diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2019-03-09 15:53:03 -0800 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2019-03-09 15:53:03 -0800 |
commit | a50243b1ddcdd766d0d17fbfeeb1a22e62fdc461 (patch) | |
tree | 3dbf847105558eaac3658a46c4934df503c866a2 /include/rdma | |
parent | 2901752c14b8e1b7dd898d2e5245c93e531aa624 (diff) | |
parent | fca22e7e595f1799cfbfdfa13e16d48ece0d136c (diff) | |
download | linux-stable-a50243b1ddcdd766d0d17fbfeeb1a22e62fdc461.tar.gz linux-stable-a50243b1ddcdd766d0d17fbfeeb1a22e62fdc461.tar.bz2 linux-stable-a50243b1ddcdd766d0d17fbfeeb1a22e62fdc461.zip |
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull rdma updates from Jason Gunthorpe:
"This has been a slightly more active cycle than normal with ongoing
core changes and quite a lot of collected driver updates.
- Various driver fixes for bnxt_re, cxgb4, hns, mlx5, pvrdma, rxe
- A new data transfer mode for HFI1 giving higher performance
- Significant functional and bug fix update to the mlx5
On-Demand-Paging MR feature
- A chip hang reset recovery system for hns
- Change mm->pinned_vm to an atomic64
- Update bnxt_re to support a new 57500 chip
- A sane netlink 'rdma link add' method for creating rxe devices and
fixing the various unregistration race conditions in rxe's
unregister flow
- Allow lookup up objects by an ID over netlink
- Various reworking of the core to driver interface:
- drivers should not assume umem SGLs are in PAGE_SIZE chunks
- ucontext is accessed via udata not other means
- start to make the core code responsible for object memory
allocation
- drivers should convert struct device to struct ib_device via a
helper
- drivers have more tools to avoid use after unregister problems"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (280 commits)
net/mlx5: ODP support for XRC transport is not enabled by default in FW
IB/hfi1: Close race condition on user context disable and close
RDMA/umem: Revert broken 'off by one' fix
RDMA/umem: minor bug fix in error handling path
RDMA/hns: Use GFP_ATOMIC in hns_roce_v2_modify_qp
cxgb4: kfree mhp after the debug print
IB/rdmavt: Fix concurrency panics in QP post_send and modify to error
IB/rdmavt: Fix loopback send with invalidate ordering
IB/iser: Fix dma_nents type definition
IB/mlx5: Set correct write permissions for implicit ODP MR
bnxt_re: Clean cq for kernel consumers only
RDMA/uverbs: Don't do double free of allocated PD
RDMA: Handle ucontext allocations by IB/core
RDMA/core: Fix a WARN() message
bnxt_re: fix the regression due to changes in alloc_pbl
IB/mlx4: Increase the timeout for CM cache
IB/core: Abort page fault handler silently during owning process exit
IB/mlx5: Validate correct PD before prefetch MR
IB/mlx5: Protect against prefetch of invalid MR
RDMA/uverbs: Store PR pointer before it is overwritten
...
Diffstat (limited to 'include/rdma')
-rw-r--r-- | include/rdma/ib_hdrs.h | 14 | ||||
-rw-r--r-- | include/rdma/ib_mad.h | 5 | ||||
-rw-r--r-- | include/rdma/ib_umem.h | 8 | ||||
-rw-r--r-- | include/rdma/ib_umem_odp.h | 34 | ||||
-rw-r--r-- | include/rdma/ib_verbs.h | 274 | ||||
-rw-r--r-- | include/rdma/iw_cm.h | 16 | ||||
-rw-r--r-- | include/rdma/iw_portmap.h | 144 | ||||
-rw-r--r-- | include/rdma/rdma_cm.h | 1 | ||||
-rw-r--r-- | include/rdma/rdma_netlink.h | 11 | ||||
-rw-r--r-- | include/rdma/rdma_vt.h | 30 | ||||
-rw-r--r-- | include/rdma/rdmavt_qp.h | 20 | ||||
-rw-r--r-- | include/rdma/restrack.h | 58 | ||||
-rw-r--r-- | include/rdma/tid_rdma_defs.h | 108 | ||||
-rw-r--r-- | include/rdma/uverbs_ioctl.h | 18 | ||||
-rw-r--r-- | include/rdma/uverbs_std_types.h | 18 | ||||
-rw-r--r-- | include/rdma/uverbs_types.h | 1 |
16 files changed, 446 insertions, 314 deletions
diff --git a/include/rdma/ib_hdrs.h b/include/rdma/ib_hdrs.h index 6e35416170a3..9a90bd031e8c 100644 --- a/include/rdma/ib_hdrs.h +++ b/include/rdma/ib_hdrs.h @@ -1,5 +1,5 @@ /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2016 - 2018 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. @@ -100,6 +100,8 @@ struct ib_atomic_eth { __be64 compare_data; /* potentially unaligned */ } __packed; +#include <rdma/tid_rdma_defs.h> + union ib_ehdrs { struct { __be32 deth[2]; @@ -117,6 +119,16 @@ union ib_ehdrs { __be32 aeth; __be32 ieth; struct ib_atomic_eth atomic_eth; + /* TID RDMA headers */ + union { + struct tid_rdma_read_req r_req; + struct tid_rdma_read_resp r_rsp; + struct tid_rdma_write_req w_req; + struct tid_rdma_write_resp w_rsp; + struct tid_rdma_write_data w_data; + struct tid_rdma_resync resync; + struct tid_rdma_ack ack; + } tid_rdma; } __packed; struct ib_other_headers { diff --git a/include/rdma/ib_mad.h b/include/rdma/ib_mad.h index fdef558e3a2d..79ba8219e7dc 100644 --- a/include/rdma/ib_mad.h +++ b/include/rdma/ib_mad.h @@ -616,12 +616,11 @@ struct ib_mad_agent { void *context; u32 hi_tid; u32 flags; + void *security; + struct list_head mad_agent_sec_list; u8 port_num; u8 rmpp_version; - void *security; bool smp_allowed; - bool lsm_nb_reg; - struct notifier_block lsm_nb; }; /** diff --git a/include/rdma/ib_umem.h b/include/rdma/ib_umem.h index 5d3755ec5afa..73af05db04c7 100644 --- a/include/rdma/ib_umem.h +++ b/include/rdma/ib_umem.h @@ -36,6 +36,7 @@ #include <linux/list.h> #include <linux/scatterlist.h> #include <linux/workqueue.h> +#include <rdma/ib_verbs.h> struct ib_ucontext; struct ib_umem_odp; @@ -80,7 +81,7 @@ static inline size_t ib_umem_num_pages(struct ib_umem *umem) #ifdef CONFIG_INFINIBAND_USER_MEM -struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, +struct ib_umem *ib_umem_get(struct ib_udata *udata, unsigned long addr, size_t size, int access, int dmasync); void ib_umem_release(struct ib_umem *umem); int ib_umem_page_count(struct ib_umem *umem); @@ -91,9 +92,10 @@ int ib_umem_copy_from(void *dst, struct ib_umem *umem, size_t offset, #include <linux/err.h> -static inline struct ib_umem *ib_umem_get(struct ib_ucontext *context, +static inline struct ib_umem *ib_umem_get(struct ib_udata *udata, unsigned long addr, size_t size, - int access, int dmasync) { + int access, int dmasync) +{ return ERR_PTR(-EINVAL); } static inline void ib_umem_release(struct ib_umem *umem) { } diff --git a/include/rdma/ib_umem_odp.h b/include/rdma/ib_umem_odp.h index 0b1446fe2fab..dadc96dea39c 100644 --- a/include/rdma/ib_umem_odp.h +++ b/include/rdma/ib_umem_odp.h @@ -83,6 +83,19 @@ static inline struct ib_umem_odp *to_ib_umem_odp(struct ib_umem *umem) return container_of(umem, struct ib_umem_odp, umem); } +/* + * The lower 2 bits of the DMA address signal the R/W permissions for + * the entry. To upgrade the permissions, provide the appropriate + * bitmask to the map_dma_pages function. + * + * Be aware that upgrading a mapped address might result in change of + * the DMA address for the page. + */ +#define ODP_READ_ALLOWED_BIT (1<<0ULL) +#define ODP_WRITE_ALLOWED_BIT (1<<1ULL) + +#define ODP_DMA_ADDR_MASK (~(ODP_READ_ALLOWED_BIT | ODP_WRITE_ALLOWED_BIT)) + #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING struct ib_ucontext_per_mm { @@ -103,23 +116,10 @@ struct ib_ucontext_per_mm { }; int ib_umem_odp_get(struct ib_umem_odp *umem_odp, int access); -struct ib_umem_odp *ib_alloc_odp_umem(struct ib_ucontext_per_mm *per_mm, +struct ib_umem_odp *ib_alloc_odp_umem(struct ib_umem_odp *root_umem, unsigned long addr, size_t size); void ib_umem_odp_release(struct ib_umem_odp *umem_odp); -/* - * The lower 2 bits of the DMA address signal the R/W permissions for - * the entry. To upgrade the permissions, provide the appropriate - * bitmask to the map_dma_pages function. - * - * Be aware that upgrading a mapped address might result in change of - * the DMA address for the page. - */ -#define ODP_READ_ALLOWED_BIT (1<<0ULL) -#define ODP_WRITE_ALLOWED_BIT (1<<1ULL) - -#define ODP_DMA_ADDR_MASK (~(ODP_READ_ALLOWED_BIT | ODP_WRITE_ALLOWED_BIT)) - int ib_umem_odp_map_dma_pages(struct ib_umem_odp *umem_odp, u64 start_offset, u64 bcnt, u64 access_mask, unsigned long current_seq); @@ -169,12 +169,6 @@ static inline int ib_umem_odp_get(struct ib_umem_odp *umem_odp, int access) return -EINVAL; } -static inline struct ib_umem_odp * -ib_alloc_odp_umem(struct ib_ucontext *context, unsigned long addr, size_t size) -{ - return ERR_PTR(-EINVAL); -} - static inline void ib_umem_odp_release(struct ib_umem_odp *umem_odp) {} #endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 80debf5982ac..9b9e17bcc201 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -238,6 +238,7 @@ enum ib_device_cap_flags { IB_DEVICE_RDMA_NETDEV_OPA_VNIC = (1ULL << 35), /* The device supports padding incoming writes to cacheline. */ IB_DEVICE_PCI_WRITE_END_PADDING = (1ULL << 36), + IB_DEVICE_ALLOW_USER_UNREG = (1ULL << 37), }; enum ib_signature_prot_cap { @@ -268,6 +269,7 @@ enum ib_odp_transport_cap_bits { IB_ODP_SUPPORT_WRITE = 1 << 2, IB_ODP_SUPPORT_READ = 1 << 3, IB_ODP_SUPPORT_ATOMIC = 1 << 4, + IB_ODP_SUPPORT_SRQ_RECV = 1 << 5, }; struct ib_odp_caps { @@ -276,6 +278,7 @@ struct ib_odp_caps { uint32_t rc_odp_caps; uint32_t uc_odp_caps; uint32_t ud_odp_caps; + uint32_t xrc_odp_caps; } per_transport_caps; }; @@ -1504,12 +1507,10 @@ struct ib_ucontext { bool cleanup_retryable; -#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING void (*invalidate_range)(struct ib_umem_odp *umem_odp, unsigned long start, unsigned long end); struct mutex per_mm_list_lock; struct list_head per_mm_list; -#endif struct ib_rdmacg_object cg_obj; /* @@ -2186,7 +2187,6 @@ struct ib_port_cache { struct ib_cache { rwlock_t lock; struct ib_event_handler event_handler; - struct ib_port_cache *ports; }; struct iw_cm_verbs; @@ -2198,6 +2198,21 @@ struct ib_port_immutable { u32 max_mad_size; }; +struct ib_port_data { + struct ib_device *ib_dev; + + struct ib_port_immutable immutable; + + spinlock_t pkey_list_lock; + struct list_head pkey_list; + + struct ib_port_cache cache; + + spinlock_t netdev_lock; + struct net_device __rcu *netdev; + struct hlist_node ndev_hash_link; +}; + /* rdma netdev type - specifies protocol type */ enum rdma_netdev_t { RDMA_NETDEV_OPA_VNIC, @@ -2243,12 +2258,6 @@ struct rdma_netdev_alloc_params { struct net_device *netdev, void *param); }; -struct ib_port_pkey_list { - /* Lock to hold while modifying the list. */ - spinlock_t list_lock; - struct list_head pkey_list; -}; - struct ib_counters { struct ib_device *device; struct ib_uobject *uobject; @@ -2264,6 +2273,19 @@ struct ib_counters_read_attr { struct uverbs_attr_bundle; +#define INIT_RDMA_OBJ_SIZE(ib_struct, drv_struct, member) \ + .size_##ib_struct = \ + (sizeof(struct drv_struct) + \ + BUILD_BUG_ON_ZERO(offsetof(struct drv_struct, member)) + \ + BUILD_BUG_ON_ZERO( \ + !__same_type(((struct drv_struct *)NULL)->member, \ + struct ib_struct))) + +#define rdma_zalloc_drv_obj(ib_dev, ib_type) \ + ((struct ib_type *)kzalloc(ib_dev->ops.size_##ib_type, GFP_KERNEL)) + +#define DECLARE_RDMA_OBJ_SIZE(ib_struct) size_t size_##ib_struct + /** * struct ib_device_ops - InfiniBand device operations * This structure defines all the InfiniBand device operations, providers will @@ -2367,15 +2389,14 @@ struct ib_device_ops { int (*del_gid)(const struct ib_gid_attr *attr, void **context); int (*query_pkey)(struct ib_device *device, u8 port_num, u16 index, u16 *pkey); - struct ib_ucontext *(*alloc_ucontext)(struct ib_device *device, - struct ib_udata *udata); - int (*dealloc_ucontext)(struct ib_ucontext *context); + int (*alloc_ucontext)(struct ib_ucontext *context, + struct ib_udata *udata); + void (*dealloc_ucontext)(struct ib_ucontext *context); int (*mmap)(struct ib_ucontext *context, struct vm_area_struct *vma); void (*disassociate_ucontext)(struct ib_ucontext *ibcontext); - struct ib_pd *(*alloc_pd)(struct ib_device *device, - struct ib_ucontext *context, - struct ib_udata *udata); - int (*dealloc_pd)(struct ib_pd *pd); + int (*alloc_pd)(struct ib_pd *pd, struct ib_ucontext *context, + struct ib_udata *udata); + void (*dealloc_pd)(struct ib_pd *pd); struct ib_ah *(*create_ah)(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, u32 flags, struct ib_udata *udata); @@ -2506,34 +2527,57 @@ struct ib_device_ops { */ int (*get_hw_stats)(struct ib_device *device, struct rdma_hw_stats *stats, u8 port, int index); + /* + * This function is called once for each port when a ib device is + * registered. + */ + int (*init_port)(struct ib_device *device, u8 port_num, + struct kobject *port_sysfs); + /** + * Allows rdma drivers to add their own restrack attributes. + */ + int (*fill_res_entry)(struct sk_buff *msg, + struct rdma_restrack_entry *entry); + + /* Device lifecycle callbacks */ + /* + * Called after the device becomes registered, before clients are + * attached + */ + int (*enable_driver)(struct ib_device *dev); + /* + * This is called as part of ib_dealloc_device(). + */ + void (*dealloc_driver)(struct ib_device *dev); + + DECLARE_RDMA_OBJ_SIZE(ib_pd); + DECLARE_RDMA_OBJ_SIZE(ib_ucontext); }; +struct rdma_restrack_root; + struct ib_device { /* Do not access @dma_device directly from ULP nor from HW drivers. */ struct device *dma_device; struct ib_device_ops ops; char name[IB_DEVICE_NAME_MAX]; + struct rcu_head rcu_head; struct list_head event_handler_list; spinlock_t event_handler_lock; - rwlock_t client_data_lock; - struct list_head core_list; - /* Access to the client_data_list is protected by the client_data_lock - * rwlock and the lists_rwsem read-write semaphore - */ - struct list_head client_data_list; + struct rw_semaphore client_data_rwsem; + struct xarray client_data; + struct mutex unregistration_lock; struct ib_cache cache; /** - * port_immutable is indexed by port number + * port_data is indexed by port number */ - struct ib_port_immutable *port_immutable; + struct ib_port_data *port_data; int num_comp_vectors; - struct ib_port_pkey_list *port_pkey_list; - struct iw_cm_verbs *iwcm; struct module *owner; @@ -2547,12 +2591,6 @@ struct ib_device { struct kobject *ports_kobj; struct list_head port_list; - enum { - IB_DEV_UNINITIALIZED, - IB_DEV_REGISTERED, - IB_DEV_UNREGISTERED - } reg_state; - int uverbs_abi_ver; u64 uverbs_cmd_mask; u64 uverbs_ex_cmd_mask; @@ -2561,6 +2599,8 @@ struct ib_device { __be64 node_guid; u32 local_dma_lkey; u16 is_switch:1; + /* Indicates kernel verbs support, should not be used in drivers */ + u16 kverbs_provider:1; u8 node_type; u8 phys_port_cnt; struct ib_device_attr attrs; @@ -2572,10 +2612,7 @@ struct ib_device { #endif u32 index; - /* - * Implementation details of the RDMA core, don't use in drivers - */ - struct rdma_restrack_root res; + struct rdma_restrack_root *res; const struct uapi_definition *driver_def; enum rdma_driver_id driver_id; @@ -2586,10 +2623,13 @@ struct ib_device { */ refcount_t refcount; struct completion unreg_completion; + struct work_struct unregistration_work; + + const struct rdma_link_ops *link_ops; }; struct ib_client { - char *name; + const char *name; void (*add) (struct ib_device *); void (*remove)(struct ib_device *, void *client_data); @@ -2616,22 +2656,47 @@ struct ib_client { const struct sockaddr *addr, void *client_data); struct list_head list; + u32 client_id; + + /* kverbs are not required by the client */ + u8 no_kverbs_req:1; }; -struct ib_device *ib_alloc_device(size_t size); +struct ib_device *_ib_alloc_device(size_t size); +#define ib_alloc_device(drv_struct, member) \ + container_of(_ib_alloc_device(sizeof(struct drv_struct) + \ + BUILD_BUG_ON_ZERO(offsetof( \ + struct drv_struct, member))), \ + struct drv_struct, member) + void ib_dealloc_device(struct ib_device *device); void ib_get_device_fw_str(struct ib_device *device, char *str); -int ib_register_device(struct ib_device *device, const char *name, - int (*port_callback)(struct ib_device *, u8, - struct kobject *)); +int ib_register_device(struct ib_device *device, const char *name); void ib_unregister_device(struct ib_device *device); +void ib_unregister_driver(enum rdma_driver_id driver_id); +void ib_unregister_device_and_put(struct ib_device *device); +void ib_unregister_device_queued(struct ib_device *ib_dev); int ib_register_client (struct ib_client *client); void ib_unregister_client(struct ib_client *client); -void *ib_get_client_data(struct ib_device *device, struct ib_client *client); +/** + * ib_get_client_data - Get IB client context + * @device:Device to get context for + * @client:Client to get context for + * + * ib_get_client_data() returns the client context data set with + * ib_set_client_data(). This can only be called while the client is + * registered to the device, once the ib_client remove() callback returns this + * cannot be called. + */ +static inline void *ib_get_client_data(struct ib_device *device, + struct ib_client *client) +{ + return xa_load(&device->client_data, client->client_id); +} void ib_set_client_data(struct ib_device *device, struct ib_client *client, void *data); void ib_set_device_ops(struct ib_device *device, @@ -2790,6 +2855,16 @@ static inline u8 rdma_start_port(const struct ib_device *device) } /** + * rdma_for_each_port - Iterate over all valid port numbers of the IB device + * @device - The struct ib_device * to iterate over + * @iter - The unsigned int to store the port number + */ +#define rdma_for_each_port(device, iter) \ + for (iter = rdma_start_port(device + BUILD_BUG_ON_ZERO(!__same_type( \ + unsigned int, iter))); \ + iter <= rdma_end_port(device); (iter)++) + +/** * rdma_end_port - Return the last valid port number for the device * specified * @@ -2812,34 +2887,38 @@ static inline int rdma_is_port_valid(const struct ib_device *device, static inline bool rdma_is_grh_required(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & - RDMA_CORE_PORT_IB_GRH_REQUIRED; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_PORT_IB_GRH_REQUIRED; } 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; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_IB; } static inline bool rdma_protocol_roce(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & - (RDMA_CORE_CAP_PROT_ROCE | RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP); + return device->port_data[port_num].immutable.core_cap_flags & + (RDMA_CORE_CAP_PROT_ROCE | RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP); } static inline bool rdma_protocol_roce_udp_encap(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; } static inline bool rdma_protocol_roce_eth_encap(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_ROCE; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_ROCE; } static inline bool rdma_protocol_iwarp(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_IWARP; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_IWARP; } static inline bool rdma_ib_or_roce(const struct ib_device *device, u8 port_num) @@ -2850,12 +2929,14 @@ static inline bool rdma_ib_or_roce(const struct ib_device *device, u8 port_num) static inline bool rdma_protocol_raw_packet(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_RAW_PACKET; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_RAW_PACKET; } static inline bool rdma_protocol_usnic(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_USNIC; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_PROT_USNIC; } /** @@ -2872,7 +2953,8 @@ static inline bool rdma_protocol_usnic(const struct ib_device *device, u8 port_n */ static inline bool rdma_cap_ib_mad(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_MAD; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_IB_MAD; } /** @@ -2896,8 +2978,8 @@ static inline bool rdma_cap_ib_mad(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_opa_mad(struct ib_device *device, u8 port_num) { - return (device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_OPA_MAD) - == RDMA_CORE_CAP_OPA_MAD; + return (device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_OPA_MAD) == RDMA_CORE_CAP_OPA_MAD; } /** @@ -2922,7 +3004,8 @@ static inline bool rdma_cap_opa_mad(struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_ib_smi(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_SMI; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_IB_SMI; } /** @@ -2942,7 +3025,8 @@ static inline bool rdma_cap_ib_smi(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_ib_cm(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_CM; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_IB_CM; } /** @@ -2959,7 +3043,8 @@ static inline bool rdma_cap_ib_cm(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_iw_cm(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IW_CM; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_IW_CM; } /** @@ -2979,7 +3064,8 @@ static inline bool rdma_cap_iw_cm(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_ib_sa(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_SA; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_IB_SA; } /** @@ -3019,7 +3105,8 @@ static inline bool rdma_cap_ib_mcast(const struct ib_device *device, u8 port_num */ static inline bool rdma_cap_af_ib(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_AF_IB; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_AF_IB; } /** @@ -3040,7 +3127,8 @@ static inline bool rdma_cap_af_ib(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_eth_ah(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_ETH_AH; + return device->port_data[port_num].immutable.core_cap_flags & + RDMA_CORE_CAP_ETH_AH; } /** @@ -3054,7 +3142,7 @@ static inline bool rdma_cap_eth_ah(const struct ib_device *device, u8 port_num) */ static inline bool rdma_cap_opa_ah(struct ib_device *device, u8 port_num) { - return (device->port_immutable[port_num].core_cap_flags & + return (device->port_data[port_num].immutable.core_cap_flags & RDMA_CORE_CAP_OPA_AH) == RDMA_CORE_CAP_OPA_AH; } @@ -3072,7 +3160,7 @@ static inline bool rdma_cap_opa_ah(struct ib_device *device, u8 port_num) */ static inline size_t rdma_max_mad_size(const struct ib_device *device, u8 port_num) { - return device->port_immutable[port_num].max_mad_size; + return device->port_data[port_num].immutable.max_mad_size; } /** @@ -3686,32 +3774,18 @@ static inline void ib_dma_unmap_sg_attrs(struct ib_device *dev, { dma_unmap_sg_attrs(dev->dma_device, sg, nents, direction, dma_attrs); } -/** - * ib_sg_dma_address - Return the DMA address from a scatter/gather entry - * @dev: The device for which the DMA addresses were created - * @sg: The scatter/gather entry - * - * Note: this function is obsolete. To do: change all occurrences of - * ib_sg_dma_address() into sg_dma_address(). - */ -static inline u64 ib_sg_dma_address(struct ib_device *dev, - struct scatterlist *sg) -{ - return sg_dma_address(sg); -} /** - * ib_sg_dma_len - Return the DMA length from a scatter/gather entry - * @dev: The device for which the DMA addresses were created - * @sg: The scatter/gather entry + * ib_dma_max_seg_size - Return the size limit of a single DMA transfer + * @dev: The device to query * - * Note: this function is obsolete. To do: change all occurrences of - * ib_sg_dma_len() into sg_dma_len(). + * The returned value represents a size in bytes. */ -static inline unsigned int ib_sg_dma_len(struct ib_device *dev, - struct scatterlist *sg) +static inline unsigned int ib_dma_max_seg_size(struct ib_device *dev) { - return sg_dma_len(sg); + struct device_dma_parameters *p = dev->dma_device->dma_parms; + + return p ? p->max_segment_size : UINT_MAX; } /** @@ -3946,9 +4020,17 @@ static inline bool ib_device_try_get(struct ib_device *dev) } void ib_device_put(struct ib_device *device); +struct ib_device *ib_device_get_by_netdev(struct net_device *ndev, + enum rdma_driver_id driver_id); +struct ib_device *ib_device_get_by_name(const char *name, + enum rdma_driver_id driver_id); struct net_device *ib_get_net_dev_by_params(struct ib_device *dev, u8 port, u16 pkey, const union ib_gid *gid, const struct sockaddr *addr); +int ib_device_set_netdev(struct ib_device *ib_dev, struct net_device *ndev, + unsigned int port); +struct net_device *ib_device_netdev(struct ib_device *dev, u8 port); + struct ib_wq *ib_create_wq(struct ib_pd *pd, struct ib_wq_init_attr *init_attr); int ib_destroy_wq(struct ib_wq *wq); @@ -4222,7 +4304,6 @@ void rdma_roce_rescan_device(struct ib_device *ibdev); struct ib_ucontext *ib_uverbs_get_ucontext_file(struct ib_uverbs_file *ufile); - int uverbs_destroy_def_handler(struct uverbs_attr_bundle *attrs); struct net_device *rdma_alloc_netdev(struct ib_device *device, u8 port_num, @@ -4258,4 +4339,27 @@ rdma_set_device_sysfs_group(struct ib_device *dev, dev->groups[1] = group; } +/** + * rdma_device_to_ibdev - Get ib_device pointer from device pointer + * + * @device: device pointer for which ib_device pointer to retrieve + * + * rdma_device_to_ibdev() retrieves ib_device pointer from device. + * + */ +static inline struct ib_device *rdma_device_to_ibdev(struct device *device) +{ + return container_of(device, struct ib_device, dev); +} + +/** + * rdma_device_to_drv_device - Helper macro to reach back to driver's + * ib_device holder structure from device pointer. + * + * NOTE: New drivers should not make use of this API; This API is only for + * existing drivers who have exposed sysfs entries using + * rdma_set_device_sysfs_group(). + */ +#define rdma_device_to_drv_device(dev, drv_dev_struct, ibdev_member) \ + container_of(rdma_device_to_ibdev(dev), drv_dev_struct, ibdev_member) #endif /* IB_VERBS_H */ diff --git a/include/rdma/iw_cm.h b/include/rdma/iw_cm.h index 5cd7701db148..0e1f02815643 100644 --- a/include/rdma/iw_cm.h +++ b/include/rdma/iw_cm.h @@ -94,7 +94,8 @@ struct iw_cm_id { void (*add_ref)(struct iw_cm_id *); void (*rem_ref)(struct iw_cm_id *); u8 tos; - bool mapped; + bool tos_set:1; + bool mapped:1; }; struct iw_cm_conn_param { @@ -105,6 +106,18 @@ struct iw_cm_conn_param { u32 qpn; }; +enum iw_flags { + + /* + * This flag allows the iwcm and iwpmd to still advertise + * mappings but the real and mapped port numbers are the + * same. Further, iwpmd will not bind any user socket to + * reserve the port. This is required for soft iwarp + * to play in the port mapped iwarp space. + */ + IW_F_NO_PORT_MAP = (1 << 0), +}; + struct iw_cm_verbs { void (*add_ref)(struct ib_qp *qp); @@ -127,6 +140,7 @@ struct iw_cm_verbs { int (*destroy_listen)(struct iw_cm_id *cm_id); char ifname[IFNAMSIZ]; + enum iw_flags driver_flags; }; /** diff --git a/include/rdma/iw_portmap.h b/include/rdma/iw_portmap.h index fda31673a562..b9fee7feeeb5 100644 --- a/include/rdma/iw_portmap.h +++ b/include/rdma/iw_portmap.h @@ -58,167 +58,31 @@ struct iwpm_sa_data { struct sockaddr_storage mapped_loc_addr; struct sockaddr_storage rem_addr; struct sockaddr_storage mapped_rem_addr; + u32 flags; }; -/** - * iwpm_init - Allocate resources for the iwarp port mapper - * - * Should be called when network interface goes up. - */ int iwpm_init(u8); - -/** - * iwpm_exit - Deallocate resources for the iwarp port mapper - * - * Should be called when network interface goes down. - */ int iwpm_exit(u8); - -/** - * iwpm_valid_pid - Check if the userspace iwarp port mapper pid is valid - * - * Returns true if the pid is greater than zero, otherwise returns false - */ int iwpm_valid_pid(void); - -/** - * iwpm_register_pid - Send a netlink query to userspace - * to get the iwarp port mapper pid - * @pm_msg: Contains driver info to send to the userspace port mapper - * @nl_client: The index of the netlink client - */ int iwpm_register_pid(struct iwpm_dev_data *pm_msg, u8 nl_client); - -/** - * iwpm_add_mapping - Send a netlink add mapping request to - * the userspace port mapper - * @pm_msg: Contains the local ip/tcp address info to send - * @nl_client: The index of the netlink client - * - * If the request is successful, the pm_msg stores - * the port mapper response (mapped address info) - */ int iwpm_add_mapping(struct iwpm_sa_data *pm_msg, u8 nl_client); - -/** - * iwpm_add_and_query_mapping - Send a netlink add and query mapping request - * to the userspace port mapper - * @pm_msg: Contains the local and remote ip/tcp address info to send - * @nl_client: The index of the netlink client - * - * If the request is successful, the pm_msg stores the - * port mapper response (mapped local and remote address info) - */ int iwpm_add_and_query_mapping(struct iwpm_sa_data *pm_msg, u8 nl_client); - -/** - * iwpm_remove_mapping - Send a netlink remove mapping request - * to the userspace port mapper - * - * @local_addr: Local ip/tcp address to remove - * @nl_client: The index of the netlink client - */ int iwpm_remove_mapping(struct sockaddr_storage *local_addr, u8 nl_client); - -/** - * iwpm_register_pid_cb - Process the port mapper response to - * iwpm_register_pid query - * @skb: - * @cb: Contains the received message (payload and netlink header) - * - * If successful, the function receives the userspace port mapper pid - * which is used in future communication with the port mapper - */ int iwpm_register_pid_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_add_mapping_cb - Process the port mapper response to - * iwpm_add_mapping request - * @skb: - * @cb: Contains the received message (payload and netlink header) - */ int iwpm_add_mapping_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_add_and_query_mapping_cb - Process the port mapper response to - * iwpm_add_and_query_mapping request - * @skb: - * @cb: Contains the received message (payload and netlink header) - */ int iwpm_add_and_query_mapping_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_remote_info_cb - Process remote connecting peer address info, which - * the port mapper has received from the connecting peer - * - * @cb: Contains the received message (payload and netlink header) - * - * Stores the IPv4/IPv6 address info in a hash table - */ int iwpm_remote_info_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_mapping_error_cb - Process port mapper notification for error - * - * @skb: - * @cb: Contains the received message (payload and netlink header) - */ int iwpm_mapping_error_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_mapping_info_cb - Process a notification that the userspace - * port mapper daemon is started - * @skb: - * @cb: Contains the received message (payload and netlink header) - * - * Using the received port mapper pid, send all the local mapping - * info records to the userspace port mapper - */ int iwpm_mapping_info_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_ack_mapping_info_cb - Process the port mapper ack for - * the provided local mapping info records - * @skb: - * @cb: Contains the received message (payload and netlink header) - */ int iwpm_ack_mapping_info_cb(struct sk_buff *, struct netlink_callback *); - -/** - * iwpm_get_remote_info - Get the remote connecting peer address info - * - * @mapped_loc_addr: Mapped local address of the listening peer - * @mapped_rem_addr: Mapped remote address of the connecting peer - * @remote_addr: To store the remote address of the connecting peer - * @nl_client: The index of the netlink client - * - * The remote address info is retrieved and provided to the client in - * the remote_addr. After that it is removed from the hash table - */ int iwpm_get_remote_info(struct sockaddr_storage *mapped_loc_addr, struct sockaddr_storage *mapped_rem_addr, struct sockaddr_storage *remote_addr, u8 nl_client); - -/** - * iwpm_create_mapinfo - Store local and mapped IPv4/IPv6 address - * info in a hash table - * @local_addr: Local ip/tcp address - * @mapped_addr: Mapped local ip/tcp address - * @nl_client: The index of the netlink client - */ int iwpm_create_mapinfo(struct sockaddr_storage *local_addr, - struct sockaddr_storage *mapped_addr, u8 nl_client); - -/** - * iwpm_remove_mapinfo - Remove local and mapped IPv4/IPv6 address - * info from the hash table - * @local_addr: Local ip/tcp address - * @mapped_addr: Mapped local ip/tcp address - * - * Returns err code if mapping info is not found in the hash table, - * otherwise returns 0 - */ + struct sockaddr_storage *mapped_addr, u8 nl_client, + u32 map_flags); int iwpm_remove_mapinfo(struct sockaddr_storage *local_addr, struct sockaddr_storage *mapped_addr); +int iwpm_hello_cb(struct sk_buff *skb, struct netlink_callback *cb); #endif /* _IW_PORTMAP_H */ diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h index 60987a5903b7..71f48cfdc24c 100644 --- a/include/rdma/rdma_cm.h +++ b/include/rdma/rdma_cm.h @@ -374,6 +374,7 @@ int rdma_set_reuseaddr(struct rdma_cm_id *id, int reuse); */ int rdma_set_afonly(struct rdma_cm_id *id, int afonly); +int rdma_set_ack_timeout(struct rdma_cm_id *id, u8 timeout); /** * rdma_get_service_id - Return the IB service ID for a specified address. * @id: Communication identifier associated with the address. diff --git a/include/rdma/rdma_netlink.h b/include/rdma/rdma_netlink.h index 70218e6b5187..10732ab31ba2 100644 --- a/include/rdma/rdma_netlink.h +++ b/include/rdma/rdma_netlink.h @@ -99,4 +99,15 @@ int rdma_nl_multicast(struct sk_buff *skb, unsigned int group, gfp_t flags); * Returns true on success or false if no listeners. */ bool rdma_nl_chk_listeners(unsigned int group); + +struct rdma_link_ops { + struct list_head list; + const char *type; + int (*newlink)(const char *ibdev_name, struct net_device *ndev); +}; + +void rdma_link_register(struct rdma_link_ops *ops); +void rdma_link_unregister(struct rdma_link_ops *ops); + +#define MODULE_ALIAS_RDMA_LINK(type) MODULE_ALIAS("rdma-link-" type) #endif /* _RDMA_NETLINK_H */ diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h index dd0ed8048bb4..4c257aff7d32 100644 --- a/include/rdma/rdma_vt.h +++ b/include/rdma/rdma_vt.h @@ -182,9 +182,15 @@ struct rvt_driver_params { u32 max_mad_size; u8 qos_shift; u8 max_rdma_atomic; + u8 extra_rdma_atomic; u8 reserved_operations; }; +/* User context */ +struct rvt_ucontext { + struct ib_ucontext ibucontext; +}; + /* Protection domain */ struct rvt_pd { struct ib_pd ibpd; @@ -250,9 +256,6 @@ struct rvt_driver_provided { */ void (*do_send)(struct rvt_qp *qp); - /* Passed to ib core registration. Callback to create syfs files */ - int (*port_callback)(struct ib_device *, u8, struct kobject *); - /* * Returns a pointer to the undelying hardware's PCI device. This is * used to display information as to what hardware is being referenced @@ -522,7 +525,14 @@ static inline unsigned rvt_get_npkeys(struct rvt_dev_info *rdi) */ static inline unsigned int rvt_max_atomic(struct rvt_dev_info *rdi) { - return rdi->dparms.max_rdma_atomic + 1; + return rdi->dparms.max_rdma_atomic + + rdi->dparms.extra_rdma_atomic + 1; +} + +static inline unsigned int rvt_size_atomic(struct rvt_dev_info *rdi) +{ + return rdi->dparms.max_rdma_atomic + + rdi->dparms.extra_rdma_atomic; } /* @@ -569,9 +579,10 @@ static inline struct rvt_qp *rvt_lookup_qpn(struct rvt_dev_info *rdi, /** * rvt_mod_retry_timer - mod a retry timer * @qp - the QP + * @shift - timeout shift to wait for multiple packets * Modify a potentially already running retry timer */ -static inline void rvt_mod_retry_timer(struct rvt_qp *qp) +static inline void rvt_mod_retry_timer_ext(struct rvt_qp *qp, u8 shift) { struct ib_qp *ibqp = &qp->ibqp; struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); @@ -579,8 +590,13 @@ static inline void rvt_mod_retry_timer(struct rvt_qp *qp) 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); + mod_timer(&qp->s_timer, jiffies + rdi->busy_jiffies + + (qp->timeout_jiffies << shift)); +} + +static inline void rvt_mod_retry_timer(struct rvt_qp *qp) +{ + return rvt_mod_retry_timer_ext(qp, 0); } struct rvt_dev_info *rvt_alloc_device(size_t size, int nports); diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index cbafb1878669..f0fbd4063fef 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -174,6 +174,7 @@ struct rvt_swqe { u32 lpsn; /* last packet sequence number */ u32 ssn; /* send sequence number */ u32 length; /* total length of data in sg_list */ + void *priv; /* driver dependent field */ struct rvt_sge sg_list[0]; }; @@ -235,6 +236,7 @@ struct rvt_ack_entry { u32 lpsn; u8 opcode; u8 sent; + void *priv; }; #define RC_QP_SCALING_INTERVAL 5 @@ -244,6 +246,7 @@ struct rvt_ack_entry { #define RVT_OPERATION_ATOMIC_SGE 0x00000004 #define RVT_OPERATION_LOCAL 0x00000008 #define RVT_OPERATION_USE_RESERVE 0x00000010 +#define RVT_OPERATION_IGN_RNR_CNT 0x00000020 #define RVT_OPERATION_MAX (IB_WR_RESERVED10 + 1) @@ -373,6 +376,7 @@ struct rvt_qp { u8 s_rnr_retry; /* requester RNR retry counter */ u8 s_num_rd_atomic; /* number of RDMA read/atomic pending */ u8 s_tail_ack_queue; /* index into s_ack_queue[] */ + u8 s_acked_ack_queue; /* index into s_ack_queue[] */ struct rvt_sge_state s_ack_rdma_sge; struct timer_list s_timer; @@ -629,6 +633,16 @@ __be32 rvt_compute_aeth(struct rvt_qp *qp); void rvt_get_credit(struct rvt_qp *qp, u32 aeth); /** + * rvt_restart_sge - rewind the sge state for a wqe + * @ss: the sge state pointer + * @wqe: the wqe to rewind + * @len: the data length from the start of the wqe in bytes + * + * Returns the remaining data length. + */ +u32 rvt_restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, u32 len); + +/** * @qp - the qp pair * @len - the length * @@ -676,7 +690,11 @@ 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); +void rvt_add_retry_timer_ext(struct rvt_qp *qp, u8 shift); +static inline void rvt_add_retry_timer(struct rvt_qp *qp) +{ + rvt_add_retry_timer_ext(qp, 0); +} void rvt_copy_sge(struct rvt_qp *qp, struct rvt_sge_state *ss, void *data, u32 length, diff --git a/include/rdma/restrack.h b/include/rdma/restrack.h index 8f179be9d9a9..ecf3c7702a4f 100644 --- a/include/rdma/restrack.h +++ b/include/rdma/restrack.h @@ -7,12 +7,12 @@ #define _RDMA_RESTRACK_H_ #include <linux/typecheck.h> -#include <linux/rwsem.h> #include <linux/sched.h> #include <linux/kref.h> #include <linux/completion.h> #include <linux/sched/task.h> #include <uapi/rdma/rdma_netlink.h> +#include <linux/xarray.h> /** * enum rdma_restrack_type - HW objects to track @@ -48,30 +48,7 @@ enum rdma_restrack_type { RDMA_RESTRACK_MAX }; -#define RDMA_RESTRACK_HASH_BITS 8 -struct rdma_restrack_entry; - -/** - * struct rdma_restrack_root - main resource tracking management - * entity, per-device - */ -struct rdma_restrack_root { - /* - * @rwsem: Read/write lock to protect lists - */ - struct rw_semaphore rwsem; - /** - * @hash: global database for all resources per-device - */ - DECLARE_HASHTABLE(hash, RDMA_RESTRACK_HASH_BITS); - /** - * @fill_res_entry: driver-specific fill function - * - * Allows rdma drivers to add their own restrack attributes. - */ - int (*fill_res_entry)(struct sk_buff *msg, - struct rdma_restrack_entry *entry); -}; +struct ib_device; /** * struct rdma_restrack_entry - metadata per-entry @@ -109,10 +86,6 @@ struct rdma_restrack_entry { */ const char *kern_name; /** - * @node: hash table entry - */ - struct hlist_node node; - /** * @type: various objects in restrack database */ enum rdma_restrack_type type; @@ -120,27 +93,13 @@ struct rdma_restrack_entry { * @user: user resource */ bool user; + /** + * @id: ID to expose to users + */ + u32 id; }; -/** - * rdma_restrack_init() - initialize resource tracking - * @res: resource tracking root - */ -void rdma_restrack_init(struct rdma_restrack_root *res); - -/** - * rdma_restrack_clean() - clean resource tracking - * @res: resource tracking root - */ -void rdma_restrack_clean(struct rdma_restrack_root *res); - -/** - * rdma_restrack_count() - the current usage of specific object - * @res: resource entry - * @type: actual type of object to operate - * @ns: PID namespace - */ -int rdma_restrack_count(struct rdma_restrack_root *res, +int rdma_restrack_count(struct ib_device *dev, enum rdma_restrack_type type, struct pid_namespace *ns); @@ -193,4 +152,7 @@ int rdma_nl_put_driver_u32_hex(struct sk_buff *msg, const char *name, int rdma_nl_put_driver_u64(struct sk_buff *msg, const char *name, u64 value); int rdma_nl_put_driver_u64_hex(struct sk_buff *msg, const char *name, u64 value); +struct rdma_restrack_entry *rdma_restrack_get_byid(struct ib_device *dev, + enum rdma_restrack_type type, + u32 id); #endif /* _RDMA_RESTRACK_H_ */ diff --git a/include/rdma/tid_rdma_defs.h b/include/rdma/tid_rdma_defs.h new file mode 100644 index 000000000000..08fe47c7ad2c --- /dev/null +++ b/include/rdma/tid_rdma_defs.h @@ -0,0 +1,108 @@ +/* SPDX-License-Identifier: (GPL-2.0 OR BSD-3-Clause) */ +/* + * Copyright(c) 2018 Intel Corporation. + * + */ + +#ifndef TID_RDMA_DEFS_H +#define TID_RDMA_DEFS_H + +#include <rdma/ib_pack.h> + +struct tid_rdma_read_req { + __le32 kdeth0; + __le32 kdeth1; + struct ib_reth reth; + __be32 tid_flow_psn; + __be32 tid_flow_qp; + __be32 verbs_qp; +}; + +struct tid_rdma_read_resp { + __le32 kdeth0; + __le32 kdeth1; + __be32 aeth; + __be32 reserved[4]; + __be32 verbs_psn; + __be32 verbs_qp; +}; + +struct tid_rdma_write_req { + __le32 kdeth0; + __le32 kdeth1; + struct ib_reth reth; + __be32 reserved[2]; + __be32 verbs_qp; +}; + +struct tid_rdma_write_resp { + __le32 kdeth0; + __le32 kdeth1; + __be32 aeth; + __be32 reserved[3]; + __be32 tid_flow_psn; + __be32 tid_flow_qp; + __be32 verbs_qp; +}; + +struct tid_rdma_write_data { + __le32 kdeth0; + __le32 kdeth1; + __be32 reserved[6]; + __be32 verbs_qp; +}; + +struct tid_rdma_resync { + __le32 kdeth0; + __le32 kdeth1; + __be32 reserved[6]; + __be32 verbs_qp; +}; + +struct tid_rdma_ack { + __le32 kdeth0; + __le32 kdeth1; + __be32 aeth; + __be32 reserved[2]; + __be32 tid_flow_psn; + __be32 verbs_psn; + __be32 tid_flow_qp; + __be32 verbs_qp; +}; + +/* + * TID RDMA Opcodes + */ +#define IB_OPCODE_TID_RDMA 0xe0 +enum { + IB_OPCODE_WRITE_REQ = 0x0, + IB_OPCODE_WRITE_RESP = 0x1, + IB_OPCODE_WRITE_DATA = 0x2, + IB_OPCODE_WRITE_DATA_LAST = 0x3, + IB_OPCODE_READ_REQ = 0x4, + IB_OPCODE_READ_RESP = 0x5, + IB_OPCODE_RESYNC = 0x6, + IB_OPCODE_ACK = 0x7, + + IB_OPCODE(TID_RDMA, WRITE_REQ), + IB_OPCODE(TID_RDMA, WRITE_RESP), + IB_OPCODE(TID_RDMA, WRITE_DATA), + IB_OPCODE(TID_RDMA, WRITE_DATA_LAST), + IB_OPCODE(TID_RDMA, READ_REQ), + IB_OPCODE(TID_RDMA, READ_RESP), + IB_OPCODE(TID_RDMA, RESYNC), + IB_OPCODE(TID_RDMA, ACK), +}; + +#define TID_OP(x) IB_OPCODE_TID_RDMA_##x + +/* + * Define TID RDMA specific WR opcodes. The ib_wr_opcode + * enum already provides some reserved values for use by + * low level drivers. Two of those are used but renamed + * to be more descriptive. + */ +#define IB_WR_TID_RDMA_WRITE IB_WR_RESERVED1 +#define IB_WR_TID_RDMA_READ IB_WR_RESERVED2 + +#endif /* TID_RDMA_DEFS_H */ diff --git a/include/rdma/uverbs_ioctl.h b/include/rdma/uverbs_ioctl.h index 27da906beea7..28570ac2b6a0 100644 --- a/include/rdma/uverbs_ioctl.h +++ b/include/rdma/uverbs_ioctl.h @@ -652,6 +652,7 @@ struct uverbs_attr_bundle { struct ib_udata driver_udata; struct ib_udata ucore; struct ib_uverbs_file *ufile; + struct ib_ucontext *context; DECLARE_BITMAP(attr_present, UVERBS_API_ATTR_BKEY_LEN); struct uverbs_attr attrs[]; }; @@ -663,6 +664,23 @@ static inline bool uverbs_attr_is_valid(const struct uverbs_attr_bundle *attrs_b attrs_bundle->attr_present); } +/** + * rdma_udata_to_drv_context - Helper macro to get the driver's context out of + * ib_udata which is embedded in uverbs_attr_bundle. + * + * If udata is not NULL this cannot fail. Otherwise a NULL udata will result + * in a NULL ucontext pointer, as a safety precaution. Callers should be using + * 'udata' to determine if the driver call is in user or kernel mode, not + * 'ucontext'. + * + */ +#define rdma_udata_to_drv_context(udata, drv_dev_struct, member) \ + (udata ? container_of(container_of(udata, struct uverbs_attr_bundle, \ + driver_udata) \ + ->context, \ + drv_dev_struct, member) : \ + (drv_dev_struct *)NULL) + #define IS_UVERBS_COPY_ERR(_ret) ((_ret) && (_ret) != -ENOENT) static inline const struct uverbs_attr *uverbs_attr_get(const struct uverbs_attr_bundle *attrs_bundle, diff --git a/include/rdma/uverbs_std_types.h b/include/rdma/uverbs_std_types.h index 883abcf6d36e..794c47565971 100644 --- a/include/rdma/uverbs_std_types.h +++ b/include/rdma/uverbs_std_types.h @@ -48,9 +48,12 @@ #define uobj_get_type(_attrs, _object) \ uapi_get_object((_attrs)->ufile->device->uapi, _object) +struct ib_uobject *_uobj_get_read(enum uverbs_default_objects type, + u32 object_id, + struct uverbs_attr_bundle *attrs); + #define uobj_get_read(_type, _id, _attrs) \ - rdma_lookup_get_uobject(uobj_get_type(_attrs, _type), (_attrs)->ufile, \ - _uobj_check_id(_id), UVERBS_LOOKUP_READ) + _uobj_get_read(_type, _uobj_check_id(_id), _attrs) #define ufd_get_read(_type, _fdnum, _attrs) \ rdma_lookup_get_uobject(uobj_get_type(_attrs, _type), (_attrs)->ufile, \ @@ -67,9 +70,12 @@ static inline void *_uobj_get_obj_read(struct ib_uobject *uobj) ((struct ib_##_object *)_uobj_get_obj_read( \ uobj_get_read(_type, _id, _attrs))) +struct ib_uobject *_uobj_get_write(enum uverbs_default_objects type, + u32 object_id, + struct uverbs_attr_bundle *attrs); + #define uobj_get_write(_type, _id, _attrs) \ - rdma_lookup_get_uobject(uobj_get_type(_attrs, _type), (_attrs)->ufile, \ - _uobj_check_id(_id), UVERBS_LOOKUP_WRITE) + _uobj_get_write(_type, _uobj_check_id(_id), _attrs) int __uobj_perform_destroy(const struct uverbs_api_object *obj, u32 id, const struct uverbs_attr_bundle *attrs); @@ -123,8 +129,10 @@ __uobj_alloc(const struct uverbs_api_object *obj, { struct ib_uobject *uobj = rdma_alloc_begin_uobject(obj, attrs->ufile); - if (!IS_ERR(uobj)) + if (!IS_ERR(uobj)) { *ib_dev = uobj->context->device; + attrs->context = uobj->context; + } return uobj; } diff --git a/include/rdma/uverbs_types.h b/include/rdma/uverbs_types.h index acb1bfa3cc99..175d761695e1 100644 --- a/include/rdma/uverbs_types.h +++ b/include/rdma/uverbs_types.h @@ -157,6 +157,7 @@ struct uverbs_obj_fd_type { extern const struct uverbs_obj_type_class uverbs_idr_class; extern const struct uverbs_obj_type_class uverbs_fd_class; +void uverbs_close_fd(struct file *f); #define UVERBS_BUILD_BUG_ON(cond) (sizeof(char[1 - 2 * !!(cond)]) - \ sizeof(char)) |