rvt_dev_info *rdi = &ppd->dd->verbs_dev.rdi; opcode_handler packet_handler; unsigned long flags; @@ -1910,7 +1910,7 @@ void hfi1_unregister_ib_device(struct hfi1_devdata *dd) void hfi1_cnp_rcv(struct hfi1_packet *packet) { - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); struct ib_header *hdr = packet->hdr; struct rvt_qp *qp = packet->qp; -- cgit v1.2.3 From 338adfdddf6abe89726e1146ad3102ce9663a634 Mon Sep 17 00:00:00 2001 From: Sebastian Sanchez Date: Wed, 8 Feb 2017 05:26:31 -0800 Subject: IB/rdmavt: Use per-CPU reference count for MRs Having per-CPU reference count for each MR prevents cache-line bouncing across the system. Thus, it prevents bottlenecks. Use per-CPU reference counts per MR. The per-CPU reference count for FMRs is used in atomic mode to allow accurate testing of the busy state. Other MR types run in per-CPU mode MR until they're freed. Reviewed-by: Mike Marciniszyn Signed-off-by: Sebastian Sanchez Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rdmavt/mr.c | 59 +++++++++++++++++++++++++-------------- include/rdma/rdmavt_mr.h | 10 +++---- 2 files changed, 43 insertions(+), 26 deletions(-) diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c index 52fd15276ee6..c80a69b1ffcb 100644 --- a/drivers/infiniband/sw/rdmavt/mr.c +++ b/drivers/infiniband/sw/rdmavt/mr.c @@ -120,10 +120,19 @@ static void rvt_deinit_mregion(struct rvt_mregion *mr) mr->mapsz = 0; while (i) kfree(mr->map[--i]); + percpu_ref_exit(&mr->refcount); +} + +static void __rvt_mregion_complete(struct percpu_ref *ref) +{ + struct rvt_mregion *mr = container_of(ref, struct rvt_mregion, + refcount); + + complete(&mr->comp); } static int rvt_init_mregion(struct rvt_mregion *mr, struct ib_pd *pd, - int count) + int count, unsigned int percpu_flags) { int m, i = 0; struct rvt_dev_info *dev = ib_to_rvt(pd->device); @@ -133,19 +142,23 @@ static int rvt_init_mregion(struct rvt_mregion *mr, struct ib_pd *pd, for (; i < m; i++) { mr->map[i] = kzalloc_node(sizeof(*mr->map[0]), GFP_KERNEL, dev->dparms.node); - if (!mr->map[i]) { - rvt_deinit_mregion(mr); - return -ENOMEM; - } + if (!mr->map[i]) + goto bail; mr->mapsz++; } init_completion(&mr->comp); /* count returning the ptr to user */ - atomic_set(&mr->refcount, 1); + if (percpu_ref_init(&mr->refcount, &__rvt_mregion_complete, + percpu_flags, GFP_KERNEL)) + goto bail; + atomic_set(&mr->lkey_invalid, 0); mr->pd = pd; mr->max_segs = count; return 0; +bail: + rvt_deinit_mregion(mr); + return -ENOMEM; } /** @@ -180,8 +193,7 @@ static int rvt_alloc_lkey(struct rvt_mregion *mr, int dma_region) if (!tmr) { rcu_assign_pointer(dev->dma_mr, mr); mr->lkey_published = 1; - } else { - rvt_put_mr(mr); + rvt_get_mr(mr); } goto success; } @@ -239,11 +251,14 @@ static void rvt_free_lkey(struct rvt_mregion *mr) int freed = 0; spin_lock_irqsave(&rkt->lock, flags); - if (!mr->lkey_published) - goto out; - if (lkey == 0) { - RCU_INIT_POINTER(dev->dma_mr, NULL); + if (!lkey) { + if (mr->lkey_published) { + RCU_INIT_POINTER(dev->dma_mr, NULL); + rvt_put_mr(mr); + } } else { + if (!mr->lkey_published) + goto out; r = lkey >> (32 - dev->dparms.lkey_table_size); RCU_INIT_POINTER(rkt->table[r], NULL); } @@ -253,7 +268,7 @@ out: spin_unlock_irqrestore(&rkt->lock, flags); if (freed) { synchronize_rcu(); - rvt_put_mr(mr); + percpu_ref_kill(&mr->refcount); } } @@ -269,7 +284,7 @@ static struct rvt_mr *__rvt_alloc_mr(int count, struct ib_pd *pd) if (!mr) goto bail; - rval = rvt_init_mregion(&mr->mr, pd, count); + rval = rvt_init_mregion(&mr->mr, pd, count, 0); if (rval) goto bail; /* @@ -294,8 +309,8 @@ bail: static void __rvt_free_mr(struct rvt_mr *mr) { - rvt_deinit_mregion(&mr->mr); rvt_free_lkey(&mr->mr); + rvt_deinit_mregion(&mr->mr); kfree(mr); } @@ -323,7 +338,7 @@ struct ib_mr *rvt_get_dma_mr(struct ib_pd *pd, int acc) goto bail; } - rval = rvt_init_mregion(&mr->mr, pd, 0); + rval = rvt_init_mregion(&mr->mr, pd, 0, 0); if (rval) { ret = ERR_PTR(rval); goto bail; @@ -445,8 +460,8 @@ int rvt_dereg_mr(struct ib_mr *ibmr) timeout = wait_for_completion_timeout(&mr->mr.comp, 5 * HZ); if (!timeout) { rvt_pr_err(rdi, - "rvt_dereg_mr timeout mr %p pd %p refcount %u\n", - mr, mr->mr.pd, atomic_read(&mr->mr.refcount)); + "rvt_dereg_mr timeout mr %p pd %p\n", + mr, mr->mr.pd); rvt_get_mr(&mr->mr); ret = -EBUSY; goto out; @@ -623,7 +638,8 @@ struct ib_fmr *rvt_alloc_fmr(struct ib_pd *pd, int mr_access_flags, if (!fmr) goto bail; - rval = rvt_init_mregion(&fmr->mr, pd, fmr_attr->max_pages); + rval = rvt_init_mregion(&fmr->mr, pd, fmr_attr->max_pages, + PERCPU_REF_INIT_ATOMIC); if (rval) goto bail; @@ -674,11 +690,12 @@ int rvt_map_phys_fmr(struct ib_fmr *ibfmr, u64 *page_list, struct rvt_fmr *fmr = to_ifmr(ibfmr); struct rvt_lkey_table *rkt; unsigned long flags; - int m, n, i; + int m, n; + unsigned long i; u32 ps; struct rvt_dev_info *rdi = ib_to_rvt(ibfmr->device); - i = atomic_read(&fmr->mr.refcount); + i = atomic_long_read(&fmr->mr.refcount.count); if (i > 2) return -EBUSY; diff --git a/include/rdma/rdmavt_mr.h b/include/rdma/rdmavt_mr.h index de59de28b6a2..05698d8d9c6f 100644 --- a/include/rdma/rdmavt_mr.h +++ b/include/rdma/rdmavt_mr.h @@ -52,6 +52,7 @@ * For Memory Regions. This stuff should probably be moved into rdmavt/mr.h once * drivers no longer need access to the MR directly. */ +#include /* * A segment is a linear region of low physical memory. @@ -79,11 +80,11 @@ struct rvt_mregion { int access_flags; u32 max_segs; /* number of rvt_segs in all the arrays */ u32 mapsz; /* size of the map array */ + atomic_t lkey_invalid; /* true if current lkey is invalid */ u8 page_shift; /* 0 - non unform/non powerof2 sizes */ u8 lkey_published; /* in global table */ - atomic_t lkey_invalid; /* true if current lkey is invalid */ + struct percpu_ref refcount; struct completion comp; /* complete when refcount goes to zero */ - atomic_t refcount; struct rvt_segarray *map[0]; /* the segments */ }; @@ -123,13 +124,12 @@ struct rvt_sge_state { static inline void rvt_put_mr(struct rvt_mregion *mr) { - if (unlikely(atomic_dec_and_test(&mr->refcount))) - complete(&mr->comp); + percpu_ref_put(&mr->refcount); } static inline void rvt_get_mr(struct rvt_mregion *mr) { - atomic_inc(&mr->refcount); + percpu_ref_get(&mr->refcount); } static inline void rvt_put_ss(struct rvt_sge_state *ss) -- cgit v1.2.3 From b448bf9a0df6093dbadac36979a55ce4e012a677 Mon Sep 17 00:00:00 2001 From: Sebastian Sanchez Date: Wed, 8 Feb 2017 05:26:37 -0800 Subject: IB/hfi1: Allocate context data on memory node There are some memory allocation calls in hfi1_create_ctxtdata() that do not use the numa function parameter. This can cause cache lines to be filled over QPI. Reviewed-by: Mike Marciniszyn Signed-off-by: Sebastian Sanchez Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/init.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c index e3b5bc93bc70..f40864e9a3b2 100644 --- a/drivers/infiniband/hw/hfi1/init.c +++ b/drivers/infiniband/hw/hfi1/init.c @@ -297,14 +297,15 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, * The resulting value will be rounded down to the closest * multiple of dd->rcv_entries.group_size. */ - rcd->egrbufs.buffers = kcalloc(rcd->egrbufs.count, - sizeof(*rcd->egrbufs.buffers), - GFP_KERNEL); + rcd->egrbufs.buffers = kzalloc_node( + rcd->egrbufs.count * sizeof(*rcd->egrbufs.buffers), + GFP_KERNEL, numa); if (!rcd->egrbufs.buffers) goto bail; - rcd->egrbufs.rcvtids = kcalloc(rcd->egrbufs.count, - sizeof(*rcd->egrbufs.rcvtids), - GFP_KERNEL); + rcd->egrbufs.rcvtids = kzalloc_node( + rcd->egrbufs.count * + sizeof(*rcd->egrbufs.rcvtids), + GFP_KERNEL, numa); if (!rcd->egrbufs.rcvtids) goto bail; rcd->egrbufs.size = eager_buffer_size; @@ -322,8 +323,8 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, rcd->egrbufs.rcvtid_size = HFI1_MAX_EAGER_BUFFER_SIZE; if (ctxt < dd->first_user_ctxt) { /* N/A for PSM contexts */ - rcd->opstats = kzalloc(sizeof(*rcd->opstats), - GFP_KERNEL); + rcd->opstats = kzalloc_node(sizeof(*rcd->opstats), + GFP_KERNEL, numa); if (!rcd->opstats) goto bail; } -- cgit v1.2.3 From d7c76e91aa6c10775ea172263828c3119e4335a0 Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:26:43 -0800 Subject: IB/hfi1: Add additional fields to qp_stats The r_psn and s_rnr_retry are missing. Add with this patch. Reviewed-by: Kaike Wan Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/qp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index c2a166d473b2..48d59e7175c1 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -744,7 +744,7 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) wqe = rvt_get_swqe_ptr(qp, qp->s_last); send_context = qp_to_send_context(qp, priv->s_sc); seq_printf(s, - "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u PSN %x %x %x %x %x (%u %u %u %u %u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d\n", + "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u SPSN %x %x %x %x %x RPSN %x (%u %u %u %u %u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d\n", iter->n, qp_idle(qp) ? "I" : "B", qp->ibqp.qp_num, @@ -763,6 +763,7 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) qp->s_last_psn, qp->s_psn, qp->s_next_psn, qp->s_sending_psn, qp->s_sending_hpsn, + qp->r_psn, qp->s_last, qp->s_acked, qp->s_cur, qp->s_tail, qp->s_head, qp->s_size, qp->s_avail, @@ -773,6 +774,7 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) qp->s_retry, qp->s_retry_cnt, qp->s_rnr_retry_cnt, + qp->s_rnr_retry, sde, sde ? sde->this_idx : 0, send_context, -- cgit v1.2.3 From 76327627be1d9bd6a360b645ac7675c02a39096f Mon Sep 17 00:00:00 2001 From: Sebastian Sanchez Date: Wed, 8 Feb 2017 05:26:49 -0800 Subject: IB/hfi1: Reduce oversized fields in struct hfi1_packet Some fields in struct hfi1_packet are oversized. Reduce them. Reviewed-by: Mike Marciniszyn Signed-off-by: Sebastian Sanchez Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/hfi.h | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 1d0873914a6f..fcfd5f1b9017 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -356,12 +356,11 @@ struct hfi1_packet { u64 rhf; u32 maxcnt; u32 rhqoff; - u32 hdrqtail; - int numpkt; u16 tlen; - u16 hlen; s16 etail; - u16 rsize; + u8 hlen; + u8 numpkt; + u8 rsize; u8 updegr; u8 rcv_flags; u8 etype; -- cgit v1.2.3 From c03c08d50b3d44cd331319b4b2882315413ee281 Mon Sep 17 00:00:00 2001 From: Sebastian Sanchez Date: Wed, 8 Feb 2017 05:26:55 -0800 Subject: IB/hfi1: Check upper-case EFI variables The EFI variable that provides board ID is named by the PCI address of the device, which is published in upper-case, while the HFI1 driver reads the EFI variable in lower-case. This prevents returning the correct board id when queried through sysfs. Read EFI variables in upper-case if the lower-case read fails. Reviewed-by: Easwar Hariharan Signed-off-by: Sebastian Sanchez Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/efivar.c | 26 ++++++++++++++++++++++---- 1 file changed, 22 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/efivar.c b/drivers/infiniband/hw/hfi1/efivar.c index 106349fc1fb9..d106d23016ba 100644 --- a/drivers/infiniband/hw/hfi1/efivar.c +++ b/drivers/infiniband/hw/hfi1/efivar.c @@ -45,6 +45,7 @@ * */ +#include #include "efivar.h" /* GUID for HFI1 variables in EFI */ @@ -150,15 +151,32 @@ fail: int read_hfi1_efi_var(struct hfi1_devdata *dd, const char *kind, unsigned long *size, void **return_data) { + char prefix_name[64]; char name[64]; + int result; + int i; /* create a common prefix */ - snprintf(name, sizeof(name), "%04x:%02x:%02x.%x-%s", + snprintf(prefix_name, sizeof(prefix_name), "%04x:%02x:%02x.%x", pci_domain_nr(dd->pcidev->bus), dd->pcidev->bus->number, PCI_SLOT(dd->pcidev->devfn), - PCI_FUNC(dd->pcidev->devfn), - kind); + PCI_FUNC(dd->pcidev->devfn)); + snprintf(name, sizeof(name), "%s-%s", prefix_name, kind); + result = read_efi_var(name, size, return_data); + + /* + * If reading the lowercase EFI variable fail, read the uppercase + * variable. + */ + if (result) { + /* Converting to uppercase */ + for (i = 0; prefix_name[i]; i++) + if (isalpha(prefix_name[i])) + prefix_name[i] = toupper(prefix_name[i]); + snprintf(name, sizeof(name), "%s-%s", prefix_name, kind); + result = read_efi_var(name, size, return_data); + } - return read_efi_var(name, size, return_data); + return result; } -- cgit v1.2.3 From beb5a0426794c9698c4e0349c626d819b5f3b2c7 Mon Sep 17 00:00:00 2001 From: Brian Welty Date: Wed, 8 Feb 2017 05:27:01 -0800 Subject: IB/hfi1, qib, rdmavt: Move two IB event functions into rdmavt Add rvt_rc_error() and rvt_comm_est() as shared functions in rdmavt, moved from hfi1/qib logic. Reviewed-by: Dennis Dalessandro Signed-off-by: Brian Welty Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/qp.c | 13 ------------ drivers/infiniband/hw/hfi1/qp.h | 6 ------ drivers/infiniband/hw/hfi1/rc.c | 27 ++++--------------------- drivers/infiniband/hw/hfi1/ruc.c | 2 +- drivers/infiniband/hw/hfi1/uc.c | 4 ++-- drivers/infiniband/hw/hfi1/ud.c | 4 ++-- drivers/infiniband/hw/hfi1/verbs.h | 2 -- drivers/infiniband/hw/qib/qib_rc.c | 38 +++++------------------------------ drivers/infiniband/hw/qib/qib_ruc.c | 2 +- drivers/infiniband/hw/qib/qib_uc.c | 15 +++----------- drivers/infiniband/hw/qib/qib_ud.c | 4 ++-- drivers/infiniband/hw/qib/qib_verbs.h | 2 -- drivers/infiniband/sw/rdmavt/qp.c | 38 +++++++++++++++++++++++++++++++++++ include/rdma/rdmavt_qp.h | 2 ++ 14 files changed, 60 insertions(+), 99 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index 48d59e7175c1..8f66c58b3afb 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -784,19 +784,6 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) qp->pid); } -void qp_comm_est(struct rvt_qp *qp) -{ - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, gfp_t gfp) { diff --git a/drivers/infiniband/hw/hfi1/qp.h b/drivers/infiniband/hw/hfi1/qp.h index 587d84d65bb8..4012fc5e1d6e 100644 --- a/drivers/infiniband/hw/hfi1/qp.h +++ b/drivers/infiniband/hw/hfi1/qp.h @@ -131,12 +131,6 @@ int qp_iter_next(struct qp_iter *iter); */ void qp_iter_print(struct seq_file *s, struct qp_iter *iter); -/** - * qp_comm_est - handle trap with QP established - * @qp: the QP - */ -void qp_comm_est(struct rvt_qp *qp); - void _hfi1_schedule_send(struct rvt_qp *qp); void hfi1_schedule_send(struct rvt_qp *qp); diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index abea4b7c92f4..bb5271545127 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -1966,25 +1966,6 @@ send_ack: return 0; } -void hfi1_rc_error(struct rvt_qp *qp, enum ib_wc_status err) -{ - unsigned long flags; - int lastwqe; - - spin_lock_irqsave(&qp->s_lock, flags); - lastwqe = rvt_error_qp(qp, err); - spin_unlock_irqrestore(&qp->s_lock, flags); - - if (lastwqe) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_QP_LAST_WQE_REACHED; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - static inline void update_ack_queue(struct rvt_qp *qp, unsigned n) { unsigned next; @@ -2185,7 +2166,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) } if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) - qp_comm_est(qp); + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -2517,7 +2498,7 @@ rnr_nak: return; nack_op_err: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_REMOTE_OPERATIONAL_ERROR; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2527,7 +2508,7 @@ nack_op_err: nack_inv_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_inv: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_INVALID_REQUEST; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2537,7 +2518,7 @@ nack_inv: nack_acc_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_acc: - hfi1_rc_error(qp, IB_WC_LOC_PROT_ERR); + rvt_rc_error(qp, IB_WC_LOC_PROT_ERR); qp->r_nak_state = IB_NAK_REMOTE_ACCESS_ERROR; qp->r_ack_psn = qp->r_psn; send_ack: diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c index 717ed4b159d3..a1e2b6ab94fa 100644 --- a/drivers/infiniband/hw/hfi1/ruc.c +++ b/drivers/infiniband/hw/hfi1/ruc.c @@ -637,7 +637,7 @@ acc_err: wc.status = IB_WC_LOC_PROT_ERR; err: /* responder goes to error state */ - hfi1_rc_error(qp, wc.status); + rvt_rc_error(qp, wc.status); serr: spin_lock_irqsave(&sqp->s_lock, flags); diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c index 74b7b2be458c..782a0bfeedb7 100644 --- a/drivers/infiniband/hw/hfi1/uc.c +++ b/drivers/infiniband/hw/hfi1/uc.c @@ -384,7 +384,7 @@ inv: } if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) - qp_comm_est(qp); + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -584,5 +584,5 @@ drop: return; op_err: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); } diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c index 6d81d79720b0..58cd30104ec2 100644 --- a/drivers/infiniband/hw/hfi1/ud.c +++ b/drivers/infiniband/hw/hfi1/ud.c @@ -167,7 +167,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) ret = hfi1_rvt_get_rwqe(qp, 0); if (ret < 0) { - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); goto bail_unlock; } if (!ret) { @@ -796,7 +796,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) ret = hfi1_rvt_get_rwqe(qp, 0); if (ret < 0) { - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } if (!ret) { diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index e6b893010e6d..39b0074e9c43 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -327,8 +327,6 @@ void hfi1_stop_rc_timers(struct rvt_qp *qp); void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr); -void hfi1_rc_error(struct rvt_qp *qp, enum ib_wc_status err); - void hfi1_ud_rcv(struct hfi1_packet *packet); int hfi1_lookup_pkey_idx(struct hfi1_ibport *ibp, u16 pkey); diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 696bcd0d087c..5b0d01be7e50 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -1765,25 +1765,6 @@ send_ack: return 0; } -void qib_rc_error(struct rvt_qp *qp, enum ib_wc_status err) -{ - unsigned long flags; - int lastwqe; - - spin_lock_irqsave(&qp->s_lock, flags); - lastwqe = rvt_error_qp(qp, err); - spin_unlock_irqrestore(&qp->s_lock, flags); - - if (lastwqe) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_QP_LAST_WQE_REACHED; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - static inline void qib_update_ack_queue(struct rvt_qp *qp, unsigned n) { unsigned next; @@ -1895,17 +1876,8 @@ void qib_rc_rcv(struct qib_ctxtdata *rcd, struct ib_header *hdr, break; } - if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) { - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } - } + if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -2197,7 +2169,7 @@ rnr_nak: return; nack_op_err: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_REMOTE_OPERATIONAL_ERROR; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2211,7 +2183,7 @@ nack_op_err: nack_inv_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_inv: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_INVALID_REQUEST; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2225,7 +2197,7 @@ nack_inv: nack_acc_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_acc: - qib_rc_error(qp, IB_WC_LOC_PROT_ERR); + rvt_rc_error(qp, IB_WC_LOC_PROT_ERR); qp->r_nak_state = IB_NAK_REMOTE_ACCESS_ERROR; qp->r_ack_psn = qp->r_psn; send_ack: diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c index e54a2feeeb10..5e9f5e35699d 100644 --- a/drivers/infiniband/hw/qib/qib_ruc.c +++ b/drivers/infiniband/hw/qib/qib_ruc.c @@ -621,7 +621,7 @@ acc_err: wc.status = IB_WC_LOC_PROT_ERR; err: /* responder goes to error state */ - qib_rc_error(qp, wc.status); + rvt_rc_error(qp, wc.status); serr: spin_lock_irqsave(&sqp->s_lock, flags); diff --git a/drivers/infiniband/hw/qib/qib_uc.c b/drivers/infiniband/hw/qib/qib_uc.c index 5b2d483451ad..b337b60fc40d 100644 --- a/drivers/infiniband/hw/qib/qib_uc.c +++ b/drivers/infiniband/hw/qib/qib_uc.c @@ -325,17 +325,8 @@ inv: goto inv; } - if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) { - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } - } + if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -527,7 +518,7 @@ drop: return; op_err: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } diff --git a/drivers/infiniband/hw/qib/qib_ud.c b/drivers/infiniband/hw/qib/qib_ud.c index f45cad1198b0..04befa2551cd 100644 --- a/drivers/infiniband/hw/qib/qib_ud.c +++ b/drivers/infiniband/hw/qib/qib_ud.c @@ -152,7 +152,7 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) ret = qib_get_rwqe(qp, 0); if (ret < 0) { - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); goto bail_unlock; } if (!ret) { @@ -548,7 +548,7 @@ void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, ret = qib_get_rwqe(qp, 0); if (ret < 0) { - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } if (!ret) { diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index 94fd30fdedac..e399cb0fb4ee 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -326,8 +326,6 @@ void qib_rc_rnr_retry(unsigned long arg); void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr); -void qib_rc_error(struct rvt_qp *qp, enum ib_wc_status err); - int qib_post_ud_send(struct rvt_qp *qp, struct ib_send_wr *wr); void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 2a13ac660f2b..444b06cada22 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -1868,3 +1868,41 @@ int rvt_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, } return 0; } + +/** + * qp_comm_est - handle trap with QP established + * @qp: the QP + */ +void rvt_comm_est(struct rvt_qp *qp) +{ + qp->r_flags |= RVT_R_COMM_EST; + if (qp->ibqp.event_handler) { + struct ib_event ev; + + ev.device = qp->ibqp.device; + ev.element.qp = &qp->ibqp; + ev.event = IB_EVENT_COMM_EST; + qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); + } +} +EXPORT_SYMBOL(rvt_comm_est); + +void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err) +{ + unsigned long flags; + int lastwqe; + + spin_lock_irqsave(&qp->s_lock, flags); + lastwqe = rvt_error_qp(qp, err); + spin_unlock_irqrestore(&qp->s_lock, flags); + + if (lastwqe) { + struct ib_event ev; + + ev.device = qp->ibqp.device; + ev.element.qp = &qp->ibqp; + ev.event = IB_EVENT_QP_LAST_WQE_REACHED; + qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); + } +} +EXPORT_SYMBOL(rvt_rc_error); diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index f3dbd157ae5c..eaaba1b6cc57 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -607,6 +607,8 @@ static inline u32 rvt_div_mtu(struct rvt_qp *qp, u32 len) extern const int ib_rvt_state_ops[]; struct rvt_dev_info; +void rvt_comm_est(struct rvt_qp *qp); int rvt_error_qp(struct rvt_qp *qp, enum ib_wc_status err); +void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err); #endif /* DEF_RDMAVT_INCQP_H */ -- cgit v1.2.3 From 696513e8cf39ccb8ee8010fd2157f095b3af6a91 Mon Sep 17 00:00:00 2001 From: Brian Welty Date: Wed, 8 Feb 2017 05:27:07 -0800 Subject: IB/hfi1, qib, rdmavt: Move AETH credit functions into rdmavt Add rvt_compute_aeth() and rvt_get_credit() as shared functions in rdmavt, moved from hfi1/qib logic. Reviewed-by: Dennis Dalessandro Signed-off-by: Brian Welty Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/common.h | 4 - drivers/infiniband/hw/hfi1/qp.c | 137 ------------------------ drivers/infiniband/hw/hfi1/qp.h | 16 --- drivers/infiniband/hw/hfi1/rc.c | 40 +++---- drivers/infiniband/hw/hfi1/trace.c | 4 +- drivers/infiniband/hw/hfi1/verbs.h | 9 -- drivers/infiniband/hw/qib/qib_common.h | 4 - drivers/infiniband/hw/qib/qib_qp.c | 134 ----------------------- drivers/infiniband/hw/qib/qib_rc.c | 40 +++---- drivers/infiniband/hw/qib/qib_verbs.h | 4 - drivers/infiniband/sw/rdmavt/Makefile | 4 +- drivers/infiniband/sw/rdmavt/rc.c | 190 +++++++++++++++++++++++++++++++++ include/rdma/rdmavt_qp.h | 31 ++++++ 13 files changed, 265 insertions(+), 352 deletions(-) create mode 100644 drivers/infiniband/sw/rdmavt/rc.c diff --git a/drivers/infiniband/hw/hfi1/common.h b/drivers/infiniband/hw/hfi1/common.h index da7be21bedb4..1b783bbee4bb 100644 --- a/drivers/infiniband/hw/hfi1/common.h +++ b/drivers/infiniband/hw/hfi1/common.h @@ -331,10 +331,6 @@ struct diag_pkt { #define FULL_MGMT_P_KEY 0xFFFF #define DEFAULT_P_KEY LIM_MGMT_P_KEY -#define HFI1_AETH_CREDIT_SHIFT 24 -#define HFI1_AETH_CREDIT_MASK 0x1F -#define HFI1_AETH_CREDIT_INVAL 0x1F -#define HFI1_MSN_MASK 0xFFFFFF #define HFI1_FECN_SHIFT 31 #define HFI1_FECN_MASK 1 #define HFI1_FECN_SMASK BIT(HFI1_FECN_SHIFT) diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index 8f66c58b3afb..319e29656766 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -79,43 +79,6 @@ static inline unsigned mk_qpn(struct rvt_qpn_table *qpt, return (map - qpt->map) * RVT_BITS_PER_PAGE + off; } -/* - * Convert the AETH credit code into the number of credits. - */ -static const u16 credit_table[31] = { - 0, /* 0 */ - 1, /* 1 */ - 2, /* 2 */ - 3, /* 3 */ - 4, /* 4 */ - 6, /* 5 */ - 8, /* 6 */ - 12, /* 7 */ - 16, /* 8 */ - 24, /* 9 */ - 32, /* A */ - 48, /* B */ - 64, /* C */ - 96, /* D */ - 128, /* E */ - 192, /* F */ - 256, /* 10 */ - 384, /* 11 */ - 512, /* 12 */ - 768, /* 13 */ - 1024, /* 14 */ - 1536, /* 15 */ - 2048, /* 16 */ - 3072, /* 17 */ - 4096, /* 18 */ - 6144, /* 19 */ - 8192, /* 1A */ - 12288, /* 1B */ - 16384, /* 1C */ - 24576, /* 1D */ - 32768 /* 1E */ -}; - const struct rvt_operation_params hfi1_post_parms[RVT_OPERATION_MAX] = { [IB_WR_RDMA_WRITE] = { .length = sizeof(struct ib_rdma_wr), @@ -339,68 +302,6 @@ int hfi1_check_send_wqe(struct rvt_qp *qp, return wqe->length <= piothreshold; } -/** - * hfi1_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 hfi1_compute_aeth(struct rvt_qp *qp) -{ - u32 aeth = qp->r_msn & HFI1_MSN_MASK; - - if (qp->ibqp.srq) { - /* - * Shared receive queues don't generate credits. - * Set the credit field to the invalid value. - */ - aeth |= HFI1_AETH_CREDIT_INVAL << HFI1_AETH_CREDIT_SHIFT; - } else { - u32 min, max, x; - u32 credits; - struct rvt_rwq *wq = qp->r_rq.wq; - u32 head; - u32 tail; - - /* sanity check pointers before trusting them */ - head = wq->head; - if (head >= qp->r_rq.size) - head = 0; - tail = wq->tail; - if (tail >= qp->r_rq.size) - tail = 0; - /* - * Compute the number of credits available (RWQEs). - * There is a small chance that the pair of reads are - * not atomic, which is OK, since the fuzziness is - * resolved as further ACKs go out. - */ - credits = head - tail; - if ((int)credits < 0) - credits += qp->r_rq.size; - /* - * Binary search the credit table to find the code to - * use. - */ - min = 0; - max = 31; - for (;;) { - x = (min + max) / 2; - if (credit_table[x] == credits) - break; - if (credit_table[x] > credits) { - max = x; - } else { - if (min == x) - break; - min = x; - } - } - aeth |= x << HFI1_AETH_CREDIT_SHIFT; - } - return cpu_to_be32(aeth); -} - /** * _hfi1_schedule_send - schedule progress * @qp: the QP @@ -457,44 +358,6 @@ void hfi1_schedule_send(struct rvt_qp *qp) _hfi1_schedule_send(qp); } -/** - * hfi1_get_credit - handle credit in aeth - * @qp: the qp - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void hfi1_get_credit(struct rvt_qp *qp, u32 aeth) -{ - u32 credit = (aeth >> HFI1_AETH_CREDIT_SHIFT) & HFI1_AETH_CREDIT_MASK; - - lockdep_assert_held(&qp->s_lock); - /* - * If the credit is invalid, we can send - * as many packets as we like. Otherwise, we have to - * honor the credit field. - */ - if (credit == HFI1_AETH_CREDIT_INVAL) { - if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - qp->s_flags |= RVT_S_UNLIMITED_CREDIT; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - hfi1_schedule_send(qp); - } - } - } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - /* Compute new LSN (i.e., MSN + credit) */ - credit = (aeth + credit_table[credit]) & HFI1_MSN_MASK; - if (cmp_msn(credit, qp->s_lsn) > 0) { - qp->s_lsn = credit; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - hfi1_schedule_send(qp); - } - } - } -} - void hfi1_qp_wakeup(struct rvt_qp *qp, u32 flag) { unsigned long flags; diff --git a/drivers/infiniband/hw/hfi1/qp.h b/drivers/infiniband/hw/hfi1/qp.h index 4012fc5e1d6e..1eb9cd7b8c19 100644 --- a/drivers/infiniband/hw/hfi1/qp.h +++ b/drivers/infiniband/hw/hfi1/qp.h @@ -70,14 +70,6 @@ static inline void clear_ahg(struct rvt_qp *qp) qp->s_ahgidx = -1; } -/** - * hfi1_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 hfi1_compute_aeth(struct rvt_qp *qp); - /** * hfi1_create_qp - create a queue pair for a device * @ibpd: the protection domain who's device we create the queue pair for @@ -91,14 +83,6 @@ __be32 hfi1_compute_aeth(struct rvt_qp *qp); struct ib_qp *hfi1_create_qp(struct ib_pd *ibpd, struct ib_qp_init_attr *init_attr, struct ib_udata *udata); -/** - * hfi1_get_credit - flush the send work queue of a QP - * @qp: the qp who's send work queue to flush - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void hfi1_get_credit(struct rvt_qp *qp, u32 aeth); /** * hfi1_qp_wakeup - wake up on the indicated event diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index bb5271545127..b04a90327a35 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -284,7 +284,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, qp->s_ack_state = OP(RDMA_READ_RESPONSE_ONLY); e->sent = 1; } - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_rdma_psn = e->psn; bth2 = mask_psn(qp->s_ack_rdma_psn++); @@ -293,7 +293,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, ps->s_txreq->ss = NULL; len = 0; qp->s_ack_state = OP(ATOMIC_ACKNOWLEDGE); - ohdr->u.at.aeth = hfi1_compute_aeth(qp); + ohdr->u.at.aeth = rvt_compute_aeth(qp); ib_u64_put(e->atomic_data, &ohdr->u.at.atomic_ack_eth); hwords += sizeof(ohdr->u.at) / sizeof(u32); bth2 = mask_psn(e->psn); @@ -315,7 +315,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, len = pmtu; middle = HFI1_CAP_IS_KSET(SDMA_AHG); } else { - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_state = OP(RDMA_READ_RESPONSE_LAST); e = &qp->s_ack_queue[qp->s_tail_ack_queue]; @@ -338,11 +338,11 @@ normal: ps->s_txreq->ss = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & HFI1_MSN_MASK) | + cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | (qp->s_nak_state << - HFI1_AETH_CREDIT_SHIFT)); + RVT_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; len = 0; bth0 = OP(ACKNOWLEDGE) << 24; @@ -519,7 +519,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) case IB_WR_SEND_WITH_INV: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -556,7 +556,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) case IB_WR_RDMA_WRITE_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -885,11 +885,11 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & HFI1_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | (qp->r_nak_state << - HFI1_AETH_CREDIT_SHIFT)); + RVT_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; /* set PBC_DC_INFO bit (aka SC[4]) in pbc_flags */ pbc_flags |= ((!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT); @@ -1323,7 +1323,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> 29) + if (aeth >> RVT_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); ibp = rcd_to_iport(rcd); @@ -1403,7 +1403,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> 29) { + switch (aeth >> RVT_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1430,7 +1430,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_flags &= ~RVT_S_WAIT_ACK; hfi1_schedule_send(qp); } - hfi1_get_credit(qp, aeth); + rvt_get_credit(qp, aeth); qp->s_rnr_retry = qp->s_rnr_retry_cnt; qp->s_retry = qp->s_retry_cnt; update_last_psn(qp, psn); @@ -1459,8 +1459,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_flags &= ~(RVT_S_WAIT_SSN_CREDIT | RVT_S_WAIT_ACK); hfi1_stop_rc_timers(qp); to = - ib_hfi1_rnr_table[(aeth >> HFI1_AETH_CREDIT_SHIFT) & - HFI1_AETH_CREDIT_MASK]; + ib_hfi1_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & + RVT_AETH_CREDIT_MASK]; hfi1_add_rnr_timer(qp, to); return 0; @@ -1469,8 +1469,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, goto bail_stop; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> HFI1_AETH_CREDIT_SHIFT) & - HFI1_AETH_CREDIT_MASK) { + switch ((aeth >> RVT_AETH_CREDIT_SHIFT) & + RVT_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1600,8 +1600,8 @@ static void rc_rcv_resp(struct hfi1_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> 29) == 0) - hfi1_get_credit(qp, aeth); + if ((aeth >> RVT_AETH_NAK_SHIFT) == 0) + rvt_get_credit(qp, aeth); } goto ack_done; } diff --git a/drivers/infiniband/hw/hfi1/trace.c b/drivers/infiniband/hw/hfi1/trace.c index 01f525cd985a..0985b4f1b9ef 100644 --- a/drivers/infiniband/hw/hfi1/trace.c +++ b/drivers/infiniband/hw/hfi1/trace.c @@ -130,14 +130,14 @@ const char *parse_everbs_hdrs( case OP(RC, ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN, be32_to_cpu(eh->aeth) >> 24, parse_syndrome(be32_to_cpu(eh->aeth) >> 24), - be32_to_cpu(eh->aeth) & HFI1_MSN_MASK); + be32_to_cpu(eh->aeth) & RVT_MSN_MASK); break; /* aeth + atomicacketh */ case OP(RC, ATOMIC_ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN " " ATOMICACKETH_PRN, be32_to_cpu(eh->at.aeth) >> 24, parse_syndrome(be32_to_cpu(eh->at.aeth) >> 24), - be32_to_cpu(eh->at.aeth) & HFI1_MSN_MASK, + be32_to_cpu(eh->at.aeth) & RVT_MSN_MASK, ib_u64_get(&eh->at.atomic_ack_eth)); break; /* atomiceth */ diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index 39b0074e9c43..f359684d0b46 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -259,15 +259,6 @@ int hfi1_process_mad(struct ib_device *ibdev, int mad_flags, u8 port, #endif #define PSN_MODIFY_MASK 0xFFFFFF -/* - * Compare the lower 24 bits of the msn values. - * Returns an integer <, ==, or > than zero. - */ -static inline int cmp_msn(u32 a, u32 b) -{ - return (((int)a) - ((int)b)) << 8; -} - /* * Compare two PSNs * Returns an integer <, ==, or > than zero. diff --git a/drivers/infiniband/hw/qib/qib_common.h b/drivers/infiniband/hw/qib/qib_common.h index 1d6e63eb1146..a4a1f56ce824 100644 --- a/drivers/infiniband/hw/qib/qib_common.h +++ b/drivers/infiniband/hw/qib/qib_common.h @@ -742,11 +742,7 @@ struct qib_tid_session_member { #define SIZE_OF_CRC 1 #define QIB_DEFAULT_P_KEY 0xFFFF -#define QIB_AETH_CREDIT_SHIFT 24 -#define QIB_AETH_CREDIT_MASK 0x1F -#define QIB_AETH_CREDIT_INVAL 0x1F #define QIB_PSN_MASK 0xFFFFFF -#define QIB_MSN_MASK 0xFFFFFF #define QIB_EAGER_TID_ID QLOGIC_IB_I_TID_MASK #define QIB_MULTICAST_QPN 0xFFFFFF diff --git a/drivers/infiniband/hw/qib/qib_qp.c b/drivers/infiniband/hw/qib/qib_qp.c index 99d31efe4c2f..e869be01ae86 100644 --- a/drivers/infiniband/hw/qib/qib_qp.c +++ b/drivers/infiniband/hw/qib/qib_qp.c @@ -61,43 +61,6 @@ static inline unsigned find_next_offset(struct rvt_qpn_table *qpt, return off; } -/* - * Convert the AETH credit code into the number of credits. - */ -static u32 credit_table[31] = { - 0, /* 0 */ - 1, /* 1 */ - 2, /* 2 */ - 3, /* 3 */ - 4, /* 4 */ - 6, /* 5 */ - 8, /* 6 */ - 12, /* 7 */ - 16, /* 8 */ - 24, /* 9 */ - 32, /* A */ - 48, /* B */ - 64, /* C */ - 96, /* D */ - 128, /* E */ - 192, /* F */ - 256, /* 10 */ - 384, /* 11 */ - 512, /* 12 */ - 768, /* 13 */ - 1024, /* 14 */ - 1536, /* 15 */ - 2048, /* 16 */ - 3072, /* 17 */ - 4096, /* 18 */ - 6144, /* 19 */ - 8192, /* 1A */ - 12288, /* 1B */ - 16384, /* 1C */ - 24576, /* 1D */ - 32768 /* 1E */ -}; - const struct rvt_operation_params qib_post_parms[RVT_OPERATION_MAX] = { [IB_WR_RDMA_WRITE] = { .length = sizeof(struct ib_rdma_wr), @@ -354,66 +317,6 @@ u32 qib_mtu_from_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, u32 pmtu) return ib_mtu_enum_to_int(pmtu); } -/** - * qib_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 qib_compute_aeth(struct rvt_qp *qp) -{ - u32 aeth = qp->r_msn & QIB_MSN_MASK; - - if (qp->ibqp.srq) { - /* - * Shared receive queues don't generate credits. - * Set the credit field to the invalid value. - */ - aeth |= QIB_AETH_CREDIT_INVAL << QIB_AETH_CREDIT_SHIFT; - } else { - u32 min, max, x; - u32 credits; - struct rvt_rwq *wq = qp->r_rq.wq; - u32 head; - u32 tail; - - /* sanity check pointers before trusting them */ - head = wq->head; - if (head >= qp->r_rq.size) - head = 0; - tail = wq->tail; - if (tail >= qp->r_rq.size) - tail = 0; - /* - * Compute the number of credits available (RWQEs). - * XXX Not holding the r_rq.lock here so there is a small - * chance that the pair of reads are not atomic. - */ - credits = head - tail; - if ((int)credits < 0) - credits += qp->r_rq.size; - /* - * Binary search the credit table to find the code to - * use. - */ - min = 0; - max = 31; - for (;;) { - x = (min + max) / 2; - if (credit_table[x] == credits) - break; - if (credit_table[x] > credits) - max = x; - else if (min == x) - break; - else - min = x; - } - aeth |= x << QIB_AETH_CREDIT_SHIFT; - } - return cpu_to_be32(aeth); -} - void *qib_qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, gfp_t gfp) { struct qib_qp_priv *priv; @@ -473,43 +376,6 @@ void qib_flush_qp_waiters(struct rvt_qp *qp) spin_unlock(&dev->rdi.pending_lock); } -/** - * qib_get_credit - flush the send work queue of a QP - * @qp: the qp who's send work queue to flush - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void qib_get_credit(struct rvt_qp *qp, u32 aeth) -{ - u32 credit = (aeth >> QIB_AETH_CREDIT_SHIFT) & QIB_AETH_CREDIT_MASK; - - /* - * If the credit is invalid, we can send - * as many packets as we like. Otherwise, we have to - * honor the credit field. - */ - if (credit == QIB_AETH_CREDIT_INVAL) { - if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - qp->s_flags |= RVT_S_UNLIMITED_CREDIT; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - qib_schedule_send(qp); - } - } - } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - /* Compute new LSN (i.e., MSN + credit) */ - credit = (aeth + credit_table[credit]) & QIB_MSN_MASK; - if (qib_cmp24(credit, qp->s_lsn) > 0) { - qp->s_lsn = credit; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - qib_schedule_send(qp); - } - } - } -} - /** * qib_check_send_wqe - validate wr/wqe * @qp - The qp diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 5b0d01be7e50..6a189310e12d 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -144,7 +144,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, qp->s_ack_state = OP(RDMA_READ_RESPONSE_ONLY); e->sent = 1; } - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_rdma_psn = e->psn; bth2 = qp->s_ack_rdma_psn++ & QIB_PSN_MASK; @@ -153,7 +153,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, qp->s_cur_sge = NULL; len = 0; qp->s_ack_state = OP(ATOMIC_ACKNOWLEDGE); - ohdr->u.at.aeth = qib_compute_aeth(qp); + ohdr->u.at.aeth = rvt_compute_aeth(qp); ib_u64_put(e->atomic_data, &ohdr->u.at.atomic_ack_eth); hwords += sizeof(ohdr->u.at) / sizeof(u32); bth2 = e->psn & QIB_PSN_MASK; @@ -174,7 +174,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, if (len > pmtu) len = pmtu; else { - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_state = OP(RDMA_READ_RESPONSE_LAST); e = &qp->s_ack_queue[qp->s_tail_ack_queue]; @@ -197,11 +197,11 @@ normal: qp->s_cur_sge = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & QIB_MSN_MASK) | + cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | (qp->s_nak_state << - QIB_AETH_CREDIT_SHIFT)); + RVT_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; len = 0; bth0 = OP(ACKNOWLEDGE) << 24; @@ -331,7 +331,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) case IB_WR_SEND_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - qib_cmp24(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -362,7 +362,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) case IB_WR_RDMA_WRITE_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - qib_cmp24(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -658,11 +658,11 @@ void qib_send_rc_ack(struct rvt_qp *qp) if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & QIB_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | (qp->r_nak_state << - QIB_AETH_CREDIT_SHIFT)); + RVT_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); lrh0 |= ibp->sl_to_vl[qp->remote_ah_attr.sl] << 12 | qp->remote_ah_attr.sl << 4; hdr.lrh[0] = cpu_to_be16(lrh0); @@ -1098,7 +1098,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> 29) + if (aeth >> RVT_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); ibp = to_iport(qp->ibqp.device, qp->port_num); @@ -1178,7 +1178,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> 29) { + switch (aeth >> RVT_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1201,7 +1201,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_flags &= ~RVT_S_WAIT_ACK; qib_schedule_send(qp); } - qib_get_credit(qp, aeth); + rvt_get_credit(qp, aeth); qp->s_rnr_retry = qp->s_rnr_retry_cnt; qp->s_retry = qp->s_retry_cnt; update_last_psn(qp, psn); @@ -1232,8 +1232,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_flags |= RVT_S_WAIT_RNR; qp->s_timer.function = qib_rc_rnr_retry; qp->s_timer.expires = jiffies + usecs_to_jiffies( - ib_qib_rnr_table[(aeth >> QIB_AETH_CREDIT_SHIFT) & - QIB_AETH_CREDIT_MASK]); + ib_qib_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & + RVT_AETH_CREDIT_MASK]); add_timer(&qp->s_timer); goto bail; @@ -1242,8 +1242,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, goto bail; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> QIB_AETH_CREDIT_SHIFT) & - QIB_AETH_CREDIT_MASK) { + switch ((aeth >> RVT_AETH_CREDIT_SHIFT) & + RVT_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1400,8 +1400,8 @@ static void qib_rc_rcv_resp(struct qib_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> 29) == 0) - qib_get_credit(qp, aeth); + if ((aeth >> RVT_AETH_NAK_SHIFT) == 0) + rvt_get_credit(qp, aeth); } goto ack_done; } diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index e399cb0fb4ee..226e81fe39d6 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -270,8 +270,6 @@ int qib_snapshot_counters(struct qib_pportdata *ppd, u64 *swords, int qib_get_counters(struct qib_pportdata *ppd, struct qib_verbs_counters *cntrs); -__be32 qib_compute_aeth(struct rvt_qp *qp); - /* * Functions provided by qib driver for rdmavt to use */ @@ -294,8 +292,6 @@ void qib_qp_iter_print(struct seq_file *s, struct qib_qp_iter *iter); #endif -void qib_get_credit(struct rvt_qp *qp, u32 aeth); - unsigned qib_pkt_delay(u32 plen, u8 snd_mult, u8 rcv_mult); void qib_verbs_sdma_desc_avail(struct qib_pportdata *ppd, unsigned avail); diff --git a/drivers/infiniband/sw/rdmavt/Makefile b/drivers/infiniband/sw/rdmavt/Makefile index ccaa7992ac97..c33a4f84413c 100644 --- a/drivers/infiniband/sw/rdmavt/Makefile +++ b/drivers/infiniband/sw/rdmavt/Makefile @@ -7,7 +7,7 @@ # obj-$(CONFIG_INFINIBAND_RDMAVT) += rdmavt.o -rdmavt-y := vt.o ah.o cq.o dma.o mad.o mcast.o mmap.o mr.o pd.o qp.o srq.o \ - trace.o +rdmavt-y := vt.o ah.o cq.o dma.o mad.o mcast.o mmap.o mr.o pd.o qp.o \ + rc.o srq.o trace.o CFLAGS_trace.o = -I$(src) diff --git a/drivers/infiniband/sw/rdmavt/rc.c b/drivers/infiniband/sw/rdmavt/rc.c new file mode 100644 index 000000000000..b32b5a3853ac --- /dev/null +++ b/drivers/infiniband/sw/rdmavt/rc.c @@ -0,0 +1,190 @@ +/* + * Copyright(c) 2016 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +#include + +#define RVT_AETH_CREDIT_INVAL RVT_AETH_CREDIT_MASK + +/* + * Convert the AETH credit code into the number of credits. + */ +static const u16 credit_table[31] = { + 0, /* 0 */ + 1, /* 1 */ + 2, /* 2 */ + 3, /* 3 */ + 4, /* 4 */ + 6, /* 5 */ + 8, /* 6 */ + 12, /* 7 */ + 16, /* 8 */ + 24, /* 9 */ + 32, /* A */ + 48, /* B */ + 64, /* C */ + 96, /* D */ + 128, /* E */ + 192, /* F */ + 256, /* 10 */ + 384, /* 11 */ + 512, /* 12 */ + 768, /* 13 */ + 1024, /* 14 */ + 1536, /* 15 */ + 2048, /* 16 */ + 3072, /* 17 */ + 4096, /* 18 */ + 6144, /* 19 */ + 8192, /* 1A */ + 12288, /* 1B */ + 16384, /* 1C */ + 24576, /* 1D */ + 32768 /* 1E */ +}; + +/** + * rvt_compute_aeth - compute the AETH (syndrome + MSN) + * @qp: the queue pair to compute the AETH for + * + * Returns the AETH. + */ +__be32 rvt_compute_aeth(struct rvt_qp *qp) +{ + u32 aeth = qp->r_msn & RVT_MSN_MASK; + + if (qp->ibqp.srq) { + /* + * Shared receive queues don't generate credits. + * Set the credit field to the invalid value. + */ + aeth |= RVT_AETH_CREDIT_INVAL << RVT_AETH_CREDIT_SHIFT; + } else { + u32 min, max, x; + u32 credits; + struct rvt_rwq *wq = qp->r_rq.wq; + u32 head; + u32 tail; + + /* sanity check pointers before trusting them */ + head = wq->head; + if (head >= qp->r_rq.size) + head = 0; + tail = wq->tail; + if (tail >= qp->r_rq.size) + tail = 0; + /* + * Compute the number of credits available (RWQEs). + * There is a small chance that the pair of reads are + * not atomic, which is OK, since the fuzziness is + * resolved as further ACKs go out. + */ + credits = head - tail; + if ((int)credits < 0) + credits += qp->r_rq.size; + /* + * Binary search the credit table to find the code to + * use. + */ + min = 0; + max = 31; + for (;;) { + x = (min + max) / 2; + if (credit_table[x] == credits) + break; + if (credit_table[x] > credits) { + max = x; + } else { + if (min == x) + break; + min = x; + } + } + aeth |= x << RVT_AETH_CREDIT_SHIFT; + } + return cpu_to_be32(aeth); +} +EXPORT_SYMBOL(rvt_compute_aeth); + +/** + * rvt_get_credit - flush the send work queue of a QP + * @qp: the qp who's send work queue to flush + * @aeth: the Acknowledge Extended Transport Header + * + * The QP s_lock should be held. + */ +void rvt_get_credit(struct rvt_qp *qp, u32 aeth) +{ + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + u32 credit = (aeth >> RVT_AETH_CREDIT_SHIFT) & RVT_AETH_CREDIT_MASK; + + lockdep_assert_held(&qp->s_lock); + /* + * If the credit is invalid, we can send + * as many packets as we like. Otherwise, we have to + * honor the credit field. + */ + if (credit == RVT_AETH_CREDIT_INVAL) { + if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { + qp->s_flags |= RVT_S_UNLIMITED_CREDIT; + if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { + qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; + rdi->driver_f.schedule_send(qp); + } + } + } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { + /* Compute new LSN (i.e., MSN + credit) */ + credit = (aeth + credit_table[credit]) & RVT_MSN_MASK; + if (rvt_cmp_msn(credit, qp->s_lsn) > 0) { + qp->s_lsn = credit; + if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { + qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; + rdi->driver_f.schedule_send(qp); + } + } + } +} +EXPORT_SYMBOL(rvt_get_credit); diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index eaaba1b6cc57..a92e7dcfb5f5 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -582,6 +582,37 @@ static inline void rvt_qp_swqe_complete( } } +#define RVT_AETH_CREDIT_SHIFT 24 +#define RVT_AETH_CREDIT_MASK 0x1F +#define RVT_AETH_NAK_SHIFT 29 +#define RVT_MSN_MASK 0xFFFFFF + +/* + * Compare the lower 24 bits of the msn values. + * Returns an integer <, ==, or > than zero. + */ +static inline int rvt_cmp_msn(u32 a, u32 b) +{ + return (((int)a) - ((int)b)) << 8; +} + +/** + * rvt_compute_aeth - compute the AETH (syndrome + MSN) + * @qp: the queue pair to compute the AETH for + * + * Returns the AETH. + */ +__be32 rvt_compute_aeth(struct rvt_qp *qp); + +/** + * rvt_get_credit - flush the send work queue of a QP + * @qp: the qp who's send work queue to flush + * @aeth: the Acknowledge Extended Transport Header + * + * The QP s_lock should be held. + */ +void rvt_get_credit(struct rvt_qp *qp, u32 aeth); + /** * @qp - the qp pair * @len - the length -- cgit v1.2.3 From 11a10d4bc7b2640da1fce27586a617411b70f5c5 Mon Sep 17 00:00:00 2001 From: Venkata Sandeep Dhanalakota Date: Wed, 8 Feb 2017 05:27:13 -0800 Subject: IB/rdmavt: Adding timer logic to rdmavt To move common code across target to rdmavt for code reuse. Reviewed-by: Mike Marciniszyn Reviewed-by: Brian Welty Signed-off-by: Venkata Sandeep Dhanalakota Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rdmavt/qp.c | 183 +++++++++++++++++++++++++++++++++++++- include/rdma/rdma_vt.h | 19 ++++ include/rdma/rdmavt_qp.h | 6 ++ 3 files changed, 206 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 444b06cada22..db4315f8f0c8 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -55,6 +55,46 @@ #include "vt.h" #include "trace.h" +static void rvt_rc_timeout(unsigned long arg); + +/* + * Convert the AETH RNR timeout code into the number of microseconds. + */ +static const u32 ib_rvt_rnr_table[32] = { + 655360, /* 00: 655.36 */ + 10, /* 01: .01 */ + 20, /* 02 .02 */ + 30, /* 03: .03 */ + 40, /* 04: .04 */ + 60, /* 05: .06 */ + 80, /* 06: .08 */ + 120, /* 07: .12 */ + 160, /* 08: .16 */ + 240, /* 09: .24 */ + 320, /* 0A: .32 */ + 480, /* 0B: .48 */ + 640, /* 0C: .64 */ + 960, /* 0D: .96 */ + 1280, /* 0E: 1.28 */ + 1920, /* 0F: 1.92 */ + 2560, /* 10: 2.56 */ + 3840, /* 11: 3.84 */ + 5120, /* 12: 5.12 */ + 7680, /* 13: 7.68 */ + 10240, /* 14: 10.24 */ + 15360, /* 15: 15.36 */ + 20480, /* 16: 20.48 */ + 30720, /* 17: 30.72 */ + 40960, /* 18: 40.96 */ + 61440, /* 19: 61.44 */ + 81920, /* 1A: 81.92 */ + 122880, /* 1B: 122.88 */ + 163840, /* 1C: 163.84 */ + 245760, /* 1D: 245.76 */ + 327680, /* 1E: 327.68 */ + 491520 /* 1F: 491.52 */ +}; + /* * Note that it is OK to post send work requests in the SQE and ERR * states; rvt_do_send() will process them and generate error @@ -200,7 +240,8 @@ int rvt_driver_qp_init(struct rvt_dev_info *rdi) if (!rdi->driver_f.free_all_qps || !rdi->driver_f.qp_priv_alloc || !rdi->driver_f.qp_priv_free || - !rdi->driver_f.notify_qp_reset) + !rdi->driver_f.notify_qp_reset || + !rdi->driver_f.notify_restart_rc) return -EINVAL; /* allocate parent object */ @@ -587,6 +628,7 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, /* Let drivers flush their waitlist */ rdi->driver_f.flush_qp_waiters(qp); + rvt_stop_rc_timers(qp); qp->s_flags &= ~(RVT_S_TIMER | RVT_S_ANY_WAIT); spin_unlock(&qp->s_lock); spin_unlock(&qp->s_hlock); @@ -594,7 +636,7 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, /* Stop the send queue and the retry timer */ rdi->driver_f.stop_send_queue(qp); - + rvt_del_timers_sync(qp); /* Wait for things to stop */ rdi->driver_f.quiesce_qp(qp); @@ -730,6 +772,11 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, if (!qp->s_ack_queue) goto bail_qp; } + /* initialize timers needed for rc qp */ + setup_timer(&qp->s_timer, rvt_rc_timeout, (unsigned long)qp); + hrtimer_init(&qp->s_rnr_timer, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); + qp->s_rnr_timer.function = rvt_rc_rnr_retry; /* * Driver needs to set up it's private QP structure and do any @@ -1906,3 +1953,135 @@ void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err) } } EXPORT_SYMBOL(rvt_rc_error); + +static inline unsigned long rvt_aeth_to_usec(u32 aeth) +{ + return ib_rvt_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & + RVT_AETH_CREDIT_MASK]; +} + +/* + * rvt_add_retry_timer - add/start a retry timer + * @qp - the QP + * add a retry timer on the QP + */ +void rvt_add_retry_timer(struct rvt_qp *qp) +{ + struct ib_qp *ibqp = &qp->ibqp; + struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); + + lockdep_assert_held(&qp->s_lock); + qp->s_flags |= RVT_S_TIMER; + /* 4.096 usec. * (1 << qp->timeout) */ + qp->s_timer.expires = jiffies + qp->timeout_jiffies + + rdi->busy_jiffies; + add_timer(&qp->s_timer); +} +EXPORT_SYMBOL(rvt_add_retry_timer); + +/** + * rvt_add_rnr_timer - add/start an rnr timer + * @qp - the QP + * @aeth - aeth of RNR timeout, simulated aeth for loopback + * add an rnr timer on the QP + */ +void rvt_add_rnr_timer(struct rvt_qp *qp, u32 aeth) +{ + u32 to; + + lockdep_assert_held(&qp->s_lock); + qp->s_flags |= RVT_S_WAIT_RNR; + to = rvt_aeth_to_usec(aeth); + hrtimer_start(&qp->s_rnr_timer, + ns_to_ktime(1000 * to), HRTIMER_MODE_REL); +} +EXPORT_SYMBOL(rvt_add_rnr_timer); + +/** + * rvt_stop_rc_timers - stop all timers + * @qp - the QP + * stop any pending timers + */ +void rvt_stop_rc_timers(struct rvt_qp *qp) +{ + lockdep_assert_held(&qp->s_lock); + /* Remove QP from all timers */ + if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { + qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); + del_timer(&qp->s_timer); + hrtimer_try_to_cancel(&qp->s_rnr_timer); + } +} +EXPORT_SYMBOL(rvt_stop_rc_timers); + +/** + * rvt_stop_rnr_timer - stop an rnr timer + * @qp - the QP + * + * stop an rnr timer and return if the timer + * had been pending. + */ +static int rvt_stop_rnr_timer(struct rvt_qp *qp) +{ + int rval = 0; + + lockdep_assert_held(&qp->s_lock); + /* Remove QP from rnr timer */ + if (qp->s_flags & RVT_S_WAIT_RNR) { + qp->s_flags &= ~RVT_S_WAIT_RNR; + rval = hrtimer_try_to_cancel(&qp->s_rnr_timer); + } + return rval; +} + +/** + * rvt_del_timers_sync - wait for any timeout routines to exit + * @qp - the QP + */ +void rvt_del_timers_sync(struct rvt_qp *qp) +{ + del_timer_sync(&qp->s_timer); + hrtimer_cancel(&qp->s_rnr_timer); +} +EXPORT_SYMBOL(rvt_del_timers_sync); + +/** + * This is called from s_timer for missing responses. + */ +static void rvt_rc_timeout(unsigned long arg) +{ + struct rvt_qp *qp = (struct rvt_qp *)arg; + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + unsigned long flags; + + spin_lock_irqsave(&qp->r_lock, flags); + spin_lock(&qp->s_lock); + if (qp->s_flags & RVT_S_TIMER) { + qp->s_flags &= ~RVT_S_TIMER; + del_timer(&qp->s_timer); + if (rdi->driver_f.notify_restart_rc) + rdi->driver_f.notify_restart_rc(qp, + qp->s_last_psn + 1, + 1); + rdi->driver_f.schedule_send(qp); + } + spin_unlock(&qp->s_lock); + spin_unlock_irqrestore(&qp->r_lock, flags); +} + +/* + * This is called from s_timer for RNR timeouts. + */ +enum hrtimer_restart rvt_rc_rnr_retry(struct hrtimer *t) +{ + struct rvt_qp *qp = container_of(t, struct rvt_qp, s_rnr_timer); + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + unsigned long flags; + + spin_lock_irqsave(&qp->s_lock, flags); + rvt_stop_rnr_timer(qp); + rdi->driver_f.schedule_send(qp); + spin_unlock_irqrestore(&qp->s_lock, flags); + return HRTIMER_NORESTART; +} +EXPORT_SYMBOL(rvt_rc_rnr_retry); diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h index 861e23eaebda..a69dee38365f 100644 --- a/include/rdma/rdma_vt.h +++ b/include/rdma/rdma_vt.h @@ -335,6 +335,8 @@ struct rvt_driver_provided { /* Notify driver a mad agent has been removed */ void (*notify_free_mad_agent)(struct rvt_dev_info *rdi, int port_idx); + /* Notify driver to restart rc */ + void (*notify_restart_rc)(struct rvt_qp *qp, u32 psn, int wait); }; struct rvt_dev_info { @@ -483,6 +485,23 @@ static inline struct rvt_qp *rvt_lookup_qpn(struct rvt_dev_info *rdi, return qp; } +/** + * rvt_mod_retry_timer - mod a retry timer + * @qp - the QP + * Modify a potentially already running retry timer + */ +static inline void rvt_mod_retry_timer(struct rvt_qp *qp) +{ + struct ib_qp *ibqp = &qp->ibqp; + struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); + + lockdep_assert_held(&qp->s_lock); + qp->s_flags |= RVT_S_TIMER; + /* 4.096 usec. * (1 << qp->timeout) */ + mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies + + rdi->busy_jiffies); +} + struct rvt_dev_info *rvt_alloc_device(size_t size, int nports); void rvt_dealloc_device(struct rvt_dev_info *rdi); int rvt_register_device(struct rvt_dev_info *rvd); diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index a92e7dcfb5f5..0b1cbffb967c 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -370,6 +370,7 @@ struct rvt_qp { struct rvt_sge_state s_ack_rdma_sge; struct timer_list s_timer; + struct hrtimer s_rnr_timer; atomic_t local_ops_pending; /* number of fast_reg/local_inv reqs */ @@ -641,5 +642,10 @@ struct rvt_dev_info; void rvt_comm_est(struct rvt_qp *qp); int rvt_error_qp(struct rvt_qp *qp, enum ib_wc_status err); void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err); +enum hrtimer_restart rvt_rc_rnr_retry(struct hrtimer *t); +void rvt_add_rnr_timer(struct rvt_qp *qp, u32 aeth); +void rvt_del_timers_sync(struct rvt_qp *qp); +void rvt_stop_rc_timers(struct rvt_qp *qp); +void rvt_add_retry_timer(struct rvt_qp *qp); #endif /* DEF_RDMAVT_INCQP_H */ -- cgit v1.2.3 From 56acbbfb46d9d613858de2bffcc7dee4205682b7 Mon Sep 17 00:00:00 2001 From: Venkata Sandeep Dhanalakota Date: Wed, 8 Feb 2017 05:27:19 -0800 Subject: IB/hfi1: Use new rdmavt timers Reduce hfi1 code footprint by using the rdmavt timers. Reviewed-by: Mike Marciniszyn Reviewed-by: Brian Welty Signed-off-by: Venkata Sandeep Dhanalakota Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/qp.c | 4 - drivers/infiniband/hw/hfi1/rc.c | 194 +++---------------------------------- drivers/infiniband/hw/hfi1/ruc.c | 43 +------- drivers/infiniband/hw/hfi1/verbs.c | 1 + drivers/infiniband/hw/hfi1/verbs.h | 9 +- 5 files changed, 16 insertions(+), 235 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index 319e29656766..c4ebd097b20f 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -671,8 +671,6 @@ void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, iowait_sleep, iowait_wakeup, iowait_sdma_drained); - setup_timer(&priv->s_rnr_timer, hfi1_rc_rnr_retry, (unsigned long)qp); - qp->s_timer.function = hfi1_rc_timeout; return priv; } @@ -713,7 +711,6 @@ void flush_qp_waiters(struct rvt_qp *qp) { lockdep_assert_held(&qp->s_lock); flush_iowait(qp); - hfi1_stop_rc_timers(qp); } void stop_send_queue(struct rvt_qp *qp) @@ -721,7 +718,6 @@ void stop_send_queue(struct rvt_qp *qp) struct hfi1_qp_priv *priv = qp->priv; cancel_work_sync(&priv->s_iowait.iowork); - hfi1_del_timers_sync(qp); } void quiesce_qp(struct rvt_qp *qp) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index b04a90327a35..c2f1a6f576fc 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -57,133 +57,6 @@ /* cut down ridiculously long IB macro names */ #define OP(x) RC_OP(x) -/** - * hfi1_add_retry_timer - add/start a retry timer - * @qp - the QP - * - * add a retry timer on the QP - */ -static inline void hfi1_add_retry_timer(struct rvt_qp *qp) -{ - struct ib_qp *ibqp = &qp->ibqp; - struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_TIMER; - /* 4.096 usec. * (1 << qp->timeout) */ - qp->s_timer.expires = jiffies + qp->timeout_jiffies + - rdi->busy_jiffies; - add_timer(&qp->s_timer); -} - -/** - * hfi1_add_rnr_timer - add/start an rnr timer - * @qp - the QP - * @to - timeout in usecs - * - * add an rnr timer on the QP - */ -void hfi1_add_rnr_timer(struct rvt_qp *qp, u32 to) -{ - struct hfi1_qp_priv *priv = qp->priv; - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_WAIT_RNR; - priv->s_rnr_timer.expires = jiffies + usecs_to_jiffies(to); - add_timer(&priv->s_rnr_timer); -} - -/** - * hfi1_mod_retry_timer - mod a retry timer - * @qp - the QP - * - * Modify a potentially already running retry - * timer - */ -static inline void hfi1_mod_retry_timer(struct rvt_qp *qp) -{ - struct ib_qp *ibqp = &qp->ibqp; - struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_TIMER; - /* 4.096 usec. * (1 << qp->timeout) */ - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies + - rdi->busy_jiffies); -} - -/** - * hfi1_stop_retry_timer - stop a retry timer - * @qp - the QP - * - * stop a retry timer and return if the timer - * had been pending. - */ -static inline int hfi1_stop_retry_timer(struct rvt_qp *qp) -{ - int rval = 0; - - lockdep_assert_held(&qp->s_lock); - /* Remove QP from retry */ - if (qp->s_flags & RVT_S_TIMER) { - qp->s_flags &= ~RVT_S_TIMER; - rval = del_timer(&qp->s_timer); - } - return rval; -} - -/** - * hfi1_stop_rc_timers - stop all timers - * @qp - the QP - * - * stop any pending timers - */ -void hfi1_stop_rc_timers(struct rvt_qp *qp) -{ - struct hfi1_qp_priv *priv = qp->priv; - - lockdep_assert_held(&qp->s_lock); - /* Remove QP from all timers */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - del_timer(&priv->s_rnr_timer); - } -} - -/** - * hfi1_stop_rnr_timer - stop an rnr timer - * @qp - the QP - * - * stop an rnr timer and return if the timer - * had been pending. - */ -static inline int hfi1_stop_rnr_timer(struct rvt_qp *qp) -{ - int rval = 0; - struct hfi1_qp_priv *priv = qp->priv; - - lockdep_assert_held(&qp->s_lock); - /* Remove QP from rnr timer */ - if (qp->s_flags & RVT_S_WAIT_RNR) { - qp->s_flags &= ~RVT_S_WAIT_RNR; - rval = del_timer(&priv->s_rnr_timer); - } - return rval; -} - -/** - * hfi1_del_timers_sync - wait for any timeout routines to exit - * @qp - the QP - */ -void hfi1_del_timers_sync(struct rvt_qp *qp) -{ - struct hfi1_qp_priv *priv = qp->priv; - - del_timer_sync(&qp->s_timer); - del_timer_sync(&priv->s_rnr_timer); -} - static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, u32 psn, u32 pmtu) { @@ -1043,7 +916,7 @@ done: * Back up requester to resend the last un-ACKed request. * The QP r_lock and s_lock should be held and interrupts disabled. */ -static void restart_rc(struct rvt_qp *qp, u32 psn, int wait) +void hfi1_restart_rc(struct rvt_qp *qp, u32 psn, int wait) { struct rvt_swqe *wqe = rvt_get_swqe_ptr(qp, qp->s_acked); struct hfi1_ibport *ibp; @@ -1079,44 +952,6 @@ static void restart_rc(struct rvt_qp *qp, u32 psn, int wait) reset_psn(qp, psn); } -/* - * This is called from s_timer for missing responses. - */ -void hfi1_rc_timeout(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - struct hfi1_ibport *ibp; - unsigned long flags; - - spin_lock_irqsave(&qp->r_lock, flags); - spin_lock(&qp->s_lock); - if (qp->s_flags & RVT_S_TIMER) { - ibp = to_iport(qp->ibqp.device, qp->port_num); - ibp->rvp.n_rc_timeouts++; - qp->s_flags &= ~RVT_S_TIMER; - del_timer(&qp->s_timer); - trace_hfi1_timeout(qp, qp->s_last_psn + 1); - restart_rc(qp, qp->s_last_psn + 1, 1); - hfi1_schedule_send(qp); - } - spin_unlock(&qp->s_lock); - spin_unlock_irqrestore(&qp->r_lock, flags); -} - -/* - * This is called from s_timer for RNR timeouts. - */ -void hfi1_rc_rnr_retry(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - unsigned long flags; - - spin_lock_irqsave(&qp->s_lock, flags); - hfi1_stop_rnr_timer(qp); - hfi1_schedule_send(qp); - spin_unlock_irqrestore(&qp->s_lock, flags); -} - /* * Set qp->s_sending_psn to the next PSN after the given one. * This would be psn+1 except when RDMA reads are present. @@ -1183,7 +1018,7 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) !(qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR | RVT_S_WAIT_PSN)) && (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK)) - hfi1_add_retry_timer(qp); + rvt_add_retry_timer(qp); while (qp->s_last != qp->s_acked) { u32 s_last; @@ -1313,7 +1148,6 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, int ret = 0; u32 ack_psn; int diff; - unsigned long to; lockdep_assert_held(&qp->s_lock); /* @@ -1362,7 +1196,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, /* Retry this request. */ if (!(qp->r_flags & RVT_R_RDMAR_SEQ)) { qp->r_flags |= RVT_R_RDMAR_SEQ; - restart_rc(qp, qp->s_last_psn + 1, 0); + hfi1_restart_rc(qp, qp->s_last_psn + 1, 0); if (list_empty(&qp->rspwait)) { qp->r_flags |= RVT_R_RSP_SEND; rvt_get_qp(qp); @@ -1411,7 +1245,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * We are expecting more ACKs so * mod the retry timer. */ - hfi1_mod_retry_timer(qp); + rvt_mod_retry_timer(qp); /* * We can stop re-sending the earlier packets and * continue with the next packet the receiver wants. @@ -1420,7 +1254,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn + 1); } else { /* No more acks - kill all timers */ - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); if (cmp_psn(qp->s_psn, psn) <= 0) { qp->s_state = OP(SEND_LAST); qp->s_psn = psn + 1; @@ -1457,11 +1291,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn); qp->s_flags &= ~(RVT_S_WAIT_SSN_CREDIT | RVT_S_WAIT_ACK); - hfi1_stop_rc_timers(qp); - to = - ib_hfi1_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & - RVT_AETH_CREDIT_MASK]; - hfi1_add_rnr_timer(qp, to); + rvt_stop_rc_timers(qp); + rvt_add_rnr_timer(qp, aeth); return 0; case 3: /* NAK */ @@ -1479,7 +1310,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * RDMA READ response which terminates the RDMA * READ. */ - restart_rc(qp, psn, 0); + hfi1_restart_rc(qp, psn, 0); hfi1_schedule_send(qp); break; @@ -1518,7 +1349,7 @@ reserved: } /* cannot be reached */ bail_stop: - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); return ret; } @@ -1533,7 +1364,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct hfi1_ibport *ibp, u32 psn, lockdep_assert_held(&qp->s_lock); /* Remove QP from retry timer */ - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); wqe = rvt_get_swqe_ptr(qp, qp->s_acked); @@ -1547,7 +1378,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct hfi1_ibport *ibp, u32 psn, ibp->rvp.n_rdma_seq++; qp->r_flags |= RVT_R_RDMAR_SEQ; - restart_rc(qp, qp->s_last_psn + 1, 0); + hfi1_restart_rc(qp, qp->s_last_psn + 1, 0); if (list_empty(&qp->rspwait)) { qp->r_flags |= RVT_R_RSP_SEND; rvt_get_qp(qp); @@ -1661,8 +1492,7 @@ read_middle: * We got a response so update the timeout. * 4.096 usec. * (1 << qp->timeout) */ - qp->s_flags |= RVT_S_TIMER; - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies); + rvt_mod_retry_timer(qp); if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; hfi1_schedule_send(qp); diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c index a1e2b6ab94fa..dc41b2c5628a 100644 --- a/drivers/infiniband/hw/hfi1/ruc.c +++ b/drivers/infiniband/hw/hfi1/ruc.c @@ -53,44 +53,6 @@ #include "verbs_txreq.h" #include "trace.h" -/* - * Convert the AETH RNR timeout code into the number of microseconds. - */ -const u32 ib_hfi1_rnr_table[32] = { - 655360, /* 00: 655.36 */ - 10, /* 01: .01 */ - 20, /* 02 .02 */ - 30, /* 03: .03 */ - 40, /* 04: .04 */ - 60, /* 05: .06 */ - 80, /* 06: .08 */ - 120, /* 07: .12 */ - 160, /* 08: .16 */ - 240, /* 09: .24 */ - 320, /* 0A: .32 */ - 480, /* 0B: .48 */ - 640, /* 0C: .64 */ - 960, /* 0D: .96 */ - 1280, /* 0E: 1.28 */ - 1920, /* 0F: 1.92 */ - 2560, /* 10: 2.56 */ - 3840, /* 11: 3.84 */ - 5120, /* 12: 5.12 */ - 7680, /* 13: 7.68 */ - 10240, /* 14: 10.24 */ - 15360, /* 15: 15.36 */ - 20480, /* 16: 20.48 */ - 30720, /* 17: 30.72 */ - 40960, /* 18: 40.96 */ - 61440, /* 19: 61.44 */ - 81920, /* 1A: 81.92 */ - 122880, /* 1B: 122.88 */ - 163840, /* 1C: 163.84 */ - 245760, /* 1D: 245.76 */ - 327680, /* 1E: 327.68 */ - 491520 /* 1F: 491.52 */ -}; - /* * Validate a RWQE and fill in the SGE state. * Return 1 if OK. @@ -361,7 +323,6 @@ static void ruc_loopback(struct rvt_qp *sqp) int release; int ret; int copy_last = 0; - u32 to; int local_ops = 0; rcu_read_lock(); @@ -618,8 +579,8 @@ rnr_nak: spin_lock_irqsave(&sqp->s_lock, flags); if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; - to = ib_hfi1_rnr_table[qp->r_min_rnr_timer]; - hfi1_add_rnr_timer(sqp, to); + rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << + RVT_AETH_CREDIT_SHIFT); goto clr_busy; op_err: diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index b937a23efc75..cae32304eaba 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -1845,6 +1845,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) dd->verbs_dev.rdi.driver_f.mtu_to_path_mtu = mtu_to_path_mtu; dd->verbs_dev.rdi.driver_f.check_modify_qp = hfi1_check_modify_qp; dd->verbs_dev.rdi.driver_f.modify_qp = hfi1_modify_qp; + dd->verbs_dev.rdi.driver_f.notify_restart_rc = hfi1_restart_rc; dd->verbs_dev.rdi.driver_f.check_send_wqe = hfi1_check_send_wqe; /* completeion queue */ diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index f359684d0b46..6e9e4bbf8ae4 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -127,7 +127,6 @@ struct hfi1_qp_priv { u8 s_sc; /* SC[0..4] for next packet */ u8 r_adefered; /* number of acks defered */ struct iowait s_iowait; - struct timer_list s_rnr_timer; struct rvt_qp *owner; }; @@ -310,12 +309,6 @@ u8 ah_to_sc(struct ib_device *ibdev, struct ib_ah_attr *ah_attr); struct ib_ah *hfi1_create_qp0_ah(struct hfi1_ibport *ibp, u16 dlid); -void hfi1_rc_rnr_retry(unsigned long arg); -void hfi1_add_rnr_timer(struct rvt_qp *qp, u32 to); -void hfi1_rc_timeout(unsigned long arg); -void hfi1_del_timers_sync(struct rvt_qp *qp); -void hfi1_stop_rc_timers(struct rvt_qp *qp); - void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr); void hfi1_ud_rcv(struct hfi1_packet *packet); @@ -331,7 +324,7 @@ int hfi1_check_modify_qp(struct rvt_qp *qp, struct ib_qp_attr *attr, void hfi1_modify_qp(struct rvt_qp *qp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata); - +void hfi1_restart_rc(struct rvt_qp *qp, u32 psn, int wait); int hfi1_check_send_wqe(struct rvt_qp *qp, struct rvt_swqe *wqe); extern const u32 rc_only_opcode; -- cgit v1.2.3 From b4238e70579cb1edf32c56ee512f84cbad01ac27 Mon Sep 17 00:00:00 2001 From: Venkata Sandeep Dhanalakota Date: Wed, 8 Feb 2017 05:27:25 -0800 Subject: IB/qib: Use new rdmavt timers Reduce qib code footprint by using the rdmavt timers. Reviewed-by: Mike Marciniszyn Reviewed-by: Brian Welty Signed-off-by: Venkata Sandeep Dhanalakota Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qib/qib_qp.c | 1 - drivers/infiniband/hw/qib/qib_rc.c | 94 +++++++---------------------------- drivers/infiniband/hw/qib/qib_ruc.c | 45 +---------------- drivers/infiniband/hw/qib/qib_verbs.c | 1 + drivers/infiniband/hw/qib/qib_verbs.h | 2 +- 5 files changed, 21 insertions(+), 122 deletions(-) diff --git a/drivers/infiniband/hw/qib/qib_qp.c b/drivers/infiniband/hw/qib/qib_qp.c index e869be01ae86..2ac0c0f79e74 100644 --- a/drivers/infiniband/hw/qib/qib_qp.c +++ b/drivers/infiniband/hw/qib/qib_qp.c @@ -351,7 +351,6 @@ void qib_stop_send_queue(struct rvt_qp *qp) struct qib_qp_priv *priv = qp->priv; cancel_work_sync(&priv->s_work); - del_timer_sync(&qp->s_timer); } void qib_quiesce_qp(struct rvt_qp *qp) diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 6a189310e12d..ea0bc6bc09fa 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -38,7 +38,6 @@ /* cut down ridiculously long IB macro names */ #define OP(x) IB_OPCODE_RC_##x -static void rc_timeout(unsigned long arg); static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, u32 psn, u32 pmtu) @@ -54,15 +53,6 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, return wqe->length - len; } -static void start_timer(struct rvt_qp *qp) -{ - qp->s_flags |= RVT_S_TIMER; - qp->s_timer.function = rc_timeout; - /* 4.096 usec. * (1 << qp->timeout) */ - qp->s_timer.expires = jiffies + qp->timeout_jiffies; - add_timer(&qp->s_timer); -} - /** * qib_make_rc_ack - construct a response packet (ACK, NAK, or RDMA read) * @dev: the device for this QP @@ -837,7 +827,7 @@ done: * Back up requester to resend the last un-ACKed request. * The QP r_lock and s_lock should be held and interrupts disabled. */ -static void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) +void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) { struct rvt_swqe *wqe = rvt_get_swqe_ptr(qp, qp->s_acked); struct qib_ibport *ibp; @@ -869,46 +859,6 @@ static void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) reset_psn(qp, psn); } -/* - * This is called from s_timer for missing responses. - */ -static void rc_timeout(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - struct qib_ibport *ibp; - unsigned long flags; - - spin_lock_irqsave(&qp->r_lock, flags); - spin_lock(&qp->s_lock); - if (qp->s_flags & RVT_S_TIMER) { - ibp = to_iport(qp->ibqp.device, qp->port_num); - ibp->rvp.n_rc_timeouts++; - qp->s_flags &= ~RVT_S_TIMER; - del_timer(&qp->s_timer); - qib_restart_rc(qp, qp->s_last_psn + 1, 1); - qib_schedule_send(qp); - } - spin_unlock(&qp->s_lock); - spin_unlock_irqrestore(&qp->r_lock, flags); -} - -/* - * This is called from s_timer for RNR timeouts. - */ -void qib_rc_rnr_retry(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - unsigned long flags; - - spin_lock_irqsave(&qp->s_lock, flags); - if (qp->s_flags & RVT_S_WAIT_RNR) { - qp->s_flags &= ~RVT_S_WAIT_RNR; - del_timer(&qp->s_timer); - qib_schedule_send(qp); - } - spin_unlock_irqrestore(&qp->s_lock, flags); -} - /* * Set qp->s_sending_psn to the next PSN after the given one. * This would be psn+1 except when RDMA reads are present. @@ -972,7 +922,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) if ((psn & IB_BTH_REQ_ACK) && qp->s_acked != qp->s_tail && !(qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR | RVT_S_WAIT_PSN)) && (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK)) - start_timer(qp); + rvt_add_retry_timer(qp); while (qp->s_last != qp->s_acked) { u32 s_last; @@ -1085,12 +1035,6 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, u32 ack_psn; int diff; - /* Remove QP from retry timer */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - } - /* * Note that NAKs implicitly ACK outstanding SEND and RDMA write * requests and implicitly NAK RDMA read and atomic requests issued @@ -1186,16 +1130,20 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * We are expecting more ACKs so * reset the retransmit timer. */ - start_timer(qp); + rvt_mod_retry_timer(qp); /* * We can stop resending the earlier packets and * continue with the next packet the receiver wants. */ if (qib_cmp24(qp->s_psn, psn) <= 0) reset_psn(qp, psn + 1); - } else if (qib_cmp24(qp->s_psn, psn) <= 0) { - qp->s_state = OP(SEND_LAST); - qp->s_psn = psn + 1; + } else { + /* No more acks - kill all timers */ + rvt_stop_rc_timers(qp); + if (qib_cmp24(qp->s_psn, psn) <= 0) { + qp->s_state = OP(SEND_LAST); + qp->s_psn = psn + 1; + } } if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; @@ -1205,8 +1153,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_rnr_retry = qp->s_rnr_retry_cnt; qp->s_retry = qp->s_retry_cnt; update_last_psn(qp, psn); - ret = 1; - goto bail; + return 1; case 1: /* RNR NAK */ ibp->rvp.n_rnr_naks++; @@ -1229,13 +1176,9 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn); qp->s_flags &= ~(RVT_S_WAIT_SSN_CREDIT | RVT_S_WAIT_ACK); - qp->s_flags |= RVT_S_WAIT_RNR; - qp->s_timer.function = qib_rc_rnr_retry; - qp->s_timer.expires = jiffies + usecs_to_jiffies( - ib_qib_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & - RVT_AETH_CREDIT_MASK]); - add_timer(&qp->s_timer); - goto bail; + rvt_stop_rc_timers(qp); + rvt_add_rnr_timer(qp, aeth); + return 0; case 3: /* NAK */ if (qp->s_acked == qp->s_tail) @@ -1291,6 +1234,7 @@ reserved: } bail: + rvt_stop_rc_timers(qp); return ret; } @@ -1304,10 +1248,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct qib_ibport *ibp, u32 psn, struct rvt_swqe *wqe; /* Remove QP from retry timer */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - } + rvt_stop_rc_timers(qp); wqe = rvt_get_swqe_ptr(qp, qp->s_acked); @@ -1462,8 +1403,7 @@ read_middle: * We got a response so update the timeout. * 4.096 usec. * (1 << qp->timeout) */ - qp->s_flags |= RVT_S_TIMER; - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies); + rvt_mod_retry_timer(qp); if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; qib_schedule_send(qp); diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c index 5e9f5e35699d..58b28412121d 100644 --- a/drivers/infiniband/hw/qib/qib_ruc.c +++ b/drivers/infiniband/hw/qib/qib_ruc.c @@ -37,44 +37,6 @@ #include "qib.h" #include "qib_mad.h" -/* - * Convert the AETH RNR timeout code into the number of microseconds. - */ -const u32 ib_qib_rnr_table[32] = { - 655360, /* 00: 655.36 */ - 10, /* 01: .01 */ - 20, /* 02 .02 */ - 30, /* 03: .03 */ - 40, /* 04: .04 */ - 60, /* 05: .06 */ - 80, /* 06: .08 */ - 120, /* 07: .12 */ - 160, /* 08: .16 */ - 240, /* 09: .24 */ - 320, /* 0A: .32 */ - 480, /* 0B: .48 */ - 640, /* 0C: .64 */ - 960, /* 0D: .96 */ - 1280, /* 0E: 1.28 */ - 1920, /* 0F: 1.92 */ - 2560, /* 10: 2.56 */ - 3840, /* 11: 3.84 */ - 5120, /* 12: 5.12 */ - 7680, /* 13: 7.68 */ - 10240, /* 14: 10.24 */ - 15360, /* 15: 15.36 */ - 20480, /* 16: 20.48 */ - 30720, /* 17: 30.72 */ - 40960, /* 18: 40.96 */ - 61440, /* 19: 61.44 */ - 81920, /* 1A: 81.92 */ - 122880, /* 1B: 122.88 */ - 163840, /* 1C: 163.84 */ - 245760, /* 1D: 245.76 */ - 327680, /* 1E: 327.68 */ - 491520 /* 1F: 491.52 */ -}; - /* * Validate a RWQE and fill in the SGE state. * Return 1 if OK. @@ -599,11 +561,8 @@ rnr_nak: spin_lock_irqsave(&sqp->s_lock, flags); if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; - sqp->s_flags |= RVT_S_WAIT_RNR; - sqp->s_timer.function = qib_rc_rnr_retry; - sqp->s_timer.expires = jiffies + - usecs_to_jiffies(ib_qib_rnr_table[qp->r_min_rnr_timer]); - add_timer(&sqp->s_timer); + rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << + RVT_AETH_CREDIT_SHIFT); goto clr_busy; op_err: diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c index 4b54c0ddd08a..ee7741e55817 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.c +++ b/drivers/infiniband/hw/qib/qib_verbs.c @@ -1659,6 +1659,7 @@ int qib_register_ib_device(struct qib_devdata *dd) dd->verbs_dev.rdi.driver_f.stop_send_queue = qib_stop_send_queue; dd->verbs_dev.rdi.driver_f.flush_qp_waiters = qib_flush_qp_waiters; dd->verbs_dev.rdi.driver_f.notify_error_qp = qib_notify_error_qp; + dd->verbs_dev.rdi.driver_f.notify_restart_rc = qib_restart_rc; dd->verbs_dev.rdi.driver_f.mtu_to_path_mtu = qib_mtu_to_path_mtu; dd->verbs_dev.rdi.driver_f.mtu_from_qp = qib_mtu_from_qp; dd->verbs_dev.rdi.driver_f.get_pmtu_from_attr = qib_get_pmtu_from_attr; diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index 226e81fe39d6..1bdfa1ceb563 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -279,7 +279,7 @@ void qib_qp_priv_free(struct rvt_dev_info *rdi, struct rvt_qp *qp); void qib_notify_qp_reset(struct rvt_qp *qp); int qib_alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, enum ib_qp_type type, u8 port, gfp_t gfp); - +void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait); #ifdef CONFIG_DEBUG_FS struct qib_qp_iter; -- cgit v1.2.3 From 0128fceaf934dbfca4537d4eb8c3a5f7e84562c8 Mon Sep 17 00:00:00 2001 From: Brian Welty Date: Wed, 8 Feb 2017 05:27:31 -0800 Subject: IB/hfi1, rdmavt: Update copy_sge to use boolean arguments Convert copy_sge and related SGE state functions to use boolean. For determining if QP is in user mode, add helper function in rdmavt_qp.h. This is used to determine if QP needs the last byte ordering. While here, change rvt_pd.user to a boolean. Reviewed-by: Mike Marciniszyn Reviewed-by: Dean Luick Signed-off-by: Brian Welty Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/rc.c | 16 ++++++++-------- drivers/infiniband/hw/hfi1/ruc.c | 10 +++++----- drivers/infiniband/hw/hfi1/uc.c | 10 +++++----- drivers/infiniband/hw/hfi1/ud.c | 12 ++++++------ drivers/infiniband/hw/hfi1/verbs.c | 21 +++++++++++---------- drivers/infiniband/hw/hfi1/verbs.h | 4 ++-- drivers/infiniband/sw/rdmavt/pd.c | 2 +- include/rdma/rdma_vt.h | 2 +- include/rdma/rdmavt_qp.h | 9 +++++++++ 9 files changed, 48 insertions(+), 38 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index c2f1a6f576fc..b1c350dd3e8b 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -67,7 +67,7 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, ss->sg_list = wqe->sg_list + 1; ss->num_sge = wqe->wr.num_sge; ss->total_len = wqe->length; - hfi1_skip_sge(ss, len, 0); + hfi1_skip_sge(ss, len, false); return wqe->length - len; } @@ -1508,7 +1508,7 @@ read_middle: qp->s_rdma_read_len -= pmtu; update_last_psn(qp, psn); spin_unlock_irqrestore(&qp->s_lock, flags); - hfi1_copy_sge(&qp->s_rdma_read_sge, data, pmtu, 0, 0); + hfi1_copy_sge(&qp->s_rdma_read_sge, data, pmtu, false, false); goto bail; case OP(RDMA_READ_RESPONSE_ONLY): @@ -1552,7 +1552,7 @@ read_last: if (unlikely(tlen != qp->s_rdma_read_len)) goto ack_len_err; aeth = be32_to_cpu(ohdr->u.aeth); - hfi1_copy_sge(&qp->s_rdma_read_sge, data, tlen, 0, 0); + hfi1_copy_sge(&qp->s_rdma_read_sge, data, tlen, false, false); WARN_ON(qp->s_rdma_read_sge.num_sge); (void)do_rc_ack(qp, aeth, psn, OP(RDMA_READ_RESPONSE_LAST), 0, rcd); @@ -1923,7 +1923,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) struct ib_reth *reth; unsigned long flags; int ret, is_fecn = 0; - int copy_last = 0; + bool copy_last = false; u32 rkey; lockdep_assert_held(&qp->r_lock); @@ -2017,7 +2017,7 @@ send_middle: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto nack_inv; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, true, false); break; case OP(RDMA_WRITE_LAST_WITH_IMMEDIATE): @@ -2057,7 +2057,7 @@ send_last_inv: wc.wc_flags = IB_WC_WITH_INVALIDATE; goto send_last; case OP(RDMA_WRITE_LAST): - copy_last = ibpd_to_rvtpd(qp->ibqp.pd)->user; + copy_last = rvt_is_user_qp(qp); /* fall through */ case OP(SEND_LAST): no_immediate_data: @@ -2075,7 +2075,7 @@ send_last: wc.byte_len = tlen + qp->r_rcv_len; if (unlikely(wc.byte_len > qp->r_len)) goto nack_inv; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, copy_last); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, copy_last); rvt_put_ss(&qp->r_sge); qp->r_msn++; if (!__test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) @@ -2113,7 +2113,7 @@ send_last: break; case OP(RDMA_WRITE_ONLY): - copy_last = 1; + copy_last = rvt_is_user_qp(qp); /* fall through */ case OP(RDMA_WRITE_FIRST): case OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE): diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c index dc41b2c5628a..bdf3697de881 100644 --- a/drivers/infiniband/hw/hfi1/ruc.c +++ b/drivers/infiniband/hw/hfi1/ruc.c @@ -320,9 +320,9 @@ static void ruc_loopback(struct rvt_qp *sqp) u64 sdata; atomic64_t *maddr; enum ib_wc_status send_status; - int release; + bool release; int ret; - int copy_last = 0; + bool copy_last = false; int local_ops = 0; rcu_read_lock(); @@ -386,7 +386,7 @@ again: memset(&wc, 0, sizeof(wc)); send_status = IB_WC_SUCCESS; - release = 1; + release = true; sqp->s_sge.sge = wqe->sg_list[0]; sqp->s_sge.sg_list = wqe->sg_list + 1; sqp->s_sge.num_sge = wqe->wr.num_sge; @@ -437,7 +437,7 @@ send: /* skip copy_last set and qp_access_flags recheck */ goto do_write; case IB_WR_RDMA_WRITE: - copy_last = ibpd_to_rvtpd(qp->ibqp.pd)->user; + copy_last = rvt_is_user_qp(qp); if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_WRITE))) goto inv_err; do_write: @@ -461,7 +461,7 @@ do_write: wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_READ))) goto acc_err; - release = 0; + release = false; sqp->s_sge.sg_list = NULL; sqp->s_sge.num_sge = 1; qp->r_sge.sge = wqe->sg_list[0]; diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c index 782a0bfeedb7..4b2a8400c823 100644 --- a/drivers/infiniband/hw/hfi1/uc.c +++ b/drivers/infiniband/hw/hfi1/uc.c @@ -419,7 +419,7 @@ send_first: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto rewind; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 0, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, false, false); break; case OP(SEND_LAST_WITH_IMMEDIATE): @@ -444,7 +444,7 @@ send_last: if (unlikely(wc.byte_len > qp->r_len)) goto rewind; wc.opcode = IB_WC_RECV; - hfi1_copy_sge(&qp->r_sge, data, tlen, 0, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, false, false); rvt_put_ss(&qp->s_rdma_read_sge); last_imm: wc.wr_id = qp->r_wr_id; @@ -519,7 +519,7 @@ rdma_first: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto drop; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, true, false); break; case OP(RDMA_WRITE_LAST_WITH_IMMEDIATE): @@ -548,7 +548,7 @@ rdma_last_imm: } wc.byte_len = qp->r_len; wc.opcode = IB_WC_RECV_RDMA_WITH_IMM; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, false); rvt_put_ss(&qp->r_sge); goto last_imm; @@ -564,7 +564,7 @@ rdma_last: tlen -= (hdrsize + pad + 4); if (unlikely(tlen + qp->r_rcv_len != qp->r_len)) goto drop; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, false); rvt_put_ss(&qp->r_sge); break; diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c index 58cd30104ec2..9329c16a1f6f 100644 --- a/drivers/infiniband/hw/hfi1/ud.c +++ b/drivers/infiniband/hw/hfi1/ud.c @@ -189,10 +189,10 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) hfi1_make_grh(ibp, &grh, &grd, 0, 0); hfi1_copy_sge(&qp->r_sge, &grh, - sizeof(grh), 1, 0); + sizeof(grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } ssge.sg_list = swqe->sg_list + 1; ssge.sge = *swqe->sg_list; @@ -206,7 +206,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) if (len > sge->sge_length) len = sge->sge_length; WARN_ON_ONCE(len == 0); - hfi1_copy_sge(&qp->r_sge, sge->vaddr, len, 1, 0); + hfi1_copy_sge(&qp->r_sge, sge->vaddr, len, true, false); sge->vaddr += len; sge->length -= len; sge->sge_length -= len; @@ -812,13 +812,13 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) } if (has_grh) { hfi1_copy_sge(&qp->r_sge, &hdr->u.l.grh, - sizeof(struct ib_grh), 1, 0); + sizeof(struct ib_grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } hfi1_copy_sge(&qp->r_sge, data, wc.byte_len - sizeof(struct ib_grh), - 1, 0); + true, false); rvt_put_ss(&qp->r_sge); if (!test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) return; diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index cae32304eaba..b10c2dcb4ba5 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -291,7 +291,7 @@ static void wss_insert(void *address) /* * Is the working set larger than the threshold? */ -static inline int wss_exceeds_threshold(void) +static inline bool wss_exceeds_threshold(void) { return atomic_read(&wss.total_count) >= wss.threshold; } @@ -419,18 +419,19 @@ __be64 ib_hfi1_sys_image_guid; * @ss: the SGE state * @data: the data to copy * @length: the length of the data + * @release: boolean to release MR * @copy_last: do a separate copy of the last 8 bytes */ void hfi1_copy_sge( struct rvt_sge_state *ss, void *data, u32 length, - int release, - int copy_last) + bool release, + bool copy_last) { struct rvt_sge *sge = &ss->sge; - int in_last = 0; int i; - int cacheless_copy = 0; + bool in_last = false; + bool cacheless_copy = false; if (sge_copy_mode == COPY_CACHELESS) { cacheless_copy = length >= PAGE_SIZE; @@ -454,8 +455,8 @@ void hfi1_copy_sge( if (length > 8) { length -= 8; } else { - copy_last = 0; - in_last = 1; + copy_last = false; + in_last = true; } } @@ -501,8 +502,8 @@ again: } if (copy_last) { - copy_last = 0; - in_last = 1; + copy_last = false; + in_last = true; length = 8; goto again; } @@ -513,7 +514,7 @@ again: * @ss: the SGE state * @length: the number of bytes to skip */ -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, int release) +void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, bool release) { struct rvt_sge *sge = &ss->sge; diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index 6e9e4bbf8ae4..9fab6b3d12f6 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -289,9 +289,9 @@ void hfi1_put_txreq(struct verbs_txreq *tx); int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps); void hfi1_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, - int release, int copy_last); + bool release, bool copy_last); -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, int release); +void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, bool release); void hfi1_cnp_rcv(struct hfi1_packet *packet); diff --git a/drivers/infiniband/sw/rdmavt/pd.c b/drivers/infiniband/sw/rdmavt/pd.c index d1292f324c67..8a89afff3363 100644 --- a/drivers/infiniband/sw/rdmavt/pd.c +++ b/drivers/infiniband/sw/rdmavt/pd.c @@ -90,7 +90,7 @@ struct ib_pd *rvt_alloc_pd(struct ib_device *ibdev, spin_unlock(&dev->n_pds_lock); /* ib_alloc_pd() will initialize pd->ibpd. */ - pd->user = udata ? 1 : 0; + pd->user = !!udata; ret = &pd->ibpd; diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h index a69dee38365f..8fc1ca7b6f23 100644 --- a/include/rdma/rdma_vt.h +++ b/include/rdma/rdma_vt.h @@ -164,7 +164,7 @@ struct rvt_driver_params { /* Protection domain */ struct rvt_pd { struct ib_pd ibpd; - int user; /* non-zero if created from user space */ + bool user; }; /* Address handle */ diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index 0b1cbffb967c..561b6c876086 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -467,6 +467,15 @@ static inline struct rvt_rwqe *rvt_get_rwqe_ptr(struct rvt_rq *rq, unsigned n) rq->max_sge * sizeof(struct ib_sge)) * n); } +/** + * rvt_is_user_qp - return if this is user mode QP + * @qp - the target QP + */ +static inline bool rvt_is_user_qp(struct rvt_qp *qp) +{ + return !!qp->pid; +} + /** * rvt_get_qp - get a QP reference * @qp - the QP to hold -- cgit v1.2.3 From 1198fcea8a7810d06069a85256d1ef9ec97d434b Mon Sep 17 00:00:00 2001 From: Brian Welty Date: Wed, 8 Feb 2017 05:27:37 -0800 Subject: IB/hfi1, rdmavt: Move SGE state helper routines into rdmavt To improve code reuse, add small SGE state helper routines to rdmavt_mr.h. Leverage these in hfi1, including refactoring of hfi1_copy_sge. Reviewed-by: Mike Marciniszyn Signed-off-by: Brian Welty Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/hfi.h | 2 - drivers/infiniband/hw/hfi1/rc.c | 2 +- drivers/infiniband/hw/hfi1/ud.c | 4 +- drivers/infiniband/hw/hfi1/verbs.c | 91 ++------------------------------------ drivers/infiniband/hw/hfi1/verbs.h | 2 - include/rdma/rdmavt_mr.h | 50 +++++++++++++++++++++ 6 files changed, 57 insertions(+), 94 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index fcfd5f1b9017..9c3be919c0b7 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1797,8 +1797,6 @@ int kdeth_process_expected(struct hfi1_packet *packet); int kdeth_process_eager(struct hfi1_packet *packet); int process_receive_invalid(struct hfi1_packet *packet); -void update_sge(struct rvt_sge_state *ss, u32 length); - /* global module parameter variables */ extern unsigned int hfi1_max_mtu; extern unsigned int hfi1_cu; diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index b1c350dd3e8b..85e7bf6dcc11 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -67,7 +67,7 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, ss->sg_list = wqe->sg_list + 1; ss->num_sge = wqe->wr.num_sge; ss->total_len = wqe->length; - hfi1_skip_sge(ss, len, false); + rvt_skip_sge(ss, len, false); return wqe->length - len; } diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c index 9329c16a1f6f..13ea4eb6ef3d 100644 --- a/drivers/infiniband/hw/hfi1/ud.c +++ b/drivers/infiniband/hw/hfi1/ud.c @@ -192,7 +192,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) sizeof(grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } ssge.sg_list = swqe->sg_list + 1; ssge.sge = *swqe->sg_list; @@ -815,7 +815,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) sizeof(struct ib_grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } hfi1_copy_sge(&qp->r_sge, data, wc.byte_len - sizeof(struct ib_grh), true, false); diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index b10c2dcb4ba5..72f459e1fdde 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -462,12 +462,8 @@ void hfi1_copy_sge( again: while (length) { - u32 len = sge->length; + u32 len = rvt_get_sge_length(sge, length); - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; WARN_ON_ONCE(len == 0); if (unlikely(in_last)) { /* enforce byte transfer ordering */ @@ -478,25 +474,7 @@ again: } else { memcpy(sge->vaddr, data, len); } - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } + rvt_update_sge(ss, len, release); data += len; length -= len; } @@ -509,46 +487,6 @@ again: } } -/** - * hfi1_skip_sge - skip over SGE memory - * @ss: the SGE state - * @length: the number of bytes to skip - */ -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, bool release) -{ - struct rvt_sge *sge = &ss->sge; - - while (length) { - u32 len = sge->length; - - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - WARN_ON_ONCE(len == 0); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } - length -= len; - } -} - /* * Make sure the QP is ready and able to accept the given opcode. */ @@ -690,27 +628,6 @@ static void mem_timer(unsigned long data) hfi1_qp_wakeup(qp, RVT_S_WAIT_KMEM); } -void update_sge(struct rvt_sge_state *ss, u32 length) -{ - struct rvt_sge *sge = &ss->sge; - - sge->vaddr += length; - sge->length -= length; - sge->sge_length -= length; - if (sge->sge_length == 0) { - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - return; - sge->n = 0; - } - sge->vaddr = sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = sge->mr->map[sge->m]->segs[sge->n].length; - } -} - /* * This is called with progress side lock held. */ @@ -799,7 +716,7 @@ static noinline int build_verbs_ulp_payload( len); if (ret) goto bail_txadd; - update_sge(ss, len); + rvt_update_sge(ss, len, false); length -= len; } return ret; @@ -1074,7 +991,7 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps, if (slen > len) slen = len; - update_sge(ss, slen); + rvt_update_sge(ss, slen, false); seg_pio_copy_mid(pbuf, addr, slen); len -= slen; } diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index 9fab6b3d12f6..3a0b589e41c2 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -291,8 +291,6 @@ int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps); void hfi1_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, bool release, bool copy_last); -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, bool release); - void hfi1_cnp_rcv(struct hfi1_packet *packet); void hfi1_uc_rcv(struct hfi1_packet *packet); diff --git a/include/rdma/rdmavt_mr.h b/include/rdma/rdmavt_mr.h index 05698d8d9c6f..f418bd5571a5 100644 --- a/include/rdma/rdmavt_mr.h +++ b/include/rdma/rdmavt_mr.h @@ -141,4 +141,54 @@ static inline void rvt_put_ss(struct rvt_sge_state *ss) } } +static inline u32 rvt_get_sge_length(struct rvt_sge *sge, u32 length) +{ + u32 len = sge->length; + + if (len > length) + len = length; + if (len > sge->sge_length) + len = sge->sge_length; + + return len; +} + +static inline void rvt_update_sge(struct rvt_sge_state *ss, u32 length, + bool release) +{ + struct rvt_sge *sge = &ss->sge; + + sge->vaddr += length; + sge->length -= length; + sge->sge_length -= length; + if (sge->sge_length == 0) { + if (release) + rvt_put_mr(sge->mr); + if (--ss->num_sge) + *sge = *ss->sg_list++; + } else if (sge->length == 0 && sge->mr->lkey) { + if (++sge->n >= RVT_SEGSZ) { + if (++sge->m >= sge->mr->mapsz) + return; + sge->n = 0; + } + sge->vaddr = sge->mr->map[sge->m]->segs[sge->n].vaddr; + sge->length = sge->mr->map[sge->m]->segs[sge->n].length; + } +} + +static inline void rvt_skip_sge(struct rvt_sge_state *ss, u32 length, + bool release) +{ + struct rvt_sge *sge = &ss->sge; + + while (length) { + u32 len = rvt_get_sge_length(sge, length); + + WARN_ON_ONCE(len == 0); + rvt_update_sge(ss, len, release); + length -= len; + } +} + #endif /* DEF_RDMAVT_INCMRH */ -- cgit v1.2.3 From 3fc4a0906f9c49839f03de4b6500d82fa5b3254f Mon Sep 17 00:00:00 2001 From: Brian Welty Date: Wed, 8 Feb 2017 05:27:43 -0800 Subject: IB/qib: Updates to use rdmavt's SGE helper routines Reviewed-by: Mike Marciniszyn Signed-off-by: Brian Welty Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qib/qib_rc.c | 2 +- drivers/infiniband/hw/qib/qib_ud.c | 4 +- drivers/infiniband/hw/qib/qib_verbs.c | 95 +++-------------------------------- drivers/infiniband/hw/qib/qib_verbs.h | 2 - 4 files changed, 9 insertions(+), 94 deletions(-) diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index ea0bc6bc09fa..d712043cbccc 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -49,7 +49,7 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, ss->sg_list = wqe->sg_list + 1; ss->num_sge = wqe->wr.num_sge; ss->total_len = wqe->length; - qib_skip_sge(ss, len, 0); + rvt_skip_sge(ss, len, false); return wqe->length - len; } diff --git a/drivers/infiniband/hw/qib/qib_ud.c b/drivers/infiniband/hw/qib/qib_ud.c index 04befa2551cd..ddd4e7458750 100644 --- a/drivers/infiniband/hw/qib/qib_ud.c +++ b/drivers/infiniband/hw/qib/qib_ud.c @@ -177,7 +177,7 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) sizeof(grh), 1); wc.wc_flags |= IB_WC_GRH; } else - qib_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); ssge.sg_list = swqe->sg_list + 1; ssge.sge = *swqe->sg_list; ssge.num_sge = swqe->wr.num_sge; @@ -567,7 +567,7 @@ void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, sizeof(struct ib_grh), 1); wc.wc_flags |= IB_WC_GRH; } else - qib_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); qib_copy_sge(&qp->r_sge, data, wc.byte_len - sizeof(struct ib_grh), 1); rvt_put_ss(&qp->r_sge); if (!test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c index ee7741e55817..b0b78e1cec92 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.c +++ b/drivers/infiniband/hw/qib/qib_verbs.c @@ -129,78 +129,16 @@ void qib_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, int release) struct rvt_sge *sge = &ss->sge; while (length) { - u32 len = sge->length; + u32 len = rvt_get_sge_length(sge, length); - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - BUG_ON(len == 0); + WARN_ON_ONCE(len == 0); memcpy(sge->vaddr, data, len); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } + rvt_update_sge(ss, len, release); data += len; length -= len; } } -/** - * qib_skip_sge - skip over SGE memory - XXX almost dup of prev func - * @ss: the SGE state - * @length: the number of bytes to skip - */ -void qib_skip_sge(struct rvt_sge_state *ss, u32 length, int release) -{ - struct rvt_sge *sge = &ss->sge; - - while (length) { - u32 len = sge->length; - - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - BUG_ON(len == 0); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } - length -= len; - } -} - /* * Count the number of DMA descriptors needed to send length bytes of data. * Don't modify the qib_sge_state to get the count. @@ -468,27 +406,6 @@ static void mem_timer(unsigned long data) } } -static void update_sge(struct rvt_sge_state *ss, u32 length) -{ - struct rvt_sge *sge = &ss->sge; - - sge->vaddr += length; - sge->length -= length; - sge->sge_length -= length; - if (sge->sge_length == 0) { - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - return; - sge->n = 0; - } - sge->vaddr = sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = sge->mr->map[sge->m]->segs[sge->n].length; - } -} - #ifdef __LITTLE_ENDIAN static inline u32 get_upper_bits(u32 data, u32 shift) { @@ -646,11 +563,11 @@ static void copy_io(u32 __iomem *piobuf, struct rvt_sge_state *ss, data = clear_upper_bytes(v, extra, 0); } } - update_sge(ss, len); + rvt_update_sge(ss, len, false); length -= len; } /* Update address before sending packet. */ - update_sge(ss, length); + rvt_update_sge(ss, length, false); if (flush_wc) { /* must flush early everything before trigger word */ qib_flush_wc(); @@ -1069,7 +986,7 @@ static int qib_verbs_send_pio(struct rvt_qp *qp, struct ib_header *ibhdr, u32 *addr = (u32 *) ss->sge.vaddr; /* Update address before sending packet. */ - update_sge(ss, len); + rvt_update_sge(ss, len, false); if (flush_wc) { qib_pio_copy(piobuf, addr, dwords - 1); /* must flush early everything before trigger word */ diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index 1bdfa1ceb563..212e8ce71be8 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -304,8 +304,6 @@ int qib_verbs_send(struct rvt_qp *qp, struct ib_header *hdr, void qib_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, int release); -void qib_skip_sge(struct rvt_sge_state *ss, u32 length, int release); - void qib_uc_rcv(struct qib_ibport *ibp, struct ib_header *hdr, int has_grh, void *data, u32 tlen, struct rvt_qp *qp); -- cgit v1.2.3 From f9215b5e536b0c598dff4041fc7d6136cd599981 Mon Sep 17 00:00:00 2001 From: Mike Marciniszyn Date: Wed, 8 Feb 2017 05:27:49 -0800 Subject: IB/rdmavt, IB/hfi1, IB/qib: Correct ack count for passive (RTR) QPs The send complete for RC QPs mismanages the ack count when the responder side is only in RTR. A QP in that state cannot send requests, but it can be the target for operations that elicit responses. Adjust the RC completion logic to correct the count maintenance by reflecting RECV_OK in a new state test. Reviewed-by: Kaike Wan Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/rc.c | 2 +- drivers/infiniband/hw/qib/qib_rc.c | 2 +- include/rdma/rdmavt_qp.h | 2 ++ 3 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 85e7bf6dcc11..3cddf7d03e70 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -990,7 +990,7 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) u32 psn; lockdep_assert_held(&qp->s_lock); - if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND)) + if (!(ib_rvt_state_ops[qp->state] & RVT_SEND_OR_FLUSH_OR_RECV_OK)) return; /* Find out where the BTH is */ diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index d712043cbccc..2a495fbc6508 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -895,7 +895,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) u32 opcode; u32 psn; - if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND)) + if (!(ib_rvt_state_ops[qp->state] & RVT_SEND_OR_FLUSH_OR_RECV_OK)) return; /* Find out where the BTH is */ diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index 561b6c876086..2f91fae8d23f 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -144,6 +144,8 @@ #define RVT_FLUSH_RECV 0x40 #define RVT_PROCESS_OR_FLUSH_SEND \ (RVT_PROCESS_SEND_OK | RVT_FLUSH_SEND) +#define RVT_SEND_OR_FLUSH_OR_RECV_OK \ + (RVT_PROCESS_SEND_OK | RVT_FLUSH_SEND | RVT_PROCESS_RECV_OK) /* * Internal send flags -- cgit v1.2.3 From c27aad00d191e8ba88fe755db53b9f2b8cefc0e9 Mon Sep 17 00:00:00 2001 From: Jakub Byczkowski Date: Wed, 8 Feb 2017 05:27:55 -0800 Subject: IB/hfi1: Modify logging frequency of DCC errors Use rate-limit state to limit number of messages logged to kernel message buffer for DCC errors. Add new macro dd_dev_info_ratelimited for that propose. Replace all dd_dev_info calls in handle_dcc_err function with rate-limited version. Reviewed-by: Dennis Dalessandro Signed-off-by: Jakub Byczkowski Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/chip.c | 20 +++++++++++--------- drivers/infiniband/hw/hfi1/hfi.h | 4 ++++ 2 files changed, 15 insertions(+), 9 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index ef72bc2a9e1d..b9d491c9071a 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -7827,7 +7827,8 @@ static void handle_dcc_err(struct hfi1_devdata *dd, u32 unused, u64 reg) } /* just report this */ - dd_dev_info(dd, "DCC Error: fmconfig error: %s\n", extra); + dd_dev_info_ratelimited(dd, "DCC Error: fmconfig error: %s\n", + extra); reg &= ~DCC_ERR_FLG_FMCONFIG_ERR_SMASK; } @@ -7878,34 +7879,35 @@ static void handle_dcc_err(struct hfi1_devdata *dd, u32 unused, u64 reg) } /* just report this */ - dd_dev_info(dd, "DCC Error: PortRcv error: %s\n", extra); - dd_dev_info(dd, " hdr0 0x%llx, hdr1 0x%llx\n", - hdr0, hdr1); + dd_dev_info_ratelimited(dd, "DCC Error: PortRcv error: %s\n" + " hdr0 0x%llx, hdr1 0x%llx\n", + extra, hdr0, hdr1); reg &= ~DCC_ERR_FLG_RCVPORT_ERR_SMASK; } if (reg & DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_UC_SMASK) { /* informative only */ - dd_dev_info(dd, "8051 access to LCB blocked\n"); + dd_dev_info_ratelimited(dd, "8051 access to LCB blocked\n"); reg &= ~DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_UC_SMASK; } if (reg & DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_HOST_SMASK) { /* informative only */ - dd_dev_info(dd, "host access to LCB blocked\n"); + dd_dev_info_ratelimited(dd, "host access to LCB blocked\n"); reg &= ~DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_HOST_SMASK; } /* report any remaining errors */ if (reg) - dd_dev_info(dd, "DCC Error: %s\n", - dcc_err_string(buf, sizeof(buf), reg)); + dd_dev_info_ratelimited(dd, "DCC Error: %s\n", + dcc_err_string(buf, sizeof(buf), reg)); if (lcl_reason == 0) lcl_reason = OPA_LINKDOWN_REASON_UNKNOWN; if (do_bounce) { - dd_dev_info(dd, "%s: PortErrorAction bounce\n", __func__); + dd_dev_info_ratelimited(dd, "%s: PortErrorAction bounce\n", + __func__); set_link_down_reason(ppd, lcl_reason, 0, lcl_reason); queue_work(ppd->hfi1_wq, &ppd->link_bounce_work); } diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 9c3be919c0b7..0808e3c3ba39 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1942,6 +1942,10 @@ static inline u64 hfi1_pkt_base_sdma_integrity(struct hfi1_devdata *dd) dev_info(&(dd)->pcidev->dev, "%s: " fmt, \ get_unit_name((dd)->unit), ##__VA_ARGS__) +#define dd_dev_info_ratelimited(dd, fmt, ...) \ + dev_info_ratelimited(&(dd)->pcidev->dev, "%s: " fmt, \ + get_unit_name((dd)->unit), ##__VA_ARGS__) + #define dd_dev_dbg(dd, fmt, ...) \ dev_dbg(&(dd)->pcidev->dev, "%s: " fmt, \ get_unit_name((dd)->unit), ##__VA_ARGS__) -- cgit v1.2.3 From db069ecb5d4dcd297628d5ebc56dd22bdbab9729 Mon Sep 17 00:00:00 2001 From: "Michael J. Ruhl" Date: Wed, 8 Feb 2017 05:28:13 -0800 Subject: IB/hfi1: Do not set physical link state if DC is in the shutdown state If the DC is in shutdown state, the set link state function will return an error. Since this is not a failure in this state, make sure to only call set link state if the DC is on. Reviewed-by: Easwar Hariharan Signed-off-by: Michael J. Ruhl Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/chip.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index b9d491c9071a..121a4c920f1b 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -10510,16 +10510,18 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state) ppd->remote_link_down_reason = 0; } - ret1 = set_physical_link_state(dd, PLS_DISABLED); - if (ret1 != HCMD_SUCCESS) { - dd_dev_err(dd, - "Failed to transition to Disabled link state, return 0x%x\n", - ret1); - ret = -EINVAL; - break; + if (!dd->dc_shutdown) { + ret1 = set_physical_link_state(dd, PLS_DISABLED); + if (ret1 != HCMD_SUCCESS) { + dd_dev_err(dd, + "Failed to transition to Disabled link state, return 0x%x\n", + ret1); + ret = -EINVAL; + break; + } + dc_shutdown(dd); } ppd->host_link_state = HLS_DN_DISABLE; - dc_shutdown(dd); break; case HLS_DN_OFFLINE: if (ppd->host_link_state == HLS_DN_DISABLE) -- cgit v1.2.3 From 881fccb8640bdd3533f5a98a99991aae23a5106b Mon Sep 17 00:00:00 2001 From: Don Hiatt Date: Wed, 8 Feb 2017 05:28:19 -0800 Subject: IB/hfi1: Add rvt_rnr_tbl_to_usec function Return usec from an index into ib_rvt_rnr_table. Reviewed-by: Mike Marciniszyn Signed-off-by: Don Hiatt Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/sw/rdmavt/qp.c | 11 +++++++++++ include/rdma/rdmavt_qp.h | 1 + 2 files changed, 12 insertions(+) diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index db4315f8f0c8..0b97598136b4 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -1954,6 +1954,17 @@ void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err) } EXPORT_SYMBOL(rvt_rc_error); +/* + * rvt_rnr_tbl_to_usec - return index into ib_rvt_rnr_table + * @index - the index + * return usec from an index into ib_rvt_rnr_table + */ +unsigned long rvt_rnr_tbl_to_usec(u32 index) +{ + return ib_rvt_rnr_table[(index & RVT_AETH_CREDIT_MASK)]; +} +EXPORT_SYMBOL(rvt_rnr_tbl_to_usec); + static inline unsigned long rvt_aeth_to_usec(u32 aeth) { return ib_rvt_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index 2f91fae8d23f..9767549ab42c 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -653,6 +653,7 @@ struct rvt_dev_info; void rvt_comm_est(struct rvt_qp *qp); int rvt_error_qp(struct rvt_qp *qp, enum ib_wc_status err); void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err); +unsigned long rvt_rnr_tbl_to_usec(u32 index); enum hrtimer_restart rvt_rc_rnr_retry(struct hrtimer *t); void rvt_add_rnr_timer(struct rvt_qp *qp, u32 aeth); void rvt_del_timers_sync(struct rvt_qp *qp); -- cgit v1.2.3 From 832666c163f04306fa6823b8974bccf7bb5e5ad3 Mon Sep 17 00:00:00 2001 From: Don Hiatt Date: Wed, 8 Feb 2017 05:28:25 -0800 Subject: IB/hfi1, qib, rdmavt: Move AETH defines to rdma/ib_hdrs.h Rename RVT AETH defines and export in rdma/ib_hdrs.h Reviewed-by: Mike Marciniszyn Signed-off-by: Don Hiatt Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/rc.c | 18 +++++++++--------- drivers/infiniband/hw/hfi1/ruc.c | 2 +- drivers/infiniband/hw/hfi1/trace.c | 4 ++-- drivers/infiniband/hw/qib/qib_rc.c | 18 +++++++++--------- drivers/infiniband/hw/qib/qib_ruc.c | 2 +- drivers/infiniband/sw/rdmavt/qp.c | 7 ++++--- drivers/infiniband/sw/rdmavt/rc.c | 15 +++++++-------- include/rdma/ib_hdrs.h | 6 ++++++ include/rdma/rdmavt_qp.h | 5 ----- 9 files changed, 39 insertions(+), 38 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 3cddf7d03e70..7382be11afca 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -211,9 +211,9 @@ normal: ps->s_txreq->ss = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | + cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->s_nak_state << - RVT_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; @@ -758,9 +758,9 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->r_nak_state << - RVT_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else ohdr->u.aeth = rvt_compute_aeth(qp); sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; @@ -1157,7 +1157,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> RVT_AETH_NAK_SHIFT) + if (aeth >> IB_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); ibp = rcd_to_iport(rcd); @@ -1237,7 +1237,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> RVT_AETH_NAK_SHIFT) { + switch (aeth >> IB_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1300,8 +1300,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, goto bail_stop; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> RVT_AETH_CREDIT_SHIFT) & - RVT_AETH_CREDIT_MASK) { + switch ((aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1431,7 +1431,7 @@ static void rc_rcv_resp(struct hfi1_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> RVT_AETH_NAK_SHIFT) == 0) + if ((aeth >> IB_AETH_NAK_SHIFT) == 0) rvt_get_credit(qp, aeth); } goto ack_done; diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c index bdf3697de881..aa15bcbfb079 100644 --- a/drivers/infiniband/hw/hfi1/ruc.c +++ b/drivers/infiniband/hw/hfi1/ruc.c @@ -580,7 +580,7 @@ rnr_nak: if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << - RVT_AETH_CREDIT_SHIFT); + IB_AETH_CREDIT_SHIFT); goto clr_busy; op_err: diff --git a/drivers/infiniband/hw/hfi1/trace.c b/drivers/infiniband/hw/hfi1/trace.c index 0985b4f1b9ef..e86798af6903 100644 --- a/drivers/infiniband/hw/hfi1/trace.c +++ b/drivers/infiniband/hw/hfi1/trace.c @@ -130,14 +130,14 @@ const char *parse_everbs_hdrs( case OP(RC, ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN, be32_to_cpu(eh->aeth) >> 24, parse_syndrome(be32_to_cpu(eh->aeth) >> 24), - be32_to_cpu(eh->aeth) & RVT_MSN_MASK); + be32_to_cpu(eh->aeth) & IB_MSN_MASK); break; /* aeth + atomicacketh */ case OP(RC, ATOMIC_ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN " " ATOMICACKETH_PRN, be32_to_cpu(eh->at.aeth) >> 24, parse_syndrome(be32_to_cpu(eh->at.aeth) >> 24), - be32_to_cpu(eh->at.aeth) & RVT_MSN_MASK, + be32_to_cpu(eh->at.aeth) & IB_MSN_MASK, ib_u64_get(&eh->at.atomic_ack_eth)); break; /* atomiceth */ diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 2a495fbc6508..12658e3fe154 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -187,9 +187,9 @@ normal: qp->s_cur_sge = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | + cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->s_nak_state << - RVT_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; @@ -648,9 +648,9 @@ void qib_send_rc_ack(struct rvt_qp *qp) if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & RVT_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->r_nak_state << - RVT_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else ohdr->u.aeth = rvt_compute_aeth(qp); lrh0 |= ibp->sl_to_vl[qp->remote_ah_attr.sl] << 12 | @@ -1042,7 +1042,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> RVT_AETH_NAK_SHIFT) + if (aeth >> IB_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); ibp = to_iport(qp->ibqp.device, qp->port_num); @@ -1122,7 +1122,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> RVT_AETH_NAK_SHIFT) { + switch (aeth >> IB_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1185,8 +1185,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, goto bail; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> RVT_AETH_CREDIT_SHIFT) & - RVT_AETH_CREDIT_MASK) { + switch ((aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1341,7 +1341,7 @@ static void qib_rc_rcv_resp(struct qib_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> RVT_AETH_NAK_SHIFT) == 0) + if ((aeth >> IB_AETH_NAK_SHIFT) == 0) rvt_get_credit(qp, aeth); } goto ack_done; diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c index 58b28412121d..17655cc3e6fe 100644 --- a/drivers/infiniband/hw/qib/qib_ruc.c +++ b/drivers/infiniband/hw/qib/qib_ruc.c @@ -562,7 +562,7 @@ rnr_nak: if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << - RVT_AETH_CREDIT_SHIFT); + IB_AETH_CREDIT_SHIFT); goto clr_busy; op_err: diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 0b97598136b4..f5ad8d4bfb39 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -51,6 +51,7 @@ #include #include #include +#include #include "qp.h" #include "vt.h" #include "trace.h" @@ -1961,14 +1962,14 @@ EXPORT_SYMBOL(rvt_rc_error); */ unsigned long rvt_rnr_tbl_to_usec(u32 index) { - return ib_rvt_rnr_table[(index & RVT_AETH_CREDIT_MASK)]; + return ib_rvt_rnr_table[(index & IB_AETH_CREDIT_MASK)]; } EXPORT_SYMBOL(rvt_rnr_tbl_to_usec); static inline unsigned long rvt_aeth_to_usec(u32 aeth) { - return ib_rvt_rnr_table[(aeth >> RVT_AETH_CREDIT_SHIFT) & - RVT_AETH_CREDIT_MASK]; + return ib_rvt_rnr_table[(aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK]; } /* diff --git a/drivers/infiniband/sw/rdmavt/rc.c b/drivers/infiniband/sw/rdmavt/rc.c index b32b5a3853ac..6131cc558bdb 100644 --- a/drivers/infiniband/sw/rdmavt/rc.c +++ b/drivers/infiniband/sw/rdmavt/rc.c @@ -46,8 +46,7 @@ */ #include - -#define RVT_AETH_CREDIT_INVAL RVT_AETH_CREDIT_MASK +#include /* * Convert the AETH credit code into the number of credits. @@ -94,14 +93,14 @@ static const u16 credit_table[31] = { */ __be32 rvt_compute_aeth(struct rvt_qp *qp) { - u32 aeth = qp->r_msn & RVT_MSN_MASK; + u32 aeth = qp->r_msn & IB_MSN_MASK; if (qp->ibqp.srq) { /* * Shared receive queues don't generate credits. * Set the credit field to the invalid value. */ - aeth |= RVT_AETH_CREDIT_INVAL << RVT_AETH_CREDIT_SHIFT; + aeth |= IB_AETH_CREDIT_INVAL << IB_AETH_CREDIT_SHIFT; } else { u32 min, max, x; u32 credits; @@ -143,7 +142,7 @@ __be32 rvt_compute_aeth(struct rvt_qp *qp) min = x; } } - aeth |= x << RVT_AETH_CREDIT_SHIFT; + aeth |= x << IB_AETH_CREDIT_SHIFT; } return cpu_to_be32(aeth); } @@ -159,7 +158,7 @@ EXPORT_SYMBOL(rvt_compute_aeth); void rvt_get_credit(struct rvt_qp *qp, u32 aeth) { struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); - u32 credit = (aeth >> RVT_AETH_CREDIT_SHIFT) & RVT_AETH_CREDIT_MASK; + u32 credit = (aeth >> IB_AETH_CREDIT_SHIFT) & IB_AETH_CREDIT_MASK; lockdep_assert_held(&qp->s_lock); /* @@ -167,7 +166,7 @@ void rvt_get_credit(struct rvt_qp *qp, u32 aeth) * as many packets as we like. Otherwise, we have to * honor the credit field. */ - if (credit == RVT_AETH_CREDIT_INVAL) { + if (credit == IB_AETH_CREDIT_INVAL) { if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { qp->s_flags |= RVT_S_UNLIMITED_CREDIT; if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { @@ -177,7 +176,7 @@ void rvt_get_credit(struct rvt_qp *qp, u32 aeth) } } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { /* Compute new LSN (i.e., MSN + credit) */ - credit = (aeth + credit_table[credit]) & RVT_MSN_MASK; + credit = (aeth + credit_table[credit]) & IB_MSN_MASK; if (rvt_cmp_msn(credit, qp->s_lsn) > 0) { qp->s_lsn = credit; if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { diff --git a/include/rdma/ib_hdrs.h b/include/rdma/ib_hdrs.h index 408439fe911e..c755325f0831 100644 --- a/include/rdma/ib_hdrs.h +++ b/include/rdma/ib_hdrs.h @@ -75,6 +75,12 @@ #define IB_GRH_FLOW_SHIFT 0 #define IB_GRH_NEXT_HDR 0x1B +#define IB_AETH_CREDIT_SHIFT 24 +#define IB_AETH_CREDIT_MASK 0x1F +#define IB_AETH_CREDIT_INVAL 0x1F +#define IB_AETH_NAK_SHIFT 29 +#define IB_MSN_MASK 0xFFFFFF + struct ib_reth { __be64 vaddr; /* potentially unaligned */ __be32 rkey; diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index 9767549ab42c..f3816396c76a 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -594,11 +594,6 @@ static inline void rvt_qp_swqe_complete( } } -#define RVT_AETH_CREDIT_SHIFT 24 -#define RVT_AETH_CREDIT_MASK 0x1F -#define RVT_AETH_NAK_SHIFT 29 -#define RVT_MSN_MASK 0xFFFFFF - /* * Compare the lower 24 bits of the msn values. * Returns an integer <, ==, or > than zero. -- cgit v1.2.3 From 1bb0d7b781b1ca647f1ae69665c0ad1de09a1174 Mon Sep 17 00:00:00 2001 From: "Michael J. Ruhl" Date: Wed, 8 Feb 2017 05:28:31 -0800 Subject: IB/hfi1: Code reuse with memdup_copy Update several usages of kmalloc/user_copy to memdup_copy and memdup_copy_nul. Reviewed-by: Dennis Dalessandro Reviewed-by: Mike Marciniszyn Signed-off-by: Michael J. Ruhl Signed-off-by: Dennis Dalessandro Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/debugfs.c | 39 +++++++++---------------------- drivers/infiniband/hw/hfi1/user_exp_rcv.c | 17 +++++--------- drivers/infiniband/hw/hfi1/user_sdma.c | 17 +++++++------- 3 files changed, 25 insertions(+), 48 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/debugfs.c b/drivers/infiniband/hw/hfi1/debugfs.c index 8725f4c086cf..7fe9dd885746 100644 --- a/drivers/infiniband/hw/hfi1/debugfs.c +++ b/drivers/infiniband/hw/hfi1/debugfs.c @@ -50,6 +50,7 @@ #include #include #include +#include #include "hfi.h" #include "debugfs.h" @@ -503,18 +504,11 @@ static ssize_t asic_flags_write(struct file *file, const char __user *buf, ppd = private2ppd(file); dd = ppd->dd; - buff = kmalloc(count + 1, GFP_KERNEL); - if (!buff) - return -ENOMEM; - - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto do_free; - } - /* zero terminate and read the expected integer */ - buff[count] = 0; + buff = memdup_user_nul(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); + ret = kstrtoull(buff, 0, &value); if (ret) goto do_free; @@ -692,15 +686,9 @@ static ssize_t __i2c_debugfs_write(struct file *file, const char __user *buf, if (i2c_addr == 0) return -EINVAL; - buff = kmalloc(count, GFP_KERNEL); - if (!buff) - return -ENOMEM; - - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto _free; - } + buff = memdup_user(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); total_written = i2c_write(ppd, target, i2c_addr, offset, buff, count); if (total_written < 0) { @@ -805,15 +793,10 @@ static ssize_t __qsfp_debugfs_write(struct file *file, const char __user *buf, ppd = private2ppd(file); - buff = kmalloc(count, GFP_KERNEL); - if (!buff) - return -ENOMEM; + buff = memdup_user(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto _free; - } total_written = qsfp_write(ppd, target, *ppos, buff, count); if (total_written < 0) { ret = total_written; diff --git a/drivers/infiniband/hw/hfi1/user_exp_rcv.c b/drivers/infiniband/hw/hfi1/user_exp_rcv.c index 64d26525435a..4a8295399e71 100644 --- a/drivers/infiniband/hw/hfi1/user_exp_rcv.c +++ b/drivers/infiniband/hw/hfi1/user_exp_rcv.c @@ -45,6 +45,7 @@ * */ #include +#include #include "user_exp_rcv.h" #include "trace.h" @@ -577,16 +578,10 @@ int hfi1_user_exp_rcv_clear(struct file *fp, struct hfi1_tid_info *tinfo) u32 *tidinfo; unsigned tididx; - tidinfo = kcalloc(tinfo->tidcnt, sizeof(*tidinfo), GFP_KERNEL); - if (!tidinfo) - return -ENOMEM; - - if (copy_from_user(tidinfo, (void __user *)(unsigned long) - tinfo->tidlist, sizeof(tidinfo[0]) * - tinfo->tidcnt)) { - ret = -EFAULT; - goto done; - } + tidinfo = memdup_user((void __user *)(unsigned long)tinfo->tidlist, + sizeof(tidinfo[0]) * tinfo->tidcnt); + if (IS_ERR(tidinfo)) + return PTR_ERR(tidinfo); mutex_lock(&uctxt->exp_lock); for (tididx = 0; tididx < tinfo->tidcnt; tididx++) { @@ -602,7 +597,7 @@ int hfi1_user_exp_rcv_clear(struct file *fp, struct hfi1_tid_info *tinfo) spin_unlock(&fd->tid_lock); tinfo->tidcnt = tididx; mutex_unlock(&uctxt->exp_lock); -done: + kfree(tidinfo); return ret; } diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index 7d22f8ee98ef..e6811c4edc73 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -60,6 +60,7 @@ #include #include #include +#include #include "hfi.h" #include "sdma.h" @@ -725,30 +726,28 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, */ if (req_opcode(req->info.ctrl) == EXPECTED) { u16 ntids = iovec[idx].iov_len / sizeof(*req->tids); + u32 *tmp; if (!ntids || ntids > MAX_TID_PAIR_ENTRIES) { ret = -EINVAL; goto free_req; } - req->tids = kcalloc(ntids, sizeof(*req->tids), GFP_KERNEL); - if (!req->tids) { - ret = -ENOMEM; - goto free_req; - } + /* * We have to copy all of the tids because they may vary * in size and, therefore, the TID count might not be * equal to the pkt count. However, there is no way to * tell at this point. */ - ret = copy_from_user(req->tids, iovec[idx].iov_base, - ntids * sizeof(*req->tids)); - if (ret) { + tmp = memdup_user(iovec[idx].iov_base, + ntids * sizeof(*req->tids)); + if (IS_ERR(tmp)) { + ret = PTR_ERR(tmp); SDMA_DBG(req, "Failed to copy %d TIDs (%d)", ntids, ret); - ret = -EFAULT; goto free_req; } + req->tids = tmp; req->n_tids = ntids; idx++; } -- cgit v1.2.3 From f50cccdd03ed4b22960308a5d2499519df13ee9b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 9 Feb 2017 16:10:58 +0100 Subject: IB/mthca: switch to pci_alloc_irq_vectors Trivial switch to the new API for this driver. Signed-off-by: Christoph Hellwig Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mthca/mthca_main.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index ded76c101dde..c309e5c96383 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -851,20 +851,18 @@ err_uar_table_free: static int mthca_enable_msi_x(struct mthca_dev *mdev) { - struct msix_entry entries[3]; int err; - entries[0].entry = 0; - entries[1].entry = 1; - entries[2].entry = 2; - - err = pci_enable_msix_exact(mdev->pdev, entries, ARRAY_SIZE(entries)); - if (err) + err = pci_alloc_irq_vectors(mdev->pdev, 3, 3, PCI_IRQ_MSIX); + if (err < 0) return err; - mdev->eq_table.eq[MTHCA_EQ_COMP ].msi_x_vector = entries[0].vector; - mdev->eq_table.eq[MTHCA_EQ_ASYNC].msi_x_vector = entries[1].vector; - mdev->eq_table.eq[MTHCA_EQ_CMD ].msi_x_vector = entries[2].vector; + mdev->eq_table.eq[MTHCA_EQ_COMP ].msi_x_vector = + pci_irq_vector(mdev->pdev, 0); + mdev->eq_table.eq[MTHCA_EQ_ASYNC].msi_x_vector = + pci_irq_vector(mdev->pdev, 1); + mdev->eq_table.eq[MTHCA_EQ_CMD ].msi_x_vector = + pci_irq_vector(mdev->pdev, 2); return 0; } @@ -1018,7 +1016,7 @@ static int __mthca_init_one(struct pci_dev *pdev, int hca_type) err = mthca_setup_hca(mdev); if (err == -EBUSY && (mdev->mthca_flags & MTHCA_FLAG_MSI_X)) { if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); mdev->mthca_flags &= ~MTHCA_FLAG_MSI_X; err = mthca_setup_hca(mdev); @@ -1062,7 +1060,7 @@ err_cleanup: err_close: if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); mthca_close_hca(mdev); @@ -1113,7 +1111,7 @@ static void __mthca_remove_one(struct pci_dev *pdev) mthca_cmd_cleanup(mdev); if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); ib_dealloc_device(&mdev->ib_dev); pci_release_regions(pdev); -- cgit v1.2.3 From c5e8f57b0b560b9cc9ea9166975858d023f81030 Mon Sep 17 00:00:00 2001 From: Zhu Yanjun Date: Fri, 10 Feb 2017 01:48:42 -0500 Subject: IB/ipoib: remove the unnecessary memory free In the function ipoib_cm_nonsrq_init_rx, the memory is not allocated successfully. It is not necessary to free it. CC: Joe Jin CC: Junxiao Bi Signed-off-by: Zhu Yanjun Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 53d69d7d9ad4..eb8cafe80115 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -363,7 +363,7 @@ static int ipoib_cm_nonsrq_init_rx(struct net_device *dev, struct ib_cm_id *cm_i t = kmalloc(sizeof *t, GFP_KERNEL); if (!t) { ret = -ENOMEM; - goto err_free; + goto err_free_1; } ipoib_cm_init_rx_wr(dev, &t->wr, t->sge); @@ -410,6 +410,8 @@ err_count: err_free: kfree(t); + +err_free_1: ipoib_cm_free_rx_ring(dev, rx->rx_ring); return ret; -- cgit v1.2.3 From 23536dfaa3ef5b4e61272ea65ed6c1b15db022ec Mon Sep 17 00:00:00 2001 From: Zhu Yanjun Date: Tue, 14 Feb 2017 03:43:12 -0500 Subject: IB/ipoib: Remove redudant label There are 2 labels to mark the same statements. Replace the 2 labels with one versatile labe. Cc: Joe Jin Cc: Junxiao Bi Signed-off-by: Zhu Yanjun Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index eb8cafe80115..a6d6c617b597 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1157,13 +1157,13 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, ret = ipoib_cm_modify_tx_init(p->dev, p->id, p->qp); if (ret) { ipoib_warn(priv, "failed to modify tx qp to rtr: %d\n", ret); - goto err_modify; + goto err_modify_send; } ret = ipoib_cm_send_req(p->dev, p->id, p->qp, qpn, pathrec); if (ret) { ipoib_warn(priv, "failed to send cm req: %d\n", ret); - goto err_send_cm; + goto err_modify_send; } ipoib_dbg(priv, "Request connection 0x%x for gid %pI6 qpn 0x%x\n", @@ -1171,8 +1171,7 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, return 0; -err_send_cm: -err_modify: +err_modify_send: ib_destroy_cm_id(p->id); err_id: p->id = NULL; -- cgit v1.2.3 From 64b2ae74e83c0294f66ab8db9c99c274fa1a5f1a Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 14 Feb 2017 22:23:07 +0100 Subject: IB/hfi1: use size_t for passing array length gcc-7 produces a mysterious warning about the size argument being potentially out of range: drivers/infiniband/hw/hfi1/verbs.c: In function 'init_cntr_names': drivers/infiniband/hw/hfi1/verbs.c:1644:2: error: 'memcpy': specified size between 18446744071562067968 and 18446744073709551615 exceeds maximum object size 9223372036854775807 [-Werror=stringop-overflow=] This seems to refer to a the case where an 64-bit size_t gets truncated into a negative 'int' and subsequently turned into a high 64-bit number again. The fix is clearly to use size_t here, which matches the type that gets used for this value elsewhere. Fixes: b7481944b06e ("IB/hfi1: Show statistics counters under IB stats interface") Signed-off-by: Arnd Bergmann Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/verbs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index 72f459e1fdde..5ba4c0dec348 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -1536,7 +1536,7 @@ static int cntr_names_initialized; * external strings. */ static int init_cntr_names(const char *names_in, - const int names_len, + const size_t names_len, int num_extra_names, int *num_cntrs, const char ***cntr_names) -- cgit v1.2.3 From 7bf3976d6cfd77c956484186e8bed8f4aab677ce Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 15 Feb 2017 08:47:04 +0100 Subject: vmw_pvrdma: switch to pci_alloc_irq_vectors .. and greatly clean up the irq handling boilerplate code. Signed-off-by: Christoph Hellwig Acked-by: Adit Ranadive Signed-off-by: Doug Ledford --- drivers/infiniband/hw/vmw_pvrdma/pvrdma.h | 8 +- drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h | 6 - drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c | 162 +++++----------------- 3 files changed, 34 insertions(+), 142 deletions(-) diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h index 71e1d55d69d6..3cd96c1b9502 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h @@ -196,13 +196,7 @@ struct pvrdma_dev { spinlock_t cmd_lock; /* Command lock. */ struct semaphore cmd_sema; struct completion cmd_done; - struct { - enum pvrdma_intr_type type; /* Intr type */ - struct msix_entry msix_entry[PVRDMA_MAX_INTERRUPTS]; - irq_handler_t handler[PVRDMA_MAX_INTERRUPTS]; - u8 enabled[PVRDMA_MAX_INTERRUPTS]; - u8 size; - } intr; + unsigned int nr_vectors; /* RDMA-related device information. */ union ib_gid *sgid_tbl; diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h index c06768635d65..e69d6f3cae32 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h @@ -149,12 +149,6 @@ enum pvrdma_intr_cause { PVRDMA_INTR_CAUSE_CQ = (1 << PVRDMA_INTR_VECTOR_CQ), }; -enum pvrdma_intr_type { - PVRDMA_INTR_TYPE_INTX, /* Legacy. */ - PVRDMA_INTR_TYPE_MSI, /* MSI. */ - PVRDMA_INTR_TYPE_MSIX, /* MSI-X. */ -}; - enum pvrdma_gos_bits { PVRDMA_GOS_BITS_UNK, /* Unknown. */ PVRDMA_GOS_BITS_32, /* 32-bit. */ diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c index bd8fbd3d2032..60cdb7719565 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c @@ -282,7 +282,7 @@ static irqreturn_t pvrdma_intr0_handler(int irq, void *dev_id) dev_dbg(&dev->pdev->dev, "interrupt 0 (response) handler\n"); - if (dev->intr.type != PVRDMA_INTR_TYPE_MSIX) { + if (!dev->pdev->msix_enabled) { /* Legacy intr */ icr = pvrdma_read_reg(dev, PVRDMA_REG_ICR); if (icr == 0) @@ -489,31 +489,13 @@ static irqreturn_t pvrdma_intrx_handler(int irq, void *dev_id) return IRQ_HANDLED; } -static void pvrdma_disable_msi_all(struct pvrdma_dev *dev) -{ - if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX) - pci_disable_msix(dev->pdev); - else if (dev->intr.type == PVRDMA_INTR_TYPE_MSI) - pci_disable_msi(dev->pdev); -} - static void pvrdma_free_irq(struct pvrdma_dev *dev) { int i; dev_dbg(&dev->pdev->dev, "freeing interrupts\n"); - - if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX) { - for (i = 0; i < dev->intr.size; i++) { - if (dev->intr.enabled[i]) { - free_irq(dev->intr.msix_entry[i].vector, dev); - dev->intr.enabled[i] = 0; - } - } - } else if (dev->intr.type == PVRDMA_INTR_TYPE_INTX || - dev->intr.type == PVRDMA_INTR_TYPE_MSI) { - free_irq(dev->pdev->irq, dev); - } + for (i = 0; i < dev->nr_vectors; i++) + free_irq(pci_irq_vector(dev->pdev, i), dev); } static void pvrdma_enable_intrs(struct pvrdma_dev *dev) @@ -528,126 +510,48 @@ static void pvrdma_disable_intrs(struct pvrdma_dev *dev) pvrdma_write_reg(dev, PVRDMA_REG_IMR, ~0); } -static int pvrdma_enable_msix(struct pci_dev *pdev, struct pvrdma_dev *dev) -{ - int i; - int ret; - - for (i = 0; i < PVRDMA_MAX_INTERRUPTS; i++) { - dev->intr.msix_entry[i].entry = i; - dev->intr.msix_entry[i].vector = i; - - switch (i) { - case 0: - /* CMD ring handler */ - dev->intr.handler[i] = pvrdma_intr0_handler; - break; - case 1: - /* Async event ring handler */ - dev->intr.handler[i] = pvrdma_intr1_handler; - break; - default: - /* Completion queue handler */ - dev->intr.handler[i] = pvrdma_intrx_handler; - break; - } - } - - ret = pci_enable_msix(pdev, dev->intr.msix_entry, - PVRDMA_MAX_INTERRUPTS); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSIX; - dev->intr.size = PVRDMA_MAX_INTERRUPTS; - } else if (ret > 0) { - ret = pci_enable_msix(pdev, dev->intr.msix_entry, ret); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSIX; - dev->intr.size = ret; - } else { - dev->intr.size = 0; - } - } - - dev_dbg(&pdev->dev, "using interrupt type %d, size %d\n", - dev->intr.type, dev->intr.size); - - return ret; -} - static int pvrdma_alloc_intrs(struct pvrdma_dev *dev) { - int ret = 0; - int i; + struct pci_dev *pdev = dev->pdev; + int ret = 0, i; - if (pci_find_capability(dev->pdev, PCI_CAP_ID_MSIX) && - pvrdma_enable_msix(dev->pdev, dev)) { - /* Try MSI */ - ret = pci_enable_msi(dev->pdev); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSI; - } else { - /* Legacy INTR */ - dev->intr.type = PVRDMA_INTR_TYPE_INTX; - } + ret = pci_alloc_irq_vectors(pdev, 1, PVRDMA_MAX_INTERRUPTS, + PCI_IRQ_MSIX); + if (ret < 0) { + ret = pci_alloc_irq_vectors(pdev, 1, 1, + PCI_IRQ_MSI | PCI_IRQ_LEGACY); + if (ret < 0) + return ret; } + dev->nr_vectors = ret; - /* Request First IRQ */ - switch (dev->intr.type) { - case PVRDMA_INTR_TYPE_INTX: - case PVRDMA_INTR_TYPE_MSI: - ret = request_irq(dev->pdev->irq, pvrdma_intr0_handler, - IRQF_SHARED, DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt\n"); - goto disable_msi; - } - break; - case PVRDMA_INTR_TYPE_MSIX: - ret = request_irq(dev->intr.msix_entry[0].vector, - pvrdma_intr0_handler, 0, DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt 0\n"); - goto disable_msi; - } - dev->intr.enabled[0] = 1; - break; - default: - /* Not reached */ - break; + ret = request_irq(pci_irq_vector(dev->pdev, 0), pvrdma_intr0_handler, + pdev->msix_enabled ? 0 : IRQF_SHARED, DRV_NAME, dev); + if (ret) { + dev_err(&dev->pdev->dev, + "failed to request interrupt 0\n"); + goto out_free_vectors; } - /* For MSIX: request intr for each vector */ - if (dev->intr.size > 1) { - ret = request_irq(dev->intr.msix_entry[1].vector, - pvrdma_intr1_handler, 0, DRV_NAME, dev); + for (i = 1; i < dev->nr_vectors; i++) { + ret = request_irq(pci_irq_vector(dev->pdev, i), + i == 1 ? pvrdma_intr1_handler : + pvrdma_intrx_handler, + 0, DRV_NAME, dev); if (ret) { dev_err(&dev->pdev->dev, - "failed to request interrupt 1\n"); - goto free_irq; - } - dev->intr.enabled[1] = 1; - - for (i = 2; i < dev->intr.size; i++) { - ret = request_irq(dev->intr.msix_entry[i].vector, - pvrdma_intrx_handler, 0, - DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt %d\n", i); - goto free_irq; - } - dev->intr.enabled[i] = 1; + "failed to request interrupt %d\n", i); + goto free_irqs; } } return 0; -free_irq: - pvrdma_free_irq(dev); -disable_msi: - pvrdma_disable_msi_all(dev); +free_irqs: + while (--i >= 0) + free_irq(pci_irq_vector(dev->pdev, i), dev); +out_free_vectors: + pci_free_irq_vectors(pdev); return ret; } @@ -1091,7 +995,7 @@ err_free_uar_table: pvrdma_uar_table_cleanup(dev); err_free_intrs: pvrdma_free_irq(dev); - pvrdma_disable_msi_all(dev); + pci_free_irq_vectors(pdev); err_free_cq_ring: pvrdma_page_dir_cleanup(dev, &dev->cq_pdir); err_free_async_ring: @@ -1141,7 +1045,7 @@ static void pvrdma_pci_remove(struct pci_dev *pdev) pvrdma_disable_intrs(dev); pvrdma_free_irq(dev); - pvrdma_disable_msi_all(dev); + pci_free_irq_vectors(pdev); /* Deactivate pvrdma device */ pvrdma_write_reg(dev, PVRDMA_REG_CTL, PVRDMA_DEVICE_CTL_RESET); -- cgit v1.2.3 From c67294b70b5aafa2769e8c677a044243ce974c3a Mon Sep 17 00:00:00 2001 From: Yuval Shaia Date: Thu, 16 Feb 2017 14:05:05 +0200 Subject: IB/vmw_pvrdma: Expose vendor error to ULPs Signed-off-by: Yuval Shaia Acked-by: Adit Ranadive Signed-off-by: Doug Ledford --- drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c index e429ca5b16aa..69bda611d313 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c @@ -373,7 +373,7 @@ retry: wc->sl = cqe->sl; wc->dlid_path_bits = cqe->dlid_path_bits; wc->port_num = cqe->port_num; - wc->vendor_err = 0; + wc->vendor_err = cqe->vendor_err; /* Update shared ring state */ pvrdma_idx_ring_inc(&cq->ring_state->rx.cons_head, cq->ibcq.cqe); -- cgit v1.2.3 From 3ecc16c82c068885cef5697d2f056a8fc317cf45 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 17 Feb 2017 15:38:26 +0100 Subject: IB/hns: include linux/module.h I ran into a build error on arm64 randconfig testing: infiniband/hw/hns/hns_roce_main.c:539:1: error: data definition has no type or storage class [-Werror] infiniband/hw/hns/hns_roce_main.c:539:1: error: type defaults to 'int' in declaration of 'MODULE_DEVICE_TABLE' [-Werror=implicit-int] infiniband/hw/hns/hns_roce_main.c:539:1: error: parameter names (without types) in function declaration [-Werror] infiniband/hw/hns/hns_roce_main.c:979:226: error: data definition has no type or storage class [-Werror] infiniband/hw/hns/hns_roce_main.c:979:226: error: type defaults to 'int' in declaration of 'module_init' [-Werror=implicit-int] infiniband/hw/hns/hns_roce_main.c:979:1: error: parameter names (without types) in function declaration [-Werror] Including the module.h makes it build again. Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver") Signed-off-by: Arnd Bergmann Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hns/hns_roce_main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c index 4953d9cb83a7..cf14679664ca 100644 --- a/drivers/infiniband/hw/hns/hns_roce_main.c +++ b/drivers/infiniband/hw/hns/hns_roce_main.c @@ -32,6 +32,7 @@ */ #include #include +#include #include #include #include -- cgit v1.2.3 From e57f774db1fef3d983fd77db0f48f3f39875c65f Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 17 Feb 2017 15:40:21 +0100 Subject: RDMA/bnxt_re: add DCB dependency When CONFIG_DCB is disabled, we get a link error: drivers/infiniband/built-in.o: In function `bnxt_re_setup_qos': trace.c:(.text+0x155774): undefined reference to `dcb_ieee_getapp_mask' trace.c:(.text+0x155774): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `dcb_ieee_getapp_mask' trace.c:(.text+0x155794): undefined reference to `dcb_ieee_getapp_mask' trace.c:(.text+0x155794): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `dcb_ieee_getapp_mask' Like the other drivers that use this function, a Kconfig dependency is the correct fix. Signed-off-by: Arnd Bergmann Signed-off-by: Doug Ledford --- drivers/infiniband/hw/bnxt_re/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/bnxt_re/Kconfig b/drivers/infiniband/hw/bnxt_re/Kconfig index cd0175e32584..19982a4a9bba 100644 --- a/drivers/infiniband/hw/bnxt_re/Kconfig +++ b/drivers/infiniband/hw/bnxt_re/Kconfig @@ -1,6 +1,6 @@ config INFINIBAND_BNXT_RE tristate "Broadcom Netxtreme HCA support" - depends on ETHERNET && NETDEVICES && PCI && INET + depends on ETHERNET && NETDEVICES && PCI && INET && DCB select NET_VENDOR_BROADCOM select BNXT ---help--- -- cgit v1.2.3 From 4cd33aafe45866860b7c72401297d1efcd7bd8aa Mon Sep 17 00:00:00 2001 From: Christophe Jaillet Date: Sat, 18 Feb 2017 12:28:15 +0100 Subject: RDMA/qedr: Fix some error handling 'qedr_alloc_pbl_tbl()' can not return NULL. In qedr_init_user_queue(): - simplify the test for the return value, no need to test for NULL - propagate the error pointer if needed, otherwise 0 (success) is returned. This is spurious. In init_mr_info(): - test the return value with IS_ERR - propagate the error pointer if needed instead of an exlictit -ENOMEM. This is a no-op as the only error pointer that we can have here is already -ENOMEM Signed-off-by: Christophe JAILLET Acked-by: Ram Amrani Signed-off-by: Doug Ledford --- drivers/infiniband/hw/qedr/verbs.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c index ef83a3f322d6..0c51657af151 100644 --- a/drivers/infiniband/hw/qedr/verbs.c +++ b/drivers/infiniband/hw/qedr/verbs.c @@ -771,8 +771,10 @@ static inline int qedr_init_user_queue(struct ib_ucontext *ib_ctx, goto err0; q->pbl_tbl = qedr_alloc_pbl_tbl(dev, &q->pbl_info, GFP_KERNEL); - if (IS_ERR_OR_NULL(q->pbl_tbl)) + if (IS_ERR(q->pbl_tbl)) { + rc = PTR_ERR(q->pbl_tbl); goto err0; + } qedr_populate_pbls(dev, q->umem, q->pbl_tbl, &q->pbl_info); @@ -2105,8 +2107,8 @@ static int init_mr_info(struct qedr_dev *dev, struct mr_info *info, goto done; info->pbl_table = qedr_alloc_pbl_tbl(dev, &info->pbl_info, GFP_KERNEL); - if (!info->pbl_table) { - rc = -ENOMEM; + if (IS_ERR(info->pbl_table)) { + rc = PTR_ERR(info->pbl_table); goto done; } @@ -2117,7 +2119,7 @@ static int init_mr_info(struct qedr_dev *dev, struct mr_info *info, * list and allocating another one */ tmp = qedr_alloc_pbl_tbl(dev, &info->pbl_info, GFP_KERNEL); - if (!tmp) { + if (IS_ERR(tmp)) { DP_DEBUG(dev, QEDR_MSG_MR, "Extra PBL is not allocated\n"); goto done; } -- cgit v1.2.3 From d6c58dc40fec35ff6cdb350b53bce0fcf9143709 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:29 -0800 Subject: IB/SRP: Avoid using IB_MR_TYPE_SG_GAPS Tests have shown that the following error message is reported when using SG-GAPS registration with an mlx5 adapter: scsi host1: ib_srp: failed RECV status WR flushed (5) for CQE ffff880bd4270eb0 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 0f007806 2500002a ad9fafd1 scsi host1: ib_srp: reconnect succeeded mlx5_0:dump_cqe:262:(pid 7369): dump error cqe 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 0f007806 25000032 00105dd0 scsi host1: ib_srp: failed FAST REG status memory management operation error (6) for CQE ffff880b92860138 Hence avoid using SG-GAPS memory registrations. Additionally, always configure the blk_queue_virt_boundary() to avoid to trigger a mapping failure when using adapters that support SG-GAPS (e.g. mlx5). Fixes: commit ad8e66b4a801 ("IB/srp: fix mr allocation when the device supports sg gaps") Fixes: commit 509c5f33f4f6 ("IB/srp: Prevent mapping failures") Reported-by: Laurence Oberman Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Leon Romanovsky Cc: Mark Bloch Cc: Yuval Shaia Cc: # 4.7+ Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 79bf48477ddb..6e4467150049 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -371,7 +371,6 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, struct srp_fr_desc *d; struct ib_mr *mr; int i, ret = -EINVAL; - enum ib_mr_type mr_type; if (pool_size <= 0) goto err; @@ -385,13 +384,9 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, spin_lock_init(&pool->lock); INIT_LIST_HEAD(&pool->free_list); - if (device->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG) - mr_type = IB_MR_TYPE_SG_GAPS; - else - mr_type = IB_MR_TYPE_MEM_REG; - for (i = 0, d = &pool->desc[0]; i < pool->size; i++, d++) { - mr = ib_alloc_mr(pd, mr_type, max_page_list_len); + mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, + max_page_list_len); if (IS_ERR(mr)) { ret = PTR_ERR(mr); if (ret == -ENOMEM) @@ -2664,9 +2659,8 @@ static int srp_slave_alloc(struct scsi_device *sdev) struct Scsi_Host *shost = sdev->host; struct srp_target_port *target = host_to_target(shost); struct srp_device *srp_dev = target->srp_host->srp_dev; - struct ib_device *ibdev = srp_dev->dev; - if (!(ibdev->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG)) + if (true) blk_queue_virt_boundary(sdev->request_queue, ~srp_dev->mr_page_mask); -- cgit v1.2.3 From 6cb72bc1b40bb2c1750ee7a5ebade93bed49a5fb Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:30 -0800 Subject: IB/srp: Avoid that duplicate responses trigger a kernel bug After srp_process_rsp() returns there is a short time during which the scsi_host_find_tag() call will return a pointer to the SCSI command that is being completed. If during that time a duplicate response is received, avoid that the following call stack appears: BUG: unable to handle kernel NULL pointer dereference at (null) IP: srp_recv_done+0x450/0x6b0 [ib_srp] Oops: 0000 [#1] SMP CPU: 10 PID: 0 Comm: swapper/10 Not tainted 4.10.0-rc7-dbg+ #1 Call Trace: __ib_process_cq+0x4b/0xd0 [ib_core] ib_poll_handler+0x1d/0x70 [ib_core] irq_poll_softirq+0xba/0x120 __do_softirq+0xba/0x4c0 irq_exit+0xbe/0xd0 smp_apic_timer_interrupt+0x38/0x50 apic_timer_interrupt+0x90/0xa0 RIP: srp_recv_done+0x450/0x6b0 [ib_srp] RSP: ffff88046f483e20 Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Cc: Steve Feeley Cc: Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 6e4467150049..a49289f600b1 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1892,9 +1892,11 @@ static void srp_process_rsp(struct srp_rdma_ch *ch, struct srp_rsp *rsp) complete(&ch->tsk_mgmt_done); } else { scmnd = scsi_host_find_tag(target->scsi_host, rsp->tag); - if (scmnd) { + if (scmnd && scmnd->host_scribble) { req = (void *)scmnd->host_scribble; scmnd = srp_claim_req(ch, req, NULL, scmnd); + } else { + scmnd = NULL; } if (!scmnd) { shost_printk(KERN_ERR, target->scsi_host, -- cgit v1.2.3 From 0a6fdbdeb1c25e31763c1fb333fa2723a7d2aba6 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:31 -0800 Subject: IB/srp: Fix race conditions related to task management Avoid that srp_process_rsp() overwrites the status information in ch if the SRP target response timed out and processing of another task management function has already started. Avoid that issuing multiple task management functions concurrently triggers list corruption. This patch prevents that the following stack trace appears in the system log: WARNING: CPU: 8 PID: 9269 at lib/list_debug.c:52 __list_del_entry_valid+0xbc/0xc0 list_del corruption. prev->next should be ffffc90004bb7b00, but was ffff8804052ecc68 CPU: 8 PID: 9269 Comm: sg_reset Tainted: G W 4.10.0-rc7-dbg+ #3 Call Trace: dump_stack+0x68/0x93 __warn+0xc6/0xe0 warn_slowpath_fmt+0x4a/0x50 __list_del_entry_valid+0xbc/0xc0 wait_for_completion_timeout+0x12e/0x170 srp_send_tsk_mgmt+0x1ef/0x2d0 [ib_srp] srp_reset_device+0x5b/0x110 [ib_srp] scsi_ioctl_reset+0x1c7/0x290 scsi_ioctl+0x12a/0x420 sd_ioctl+0x9d/0x100 blkdev_ioctl+0x51e/0x9f0 block_ioctl+0x38/0x40 do_vfs_ioctl+0x8f/0x700 SyS_ioctl+0x3c/0x70 entry_SYSCALL_64_fastpath+0x18/0xad Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Cc: Steve Feeley Cc: Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 45 ++++++++++++++++++++++++------------- drivers/infiniband/ulp/srp/ib_srp.h | 1 + 2 files changed, 30 insertions(+), 16 deletions(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index a49289f600b1..d9b57f5958b5 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1884,12 +1884,17 @@ static void srp_process_rsp(struct srp_rdma_ch *ch, struct srp_rsp *rsp) if (unlikely(rsp->tag & SRP_TAG_TSK_MGMT)) { spin_lock_irqsave(&ch->lock, flags); ch->req_lim += be32_to_cpu(rsp->req_lim_delta); + if (rsp->tag == ch->tsk_mgmt_tag) { + ch->tsk_mgmt_status = -1; + if (be32_to_cpu(rsp->resp_data_len) >= 4) + ch->tsk_mgmt_status = rsp->data[3]; + complete(&ch->tsk_mgmt_done); + } else { + shost_printk(KERN_ERR, target->scsi_host, + "Received tsk mgmt response too late for tag %#llx\n", + rsp->tag); + } spin_unlock_irqrestore(&ch->lock, flags); - - ch->tsk_mgmt_status = -1; - if (be32_to_cpu(rsp->resp_data_len) >= 4) - ch->tsk_mgmt_status = rsp->data[3]; - complete(&ch->tsk_mgmt_done); } else { scmnd = scsi_host_find_tag(target->scsi_host, rsp->tag); if (scmnd && scmnd->host_scribble) { @@ -2528,19 +2533,18 @@ srp_change_queue_depth(struct scsi_device *sdev, int qdepth) } static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, - u8 func) + u8 func, u8 *status) { struct srp_target_port *target = ch->target; struct srp_rport *rport = target->rport; struct ib_device *dev = target->srp_host->srp_dev->dev; struct srp_iu *iu; struct srp_tsk_mgmt *tsk_mgmt; + int res; if (!ch->connected || target->qp_in_error) return -1; - init_completion(&ch->tsk_mgmt_done); - /* * Lock the rport mutex to avoid that srp_create_ch_ib() is * invoked while a task management function is being sent. @@ -2563,10 +2567,16 @@ static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, tsk_mgmt->opcode = SRP_TSK_MGMT; int_to_scsilun(lun, &tsk_mgmt->lun); - tsk_mgmt->tag = req_tag | SRP_TAG_TSK_MGMT; tsk_mgmt->tsk_mgmt_func = func; tsk_mgmt->task_tag = req_tag; + spin_lock_irq(&ch->lock); + ch->tsk_mgmt_tag = (ch->tsk_mgmt_tag + 1) | SRP_TAG_TSK_MGMT; + tsk_mgmt->tag = ch->tsk_mgmt_tag; + spin_unlock_irq(&ch->lock); + + init_completion(&ch->tsk_mgmt_done); + ib_dma_sync_single_for_device(dev, iu->dma, sizeof *tsk_mgmt, DMA_TO_DEVICE); if (srp_post_send(ch, iu, sizeof(*tsk_mgmt))) { @@ -2575,13 +2585,15 @@ static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, return -1; } + res = wait_for_completion_timeout(&ch->tsk_mgmt_done, + msecs_to_jiffies(SRP_ABORT_TIMEOUT_MS)); + if (res > 0 && status) + *status = ch->tsk_mgmt_status; mutex_unlock(&rport->mutex); - if (!wait_for_completion_timeout(&ch->tsk_mgmt_done, - msecs_to_jiffies(SRP_ABORT_TIMEOUT_MS))) - return -1; + WARN_ON_ONCE(res < 0); - return 0; + return res > 0 ? 0 : -1; } static int srp_abort(struct scsi_cmnd *scmnd) @@ -2607,7 +2619,7 @@ static int srp_abort(struct scsi_cmnd *scmnd) shost_printk(KERN_ERR, target->scsi_host, "Sending SRP abort for tag %#x\n", tag); if (srp_send_tsk_mgmt(ch, tag, scmnd->device->lun, - SRP_TSK_ABORT_TASK) == 0) + SRP_TSK_ABORT_TASK, NULL) == 0) ret = SUCCESS; else if (target->rport->state == SRP_RPORT_LOST) ret = FAST_IO_FAIL; @@ -2625,14 +2637,15 @@ static int srp_reset_device(struct scsi_cmnd *scmnd) struct srp_target_port *target = host_to_target(scmnd->device->host); struct srp_rdma_ch *ch; int i; + u8 status; shost_printk(KERN_ERR, target->scsi_host, "SRP reset_device called\n"); ch = &target->ch[0]; if (srp_send_tsk_mgmt(ch, SRP_TAG_NO_REQ, scmnd->device->lun, - SRP_TSK_LUN_RESET)) + SRP_TSK_LUN_RESET, &status)) return FAILED; - if (ch->tsk_mgmt_status) + if (status) return FAILED; for (i = 0; i < target->ch_count; i++) { diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h index 21c69695f9d4..32ed40db3ca2 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.h +++ b/drivers/infiniband/ulp/srp/ib_srp.h @@ -163,6 +163,7 @@ struct srp_rdma_ch { int max_ti_iu_len; int comp_vector; + u64 tsk_mgmt_tag; struct completion tsk_mgmt_done; u8 tsk_mgmt_status; bool connected; -- cgit v1.2.3 From 93c76dbb749eb0c791c2680f8fdaeae51240fdf5 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:32 -0800 Subject: IB/srp: Document locking conventions Use lockdep_assert_held() statements to verify at run-time whether the proper locks are held. Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index d9b57f5958b5..862ee2ff5bb9 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -40,6 +40,7 @@ #include #include #include +#include #include #include @@ -1799,6 +1800,8 @@ static struct srp_iu *__srp_get_tx_iu(struct srp_rdma_ch *ch, s32 rsv = (iu_type == SRP_IU_TSK_MGMT) ? 0 : SRP_TSK_MGMT_SQ_SIZE; struct srp_iu *iu; + lockdep_assert_held(&ch->lock); + ib_process_cq_direct(ch->send_cq, -1); if (list_empty(&ch->free_tx)) @@ -1829,6 +1832,8 @@ static void srp_send_done(struct ib_cq *cq, struct ib_wc *wc) return; } + lockdep_assert_held(&ch->lock); + list_add(&iu->list, &ch->free_tx); } -- cgit v1.2.3 From a7139ca80f7cd79e034a5c5f9c242bfb0471545c Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:33 -0800 Subject: IB/srp: Make a diagnostic message more informative Report the destination port GID if connecting fails. Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 862ee2ff5bb9..33dd0920c248 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3435,9 +3435,10 @@ static ssize_t srp_create_target(struct device *dev, ret = srp_connect_ch(ch, multich); if (ret) { shost_printk(KERN_ERR, target->scsi_host, - PFX "Connection %d/%d failed\n", + PFX "Connection %d/%d to %pI6 failed\n", ch_start + cpu_idx, - target->ch_count); + target->ch_count, + ch->target->orig_dgid.raw); if (node_idx == 0 && cpu_idx == 0) { goto err_disconnect; } else { -- cgit v1.2.3 From b02c15360b17dde352fbf4003e20dabcd3ae9157 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:34 -0800 Subject: IB/srp: Improve an error path Avoid that the following message is printed if login fails: scsi host0: ib_srp: Sending CM DREQ failed Signed-off-by: Bart Van Assche Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 33dd0920c248..e8225cc8b938 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3440,7 +3440,7 @@ static ssize_t srp_create_target(struct device *dev, target->ch_count, ch->target->orig_dgid.raw); if (node_idx == 0 && cpu_idx == 0) { - goto err_disconnect; + goto free_ch; } else { srp_free_ch_ib(target, ch); srp_free_req_data(target, ch); @@ -3487,6 +3487,7 @@ put: err_disconnect: srp_disconnect_target(target); +free_ch: for (i = 0; i < target->ch_count; i++) { ch = &target->ch[i]; srp_free_ch_ib(target, ch); -- cgit v1.2.3 From f039f44fc331a7c6f828dfed97d5df0588602fd8 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:35 -0800 Subject: IB/core: Add support for draining IB_POLL_DIRECT completion queues Signed-off-by: Bart Van Assche Cc: Steve Wise Cc: Chuck Lever Cc: Christoph Hellwig Cc: Max Gurtovoy Reviewed-by: Steve Wise Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/cq.c | 4 ++-- drivers/infiniband/core/verbs.c | 35 +++++++++++++++-------------------- 2 files changed, 17 insertions(+), 22 deletions(-) diff --git a/drivers/infiniband/core/cq.c b/drivers/infiniband/core/cq.c index 838147ef727f..e95510117a6d 100644 --- a/drivers/infiniband/core/cq.c +++ b/drivers/infiniband/core/cq.c @@ -58,8 +58,8 @@ static int __ib_process_cq(struct ib_cq *cq, int budget) * %IB_POLL_DIRECT CQ. It does not offload CQ processing to a different * context and does not ask for completion interrupts from the HCA. * - * Note: for compatibility reasons -1 can be passed in %budget for unlimited - * polling. Do not use this feature in new code, it will be removed soon. + * Note: do not pass -1 as %budget unless it is guaranteed that the number + * of completions that will be processed is small. */ int ib_process_cq_direct(struct ib_cq *cq, int budget) { diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 9b77fbc86903..85ed5051fdfd 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1948,17 +1948,12 @@ static void ib_drain_qp_done(struct ib_cq *cq, struct ib_wc *wc) */ static void __ib_drain_sq(struct ib_qp *qp) { + struct ib_cq *cq = qp->send_cq; struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR }; struct ib_drain_cqe sdrain; struct ib_send_wr swr = {}, *bad_swr; int ret; - if (qp->send_cq->poll_ctx == IB_POLL_DIRECT) { - WARN_ONCE(qp->send_cq->poll_ctx == IB_POLL_DIRECT, - "IB_POLL_DIRECT poll_ctx not supported for drain\n"); - return; - } - swr.wr_cqe = &sdrain.cqe; sdrain.cqe.done = ib_drain_qp_done; init_completion(&sdrain.done); @@ -1975,7 +1970,11 @@ static void __ib_drain_sq(struct ib_qp *qp) return; } - wait_for_completion(&sdrain.done); + if (cq->poll_ctx == IB_POLL_DIRECT) + while (wait_for_completion_timeout(&sdrain.done, HZ / 10) <= 0) + ib_process_cq_direct(cq, -1); + else + wait_for_completion(&sdrain.done); } /* @@ -1983,17 +1982,12 @@ static void __ib_drain_sq(struct ib_qp *qp) */ static void __ib_drain_rq(struct ib_qp *qp) { + struct ib_cq *cq = qp->recv_cq; struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR }; struct ib_drain_cqe rdrain; struct ib_recv_wr rwr = {}, *bad_rwr; int ret; - if (qp->recv_cq->poll_ctx == IB_POLL_DIRECT) { - WARN_ONCE(qp->recv_cq->poll_ctx == IB_POLL_DIRECT, - "IB_POLL_DIRECT poll_ctx not supported for drain\n"); - return; - } - rwr.wr_cqe = &rdrain.cqe; rdrain.cqe.done = ib_drain_qp_done; init_completion(&rdrain.done); @@ -2010,7 +2004,11 @@ static void __ib_drain_rq(struct ib_qp *qp) return; } - wait_for_completion(&rdrain.done); + if (cq->poll_ctx == IB_POLL_DIRECT) + while (wait_for_completion_timeout(&rdrain.done, HZ / 10) <= 0) + ib_process_cq_direct(cq, -1); + else + wait_for_completion(&rdrain.done); } /** @@ -2027,8 +2025,7 @@ static void __ib_drain_rq(struct ib_qp *qp) * ensure there is room in the CQ and SQ for the drain work request and * completion. * - * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQ using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. @@ -2056,8 +2053,7 @@ EXPORT_SYMBOL(ib_drain_sq); * ensure there is room in the CQ and RQ for the drain work request and * completion. * - * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQ using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. @@ -2081,8 +2077,7 @@ EXPORT_SYMBOL(ib_drain_rq); * ensure there is room in the CQ(s), SQ, and RQ for drain work requests * and completions. * - * allocate the CQs using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQs using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. -- cgit v1.2.3 From 9294000d6d895ad609f3cc4aff98c9c6175b466f Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 14 Feb 2017 10:56:36 -0800 Subject: IB/srp: Drain the send queue before destroying a QP A quote from the IB spec: However, if the Consumer does not wait for the Affiliated Asynchronous Last WQE Reached Event, then WQE and Data Segment leakage may occur. Therefore, it is good programming practice to tear down a QP that is associated with an SRQ by using the following process: * Put the QP in the Error State; * wait for the Affiliated Asynchronous Last WQE Reached Event; * either: * drain the CQ by invoking the Poll CQ verb and either wait for CQ to be empty or the number of Poll CQ operations has exceeded CQ capacity size; or * post another WR that completes on the same CQ and wait for this WR to return as a WC; * and then invoke a Destroy QP or Reset QP. Signed-off-by: Bart Van Assche Cc: Christoph Hellwig Cc: Israel Rukshin Cc: Max Gurtovoy Cc: Laurence Oberman Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/srp/ib_srp.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index e8225cc8b938..c7d97097d55c 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -466,9 +466,13 @@ static struct srp_fr_pool *srp_alloc_fr_pool(struct srp_target_port *target) * completion handler can access the queue pair while it is * being destroyed. */ -static void srp_destroy_qp(struct ib_qp *qp) +static void srp_destroy_qp(struct srp_rdma_ch *ch, struct ib_qp *qp) { - ib_drain_rq(qp); + spin_lock_irq(&ch->lock); + ib_process_cq_direct(ch->send_cq, -1); + spin_unlock_irq(&ch->lock); + + ib_drain_qp(qp); ib_destroy_qp(qp); } @@ -542,7 +546,7 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch) } if (ch->qp) - srp_destroy_qp(ch->qp); + srp_destroy_qp(ch, ch->qp); if (ch->recv_cq) ib_free_cq(ch->recv_cq); if (ch->send_cq) @@ -566,7 +570,7 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch) return 0; err_qp: - srp_destroy_qp(qp); + srp_destroy_qp(ch, qp); err_send_cq: ib_free_cq(send_cq); @@ -609,7 +613,7 @@ static void srp_free_ch_ib(struct srp_target_port *target, ib_destroy_fmr_pool(ch->fmr_pool); } - srp_destroy_qp(ch->qp); + srp_destroy_qp(ch, ch->qp); ib_free_cq(ch->send_cq); ib_free_cq(ch->recv_cq); @@ -1822,6 +1826,11 @@ static struct srp_iu *__srp_get_tx_iu(struct srp_rdma_ch *ch, return iu; } +/* + * Note: if this function is called from inside ib_drain_sq() then it will + * be called without ch->lock being held. If ib_drain_sq() dequeues a WQE + * with status IB_WC_SUCCESS then that's a bug. + */ static void srp_send_done(struct ib_cq *cq, struct ib_wc *wc) { struct srp_iu *iu = container_of(wc->wr_cqe, struct srp_iu, cqe); -- cgit v1.2.3 From f2625f7db4dd0bbd16a9c7d2950e7621f9aa57ad Mon Sep 17 00:00:00 2001 From: Steve Wise Date: Tue, 21 Feb 2017 11:21:57 -0800 Subject: rdma_cm: fail iwarp accepts w/o connection params cma_accept_iw() needs to return an error if conn_params is NULL. Since this is coming from user space, we can crash. Reported-by: Shaobo He Acked-by: Sean Hefty Signed-off-by: Steve Wise Cc: stable@vger.kernel.org Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index dab2a96ea032..5ed6ec9d6a93 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -3683,6 +3683,9 @@ static int cma_accept_iw(struct rdma_id_private *id_priv, struct iw_cm_conn_param iw_param; int ret; + if (!conn_param) + return -EINVAL; + ret = cma_modify_qp_rtr(id_priv, conn_param); if (ret) return ret; -- cgit v1.2.3 From db690328a7df0b507f7d59de0c7e1bbe8f4b9e6a Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 15 Feb 2017 11:30:03 +1100 Subject: RDMA/bnxt_re: fix for "bnxt_en: Update to firmware interface spec 1.7.0." When the firmware interface spec was updated, a constant element was renamed. The rename missed the instances in the bnxt_re driver because it wasn't upstream yet. This updates the bnxt_re driver with the rename. Signed-off-by: Stephen Rothwell Signed-off-by: Doug Ledford --- drivers/infiniband/hw/bnxt_re/main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/bnxt_re/main.c b/drivers/infiniband/hw/bnxt_re/main.c index 6b9f1178050f..bd452a92b386 100644 --- a/drivers/infiniband/hw/bnxt_re/main.c +++ b/drivers/infiniband/hw/bnxt_re/main.c @@ -228,7 +228,7 @@ static int bnxt_re_net_ring_free(struct bnxt_re_dev *rdev, u16 fw_ring_id, } bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_RING_FREE, -1, -1); - req.ring_type = RING_ALLOC_REQ_RING_TYPE_CMPL; + req.ring_type = RING_ALLOC_REQ_RING_TYPE_L2_CMPL; req.ring_id = cpu_to_le16(fw_ring_id); bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); @@ -268,7 +268,7 @@ static int bnxt_re_net_ring_alloc(struct bnxt_re_dev *rdev, dma_addr_t *dma_arr, /* Association of ring index with doorbell index and MSIX number */ req.logical_id = cpu_to_le16(map_index); req.length = cpu_to_le32(ring_mask + 1); - req.ring_type = RING_ALLOC_REQ_RING_TYPE_CMPL; + req.ring_type = RING_ALLOC_REQ_RING_TYPE_L2_CMPL; req.int_mode = RING_ALLOC_REQ_INT_MODE_MSIX; bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); -- cgit v1.2.3