diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2020-06-01 12:00:10 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2020-06-01 12:00:10 -0700 |
commit | 81e8c10dac62c427b25f6bbb07936806e4dd9e8a (patch) | |
tree | 9eea9ad000cfdb728e4385873ebba73222c80d93 | |
parent | 729ea4e064202aeec149b034b459501ef0a5060e (diff) | |
parent | 58ca0060ec4e51208d2eee12198fc55fd9e4feb3 (diff) | |
download | linux-81e8c10dac62c427b25f6bbb07936806e4dd9e8a.tar.gz linux-81e8c10dac62c427b25f6bbb07936806e4dd9e8a.tar.bz2 linux-81e8c10dac62c427b25f6bbb07936806e4dd9e8a.zip |
Merge branch 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Pull crypto updates from Herbert Xu:
"API:
- Introduce crypto_shash_tfm_digest() and use it wherever possible.
- Fix use-after-free and race in crypto_spawn_alg.
- Add support for parallel and batch requests to crypto_engine.
Algorithms:
- Update jitter RNG for SP800-90B compliance.
- Always use jitter RNG as seed in drbg.
Drivers:
- Add Arm CryptoCell driver cctrng.
- Add support for SEV-ES to the PSP driver in ccp"
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (114 commits)
crypto: hisilicon - fix driver compatibility issue with different versions of devices
crypto: engine - do not requeue in case of fatal error
crypto: cavium/nitrox - Fix a typo in a comment
crypto: hisilicon/qm - change debugfs file name from qm_regs to regs
crypto: hisilicon/qm - add DebugFS for xQC and xQE dump
crypto: hisilicon/zip - add debugfs for Hisilicon ZIP
crypto: hisilicon/hpre - add debugfs for Hisilicon HPRE
crypto: hisilicon/sec2 - add debugfs for Hisilicon SEC
crypto: hisilicon/qm - add debugfs to the QM state machine
crypto: hisilicon/qm - add debugfs for QM
crypto: stm32/crc32 - protect from concurrent accesses
crypto: stm32/crc32 - don't sleep in runtime pm
crypto: stm32/crc32 - fix multi-instance
crypto: stm32/crc32 - fix run-time self test issue.
crypto: stm32/crc32 - fix ext4 chksum BUG_ON()
crypto: hisilicon/zip - Use temporary sqe when doing work
crypto: hisilicon - add device error report through abnormal irq
crypto: hisilicon - remove codes of directly report device errors through MSI
crypto: hisilicon - QM memory management optimization
crypto: hisilicon - unify initial value assignment into QM
...
127 files changed, 4966 insertions, 1794 deletions
diff --git a/Documentation/ABI/testing/debugfs-hisi-hpre b/Documentation/ABI/testing/debugfs-hisi-hpre index ec4a79e3a807..b4be5f1db4b7 100644 --- a/Documentation/ABI/testing/debugfs-hisi-hpre +++ b/Documentation/ABI/testing/debugfs-hisi-hpre @@ -33,7 +33,7 @@ Contact: linux-crypto@vger.kernel.org Description: Dump debug registers from the HPRE. Only available for PF. -What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/qm_regs +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/regs Date: Sep 2019 Contact: linux-crypto@vger.kernel.org Description: Dump debug registers from the QM. @@ -44,14 +44,97 @@ What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/current_q Date: Sep 2019 Contact: linux-crypto@vger.kernel.org Description: One QM may contain multiple queues. Select specific queue to - show its debug registers in above qm_regs. + show its debug registers in above regs. Only available for PF. What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/clear_enable Date: Sep 2019 Contact: linux-crypto@vger.kernel.org -Description: QM debug registers(qm_regs) read clear control. 1 means enable +Description: QM debug registers(regs) read clear control. 1 means enable register read clear, otherwise 0. Writing to this file has no functional effect, only enable or disable counters clear after reading of these registers. Only available for PF. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/err_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of invalid interrupts for + QM task completion. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/aeq_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of QM async event queue interrupts. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/abnormal_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of interrupts for QM abnormal event. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/create_qp_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of queue allocation errors. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/mb_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of failed QM mailbox commands. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/qm/status +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the status of the QM. + Four states: initiated, started, stopped and closed. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/send_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of sent requests. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/recv_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of received requests. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/send_busy_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of requests sent + with returning busy. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/send_fail_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of completed but error requests. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/invalid_req_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of invalid requests being received. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/overtime_thrhld +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Set the threshold time for counting the request which is + processed longer than the threshold. + 0: disable(default), 1: 1 microsecond. + Available for both PF and VF, and take no other effect on HPRE. + +What: /sys/kernel/debug/hisi_hpre/<bdf>/hpre_dfx/over_thrhld_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of time out requests. + Available for both PF and VF, and take no other effect on HPRE. diff --git a/Documentation/ABI/testing/debugfs-hisi-sec b/Documentation/ABI/testing/debugfs-hisi-sec index 06adb899495e..85feb4408e0f 100644 --- a/Documentation/ABI/testing/debugfs-hisi-sec +++ b/Documentation/ABI/testing/debugfs-hisi-sec @@ -1,10 +1,4 @@ -What: /sys/kernel/debug/hisi_sec/<bdf>/sec_dfx -Date: Oct 2019 -Contact: linux-crypto@vger.kernel.org -Description: Dump the debug registers of SEC cores. - Only available for PF. - -What: /sys/kernel/debug/hisi_sec/<bdf>/clear_enable +What: /sys/kernel/debug/hisi_sec2/<bdf>/clear_enable Date: Oct 2019 Contact: linux-crypto@vger.kernel.org Description: Enabling/disabling of clear action after reading @@ -12,7 +6,7 @@ Description: Enabling/disabling of clear action after reading 0: disable, 1: enable. Only available for PF, and take no other effect on SEC. -What: /sys/kernel/debug/hisi_sec/<bdf>/current_qm +What: /sys/kernel/debug/hisi_sec2/<bdf>/current_qm Date: Oct 2019 Contact: linux-crypto@vger.kernel.org Description: One SEC controller has one PF and multiple VFs, each function @@ -20,24 +14,100 @@ Description: One SEC controller has one PF and multiple VFs, each function qm refers to. Only available for PF. -What: /sys/kernel/debug/hisi_sec/<bdf>/qm/qm_regs +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/qm_regs Date: Oct 2019 Contact: linux-crypto@vger.kernel.org Description: Dump of QM related debug registers. Available for PF and VF in host. VF in guest currently only has one debug register. -What: /sys/kernel/debug/hisi_sec/<bdf>/qm/current_q +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/current_q Date: Oct 2019 Contact: linux-crypto@vger.kernel.org Description: One QM of SEC may contain multiple queues. Select specific - queue to show its debug registers in above 'qm_regs'. + queue to show its debug registers in above 'regs'. Only available for PF. -What: /sys/kernel/debug/hisi_sec/<bdf>/qm/clear_enable +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/clear_enable Date: Oct 2019 Contact: linux-crypto@vger.kernel.org Description: Enabling/disabling of clear action after reading the SEC's QM debug registers. 0: disable, 1: enable. Only available for PF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/err_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of invalid interrupts for + QM task completion. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/aeq_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of QM async event queue interrupts. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/abnormal_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of interrupts for QM abnormal event. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/create_qp_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of queue allocation errors. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/mb_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of failed QM mailbox commands. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/qm/status +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the status of the QM. + Four states: initiated, started, stopped and closed. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/send_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of sent requests. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/recv_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of received requests. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/send_busy_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of requests sent with returning busy. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/err_bd_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of BD type error requests + to be received. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/invalid_req_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of invalid requests being received. + Available for both PF and VF, and take no other effect on SEC. + +What: /sys/kernel/debug/hisi_sec2/<bdf>/sec_dfx/done_flag_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of completed but marked error requests + to be received. + Available for both PF and VF, and take no other effect on SEC. diff --git a/Documentation/ABI/testing/debugfs-hisi-zip b/Documentation/ABI/testing/debugfs-hisi-zip index a7c63e6c4bc3..3034a2bf99ca 100644 --- a/Documentation/ABI/testing/debugfs-hisi-zip +++ b/Documentation/ABI/testing/debugfs-hisi-zip @@ -26,7 +26,7 @@ Description: One ZIP controller has one PF and multiple VFs, each function has a QM. Select the QM which below qm refers to. Only available for PF. -What: /sys/kernel/debug/hisi_zip/<bdf>/qm/qm_regs +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/regs Date: Nov 2018 Contact: linux-crypto@vger.kernel.org Description: Dump of QM related debug registers. @@ -37,14 +37,78 @@ What: /sys/kernel/debug/hisi_zip/<bdf>/qm/current_q Date: Nov 2018 Contact: linux-crypto@vger.kernel.org Description: One QM may contain multiple queues. Select specific queue to - show its debug registers in above qm_regs. + show its debug registers in above regs. Only available for PF. What: /sys/kernel/debug/hisi_zip/<bdf>/qm/clear_enable Date: Nov 2018 Contact: linux-crypto@vger.kernel.org -Description: QM debug registers(qm_regs) read clear control. 1 means enable +Description: QM debug registers(regs) read clear control. 1 means enable register read clear, otherwise 0. Writing to this file has no functional effect, only enable or disable counters clear after reading of these registers. Only available for PF. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/err_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of invalid interrupts for + QM task completion. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/aeq_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of QM async event queue interrupts. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/abnormal_irq +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of interrupts for QM abnormal event. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/create_qp_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of queue allocation errors. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/mb_err +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the number of failed QM mailbox commands. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/qm/status +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the status of the QM. + Four states: initiated, started, stopped and closed. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/zip_dfx/send_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of sent requests. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/zip_dfx/recv_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of received requests. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/zip_dfx/send_busy_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of requests received + with returning busy. + Available for both PF and VF, and take no other effect on ZIP. + +What: /sys/kernel/debug/hisi_zip/<bdf>/zip_dfx/err_bd_cnt +Date: Apr 2020 +Contact: linux-crypto@vger.kernel.org +Description: Dump the total number of BD type error requests + to be received. + Available for both PF and VF, and take no other effect on ZIP. diff --git a/Documentation/devicetree/bindings/rng/arm-cctrng.yaml b/Documentation/devicetree/bindings/rng/arm-cctrng.yaml new file mode 100644 index 000000000000..ca6aad19b6ba --- /dev/null +++ b/Documentation/devicetree/bindings/rng/arm-cctrng.yaml @@ -0,0 +1,54 @@ +# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/rng/arm-cctrng.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Arm TrustZone CryptoCell TRNG engine + +maintainers: + - Hadar Gat <hadar.gat@arm.com> + +description: |+ + Arm TrustZone CryptoCell TRNG (True Random Number Generator) engine. + +properties: + compatible: + enum: + - arm,cryptocell-713-trng + - arm,cryptocell-703-trng + + interrupts: + maxItems: 1 + + reg: + maxItems: 1 + + arm,rosc-ratio: + description: + Arm TrustZone CryptoCell TRNG engine has 4 ring oscillators. + Sampling ratio values for these 4 ring oscillators. (from calibration) + allOf: + - $ref: /schemas/types.yaml#/definitions/uint32-array + - items: + maxItems: 4 + + clocks: + maxItems: 1 + +required: + - compatible + - interrupts + - reg + - arm,rosc-ratio + +additionalProperties: false + +examples: + - | + arm_cctrng: rng@60000000 { + compatible = "arm,cryptocell-713-trng"; + interrupts = <0 29 4>; + reg = <0x60000000 0x10000>; + arm,rosc-ratio = <5000 1000 500 0>; + }; diff --git a/Documentation/security/siphash.rst b/Documentation/security/siphash.rst index 4eba68cdf0a1..bd9363025fcb 100644 --- a/Documentation/security/siphash.rst +++ b/Documentation/security/siphash.rst @@ -7,7 +7,7 @@ SipHash - a short input PRF SipHash is a cryptographically secure PRF -- a keyed hash function -- that performs very well for short inputs, hence the name. It was designed by cryptographers Daniel J. Bernstein and Jean-Philippe Aumasson. It is intended -as a replacement for some uses of: `jhash`, `md5_transform`, `sha_transform`, +as a replacement for some uses of: `jhash`, `md5_transform`, `sha1_transform`, and so forth. SipHash takes a secret key filled with randomly generated numbers and either diff --git a/MAINTAINERS b/MAINTAINERS index cbfdf73029c6..2b17e667c52b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3908,6 +3908,15 @@ S: Supported W: https://developer.arm.com/products/system-ip/trustzone-cryptocell/cryptocell-700-family F: drivers/crypto/ccree/ +CCTRNG ARM TRUSTZONE CRYPTOCELL TRUE RANDOM NUMBER GENERATOR (TRNG) DRIVER +M: Hadar Gat <hadar.gat@arm.com> +L: linux-crypto@vger.kernel.org +S: Supported +F: drivers/char/hw_random/cctrng.c +F: drivers/char/hw_random/cctrng.h +F: Documentation/devicetree/bindings/rng/arm-cctrng.txt +W: https://developer.arm.com/products/system-ip/trustzone-cryptocell/cryptocell-700-family + CEC FRAMEWORK M: Hans Verkuil <hverkuil-cisco@xs4all.nl> L: linux-media@vger.kernel.org diff --git a/arch/arm/crypto/sha1_glue.c b/arch/arm/crypto/sha1_glue.c index c80b0ebfd02f..4e954b3f7ecd 100644 --- a/arch/arm/crypto/sha1_glue.c +++ b/arch/arm/crypto/sha1_glue.c @@ -14,7 +14,6 @@ #include <crypto/internal/hash.h> #include <linux/init.h> #include <linux/module.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <crypto/sha1_base.h> diff --git a/arch/arm/crypto/sha1_neon_glue.c b/arch/arm/crypto/sha1_neon_glue.c index 2c3627334335..0071e5e4411a 100644 --- a/arch/arm/crypto/sha1_neon_glue.c +++ b/arch/arm/crypto/sha1_neon_glue.c @@ -18,7 +18,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <crypto/sha1_base.h> diff --git a/arch/arm/crypto/sha256_glue.c b/arch/arm/crypto/sha256_glue.c index 215497f011f2..b8a4f79020cf 100644 --- a/arch/arm/crypto/sha256_glue.c +++ b/arch/arm/crypto/sha256_glue.c @@ -15,7 +15,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <linux/string.h> #include <crypto/sha.h> diff --git a/arch/arm/crypto/sha256_neon_glue.c b/arch/arm/crypto/sha256_neon_glue.c index 38645e415196..79820b9e2541 100644 --- a/arch/arm/crypto/sha256_neon_glue.c +++ b/arch/arm/crypto/sha256_neon_glue.c @@ -11,7 +11,6 @@ #include <crypto/internal/hash.h> #include <crypto/internal/simd.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <linux/string.h> #include <crypto/sha.h> diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c index 98bdea51089d..82e96ac83684 100644 --- a/arch/arm/kernel/armksyms.c +++ b/arch/arm/kernel/armksyms.c @@ -7,7 +7,6 @@ #include <linux/export.h> #include <linux/sched.h> #include <linux/string.h> -#include <linux/cryptohash.h> #include <linux/delay.h> #include <linux/in6.h> #include <linux/syscalls.h> diff --git a/arch/arm64/crypto/aes-glue.c b/arch/arm64/crypto/aes-glue.c index ed5409c6abf4..395bbf64b2ab 100644 --- a/arch/arm64/crypto/aes-glue.c +++ b/arch/arm64/crypto/aes-glue.c @@ -158,7 +158,6 @@ static int __maybe_unused essiv_cbc_set_key(struct crypto_skcipher *tfm, unsigned int key_len) { struct crypto_aes_essiv_cbc_ctx *ctx = crypto_skcipher_ctx(tfm); - SHASH_DESC_ON_STACK(desc, ctx->hash); u8 digest[SHA256_DIGEST_SIZE]; int ret; @@ -166,8 +165,7 @@ static int __maybe_unused essiv_cbc_set_key(struct crypto_skcipher *tfm, if (ret) return ret; - desc->tfm = ctx->hash; - crypto_shash_digest(desc, in_key, key_len, digest); + crypto_shash_tfm_digest(ctx->hash, in_key, key_len, digest); return aes_expandkey(&ctx->key2, digest, sizeof(digest)); } diff --git a/arch/arm64/crypto/crct10dif-ce-core.S b/arch/arm64/crypto/crct10dif-ce-core.S index 5a95c2628fbf..111d9c9abddd 100644 --- a/arch/arm64/crypto/crct10dif-ce-core.S +++ b/arch/arm64/crypto/crct10dif-ce-core.S @@ -66,7 +66,7 @@ #include <asm/assembler.h> .text - .cpu generic+crypto + .arch armv8-a+crypto init_crc .req w19 buf .req x20 diff --git a/arch/arm64/crypto/sha256-glue.c b/arch/arm64/crypto/sha256-glue.c index ddf4a0d85c1c..77bc6e72abae 100644 --- a/arch/arm64/crypto/sha256-glue.c +++ b/arch/arm64/crypto/sha256-glue.c @@ -12,7 +12,6 @@ #include <crypto/internal/simd.h> #include <crypto/sha.h> #include <crypto/sha256_base.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <linux/string.h> diff --git a/arch/arm64/crypto/sha512-glue.c b/arch/arm64/crypto/sha512-glue.c index 78d3083de6b7..370ccb29602f 100644 --- a/arch/arm64/crypto/sha512-glue.c +++ b/arch/arm64/crypto/sha512-glue.c @@ -6,7 +6,6 @@ */ #include <crypto/internal/hash.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <linux/string.h> #include <crypto/sha.h> diff --git a/arch/microblaze/kernel/microblaze_ksyms.c b/arch/microblaze/kernel/microblaze_ksyms.c index 92e12c2c2ec1..51c43ee5e380 100644 --- a/arch/microblaze/kernel/microblaze_ksyms.c +++ b/arch/microblaze/kernel/microblaze_ksyms.c @@ -6,7 +6,6 @@ #include <linux/export.h> #include <linux/string.h> -#include <linux/cryptohash.h> #include <linux/delay.h> #include <linux/in6.h> #include <linux/syscalls.h> diff --git a/arch/mips/cavium-octeon/crypto/octeon-md5.c b/arch/mips/cavium-octeon/crypto/octeon-md5.c index d1ed066e1a17..8c8ea139653e 100644 --- a/arch/mips/cavium-octeon/crypto/octeon-md5.c +++ b/arch/mips/cavium-octeon/crypto/octeon-md5.c @@ -25,7 +25,6 @@ #include <linux/module.h> #include <linux/string.h> #include <asm/byteorder.h> -#include <linux/cryptohash.h> #include <asm/octeon/octeon.h> #include <crypto/internal/hash.h> diff --git a/arch/powerpc/crypto/md5-glue.c b/arch/powerpc/crypto/md5-glue.c index 7d1bf2fcf668..c24f605033bd 100644 --- a/arch/powerpc/crypto/md5-glue.c +++ b/arch/powerpc/crypto/md5-glue.c @@ -11,7 +11,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/md5.h> #include <asm/byteorder.h> diff --git a/arch/powerpc/crypto/sha1-spe-glue.c b/arch/powerpc/crypto/sha1-spe-glue.c index 6379990bd604..cb57be4ada61 100644 --- a/arch/powerpc/crypto/sha1-spe-glue.c +++ b/arch/powerpc/crypto/sha1-spe-glue.c @@ -11,7 +11,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <asm/byteorder.h> diff --git a/arch/powerpc/crypto/sha1.c b/arch/powerpc/crypto/sha1.c index 7b43fc352089..b40dc50a6908 100644 --- a/arch/powerpc/crypto/sha1.c +++ b/arch/powerpc/crypto/sha1.c @@ -16,14 +16,13 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <asm/byteorder.h> -extern void powerpc_sha_transform(u32 *state, const u8 *src, u32 *temp); +void powerpc_sha_transform(u32 *state, const u8 *src); -static int sha1_init(struct shash_desc *desc) +static int powerpc_sha1_init(struct shash_desc *desc) { struct sha1_state *sctx = shash_desc_ctx(desc); @@ -34,8 +33,8 @@ static int sha1_init(struct shash_desc *desc) return 0; } -static int sha1_update(struct shash_desc *desc, const u8 *data, - unsigned int len) +static int powerpc_sha1_update(struct shash_desc *desc, const u8 *data, + unsigned int len) { struct sha1_state *sctx = shash_desc_ctx(desc); unsigned int partial, done; @@ -47,7 +46,6 @@ static int sha1_update(struct shash_desc *desc, const u8 *data, src = data; if ((partial + len) > 63) { - u32 temp[SHA_WORKSPACE_WORDS]; if (partial) { done = -partial; @@ -56,12 +54,11 @@ static int sha1_update(struct shash_desc *desc, const u8 *data, } do { - powerpc_sha_transform(sctx->state, src, temp); + powerpc_sha_transform(sctx->state, src); done += 64; src = data + done; } while (done + 63 < len); - memzero_explicit(temp, sizeof(temp)); partial = 0; } memcpy(sctx->buffer + partial, src, len - done); @@ -71,7 +68,7 @@ static int sha1_update(struct shash_desc *desc, const u8 *data, /* Add padding and return the message digest. */ -static int sha1_final(struct shash_desc *desc, u8 *out) +static int powerpc_sha1_final(struct shash_desc *desc, u8 *out) { struct sha1_state *sctx = shash_desc_ctx(desc); __be32 *dst = (__be32 *)out; @@ -84,10 +81,10 @@ static int sha1_final(struct shash_desc *desc, u8 *out) /* Pad out to 56 mod 64 */ index = sctx->count & 0x3f; padlen = (index < 56) ? (56 - index) : ((64+56) - index); - sha1_update(desc, padding, padlen); + powerpc_sha1_update(desc, padding, padlen); /* Append length */ - sha1_update(desc, (const u8 *)&bits, sizeof(bits)); + powerpc_sha1_update(desc, (const u8 *)&bits, sizeof(bits)); /* Store state in digest */ for (i = 0; i < 5; i++) @@ -99,7 +96,7 @@ static int sha1_final(struct shash_desc *desc, u8 *out) return 0; } -static int sha1_export(struct shash_desc *desc, void *out) +static int powerpc_sha1_export(struct shash_desc *desc, void *out) { struct sha1_state *sctx = shash_desc_ctx(desc); @@ -107,7 +104,7 @@ static int sha1_export(struct shash_desc *desc, void *out) return 0; } -static int sha1_import(struct shash_desc *desc, const void *in) +static int powerpc_sha1_import(struct shash_desc *desc, const void *in) { struct sha1_state *sctx = shash_desc_ctx(desc); @@ -117,11 +114,11 @@ static int sha1_import(struct shash_desc *desc, const void *in) static struct shash_alg alg = { .digestsize = SHA1_DIGEST_SIZE, - .init = sha1_init, - .update = sha1_update, - .final = sha1_final, - .export = sha1_export, - .import = sha1_import, + .init = powerpc_sha1_init, + .update = powerpc_sha1_update, + .final = powerpc_sha1_final, + .export = powerpc_sha1_export, + .import = powerpc_sha1_import, .descsize = sizeof(struct sha1_state), .statesize = sizeof(struct sha1_state), .base = { diff --git a/arch/powerpc/crypto/sha256-spe-glue.c b/arch/powerpc/crypto/sha256-spe-glue.c index 84939e563b81..ceb0b6c980b3 100644 --- a/arch/powerpc/crypto/sha256-spe-glue.c +++ b/arch/powerpc/crypto/sha256-spe-glue.c @@ -12,7 +12,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <asm/byteorder.h> diff --git a/arch/s390/crypto/sha1_s390.c b/arch/s390/crypto/sha1_s390.c index 7c15542d3685..698b1e6d3c14 100644 --- a/arch/s390/crypto/sha1_s390.c +++ b/arch/s390/crypto/sha1_s390.c @@ -27,7 +27,7 @@ #include "sha.h" -static int sha1_init(struct shash_desc *desc) +static int s390_sha1_init(struct shash_desc *desc) { struct s390_sha_ctx *sctx = shash_desc_ctx(desc); @@ -42,7 +42,7 @@ static int sha1_init(struct shash_desc *desc) return 0; } -static int sha1_export(struct shash_desc *desc, void *out) +static int s390_sha1_export(struct shash_desc *desc, void *out) { struct s390_sha_ctx *sctx = shash_desc_ctx(desc); struct sha1_state *octx = out; @@ -53,7 +53,7 @@ static int sha1_export(struct shash_desc *desc, void *out) return 0; } -static int sha1_import(struct shash_desc *desc, const void *in) +static int s390_sha1_import(struct shash_desc *desc, const void *in) { struct s390_sha_ctx *sctx = shash_desc_ctx(desc); const struct sha1_state *ictx = in; @@ -67,11 +67,11 @@ static int sha1_import(struct shash_desc *desc, const void *in) static struct shash_alg alg = { .digestsize = SHA1_DIGEST_SIZE, - .init = sha1_init, + .init = s390_sha1_init, .update = s390_sha_update, .final = s390_sha_final, - .export = sha1_export, - .import = sha1_import, + .export = s390_sha1_export, + .import = s390_sha1_import, .descsize = sizeof(struct s390_sha_ctx), .statesize = sizeof(struct sha1_state), .base = { diff --git a/arch/sparc/crypto/md5_glue.c b/arch/sparc/crypto/md5_glue.c index 14f6c15be6ae..111283fe837e 100644 --- a/arch/sparc/crypto/md5_glue.c +++ b/arch/sparc/crypto/md5_glue.c @@ -18,7 +18,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/md5.h> diff --git a/arch/sparc/crypto/sha1_glue.c b/arch/sparc/crypto/sha1_glue.c index 7c1666304441..dc017782be52 100644 --- a/arch/sparc/crypto/sha1_glue.c +++ b/arch/sparc/crypto/sha1_glue.c @@ -15,7 +15,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> diff --git a/arch/sparc/crypto/sha256_glue.c b/arch/sparc/crypto/sha256_glue.c index f403ce9ba6e4..286bc8ecf15b 100644 --- a/arch/sparc/crypto/sha256_glue.c +++ b/arch/sparc/crypto/sha256_glue.c @@ -15,7 +15,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> diff --git a/arch/sparc/crypto/sha512_glue.c b/arch/sparc/crypto/sha512_glue.c index a3b532e43c07..3b2ca732ff7a 100644 --- a/arch/sparc/crypto/sha512_glue.c +++ b/arch/sparc/crypto/sha512_glue.c @@ -14,7 +14,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> diff --git a/arch/unicore32/kernel/ksyms.c b/arch/unicore32/kernel/ksyms.c index f4b84872d640..731445008932 100644 --- a/arch/unicore32/kernel/ksyms.c +++ b/arch/unicore32/kernel/ksyms.c @@ -9,7 +9,6 @@ #include <linux/module.h> #include <linux/sched.h> #include <linux/string.h> -#include <linux/cryptohash.h> #include <linux/delay.h> #include <linux/in6.h> #include <linux/syscalls.h> diff --git a/arch/x86/crypto/sha1_ssse3_glue.c b/arch/x86/crypto/sha1_ssse3_glue.c index a801ffc10cbb..18200135603f 100644 --- a/arch/x86/crypto/sha1_ssse3_glue.c +++ b/arch/x86/crypto/sha1_ssse3_glue.c @@ -21,7 +21,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <crypto/sha1_base.h> diff --git a/arch/x86/crypto/sha256_ssse3_glue.c b/arch/x86/crypto/sha256_ssse3_glue.c index 6394b5fe8db6..dd06249229e1 100644 --- a/arch/x86/crypto/sha256_ssse3_glue.c +++ b/arch/x86/crypto/sha256_ssse3_glue.c @@ -34,7 +34,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <crypto/sha256_base.h> diff --git a/arch/x86/crypto/sha512_ssse3_glue.c b/arch/x86/crypto/sha512_ssse3_glue.c index 82cc1b3ced1d..b0b05c93409e 100644 --- a/arch/x86/crypto/sha512_ssse3_glue.c +++ b/arch/x86/crypto/sha512_ssse3_glue.c @@ -32,7 +32,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/string.h> #include <linux/types.h> #include <crypto/sha.h> diff --git a/crypto/Kconfig b/crypto/Kconfig index c24a47406f8f..d5daf35431e3 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -370,7 +370,6 @@ config CRYPTO_CFB config CRYPTO_CTR tristate "CTR support" select CRYPTO_SKCIPHER - select CRYPTO_SEQIV select CRYPTO_MANAGER help CTR: Counter mode @@ -1820,7 +1819,7 @@ config CRYPTO_DRBG_HASH config CRYPTO_DRBG_CTR bool "Enable CTR DRBG" select CRYPTO_AES - depends on CRYPTO_CTR + select CRYPTO_CTR help Enable the CTR DRBG variant as defined in NIST SP800-90A. diff --git a/crypto/algapi.c b/crypto/algapi.c index 69605e21af92..92abdf675992 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -403,7 +403,7 @@ static void crypto_wait_for_test(struct crypto_larval *larval) err = wait_for_completion_killable(&larval->completion); WARN_ON(err); if (!err) - crypto_probing_notify(CRYPTO_MSG_ALG_LOADED, larval); + crypto_notify(CRYPTO_MSG_ALG_LOADED, larval); out: crypto_larval_kill(&larval->alg); @@ -716,17 +716,27 @@ EXPORT_SYMBOL_GPL(crypto_drop_spawn); static struct crypto_alg *crypto_spawn_alg(struct crypto_spawn *spawn) { - struct crypto_alg *alg; + struct crypto_alg *alg = ERR_PTR(-EAGAIN); + struct crypto_alg *target; + bool shoot = false; down_read(&crypto_alg_sem); - alg = spawn->alg; - if (!spawn->dead && !crypto_mod_get(alg)) { - alg->cra_flags |= CRYPTO_ALG_DYING; - alg = NULL; + if (!spawn->dead) { + alg = spawn->alg; + if (!crypto_mod_get(alg)) { + target = crypto_alg_get(alg); + shoot = true; + alg = ERR_PTR(-EAGAIN); + } } up_read(&crypto_alg_sem); - return alg ?: ERR_PTR(-EAGAIN); + if (shoot) { + crypto_shoot_alg(target); + crypto_alg_put(target); + } + + return alg; } struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn, u32 type, @@ -904,6 +914,14 @@ out: } EXPORT_SYMBOL_GPL(crypto_enqueue_request); +void crypto_enqueue_request_head(struct crypto_queue *queue, + struct crypto_async_request *request) +{ + queue->qlen++; + list_add(&request->list, &queue->list); +} +EXPORT_SYMBOL_GPL(crypto_enqueue_request_head); + struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue) { struct list_head *request; diff --git a/crypto/algif_rng.c b/crypto/algif_rng.c index 22df3799a17b..087c0ad09d38 100644 --- a/crypto/algif_rng.c +++ b/crypto/algif_rng.c @@ -61,7 +61,7 @@ static int rng_recvmsg(struct socket *sock, struct msghdr *msg, size_t len, struct sock *sk = sock->sk; struct alg_sock *ask = alg_sk(sk); struct rng_ctx *ctx = ask->private; - int err = -EFAULT; + int err; int genlen = 0; u8 result[MAXSIZE]; diff --git a/crypto/api.c b/crypto/api.c index 7d71a9b10e5f..edcf690800d4 100644 --- a/crypto/api.c +++ b/crypto/api.c @@ -333,12 +333,13 @@ static unsigned int crypto_ctxsize(struct crypto_alg *alg, u32 type, u32 mask) return len; } -static void crypto_shoot_alg(struct crypto_alg *alg) +void crypto_shoot_alg(struct crypto_alg *alg) { down_write(&crypto_alg_sem); alg->cra_flags |= CRYPTO_ALG_DYING; up_write(&crypto_alg_sem); } +EXPORT_SYMBOL_GPL(crypto_shoot_alg); struct crypto_tfm *__crypto_alloc_tfm(struct crypto_alg *alg, u32 type, u32 mask) diff --git a/crypto/blake2b_generic.c b/crypto/blake2b_generic.c index 1d262374fa4e..0ffd8d92e308 100644 --- a/crypto/blake2b_generic.c +++ b/crypto/blake2b_generic.c @@ -129,7 +129,9 @@ static void blake2b_compress(struct blake2b_state *S, ROUND(9); ROUND(10); ROUND(11); - +#ifdef CONFIG_CC_IS_CLANG +#pragma nounroll /* https://bugs.llvm.org/show_bug.cgi?id=45803 */ +#endif for (i = 0; i < 8; ++i) S->h[i] = S->h[i] ^ v[i] ^ v[i + 8]; } diff --git a/crypto/crypto_engine.c b/crypto/crypto_engine.c index eb029ff1e05a..3655d9d3f5df 100644 --- a/crypto/crypto_engine.c +++ b/crypto/crypto_engine.c @@ -22,32 +22,36 @@ * @err: error number */ static void crypto_finalize_request(struct crypto_engine *engine, - struct crypto_async_request *req, int err) + struct crypto_async_request *req, int err) { unsigned long flags; - bool finalize_cur_req = false; + bool finalize_req = false; int ret; struct crypto_engine_ctx *enginectx; - spin_lock_irqsave(&engine->queue_lock, flags); - if (engine->cur_req == req) - finalize_cur_req = true; - spin_unlock_irqrestore(&engine->queue_lock, flags); + /* + * If hardware cannot enqueue more requests + * and retry mechanism is not supported + * make sure we are completing the current request + */ + if (!engine->retry_support) { + spin_lock_irqsave(&engine->queue_lock, flags); + if (engine->cur_req == req) { + finalize_req = true; + engine->cur_req = NULL; + } + spin_unlock_irqrestore(&engine->queue_lock, flags); + } - if (finalize_cur_req) { + if (finalize_req || engine->retry_support) { enginectx = crypto_tfm_ctx(req->tfm); - if (engine->cur_req_prepared && + if (enginectx->op.prepare_request && enginectx->op.unprepare_request) { ret = enginectx->op.unprepare_request(engine, req); if (ret) dev_err(engine->dev, "failed to unprepare request\n"); } - spin_lock_irqsave(&engine->queue_lock, flags); - engine->cur_req = NULL; - engine->cur_req_prepared = false; - spin_unlock_irqrestore(&engine->queue_lock, flags); } - req->complete(req, err); kthread_queue_work(engine->kworker, &engine->pump_requests); @@ -74,7 +78,7 @@ static void crypto_pump_requests(struct crypto_engine *engine, spin_lock_irqsave(&engine->queue_lock, flags); /* Make sure we are not already running a request */ - if (engine->cur_req) + if (!engine->retry_support && engine->cur_req) goto out; /* If another context is idling then defer */ @@ -108,13 +112,21 @@ static void crypto_pump_requests(struct crypto_engine *engine, goto out; } +start_request: /* Get the fist request from the engine queue to handle */ backlog = crypto_get_backlog(&engine->queue); async_req = crypto_dequeue_request(&engine->queue); if (!async_req) goto out; - engine->cur_req = async_req; + /* + * If hardware doesn't support the retry mechanism, + * keep track of the request we are processing now. + * We'll need it on completion (crypto_finalize_request). + */ + if (!engine->retry_support) + engine->cur_req = async_req; + if (backlog) backlog->complete(backlog, -EINPROGRESS); @@ -130,7 +142,7 @@ static void crypto_pump_requests(struct crypto_engine *engine, ret = engine->prepare_crypt_hardware(engine); if (ret) { dev_err(engine->dev, "failed to prepare crypt hardware\n"); - goto req_err; + goto req_err_2; } } @@ -141,28 +153,90 @@ static void crypto_pump_requests(struct crypto_engine *engine, if (ret) { dev_err(engine->dev, "failed to prepare request: %d\n", ret); - goto req_err; + goto req_err_2; } - engine->cur_req_prepared = true; } if (!enginectx->op.do_one_request) { dev_err(engine->dev, "failed to do request\n"); ret = -EINVAL; - goto req_err; + goto req_err_1; } + ret = enginectx->op.do_one_request(engine, async_req); - if (ret) { - dev_err(engine->dev, "Failed to do one request from queue: %d\n", ret); - goto req_err; + + /* Request unsuccessfully executed by hardware */ + if (ret < 0) { + /* + * If hardware queue is full (-ENOSPC), requeue request + * regardless of backlog flag. + * Otherwise, unprepare and complete the request. + */ + if (!engine->retry_support || + (ret != -ENOSPC)) { + dev_err(engine->dev, + "Failed to do one request from queue: %d\n", + ret); + goto req_err_1; + } + /* + * If retry mechanism is supported, + * unprepare current request and + * enqueue it back into crypto-engine queue. + */ + if (enginectx->op.unprepare_request) { + ret = enginectx->op.unprepare_request(engine, + async_req); + if (ret) + dev_err(engine->dev, + "failed to unprepare request\n"); + } + spin_lock_irqsave(&engine->queue_lock, flags); + /* + * If hardware was unable to execute request, enqueue it + * back in front of crypto-engine queue, to keep the order + * of requests. + */ + crypto_enqueue_request_head(&engine->queue, async_req); + + kthread_queue_work(engine->kworker, &engine->pump_requests); + goto out; } - return; -req_err: - crypto_finalize_request(engine, async_req, ret); + goto retry; + +req_err_1: + if (enginectx->op.unprepare_request) { + ret = enginectx->op.unprepare_request(engine, async_req); + if (ret) + dev_err(engine->dev, "failed to unprepare request\n"); + } + +req_err_2: + async_req->complete(async_req, ret); + +retry: + /* If retry mechanism is supported, send new requests to engine */ + if (engine->retry_support) { + spin_lock_irqsave(&engine->queue_lock, flags); + goto start_request; + } return; out: spin_unlock_irqrestore(&engine->queue_lock, flags); + + /* + * Batch requests is possible only if + * hardware can enqueue multiple requests + */ + if (engine->do_batch_requests) { + ret = engine->do_batch_requests(engine); + if (ret) + dev_err(engine->dev, "failed to do batch requests: %d\n", + ret); + } + + return; } static void crypto_pump_work(struct kthread_work *work) @@ -386,15 +460,27 @@ int crypto_engine_stop(struct crypto_engine *engine) EXPORT_SYMBOL_GPL(crypto_engine_stop); /** - * crypto_engine_alloc_init - allocate crypto hardware engine structure and - * initialize it. + * crypto_engine_alloc_init_and_set - allocate crypto hardware engine structure + * and initialize it by setting the maximum number of entries in the software + * crypto-engine queue. * @dev: the device attached with one hardware engine + * @retry_support: whether hardware has support for retry mechanism + * @cbk_do_batch: pointer to a callback function to be invoked when executing a + * a batch of requests. + * This has the form: + * callback(struct crypto_engine *engine) + * where: + * @engine: the crypto engine structure. * @rt: whether this queue is set to run as a realtime task + * @qlen: maximum size of the crypto-engine queue * * This must be called from context that can sleep. * Return: the crypto engine structure on success, else NULL. */ -struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt) +struct crypto_engine *crypto_engine_alloc_init_and_set(struct device *dev, + bool retry_support, + int (*cbk_do_batch)(struct crypto_engine *engine), + bool rt, int qlen) { struct sched_param param = { .sched_priority = MAX_RT_PRIO / 2 }; struct crypto_engine *engine; @@ -411,12 +497,18 @@ struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt) engine->running = false; engine->busy = false; engine->idling = false; - engine->cur_req_prepared = false; + engine->retry_support = retry_support; engine->priv_data = dev; + /* + * Batch requests is possible only if + * hardware has support for retry mechanism. + */ + engine->do_batch_requests = retry_support ? cbk_do_batch : NULL; + snprintf(engine->name, sizeof(engine->name), "%s-engine", dev_name(dev)); - crypto_init_queue(&engine->queue, CRYPTO_ENGINE_MAX_QLEN); + crypto_init_queue(&engine->queue, qlen); spin_lock_init(&engine->queue_lock); engine->kworker = kthread_create_worker(0, "%s", engine->name); @@ -433,6 +525,22 @@ struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt) return engine; } +EXPORT_SYMBOL_GPL(crypto_engine_alloc_init_and_set); + +/** + * crypto_engine_alloc_init - allocate crypto hardware engine structure and + * initialize it. + * @dev: the device attached with one hardware engine + * @rt: whether this queue is set to run as a realtime task + * + * This must be called from context that can sleep. + * Return: the crypto engine structure on success, else NULL. + */ +struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt) +{ + return crypto_engine_alloc_init_and_set(dev, false, NULL, rt, + CRYPTO_ENGINE_MAX_QLEN); +} EXPORT_SYMBOL_GPL(crypto_engine_alloc_init); /** diff --git a/crypto/drbg.c b/crypto/drbg.c index b6929eb5f565..37526eb8c5d5 100644 --- a/crypto/drbg.c +++ b/crypto/drbg.c @@ -1087,10 +1087,6 @@ static void drbg_async_seed(struct work_struct *work) if (ret) goto unlock; - /* If nonblocking pool is initialized, deactivate Jitter RNG */ - crypto_free_rng(drbg->jent); - drbg->jent = NULL; - /* Set seeded to false so that if __drbg_seed fails the * next generate call will trigger a reseed. */ @@ -1168,7 +1164,23 @@ static int drbg_seed(struct drbg_state *drbg, struct drbg_string *pers, entropylen); if (ret) { pr_devel("DRBG: jent failed with %d\n", ret); - goto out; + + /* + * Do not treat the transient failure of the + * Jitter RNG as an error that needs to be + * reported. The combined number of the + * maximum reseed threshold times the maximum + * number of Jitter RNG transient errors is + * less than the reseed threshold required by + * SP800-90A allowing us to treat the + * transient errors as such. + * + * However, we mandate that at least the first + * seeding operation must succeed with the + * Jitter RNG. + */ + if (!reseed || ret != -EAGAIN) + goto out; } drbg_string_fill(&data1, entropy, entropylen * 2); @@ -1294,8 +1306,10 @@ static inline int drbg_alloc_state(struct drbg_state *drbg) if (IS_ENABLED(CONFIG_CRYPTO_FIPS)) { drbg->prev = kzalloc(drbg_sec_strength(drbg->core->flags), GFP_KERNEL); - if (!drbg->prev) + if (!drbg->prev) { + ret = -ENOMEM; goto fini; + } drbg->fips_primed = false; } @@ -1492,6 +1506,8 @@ static int drbg_prepare_hrng(struct drbg_state *drbg) if (list_empty(&drbg->test_data.list)) return 0; + drbg->jent = crypto_alloc_rng("jitterentropy_rng", 0, 0); + INIT_WORK(&drbg->seed_work, drbg_async_seed); drbg->random_ready.owner = THIS_MODULE; @@ -1512,8 +1528,6 @@ static int drbg_prepare_hrng(struct drbg_state *drbg) return err; } - drbg->jent = crypto_alloc_rng("jitterentropy_rng", 0, 0); - /* * Require frequent reseeds until the seed source is fully * initialized. diff --git a/crypto/essiv.c b/crypto/essiv.c index 465a89c9d1ef..a7f45dbc4ee2 100644 --- a/crypto/essiv.c +++ b/crypto/essiv.c @@ -66,7 +66,6 @@ static int essiv_skcipher_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { struct essiv_tfm_ctx *tctx = crypto_skcipher_ctx(tfm); - SHASH_DESC_ON_STACK(desc, tctx->hash); u8 salt[HASH_MAX_DIGESTSIZE]; int err; @@ -78,8 +77,7 @@ static int essiv_skcipher_setkey(struct crypto_skcipher *tfm, if (err) return err; - desc->tfm = tctx->hash; - err = crypto_shash_digest(desc, key, keylen, salt); + err = crypto_shash_tfm_digest(tctx->hash, key, keylen, salt); if (err) return err; diff --git a/crypto/internal.h b/crypto/internal.h index d5ebc60c5143..ff06a3bd1ca1 100644 --- a/crypto/internal.h +++ b/crypto/internal.h @@ -65,6 +65,7 @@ void crypto_alg_tested(const char *name, int err); void crypto_remove_spawns(struct crypto_alg *alg, struct list_head *list, struct crypto_alg *nalg); void crypto_remove_final(struct list_head *list); +void crypto_shoot_alg(struct crypto_alg *alg); struct crypto_tfm *__crypto_alloc_tfm(struct crypto_alg *alg, u32 type, u32 mask); void *crypto_create_tfm(struct crypto_alg *alg, diff --git a/crypto/jitterentropy-kcapi.c b/crypto/jitterentropy-kcapi.c index a5ce8f96790f..b43684c0dade 100644 --- a/crypto/jitterentropy-kcapi.c +++ b/crypto/jitterentropy-kcapi.c @@ -108,6 +108,7 @@ void jent_get_nstime(__u64 *out) struct jitterentropy { spinlock_t jent_lock; struct rand_data *entropy_collector; + unsigned int reset_cnt; }; static int jent_kcapi_init(struct crypto_tfm *tfm) @@ -142,7 +143,33 @@ static int jent_kcapi_random(struct crypto_rng *tfm, int ret = 0; spin_lock(&rng->jent_lock); + + /* Return a permanent error in case we had too many resets in a row. */ + if (rng->reset_cnt > (1<<10)) { + ret = -EFAULT; + goto out; + } + ret = jent_read_entropy(rng->entropy_collector, rdata, dlen); + + /* Reset RNG in case of health failures */ + if (ret < -1) { + pr_warn_ratelimited("Reset Jitter RNG due to health test failure: %s failure\n", + (ret == -2) ? "Repetition Count Test" : + "Adaptive Proportion Test"); + + rng->reset_cnt++; + + ret = -EAGAIN; + } else { + rng->reset_cnt = 0; + + /* Convert the Jitter RNG error into a usable error code */ + if (ret == -1) + ret = -EINVAL; + } + +out: spin_unlock(&rng->jent_lock); return ret; diff --git a/crypto/jitterentropy.c b/crypto/jitterentropy.c index 042157f0d28b..57f4a1ac738b 100644 --- a/crypto/jitterentropy.c +++ b/crypto/jitterentropy.c @@ -2,7 +2,7 @@ * Non-physical true random number generator based on timing jitter -- * Jitter RNG standalone code. * - * Copyright Stephan Mueller <smueller@chronox.de>, 2015 - 2019 + * Copyright Stephan Mueller <smueller@chronox.de>, 2015 - 2020 * * Design * ====== @@ -47,7 +47,7 @@ /* * This Jitterentropy RNG is based on the jitterentropy library - * version 2.1.2 provided at http://www.chronox.de/jent.html + * version 2.2.0 provided at http://www.chronox.de/jent.html */ #ifdef __OPTIMIZE__ @@ -83,6 +83,22 @@ struct rand_data { unsigned int memblocksize; /* Size of one memory block in bytes */ unsigned int memaccessloops; /* Number of memory accesses per random * bit generation */ + + /* Repetition Count Test */ + int rct_count; /* Number of stuck values */ + + /* Adaptive Proportion Test for a significance level of 2^-30 */ +#define JENT_APT_CUTOFF 325 /* Taken from SP800-90B sec 4.4.2 */ +#define JENT_APT_WINDOW_SIZE 512 /* Data window size */ + /* LSB of time stamp to process */ +#define JENT_APT_LSB 16 +#define JENT_APT_WORD_MASK (JENT_APT_LSB - 1) + unsigned int apt_observations; /* Number of collected observations */ + unsigned int apt_count; /* APT counter */ + unsigned int apt_base; /* APT base reference */ + unsigned int apt_base_set:1; /* APT base reference set? */ + + unsigned int health_failure:1; /* Permanent health failure */ }; /* Flags that can be used to initialize the RNG */ @@ -98,12 +114,201 @@ struct rand_data { * variations (2nd derivation of time is * zero). */ #define JENT_ESTUCK 8 /* Too many stuck results during init. */ +#define JENT_EHEALTH 9 /* Health test failed during initialization */ +#define JENT_ERCT 10 /* RCT failed during initialization */ + +#include "jitterentropy.h" /*************************************************************************** - * Helper functions + * Adaptive Proportion Test + * + * This test complies with SP800-90B section 4.4.2. ***************************************************************************/ -#include "jitterentropy.h" +/** + * Reset the APT counter + * + * @ec [in] Reference to entropy collector + */ +static void jent_apt_reset(struct rand_data *ec, unsigned int delta_masked) +{ + /* Reset APT counter */ + ec->apt_count = 0; + ec->apt_base = delta_masked; + ec->apt_observations = 0; +} + +/** + * Insert a new entropy event into APT + * + * @ec [in] Reference to entropy collector + * @delta_masked [in] Masked time delta to process + */ +static void jent_apt_insert(struct rand_data *ec, unsigned int delta_masked) +{ + /* Initialize the base reference */ + if (!ec->apt_base_set) { + ec->apt_base = delta_masked; + ec->apt_base_set = 1; + return; + } + + if (delta_masked == ec->apt_base) { + ec->apt_count++; + + if (ec->apt_count >= JENT_APT_CUTOFF) + ec->health_failure = 1; + } + + ec->apt_observations++; + + if (ec->apt_observations >= JENT_APT_WINDOW_SIZE) + jent_apt_reset(ec, delta_masked); +} + +/*************************************************************************** + * Stuck Test and its use as Repetition Count Test + * + * The Jitter RNG uses an enhanced version of the Repetition Count Test + * (RCT) specified in SP800-90B section 4.4.1. Instead of counting identical + * back-to-back values, the input to the RCT is the counting of the stuck + * values during the generation of one Jitter RNG output block. + * + * The RCT is applied with an alpha of 2^{-30} compliant to FIPS 140-2 IG 9.8. + * + * During the counting operation, the Jitter RNG always calculates the RCT + * cut-off value of C. If that value exceeds the allowed cut-off value, + * the Jitter RNG output block will be calculated completely but discarded at + * the end. The caller of the Jitter RNG is informed with an error code. + ***************************************************************************/ + +/** + * Repetition Count Test as defined in SP800-90B section 4.4.1 + * + * @ec [in] Reference to entropy collector + * @stuck [in] Indicator whether the value is stuck + */ +static void jent_rct_insert(struct rand_data *ec, int stuck) +{ + /* + * If we have a count less than zero, a previous RCT round identified + * a failure. We will not overwrite it. + */ + if (ec->rct_count < 0) + return; + + if (stuck) { + ec->rct_count++; + + /* + * The cutoff value is based on the following consideration: + * alpha = 2^-30 as recommended in FIPS 140-2 IG 9.8. + * In addition, we require an entropy value H of 1/OSR as this + * is the minimum entropy required to provide full entropy. + * Note, we collect 64 * OSR deltas for inserting them into + * the entropy pool which should then have (close to) 64 bits + * of entropy. + * + * Note, ec->rct_count (which equals to value B in the pseudo + * code of SP800-90B section 4.4.1) starts with zero. Hence + * we need to subtract one from the cutoff value as calculated + * following SP800-90B. + */ + if ((unsigned int)ec->rct_count >= (31 * ec->osr)) { + ec->rct_count = -1; + ec->health_failure = 1; + } + } else { + ec->rct_count = 0; + } +} + +/** + * Is there an RCT health test failure? + * + * @ec [in] Reference to entropy collector + * + * @return + * 0 No health test failure + * 1 Permanent health test failure + */ +static int jent_rct_failure(struct rand_data *ec) +{ + if (ec->rct_count < 0) + return 1; + return 0; +} + +static inline __u64 jent_delta(__u64 prev, __u64 next) +{ +#define JENT_UINT64_MAX (__u64)(~((__u64) 0)) + return (prev < next) ? (next - prev) : + (JENT_UINT64_MAX - prev + 1 + next); +} + +/** + * Stuck test by checking the: + * 1st derivative of the jitter measurement (time delta) + * 2nd derivative of the jitter measurement (delta of time deltas) + * 3rd derivative of the jitter measurement (delta of delta of time deltas) + * + * All values must always be non-zero. + * + * @ec [in] Reference to entropy collector + * @current_delta [in] Jitter time delta + * + * @return + * 0 jitter measurement not stuck (good bit) + * 1 jitter measurement stuck (reject bit) + */ +static int jent_stuck(struct rand_data *ec, __u64 current_delta) +{ + __u64 delta2 = jent_delta(ec->last_delta, current_delta); + __u64 delta3 = jent_delta(ec->last_delta2, delta2); + unsigned int delta_masked = current_delta & JENT_APT_WORD_MASK; + + ec->last_delta = current_delta; + ec->last_delta2 = delta2; + + /* + * Insert the result of the comparison of two back-to-back time + * deltas. + */ + jent_apt_insert(ec, delta_masked); + + if (!current_delta || !delta2 || !delta3) { + /* RCT with a stuck bit */ + jent_rct_insert(ec, 1); + return 1; + } + + /* RCT with a non-stuck bit */ + jent_rct_insert(ec, 0); + + return 0; +} + +/** + * Report any health test failures + * + * @ec [in] Reference to entropy collector + * + * @return + * 0 No health test failure + * 1 Permanent health test failure + */ +static int jent_health_failure(struct rand_data *ec) +{ + /* Test is only enabled in FIPS mode */ + if (!jent_fips_enabled()) + return 0; + + return ec->health_failure; +} + +/*************************************************************************** + * Noise sources + ***************************************************************************/ /** * Update of the loop count used for the next round of @@ -148,10 +353,6 @@ static __u64 jent_loop_shuffle(struct rand_data *ec, return (shuffle + (1<<min)); } -/*************************************************************************** - * Noise sources - ***************************************************************************/ - /** * CPU Jitter noise source -- this is the noise source based on the CPU * execution time jitter @@ -166,18 +367,19 @@ static __u64 jent_loop_shuffle(struct rand_data *ec, * the CPU execution time jitter. Any change to the loop in this function * implies that careful retesting must be done. * - * Input: - * @ec entropy collector struct - * @time time stamp to be injected - * @loop_cnt if a value not equal to 0 is set, use the given value as number of - * loops to perform the folding + * @ec [in] entropy collector struct + * @time [in] time stamp to be injected + * @loop_cnt [in] if a value not equal to 0 is set, use the given value as + * number of loops to perform the folding + * @stuck [in] Is the time stamp identified as stuck? * * Output: * updated ec->data * * @return Number of loops the folding operation is performed */ -static __u64 jent_lfsr_time(struct rand_data *ec, __u64 time, __u64 loop_cnt) +static void jent_lfsr_time(struct rand_data *ec, __u64 time, __u64 loop_cnt, + int stuck) { unsigned int i; __u64 j = 0; @@ -220,9 +422,17 @@ static __u64 jent_lfsr_time(struct rand_data *ec, __u64 time, __u64 loop_cnt) new ^= tmp; } } - ec->data = new; - return fold_loop_cnt; + /* + * If the time stamp is stuck, do not finally insert the value into + * the entropy pool. Although this operation should not do any harm + * even when the time stamp has no entropy, SP800-90B requires that + * any conditioning operation (SP800-90B considers the LFSR to be a + * conditioning operation) to have an identical amount of input + * data according to section 3.1.5. + */ + if (!stuck) + ec->data = new; } /** @@ -243,16 +453,13 @@ static __u64 jent_lfsr_time(struct rand_data *ec, __u64 time, __u64 loop_cnt) * to reliably access either L3 or memory, the ec->mem memory must be quite * large which is usually not desirable. * - * Input: - * @ec Reference to the entropy collector with the memory access data -- if - * the reference to the memory block to be accessed is NULL, this noise - * source is disabled - * @loop_cnt if a value not equal to 0 is set, use the given value as number of - * loops to perform the folding - * - * @return Number of memory access operations + * @ec [in] Reference to the entropy collector with the memory access data -- if + * the reference to the memory block to be accessed is NULL, this noise + * source is disabled + * @loop_cnt [in] if a value not equal to 0 is set, use the given value + * number of loops to perform the LFSR */ -static unsigned int jent_memaccess(struct rand_data *ec, __u64 loop_cnt) +static void jent_memaccess(struct rand_data *ec, __u64 loop_cnt) { unsigned int wrap = 0; __u64 i = 0; @@ -262,7 +469,7 @@ static unsigned int jent_memaccess(struct rand_data *ec, __u64 loop_cnt) jent_loop_shuffle(ec, MAX_ACC_LOOP_BIT, MIN_ACC_LOOP_BIT); if (NULL == ec || NULL == ec->mem) - return 0; + return; wrap = ec->memblocksize * ec->memblocks; /* @@ -288,43 +495,11 @@ static unsigned int jent_memaccess(struct rand_data *ec, __u64 loop_cnt) ec->memlocation = ec->memlocation + ec->memblocksize - 1; ec->memlocation = ec->memlocation % wrap; } - return i; } /*************************************************************************** * Start of entropy processing logic ***************************************************************************/ - -/** - * Stuck test by checking the: - * 1st derivation of the jitter measurement (time delta) - * 2nd derivation of the jitter measurement (delta of time deltas) - * 3rd derivation of the jitter measurement (delta of delta of time deltas) - * - * All values must always be non-zero. - * - * Input: - * @ec Reference to entropy collector - * @current_delta Jitter time delta - * - * @return - * 0 jitter measurement not stuck (good bit) - * 1 jitter measurement stuck (reject bit) - */ -static int jent_stuck(struct rand_data *ec, __u64 current_delta) -{ - __s64 delta2 = ec->last_delta - current_delta; - __s64 delta3 = delta2 - ec->last_delta2; - - ec->last_delta = current_delta; - ec->last_delta2 = delta2; - - if (!current_delta || !delta2 || !delta3) - return 1; - - return 0; -} - /** * This is the heart of the entropy generation: calculate time deltas and * use the CPU jitter in the time deltas. The jitter is injected into the @@ -334,8 +509,7 @@ static int jent_stuck(struct rand_data *ec, __u64 current_delta) * of this function! This can be done by calling this function * and not using its result. * - * Input: - * @entropy_collector Reference to entropy collector + * @ec [in] Reference to entropy collector * * @return result of stuck test */ @@ -343,6 +517,7 @@ static int jent_measure_jitter(struct rand_data *ec) { __u64 time = 0; __u64 current_delta = 0; + int stuck; /* Invoke one noise source before time measurement to add variations */ jent_memaccess(ec, 0); @@ -352,22 +527,23 @@ static int jent_measure_jitter(struct rand_data *ec) * invocation to measure the timing variations */ jent_get_nstime(&time); - current_delta = time - ec->prev_time; + current_delta = jent_delta(ec->prev_time, time); ec->prev_time = time; + /* Check whether we have a stuck measurement. */ + stuck = jent_stuck(ec, current_delta); + /* Now call the next noise sources which also injects the data */ - jent_lfsr_time(ec, current_delta, 0); + jent_lfsr_time(ec, current_delta, 0, stuck); - /* Check whether we have a stuck measurement. */ - return jent_stuck(ec, current_delta); + return stuck; } /** * Generator of one 64 bit random number * Function fills rand_data->data * - * Input: - * @ec Reference to entropy collector + * @ec [in] Reference to entropy collector */ static void jent_gen_entropy(struct rand_data *ec) { @@ -391,31 +567,6 @@ static void jent_gen_entropy(struct rand_data *ec) } /** - * The continuous test required by FIPS 140-2 -- the function automatically - * primes the test if needed. - * - * Return: - * returns normally if FIPS test passed - * panics the kernel if FIPS test failed - */ -static void jent_fips_test(struct rand_data *ec) -{ - if (!jent_fips_enabled()) - return; - - /* prime the FIPS test */ - if (!ec->old_data) { - ec->old_data = ec->data; - jent_gen_entropy(ec); - } - - if (ec->data == ec->old_data) - jent_panic("jitterentropy: Duplicate output detected\n"); - - ec->old_data = ec->data; -} - -/** * Entry function: Obtain entropy for the caller. * * This function invokes the entropy gathering logic as often to generate @@ -425,17 +576,18 @@ static void jent_fips_test(struct rand_data *ec) * This function truncates the last 64 bit entropy value output to the exact * size specified by the caller. * - * Input: - * @ec Reference to entropy collector - * @data pointer to buffer for storing random data -- buffer must already - * exist - * @len size of the buffer, specifying also the requested number of random - * in bytes + * @ec [in] Reference to entropy collector + * @data [in] pointer to buffer for storing random data -- buffer must already + * exist + * @len [in] size of the buffer, specifying also the requested number of random + * in bytes * * @return 0 when request is fulfilled or an error * * The following error codes can occur: * -1 entropy_collector is NULL + * -2 RCT failed + * -3 APT test failed */ int jent_read_entropy(struct rand_data *ec, unsigned char *data, unsigned int len) @@ -449,7 +601,42 @@ int jent_read_entropy(struct rand_data *ec, unsigned char *data, unsigned int tocopy; jent_gen_entropy(ec); - jent_fips_test(ec); + + if (jent_health_failure(ec)) { + int ret; + + if (jent_rct_failure(ec)) + ret = -2; + else + ret = -3; + + /* + * Re-initialize the noise source + * + * If the health test fails, the Jitter RNG remains + * in failure state and will return a health failure + * during next invocation. + */ + if (jent_entropy_init()) + return ret; + + /* Set APT to initial state */ + jent_apt_reset(ec, 0); + ec->apt_base_set = 0; + + /* Set RCT to initial state */ + ec->rct_count = 0; + + /* Re-enable Jitter RNG */ + ec->health_failure = 0; + + /* + * Return the health test failure status to the + * caller as the generated value is not appropriate. + */ + return ret; + } + if ((DATA_SIZE_BITS / 8) < len) tocopy = (DATA_SIZE_BITS / 8); else @@ -513,11 +700,15 @@ int jent_entropy_init(void) int i; __u64 delta_sum = 0; __u64 old_delta = 0; + unsigned int nonstuck = 0; int time_backwards = 0; int count_mod = 0; int count_stuck = 0; struct rand_data ec = { 0 }; + /* Required for RCT */ + ec.osr = 1; + /* We could perform statistical tests here, but the problem is * that we only have a few loop counts to do testing. These * loop counts may show some slight skew and we produce @@ -539,8 +730,10 @@ int jent_entropy_init(void) /* * TESTLOOPCOUNT needs some loops to identify edge systems. 100 is * definitely too little. + * + * SP800-90B requires at least 1024 initial test cycles. */ -#define TESTLOOPCOUNT 300 +#define TESTLOOPCOUNT 1024 #define CLEARCACHE 100 for (i = 0; (TESTLOOPCOUNT + CLEARCACHE) > i; i++) { __u64 time = 0; @@ -552,13 +745,13 @@ int jent_entropy_init(void) /* Invoke core entropy collection logic */ jent_get_nstime(&time); ec.prev_time = time; - jent_lfsr_time(&ec, time, 0); + jent_lfsr_time(&ec, time, 0, 0); jent_get_nstime(&time2); /* test whether timer works */ if (!time || !time2) return JENT_ENOTIME; - delta = time2 - time; + delta = jent_delta(time, time2); /* * test whether timer is fine grained enough to provide * delta even when called shortly after each other -- this @@ -581,6 +774,28 @@ int jent_entropy_init(void) if (stuck) count_stuck++; + else { + nonstuck++; + + /* + * Ensure that the APT succeeded. + * + * With the check below that count_stuck must be less + * than 10% of the overall generated raw entropy values + * it is guaranteed that the APT is invoked at + * floor((TESTLOOPCOUNT * 0.9) / 64) == 14 times. + */ + if ((nonstuck % JENT_APT_WINDOW_SIZE) == 0) { + jent_apt_reset(&ec, + delta & JENT_APT_WORD_MASK); + if (jent_health_failure(&ec)) + return JENT_EHEALTH; + } + } + + /* Validate RCT */ + if (jent_rct_failure(&ec)) + return JENT_ERCT; /* test whether we have an increasing timer */ if (!(time2 > time)) diff --git a/crypto/lrw.c b/crypto/lrw.c index 3c734b81b3a2..5b07a7c09296 100644 --- a/crypto/lrw.c +++ b/crypto/lrw.c @@ -287,7 +287,7 @@ static void exit_tfm(struct crypto_skcipher *tfm) crypto_free_skcipher(ctx->child); } -static void free_inst(struct skcipher_instance *inst) +static void crypto_lrw_free(struct skcipher_instance *inst) { crypto_drop_skcipher(skcipher_instance_ctx(inst)); kfree(inst); @@ -400,12 +400,12 @@ static int create(struct crypto_template *tmpl, struct rtattr **tb) inst->alg.encrypt = encrypt; inst->alg.decrypt = decrypt; - inst->free = free_inst; + inst->free = crypto_lrw_free; err = skcipher_register_instance(tmpl, inst); if (err) { err_free_inst: - free_inst(inst); + crypto_lrw_free(inst); } return err; } diff --git a/crypto/sha1_generic.c b/crypto/sha1_generic.c index 7c57b844c382..1d43472fecbd 100644 --- a/crypto/sha1_generic.c +++ b/crypto/sha1_generic.c @@ -15,7 +15,6 @@ #include <linux/init.h> #include <linux/module.h> #include <linux/mm.h> -#include <linux/cryptohash.h> #include <linux/types.h> #include <crypto/sha.h> #include <crypto/sha1_base.h> @@ -31,10 +30,10 @@ EXPORT_SYMBOL_GPL(sha1_zero_message_hash); static void sha1_generic_block_fn(struct sha1_state *sst, u8 const *src, int blocks) { - u32 temp[SHA_WORKSPACE_WORDS]; + u32 temp[SHA1_WORKSPACE_WORDS]; while (blocks--) { - sha_transform(sst->state, src, temp); + sha1_transform(sst->state, src, temp); src += SHA1_BLOCK_SIZE; } memzero_explicit(temp, sizeof(temp)); diff --git a/crypto/sha256_generic.c b/crypto/sha256_generic.c index f2d7095d4f2d..88156e3e2a33 100644 --- a/crypto/sha256_generic.c +++ b/crypto/sha256_generic.c @@ -35,27 +35,31 @@ EXPORT_SYMBOL_GPL(sha256_zero_message_hash); static int crypto_sha256_init(struct shash_desc *desc) { - return sha256_init(shash_desc_ctx(desc)); + sha256_init(shash_desc_ctx(desc)); + return 0; } static int crypto_sha224_init(struct shash_desc *desc) { - return sha224_init(shash_desc_ctx(desc)); + sha224_init(shash_desc_ctx(desc)); + return 0; } int crypto_sha256_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - return sha256_update(shash_desc_ctx(desc), data, len); + sha256_update(shash_desc_ctx(desc), data, len); + return 0; } EXPORT_SYMBOL(crypto_sha256_update); static int crypto_sha256_final(struct shash_desc *desc, u8 *out) { if (crypto_shash_digestsize(desc->tfm) == SHA224_DIGEST_SIZE) - return sha224_final(shash_desc_ctx(desc), out); + sha224_final(shash_desc_ctx(desc), out); else - return sha256_final(shash_desc_ctx(desc), out); + sha256_final(shash_desc_ctx(desc), out); + return 0; } int crypto_sha256_finup(struct shash_desc *desc, const u8 *data, diff --git a/crypto/shash.c b/crypto/shash.c index c075b26c2a1d..e6a4b5f39b8c 100644 --- a/crypto/shash.c +++ b/crypto/shash.c @@ -206,6 +206,22 @@ int crypto_shash_digest(struct shash_desc *desc, const u8 *data, } EXPORT_SYMBOL_GPL(crypto_shash_digest); +int crypto_shash_tfm_digest(struct crypto_shash *tfm, const u8 *data, + unsigned int len, u8 *out) +{ + SHASH_DESC_ON_STACK(desc, tfm); + int err; + + desc->tfm = tfm; + + err = crypto_shash_digest(desc, data, len, out); + + shash_desc_zero(desc); + + return err; +} +EXPORT_SYMBOL_GPL(crypto_shash_tfm_digest); + static int shash_default_export(struct shash_desc *desc, void *out) { memcpy(out, shash_desc_ctx(desc), crypto_shash_descsize(desc->tfm)); diff --git a/crypto/xts.c b/crypto/xts.c index 6d8cea94b3cf..3565f3b863a6 100644 --- a/crypto/xts.c +++ b/crypto/xts.c @@ -322,7 +322,7 @@ static void exit_tfm(struct crypto_skcipher *tfm) crypto_free_cipher(ctx->tweak); } -static void free_inst(struct skcipher_instance *inst) +static void crypto_xts_free(struct skcipher_instance *inst) { crypto_drop_skcipher(skcipher_instance_ctx(inst)); kfree(inst); @@ -434,12 +434,12 @@ static int create(struct crypto_template *tmpl, struct rtattr **tb) inst->alg.encrypt = encrypt; inst->alg.decrypt = decrypt; - inst->free = free_inst; + inst->free = crypto_xts_free; err = skcipher_register_instance(tmpl, inst); if (err) { err_free_inst: - free_inst(inst); + crypto_xts_free(inst); } return err; } diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig index 9bc46da8d77a..ac00d78ee9cc 100644 --- a/drivers/char/hw_random/Kconfig +++ b/drivers/char/hw_random/Kconfig @@ -474,6 +474,19 @@ config HW_RANDOM_KEYSTONE help This option enables Keystone's hardware random generator. +config HW_RANDOM_CCTRNG + tristate "Arm CryptoCell True Random Number Generator support" + depends on HAS_IOMEM && OF + help + Say 'Y' to enable the True Random Number Generator driver for the + Arm TrustZone CryptoCell family of processors. + Currently the CryptoCell 713 and 703 are supported. + The driver is supported only in SoC where Trusted Execution + Environment is not used. + Choose 'M' to compile this driver as a module. The module + will be called cctrng. + If unsure, say 'N'. + endif # HW_RANDOM config UML_RANDOM diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile index a7801b49ce6c..2c6724735345 100644 --- a/drivers/char/hw_random/Makefile +++ b/drivers/char/hw_random/Makefile @@ -41,3 +41,4 @@ obj-$(CONFIG_HW_RANDOM_S390) += s390-trng.o obj-$(CONFIG_HW_RANDOM_KEYSTONE) += ks-sa-rng.o obj-$(CONFIG_HW_RANDOM_OPTEE) += optee-rng.o obj-$(CONFIG_HW_RANDOM_NPCM) += npcm-rng.o +obj-$(CONFIG_HW_RANDOM_CCTRNG) += cctrng.o diff --git a/drivers/char/hw_random/cctrng.c b/drivers/char/hw_random/cctrng.c new file mode 100644 index 000000000000..619148fb2dc9 --- /dev/null +++ b/drivers/char/hw_random/cctrng.c @@ -0,0 +1,735 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (C) 2019-2020 ARM Limited or its affiliates. */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/clk.h> +#include <linux/hw_random.h> +#include <linux/io.h> +#include <linux/platform_device.h> +#include <linux/pm_runtime.h> +#include <linux/interrupt.h> +#include <linux/irqreturn.h> +#include <linux/workqueue.h> +#include <linux/circ_buf.h> +#include <linux/completion.h> +#include <linux/of.h> +#include <linux/bitfield.h> +#include <linux/fips.h> + +#include "cctrng.h" + +#define CC_REG_LOW(name) (name ## _BIT_SHIFT) +#define CC_REG_HIGH(name) (CC_REG_LOW(name) + name ## _BIT_SIZE - 1) +#define CC_GENMASK(name) GENMASK(CC_REG_HIGH(name), CC_REG_LOW(name)) + +#define CC_REG_FLD_GET(reg_name, fld_name, reg_val) \ + (FIELD_GET(CC_GENMASK(CC_ ## reg_name ## _ ## fld_name), reg_val)) + +#define CC_HW_RESET_LOOP_COUNT 10 +#define CC_TRNG_SUSPEND_TIMEOUT 3000 + +/* data circular buffer in words must be: + * - of a power-of-2 size (limitation of circ_buf.h macros) + * - at least 6, the size generated in the EHR according to HW implementation + */ +#define CCTRNG_DATA_BUF_WORDS 32 + +/* The timeout for the TRNG operation should be calculated with the formula: + * Timeout = EHR_NUM * VN_COEFF * EHR_LENGTH * SAMPLE_CNT * SCALE_VALUE + * while: + * - SAMPLE_CNT is input value from the characterisation process + * - all the rest are constants + */ +#define EHR_NUM 1 +#define VN_COEFF 4 +#define EHR_LENGTH CC_TRNG_EHR_IN_BITS +#define SCALE_VALUE 2 +#define CCTRNG_TIMEOUT(smpl_cnt) \ + (EHR_NUM * VN_COEFF * EHR_LENGTH * smpl_cnt * SCALE_VALUE) + +struct cctrng_drvdata { + struct platform_device *pdev; + void __iomem *cc_base; + struct clk *clk; + struct hwrng rng; + u32 active_rosc; + /* Sampling interval for each ring oscillator: + * count of ring oscillator cycles between consecutive bits sampling. + * Value of 0 indicates non-valid rosc + */ + u32 smpl_ratio[CC_TRNG_NUM_OF_ROSCS]; + + u32 data_buf[CCTRNG_DATA_BUF_WORDS]; + struct circ_buf circ; + struct work_struct compwork; + struct work_struct startwork; + + /* pending_hw - 1 when HW is pending, 0 when it is idle */ + atomic_t pending_hw; + + /* protects against multiple concurrent consumers of data_buf */ + spinlock_t read_lock; +}; + + +/* functions for write/read CC registers */ +static inline void cc_iowrite(struct cctrng_drvdata *drvdata, u32 reg, u32 val) +{ + iowrite32(val, (drvdata->cc_base + reg)); +} +static inline u32 cc_ioread(struct cctrng_drvdata *drvdata, u32 reg) +{ + return ioread32(drvdata->cc_base + reg); +} + + +static int cc_trng_pm_get(struct device *dev) +{ + int rc = 0; + + rc = pm_runtime_get_sync(dev); + + /* pm_runtime_get_sync() can return 1 as a valid return code */ + return (rc == 1 ? 0 : rc); +} + +static void cc_trng_pm_put_suspend(struct device *dev) +{ + int rc = 0; + + pm_runtime_mark_last_busy(dev); + rc = pm_runtime_put_autosuspend(dev); + if (rc) + dev_err(dev, "pm_runtime_put_autosuspend returned %x\n", rc); +} + +static int cc_trng_pm_init(struct cctrng_drvdata *drvdata) +{ + struct device *dev = &(drvdata->pdev->dev); + + /* must be before the enabling to avoid redundant suspending */ + pm_runtime_set_autosuspend_delay(dev, CC_TRNG_SUSPEND_TIMEOUT); + pm_runtime_use_autosuspend(dev); + /* set us as active - note we won't do PM ops until cc_trng_pm_go()! */ + return pm_runtime_set_active(dev); +} + +static void cc_trng_pm_go(struct cctrng_drvdata *drvdata) +{ + struct device *dev = &(drvdata->pdev->dev); + + /* enable the PM module*/ + pm_runtime_enable(dev); +} + +static void cc_trng_pm_fini(struct cctrng_drvdata *drvdata) +{ + struct device *dev = &(drvdata->pdev->dev); + + pm_runtime_disable(dev); +} + + +static inline int cc_trng_parse_sampling_ratio(struct cctrng_drvdata *drvdata) +{ + struct device *dev = &(drvdata->pdev->dev); + struct device_node *np = drvdata->pdev->dev.of_node; + int rc; + int i; + /* ret will be set to 0 if at least one rosc has (sampling ratio > 0) */ + int ret = -EINVAL; + + rc = of_property_read_u32_array(np, "arm,rosc-ratio", + drvdata->smpl_ratio, + CC_TRNG_NUM_OF_ROSCS); + if (rc) { + /* arm,rosc-ratio was not found in device tree */ + return rc; + } + + /* verify that at least one rosc has (sampling ratio > 0) */ + for (i = 0; i < CC_TRNG_NUM_OF_ROSCS; ++i) { + dev_dbg(dev, "rosc %d sampling ratio %u", + i, drvdata->smpl_ratio[i]); + + if (drvdata->smpl_ratio[i] > 0) + ret = 0; + } + + return ret; +} + +static int cc_trng_change_rosc(struct cctrng_drvdata *drvdata) +{ + struct device *dev = &(drvdata->pdev->dev); + + dev_dbg(dev, "cctrng change rosc (was %d)\n", drvdata->active_rosc); + drvdata->active_rosc += 1; + + while (drvdata->active_rosc < CC_TRNG_NUM_OF_ROSCS) { + if (drvdata->smpl_ratio[drvdata->active_rosc] > 0) + return 0; + + drvdata->active_rosc += 1; + } + return -EINVAL; +} + + +static void cc_trng_enable_rnd_source(struct cctrng_drvdata *drvdata) +{ + u32 max_cycles; + + /* Set watchdog threshold to maximal allowed time (in CPU cycles) */ + max_cycles = CCTRNG_TIMEOUT(drvdata->smpl_ratio[drvdata->active_rosc]); + cc_iowrite(drvdata, CC_RNG_WATCHDOG_VAL_REG_OFFSET, max_cycles); + + /* enable the RND source */ + cc_iowrite(drvdata, CC_RND_SOURCE_ENABLE_REG_OFFSET, 0x1); + + /* unmask RNG interrupts */ + cc_iowrite(drvdata, CC_RNG_IMR_REG_OFFSET, (u32)~CC_RNG_INT_MASK); +} + + +/* increase circular data buffer index (head/tail) */ +static inline void circ_idx_inc(int *idx, int bytes) +{ + *idx += (bytes + 3) >> 2; + *idx &= (CCTRNG_DATA_BUF_WORDS - 1); +} + +static inline size_t circ_buf_space(struct cctrng_drvdata *drvdata) +{ + return CIRC_SPACE(drvdata->circ.head, + drvdata->circ.tail, CCTRNG_DATA_BUF_WORDS); + +} + +static int cctrng_read(struct hwrng *rng, void *data, size_t max, bool wait) +{ + /* current implementation ignores "wait" */ + + struct cctrng_drvdata *drvdata = (struct cctrng_drvdata *)rng->priv; + struct device *dev = &(drvdata->pdev->dev); + u32 *buf = (u32 *)drvdata->circ.buf; + size_t copied = 0; + size_t cnt_w; + size_t size; + size_t left; + + if (!spin_trylock(&drvdata->read_lock)) { + /* concurrent consumers from data_buf cannot be served */ + dev_dbg_ratelimited(dev, "unable to hold lock\n"); + return 0; + } + + /* copy till end of data buffer (without wrap back) */ + cnt_w = CIRC_CNT_TO_END(drvdata->circ.head, + drvdata->circ.tail, CCTRNG_DATA_BUF_WORDS); + size = min((cnt_w<<2), max); + memcpy(data, &(buf[drvdata->circ.tail]), size); + copied = size; + circ_idx_inc(&drvdata->circ.tail, size); + /* copy rest of data in data buffer */ + left = max - copied; + if (left > 0) { + cnt_w = CIRC_CNT(drvdata->circ.head, + drvdata->circ.tail, CCTRNG_DATA_BUF_WORDS); + size = min((cnt_w<<2), left); + memcpy(data, &(buf[drvdata->circ.tail]), size); + copied += size; + circ_idx_inc(&drvdata->circ.tail, size); + } + + spin_unlock(&drvdata->read_lock); + + if (circ_buf_space(drvdata) >= CC_TRNG_EHR_IN_WORDS) { + if (atomic_cmpxchg(&drvdata->pending_hw, 0, 1) == 0) { + /* re-check space in buffer to avoid potential race */ + if (circ_buf_space(drvdata) >= CC_TRNG_EHR_IN_WORDS) { + /* increment device's usage counter */ + int rc = cc_trng_pm_get(dev); + + if (rc) { + dev_err(dev, + "cc_trng_pm_get returned %x\n", + rc); + return rc; + } + + /* schedule execution of deferred work handler + * for filling of data buffer + */ + schedule_work(&drvdata->startwork); + } else { + atomic_set(&drvdata->pending_hw, 0); + } + } + } + + return copied; +} + +static void cc_trng_hw_trigger(struct cctrng_drvdata *drvdata) +{ + u32 tmp_smpl_cnt = 0; + struct device *dev = &(drvdata->pdev->dev); + + dev_dbg(dev, "cctrng hw trigger.\n"); + + /* enable the HW RND clock */ + cc_iowrite(drvdata, CC_RNG_CLK_ENABLE_REG_OFFSET, 0x1); + + /* do software reset */ + cc_iowrite(drvdata, CC_RNG_SW_RESET_REG_OFFSET, 0x1); + /* in order to verify that the reset has completed, + * the sample count need to be verified + */ + do { + /* enable the HW RND clock */ + cc_iowrite(drvdata, CC_RNG_CLK_ENABLE_REG_OFFSET, 0x1); + + /* set sampling ratio (rng_clocks) between consecutive bits */ + cc_iowrite(drvdata, CC_SAMPLE_CNT1_REG_OFFSET, + drvdata->smpl_ratio[drvdata->active_rosc]); + + /* read the sampling ratio */ + tmp_smpl_cnt = cc_ioread(drvdata, CC_SAMPLE_CNT1_REG_OFFSET); + + } while (tmp_smpl_cnt != drvdata->smpl_ratio[drvdata->active_rosc]); + + /* disable the RND source for setting new parameters in HW */ + cc_iowrite(drvdata, CC_RND_SOURCE_ENABLE_REG_OFFSET, 0); + + cc_iowrite(drvdata, CC_RNG_ICR_REG_OFFSET, 0xFFFFFFFF); + + cc_iowrite(drvdata, CC_TRNG_CONFIG_REG_OFFSET, drvdata->active_rosc); + + /* Debug Control register: set to 0 - no bypasses */ + cc_iowrite(drvdata, CC_TRNG_DEBUG_CONTROL_REG_OFFSET, 0); + + cc_trng_enable_rnd_source(drvdata); +} + +static void cc_trng_compwork_handler(struct work_struct *w) +{ + u32 isr = 0; + u32 ehr_valid = 0; + struct cctrng_drvdata *drvdata = + container_of(w, struct cctrng_drvdata, compwork); + struct device *dev = &(drvdata->pdev->dev); + int i; + + /* stop DMA and the RNG source */ + cc_iowrite(drvdata, CC_RNG_DMA_ENABLE_REG_OFFSET, 0); + cc_iowrite(drvdata, CC_RND_SOURCE_ENABLE_REG_OFFSET, 0); + + /* read RNG_ISR and check for errors */ + isr = cc_ioread(drvdata, CC_RNG_ISR_REG_OFFSET); + ehr_valid = CC_REG_FLD_GET(RNG_ISR, EHR_VALID, isr); + dev_dbg(dev, "Got RNG_ISR=0x%08X (EHR_VALID=%u)\n", isr, ehr_valid); + + if (fips_enabled && CC_REG_FLD_GET(RNG_ISR, CRNGT_ERR, isr)) { + fips_fail_notify(); + /* FIPS error is fatal */ + panic("Got HW CRNGT error while fips is enabled!\n"); + } + + /* Clear all pending RNG interrupts */ + cc_iowrite(drvdata, CC_RNG_ICR_REG_OFFSET, isr); + + + if (!ehr_valid) { + /* in case of AUTOCORR/TIMEOUT error, try the next ROSC */ + if (CC_REG_FLD_GET(RNG_ISR, AUTOCORR_ERR, isr) || + CC_REG_FLD_GET(RNG_ISR, WATCHDOG, isr)) { + dev_dbg(dev, "cctrng autocorr/timeout error.\n"); + goto next_rosc; + } + + /* in case of VN error, ignore it */ + } + + /* read EHR data from registers */ + for (i = 0; i < CC_TRNG_EHR_IN_WORDS; i++) { + /* calc word ptr in data_buf */ + u32 *buf = (u32 *)drvdata->circ.buf; + + buf[drvdata->circ.head] = cc_ioread(drvdata, + CC_EHR_DATA_0_REG_OFFSET + (i*sizeof(u32))); + + /* EHR_DATA registers are cleared on read. In case 0 value was + * returned, restart the entropy collection. + */ + if (buf[drvdata->circ.head] == 0) { + dev_dbg(dev, "Got 0 value in EHR. active_rosc %u\n", + drvdata->active_rosc); + goto next_rosc; + } + + circ_idx_inc(&drvdata->circ.head, 1<<2); + } + + atomic_set(&drvdata->pending_hw, 0); + + /* continue to fill data buffer if needed */ + if (circ_buf_space(drvdata) >= CC_TRNG_EHR_IN_WORDS) { + if (atomic_cmpxchg(&drvdata->pending_hw, 0, 1) == 0) { + /* Re-enable rnd source */ + cc_trng_enable_rnd_source(drvdata); + return; + } + } + + cc_trng_pm_put_suspend(dev); + + dev_dbg(dev, "compwork handler done\n"); + return; + +next_rosc: + if ((circ_buf_space(drvdata) >= CC_TRNG_EHR_IN_WORDS) && + (cc_trng_change_rosc(drvdata) == 0)) { + /* trigger trng hw with next rosc */ + cc_trng_hw_trigger(drvdata); + } else { + atomic_set(&drvdata->pending_hw, 0); + cc_trng_pm_put_suspend(dev); + } +} + +static irqreturn_t cc_isr(int irq, void *dev_id) +{ + struct cctrng_drvdata *drvdata = (struct cctrng_drvdata *)dev_id; + struct device *dev = &(drvdata->pdev->dev); + u32 irr; + + /* if driver suspended return, probably shared interrupt */ + if (pm_runtime_suspended(dev)) + return IRQ_NONE; + + /* read the interrupt status */ + irr = cc_ioread(drvdata, CC_HOST_RGF_IRR_REG_OFFSET); + dev_dbg(dev, "Got IRR=0x%08X\n", irr); + + if (irr == 0) /* Probably shared interrupt line */ + return IRQ_NONE; + + /* clear interrupt - must be before processing events */ + cc_iowrite(drvdata, CC_HOST_RGF_ICR_REG_OFFSET, irr); + + /* RNG interrupt - most probable */ + if (irr & CC_HOST_RNG_IRQ_MASK) { + /* Mask RNG interrupts - will be unmasked in deferred work */ + cc_iowrite(drvdata, CC_RNG_IMR_REG_OFFSET, 0xFFFFFFFF); + + /* We clear RNG interrupt here, + * to avoid it from firing as we'll unmask RNG interrupts. + */ + cc_iowrite(drvdata, CC_HOST_RGF_ICR_REG_OFFSET, + CC_HOST_RNG_IRQ_MASK); + + irr &= ~CC_HOST_RNG_IRQ_MASK; + + /* schedule execution of deferred work handler */ + schedule_work(&drvdata->compwork); + } + + if (irr) { + dev_dbg_ratelimited(dev, + "IRR includes unknown cause bits (0x%08X)\n", + irr); + /* Just warning */ + } + + return IRQ_HANDLED; +} + +static void cc_trng_startwork_handler(struct work_struct *w) +{ + struct cctrng_drvdata *drvdata = + container_of(w, struct cctrng_drvdata, startwork); + + drvdata->active_rosc = 0; + cc_trng_hw_trigger(drvdata); +} + + +static int cc_trng_clk_init(struct cctrng_drvdata *drvdata) +{ + struct clk *clk; + struct device *dev = &(drvdata->pdev->dev); + int rc = 0; + + clk = devm_clk_get_optional(dev, NULL); + if (IS_ERR(clk)) { + if (PTR_ERR(clk) != -EPROBE_DEFER) + dev_err(dev, "Error getting clock: %pe\n", clk); + return PTR_ERR(clk); + } + drvdata->clk = clk; + + rc = clk_prepare_enable(drvdata->clk); + if (rc) { + dev_err(dev, "Failed to enable clock\n"); + return rc; + } + + return 0; +} + +static void cc_trng_clk_fini(struct cctrng_drvdata *drvdata) +{ + clk_disable_unprepare(drvdata->clk); +} + + +static int cctrng_probe(struct platform_device *pdev) +{ + struct resource *req_mem_cc_regs = NULL; + struct cctrng_drvdata *drvdata; + struct device *dev = &pdev->dev; + int rc = 0; + u32 val; + int irq; + + drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL); + if (!drvdata) + return -ENOMEM; + + drvdata->rng.name = devm_kstrdup(dev, dev_name(dev), GFP_KERNEL); + if (!drvdata->rng.name) + return -ENOMEM; + + drvdata->rng.read = cctrng_read; + drvdata->rng.priv = (unsigned long)drvdata; + drvdata->rng.quality = CC_TRNG_QUALITY; + + platform_set_drvdata(pdev, drvdata); + drvdata->pdev = pdev; + + drvdata->circ.buf = (char *)drvdata->data_buf; + + /* Get device resources */ + /* First CC registers space */ + req_mem_cc_regs = platform_get_resource(pdev, IORESOURCE_MEM, 0); + /* Map registers space */ + drvdata->cc_base = devm_ioremap_resource(dev, req_mem_cc_regs); + if (IS_ERR(drvdata->cc_base)) { + dev_err(dev, "Failed to ioremap registers"); + return PTR_ERR(drvdata->cc_base); + } + + dev_dbg(dev, "Got MEM resource (%s): %pR\n", req_mem_cc_regs->name, + req_mem_cc_regs); + dev_dbg(dev, "CC registers mapped from %pa to 0x%p\n", + &req_mem_cc_regs->start, drvdata->cc_base); + + /* Then IRQ */ + irq = platform_get_irq(pdev, 0); + if (irq < 0) { + dev_err(dev, "Failed getting IRQ resource\n"); + return irq; + } + + /* parse sampling rate from device tree */ + rc = cc_trng_parse_sampling_ratio(drvdata); + if (rc) { + dev_err(dev, "Failed to get legal sampling ratio for rosc\n"); + return rc; + } + + rc = cc_trng_clk_init(drvdata); + if (rc) { + dev_err(dev, "cc_trng_clk_init failed\n"); + return rc; + } + + INIT_WORK(&drvdata->compwork, cc_trng_compwork_handler); + INIT_WORK(&drvdata->startwork, cc_trng_startwork_handler); + spin_lock_init(&drvdata->read_lock); + + /* register the driver isr function */ + rc = devm_request_irq(dev, irq, cc_isr, IRQF_SHARED, "cctrng", drvdata); + if (rc) { + dev_err(dev, "Could not register to interrupt %d\n", irq); + goto post_clk_err; + } + dev_dbg(dev, "Registered to IRQ: %d\n", irq); + + /* Clear all pending interrupts */ + val = cc_ioread(drvdata, CC_HOST_RGF_IRR_REG_OFFSET); + dev_dbg(dev, "IRR=0x%08X\n", val); + cc_iowrite(drvdata, CC_HOST_RGF_ICR_REG_OFFSET, val); + + /* unmask HOST RNG interrupt */ + cc_iowrite(drvdata, CC_HOST_RGF_IMR_REG_OFFSET, + cc_ioread(drvdata, CC_HOST_RGF_IMR_REG_OFFSET) & + ~CC_HOST_RNG_IRQ_MASK); + + /* init PM */ + rc = cc_trng_pm_init(drvdata); + if (rc) { + dev_err(dev, "cc_trng_pm_init failed\n"); + goto post_clk_err; + } + + /* increment device's usage counter */ + rc = cc_trng_pm_get(dev); + if (rc) { + dev_err(dev, "cc_trng_pm_get returned %x\n", rc); + goto post_pm_err; + } + + /* set pending_hw to verify that HW won't be triggered from read */ + atomic_set(&drvdata->pending_hw, 1); + + /* registration of the hwrng device */ + rc = hwrng_register(&drvdata->rng); + if (rc) { + dev_err(dev, "Could not register hwrng device.\n"); + goto post_pm_err; + } + + /* trigger HW to start generate data */ + drvdata->active_rosc = 0; + cc_trng_hw_trigger(drvdata); + + /* All set, we can allow auto-suspend */ + cc_trng_pm_go(drvdata); + + dev_info(dev, "ARM cctrng device initialized\n"); + + return 0; + +post_pm_err: + cc_trng_pm_fini(drvdata); + +post_clk_err: + cc_trng_clk_fini(drvdata); + + return rc; +} + +static int cctrng_remove(struct platform_device *pdev) +{ + struct cctrng_drvdata *drvdata = platform_get_drvdata(pdev); + struct device *dev = &pdev->dev; + + dev_dbg(dev, "Releasing cctrng resources...\n"); + + hwrng_unregister(&drvdata->rng); + + cc_trng_pm_fini(drvdata); + + cc_trng_clk_fini(drvdata); + + dev_info(dev, "ARM cctrng device terminated\n"); + + return 0; +} + +static int __maybe_unused cctrng_suspend(struct device *dev) +{ + struct cctrng_drvdata *drvdata = dev_get_drvdata(dev); + + dev_dbg(dev, "set HOST_POWER_DOWN_EN\n"); + cc_iowrite(drvdata, CC_HOST_POWER_DOWN_EN_REG_OFFSET, + POWER_DOWN_ENABLE); + + clk_disable_unprepare(drvdata->clk); + + return 0; +} + +static bool cctrng_wait_for_reset_completion(struct cctrng_drvdata *drvdata) +{ + unsigned int val; + unsigned int i; + + for (i = 0; i < CC_HW_RESET_LOOP_COUNT; i++) { + /* in cc7x3 NVM_IS_IDLE indicates that CC reset is + * completed and device is fully functional + */ + val = cc_ioread(drvdata, CC_NVM_IS_IDLE_REG_OFFSET); + if (val & BIT(CC_NVM_IS_IDLE_VALUE_BIT_SHIFT)) { + /* hw indicate reset completed */ + return true; + } + /* allow scheduling other process on the processor */ + schedule(); + } + /* reset not completed */ + return false; +} + +static int __maybe_unused cctrng_resume(struct device *dev) +{ + struct cctrng_drvdata *drvdata = dev_get_drvdata(dev); + int rc; + + dev_dbg(dev, "unset HOST_POWER_DOWN_EN\n"); + /* Enables the device source clk */ + rc = clk_prepare_enable(drvdata->clk); + if (rc) { + dev_err(dev, "failed getting clock back on. We're toast.\n"); + return rc; + } + + /* wait for Cryptocell reset completion */ + if (!cctrng_wait_for_reset_completion(drvdata)) { + dev_err(dev, "Cryptocell reset not completed"); + return -EBUSY; + } + + /* unmask HOST RNG interrupt */ + cc_iowrite(drvdata, CC_HOST_RGF_IMR_REG_OFFSET, + cc_ioread(drvdata, CC_HOST_RGF_IMR_REG_OFFSET) & + ~CC_HOST_RNG_IRQ_MASK); + + cc_iowrite(drvdata, CC_HOST_POWER_DOWN_EN_REG_OFFSET, + POWER_DOWN_DISABLE); + + return 0; +} + +static UNIVERSAL_DEV_PM_OPS(cctrng_pm, cctrng_suspend, cctrng_resume, NULL); + +static const struct of_device_id arm_cctrng_dt_match[] = { + { .compatible = "arm,cryptocell-713-trng", }, + { .compatible = "arm,cryptocell-703-trng", }, + {}, +}; +MODULE_DEVICE_TABLE(of, arm_cctrng_dt_match); + +static struct platform_driver cctrng_driver = { + .driver = { + .name = "cctrng", + .of_match_table = arm_cctrng_dt_match, + .pm = &cctrng_pm, + }, + .probe = cctrng_probe, + .remove = cctrng_remove, +}; + +static int __init cctrng_mod_init(void) +{ + /* Compile time assertion checks */ + BUILD_BUG_ON(CCTRNG_DATA_BUF_WORDS < 6); + BUILD_BUG_ON((CCTRNG_DATA_BUF_WORDS & (CCTRNG_DATA_BUF_WORDS-1)) != 0); + + return platform_driver_register(&cctrng_driver); +} +module_init(cctrng_mod_init); + +static void __exit cctrng_mod_exit(void) +{ + platform_driver_unregister(&cctrng_driver); +} +module_exit(cctrng_mod_exit); + +/* Module description */ +MODULE_DESCRIPTION("ARM CryptoCell TRNG Driver"); +MODULE_AUTHOR("ARM"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/char/hw_random/cctrng.h b/drivers/char/hw_random/cctrng.h new file mode 100644 index 000000000000..1f2fde95adcb --- /dev/null +++ b/drivers/char/hw_random/cctrng.h @@ -0,0 +1,72 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright (C) 2019-2020 ARM Limited or its affiliates. */ + +#include <linux/bitops.h> + +#define POWER_DOWN_ENABLE 0x01 +#define POWER_DOWN_DISABLE 0x00 + +/* hwrng quality: bits of true entropy per 1024 bits of input */ +#define CC_TRNG_QUALITY 1024 + +/* CryptoCell TRNG HW definitions */ +#define CC_TRNG_NUM_OF_ROSCS 4 +/* The number of words generated in the entropy holding register (EHR) + * 6 words (192 bit) according to HW implementation + */ +#define CC_TRNG_EHR_IN_WORDS 6 +#define CC_TRNG_EHR_IN_BITS (CC_TRNG_EHR_IN_WORDS * BITS_PER_TYPE(u32)) + +#define CC_HOST_RNG_IRQ_MASK BIT(CC_HOST_RGF_IRR_RNG_INT_BIT_SHIFT) + +/* RNG interrupt mask */ +#define CC_RNG_INT_MASK (BIT(CC_RNG_IMR_EHR_VALID_INT_MASK_BIT_SHIFT) | \ + BIT(CC_RNG_IMR_AUTOCORR_ERR_INT_MASK_BIT_SHIFT) | \ + BIT(CC_RNG_IMR_CRNGT_ERR_INT_MASK_BIT_SHIFT) | \ + BIT(CC_RNG_IMR_VN_ERR_INT_MASK_BIT_SHIFT) | \ + BIT(CC_RNG_IMR_WATCHDOG_INT_MASK_BIT_SHIFT)) + +// -------------------------------------- +// BLOCK: RNG +// -------------------------------------- +#define CC_RNG_IMR_REG_OFFSET 0x0100UL +#define CC_RNG_IMR_EHR_VALID_INT_MASK_BIT_SHIFT 0x0UL +#define CC_RNG_IMR_AUTOCORR_ERR_INT_MASK_BIT_SHIFT 0x1UL +#define CC_RNG_IMR_CRNGT_ERR_INT_MASK_BIT_SHIFT 0x2UL +#define CC_RNG_IMR_VN_ERR_INT_MASK_BIT_SHIFT 0x3UL +#define CC_RNG_IMR_WATCHDOG_INT_MASK_BIT_SHIFT 0x4UL +#define CC_RNG_ISR_REG_OFFSET 0x0104UL +#define CC_RNG_ISR_EHR_VALID_BIT_SHIFT 0x0UL +#define CC_RNG_ISR_EHR_VALID_BIT_SIZE 0x1UL +#define CC_RNG_ISR_AUTOCORR_ERR_BIT_SHIFT 0x1UL +#define CC_RNG_ISR_AUTOCORR_ERR_BIT_SIZE 0x1UL +#define CC_RNG_ISR_CRNGT_ERR_BIT_SHIFT 0x2UL +#define CC_RNG_ISR_CRNGT_ERR_BIT_SIZE 0x1UL +#define CC_RNG_ISR_WATCHDOG_BIT_SHIFT 0x4UL +#define CC_RNG_ISR_WATCHDOG_BIT_SIZE 0x1UL +#define CC_RNG_ICR_REG_OFFSET 0x0108UL +#define CC_TRNG_CONFIG_REG_OFFSET 0x010CUL +#define CC_EHR_DATA_0_REG_OFFSET 0x0114UL +#define CC_RND_SOURCE_ENABLE_REG_OFFSET 0x012CUL +#define CC_SAMPLE_CNT1_REG_OFFSET 0x0130UL +#define CC_TRNG_DEBUG_CONTROL_REG_OFFSET 0x0138UL +#define CC_RNG_SW_RESET_REG_OFFSET 0x0140UL +#define CC_RNG_CLK_ENABLE_REG_OFFSET 0x01C4UL +#define CC_RNG_DMA_ENABLE_REG_OFFSET 0x01C8UL +#define CC_RNG_WATCHDOG_VAL_REG_OFFSET 0x01D8UL +// -------------------------------------- +// BLOCK: SEC_HOST_RGF +// -------------------------------------- +#define CC_HOST_RGF_IRR_REG_OFFSET 0x0A00UL +#define CC_HOST_RGF_IRR_RNG_INT_BIT_SHIFT 0xAUL +#define CC_HOST_RGF_IMR_REG_OFFSET 0x0A04UL +#define CC_HOST_RGF_ICR_REG_OFFSET 0x0A08UL + +#define CC_HOST_POWER_DOWN_EN_REG_OFFSET 0x0A78UL + +// -------------------------------------- +// BLOCK: NVM +// -------------------------------------- +#define CC_NVM_IS_IDLE_REG_OFFSET 0x0F10UL +#define CC_NVM_IS_IDLE_VALUE_BIT_SHIFT 0x0UL +#define CC_NVM_IS_IDLE_VALUE_BIT_SIZE 0x1UL diff --git a/drivers/char/hw_random/omap-rng.c b/drivers/char/hw_random/omap-rng.c index 65952393e1bb..7290c603fcb8 100644 --- a/drivers/char/hw_random/omap-rng.c +++ b/drivers/char/hw_random/omap-rng.c @@ -392,11 +392,8 @@ static int of_get_omap_rng_device_details(struct omap_rng_dev *priv, if (of_device_is_compatible(dev->of_node, "ti,omap4-rng") || of_device_is_compatible(dev->of_node, "inside-secure,safexcel-eip76")) { irq = platform_get_irq(pdev, 0); - if (irq < 0) { - dev_err(dev, "%s: error getting IRQ resource - %d\n", - __func__, irq); + if (irq < 0) return irq; - } err = devm_request_irq(dev, irq, omap4_rng_irq, IRQF_TRIGGER_NONE, dev_name(dev), priv); diff --git a/drivers/char/hw_random/optee-rng.c b/drivers/char/hw_random/optee-rng.c index ddfbabaa5f8f..49b2e02537dd 100644 --- a/drivers/char/hw_random/optee-rng.c +++ b/drivers/char/hw_random/optee-rng.c @@ -226,7 +226,7 @@ static int optee_rng_probe(struct device *dev) return -ENODEV; /* Open session with hwrng Trusted App */ - memcpy(sess_arg.uuid, rng_device->id.uuid.b, TEE_IOCTL_UUID_LEN); + export_uuid(sess_arg.uuid, &rng_device->id.uuid); sess_arg.clnt_login = TEE_IOCTL_LOGIN_PUBLIC; sess_arg.num_params = 0; diff --git a/drivers/char/hw_random/xgene-rng.c b/drivers/char/hw_random/xgene-rng.c index d7516a446987..008e6db9ce01 100644 --- a/drivers/char/hw_random/xgene-rng.c +++ b/drivers/char/hw_random/xgene-rng.c @@ -328,10 +328,8 @@ static int xgene_rng_probe(struct platform_device *pdev) return PTR_ERR(ctx->csr_base); rc = platform_get_irq(pdev, 0); - if (rc < 0) { - dev_err(&pdev->dev, "No IRQ resource\n"); + if (rc < 0) return rc; - } ctx->irq = rc; dev_dbg(&pdev->dev, "APM X-Gene RNG BASE %p ALARM IRQ %d", diff --git a/drivers/char/random.c b/drivers/char/random.c index 0d10e31fd342..cae02b2a871c 100644 --- a/drivers/char/random.c +++ b/drivers/char/random.c @@ -327,7 +327,6 @@ #include <linux/spinlock.h> #include <linux/kthread.h> #include <linux/percpu.h> -#include <linux/cryptohash.h> #include <linux/fips.h> #include <linux/ptrace.h> #include <linux/workqueue.h> @@ -337,6 +336,7 @@ #include <linux/completion.h> #include <linux/uuid.h> #include <crypto/chacha.h> +#include <crypto/sha.h> #include <asm/processor.h> #include <linux/uaccess.h> @@ -1397,14 +1397,14 @@ static void extract_buf(struct entropy_store *r, __u8 *out) __u32 w[5]; unsigned long l[LONGS(20)]; } hash; - __u32 workspace[SHA_WORKSPACE_WORDS]; + __u32 workspace[SHA1_WORKSPACE_WORDS]; unsigned long flags; /* * If we have an architectural hardware random number * generator, use it for SHA's initial vector */ - sha_init(hash.w); + sha1_init(hash.w); for (i = 0; i < LONGS(20); i++) { unsigned long v; if (!arch_get_random_long(&v)) @@ -1415,7 +1415,7 @@ static void extract_buf(struct entropy_store *r, __u8 *out) /* Generate a hash across the pool, 16 words (512 bits) at a time */ spin_lock_irqsave(&r->lock, flags); for (i = 0; i < r->poolinfo->poolwords; i += 16) - sha_transform(hash.w, (__u8 *)(r->pool + i), workspace); + sha1_transform(hash.w, (__u8 *)(r->pool + i), workspace); /* * We mix the hash back into the pool to prevent backtracking diff --git a/drivers/crypto/allwinner/sun8i-ss/sun8i-ss-core.c b/drivers/crypto/allwinner/sun8i-ss/sun8i-ss-core.c index 6b301afffd11..a1fb2fbdbe7b 100644 --- a/drivers/crypto/allwinner/sun8i-ss/sun8i-ss-core.c +++ b/drivers/crypto/allwinner/sun8i-ss/sun8i-ss-core.c @@ -537,10 +537,8 @@ static int sun8i_ss_probe(struct platform_device *pdev) return err; irq = platform_get_irq(pdev, 0); - if (irq < 0) { - dev_err(ss->dev, "Cannot get SecuritySystem IRQ\n"); + if (irq < 0) return irq; - } ss->reset = devm_reset_control_get(&pdev->dev, NULL); if (IS_ERR(ss->reset)) { diff --git a/drivers/crypto/amlogic/amlogic-gxl-core.c b/drivers/crypto/amlogic/amlogic-gxl-core.c index 9d4ead2f7ebb..411857fad8ba 100644 --- a/drivers/crypto/amlogic/amlogic-gxl-core.c +++ b/drivers/crypto/amlogic/amlogic-gxl-core.c @@ -253,10 +253,8 @@ static int meson_crypto_probe(struct platform_device *pdev) mc->irqs = devm_kcalloc(mc->dev, MAXFLOW, sizeof(int), GFP_KERNEL); for (i = 0; i < MAXFLOW; i++) { mc->irqs[i] = platform_get_irq(pdev, i); - if (mc->irqs[i] < 0) { - dev_err(mc->dev, "Cannot get IRQ for flow %d\n", i); + if (mc->irqs[i] < 0) return mc->irqs[i]; - } err = devm_request_irq(&pdev->dev, mc->irqs[i], meson_irq_handler, 0, "gxl-crypto", mc); diff --git a/drivers/crypto/atmel-sha.c b/drivers/crypto/atmel-sha.c index e536e2a6bbd8..75ccf41a7cb9 100644 --- a/drivers/crypto/atmel-sha.c +++ b/drivers/crypto/atmel-sha.c @@ -31,7 +31,6 @@ #include <linux/of_device.h> #include <linux/delay.h> #include <linux/crypto.h> -#include <linux/cryptohash.h> #include <crypto/scatterwalk.h> #include <crypto/algapi.h> #include <crypto/sha.h> diff --git a/drivers/crypto/axis/artpec6_crypto.c b/drivers/crypto/axis/artpec6_crypto.c index fcf1effc7661..62ba0325a618 100644 --- a/drivers/crypto/axis/artpec6_crypto.c +++ b/drivers/crypto/axis/artpec6_crypto.c @@ -2239,16 +2239,12 @@ artpec6_crypto_hash_set_key(struct crypto_ahash *tfm, blocksize = crypto_tfm_alg_blocksize(crypto_ahash_tfm(tfm)); if (keylen > blocksize) { - SHASH_DESC_ON_STACK(hdesc, tfm_ctx->child_hash); - - hdesc->tfm = tfm_ctx->child_hash; - tfm_ctx->hmac_key_length = blocksize; - ret = crypto_shash_digest(hdesc, key, keylen, - tfm_ctx->hmac_key); + + ret = crypto_shash_tfm_digest(tfm_ctx->child_hash, key, keylen, + tfm_ctx->hmac_key); if (ret) return ret; - } else { memcpy(tfm_ctx->hmac_key, key, keylen); tfm_ctx->hmac_key_length = keylen; diff --git a/drivers/crypto/bcm/cipher.c b/drivers/crypto/bcm/cipher.c index c8b9408541a9..a353217a0d33 100644 --- a/drivers/crypto/bcm/cipher.c +++ b/drivers/crypto/bcm/cipher.c @@ -308,9 +308,9 @@ static int handle_skcipher_req(struct iproc_reqctx_s *rctx) container_of(areq, struct skcipher_request, base); struct iproc_ctx_s *ctx = rctx->ctx; struct spu_cipher_parms cipher_parms; - int err = 0; - unsigned int chunksize = 0; /* Num bytes of request to submit */ - int remaining = 0; /* Bytes of request still to process */ + int err; + unsigned int chunksize; /* Num bytes of request to submit */ + int remaining; /* Bytes of request still to process */ int chunk_start; /* Beginning of data for current SPU msg */ /* IV or ctr value to use in this SPU msg */ @@ -698,7 +698,7 @@ static int handle_ahash_req(struct iproc_reqctx_s *rctx) /* number of bytes still to be hashed in this req */ unsigned int nbytes_to_hash = 0; - int err = 0; + int err; unsigned int chunksize = 0; /* length of hash carry + new data */ /* * length of new data, not from hash carry, to be submitted in @@ -1664,7 +1664,7 @@ static void spu_rx_callback(struct mbox_client *cl, void *msg) struct spu_hw *spu = &iproc_priv.spu; struct brcm_message *mssg = msg; struct iproc_reqctx_s *rctx; - int err = 0; + int err; rctx = mssg->ctx; if (unlikely(!rctx)) { @@ -1967,7 +1967,7 @@ static int ahash_enqueue(struct ahash_request *req) struct iproc_reqctx_s *rctx = ahash_request_ctx(req); struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); struct iproc_ctx_s *ctx = crypto_ahash_ctx(tfm); - int err = 0; + int err; const char *alg_name; flow_log("ahash_enqueue() nbytes:%u\n", req->nbytes); @@ -2299,7 +2299,7 @@ ahash_finup_exit: static int ahash_digest(struct ahash_request *req) { - int err = 0; + int err; flow_log("ahash_digest() nbytes:%u\n", req->nbytes); @@ -4436,7 +4436,7 @@ static int spu_mb_init(struct device *dev) for (i = 0; i < iproc_priv.spu.num_chan; i++) { iproc_priv.mbox[i] = mbox_request_channel(mcl, i); if (IS_ERR(iproc_priv.mbox[i])) { - err = (int)PTR_ERR(iproc_priv.mbox[i]); + err = PTR_ERR(iproc_priv.mbox[i]); dev_err(dev, "Mbox channel %d request failed with err %d", i, err); @@ -4717,21 +4717,20 @@ static int spu_dt_read(struct platform_device *pdev) matched_spu_type = of_device_get_match_data(dev); if (!matched_spu_type) { - dev_err(&pdev->dev, "Failed to match device\n"); + dev_err(dev, "Failed to match device\n"); return -ENODEV; } spu->spu_type = matched_spu_type->type; spu->spu_subtype = matched_spu_type->subtype; - i = 0; for (i = 0; (i < MAX_SPUS) && ((spu_ctrl_regs = platform_get_resource(pdev, IORESOURCE_MEM, i)) != NULL); i++) { spu->reg_vbase[i] = devm_ioremap_resource(dev, spu_ctrl_regs); if (IS_ERR(spu->reg_vbase[i])) { err = PTR_ERR(spu->reg_vbase[i]); - dev_err(&pdev->dev, "Failed to map registers: %d\n", + dev_err(dev, "Failed to map registers: %d\n", err); spu->reg_vbase[i] = NULL; return err; @@ -4747,7 +4746,7 @@ static int bcm_spu_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; struct spu_hw *spu = &iproc_priv.spu; - int err = 0; + int err; iproc_priv.pdev = pdev; platform_set_drvdata(iproc_priv.pdev, @@ -4757,7 +4756,7 @@ static int bcm_spu_probe(struct platform_device *pdev) if (err < 0) goto failure; - err = spu_mb_init(&pdev->dev); + err = spu_mb_init(dev); if (err < 0) goto failure; @@ -4766,7 +4765,7 @@ static int bcm_spu_probe(struct platform_device *pdev) else if (spu->spu_type == SPU_TYPE_SPU2) iproc_priv.bcm_hdr_len = 0; - spu_functions_register(&pdev->dev, spu->spu_type, spu->spu_subtype); + spu_functions_register(dev, spu->spu_type, spu->spu_subtype); spu_counters_init(); diff --git a/drivers/crypto/cavium/nitrox/nitrox_main.c b/drivers/crypto/cavium/nitrox/nitrox_main.c index e91be9b8b083..788c6607078b 100644 --- a/drivers/crypto/cavium/nitrox/nitrox_main.c +++ b/drivers/crypto/cavium/nitrox/nitrox_main.c @@ -346,7 +346,7 @@ static void nitrox_pf_sw_cleanup(struct nitrox_device *ndev) } /** - * nitrox_bist_check - Check NITORX BIST registers status + * nitrox_bist_check - Check NITROX BIST registers status * @ndev: NITROX device */ static int nitrox_bist_check(struct nitrox_device *ndev) diff --git a/drivers/crypto/ccp/Kconfig b/drivers/crypto/ccp/Kconfig index e0a8bd15aa74..32268e239bf1 100644 --- a/drivers/crypto/ccp/Kconfig +++ b/drivers/crypto/ccp/Kconfig @@ -10,10 +10,9 @@ config CRYPTO_DEV_CCP_DD config CRYPTO_DEV_SP_CCP bool "Cryptographic Coprocessor device" default y - depends on CRYPTO_DEV_CCP_DD + depends on CRYPTO_DEV_CCP_DD && DMADEVICES select HW_RANDOM select DMA_ENGINE - select DMADEVICES select CRYPTO_SHA1 select CRYPTO_SHA256 help diff --git a/drivers/crypto/ccp/ccp-crypto-sha.c b/drivers/crypto/ccp/ccp-crypto-sha.c index 474e6f1a6a84..b0cc2bd73af8 100644 --- a/drivers/crypto/ccp/ccp-crypto-sha.c +++ b/drivers/crypto/ccp/ccp-crypto-sha.c @@ -272,9 +272,6 @@ static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key, { struct ccp_ctx *ctx = crypto_tfm_ctx(crypto_ahash_tfm(tfm)); struct crypto_shash *shash = ctx->u.sha.hmac_tfm; - - SHASH_DESC_ON_STACK(sdesc, shash); - unsigned int block_size = crypto_shash_blocksize(shash); unsigned int digest_size = crypto_shash_digestsize(shash); int i, ret; @@ -289,10 +286,8 @@ static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key, if (key_len > block_size) { /* Must hash the input key */ - sdesc->tfm = shash; - - ret = crypto_shash_digest(sdesc, key, key_len, - ctx->u.sha.key); + ret = crypto_shash_tfm_digest(shash, key, key_len, + ctx->u.sha.key); if (ret) return -EINVAL; diff --git a/drivers/crypto/ccp/sev-dev.c b/drivers/crypto/ccp/sev-dev.c index 896f190b9a50..439cd737076e 100644 --- a/drivers/crypto/ccp/sev-dev.c +++ b/drivers/crypto/ccp/sev-dev.c @@ -20,6 +20,7 @@ #include <linux/hw_random.h> #include <linux/ccp.h> #include <linux/firmware.h> +#include <linux/gfp.h> #include <asm/smp.h> @@ -44,6 +45,14 @@ MODULE_PARM_DESC(psp_probe_timeout, " default timeout value, in seconds, during static bool psp_dead; static int psp_timeout; +/* Trusted Memory Region (TMR): + * The TMR is a 1MB area that must be 1MB aligned. Use the page allocator + * to allocate the memory, which will return aligned memory for the specified + * allocation order. + */ +#define SEV_ES_TMR_SIZE (1024 * 1024) +static void *sev_es_tmr; + static inline bool sev_version_greater_or_equal(u8 maj, u8 min) { struct sev_device *sev = psp_master->sev_data; @@ -214,6 +223,20 @@ static int __sev_platform_init_locked(int *error) if (sev->state == SEV_STATE_INIT) return 0; + if (sev_es_tmr) { + u64 tmr_pa; + + /* + * Do not include the encryption mask on the physical + * address of the TMR (firmware should clear it anyway). + */ + tmr_pa = __pa(sev_es_tmr); + + sev->init_cmd_buf.flags |= SEV_INIT_FLAGS_SEV_ES; + sev->init_cmd_buf.tmr_address = tmr_pa; + sev->init_cmd_buf.tmr_len = SEV_ES_TMR_SIZE; + } + rc = __sev_do_cmd_locked(SEV_CMD_INIT, &sev->init_cmd_buf, error); if (rc) return rc; @@ -1012,6 +1035,7 @@ EXPORT_SYMBOL_GPL(sev_issue_cmd_external_user); void sev_pci_init(void) { struct sev_device *sev = psp_master->sev_data; + struct page *tmr_page; int error, rc; if (!sev) @@ -1041,6 +1065,16 @@ void sev_pci_init(void) sev_update_firmware(sev->dev) == 0) sev_get_api_version(); + /* Obtain the TMR memory area for SEV-ES use */ + tmr_page = alloc_pages(GFP_KERNEL, get_order(SEV_ES_TMR_SIZE)); + if (tmr_page) { + sev_es_tmr = page_address(tmr_page); + } else { + sev_es_tmr = NULL; + dev_warn(sev->dev, + "SEV: TMR allocation failed, SEV-ES support unavailable\n"); + } + /* Initialize the platform */ rc = sev_platform_init(&error); if (rc && (error == SEV_RET_SECURE_DATA_INVALID)) { @@ -1075,4 +1109,13 @@ void sev_pci_exit(void) return; sev_platform_shutdown(NULL); + + if (sev_es_tmr) { + /* The TMR area was encrypted, flush it from the cache */ + wbinvd_on_all_cpus(); + + free_pages((unsigned long)sev_es_tmr, + get_order(SEV_ES_TMR_SIZE)); + sev_es_tmr = NULL; + } } diff --git a/drivers/crypto/ccree/cc_cipher.c b/drivers/crypto/ccree/cc_cipher.c index a84335328f37..872ea3ff1c6b 100644 --- a/drivers/crypto/ccree/cc_cipher.c +++ b/drivers/crypto/ccree/cc_cipher.c @@ -427,12 +427,9 @@ static int cc_cipher_setkey(struct crypto_skcipher *sktfm, const u8 *key, int key_len = keylen >> 1; int err; - SHASH_DESC_ON_STACK(desc, ctx_p->shash_tfm); - - desc->tfm = ctx_p->shash_tfm; - - err = crypto_shash_digest(desc, ctx_p->user.key, key_len, - ctx_p->user.key + key_len); + err = crypto_shash_tfm_digest(ctx_p->shash_tfm, + ctx_p->user.key, key_len, + ctx_p->user.key + key_len); if (err) { dev_err(dev, "Failed to hash ESSIV key.\n"); return err; diff --git a/drivers/crypto/ccree/cc_debugfs.c b/drivers/crypto/ccree/cc_debugfs.c index c454afce7781..7083767602fc 100644 --- a/drivers/crypto/ccree/cc_debugfs.c +++ b/drivers/crypto/ccree/cc_debugfs.c @@ -26,7 +26,7 @@ static struct debugfs_reg32 ver_sig_regs[] = { { .name = "VERSION" }, /* Must be 1st */ }; -static struct debugfs_reg32 pid_cid_regs[] = { +static const struct debugfs_reg32 pid_cid_regs[] = { CC_DEBUG_REG(PERIPHERAL_ID_0), CC_DEBUG_REG(PERIPHERAL_ID_1), CC_DEBUG_REG(PERIPHERAL_ID_2), @@ -38,7 +38,7 @@ static struct debugfs_reg32 pid_cid_regs[] = { CC_DEBUG_REG(COMPONENT_ID_3), }; -static struct debugfs_reg32 debug_regs[] = { +static const struct debugfs_reg32 debug_regs[] = { CC_DEBUG_REG(HOST_IRR), CC_DEBUG_REG(HOST_POWER_DOWN_EN), CC_DEBUG_REG(AXIM_MON_ERR), diff --git a/drivers/crypto/chelsio/chcr_algo.c b/drivers/crypto/chelsio/chcr_algo.c index c29b80dd30d8..caf1136e7ef9 100644 --- a/drivers/crypto/chelsio/chcr_algo.c +++ b/drivers/crypto/chelsio/chcr_algo.c @@ -44,7 +44,6 @@ #include <linux/kernel.h> #include <linux/module.h> #include <linux/crypto.h> -#include <linux/cryptohash.h> #include <linux/skbuff.h> #include <linux/rtnetlink.h> #include <linux/highmem.h> @@ -1757,7 +1756,7 @@ static int chcr_ahash_final(struct ahash_request *req) struct uld_ctx *u_ctx = ULD_CTX(h_ctx(rtfm)); struct chcr_context *ctx = h_ctx(rtfm); u8 bs = crypto_tfm_alg_blocksize(crypto_ahash_tfm(rtfm)); - int error = -EINVAL; + int error; unsigned int cpu; cpu = get_cpu(); diff --git a/drivers/crypto/chelsio/chcr_ipsec.c b/drivers/crypto/chelsio/chcr_ipsec.c index 9fd3b9d1ec2f..25bf6d963066 100644 --- a/drivers/crypto/chelsio/chcr_ipsec.c +++ b/drivers/crypto/chelsio/chcr_ipsec.c @@ -40,7 +40,6 @@ #include <linux/kernel.h> #include <linux/module.h> #include <linux/crypto.h> -#include <linux/cryptohash.h> #include <linux/skbuff.h> #include <linux/rtnetlink.h> #include <linux/highmem.h> diff --git a/drivers/crypto/hisilicon/Kconfig b/drivers/crypto/hisilicon/Kconfig index f09c6cf7823e..9c3b3ca815e6 100644 --- a/drivers/crypto/hisilicon/Kconfig +++ b/drivers/crypto/hisilicon/Kconfig @@ -29,6 +29,7 @@ config CRYPTO_DEV_HISI_SEC2 depends on PCI && PCI_MSI depends on UACCE || UACCE=n depends on ARM64 || (COMPILE_TEST && 64BIT) + depends on ACPI help Support for HiSilicon SEC Engine of version 2 in crypto subsystem. It provides AES, SM4, and 3DES algorithms with ECB @@ -42,6 +43,7 @@ config CRYPTO_DEV_HISI_QM depends on ARM64 || COMPILE_TEST depends on PCI && PCI_MSI depends on UACCE || UACCE=n + depends on ACPI help HiSilicon accelerator engines use a common queue management interface. Specific engine driver may use this module. @@ -52,6 +54,7 @@ config CRYPTO_DEV_HISI_ZIP depends on ARM64 || (COMPILE_TEST && 64BIT) depends on !CPU_BIG_ENDIAN || COMPILE_TEST depends on UACCE || UACCE=n + depends on ACPI select CRYPTO_DEV_HISI_QM help Support for HiSilicon ZIP Driver @@ -61,6 +64,7 @@ config CRYPTO_DEV_HISI_HPRE depends on PCI && PCI_MSI depends on UACCE || UACCE=n depends on ARM64 || (COMPILE_TEST && 64BIT) + depends on ACPI select CRYPTO_DEV_HISI_QM select CRYPTO_DH select CRYPTO_RSA diff --git a/drivers/crypto/hisilicon/hpre/hpre.h b/drivers/crypto/hisilicon/hpre/hpre.h index 03d512ec6336..ed730d173e95 100644 --- a/drivers/crypto/hisilicon/hpre/hpre.h +++ b/drivers/crypto/hisilicon/hpre/hpre.h @@ -25,6 +25,17 @@ enum hpre_ctrl_dbgfs_file { HPRE_DEBUG_FILE_NUM, }; +enum hpre_dfx_dbgfs_file { + HPRE_SEND_CNT, + HPRE_RECV_CNT, + HPRE_SEND_FAIL_CNT, + HPRE_SEND_BUSY_CNT, + HPRE_OVER_THRHLD_CNT, + HPRE_OVERTIME_THRHLD, + HPRE_INVALID_REQ_CNT, + HPRE_DFX_FILE_NUM +}; + #define HPRE_DEBUGFS_FILE_NUM (HPRE_DEBUG_FILE_NUM + HPRE_CLUSTERS_NUM - 1) struct hpre_debugfs_file { @@ -34,6 +45,11 @@ struct hpre_debugfs_file { struct hpre_debug *debug; }; +struct hpre_dfx { + atomic64_t value; + enum hpre_dfx_dbgfs_file type; +}; + /* * One HPRE controller has one PF and multiple VFs, some global configurations * which PF has need this structure. @@ -41,13 +57,13 @@ struct hpre_debugfs_file { */ struct hpre_debug { struct dentry *debug_root; + struct hpre_dfx dfx[HPRE_DFX_FILE_NUM]; struct hpre_debugfs_file files[HPRE_DEBUGFS_FILE_NUM]; }; struct hpre { struct hisi_qm qm; struct hpre_debug debug; - u32 num_vfs; unsigned long status; }; diff --git a/drivers/crypto/hisilicon/hpre/hpre_crypto.c b/drivers/crypto/hisilicon/hpre/hpre_crypto.c index 65425250b2e9..7b5cb27d473d 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_crypto.c +++ b/drivers/crypto/hisilicon/hpre/hpre_crypto.c @@ -10,6 +10,7 @@ #include <linux/dma-mapping.h> #include <linux/fips.h> #include <linux/module.h> +#include <linux/time.h> #include "hpre.h" struct hpre_ctx; @@ -32,6 +33,9 @@ struct hpre_ctx; #define HPRE_SQE_DONE_SHIFT 30 #define HPRE_DH_MAX_P_SZ 512 +#define HPRE_DFX_SEC_TO_US 1000000 +#define HPRE_DFX_US_TO_NS 1000 + typedef void (*hpre_cb)(struct hpre_ctx *ctx, void *sqe); struct hpre_rsa_ctx { @@ -68,6 +72,7 @@ struct hpre_dh_ctx { struct hpre_ctx { struct hisi_qp *qp; struct hpre_asym_request **req_list; + struct hpre *hpre; spinlock_t req_lock; unsigned int key_sz; bool crt_g2_mode; @@ -90,6 +95,7 @@ struct hpre_asym_request { int err; int req_id; hpre_cb cb; + struct timespec64 req_time; }; static DEFINE_MUTEX(hpre_alg_lock); @@ -119,6 +125,7 @@ static void hpre_free_req_id(struct hpre_ctx *ctx, int req_id) static int hpre_add_req_to_ctx(struct hpre_asym_request *hpre_req) { struct hpre_ctx *ctx; + struct hpre_dfx *dfx; int id; ctx = hpre_req->ctx; @@ -129,6 +136,10 @@ static int hpre_add_req_to_ctx(struct hpre_asym_request *hpre_req) ctx->req_list[id] = hpre_req; hpre_req->req_id = id; + dfx = ctx->hpre->debug.dfx; + if (atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value)) + ktime_get_ts64(&hpre_req->req_time); + return id; } @@ -309,12 +320,16 @@ static int hpre_alg_res_post_hf(struct hpre_ctx *ctx, struct hpre_sqe *sqe, static int hpre_ctx_set(struct hpre_ctx *ctx, struct hisi_qp *qp, int qlen) { + struct hpre *hpre; + if (!ctx || !qp || qlen < 0) return -EINVAL; spin_lock_init(&ctx->req_lock); ctx->qp = qp; + hpre = container_of(ctx->qp->qm, struct hpre, qm); + ctx->hpre = hpre; ctx->req_list = kcalloc(qlen, sizeof(void *), GFP_KERNEL); if (!ctx->req_list) return -ENOMEM; @@ -337,38 +352,80 @@ static void hpre_ctx_clear(struct hpre_ctx *ctx, bool is_clear_all) ctx->key_sz = 0; } +static bool hpre_is_bd_timeout(struct hpre_asym_request *req, + u64 overtime_thrhld) +{ + struct timespec64 reply_time; + u64 time_use_us; + + ktime_get_ts64(&reply_time); + time_use_us = (reply_time.tv_sec - req->req_time.tv_sec) * + HPRE_DFX_SEC_TO_US + + (reply_time.tv_nsec - req->req_time.tv_nsec) / + HPRE_DFX_US_TO_NS; + + if (time_use_us <= overtime_thrhld) + return false; + + return true; +} + static void hpre_dh_cb(struct hpre_ctx *ctx, void *resp) { + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; struct hpre_asym_request *req; struct kpp_request *areq; + u64 overtime_thrhld; int ret; ret = hpre_alg_res_post_hf(ctx, resp, (void **)&req); areq = req->areq.dh; areq->dst_len = ctx->key_sz; + + overtime_thrhld = atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value); + if (overtime_thrhld && hpre_is_bd_timeout(req, overtime_thrhld)) + atomic64_inc(&dfx[HPRE_OVER_THRHLD_CNT].value); + hpre_hw_data_clr_all(ctx, req, areq->dst, areq->src); kpp_request_complete(areq, ret); + atomic64_inc(&dfx[HPRE_RECV_CNT].value); } static void hpre_rsa_cb(struct hpre_ctx *ctx, void *resp) { + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; struct hpre_asym_request *req; struct akcipher_request *areq; + u64 overtime_thrhld; int ret; ret = hpre_alg_res_post_hf(ctx, resp, (void **)&req); + + overtime_thrhld = atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value); + if (overtime_thrhld && hpre_is_bd_timeout(req, overtime_thrhld)) + atomic64_inc(&dfx[HPRE_OVER_THRHLD_CNT].value); + areq = req->areq.rsa; areq->dst_len = ctx->key_sz; hpre_hw_data_clr_all(ctx, req, areq->dst, areq->src); akcipher_request_complete(areq, ret); + atomic64_inc(&dfx[HPRE_RECV_CNT].value); } static void hpre_alg_cb(struct hisi_qp *qp, void *resp) { struct hpre_ctx *ctx = qp->qp_ctx; + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; struct hpre_sqe *sqe = resp; + struct hpre_asym_request *req = ctx->req_list[le16_to_cpu(sqe->tag)]; - ctx->req_list[le16_to_cpu(sqe->tag)]->cb(ctx, resp); + + if (unlikely(!req)) { + atomic64_inc(&dfx[HPRE_INVALID_REQ_CNT].value); + return; + } + + req->cb(ctx, resp); } static int hpre_ctx_init(struct hpre_ctx *ctx) @@ -436,6 +493,29 @@ static int hpre_msg_request_set(struct hpre_ctx *ctx, void *req, bool is_rsa) return 0; } +static int hpre_send(struct hpre_ctx *ctx, struct hpre_sqe *msg) +{ + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; + int ctr = 0; + int ret; + + do { + atomic64_inc(&dfx[HPRE_SEND_CNT].value); + ret = hisi_qp_send(ctx->qp, msg); + if (ret != -EBUSY) + break; + atomic64_inc(&dfx[HPRE_SEND_BUSY_CNT].value); + } while (ctr++ < HPRE_TRY_SEND_TIMES); + + if (likely(!ret)) + return ret; + + if (ret != -EBUSY) + atomic64_inc(&dfx[HPRE_SEND_FAIL_CNT].value); + + return ret; +} + #ifdef CONFIG_CRYPTO_DH static int hpre_dh_compute_value(struct kpp_request *req) { @@ -444,7 +524,6 @@ static int hpre_dh_compute_value(struct kpp_request *req) void *tmp = kpp_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; ret = hpre_msg_request_set(ctx, req, false); @@ -465,11 +544,9 @@ static int hpre_dh_compute_value(struct kpp_request *req) msg->dw0 = cpu_to_le32(le32_to_cpu(msg->dw0) | HPRE_ALG_DH_G2); else msg->dw0 = cpu_to_le32(le32_to_cpu(msg->dw0) | HPRE_ALG_DH); - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; @@ -647,7 +724,6 @@ static int hpre_rsa_enc(struct akcipher_request *req) void *tmp = akcipher_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; /* For 512 and 1536 bits key size, use soft tfm instead */ @@ -677,11 +753,8 @@ static int hpre_rsa_enc(struct akcipher_request *req) if (unlikely(ret)) goto clear_all; - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); - /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; @@ -699,7 +772,6 @@ static int hpre_rsa_dec(struct akcipher_request *req) void *tmp = akcipher_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; /* For 512 and 1536 bits key size, use soft tfm instead */ @@ -736,11 +808,8 @@ static int hpre_rsa_dec(struct akcipher_request *req) if (unlikely(ret)) goto clear_all; - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); - /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; diff --git a/drivers/crypto/hisilicon/hpre/hpre_main.c b/drivers/crypto/hisilicon/hpre/hpre_main.c index 88be53bf4a38..a3ee127a70e3 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_main.c +++ b/drivers/crypto/hisilicon/hpre/hpre_main.c @@ -59,10 +59,6 @@ #define HPRE_HAC_ECC2_CNT 0x301a08 #define HPRE_HAC_INT_STATUS 0x301800 #define HPRE_HAC_SOURCE_INT 0x301600 -#define MASTER_GLOBAL_CTRL_SHUTDOWN 1 -#define MASTER_TRANS_RETURN_RW 3 -#define HPRE_MASTER_TRANS_RETURN 0x300150 -#define HPRE_MASTER_GLOBAL_CTRL 0x300000 #define HPRE_CLSTR_ADDR_INTRVL 0x1000 #define HPRE_CLUSTER_INQURY 0x100 #define HPRE_CLSTR_ADDR_INQRY_RSLT 0x104 @@ -80,7 +76,16 @@ #define HPRE_BD_USR_MASK 0x3 #define HPRE_CLUSTER_CORE_MASK 0xf +#define HPRE_AM_OOO_SHUTDOWN_ENB 0x301044 +#define HPRE_AM_OOO_SHUTDOWN_ENABLE BIT(0) +#define HPRE_WR_MSI_PORT BIT(2) + +#define HPRE_CORE_ECC_2BIT_ERR BIT(1) +#define HPRE_OOO_ECC_2BIT_ERR BIT(5) + #define HPRE_VIA_MSI_DSM 1 +#define HPRE_SQE_MASK_OFFSET 8 +#define HPRE_SQE_MASK_LEN 24 static struct hisi_qm_list hpre_devices; static const char hpre_name[] = "hisi_hpre"; @@ -131,7 +136,7 @@ static const u64 hpre_cluster_offsets[] = { HPRE_CLSTR_BASE + HPRE_CLUSTER3 * HPRE_CLSTR_ADDR_INTRVL, }; -static struct debugfs_reg32 hpre_cluster_dfx_regs[] = { +static const struct debugfs_reg32 hpre_cluster_dfx_regs[] = { {"CORES_EN_STATUS ", HPRE_CORE_EN_OFFSET}, {"CORES_INI_CFG ", HPRE_CORE_INI_CFG_OFFSET}, {"CORES_INI_STATUS ", HPRE_CORE_INI_STATUS_OFFSET}, @@ -139,7 +144,7 @@ static struct debugfs_reg32 hpre_cluster_dfx_regs[] = { {"CORES_IS_SCHD ", HPRE_CORE_IS_SCHD_OFFSET}, }; -static struct debugfs_reg32 hpre_com_dfx_regs[] = { +static const struct debugfs_reg32 hpre_com_dfx_regs[] = { {"READ_CLR_EN ", HPRE_CTRL_CNT_CLR_CE}, {"AXQOS ", HPRE_VFG_AXQOS}, {"AWUSR_CFG ", HPRE_AWUSR_FP_CFG}, @@ -156,44 +161,38 @@ static struct debugfs_reg32 hpre_com_dfx_regs[] = { {"INT_STATUS ", HPRE_INT_STATUS}, }; -static int hpre_pf_q_num_set(const char *val, const struct kernel_param *kp) -{ - struct pci_dev *pdev; - u32 n, q_num; - u8 rev_id; - int ret; - - if (!val) - return -EINVAL; - - pdev = pci_get_device(PCI_VENDOR_ID_HUAWEI, HPRE_PCI_DEVICE_ID, NULL); - if (!pdev) { - q_num = HPRE_QUEUE_NUM_V2; - pr_info("No device found currently, suppose queue number is %d\n", - q_num); - } else { - rev_id = pdev->revision; - if (rev_id != QM_HW_V2) - return -EINVAL; - - q_num = HPRE_QUEUE_NUM_V2; - } - - ret = kstrtou32(val, 10, &n); - if (ret != 0 || n == 0 || n > q_num) - return -EINVAL; +static const char *hpre_dfx_files[HPRE_DFX_FILE_NUM] = { + "send_cnt", + "recv_cnt", + "send_fail_cnt", + "send_busy_cnt", + "over_thrhld_cnt", + "overtime_thrhld", + "invalid_req_cnt" +}; - return param_set_int(val, kp); +static int pf_q_num_set(const char *val, const struct kernel_param *kp) +{ + return q_num_set(val, kp, HPRE_PCI_DEVICE_ID); } static const struct kernel_param_ops hpre_pf_q_num_ops = { - .set = hpre_pf_q_num_set, + .set = pf_q_num_set, .get = param_get_int, }; -static u32 hpre_pf_q_num = HPRE_PF_DEF_Q_NUM; -module_param_cb(hpre_pf_q_num, &hpre_pf_q_num_ops, &hpre_pf_q_num, 0444); -MODULE_PARM_DESC(hpre_pf_q_num, "Number of queues in PF of CS(1-1024)"); +static u32 pf_q_num = HPRE_PF_DEF_Q_NUM; +module_param_cb(pf_q_num, &hpre_pf_q_num_ops, &pf_q_num, 0444); +MODULE_PARM_DESC(pf_q_num, "Number of queues in PF of CS(1-1024)"); + +static const struct kernel_param_ops vfs_num_ops = { + .set = vfs_num_set, + .get = param_get_int, +}; + +static u32 vfs_num; +module_param_cb(vfs_num, &vfs_num_ops, &vfs_num, 0444); +MODULE_PARM_DESC(vfs_num, "Number of VFs to enable(1-63), 0(default)"); struct hisi_qp *hpre_create_qp(void) { @@ -232,9 +231,8 @@ static int hpre_cfg_by_dsm(struct hisi_qm *qm) return 0; } -static int hpre_set_user_domain_and_cache(struct hpre *hpre) +static int hpre_set_user_domain_and_cache(struct hisi_qm *qm) { - struct hisi_qm *qm = &hpre->qm; struct device *dev = &qm->pdev->dev; unsigned long offset; int ret, i; @@ -324,17 +322,34 @@ static void hpre_cnt_regs_clear(struct hisi_qm *qm) static void hpre_hw_error_disable(struct hisi_qm *qm) { + u32 val; + /* disable hpre hw error interrupts */ writel(HPRE_CORE_INT_DISABLE, qm->io_base + HPRE_INT_MASK); + + /* disable HPRE block master OOO when m-bit error occur */ + val = readl(qm->io_base + HPRE_AM_OOO_SHUTDOWN_ENB); + val &= ~HPRE_AM_OOO_SHUTDOWN_ENABLE; + writel(val, qm->io_base + HPRE_AM_OOO_SHUTDOWN_ENB); } static void hpre_hw_error_enable(struct hisi_qm *qm) { + u32 val; + + /* clear HPRE hw error source if having */ + writel(HPRE_CORE_INT_DISABLE, qm->io_base + HPRE_HAC_SOURCE_INT); + /* enable hpre hw error interrupts */ writel(HPRE_CORE_INT_ENABLE, qm->io_base + HPRE_INT_MASK); writel(HPRE_HAC_RAS_CE_ENABLE, qm->io_base + HPRE_RAS_CE_ENB); writel(HPRE_HAC_RAS_NFE_ENABLE, qm->io_base + HPRE_RAS_NFE_ENB); writel(HPRE_HAC_RAS_FE_ENABLE, qm->io_base + HPRE_RAS_FE_ENB); + + /* enable HPRE block master OOO when m-bit error occur */ + val = readl(qm->io_base + HPRE_AM_OOO_SHUTDOWN_ENB); + val |= HPRE_AM_OOO_SHUTDOWN_ENABLE; + writel(val, qm->io_base + HPRE_AM_OOO_SHUTDOWN_ENB); } static inline struct hisi_qm *hpre_file_to_qm(struct hpre_debugfs_file *file) @@ -354,9 +369,7 @@ static u32 hpre_current_qm_read(struct hpre_debugfs_file *file) static int hpre_current_qm_write(struct hpre_debugfs_file *file, u32 val) { struct hisi_qm *qm = hpre_file_to_qm(file); - struct hpre_debug *debug = file->debug; - struct hpre *hpre = container_of(debug, struct hpre, debug); - u32 num_vfs = hpre->num_vfs; + u32 num_vfs = qm->vfs_num; u32 vfq_num, tmp; @@ -523,6 +536,33 @@ static const struct file_operations hpre_ctrl_debug_fops = { .write = hpre_ctrl_debug_write, }; +static int hpre_debugfs_atomic64_get(void *data, u64 *val) +{ + struct hpre_dfx *dfx_item = data; + + *val = atomic64_read(&dfx_item->value); + + return 0; +} + +static int hpre_debugfs_atomic64_set(void *data, u64 val) +{ + struct hpre_dfx *dfx_item = data; + struct hpre_dfx *hpre_dfx = dfx_item - HPRE_OVERTIME_THRHLD; + + if (val) + return -EINVAL; + + if (dfx_item->type == HPRE_OVERTIME_THRHLD) + atomic64_set(&hpre_dfx[HPRE_OVER_THRHLD_CNT].value, 0); + atomic64_set(&dfx_item->value, val); + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(hpre_atomic64_ops, hpre_debugfs_atomic64_get, + hpre_debugfs_atomic64_set, "%llu\n"); + static int hpre_create_debugfs_file(struct hpre_debug *dbg, struct dentry *dir, enum hpre_ctrl_dbgfs_file type, int indx) { @@ -620,6 +660,22 @@ static int hpre_ctrl_debug_init(struct hpre_debug *debug) return hpre_cluster_debugfs_init(debug); } +static void hpre_dfx_debug_init(struct hpre_debug *debug) +{ + struct hpre *hpre = container_of(debug, struct hpre, debug); + struct hpre_dfx *dfx = hpre->debug.dfx; + struct hisi_qm *qm = &hpre->qm; + struct dentry *parent; + int i; + + parent = debugfs_create_dir("hpre_dfx", qm->debug.debug_root); + for (i = 0; i < HPRE_DFX_FILE_NUM; i++) { + dfx[i].type = i; + debugfs_create_file(hpre_dfx_files[i], 0644, parent, &dfx[i], + &hpre_atomic64_ops); + } +} + static int hpre_debugfs_init(struct hpre *hpre) { struct hisi_qm *qm = &hpre->qm; @@ -629,6 +685,8 @@ static int hpre_debugfs_init(struct hpre *hpre) dir = debugfs_create_dir(dev_name(dev), hpre_debugfs_root); qm->debug.debug_root = dir; + qm->debug.sqe_mask_offset = HPRE_SQE_MASK_OFFSET; + qm->debug.sqe_mask_len = HPRE_SQE_MASK_LEN; ret = hisi_qm_debug_init(qm); if (ret) @@ -640,6 +698,9 @@ static int hpre_debugfs_init(struct hpre *hpre) if (ret) goto failed_to_create; } + + hpre_dfx_debug_init(&hpre->debug); + return 0; failed_to_create: @@ -654,32 +715,27 @@ static void hpre_debugfs_exit(struct hpre *hpre) debugfs_remove_recursive(qm->debug.debug_root); } -static int hpre_qm_pre_init(struct hisi_qm *qm, struct pci_dev *pdev) +static int hpre_qm_init(struct hisi_qm *qm, struct pci_dev *pdev) { - enum qm_hw_ver rev_id; - - rev_id = hisi_qm_get_hw_version(pdev); - if (rev_id < 0) - return -ENODEV; - - if (rev_id == QM_HW_V1) { + if (pdev->revision == QM_HW_V1) { pci_warn(pdev, "HPRE version 1 is not supported!\n"); return -EINVAL; } qm->pdev = pdev; - qm->ver = rev_id; + qm->ver = pdev->revision; qm->sqe_size = HPRE_SQE_SIZE; qm->dev_name = hpre_name; + qm->fun_type = (pdev->device == HPRE_PCI_DEVICE_ID) ? - QM_HW_PF : QM_HW_VF; - if (pdev->is_physfn) { + QM_HW_PF : QM_HW_VF; + if (qm->fun_type == QM_HW_PF) { qm->qp_base = HPRE_PF_DEF_Q_BASE; - qm->qp_num = hpre_pf_q_num; + qm->qp_num = pf_q_num; + qm->qm_list = &hpre_devices; } - qm->use_dma_api = true; - return 0; + return hisi_qm_init(qm); } static void hpre_log_hw_error(struct hisi_qm *qm, u32 err_sts) @@ -693,8 +749,6 @@ static void hpre_log_hw_error(struct hisi_qm *qm, u32 err_sts) err->msg, err->int_msk); err++; } - - writel(err_sts, qm->io_base + HPRE_HAC_SOURCE_INT); } static u32 hpre_get_hw_err_status(struct hisi_qm *qm) @@ -702,16 +756,38 @@ static u32 hpre_get_hw_err_status(struct hisi_qm *qm) return readl(qm->io_base + HPRE_HAC_INT_STATUS); } +static void hpre_clear_hw_err_status(struct hisi_qm *qm, u32 err_sts) +{ + writel(err_sts, qm->io_base + HPRE_HAC_SOURCE_INT); +} + +static void hpre_open_axi_master_ooo(struct hisi_qm *qm) +{ + u32 value; + + value = readl(qm->io_base + HPRE_AM_OOO_SHUTDOWN_ENB); + writel(value & ~HPRE_AM_OOO_SHUTDOWN_ENABLE, + HPRE_ADDR(qm, HPRE_AM_OOO_SHUTDOWN_ENB)); + writel(value | HPRE_AM_OOO_SHUTDOWN_ENABLE, + HPRE_ADDR(qm, HPRE_AM_OOO_SHUTDOWN_ENB)); +} + static const struct hisi_qm_err_ini hpre_err_ini = { + .hw_init = hpre_set_user_domain_and_cache, .hw_err_enable = hpre_hw_error_enable, .hw_err_disable = hpre_hw_error_disable, .get_dev_hw_err_status = hpre_get_hw_err_status, + .clear_dev_hw_err_status = hpre_clear_hw_err_status, .log_dev_hw_err = hpre_log_hw_error, + .open_axi_master_ooo = hpre_open_axi_master_ooo, .err_info = { .ce = QM_BASE_CE, .nfe = QM_BASE_NFE | QM_ACC_DO_TASK_TIMEOUT, .fe = 0, - .msi = QM_DB_RANDOM_INVALID, + .ecc_2bits_mask = HPRE_CORE_ECC_2BIT_ERR | + HPRE_OOO_ECC_2BIT_ERR, + .msi_wr_port = HPRE_WR_MSI_PORT, + .acpi_rst = "HRST", } }; @@ -722,7 +798,7 @@ static int hpre_pf_probe_init(struct hpre *hpre) qm->ctrl_qp_num = HPRE_QUEUE_NUM_V2; - ret = hpre_set_user_domain_and_cache(hpre); + ret = hpre_set_user_domain_and_cache(qm); if (ret) return ret; @@ -732,6 +808,20 @@ static int hpre_pf_probe_init(struct hpre *hpre) return 0; } +static int hpre_probe_init(struct hpre *hpre) +{ + struct hisi_qm *qm = &hpre->qm; + int ret; + + if (qm->fun_type == QM_HW_PF) { + ret = hpre_pf_probe_init(hpre); + if (ret) + return ret; + } + + return 0; +} + static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct hisi_qm *qm; @@ -742,26 +832,17 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (!hpre) return -ENOMEM; - pci_set_drvdata(pdev, hpre); - qm = &hpre->qm; - ret = hpre_qm_pre_init(qm, pdev); - if (ret) - return ret; - - ret = hisi_qm_init(qm); - if (ret) + ret = hpre_qm_init(qm, pdev); + if (ret) { + pci_err(pdev, "Failed to init HPRE QM (%d)!\n", ret); return ret; + } - if (pdev->is_physfn) { - ret = hpre_pf_probe_init(hpre); - if (ret) - goto err_with_qm_init; - } else if (qm->fun_type == QM_HW_VF && qm->ver == QM_HW_V2) { - /* v2 starts to support get vft by mailbox */ - ret = hisi_qm_get_vft(qm, &qm->qp_base, &qm->qp_num); - if (ret) - goto err_with_qm_init; + ret = hpre_probe_init(hpre); + if (ret) { + pci_err(pdev, "Failed to probe (%d)!\n", ret); + goto err_with_qm_init; } ret = hisi_qm_start(qm); @@ -779,8 +860,18 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) pci_err(pdev, "fail to register algs to crypto!\n"); goto err_with_qm_start; } + + if (qm->fun_type == QM_HW_PF && vfs_num) { + ret = hisi_qm_sriov_enable(pdev, vfs_num); + if (ret < 0) + goto err_with_crypto_register; + } + return 0; +err_with_crypto_register: + hpre_algs_unregister(); + err_with_qm_start: hisi_qm_del_from_list(qm, &hpre_devices); hisi_qm_stop(qm); @@ -794,107 +885,6 @@ err_with_qm_init: return ret; } -static int hpre_vf_q_assign(struct hpre *hpre, int num_vfs) -{ - struct hisi_qm *qm = &hpre->qm; - u32 qp_num = qm->qp_num; - int q_num, remain_q_num, i; - u32 q_base = qp_num; - int ret; - - if (!num_vfs) - return -EINVAL; - - remain_q_num = qm->ctrl_qp_num - qp_num; - - /* If remaining queues are not enough, return error. */ - if (remain_q_num < num_vfs) - return -EINVAL; - - q_num = remain_q_num / num_vfs; - for (i = 1; i <= num_vfs; i++) { - if (i == num_vfs) - q_num += remain_q_num % num_vfs; - ret = hisi_qm_set_vft(qm, i, q_base, (u32)q_num); - if (ret) - return ret; - q_base += q_num; - } - - return 0; -} - -static int hpre_clear_vft_config(struct hpre *hpre) -{ - struct hisi_qm *qm = &hpre->qm; - u32 num_vfs = hpre->num_vfs; - int ret; - u32 i; - - for (i = 1; i <= num_vfs; i++) { - ret = hisi_qm_set_vft(qm, i, 0, 0); - if (ret) - return ret; - } - hpre->num_vfs = 0; - - return 0; -} - -static int hpre_sriov_enable(struct pci_dev *pdev, int max_vfs) -{ - struct hpre *hpre = pci_get_drvdata(pdev); - int pre_existing_vfs, num_vfs, ret; - - pre_existing_vfs = pci_num_vf(pdev); - if (pre_existing_vfs) { - pci_err(pdev, - "Can't enable VF. Please disable pre-enabled VFs!\n"); - return 0; - } - - num_vfs = min_t(int, max_vfs, HPRE_VF_NUM); - ret = hpre_vf_q_assign(hpre, num_vfs); - if (ret) { - pci_err(pdev, "Can't assign queues for VF!\n"); - return ret; - } - - hpre->num_vfs = num_vfs; - - ret = pci_enable_sriov(pdev, num_vfs); - if (ret) { - pci_err(pdev, "Can't enable VF!\n"); - hpre_clear_vft_config(hpre); - return ret; - } - - return num_vfs; -} - -static int hpre_sriov_disable(struct pci_dev *pdev) -{ - struct hpre *hpre = pci_get_drvdata(pdev); - - if (pci_vfs_assigned(pdev)) { - pci_err(pdev, "Failed to disable VFs while VFs are assigned!\n"); - return -EPERM; - } - - /* remove in hpre_pci_driver will be called to free VF resources */ - pci_disable_sriov(pdev); - - return hpre_clear_vft_config(hpre); -} - -static int hpre_sriov_configure(struct pci_dev *pdev, int num_vfs) -{ - if (num_vfs) - return hpre_sriov_enable(pdev, num_vfs); - else - return hpre_sriov_disable(pdev); -} - static void hpre_remove(struct pci_dev *pdev) { struct hpre *hpre = pci_get_drvdata(pdev); @@ -903,8 +893,8 @@ static void hpre_remove(struct pci_dev *pdev) hpre_algs_unregister(); hisi_qm_del_from_list(qm, &hpre_devices); - if (qm->fun_type == QM_HW_PF && hpre->num_vfs != 0) { - ret = hpre_sriov_disable(pdev); + if (qm->fun_type == QM_HW_PF && qm->vfs_num) { + ret = hisi_qm_sriov_disable(pdev); if (ret) { pci_err(pdev, "Disable SRIOV fail!\n"); return; @@ -924,6 +914,9 @@ static void hpre_remove(struct pci_dev *pdev) static const struct pci_error_handlers hpre_err_handler = { .error_detected = hisi_qm_dev_err_detected, + .slot_reset = hisi_qm_dev_slot_reset, + .reset_prepare = hisi_qm_reset_prepare, + .reset_done = hisi_qm_reset_done, }; static struct pci_driver hpre_pci_driver = { @@ -931,7 +924,7 @@ static struct pci_driver hpre_pci_driver = { .id_table = hpre_dev_ids, .probe = hpre_probe, .remove = hpre_remove, - .sriov_configure = hpre_sriov_configure, + .sriov_configure = hisi_qm_sriov_configure, .err_handler = &hpre_err_handler, }; diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index f795fb557630..9bb263cec6c3 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -1,9 +1,12 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright (c) 2019 HiSilicon Limited. */ #include <asm/page.h> +#include <linux/acpi.h> +#include <linux/aer.h> #include <linux/bitmap.h> #include <linux/debugfs.h> #include <linux/dma-mapping.h> +#include <linux/idr.h> #include <linux/io.h> #include <linux/irqreturn.h> #include <linux/log2.h> @@ -53,6 +56,7 @@ #define QM_SQ_TYPE_SHIFT 8 #define QM_SQ_TYPE_MASK GENMASK(3, 0) +#define QM_SQ_TAIL_IDX(sqc) ((le16_to_cpu((sqc)->w11) >> 6) & 0x1) /* cqc shift */ #define QM_CQ_HOP_NUM_SHIFT 0 @@ -64,6 +68,7 @@ #define QM_CQE_PHASE(cqe) (le16_to_cpu((cqe)->w7) & 0x1) #define QM_QC_CQE_SIZE 4 +#define QM_CQ_TAIL_IDX(cqc) ((le16_to_cpu((cqc)->w11) >> 6) & 0x1) /* eqc shift */ #define QM_EQE_AEQE_SIZE (2UL << 12) @@ -122,9 +127,11 @@ #define QM_DFX_CNT_CLR_CE 0x100118 #define QM_ABNORMAL_INT_SOURCE 0x100000 +#define QM_ABNORMAL_INT_SOURCE_CLR GENMASK(12, 0) #define QM_ABNORMAL_INT_MASK 0x100004 #define QM_ABNORMAL_INT_MASK_VALUE 0x1fff #define QM_ABNORMAL_INT_STATUS 0x100008 +#define QM_ABNORMAL_INT_SET 0x10000c #define QM_ABNORMAL_INF00 0x100010 #define QM_FIFO_OVERFLOW_TYPE 0xc0 #define QM_FIFO_OVERFLOW_TYPE_SHIFT 6 @@ -140,6 +147,27 @@ #define QM_RAS_CE_TIMES_PER_IRQ 1 #define QM_RAS_MSI_INT_SEL 0x1040f4 +#define QM_DEV_RESET_FLAG 0 +#define QM_RESET_WAIT_TIMEOUT 400 +#define QM_PEH_VENDOR_ID 0x1000d8 +#define ACC_VENDOR_ID_VALUE 0x5a5a +#define QM_PEH_DFX_INFO0 0x1000fc +#define ACC_PEH_SRIOV_CTRL_VF_MSE_SHIFT 3 +#define ACC_PEH_MSI_DISABLE GENMASK(31, 0) +#define ACC_MASTER_GLOBAL_CTRL_SHUTDOWN 0x1 +#define ACC_MASTER_TRANS_RETURN_RW 3 +#define ACC_MASTER_TRANS_RETURN 0x300150 +#define ACC_MASTER_GLOBAL_CTRL 0x300000 +#define ACC_AM_CFG_PORT_WR_EN 0x30001c +#define QM_RAS_NFE_MBIT_DISABLE ~QM_ECC_MBIT +#define ACC_AM_ROB_ECC_INT_STS 0x300104 +#define ACC_ROB_ECC_ERR_MULTPL BIT(1) + +#define POLL_PERIOD 10 +#define POLL_TIMEOUT 1000 +#define WAIT_PERIOD_US_MAX 200 +#define WAIT_PERIOD_US_MIN 100 +#define MAX_WAIT_COUNTS 1000 #define QM_CACHE_WB_START 0x204 #define QM_CACHE_WB_DONE 0x208 @@ -147,7 +175,12 @@ #define QM_SQE_DATA_ALIGN_MASK GENMASK(6, 0) #define QMC_ALIGN(sz) ALIGN(sz, 32) +#define QM_DBG_READ_LEN 256 +#define QM_DBG_WRITE_LEN 1024 #define QM_DBG_TMP_BUF_LEN 22 +#define QM_PCI_COMMAND_INVALID ~0 + +#define QM_SQE_ADDR_MASK GENMASK(7, 0) #define QM_MK_CQC_DW3_V1(hop_num, pg_sz, buf_sz, cqe_sz) \ (((hop_num) << QM_CQ_HOP_NUM_SHIFT) | \ @@ -190,6 +223,12 @@ enum vft_type { CQC_VFT, }; +enum acc_err_result { + ACC_ERR_NONE, + ACC_ERR_NEED_RESET, + ACC_ERR_RECOVERED, +}; + struct qm_cqe { __le32 rsvd0; __le16 cmd_id; @@ -284,10 +323,22 @@ struct hisi_qm_hw_ops { u8 cmd, u16 index, u8 priority); u32 (*get_irq_num)(struct hisi_qm *qm); int (*debug_init)(struct hisi_qm *qm); - void (*hw_error_init)(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe, - u32 msi); + void (*hw_error_init)(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe); void (*hw_error_uninit)(struct hisi_qm *qm); - pci_ers_result_t (*hw_error_handle)(struct hisi_qm *qm); + enum acc_err_result (*hw_error_handle)(struct hisi_qm *qm); +}; + +struct qm_dfx_item { + const char *name; + u32 offset; +}; + +static struct qm_dfx_item qm_dfx_files[] = { + {"err_irq", offsetof(struct qm_dfx, err_irq_cnt)}, + {"aeq_irq", offsetof(struct qm_dfx, aeq_irq_cnt)}, + {"abnormal_irq", offsetof(struct qm_dfx, abnormal_irq_cnt)}, + {"create_qp_err", offsetof(struct qm_dfx, create_qp_err_cnt)}, + {"mb_err", offsetof(struct qm_dfx, mb_err_cnt)}, }; static const char * const qm_debug_file_name[] = { @@ -325,6 +376,93 @@ static const char * const qm_fifo_overflow[] = { "cq", "eq", "aeq", }; +static const char * const qm_s[] = { + "init", "start", "close", "stop", +}; + +static const char * const qp_s[] = { + "none", "init", "start", "stop", "close", +}; + +static bool qm_avail_state(struct hisi_qm *qm, enum qm_state new) +{ + enum qm_state curr = atomic_read(&qm->status.flags); + bool avail = false; + + switch (curr) { + case QM_INIT: + if (new == QM_START || new == QM_CLOSE) + avail = true; + break; + case QM_START: + if (new == QM_STOP) + avail = true; + break; + case QM_STOP: + if (new == QM_CLOSE || new == QM_START) + avail = true; + break; + default: + break; + } + + dev_dbg(&qm->pdev->dev, "change qm state from %s to %s\n", + qm_s[curr], qm_s[new]); + + if (!avail) + dev_warn(&qm->pdev->dev, "Can not change qm state from %s to %s\n", + qm_s[curr], qm_s[new]); + + return avail; +} + +static bool qm_qp_avail_state(struct hisi_qm *qm, struct hisi_qp *qp, + enum qp_state new) +{ + enum qm_state qm_curr = atomic_read(&qm->status.flags); + enum qp_state qp_curr = 0; + bool avail = false; + + if (qp) + qp_curr = atomic_read(&qp->qp_status.flags); + + switch (new) { + case QP_INIT: + if (qm_curr == QM_START || qm_curr == QM_INIT) + avail = true; + break; + case QP_START: + if ((qm_curr == QM_START && qp_curr == QP_INIT) || + (qm_curr == QM_START && qp_curr == QP_STOP)) + avail = true; + break; + case QP_STOP: + if ((qm_curr == QM_START && qp_curr == QP_START) || + (qp_curr == QP_INIT)) + avail = true; + break; + case QP_CLOSE: + if ((qm_curr == QM_START && qp_curr == QP_INIT) || + (qm_curr == QM_START && qp_curr == QP_STOP) || + (qm_curr == QM_STOP && qp_curr == QP_STOP) || + (qm_curr == QM_STOP && qp_curr == QP_INIT)) + avail = true; + break; + default: + break; + } + + dev_dbg(&qm->pdev->dev, "change qp state from %s to %s in QM %s\n", + qp_s[qp_curr], qp_s[new], qm_s[qm_curr]); + + if (!avail) + dev_warn(&qm->pdev->dev, + "Can not change qp state from %s to %s in QM %s\n", + qp_s[qp_curr], qp_s[new], qm_s[qm_curr]); + + return avail; +} + /* return 0 mailbox ready, -ETIMEDOUT hardware timeout */ static int qm_wait_mb_ready(struct hisi_qm *qm) { @@ -393,6 +531,8 @@ static int qm_mb(struct hisi_qm *qm, u8 cmd, dma_addr_t dma_addr, u16 queue, busy_unlock: mutex_unlock(&qm->mailbox_lock); + if (ret) + atomic64_inc(&qm->debug.dfx.mb_err_cnt); return ret; } @@ -460,7 +600,7 @@ static struct hisi_qp *qm_to_hisi_qp(struct hisi_qm *qm, struct qm_eqe *eqe) { u16 cqn = le32_to_cpu(eqe->dw0) & QM_EQE_CQN_MASK; - return qm->qp_array[cqn]; + return &qm->qp_array[cqn]; } static void qm_cq_head_update(struct hisi_qp *qp) @@ -510,8 +650,7 @@ static void qm_work_process(struct work_struct *work) while (QM_EQE_PHASE(eqe) == qm->status.eqc_phase) { eqe_num++; qp = qm_to_hisi_qp(qm, eqe); - if (qp) - qm_poll_qp(qp, qm); + qm_poll_qp(qp, qm); if (qm->status.eq_head == QM_Q_DEPTH - 1) { qm->status.eqc_phase = !qm->status.eqc_phase; @@ -551,6 +690,7 @@ static irqreturn_t qm_irq(int irq, void *data) if (readl(qm->io_base + QM_VF_EQ_INT_SOURCE)) return do_qm_irq(irq, data); + atomic64_inc(&qm->debug.dfx.err_irq_cnt); dev_err(&qm->pdev->dev, "invalid int source\n"); qm_db(qm, 0, QM_DOORBELL_CMD_EQ, qm->status.eq_head, 0); @@ -563,6 +703,7 @@ static irqreturn_t qm_aeq_irq(int irq, void *data) struct qm_aeqe *aeqe = qm->aeqe + qm->status.aeq_head; u32 type; + atomic64_inc(&qm->debug.dfx.aeq_irq_cnt); if (!readl(qm->io_base + QM_VF_AEQ_INT_SOURCE)) return IRQ_NONE; @@ -590,79 +731,20 @@ static irqreturn_t qm_aeq_irq(int irq, void *data) return IRQ_HANDLED; } -static irqreturn_t qm_abnormal_irq(int irq, void *data) -{ - const struct hisi_qm_hw_error *err = qm_hw_error; - struct hisi_qm *qm = data; - struct device *dev = &qm->pdev->dev; - u32 error_status, tmp; - - /* read err sts */ - tmp = readl(qm->io_base + QM_ABNORMAL_INT_STATUS); - error_status = qm->msi_mask & tmp; - - while (err->msg) { - if (err->int_msk & error_status) - dev_err(dev, "%s [error status=0x%x] found\n", - err->msg, err->int_msk); - - err++; - } - - /* clear err sts */ - writel(error_status, qm->io_base + QM_ABNORMAL_INT_SOURCE); - - return IRQ_HANDLED; -} - -static int qm_irq_register(struct hisi_qm *qm) -{ - struct pci_dev *pdev = qm->pdev; - int ret; - - ret = request_irq(pci_irq_vector(pdev, QM_EQ_EVENT_IRQ_VECTOR), - qm_irq, IRQF_SHARED, qm->dev_name, qm); - if (ret) - return ret; - - if (qm->ver == QM_HW_V2) { - ret = request_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), - qm_aeq_irq, IRQF_SHARED, qm->dev_name, qm); - if (ret) - goto err_aeq_irq; - - if (qm->fun_type == QM_HW_PF) { - ret = request_irq(pci_irq_vector(pdev, - QM_ABNORMAL_EVENT_IRQ_VECTOR), - qm_abnormal_irq, IRQF_SHARED, - qm->dev_name, qm); - if (ret) - goto err_abonormal_irq; - } - } - - return 0; - -err_abonormal_irq: - free_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), qm); -err_aeq_irq: - free_irq(pci_irq_vector(pdev, QM_EQ_EVENT_IRQ_VECTOR), qm); - return ret; -} - static void qm_irq_unregister(struct hisi_qm *qm) { struct pci_dev *pdev = qm->pdev; free_irq(pci_irq_vector(pdev, QM_EQ_EVENT_IRQ_VECTOR), qm); - if (qm->ver == QM_HW_V2) { - free_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), qm); + if (qm->ver == QM_HW_V1) + return; + + free_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), qm); - if (qm->fun_type == QM_HW_PF) - free_irq(pci_irq_vector(pdev, - QM_ABNORMAL_EVENT_IRQ_VECTOR), qm); - } + if (qm->fun_type == QM_HW_PF) + free_irq(pci_irq_vector(pdev, + QM_ABNORMAL_EVENT_IRQ_VECTOR), qm); } static void qm_init_qp_status(struct hisi_qp *qp) @@ -672,7 +754,7 @@ static void qm_init_qp_status(struct hisi_qp *qp) qp_status->sq_tail = 0; qp_status->cq_head = 0; qp_status->cqc_phase = true; - qp_status->flags = 0; + atomic_set(&qp_status->flags, 0); } static void qm_vft_data_cfg(struct hisi_qm *qm, enum vft_type type, u32 base, @@ -683,36 +765,26 @@ static void qm_vft_data_cfg(struct hisi_qm *qm, enum vft_type type, u32 base, if (number > 0) { switch (type) { case SQC_VFT: - switch (qm->ver) { - case QM_HW_V1: + if (qm->ver == QM_HW_V1) { tmp = QM_SQC_VFT_BUF_SIZE | QM_SQC_VFT_SQC_SIZE | QM_SQC_VFT_INDEX_NUMBER | QM_SQC_VFT_VALID | (u64)base << QM_SQC_VFT_START_SQN_SHIFT; - break; - case QM_HW_V2: + } else { tmp = (u64)base << QM_SQC_VFT_START_SQN_SHIFT | QM_SQC_VFT_VALID | (u64)(number - 1) << QM_SQC_VFT_SQN_SHIFT; - break; - case QM_HW_UNKNOWN: - break; } break; case CQC_VFT: - switch (qm->ver) { - case QM_HW_V1: + if (qm->ver == QM_HW_V1) { tmp = QM_CQC_VFT_BUF_SIZE | QM_CQC_VFT_SQC_SIZE | QM_CQC_VFT_INDEX_NUMBER | QM_CQC_VFT_VALID; - break; - case QM_HW_V2: + } else { tmp = QM_CQC_VFT_VALID; - break; - case QM_HW_UNKNOWN: - break; } break; } @@ -986,6 +1058,473 @@ static const struct file_operations qm_regs_fops = { .release = single_release, }; +static ssize_t qm_cmd_read(struct file *filp, char __user *buffer, + size_t count, loff_t *pos) +{ + char buf[QM_DBG_READ_LEN]; + int len; + + if (*pos) + return 0; + + if (count < QM_DBG_READ_LEN) + return -ENOSPC; + + len = snprintf(buf, QM_DBG_READ_LEN, "%s\n", + "Please echo help to cmd to get help information"); + + if (copy_to_user(buffer, buf, len)) + return -EFAULT; + + return (*pos = len); +} + +static void *qm_ctx_alloc(struct hisi_qm *qm, size_t ctx_size, + dma_addr_t *dma_addr) +{ + struct device *dev = &qm->pdev->dev; + void *ctx_addr; + + ctx_addr = kzalloc(ctx_size, GFP_KERNEL); + if (!ctx_addr) + return ERR_PTR(-ENOMEM); + + *dma_addr = dma_map_single(dev, ctx_addr, ctx_size, DMA_FROM_DEVICE); + if (dma_mapping_error(dev, *dma_addr)) { + dev_err(dev, "DMA mapping error!\n"); + kfree(ctx_addr); + return ERR_PTR(-ENOMEM); + } + + return ctx_addr; +} + +static void qm_ctx_free(struct hisi_qm *qm, size_t ctx_size, + const void *ctx_addr, dma_addr_t *dma_addr) +{ + struct device *dev = &qm->pdev->dev; + + dma_unmap_single(dev, *dma_addr, ctx_size, DMA_FROM_DEVICE); + kfree(ctx_addr); +} + +static int dump_show(struct hisi_qm *qm, void *info, + unsigned int info_size, char *info_name) +{ + struct device *dev = &qm->pdev->dev; + u8 *info_buf, *info_curr = info; + u32 i; +#define BYTE_PER_DW 4 + + info_buf = kzalloc(info_size, GFP_KERNEL); + if (!info_buf) + return -ENOMEM; + + for (i = 0; i < info_size; i++, info_curr++) { + if (i % BYTE_PER_DW == 0) + info_buf[i + 3UL] = *info_curr; + else if (i % BYTE_PER_DW == 1) + info_buf[i + 1UL] = *info_curr; + else if (i % BYTE_PER_DW == 2) + info_buf[i - 1] = *info_curr; + else if (i % BYTE_PER_DW == 3) + info_buf[i - 3] = *info_curr; + } + + dev_info(dev, "%s DUMP\n", info_name); + for (i = 0; i < info_size; i += BYTE_PER_DW) { + pr_info("DW%d: %02X%02X %02X%02X\n", i / BYTE_PER_DW, + info_buf[i], info_buf[i + 1UL], + info_buf[i + 2UL], info_buf[i + 3UL]); + } + + kfree(info_buf); + + return 0; +} + +static int qm_dump_sqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) +{ + return qm_mb(qm, QM_MB_CMD_SQC, dma_addr, qp_id, 1); +} + +static int qm_dump_cqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) +{ + return qm_mb(qm, QM_MB_CMD_CQC, dma_addr, qp_id, 1); +} + +static int qm_sqc_dump(struct hisi_qm *qm, const char *s) +{ + struct device *dev = &qm->pdev->dev; + struct qm_sqc *sqc, *sqc_curr; + dma_addr_t sqc_dma; + u32 qp_id; + int ret; + + if (!s) + return -EINVAL; + + ret = kstrtou32(s, 0, &qp_id); + if (ret || qp_id >= qm->qp_num) { + dev_err(dev, "Please input qp num (0-%d)", qm->qp_num - 1); + return -EINVAL; + } + + sqc = qm_ctx_alloc(qm, sizeof(*sqc), &sqc_dma); + if (IS_ERR(sqc)) + return PTR_ERR(sqc); + + ret = qm_dump_sqc_raw(qm, sqc_dma, qp_id); + if (ret) { + down_read(&qm->qps_lock); + if (qm->sqc) { + sqc_curr = qm->sqc + qp_id; + + ret = dump_show(qm, sqc_curr, sizeof(*sqc), + "SOFT SQC"); + if (ret) + dev_info(dev, "Show soft sqc failed!\n"); + } + up_read(&qm->qps_lock); + + goto err_free_ctx; + } + + ret = dump_show(qm, sqc, sizeof(*sqc), "SQC"); + if (ret) + dev_info(dev, "Show hw sqc failed!\n"); + +err_free_ctx: + qm_ctx_free(qm, sizeof(*sqc), sqc, &sqc_dma); + return ret; +} + +static int qm_cqc_dump(struct hisi_qm *qm, const char *s) +{ + struct device *dev = &qm->pdev->dev; + struct qm_cqc *cqc, *cqc_curr; + dma_addr_t cqc_dma; + u32 qp_id; + int ret; + + if (!s) + return -EINVAL; + + ret = kstrtou32(s, 0, &qp_id); + if (ret || qp_id >= qm->qp_num) { + dev_err(dev, "Please input qp num (0-%d)", qm->qp_num - 1); + return -EINVAL; + } + + cqc = qm_ctx_alloc(qm, sizeof(*cqc), &cqc_dma); + if (IS_ERR(cqc)) + return PTR_ERR(cqc); + + ret = qm_dump_cqc_raw(qm, cqc_dma, qp_id); + if (ret) { + down_read(&qm->qps_lock); + if (qm->cqc) { + cqc_curr = qm->cqc + qp_id; + + ret = dump_show(qm, cqc_curr, sizeof(*cqc), + "SOFT CQC"); + if (ret) + dev_info(dev, "Show soft cqc failed!\n"); + } + up_read(&qm->qps_lock); + + goto err_free_ctx; + } + + ret = dump_show(qm, cqc, sizeof(*cqc), "CQC"); + if (ret) + dev_info(dev, "Show hw cqc failed!\n"); + +err_free_ctx: + qm_ctx_free(qm, sizeof(*cqc), cqc, &cqc_dma); + return ret; +} + +static int qm_eqc_aeqc_dump(struct hisi_qm *qm, char *s, size_t size, + int cmd, char *name) +{ + struct device *dev = &qm->pdev->dev; + dma_addr_t xeqc_dma; + void *xeqc; + int ret; + + if (strsep(&s, " ")) { + dev_err(dev, "Please do not input extra characters!\n"); + return -EINVAL; + } + + xeqc = qm_ctx_alloc(qm, size, &xeqc_dma); + if (IS_ERR(xeqc)) + return PTR_ERR(xeqc); + + ret = qm_mb(qm, cmd, xeqc_dma, 0, 1); + if (ret) + goto err_free_ctx; + + ret = dump_show(qm, xeqc, size, name); + if (ret) + dev_info(dev, "Show hw %s failed!\n", name); + +err_free_ctx: + qm_ctx_free(qm, size, xeqc, &xeqc_dma); + return ret; +} + +static int q_dump_param_parse(struct hisi_qm *qm, char *s, + u32 *e_id, u32 *q_id) +{ + struct device *dev = &qm->pdev->dev; + unsigned int qp_num = qm->qp_num; + char *presult; + int ret; + + presult = strsep(&s, " "); + if (!presult) { + dev_err(dev, "Please input qp number!\n"); + return -EINVAL; + } + + ret = kstrtou32(presult, 0, q_id); + if (ret || *q_id >= qp_num) { + dev_err(dev, "Please input qp num (0-%d)", qp_num - 1); + return -EINVAL; + } + + presult = strsep(&s, " "); + if (!presult) { + dev_err(dev, "Please input sqe number!\n"); + return -EINVAL; + } + + ret = kstrtou32(presult, 0, e_id); + if (ret || *e_id >= QM_Q_DEPTH) { + dev_err(dev, "Please input sqe num (0-%d)", QM_Q_DEPTH - 1); + return -EINVAL; + } + + if (strsep(&s, " ")) { + dev_err(dev, "Please do not input extra characters!\n"); + return -EINVAL; + } + + return 0; +} + +static int qm_sq_dump(struct hisi_qm *qm, char *s) +{ + struct device *dev = &qm->pdev->dev; + void *sqe, *sqe_curr; + struct hisi_qp *qp; + u32 qp_id, sqe_id; + int ret; + + ret = q_dump_param_parse(qm, s, &sqe_id, &qp_id); + if (ret) + return ret; + + sqe = kzalloc(qm->sqe_size * QM_Q_DEPTH, GFP_KERNEL); + if (!sqe) + return -ENOMEM; + + qp = &qm->qp_array[qp_id]; + memcpy(sqe, qp->sqe, qm->sqe_size * QM_Q_DEPTH); + sqe_curr = sqe + (u32)(sqe_id * qm->sqe_size); + memset(sqe_curr + qm->debug.sqe_mask_offset, QM_SQE_ADDR_MASK, + qm->debug.sqe_mask_len); + + ret = dump_show(qm, sqe_curr, qm->sqe_size, "SQE"); + if (ret) + dev_info(dev, "Show sqe failed!\n"); + + kfree(sqe); + + return ret; +} + +static int qm_cq_dump(struct hisi_qm *qm, char *s) +{ + struct device *dev = &qm->pdev->dev; + struct qm_cqe *cqe_curr; + struct hisi_qp *qp; + u32 qp_id, cqe_id; + int ret; + + ret = q_dump_param_parse(qm, s, &cqe_id, &qp_id); + if (ret) + return ret; + + qp = &qm->qp_array[qp_id]; + cqe_curr = qp->cqe + cqe_id; + ret = dump_show(qm, cqe_curr, sizeof(struct qm_cqe), "CQE"); + if (ret) + dev_info(dev, "Show cqe failed!\n"); + + return ret; +} + +static int qm_eq_aeq_dump(struct hisi_qm *qm, const char *s, + size_t size, char *name) +{ + struct device *dev = &qm->pdev->dev; + void *xeqe; + u32 xeqe_id; + int ret; + + if (!s) + return -EINVAL; + + ret = kstrtou32(s, 0, &xeqe_id); + if (ret || xeqe_id >= QM_Q_DEPTH) { + dev_err(dev, "Please input aeqe num (0-%d)", QM_Q_DEPTH - 1); + return -EINVAL; + } + + down_read(&qm->qps_lock); + + if (qm->eqe && !strcmp(name, "EQE")) { + xeqe = qm->eqe + xeqe_id; + } else if (qm->aeqe && !strcmp(name, "AEQE")) { + xeqe = qm->aeqe + xeqe_id; + } else { + ret = -EINVAL; + goto err_unlock; + } + + ret = dump_show(qm, xeqe, size, name); + if (ret) + dev_info(dev, "Show %s failed!\n", name); + +err_unlock: + up_read(&qm->qps_lock); + return ret; +} + +static int qm_dbg_help(struct hisi_qm *qm, char *s) +{ + struct device *dev = &qm->pdev->dev; + + if (strsep(&s, " ")) { + dev_err(dev, "Please do not input extra characters!\n"); + return -EINVAL; + } + + dev_info(dev, "available commands:\n"); + dev_info(dev, "sqc <num>\n"); + dev_info(dev, "cqc <num>\n"); + dev_info(dev, "eqc\n"); + dev_info(dev, "aeqc\n"); + dev_info(dev, "sq <num> <e>\n"); + dev_info(dev, "cq <num> <e>\n"); + dev_info(dev, "eq <e>\n"); + dev_info(dev, "aeq <e>\n"); + + return 0; +} + +static int qm_cmd_write_dump(struct hisi_qm *qm, const char *cmd_buf) +{ + struct device *dev = &qm->pdev->dev; + char *presult, *s; + int ret; + + s = kstrdup(cmd_buf, GFP_KERNEL); + if (!s) + return -ENOMEM; + + presult = strsep(&s, " "); + if (!presult) { + kfree(s); + return -EINVAL; + } + + if (!strcmp(presult, "sqc")) + ret = qm_sqc_dump(qm, s); + else if (!strcmp(presult, "cqc")) + ret = qm_cqc_dump(qm, s); + else if (!strcmp(presult, "eqc")) + ret = qm_eqc_aeqc_dump(qm, s, sizeof(struct qm_eqc), + QM_MB_CMD_EQC, "EQC"); + else if (!strcmp(presult, "aeqc")) + ret = qm_eqc_aeqc_dump(qm, s, sizeof(struct qm_aeqc), + QM_MB_CMD_AEQC, "AEQC"); + else if (!strcmp(presult, "sq")) + ret = qm_sq_dump(qm, s); + else if (!strcmp(presult, "cq")) + ret = qm_cq_dump(qm, s); + else if (!strcmp(presult, "eq")) + ret = qm_eq_aeq_dump(qm, s, sizeof(struct qm_eqe), "EQE"); + else if (!strcmp(presult, "aeq")) + ret = qm_eq_aeq_dump(qm, s, sizeof(struct qm_aeqe), "AEQE"); + else if (!strcmp(presult, "help")) + ret = qm_dbg_help(qm, s); + else + ret = -EINVAL; + + if (ret) + dev_info(dev, "Please echo help\n"); + + kfree(s); + + return ret; +} + +static ssize_t qm_cmd_write(struct file *filp, const char __user *buffer, + size_t count, loff_t *pos) +{ + struct hisi_qm *qm = filp->private_data; + char *cmd_buf, *cmd_buf_tmp; + int ret; + + if (*pos) + return 0; + + /* Judge if the instance is being reset. */ + if (unlikely(atomic_read(&qm->status.flags) == QM_STOP)) + return 0; + + if (count > QM_DBG_WRITE_LEN) + return -ENOSPC; + + cmd_buf = kzalloc(count + 1, GFP_KERNEL); + if (!cmd_buf) + return -ENOMEM; + + if (copy_from_user(cmd_buf, buffer, count)) { + kfree(cmd_buf); + return -EFAULT; + } + + cmd_buf[count] = '\0'; + + cmd_buf_tmp = strchr(cmd_buf, '\n'); + if (cmd_buf_tmp) { + *cmd_buf_tmp = '\0'; + count = cmd_buf_tmp - cmd_buf + 1; + } + + ret = qm_cmd_write_dump(qm, cmd_buf); + if (ret) { + kfree(cmd_buf); + return ret; + } + + kfree(cmd_buf); + + return count; +} + +static const struct file_operations qm_cmd_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .read = qm_cmd_read, + .write = qm_cmd_write, +}; + static int qm_create_debugfs_file(struct hisi_qm *qm, enum qm_debug_file index) { struct dentry *qm_d = qm->debug.qm_d; @@ -1001,20 +1540,21 @@ static int qm_create_debugfs_file(struct hisi_qm *qm, enum qm_debug_file index) return 0; } -static void qm_hw_error_init_v1(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe, - u32 msi) +static void qm_hw_error_init_v1(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe) { writel(QM_ABNORMAL_INT_MASK_VALUE, qm->io_base + QM_ABNORMAL_INT_MASK); } -static void qm_hw_error_init_v2(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe, - u32 msi) +static void qm_hw_error_init_v2(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe) { - u32 irq_enable = ce | nfe | fe | msi; + u32 irq_enable = ce | nfe | fe; u32 irq_unmask = ~irq_enable; qm->error_mask = ce | nfe | fe; - qm->msi_mask = msi; + + /* clear QM hw residual error source */ + writel(QM_ABNORMAL_INT_SOURCE_CLR, + qm->io_base + QM_ABNORMAL_INT_SOURCE); /* configure error type */ writel(ce, qm->io_base + QM_RAS_CE_ENABLE); @@ -1022,9 +1562,6 @@ static void qm_hw_error_init_v2(struct hisi_qm *qm, u32 ce, u32 nfe, u32 fe, writel(nfe, qm->io_base + QM_RAS_NFE_ENABLE); writel(fe, qm->io_base + QM_RAS_FE_ENABLE); - /* use RAS irq default, so only set QM_RAS_MSI_INT_SEL for MSI */ - writel(msi, qm->io_base + QM_RAS_MSI_INT_SEL); - irq_unmask &= readl(qm->io_base + QM_ABNORMAL_INT_MASK); writel(irq_unmask, qm->io_base + QM_ABNORMAL_INT_MASK); } @@ -1071,7 +1608,7 @@ static void qm_log_hw_error(struct hisi_qm *qm, u32 error_status) } } -static pci_ers_result_t qm_hw_error_handle_v2(struct hisi_qm *qm) +static enum acc_err_result qm_hw_error_handle_v2(struct hisi_qm *qm) { u32 error_status, tmp; @@ -1080,15 +1617,20 @@ static pci_ers_result_t qm_hw_error_handle_v2(struct hisi_qm *qm) error_status = qm->error_mask & tmp; if (error_status) { - qm_log_hw_error(qm, error_status); + if (error_status & QM_ECC_MBIT) + qm->err_status.is_qm_ecc_mbit = true; - /* clear err sts */ - writel(error_status, qm->io_base + QM_ABNORMAL_INT_SOURCE); + qm_log_hw_error(qm, error_status); + if (error_status == QM_DB_RANDOM_INVALID) { + writel(error_status, qm->io_base + + QM_ABNORMAL_INT_SOURCE); + return ACC_ERR_RECOVERED; + } - return PCI_ERS_RESULT_NEED_RESET; + return ACC_ERR_NEED_RESET; } - return PCI_ERS_RESULT_RECOVERED; + return ACC_ERR_RECOVERED; } static const struct hisi_qm_hw_ops qm_hw_ops_v1 = { @@ -1117,68 +1659,61 @@ static void *qm_get_avail_sqe(struct hisi_qp *qp) return qp->sqe + sq_tail * qp->qm->sqe_size; } -/** - * hisi_qm_create_qp() - Create a queue pair from qm. - * @qm: The qm we create a qp from. - * @alg_type: Accelerator specific algorithm type in sqc. - * - * return created qp, -EBUSY if all qps in qm allocated, -ENOMEM if allocating - * qp memory fails. - */ -struct hisi_qp *hisi_qm_create_qp(struct hisi_qm *qm, u8 alg_type) +static struct hisi_qp *qm_create_qp_nolock(struct hisi_qm *qm, u8 alg_type) { struct device *dev = &qm->pdev->dev; struct hisi_qp *qp; - int qp_id, ret; - - qp = kzalloc(sizeof(*qp), GFP_KERNEL); - if (!qp) - return ERR_PTR(-ENOMEM); + int qp_id; - write_lock(&qm->qps_lock); + if (!qm_qp_avail_state(qm, NULL, QP_INIT)) + return ERR_PTR(-EPERM); - qp_id = find_first_zero_bit(qm->qp_bitmap, qm->qp_num); - if (qp_id >= qm->qp_num) { - write_unlock(&qm->qps_lock); - dev_info(&qm->pdev->dev, "QM all queues are busy!\n"); - ret = -EBUSY; - goto err_free_qp; + if (qm->qp_in_used == qm->qp_num) { + dev_info_ratelimited(dev, "All %u queues of QM are busy!\n", + qm->qp_num); + atomic64_inc(&qm->debug.dfx.create_qp_err_cnt); + return ERR_PTR(-EBUSY); } - set_bit(qp_id, qm->qp_bitmap); - qm->qp_array[qp_id] = qp; - qm->qp_in_used++; - write_unlock(&qm->qps_lock); - - qp->qm = qm; + qp_id = idr_alloc_cyclic(&qm->qp_idr, NULL, 0, qm->qp_num, GFP_ATOMIC); + if (qp_id < 0) { + dev_info_ratelimited(dev, "All %u queues of QM are busy!\n", + qm->qp_num); + atomic64_inc(&qm->debug.dfx.create_qp_err_cnt); + return ERR_PTR(-EBUSY); + } - if (qm->use_dma_api) { - qp->qdma.size = qm->sqe_size * QM_Q_DEPTH + - sizeof(struct qm_cqe) * QM_Q_DEPTH; - qp->qdma.va = dma_alloc_coherent(dev, qp->qdma.size, - &qp->qdma.dma, GFP_KERNEL); - if (!qp->qdma.va) { - ret = -ENOMEM; - goto err_clear_bit; - } + qp = &qm->qp_array[qp_id]; - dev_dbg(dev, "allocate qp dma buf(va=%pK, dma=%pad, size=%zx)\n", - qp->qdma.va, &qp->qdma.dma, qp->qdma.size); - } + memset(qp->cqe, 0, sizeof(struct qm_cqe) * QM_Q_DEPTH); + qp->event_cb = NULL; + qp->req_cb = NULL; qp->qp_id = qp_id; qp->alg_type = alg_type; + qm->qp_in_used++; + atomic_set(&qp->qp_status.flags, QP_INIT); return qp; +} + +/** + * hisi_qm_create_qp() - Create a queue pair from qm. + * @qm: The qm we create a qp from. + * @alg_type: Accelerator specific algorithm type in sqc. + * + * return created qp, -EBUSY if all qps in qm allocated, -ENOMEM if allocating + * qp memory fails. + */ +struct hisi_qp *hisi_qm_create_qp(struct hisi_qm *qm, u8 alg_type) +{ + struct hisi_qp *qp; -err_clear_bit: - write_lock(&qm->qps_lock); - qm->qp_array[qp_id] = NULL; - clear_bit(qp_id, qm->qp_bitmap); - write_unlock(&qm->qps_lock); -err_free_qp: - kfree(qp); - return ERR_PTR(ret); + down_write(&qm->qps_lock); + qp = qm_create_qp_nolock(qm, alg_type); + up_write(&qm->qps_lock); + + return qp; } EXPORT_SYMBOL_GPL(hisi_qm_create_qp); @@ -1191,19 +1726,18 @@ EXPORT_SYMBOL_GPL(hisi_qm_create_qp); void hisi_qm_release_qp(struct hisi_qp *qp) { struct hisi_qm *qm = qp->qm; - struct qm_dma *qdma = &qp->qdma; - struct device *dev = &qm->pdev->dev; - if (qm->use_dma_api && qdma->va) - dma_free_coherent(dev, qdma->size, qdma->va, qdma->dma); + down_write(&qm->qps_lock); + + if (!qm_qp_avail_state(qm, qp, QP_CLOSE)) { + up_write(&qm->qps_lock); + return; + } - write_lock(&qm->qps_lock); - qm->qp_array[qp->qp_id] = NULL; - clear_bit(qp->qp_id, qm->qp_bitmap); qm->qp_in_used--; - write_unlock(&qm->qps_lock); + idr_remove(&qm->qp_idr, qp->qp_id); - kfree(qp); + up_write(&qm->qps_lock); } EXPORT_SYMBOL_GPL(hisi_qm_release_qp); @@ -1234,7 +1768,7 @@ static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) if (ver == QM_HW_V1) { sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V1(0, 0, 0, qm->sqe_size)); sqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); - } else if (ver == QM_HW_V2) { + } else { sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V2(qm->sqe_size)); sqc->w8 = 0; /* rand_qc */ } @@ -1261,7 +1795,7 @@ static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) if (ver == QM_HW_V1) { cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V1(0, 0, 0, 4)); cqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); - } else if (ver == QM_HW_V2) { + } else { cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V2(4)); cqc->w8 = 0; } @@ -1274,6 +1808,27 @@ static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) return ret; } +static int qm_start_qp_nolock(struct hisi_qp *qp, unsigned long arg) +{ + struct hisi_qm *qm = qp->qm; + struct device *dev = &qm->pdev->dev; + int qp_id = qp->qp_id; + int pasid = arg; + int ret; + + if (!qm_qp_avail_state(qm, qp, QP_START)) + return -EPERM; + + ret = qm_qp_ctx_cfg(qp, qp_id, pasid); + if (ret) + return ret; + + atomic_set(&qp->qp_status.flags, QP_START); + dev_dbg(dev, "queue %d started\n", qp_id); + + return 0; +} + /** * hisi_qm_start_qp() - Start a qp into running. * @qp: The qp we want to start to run. @@ -1285,48 +1840,112 @@ static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) int hisi_qm_start_qp(struct hisi_qp *qp, unsigned long arg) { struct hisi_qm *qm = qp->qm; - struct device *dev = &qm->pdev->dev; - enum qm_hw_ver ver = qm->ver; - int qp_id = qp->qp_id; - int pasid = arg; - size_t off = 0; int ret; -#define QP_INIT_BUF(qp, type, size) do { \ - (qp)->type = ((qp)->qdma.va + (off)); \ - (qp)->type##_dma = (qp)->qdma.dma + (off); \ - off += (size); \ -} while (0) + down_write(&qm->qps_lock); + ret = qm_start_qp_nolock(qp, arg); + up_write(&qm->qps_lock); - if (!qp->qdma.dma) { - dev_err(dev, "cannot get qm dma buffer\n"); - return -EINVAL; + return ret; +} +EXPORT_SYMBOL_GPL(hisi_qm_start_qp); + +/** + * Determine whether the queue is cleared by judging the tail pointers of + * sq and cq. + */ +static int qm_drain_qp(struct hisi_qp *qp) +{ + size_t size = sizeof(struct qm_sqc) + sizeof(struct qm_cqc); + struct hisi_qm *qm = qp->qm; + struct device *dev = &qm->pdev->dev; + struct qm_sqc *sqc; + struct qm_cqc *cqc; + dma_addr_t dma_addr; + int ret = 0, i = 0; + void *addr; + + /* + * No need to judge if ECC multi-bit error occurs because the + * master OOO will be blocked. + */ + if (qm->err_status.is_qm_ecc_mbit || qm->err_status.is_dev_ecc_mbit) + return 0; + + addr = qm_ctx_alloc(qm, size, &dma_addr); + if (IS_ERR(addr)) { + dev_err(dev, "Failed to alloc ctx for sqc and cqc!\n"); + return -ENOMEM; } - /* sq need 128 bytes alignment */ - if (qp->qdma.dma & QM_SQE_DATA_ALIGN_MASK) { - dev_err(dev, "qm sq is not aligned to 128 byte\n"); - return -EINVAL; + while (++i) { + ret = qm_dump_sqc_raw(qm, dma_addr, qp->qp_id); + if (ret) { + dev_err_ratelimited(dev, "Failed to dump sqc!\n"); + break; + } + sqc = addr; + + ret = qm_dump_cqc_raw(qm, (dma_addr + sizeof(struct qm_sqc)), + qp->qp_id); + if (ret) { + dev_err_ratelimited(dev, "Failed to dump cqc!\n"); + break; + } + cqc = addr + sizeof(struct qm_sqc); + + if ((sqc->tail == cqc->tail) && + (QM_SQ_TAIL_IDX(sqc) == QM_CQ_TAIL_IDX(cqc))) + break; + + if (i == MAX_WAIT_COUNTS) { + dev_err(dev, "Fail to empty queue %u!\n", qp->qp_id); + ret = -EBUSY; + break; + } + + usleep_range(WAIT_PERIOD_US_MIN, WAIT_PERIOD_US_MAX); } - QP_INIT_BUF(qp, sqe, qm->sqe_size * QM_Q_DEPTH); - QP_INIT_BUF(qp, cqe, sizeof(struct qm_cqe) * QM_Q_DEPTH); + qm_ctx_free(qm, size, addr, &dma_addr); - dev_dbg(dev, "init qp buffer(v%d):\n" - " sqe (%pK, %lx)\n" - " cqe (%pK, %lx)\n", - ver, qp->sqe, (unsigned long)qp->sqe_dma, - qp->cqe, (unsigned long)qp->cqe_dma); + return ret; +} - ret = qm_qp_ctx_cfg(qp, qp_id, pasid); +static int qm_stop_qp_nolock(struct hisi_qp *qp) +{ + struct device *dev = &qp->qm->pdev->dev; + int ret; + + /* + * It is allowed to stop and release qp when reset, If the qp is + * stopped when reset but still want to be released then, the + * is_resetting flag should be set negative so that this qp will not + * be restarted after reset. + */ + if (atomic_read(&qp->qp_status.flags) == QP_STOP) { + qp->is_resetting = false; + return 0; + } + + if (!qm_qp_avail_state(qp->qm, qp, QP_STOP)) + return -EPERM; + + atomic_set(&qp->qp_status.flags, QP_STOP); + + ret = qm_drain_qp(qp); if (ret) - return ret; + dev_err(dev, "Failed to drain out data for stopping!\n"); - dev_dbg(dev, "queue %d started\n", qp_id); + if (qp->qm->wq) + flush_workqueue(qp->qm->wq); + else + flush_work(&qp->qm->work); + + dev_dbg(dev, "stop queue %u!", qp->qp_id); return 0; } -EXPORT_SYMBOL_GPL(hisi_qm_start_qp); /** * hisi_qm_stop_qp() - Stop a qp in qm. @@ -1336,27 +1955,13 @@ EXPORT_SYMBOL_GPL(hisi_qm_start_qp); */ int hisi_qm_stop_qp(struct hisi_qp *qp) { - struct device *dev = &qp->qm->pdev->dev; - int i = 0; - - /* it is stopped */ - if (test_bit(QP_STOP, &qp->qp_status.flags)) - return 0; - - while (atomic_read(&qp->qp_status.used)) { - i++; - msleep(20); - if (i == 10) { - dev_err(dev, "Cannot drain out data for stopping, Force to stop!\n"); - return 0; - } - } - - set_bit(QP_STOP, &qp->qp_status.flags); + int ret; - dev_dbg(dev, "stop queue %u!", qp->qp_id); + down_write(&qp->qm->qps_lock); + ret = qm_stop_qp_nolock(qp); + up_write(&qp->qm->qps_lock); - return 0; + return ret; } EXPORT_SYMBOL_GPL(hisi_qm_stop_qp); @@ -1367,6 +1972,13 @@ EXPORT_SYMBOL_GPL(hisi_qm_stop_qp); * * This function will return -EBUSY if qp is currently full, and -EAGAIN * if qp related qm is resetting. + * + * Note: This function may run with qm_irq_thread and ACC reset at same time. + * It has no race with qm_irq_thread. However, during hisi_qp_send, ACC + * reset may happen, we have no lock here considering performance. This + * causes current qm_db sending fail or can not receive sended sqe. QM + * sync/async receive function should handle the error sqe. ACC reset + * done function should clear used sqe to 0. */ int hisi_qp_send(struct hisi_qp *qp, const void *msg) { @@ -1375,7 +1987,9 @@ int hisi_qp_send(struct hisi_qp *qp, const void *msg) u16 sq_tail_next = (sq_tail + 1) % QM_Q_DEPTH; void *sqe = qm_get_avail_sqe(qp); - if (unlikely(test_bit(QP_STOP, &qp->qp_status.flags))) { + if (unlikely(atomic_read(&qp->qp_status.flags) == QP_STOP || + atomic_read(&qp->qm->status.flags) == QM_STOP || + qp->is_resetting)) { dev_info(&qp->qm->pdev->dev, "QP is stopped or resetting\n"); return -EAGAIN; } @@ -1397,12 +2011,13 @@ static void hisi_qm_cache_wb(struct hisi_qm *qm) { unsigned int val; - if (qm->ver == QM_HW_V2) { - writel(0x1, qm->io_base + QM_CACHE_WB_START); - if (readl_relaxed_poll_timeout(qm->io_base + QM_CACHE_WB_DONE, - val, val & BIT(0), 10, 1000)) - dev_err(&qm->pdev->dev, "QM writeback sqc cache fail!\n"); - } + if (qm->ver == QM_HW_V1) + return; + + writel(0x1, qm->io_base + QM_CACHE_WB_START); + if (readl_relaxed_poll_timeout(qm->io_base + QM_CACHE_WB_DONE, + val, val & BIT(0), 10, 1000)) + dev_err(&qm->pdev->dev, "QM writeback sqc cache fail!\n"); } static void qm_qp_event_notifier(struct hisi_qp *qp) @@ -1412,16 +2027,7 @@ static void qm_qp_event_notifier(struct hisi_qp *qp) static int hisi_qm_get_available_instances(struct uacce_device *uacce) { - int i, ret; - struct hisi_qm *qm = uacce->priv; - - read_lock(&qm->qps_lock); - for (i = 0, ret = 0; i < qm->qp_num; i++) - if (!qm->qp_array[i]) - ret++; - read_unlock(&qm->qps_lock); - - return ret; + return hisi_qm_get_free_qp_num(uacce->priv); } static int hisi_qm_uacce_get_queue(struct uacce_device *uacce, @@ -1468,12 +2074,12 @@ static int hisi_qm_uacce_mmap(struct uacce_queue *q, switch (qfr->type) { case UACCE_QFRT_MMIO: - if (qm->ver == QM_HW_V2) { - if (sz > PAGE_SIZE * (QM_DOORBELL_PAGE_NR + - QM_DOORBELL_SQ_CQ_BASE_V2 / PAGE_SIZE)) + if (qm->ver == QM_HW_V1) { + if (sz > PAGE_SIZE * QM_DOORBELL_PAGE_NR) return -EINVAL; } else { - if (sz > PAGE_SIZE * QM_DOORBELL_PAGE_NR) + if (sz > PAGE_SIZE * (QM_DOORBELL_PAGE_NR + + QM_DOORBELL_SQ_CQ_BASE_V2 / PAGE_SIZE)) return -EINVAL; } @@ -1519,9 +2125,9 @@ static int qm_set_sqctype(struct uacce_queue *q, u16 type) struct hisi_qm *qm = q->uacce->priv; struct hisi_qp *qp = q->priv; - write_lock(&qm->qps_lock); + down_write(&qm->qps_lock); qp->alg_type = type; - write_unlock(&qm->qps_lock); + up_write(&qm->qps_lock); return 0; } @@ -1623,107 +2229,121 @@ int hisi_qm_get_free_qp_num(struct hisi_qm *qm) { int ret; - read_lock(&qm->qps_lock); + down_read(&qm->qps_lock); ret = qm->qp_num - qm->qp_in_used; - read_unlock(&qm->qps_lock); + up_read(&qm->qps_lock); return ret; } EXPORT_SYMBOL_GPL(hisi_qm_get_free_qp_num); -/** - * hisi_qm_init() - Initialize configures about qm. - * @qm: The qm needing init. - * - * This function init qm, then we can call hisi_qm_start to put qm into work. - */ -int hisi_qm_init(struct hisi_qm *qm) +static void hisi_qp_memory_uninit(struct hisi_qm *qm, int num) { - struct pci_dev *pdev = qm->pdev; - struct device *dev = &pdev->dev; - unsigned int num_vec; - int ret; + struct device *dev = &qm->pdev->dev; + struct qm_dma *qdma; + int i; - switch (qm->ver) { - case QM_HW_V1: - qm->ops = &qm_hw_ops_v1; - break; - case QM_HW_V2: - qm->ops = &qm_hw_ops_v2; - break; - default: - return -EINVAL; + for (i = num - 1; i >= 0; i--) { + qdma = &qm->qp_array[i].qdma; + dma_free_coherent(dev, qdma->size, qdma->va, qdma->dma); } - ret = qm_alloc_uacce(qm); - if (ret < 0) - dev_warn(&pdev->dev, "fail to alloc uacce (%d)\n", ret); + kfree(qm->qp_array); +} - ret = pci_enable_device_mem(pdev); - if (ret < 0) { - dev_err(&pdev->dev, "Failed to enable device mem!\n"); - goto err_remove_uacce; - } +static int hisi_qp_memory_init(struct hisi_qm *qm, size_t dma_size, int id) +{ + struct device *dev = &qm->pdev->dev; + size_t off = qm->sqe_size * QM_Q_DEPTH; + struct hisi_qp *qp; - ret = pci_request_mem_regions(pdev, qm->dev_name); - if (ret < 0) { - dev_err(&pdev->dev, "Failed to request mem regions!\n"); - goto err_disable_pcidev; - } + qp = &qm->qp_array[id]; + qp->qdma.va = dma_alloc_coherent(dev, dma_size, &qp->qdma.dma, + GFP_KERNEL); + if (!qp->qdma.va) + return -ENOMEM; - qm->phys_base = pci_resource_start(pdev, PCI_BAR_2); - qm->phys_size = pci_resource_len(qm->pdev, PCI_BAR_2); - qm->io_base = ioremap(qm->phys_base, qm->phys_size); - if (!qm->io_base) { - ret = -EIO; - goto err_release_mem_regions; - } + qp->sqe = qp->qdma.va; + qp->sqe_dma = qp->qdma.dma; + qp->cqe = qp->qdma.va + off; + qp->cqe_dma = qp->qdma.dma + off; + qp->qdma.size = dma_size; + qp->qm = qm; + qp->qp_id = id; - ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64)); - if (ret < 0) - goto err_iounmap; - pci_set_master(pdev); + return 0; +} - if (!qm->ops->get_irq_num) { - ret = -EOPNOTSUPP; - goto err_iounmap; +static int hisi_qm_memory_init(struct hisi_qm *qm) +{ + struct device *dev = &qm->pdev->dev; + size_t qp_dma_size, off = 0; + int i, ret = 0; + +#define QM_INIT_BUF(qm, type, num) do { \ + (qm)->type = ((qm)->qdma.va + (off)); \ + (qm)->type##_dma = (qm)->qdma.dma + (off); \ + off += QMC_ALIGN(sizeof(struct qm_##type) * (num)); \ +} while (0) + + idr_init(&qm->qp_idr); + qm->qdma.size = QMC_ALIGN(sizeof(struct qm_eqe) * QM_Q_DEPTH) + + QMC_ALIGN(sizeof(struct qm_aeqe) * QM_Q_DEPTH) + + QMC_ALIGN(sizeof(struct qm_sqc) * qm->qp_num) + + QMC_ALIGN(sizeof(struct qm_cqc) * qm->qp_num); + qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, &qm->qdma.dma, + GFP_ATOMIC); + dev_dbg(dev, "allocate qm dma buf size=%zx)\n", qm->qdma.size); + if (!qm->qdma.va) + return -ENOMEM; + + QM_INIT_BUF(qm, eqe, QM_Q_DEPTH); + QM_INIT_BUF(qm, aeqe, QM_Q_DEPTH); + QM_INIT_BUF(qm, sqc, qm->qp_num); + QM_INIT_BUF(qm, cqc, qm->qp_num); + + qm->qp_array = kcalloc(qm->qp_num, sizeof(struct hisi_qp), GFP_KERNEL); + if (!qm->qp_array) { + ret = -ENOMEM; + goto err_alloc_qp_array; } - num_vec = qm->ops->get_irq_num(qm); - ret = pci_alloc_irq_vectors(pdev, num_vec, num_vec, PCI_IRQ_MSI); - if (ret < 0) { - dev_err(dev, "Failed to enable MSI vectors!\n"); - goto err_iounmap; + + /* one more page for device or qp statuses */ + qp_dma_size = qm->sqe_size * QM_Q_DEPTH + + sizeof(struct qm_cqe) * QM_Q_DEPTH; + qp_dma_size = PAGE_ALIGN(qp_dma_size); + for (i = 0; i < qm->qp_num; i++) { + ret = hisi_qp_memory_init(qm, qp_dma_size, i); + if (ret) + goto err_init_qp_mem; + + dev_dbg(dev, "allocate qp dma buf size=%zx)\n", qp_dma_size); } - ret = qm_irq_register(qm); - if (ret) - goto err_free_irq_vectors; + return ret; - qm->qp_in_used = 0; - mutex_init(&qm->mailbox_lock); - rwlock_init(&qm->qps_lock); - INIT_WORK(&qm->work, qm_work_process); +err_init_qp_mem: + hisi_qp_memory_uninit(qm, i); +err_alloc_qp_array: + dma_free_coherent(dev, qm->qdma.size, qm->qdma.va, qm->qdma.dma); - dev_dbg(dev, "init qm %s with %s\n", pdev->is_physfn ? "pf" : "vf", - qm->use_dma_api ? "dma api" : "iommu api"); + return ret; +} - return 0; +static void hisi_qm_pre_init(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; -err_free_irq_vectors: - pci_free_irq_vectors(pdev); -err_iounmap: - iounmap(qm->io_base); -err_release_mem_regions: - pci_release_mem_regions(pdev); -err_disable_pcidev: - pci_disable_device(pdev); -err_remove_uacce: - uacce_remove(qm->uacce); - qm->uacce = NULL; + if (qm->ver == QM_HW_V1) + qm->ops = &qm_hw_ops_v1; + else + qm->ops = &qm_hw_ops_v2; - return ret; + pci_set_drvdata(pdev, qm); + mutex_init(&qm->mailbox_lock); + init_rwsem(&qm->qps_lock); + qm->qp_in_used = 0; } -EXPORT_SYMBOL_GPL(hisi_qm_init); /** * hisi_qm_uninit() - Uninitialize qm. @@ -1736,10 +2356,20 @@ void hisi_qm_uninit(struct hisi_qm *qm) struct pci_dev *pdev = qm->pdev; struct device *dev = &pdev->dev; + down_write(&qm->qps_lock); + + if (!qm_avail_state(qm, QM_CLOSE)) { + up_write(&qm->qps_lock); + return; + } + uacce_remove(qm->uacce); qm->uacce = NULL; - if (qm->use_dma_api && qm->qdma.va) { + hisi_qp_memory_uninit(qm, qm->qp_num); + idr_destroy(&qm->qp_idr); + + if (qm->qdma.va) { hisi_qm_cache_wb(qm); dma_free_coherent(dev, qm->qdma.size, qm->qdma.va, qm->qdma.dma); @@ -1751,6 +2381,8 @@ void hisi_qm_uninit(struct hisi_qm *qm) iounmap(qm->io_base); pci_release_mem_regions(pdev); pci_disable_device(pdev); + + up_write(&qm->qps_lock); } EXPORT_SYMBOL_GPL(hisi_qm_uninit); @@ -1781,12 +2413,6 @@ int hisi_qm_get_vft(struct hisi_qm *qm, u32 *base, u32 *number) EXPORT_SYMBOL_GPL(hisi_qm_get_vft); /** - * hisi_qm_set_vft() - Set "virtual function table" for a qm. - * @fun_num: Number of operated function. - * @qm: The qm in which to set vft, alway in a PF. - * @base: The base number of queue in vft. - * @number: The number of queues in vft. 0 means invalid vft. - * * This function is alway called in PF driver, it is used to assign queues * among PF and VFs. * @@ -1794,7 +2420,7 @@ EXPORT_SYMBOL_GPL(hisi_qm_get_vft); * Assign queues A~B to VF: hisi_qm_set_vft(qm, 2, A, B - A + 1) * (VF function number 0x2) */ -int hisi_qm_set_vft(struct hisi_qm *qm, u32 fun_num, u32 base, +static int hisi_qm_set_vft(struct hisi_qm *qm, u32 fun_num, u32 base, u32 number) { u32 max_q_num = qm->ctrl_qp_num; @@ -1805,7 +2431,6 @@ int hisi_qm_set_vft(struct hisi_qm *qm, u32 fun_num, u32 base, return qm_set_sqc_cqc_vft(qm, fun_num, base, number); } -EXPORT_SYMBOL_GPL(hisi_qm_set_vft); static void qm_init_eq_aeq_status(struct hisi_qm *qm) { @@ -1872,22 +2497,10 @@ static int qm_eq_ctx_cfg(struct hisi_qm *qm) static int __hisi_qm_start(struct hisi_qm *qm) { - struct pci_dev *pdev = qm->pdev; - struct device *dev = &pdev->dev; - size_t off = 0; int ret; -#define QM_INIT_BUF(qm, type, num) do { \ - (qm)->type = ((qm)->qdma.va + (off)); \ - (qm)->type##_dma = (qm)->qdma.dma + (off); \ - off += QMC_ALIGN(sizeof(struct qm_##type) * (num)); \ -} while (0) - WARN_ON(!qm->qdma.dma); - if (qm->qp_num == 0) - return -EINVAL; - if (qm->fun_type == QM_HW_PF) { ret = qm_dev_mem_reset(qm); if (ret) @@ -1898,21 +2511,6 @@ static int __hisi_qm_start(struct hisi_qm *qm) return ret; } - QM_INIT_BUF(qm, eqe, QM_Q_DEPTH); - QM_INIT_BUF(qm, aeqe, QM_Q_DEPTH); - QM_INIT_BUF(qm, sqc, qm->qp_num); - QM_INIT_BUF(qm, cqc, qm->qp_num); - - dev_dbg(dev, "init qm buffer:\n" - " eqe (%pK, %lx)\n" - " aeqe (%pK, %lx)\n" - " sqc (%pK, %lx)\n" - " cqc (%pK, %lx)\n", - qm->eqe, (unsigned long)qm->eqe_dma, - qm->aeqe, (unsigned long)qm->aeqe_dma, - qm->sqc, (unsigned long)qm->sqc_dma, - qm->cqc, (unsigned long)qm->cqc_dma); - ret = qm_eq_ctx_cfg(qm); if (ret) return ret; @@ -1940,43 +2538,102 @@ static int __hisi_qm_start(struct hisi_qm *qm) int hisi_qm_start(struct hisi_qm *qm) { struct device *dev = &qm->pdev->dev; + int ret = 0; + + down_write(&qm->qps_lock); + + if (!qm_avail_state(qm, QM_START)) { + up_write(&qm->qps_lock); + return -EPERM; + } dev_dbg(dev, "qm start with %d queue pairs\n", qm->qp_num); if (!qm->qp_num) { dev_err(dev, "qp_num should not be 0\n"); - return -EINVAL; + ret = -EINVAL; + goto err_unlock; } - if (!qm->qp_bitmap) { - qm->qp_bitmap = devm_kcalloc(dev, BITS_TO_LONGS(qm->qp_num), - sizeof(long), GFP_KERNEL); - qm->qp_array = devm_kcalloc(dev, qm->qp_num, - sizeof(struct hisi_qp *), - GFP_KERNEL); - if (!qm->qp_bitmap || !qm->qp_array) - return -ENOMEM; + ret = __hisi_qm_start(qm); + if (!ret) + atomic_set(&qm->status.flags, QM_START); + +err_unlock: + up_write(&qm->qps_lock); + return ret; +} +EXPORT_SYMBOL_GPL(hisi_qm_start); + +static int qm_restart(struct hisi_qm *qm) +{ + struct device *dev = &qm->pdev->dev; + struct hisi_qp *qp; + int ret, i; + + ret = hisi_qm_start(qm); + if (ret < 0) + return ret; + + down_write(&qm->qps_lock); + for (i = 0; i < qm->qp_num; i++) { + qp = &qm->qp_array[i]; + if (atomic_read(&qp->qp_status.flags) == QP_STOP && + qp->is_resetting == true) { + ret = qm_start_qp_nolock(qp, 0); + if (ret < 0) { + dev_err(dev, "Failed to start qp%d!\n", i); + + up_write(&qm->qps_lock); + return ret; + } + qp->is_resetting = false; + } } + up_write(&qm->qps_lock); - if (!qm->use_dma_api) { - dev_dbg(&qm->pdev->dev, "qm delay start\n"); - return 0; - } else if (!qm->qdma.va) { - qm->qdma.size = QMC_ALIGN(sizeof(struct qm_eqe) * QM_Q_DEPTH) + - QMC_ALIGN(sizeof(struct qm_aeqe) * QM_Q_DEPTH) + - QMC_ALIGN(sizeof(struct qm_sqc) * qm->qp_num) + - QMC_ALIGN(sizeof(struct qm_cqc) * qm->qp_num); - qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, - &qm->qdma.dma, GFP_KERNEL); - dev_dbg(dev, "allocate qm dma buf(va=%pK, dma=%pad, size=%zx)\n", - qm->qdma.va, &qm->qdma.dma, qm->qdma.size); - if (!qm->qdma.va) - return -ENOMEM; + return 0; +} + +/* Stop started qps in reset flow */ +static int qm_stop_started_qp(struct hisi_qm *qm) +{ + struct device *dev = &qm->pdev->dev; + struct hisi_qp *qp; + int i, ret; + + for (i = 0; i < qm->qp_num; i++) { + qp = &qm->qp_array[i]; + if (qp && atomic_read(&qp->qp_status.flags) == QP_START) { + qp->is_resetting = true; + ret = qm_stop_qp_nolock(qp); + if (ret < 0) { + dev_err(dev, "Failed to stop qp%d!\n", i); + return ret; + } + } } - return __hisi_qm_start(qm); + return 0; +} + +/** + * This function clears all queues memory in a qm. Reset of accelerator can + * use this to clear queues. + */ +static void qm_clear_queues(struct hisi_qm *qm) +{ + struct hisi_qp *qp; + int i; + + for (i = 0; i < qm->qp_num; i++) { + qp = &qm->qp_array[i]; + if (qp->is_resetting) + memset(qp->qdma.va, 0, qp->qdma.size); + } + + memset(qm->qdma.va, 0, qm->qdma.size); } -EXPORT_SYMBOL_GPL(hisi_qm_start); /** * hisi_qm_stop() - Stop a qm. @@ -1988,43 +2645,98 @@ EXPORT_SYMBOL_GPL(hisi_qm_start); */ int hisi_qm_stop(struct hisi_qm *qm) { - struct device *dev; - struct hisi_qp *qp; - int ret = 0, i; + struct device *dev = &qm->pdev->dev; + int ret = 0; - if (!qm || !qm->pdev) { - WARN_ON(1); - return -EINVAL; + down_write(&qm->qps_lock); + + if (!qm_avail_state(qm, QM_STOP)) { + ret = -EPERM; + goto err_unlock; } - dev = &qm->pdev->dev; + if (qm->status.stop_reason == QM_SOFT_RESET || + qm->status.stop_reason == QM_FLR) { + ret = qm_stop_started_qp(qm); + if (ret < 0) { + dev_err(dev, "Failed to stop started qp!\n"); + goto err_unlock; + } + } /* Mask eq and aeq irq */ writel(0x1, qm->io_base + QM_VF_EQ_INT_MASK); writel(0x1, qm->io_base + QM_VF_AEQ_INT_MASK); - /* Stop all qps belong to this qm */ - for (i = 0; i < qm->qp_num; i++) { - qp = qm->qp_array[i]; - if (qp) { - ret = hisi_qm_stop_qp(qp); - if (ret < 0) { - dev_err(dev, "Failed to stop qp%d!\n", i); - return -EBUSY; - } - } - } - if (qm->fun_type == QM_HW_PF) { ret = hisi_qm_set_vft(qm, 0, 0, 0); - if (ret < 0) + if (ret < 0) { dev_err(dev, "Failed to set vft!\n"); + ret = -EBUSY; + goto err_unlock; + } } + qm_clear_queues(qm); + atomic_set(&qm->status.flags, QM_STOP); + +err_unlock: + up_write(&qm->qps_lock); return ret; } EXPORT_SYMBOL_GPL(hisi_qm_stop); +static ssize_t qm_status_read(struct file *filp, char __user *buffer, + size_t count, loff_t *pos) +{ + struct hisi_qm *qm = filp->private_data; + char buf[QM_DBG_READ_LEN]; + int val, cp_len, len; + + if (*pos) + return 0; + + if (count < QM_DBG_READ_LEN) + return -ENOSPC; + + val = atomic_read(&qm->status.flags); + len = snprintf(buf, QM_DBG_READ_LEN, "%s\n", qm_s[val]); + if (!len) + return -EFAULT; + + cp_len = copy_to_user(buffer, buf, len); + if (cp_len) + return -EFAULT; + + return (*pos = len); +} + +static const struct file_operations qm_status_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .read = qm_status_read, +}; + +static int qm_debugfs_atomic64_set(void *data, u64 val) +{ + if (val) + return -EINVAL; + + atomic64_set((atomic64_t *)data, 0); + + return 0; +} + +static int qm_debugfs_atomic64_get(void *data, u64 *val) +{ + *val = atomic64_read((atomic64_t *)data); + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(qm_atomic64_ops, qm_debugfs_atomic64_get, + qm_debugfs_atomic64_set, "%llu\n"); + /** * hisi_qm_debug_init() - Initialize qm related debugfs files. * @qm: The qm for which we want to add debugfs files. @@ -2033,7 +2745,9 @@ EXPORT_SYMBOL_GPL(hisi_qm_stop); */ int hisi_qm_debug_init(struct hisi_qm *qm) { + struct qm_dfx *dfx = &qm->debug.dfx; struct dentry *qm_d; + void *data; int i, ret; qm_d = debugfs_create_dir("qm", qm->debug.debug_root); @@ -2047,7 +2761,20 @@ int hisi_qm_debug_init(struct hisi_qm *qm) goto failed_to_create; } - debugfs_create_file("qm_regs", 0444, qm->debug.qm_d, qm, &qm_regs_fops); + debugfs_create_file("regs", 0444, qm->debug.qm_d, qm, &qm_regs_fops); + + debugfs_create_file("cmd", 0444, qm->debug.qm_d, qm, &qm_cmd_fops); + + debugfs_create_file("status", 0444, qm->debug.qm_d, qm, + &qm_status_fops); + for (i = 0; i < ARRAY_SIZE(qm_dfx_files); i++) { + data = (atomic64_t *)((uintptr_t)dfx + qm_dfx_files[i].offset); + debugfs_create_file(qm_dfx_files[i].name, + 0644, + qm_d, + data, + &qm_atomic64_ops); + } return 0; @@ -2095,8 +2822,7 @@ static void qm_hw_error_init(struct hisi_qm *qm) return; } - qm->ops->hw_error_init(qm, err_info->ce, err_info->nfe, - err_info->fe, err_info->msi); + qm->ops->hw_error_init(qm, err_info->ce, err_info->nfe, err_info->fe); } static void qm_hw_error_uninit(struct hisi_qm *qm) @@ -2109,36 +2835,17 @@ static void qm_hw_error_uninit(struct hisi_qm *qm) qm->ops->hw_error_uninit(qm); } -static pci_ers_result_t qm_hw_error_handle(struct hisi_qm *qm) +static enum acc_err_result qm_hw_error_handle(struct hisi_qm *qm) { if (!qm->ops->hw_error_handle) { dev_err(&qm->pdev->dev, "QM doesn't support hw error report!\n"); - return PCI_ERS_RESULT_NONE; + return ACC_ERR_NONE; } return qm->ops->hw_error_handle(qm); } /** - * hisi_qm_get_hw_version() - Get hardware version of a qm. - * @pdev: The device which hardware version we want to get. - * - * This function gets the hardware version of a qm. Return QM_HW_UNKNOWN - * if the hardware version is not supported. - */ -enum qm_hw_ver hisi_qm_get_hw_version(struct pci_dev *pdev) -{ - switch (pdev->revision) { - case QM_HW_V1: - case QM_HW_V2: - return pdev->revision; - default: - return QM_HW_UNKNOWN; - } -} -EXPORT_SYMBOL_GPL(hisi_qm_get_hw_version); - -/** * hisi_qm_dev_err_init() - Initialize device error configuration. * @qm: The qm for which we want to do error initialization. * @@ -2299,34 +3006,163 @@ err: } EXPORT_SYMBOL_GPL(hisi_qm_alloc_qps_node); -static pci_ers_result_t qm_dev_err_handle(struct hisi_qm *qm) +static int qm_vf_q_assign(struct hisi_qm *qm, u32 num_vfs) +{ + u32 remain_q_num, q_num, i, j; + u32 q_base = qm->qp_num; + int ret; + + if (!num_vfs) + return -EINVAL; + + remain_q_num = qm->ctrl_qp_num - qm->qp_num; + + /* If remain queues not enough, return error. */ + if (qm->ctrl_qp_num < qm->qp_num || remain_q_num < num_vfs) + return -EINVAL; + + q_num = remain_q_num / num_vfs; + for (i = 1; i <= num_vfs; i++) { + if (i == num_vfs) + q_num += remain_q_num % num_vfs; + ret = hisi_qm_set_vft(qm, i, q_base, q_num); + if (ret) { + for (j = i; j > 0; j--) + hisi_qm_set_vft(qm, j, 0, 0); + return ret; + } + q_base += q_num; + } + + return 0; +} + +static int qm_clear_vft_config(struct hisi_qm *qm) +{ + int ret; + u32 i; + + for (i = 1; i <= qm->vfs_num; i++) { + ret = hisi_qm_set_vft(qm, i, 0, 0); + if (ret) + return ret; + } + qm->vfs_num = 0; + + return 0; +} + +/** + * hisi_qm_sriov_enable() - enable virtual functions + * @pdev: the PCIe device + * @max_vfs: the number of virtual functions to enable + * + * Returns the number of enabled VFs. If there are VFs enabled already or + * max_vfs is more than the total number of device can be enabled, returns + * failure. + */ +int hisi_qm_sriov_enable(struct pci_dev *pdev, int max_vfs) +{ + struct hisi_qm *qm = pci_get_drvdata(pdev); + int pre_existing_vfs, num_vfs, total_vfs, ret; + + total_vfs = pci_sriov_get_totalvfs(pdev); + pre_existing_vfs = pci_num_vf(pdev); + if (pre_existing_vfs) { + pci_err(pdev, "%d VFs already enabled. Please disable pre-enabled VFs!\n", + pre_existing_vfs); + return 0; + } + + num_vfs = min_t(int, max_vfs, total_vfs); + ret = qm_vf_q_assign(qm, num_vfs); + if (ret) { + pci_err(pdev, "Can't assign queues for VF!\n"); + return ret; + } + + qm->vfs_num = num_vfs; + + ret = pci_enable_sriov(pdev, num_vfs); + if (ret) { + pci_err(pdev, "Can't enable VF!\n"); + qm_clear_vft_config(qm); + return ret; + } + + pci_info(pdev, "VF enabled, vfs_num(=%d)!\n", num_vfs); + + return num_vfs; +} +EXPORT_SYMBOL_GPL(hisi_qm_sriov_enable); + +/** + * hisi_qm_sriov_disable - disable virtual functions + * @pdev: the PCI device + * + * Return failure if there are VFs assigned already. + */ +int hisi_qm_sriov_disable(struct pci_dev *pdev) +{ + struct hisi_qm *qm = pci_get_drvdata(pdev); + + if (pci_vfs_assigned(pdev)) { + pci_err(pdev, "Failed to disable VFs as VFs are assigned!\n"); + return -EPERM; + } + + /* remove in hpre_pci_driver will be called to free VF resources */ + pci_disable_sriov(pdev); + return qm_clear_vft_config(qm); +} +EXPORT_SYMBOL_GPL(hisi_qm_sriov_disable); + +/** + * hisi_qm_sriov_configure - configure the number of VFs + * @pdev: The PCI device + * @num_vfs: The number of VFs need enabled + * + * Enable SR-IOV according to num_vfs, 0 means disable. + */ +int hisi_qm_sriov_configure(struct pci_dev *pdev, int num_vfs) +{ + if (num_vfs == 0) + return hisi_qm_sriov_disable(pdev); + else + return hisi_qm_sriov_enable(pdev, num_vfs); +} +EXPORT_SYMBOL_GPL(hisi_qm_sriov_configure); + +static enum acc_err_result qm_dev_err_handle(struct hisi_qm *qm) { u32 err_sts; if (!qm->err_ini->get_dev_hw_err_status) { dev_err(&qm->pdev->dev, "Device doesn't support get hw error status!\n"); - return PCI_ERS_RESULT_NONE; + return ACC_ERR_NONE; } /* get device hardware error status */ err_sts = qm->err_ini->get_dev_hw_err_status(qm); if (err_sts) { + if (err_sts & qm->err_ini->err_info.ecc_2bits_mask) + qm->err_status.is_dev_ecc_mbit = true; + if (!qm->err_ini->log_dev_hw_err) { dev_err(&qm->pdev->dev, "Device doesn't support log hw error!\n"); - return PCI_ERS_RESULT_NEED_RESET; + return ACC_ERR_NEED_RESET; } qm->err_ini->log_dev_hw_err(qm, err_sts); - return PCI_ERS_RESULT_NEED_RESET; + return ACC_ERR_NEED_RESET; } - return PCI_ERS_RESULT_RECOVERED; + return ACC_ERR_RECOVERED; } -static pci_ers_result_t qm_process_dev_error(struct pci_dev *pdev) +static enum acc_err_result qm_process_dev_error(struct hisi_qm *qm) { - struct hisi_qm *qm = pci_get_drvdata(pdev); - pci_ers_result_t qm_ret, dev_ret; + enum acc_err_result qm_ret, dev_ret; /* log qm error */ qm_ret = qm_hw_error_handle(qm); @@ -2334,9 +3170,9 @@ static pci_ers_result_t qm_process_dev_error(struct pci_dev *pdev) /* log device error */ dev_ret = qm_dev_err_handle(qm); - return (qm_ret == PCI_ERS_RESULT_NEED_RESET || - dev_ret == PCI_ERS_RESULT_NEED_RESET) ? - PCI_ERS_RESULT_NEED_RESET : PCI_ERS_RESULT_RECOVERED; + return (qm_ret == ACC_ERR_NEED_RESET || + dev_ret == ACC_ERR_NEED_RESET) ? + ACC_ERR_NEED_RESET : ACC_ERR_RECOVERED; } /** @@ -2350,6 +3186,9 @@ static pci_ers_result_t qm_process_dev_error(struct pci_dev *pdev) pci_ers_result_t hisi_qm_dev_err_detected(struct pci_dev *pdev, pci_channel_state_t state) { + struct hisi_qm *qm = pci_get_drvdata(pdev); + enum acc_err_result ret; + if (pdev->is_virtfn) return PCI_ERS_RESULT_NONE; @@ -2357,10 +3196,756 @@ pci_ers_result_t hisi_qm_dev_err_detected(struct pci_dev *pdev, if (state == pci_channel_io_perm_failure) return PCI_ERS_RESULT_DISCONNECT; - return qm_process_dev_error(pdev); + ret = qm_process_dev_error(qm); + if (ret == ACC_ERR_NEED_RESET) + return PCI_ERS_RESULT_NEED_RESET; + + return PCI_ERS_RESULT_RECOVERED; } EXPORT_SYMBOL_GPL(hisi_qm_dev_err_detected); +static int qm_get_hw_error_status(struct hisi_qm *qm) +{ + return readl(qm->io_base + QM_ABNORMAL_INT_STATUS); +} + +static int qm_check_req_recv(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + u32 val; + + writel(ACC_VENDOR_ID_VALUE, qm->io_base + QM_PEH_VENDOR_ID); + ret = readl_relaxed_poll_timeout(qm->io_base + QM_PEH_VENDOR_ID, val, + (val == ACC_VENDOR_ID_VALUE), + POLL_PERIOD, POLL_TIMEOUT); + if (ret) { + dev_err(&pdev->dev, "Fails to read QM reg!\n"); + return ret; + } + + writel(PCI_VENDOR_ID_HUAWEI, qm->io_base + QM_PEH_VENDOR_ID); + ret = readl_relaxed_poll_timeout(qm->io_base + QM_PEH_VENDOR_ID, val, + (val == PCI_VENDOR_ID_HUAWEI), + POLL_PERIOD, POLL_TIMEOUT); + if (ret) + dev_err(&pdev->dev, "Fails to read QM reg in the second time!\n"); + + return ret; +} + +static int qm_set_pf_mse(struct hisi_qm *qm, bool set) +{ + struct pci_dev *pdev = qm->pdev; + u16 cmd; + int i; + + pci_read_config_word(pdev, PCI_COMMAND, &cmd); + if (set) + cmd |= PCI_COMMAND_MEMORY; + else + cmd &= ~PCI_COMMAND_MEMORY; + + pci_write_config_word(pdev, PCI_COMMAND, cmd); + for (i = 0; i < MAX_WAIT_COUNTS; i++) { + pci_read_config_word(pdev, PCI_COMMAND, &cmd); + if (set == ((cmd & PCI_COMMAND_MEMORY) >> 1)) + return 0; + + udelay(1); + } + + return -ETIMEDOUT; +} + +static int qm_set_vf_mse(struct hisi_qm *qm, bool set) +{ + struct pci_dev *pdev = qm->pdev; + u16 sriov_ctrl; + int pos; + int i; + + pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_SRIOV); + pci_read_config_word(pdev, pos + PCI_SRIOV_CTRL, &sriov_ctrl); + if (set) + sriov_ctrl |= PCI_SRIOV_CTRL_MSE; + else + sriov_ctrl &= ~PCI_SRIOV_CTRL_MSE; + pci_write_config_word(pdev, pos + PCI_SRIOV_CTRL, sriov_ctrl); + + for (i = 0; i < MAX_WAIT_COUNTS; i++) { + pci_read_config_word(pdev, pos + PCI_SRIOV_CTRL, &sriov_ctrl); + if (set == (sriov_ctrl & PCI_SRIOV_CTRL_MSE) >> + ACC_PEH_SRIOV_CTRL_VF_MSE_SHIFT) + return 0; + + udelay(1); + } + + return -ETIMEDOUT; +} + +static int qm_set_msi(struct hisi_qm *qm, bool set) +{ + struct pci_dev *pdev = qm->pdev; + + if (set) { + pci_write_config_dword(pdev, pdev->msi_cap + PCI_MSI_MASK_64, + 0); + } else { + pci_write_config_dword(pdev, pdev->msi_cap + PCI_MSI_MASK_64, + ACC_PEH_MSI_DISABLE); + if (qm->err_status.is_qm_ecc_mbit || + qm->err_status.is_dev_ecc_mbit) + return 0; + + mdelay(1); + if (readl(qm->io_base + QM_PEH_DFX_INFO0)) + return -EFAULT; + } + + return 0; +} + +static int qm_vf_reset_prepare(struct hisi_qm *qm) +{ + struct hisi_qm_list *qm_list = qm->qm_list; + int stop_reason = qm->status.stop_reason; + struct pci_dev *pdev = qm->pdev; + struct pci_dev *virtfn; + struct hisi_qm *vf_qm; + int ret = 0; + + mutex_lock(&qm_list->lock); + list_for_each_entry(vf_qm, &qm_list->list, list) { + virtfn = vf_qm->pdev; + if (virtfn == pdev) + continue; + + if (pci_physfn(virtfn) == pdev) { + vf_qm->status.stop_reason = stop_reason; + ret = hisi_qm_stop(vf_qm); + if (ret) + goto stop_fail; + } + } + +stop_fail: + mutex_unlock(&qm_list->lock); + return ret; +} + +static int qm_reset_prepare_ready(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(pdev)); + int delay = 0; + + /* All reset requests need to be queued for processing */ + while (test_and_set_bit(QM_DEV_RESET_FLAG, &pf_qm->reset_flag)) { + msleep(++delay); + if (delay > QM_RESET_WAIT_TIMEOUT) + return -EBUSY; + } + + return 0; +} + +static int qm_controller_reset_prepare(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + + ret = qm_reset_prepare_ready(qm); + if (ret) { + pci_err(pdev, "Controller reset not ready!\n"); + return ret; + } + + if (qm->vfs_num) { + ret = qm_vf_reset_prepare(qm); + if (ret) { + pci_err(pdev, "Fails to stop VFs!\n"); + return ret; + } + } + + qm->status.stop_reason = QM_SOFT_RESET; + ret = hisi_qm_stop(qm); + if (ret) { + pci_err(pdev, "Fails to stop QM!\n"); + return ret; + } + + return 0; +} + +static void qm_dev_ecc_mbit_handle(struct hisi_qm *qm) +{ + u32 nfe_enb = 0; + + if (!qm->err_status.is_dev_ecc_mbit && + qm->err_status.is_qm_ecc_mbit && + qm->err_ini->close_axi_master_ooo) { + + qm->err_ini->close_axi_master_ooo(qm); + + } else if (qm->err_status.is_dev_ecc_mbit && + !qm->err_status.is_qm_ecc_mbit && + !qm->err_ini->close_axi_master_ooo) { + + nfe_enb = readl(qm->io_base + QM_RAS_NFE_ENABLE); + writel(nfe_enb & QM_RAS_NFE_MBIT_DISABLE, + qm->io_base + QM_RAS_NFE_ENABLE); + writel(QM_ECC_MBIT, qm->io_base + QM_ABNORMAL_INT_SET); + } +} + +static int qm_soft_reset(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + u32 val; + + /* Ensure all doorbells and mailboxes received by QM */ + ret = qm_check_req_recv(qm); + if (ret) + return ret; + + if (qm->vfs_num) { + ret = qm_set_vf_mse(qm, false); + if (ret) { + pci_err(pdev, "Fails to disable vf MSE bit.\n"); + return ret; + } + } + + ret = qm_set_msi(qm, false); + if (ret) { + pci_err(pdev, "Fails to disable PEH MSI bit.\n"); + return ret; + } + + qm_dev_ecc_mbit_handle(qm); + + /* OOO register set and check */ + writel(ACC_MASTER_GLOBAL_CTRL_SHUTDOWN, + qm->io_base + ACC_MASTER_GLOBAL_CTRL); + + /* If bus lock, reset chip */ + ret = readl_relaxed_poll_timeout(qm->io_base + ACC_MASTER_TRANS_RETURN, + val, + (val == ACC_MASTER_TRANS_RETURN_RW), + POLL_PERIOD, POLL_TIMEOUT); + if (ret) { + pci_emerg(pdev, "Bus lock! Please reset system.\n"); + return ret; + } + + ret = qm_set_pf_mse(qm, false); + if (ret) { + pci_err(pdev, "Fails to disable pf MSE bit.\n"); + return ret; + } + + /* The reset related sub-control registers are not in PCI BAR */ + if (ACPI_HANDLE(&pdev->dev)) { + unsigned long long value = 0; + acpi_status s; + + s = acpi_evaluate_integer(ACPI_HANDLE(&pdev->dev), + qm->err_ini->err_info.acpi_rst, + NULL, &value); + if (ACPI_FAILURE(s)) { + pci_err(pdev, "NO controller reset method!\n"); + return -EIO; + } + + if (value) { + pci_err(pdev, "Reset step %llu failed!\n", value); + return -EIO; + } + } else { + pci_err(pdev, "No reset method!\n"); + return -EINVAL; + } + + return 0; +} + +static int qm_vf_reset_done(struct hisi_qm *qm) +{ + struct hisi_qm_list *qm_list = qm->qm_list; + struct pci_dev *pdev = qm->pdev; + struct pci_dev *virtfn; + struct hisi_qm *vf_qm; + int ret = 0; + + mutex_lock(&qm_list->lock); + list_for_each_entry(vf_qm, &qm_list->list, list) { + virtfn = vf_qm->pdev; + if (virtfn == pdev) + continue; + + if (pci_physfn(virtfn) == pdev) { + ret = qm_restart(vf_qm); + if (ret) + goto restart_fail; + } + } + +restart_fail: + mutex_unlock(&qm_list->lock); + return ret; +} + +static int qm_get_dev_err_status(struct hisi_qm *qm) +{ + return qm->err_ini->get_dev_hw_err_status(qm); +} + +static int qm_dev_hw_init(struct hisi_qm *qm) +{ + return qm->err_ini->hw_init(qm); +} + +static void qm_restart_prepare(struct hisi_qm *qm) +{ + u32 value; + + if (!qm->err_status.is_qm_ecc_mbit && + !qm->err_status.is_dev_ecc_mbit) + return; + + /* temporarily close the OOO port used for PEH to write out MSI */ + value = readl(qm->io_base + ACC_AM_CFG_PORT_WR_EN); + writel(value & ~qm->err_ini->err_info.msi_wr_port, + qm->io_base + ACC_AM_CFG_PORT_WR_EN); + + /* clear dev ecc 2bit error source if having */ + value = qm_get_dev_err_status(qm) & + qm->err_ini->err_info.ecc_2bits_mask; + if (value && qm->err_ini->clear_dev_hw_err_status) + qm->err_ini->clear_dev_hw_err_status(qm, value); + + /* clear QM ecc mbit error source */ + writel(QM_ECC_MBIT, qm->io_base + QM_ABNORMAL_INT_SOURCE); + + /* clear AM Reorder Buffer ecc mbit source */ + writel(ACC_ROB_ECC_ERR_MULTPL, qm->io_base + ACC_AM_ROB_ECC_INT_STS); + + if (qm->err_ini->open_axi_master_ooo) + qm->err_ini->open_axi_master_ooo(qm); +} + +static void qm_restart_done(struct hisi_qm *qm) +{ + u32 value; + + if (!qm->err_status.is_qm_ecc_mbit && + !qm->err_status.is_dev_ecc_mbit) + return; + + /* open the OOO port for PEH to write out MSI */ + value = readl(qm->io_base + ACC_AM_CFG_PORT_WR_EN); + value |= qm->err_ini->err_info.msi_wr_port; + writel(value, qm->io_base + ACC_AM_CFG_PORT_WR_EN); + + qm->err_status.is_qm_ecc_mbit = false; + qm->err_status.is_dev_ecc_mbit = false; +} + +static int qm_controller_reset_done(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + + ret = qm_set_msi(qm, true); + if (ret) { + pci_err(pdev, "Fails to enable PEH MSI bit!\n"); + return ret; + } + + ret = qm_set_pf_mse(qm, true); + if (ret) { + pci_err(pdev, "Fails to enable pf MSE bit!\n"); + return ret; + } + + if (qm->vfs_num) { + ret = qm_set_vf_mse(qm, true); + if (ret) { + pci_err(pdev, "Fails to enable vf MSE bit!\n"); + return ret; + } + } + + ret = qm_dev_hw_init(qm); + if (ret) { + pci_err(pdev, "Failed to init device\n"); + return ret; + } + + qm_restart_prepare(qm); + + ret = qm_restart(qm); + if (ret) { + pci_err(pdev, "Failed to start QM!\n"); + return ret; + } + + if (qm->vfs_num) { + ret = qm_vf_q_assign(qm, qm->vfs_num); + if (ret) { + pci_err(pdev, "Failed to assign queue!\n"); + return ret; + } + } + + ret = qm_vf_reset_done(qm); + if (ret) { + pci_err(pdev, "Failed to start VFs!\n"); + return -EPERM; + } + + hisi_qm_dev_err_init(qm); + qm_restart_done(qm); + + clear_bit(QM_DEV_RESET_FLAG, &qm->reset_flag); + + return 0; +} + +static int qm_controller_reset(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + + pci_info(pdev, "Controller resetting...\n"); + + ret = qm_controller_reset_prepare(qm); + if (ret) + return ret; + + ret = qm_soft_reset(qm); + if (ret) { + pci_err(pdev, "Controller reset failed (%d)\n", ret); + return ret; + } + + ret = qm_controller_reset_done(qm); + if (ret) + return ret; + + pci_info(pdev, "Controller reset complete\n"); + + return 0; +} + +/** + * hisi_qm_dev_slot_reset() - slot reset + * @pdev: the PCIe device + * + * This function offers QM relate PCIe device reset interface. Drivers which + * use QM can use this function as slot_reset in its struct pci_error_handlers. + */ +pci_ers_result_t hisi_qm_dev_slot_reset(struct pci_dev *pdev) +{ + struct hisi_qm *qm = pci_get_drvdata(pdev); + int ret; + + if (pdev->is_virtfn) + return PCI_ERS_RESULT_RECOVERED; + + pci_aer_clear_nonfatal_status(pdev); + + /* reset pcie device controller */ + ret = qm_controller_reset(qm); + if (ret) { + pci_err(pdev, "Controller reset failed (%d)\n", ret); + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_RECOVERED; +} +EXPORT_SYMBOL_GPL(hisi_qm_dev_slot_reset); + +/* check the interrupt is ecc-mbit error or not */ +static int qm_check_dev_error(struct hisi_qm *qm) +{ + int ret; + + if (qm->fun_type == QM_HW_VF) + return 0; + + ret = qm_get_hw_error_status(qm) & QM_ECC_MBIT; + if (ret) + return ret; + + return (qm_get_dev_err_status(qm) & + qm->err_ini->err_info.ecc_2bits_mask); +} + +void hisi_qm_reset_prepare(struct pci_dev *pdev) +{ + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(pdev)); + struct hisi_qm *qm = pci_get_drvdata(pdev); + u32 delay = 0; + int ret; + + hisi_qm_dev_err_uninit(pf_qm); + + /* + * Check whether there is an ECC mbit error, If it occurs, need to + * wait for soft reset to fix it. + */ + while (qm_check_dev_error(pf_qm)) { + msleep(++delay); + if (delay > QM_RESET_WAIT_TIMEOUT) + return; + } + + ret = qm_reset_prepare_ready(qm); + if (ret) { + pci_err(pdev, "FLR not ready!\n"); + return; + } + + if (qm->vfs_num) { + ret = qm_vf_reset_prepare(qm); + if (ret) { + pci_err(pdev, "Failed to prepare reset, ret = %d.\n", + ret); + return; + } + } + + ret = hisi_qm_stop(qm); + if (ret) { + pci_err(pdev, "Failed to stop QM, ret = %d.\n", ret); + return; + } + + pci_info(pdev, "FLR resetting...\n"); +} +EXPORT_SYMBOL_GPL(hisi_qm_reset_prepare); + +static bool qm_flr_reset_complete(struct pci_dev *pdev) +{ + struct pci_dev *pf_pdev = pci_physfn(pdev); + struct hisi_qm *qm = pci_get_drvdata(pf_pdev); + u32 id; + + pci_read_config_dword(qm->pdev, PCI_COMMAND, &id); + if (id == QM_PCI_COMMAND_INVALID) { + pci_err(pdev, "Device can not be used!\n"); + return false; + } + + clear_bit(QM_DEV_RESET_FLAG, &qm->reset_flag); + + return true; +} + +void hisi_qm_reset_done(struct pci_dev *pdev) +{ + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(pdev)); + struct hisi_qm *qm = pci_get_drvdata(pdev); + int ret; + + hisi_qm_dev_err_init(pf_qm); + + ret = qm_restart(qm); + if (ret) { + pci_err(pdev, "Failed to start QM, ret = %d.\n", ret); + goto flr_done; + } + + if (qm->fun_type == QM_HW_PF) { + ret = qm_dev_hw_init(qm); + if (ret) { + pci_err(pdev, "Failed to init PF, ret = %d.\n", ret); + goto flr_done; + } + + if (!qm->vfs_num) + goto flr_done; + + ret = qm_vf_q_assign(qm, qm->vfs_num); + if (ret) { + pci_err(pdev, "Failed to assign VFs, ret = %d.\n", ret); + goto flr_done; + } + + ret = qm_vf_reset_done(qm); + if (ret) { + pci_err(pdev, "Failed to start VFs, ret = %d.\n", ret); + goto flr_done; + } + } + +flr_done: + if (qm_flr_reset_complete(pdev)) + pci_info(pdev, "FLR reset complete\n"); +} +EXPORT_SYMBOL_GPL(hisi_qm_reset_done); + +static irqreturn_t qm_abnormal_irq(int irq, void *data) +{ + struct hisi_qm *qm = data; + enum acc_err_result ret; + + atomic64_inc(&qm->debug.dfx.abnormal_irq_cnt); + ret = qm_process_dev_error(qm); + if (ret == ACC_ERR_NEED_RESET) + schedule_work(&qm->rst_work); + + return IRQ_HANDLED; +} + +static int qm_irq_register(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + int ret; + + ret = request_irq(pci_irq_vector(pdev, QM_EQ_EVENT_IRQ_VECTOR), + qm_irq, IRQF_SHARED, qm->dev_name, qm); + if (ret) + return ret; + + if (qm->ver != QM_HW_V1) { + ret = request_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), + qm_aeq_irq, IRQF_SHARED, qm->dev_name, qm); + if (ret) + goto err_aeq_irq; + + if (qm->fun_type == QM_HW_PF) { + ret = request_irq(pci_irq_vector(pdev, + QM_ABNORMAL_EVENT_IRQ_VECTOR), + qm_abnormal_irq, IRQF_SHARED, + qm->dev_name, qm); + if (ret) + goto err_abonormal_irq; + } + } + + return 0; + +err_abonormal_irq: + free_irq(pci_irq_vector(pdev, QM_AEQ_EVENT_IRQ_VECTOR), qm); +err_aeq_irq: + free_irq(pci_irq_vector(pdev, QM_EQ_EVENT_IRQ_VECTOR), qm); + return ret; +} + +static void hisi_qm_controller_reset(struct work_struct *rst_work) +{ + struct hisi_qm *qm = container_of(rst_work, struct hisi_qm, rst_work); + int ret; + + /* reset pcie device controller */ + ret = qm_controller_reset(qm); + if (ret) + dev_err(&qm->pdev->dev, "controller reset failed (%d)\n", ret); + +} + +/** + * hisi_qm_init() - Initialize configures about qm. + * @qm: The qm needing init. + * + * This function init qm, then we can call hisi_qm_start to put qm into work. + */ +int hisi_qm_init(struct hisi_qm *qm) +{ + struct pci_dev *pdev = qm->pdev; + struct device *dev = &pdev->dev; + unsigned int num_vec; + int ret; + + hisi_qm_pre_init(qm); + + ret = qm_alloc_uacce(qm); + if (ret < 0) + dev_warn(&pdev->dev, "fail to alloc uacce (%d)\n", ret); + + ret = pci_enable_device_mem(pdev); + if (ret < 0) { + dev_err(&pdev->dev, "Failed to enable device mem!\n"); + goto err_remove_uacce; + } + + ret = pci_request_mem_regions(pdev, qm->dev_name); + if (ret < 0) { + dev_err(&pdev->dev, "Failed to request mem regions!\n"); + goto err_disable_pcidev; + } + + qm->phys_base = pci_resource_start(pdev, PCI_BAR_2); + qm->phys_size = pci_resource_len(qm->pdev, PCI_BAR_2); + qm->io_base = ioremap(qm->phys_base, qm->phys_size); + if (!qm->io_base) { + ret = -EIO; + goto err_release_mem_regions; + } + + ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64)); + if (ret < 0) + goto err_iounmap; + pci_set_master(pdev); + + if (!qm->ops->get_irq_num) { + ret = -EOPNOTSUPP; + goto err_iounmap; + } + num_vec = qm->ops->get_irq_num(qm); + ret = pci_alloc_irq_vectors(pdev, num_vec, num_vec, PCI_IRQ_MSI); + if (ret < 0) { + dev_err(dev, "Failed to enable MSI vectors!\n"); + goto err_iounmap; + } + + ret = qm_irq_register(qm); + if (ret) + goto err_free_irq_vectors; + + if (qm->fun_type == QM_HW_VF && qm->ver != QM_HW_V1) { + /* v2 starts to support get vft by mailbox */ + ret = hisi_qm_get_vft(qm, &qm->qp_base, &qm->qp_num); + if (ret) + goto err_irq_unregister; + } + + ret = hisi_qm_memory_init(qm); + if (ret) + goto err_irq_unregister; + + INIT_WORK(&qm->work, qm_work_process); + if (qm->fun_type == QM_HW_PF) + INIT_WORK(&qm->rst_work, hisi_qm_controller_reset); + + atomic_set(&qm->status.flags, QM_INIT); + + return 0; + +err_irq_unregister: + qm_irq_unregister(qm); +err_free_irq_vectors: + pci_free_irq_vectors(pdev); +err_iounmap: + iounmap(qm->io_base); +err_release_mem_regions: + pci_release_mem_regions(pdev); +err_disable_pcidev: + pci_disable_device(pdev); +err_remove_uacce: + uacce_remove(qm->uacce); + qm->uacce = NULL; + return ret; +} +EXPORT_SYMBOL_GPL(hisi_qm_init); + + MODULE_LICENSE("GPL v2"); MODULE_AUTHOR("Zhou Wang <wangzhou1@hisilicon.com>"); MODULE_DESCRIPTION("HiSilicon Accelerator queue manager driver"); diff --git a/drivers/crypto/hisilicon/qm.h b/drivers/crypto/hisilicon/qm.h index ec5b6f48db6c..0a351de8d838 100644 --- a/drivers/crypto/hisilicon/qm.h +++ b/drivers/crypto/hisilicon/qm.h @@ -8,6 +8,10 @@ #include <linux/module.h> #include <linux/pci.h> +#define QM_QNUM_V1 4096 +#define QM_QNUM_V2 1024 +#define QM_MAX_VFS_NUM_V2 63 + /* qm user domain */ #define QM_ARUSER_M_CFG_1 0x100088 #define AXUSER_SNOOP_ENABLE BIT(30) @@ -70,7 +74,7 @@ #define QM_BASE_NFE (QM_AXI_RRESP | QM_AXI_BRESP | QM_ECC_MBIT | \ QM_ACC_GET_TASK_TIMEOUT | QM_DB_TIMEOUT | \ - QM_OF_FIFO_OF) + QM_OF_FIFO_OF | QM_DB_RANDOM_INVALID) #define QM_BASE_CE QM_ECC_1BIT #define QM_Q_DEPTH 1024 @@ -80,14 +84,31 @@ /* page number for queue file region */ #define QM_DOORBELL_PAGE_NR 1 +enum qm_stop_reason { + QM_NORMAL, + QM_SOFT_RESET, + QM_FLR, +}; + +enum qm_state { + QM_INIT = 0, + QM_START, + QM_CLOSE, + QM_STOP, +}; + enum qp_state { + QP_INIT = 1, + QP_START, QP_STOP, + QP_CLOSE, }; enum qm_hw_ver { QM_HW_UNKNOWN = -1, QM_HW_V1 = 0x20, QM_HW_V2 = 0x21, + QM_HW_V3 = 0x30, }; enum qm_fun_type { @@ -101,6 +122,14 @@ enum qm_debug_file { DEBUG_FILE_NUM, }; +struct qm_dfx { + atomic64_t err_irq_cnt; + atomic64_t aeq_irq_cnt; + atomic64_t abnormal_irq_cnt; + atomic64_t create_qp_err_cnt; + atomic64_t mb_err_cnt; +}; + struct debugfs_file { enum qm_debug_file index; struct mutex lock; @@ -109,6 +138,9 @@ struct debugfs_file { struct qm_debug { u32 curr_qm_qp_num; + u32 sqe_mask_offset; + u32 sqe_mask_len; + struct qm_dfx dfx; struct dentry *debug_root; struct dentry *qm_d; struct debugfs_file files[DEBUG_FILE_NUM]; @@ -125,22 +157,34 @@ struct hisi_qm_status { bool eqc_phase; u32 aeq_head; bool aeqc_phase; - unsigned long flags; + atomic_t flags; + int stop_reason; }; struct hisi_qm; struct hisi_qm_err_info { + char *acpi_rst; + u32 msi_wr_port; + u32 ecc_2bits_mask; u32 ce; u32 nfe; u32 fe; - u32 msi; +}; + +struct hisi_qm_err_status { + u32 is_qm_ecc_mbit; + u32 is_dev_ecc_mbit; }; struct hisi_qm_err_ini { + int (*hw_init)(struct hisi_qm *qm); void (*hw_err_enable)(struct hisi_qm *qm); void (*hw_err_disable)(struct hisi_qm *qm); u32 (*get_dev_hw_err_status)(struct hisi_qm *qm); + void (*clear_dev_hw_err_status)(struct hisi_qm *qm, u32 err_sts); + void (*open_axi_master_ooo)(struct hisi_qm *qm); + void (*close_axi_master_ooo)(struct hisi_qm *qm); void (*log_dev_hw_err)(struct hisi_qm *qm, u32 err_sts); struct hisi_qm_err_info err_info; }; @@ -161,7 +205,9 @@ struct hisi_qm { u32 qp_num; u32 qp_in_used; u32 ctrl_qp_num; + u32 vfs_num; struct list_head list; + struct hisi_qm_list *qm_list; struct qm_dma qdma; struct qm_sqc *sqc; @@ -175,10 +221,12 @@ struct hisi_qm { struct hisi_qm_status status; const struct hisi_qm_err_ini *err_ini; + struct hisi_qm_err_status err_status; + unsigned long reset_flag; - rwlock_t qps_lock; - unsigned long *qp_bitmap; - struct hisi_qp **qp_array; + struct rw_semaphore qps_lock; + struct idr qp_idr; + struct hisi_qp *qp_array; struct mutex mailbox_lock; @@ -187,13 +235,12 @@ struct hisi_qm { struct qm_debug debug; u32 error_mask; - u32 msi_mask; struct workqueue_struct *wq; struct work_struct work; + struct work_struct rst_work; const char *algs; - bool use_dma_api; bool use_sva; resource_size_t phys_base; resource_size_t phys_size; @@ -205,7 +252,7 @@ struct hisi_qp_status { u16 sq_tail; u16 cq_head; bool cqc_phase; - unsigned long flags; + atomic_t flags; }; struct hisi_qp_ops { @@ -230,10 +277,58 @@ struct hisi_qp { void (*event_cb)(struct hisi_qp *qp); struct hisi_qm *qm; + bool is_resetting; u16 pasid; struct uacce_queue *uacce_q; }; +static inline int q_num_set(const char *val, const struct kernel_param *kp, + unsigned int device) +{ + struct pci_dev *pdev = pci_get_device(PCI_VENDOR_ID_HUAWEI, + device, NULL); + u32 n, q_num; + int ret; + + if (!val) + return -EINVAL; + + if (!pdev) { + q_num = min_t(u32, QM_QNUM_V1, QM_QNUM_V2); + pr_info("No device found currently, suppose queue number is %d\n", + q_num); + } else { + if (pdev->revision == QM_HW_V1) + q_num = QM_QNUM_V1; + else + q_num = QM_QNUM_V2; + } + + ret = kstrtou32(val, 10, &n); + if (ret || !n || n > q_num) + return -EINVAL; + + return param_set_int(val, kp); +} + +static inline int vfs_num_set(const char *val, const struct kernel_param *kp) +{ + u32 n; + int ret; + + if (!val) + return -EINVAL; + + ret = kstrtou32(val, 10, &n); + if (ret < 0) + return ret; + + if (n > QM_MAX_VFS_NUM_V2) + return -EINVAL; + + return param_set_int(val, kp); +} + static inline void hisi_qm_init_list(struct hisi_qm_list *qm_list) { INIT_LIST_HEAD(&qm_list->list); @@ -267,14 +362,19 @@ void hisi_qm_release_qp(struct hisi_qp *qp); int hisi_qp_send(struct hisi_qp *qp, const void *msg); int hisi_qm_get_free_qp_num(struct hisi_qm *qm); int hisi_qm_get_vft(struct hisi_qm *qm, u32 *base, u32 *number); -int hisi_qm_set_vft(struct hisi_qm *qm, u32 fun_num, u32 base, u32 number); int hisi_qm_debug_init(struct hisi_qm *qm); enum qm_hw_ver hisi_qm_get_hw_version(struct pci_dev *pdev); void hisi_qm_debug_regs_clear(struct hisi_qm *qm); +int hisi_qm_sriov_enable(struct pci_dev *pdev, int max_vfs); +int hisi_qm_sriov_disable(struct pci_dev *pdev); +int hisi_qm_sriov_configure(struct pci_dev *pdev, int num_vfs); void hisi_qm_dev_err_init(struct hisi_qm *qm); void hisi_qm_dev_err_uninit(struct hisi_qm *qm); pci_ers_result_t hisi_qm_dev_err_detected(struct pci_dev *pdev, pci_channel_state_t state); +pci_ers_result_t hisi_qm_dev_slot_reset(struct pci_dev *pdev); +void hisi_qm_reset_prepare(struct pci_dev *pdev); +void hisi_qm_reset_done(struct pci_dev *pdev); struct hisi_acc_sgl_pool; struct hisi_acc_hw_sgl *hisi_acc_sg_buf_map_to_hw_sgl(struct device *dev, diff --git a/drivers/crypto/hisilicon/sec2/sec.h b/drivers/crypto/hisilicon/sec2/sec.h index 3598fa17beb2..7b64aca704d6 100644 --- a/drivers/crypto/hisilicon/sec2/sec.h +++ b/drivers/crypto/hisilicon/sec2/sec.h @@ -160,6 +160,10 @@ struct sec_debug_file { struct sec_dfx { atomic64_t send_cnt; atomic64_t recv_cnt; + atomic64_t send_busy_cnt; + atomic64_t err_bd_cnt; + atomic64_t invalid_req_cnt; + atomic64_t done_flag_cnt; }; struct sec_debug { @@ -172,7 +176,6 @@ struct sec_dev { struct sec_debug debug; u32 ctx_q_num; bool iommu_used; - u32 num_vfs; unsigned long status; }; diff --git a/drivers/crypto/hisilicon/sec2/sec_crypto.c b/drivers/crypto/hisilicon/sec2/sec_crypto.c index 7f1c6a31b82f..64614a9bdf21 100644 --- a/drivers/crypto/hisilicon/sec2/sec_crypto.c +++ b/drivers/crypto/hisilicon/sec2/sec_crypto.c @@ -148,6 +148,7 @@ static int sec_aead_verify(struct sec_req *req) static void sec_req_cb(struct hisi_qp *qp, void *resp) { struct sec_qp_ctx *qp_ctx = qp->qp_ctx; + struct sec_dfx *dfx = &qp_ctx->ctx->sec->debug.dfx; struct sec_sqe *bd = resp; struct sec_ctx *ctx; struct sec_req *req; @@ -157,11 +158,16 @@ static void sec_req_cb(struct hisi_qp *qp, void *resp) type = bd->type_cipher_auth & SEC_TYPE_MASK; if (unlikely(type != SEC_BD_TYPE2)) { + atomic64_inc(&dfx->err_bd_cnt); pr_err("err bd type [%d]\n", type); return; } req = qp_ctx->req_list[le16_to_cpu(bd->type2.tag)]; + if (unlikely(!req)) { + atomic64_inc(&dfx->invalid_req_cnt); + return; + } req->err_type = bd->type2.error_type; ctx = req->ctx; done = le16_to_cpu(bd->type2.done_flag) & SEC_DONE_MASK; @@ -174,12 +180,13 @@ static void sec_req_cb(struct hisi_qp *qp, void *resp) "err_type[%d],done[%d],flag[%d]\n", req->err_type, done, flag); err = -EIO; + atomic64_inc(&dfx->done_flag_cnt); } if (ctx->alg_type == SEC_AEAD && !req->c_req.encrypt) err = sec_aead_verify(req); - atomic64_inc(&ctx->sec->debug.dfx.recv_cnt); + atomic64_inc(&dfx->recv_cnt); ctx->req_op->buf_unmap(ctx, req); @@ -200,10 +207,12 @@ static int sec_bd_send(struct sec_ctx *ctx, struct sec_req *req) return -ENOBUFS; if (!ret) { - if (req->fake_busy) + if (req->fake_busy) { + atomic64_inc(&ctx->sec->debug.dfx.send_busy_cnt); ret = -EBUSY; - else + } else { ret = -EINPROGRESS; + } } return ret; @@ -832,7 +841,6 @@ static int sec_aead_auth_set_key(struct sec_auth_ctx *ctx, struct crypto_authenc_keys *keys) { struct crypto_shash *hash_tfm = ctx->hash_tfm; - SHASH_DESC_ON_STACK(shash, hash_tfm); int blocksize, ret; if (!keys->authkeylen) { @@ -842,8 +850,8 @@ static int sec_aead_auth_set_key(struct sec_auth_ctx *ctx, blocksize = crypto_shash_blocksize(hash_tfm); if (keys->authkeylen > blocksize) { - ret = crypto_shash_digest(shash, keys->authkey, - keys->authkeylen, ctx->a_key); + ret = crypto_shash_tfm_digest(hash_tfm, keys->authkey, + keys->authkeylen, ctx->a_key); if (ret) { pr_err("hisi_sec2: aead auth digest error!\n"); return -EINVAL; diff --git a/drivers/crypto/hisilicon/sec2/sec_main.c b/drivers/crypto/hisilicon/sec2/sec_main.c index 1f54ebe164b6..a4cb58b54b25 100644 --- a/drivers/crypto/hisilicon/sec2/sec_main.c +++ b/drivers/crypto/hisilicon/sec2/sec_main.c @@ -80,6 +80,9 @@ #define SEC_VF_CNT_MASK 0xffffffc0 #define SEC_DBGFS_VAL_MAX_LEN 20 +#define SEC_SQE_MASK_OFFSET 64 +#define SEC_SQE_MASK_LEN 48 + #define SEC_ADDR(qm, offset) ((qm)->io_base + (offset) + \ SEC_ENGINE_PF_CFG_OFF + SEC_ACC_COMMON_REG_OFF) @@ -88,6 +91,11 @@ struct sec_hw_error { const char *msg; }; +struct sec_dfx_item { + const char *name; + u32 offset; +}; + static const char sec_name[] = "hisi_sec2"; static struct dentry *sec_debugfs_root; static struct hisi_qm_list sec_devices; @@ -110,7 +118,16 @@ static const char * const sec_dbg_file_name[] = { [SEC_CLEAR_ENABLE] = "clear_enable", }; -static struct debugfs_reg32 sec_dfx_regs[] = { +static struct sec_dfx_item sec_dfx_labels[] = { + {"send_cnt", offsetof(struct sec_dfx, send_cnt)}, + {"recv_cnt", offsetof(struct sec_dfx, recv_cnt)}, + {"send_busy_cnt", offsetof(struct sec_dfx, send_busy_cnt)}, + {"err_bd_cnt", offsetof(struct sec_dfx, err_bd_cnt)}, + {"invalid_req_cnt", offsetof(struct sec_dfx, invalid_req_cnt)}, + {"done_flag_cnt", offsetof(struct sec_dfx, done_flag_cnt)}, +}; + +static const struct debugfs_reg32 sec_dfx_regs[] = { {"SEC_PF_ABNORMAL_INT_SOURCE ", 0x301010}, {"SEC_SAA_EN ", 0x301270}, {"SEC_BD_LATENCY_MIN ", 0x301600}, @@ -136,45 +153,14 @@ static struct debugfs_reg32 sec_dfx_regs[] = { static int sec_pf_q_num_set(const char *val, const struct kernel_param *kp) { - struct pci_dev *pdev; - u32 n, q_num; - u8 rev_id; - int ret; - - if (!val) - return -EINVAL; - - pdev = pci_get_device(PCI_VENDOR_ID_HUAWEI, - SEC_PF_PCI_DEVICE_ID, NULL); - if (!pdev) { - q_num = min_t(u32, SEC_QUEUE_NUM_V1, SEC_QUEUE_NUM_V2); - pr_info("No device, suppose queue number is %d!\n", q_num); - } else { - rev_id = pdev->revision; - - switch (rev_id) { - case QM_HW_V1: - q_num = SEC_QUEUE_NUM_V1; - break; - case QM_HW_V2: - q_num = SEC_QUEUE_NUM_V2; - break; - default: - return -EINVAL; - } - } - - ret = kstrtou32(val, 10, &n); - if (ret || !n || n > q_num) - return -EINVAL; - - return param_set_int(val, kp); + return q_num_set(val, kp, SEC_PF_PCI_DEVICE_ID); } static const struct kernel_param_ops sec_pf_q_num_ops = { .set = sec_pf_q_num_set, .get = param_get_int, }; + static u32 pf_q_num = SEC_PF_DEF_Q_NUM; module_param_cb(pf_q_num, &sec_pf_q_num_ops, &pf_q_num, 0444); MODULE_PARM_DESC(pf_q_num, "Number of queues in PF(v1 0-4096, v2 0-1024)"); @@ -207,6 +193,15 @@ static u32 ctx_q_num = SEC_CTX_Q_NUM_DEF; module_param_cb(ctx_q_num, &sec_ctx_q_num_ops, &ctx_q_num, 0444); MODULE_PARM_DESC(ctx_q_num, "Queue num in ctx (24 default, 2, 4, ..., 32)"); +static const struct kernel_param_ops vfs_num_ops = { + .set = vfs_num_set, + .get = param_get_int, +}; + +static u32 vfs_num; +module_param_cb(vfs_num, &vfs_num_ops, &vfs_num, 0444); +MODULE_PARM_DESC(vfs_num, "Number of VFs to enable(1-63), 0(default)"); + void sec_destroy_qps(struct hisi_qp **qps, int qp_num) { hisi_qm_free_qps(qps, qp_num); @@ -240,9 +235,8 @@ static const struct pci_device_id sec_dev_ids[] = { }; MODULE_DEVICE_TABLE(pci, sec_dev_ids); -static u8 sec_get_endian(struct sec_dev *sec) +static u8 sec_get_endian(struct hisi_qm *qm) { - struct hisi_qm *qm = &sec->qm; u32 reg; /* @@ -270,9 +264,8 @@ static u8 sec_get_endian(struct sec_dev *sec) return SEC_64BE; } -static int sec_engine_init(struct sec_dev *sec) +static int sec_engine_init(struct hisi_qm *qm) { - struct hisi_qm *qm = &sec->qm; int ret; u32 reg; @@ -315,7 +308,7 @@ static int sec_engine_init(struct sec_dev *sec) /* config endian */ reg = readl_relaxed(SEC_ADDR(qm, SEC_CONTROL_REG)); - reg |= sec_get_endian(sec); + reg |= sec_get_endian(qm); writel_relaxed(reg, SEC_ADDR(qm, SEC_CONTROL_REG)); /* Enable sm4 xts mode multiple iv */ @@ -325,10 +318,8 @@ static int sec_engine_init(struct sec_dev *sec) return 0; } -static int sec_set_user_domain_and_cache(struct sec_dev *sec) +static int sec_set_user_domain_and_cache(struct hisi_qm *qm) { - struct hisi_qm *qm = &sec->qm; - /* qm user domain */ writel(AXUSER_BASE, qm->io_base + QM_ARUSER_M_CFG_1); writel(ARUSER_M_CFG_ENABLE, qm->io_base + QM_ARUSER_M_CFG_ENABLE); @@ -349,7 +340,7 @@ static int sec_set_user_domain_and_cache(struct sec_dev *sec) CQC_CACHE_WB_ENABLE | FIELD_PREP(SQC_CACHE_WB_THRD, 1) | FIELD_PREP(CQC_CACHE_WB_THRD, 1), qm->io_base + QM_CACHE_CTL); - return sec_engine_init(sec); + return sec_engine_init(qm); } /* sec_debug_regs_clear() - clear the sec debug regs */ @@ -424,23 +415,22 @@ static u32 sec_current_qm_read(struct sec_debug_file *file) static int sec_current_qm_write(struct sec_debug_file *file, u32 val) { struct hisi_qm *qm = file->qm; - struct sec_dev *sec = container_of(qm, struct sec_dev, qm); u32 vfq_num; u32 tmp; - if (val > sec->num_vfs) + if (val > qm->vfs_num) return -EINVAL; /* According PF or VF Dev ID to calculation curr_qm_qp_num and store */ if (!val) { qm->debug.curr_qm_qp_num = qm->qp_num; } else { - vfq_num = (qm->ctrl_qp_num - qm->qp_num) / sec->num_vfs; + vfq_num = (qm->ctrl_qp_num - qm->qp_num) / qm->vfs_num; - if (val == sec->num_vfs) + if (val == qm->vfs_num) qm->debug.curr_qm_qp_num = qm->ctrl_qp_num - qm->qp_num - - (sec->num_vfs - 1) * vfq_num; + (qm->vfs_num - 1) * vfq_num; else qm->debug.curr_qm_qp_num = vfq_num; } @@ -570,10 +560,22 @@ static const struct file_operations sec_dbg_fops = { static int sec_debugfs_atomic64_get(void *data, u64 *val) { *val = atomic64_read((atomic64_t *)data); + + return 0; +} + +static int sec_debugfs_atomic64_set(void *data, u64 val) +{ + if (val) + return -EINVAL; + + atomic64_set((atomic64_t *)data, 0); + return 0; } + DEFINE_DEBUGFS_ATTRIBUTE(sec_atomic64_ops, sec_debugfs_atomic64_get, - NULL, "%lld\n"); + sec_debugfs_atomic64_set, "%lld\n"); static int sec_core_debug_init(struct sec_dev *sec) { @@ -582,6 +584,7 @@ static int sec_core_debug_init(struct sec_dev *sec) struct sec_dfx *dfx = &sec->debug.dfx; struct debugfs_regset32 *regset; struct dentry *tmp_d; + int i; tmp_d = debugfs_create_dir("sec_dfx", sec->qm.debug.debug_root); @@ -593,13 +596,15 @@ static int sec_core_debug_init(struct sec_dev *sec) regset->nregs = ARRAY_SIZE(sec_dfx_regs); regset->base = qm->io_base; - debugfs_create_regset32("regs", 0444, tmp_d, regset); - - debugfs_create_file("send_cnt", 0444, tmp_d, - &dfx->send_cnt, &sec_atomic64_ops); + if (qm->pdev->device == SEC_PF_PCI_DEVICE_ID) + debugfs_create_regset32("regs", 0444, tmp_d, regset); - debugfs_create_file("recv_cnt", 0444, tmp_d, - &dfx->recv_cnt, &sec_atomic64_ops); + for (i = 0; i < ARRAY_SIZE(sec_dfx_labels); i++) { + atomic64_t *data = (atomic64_t *)((uintptr_t)dfx + + sec_dfx_labels[i].offset); + debugfs_create_file(sec_dfx_labels[i].name, 0644, + tmp_d, data, &sec_atomic64_ops); + } return 0; } @@ -630,6 +635,9 @@ static int sec_debugfs_init(struct sec_dev *sec) qm->debug.debug_root = debugfs_create_dir(dev_name(dev), sec_debugfs_root); + + qm->debug.sqe_mask_offset = SEC_SQE_MASK_OFFSET; + qm->debug.sqe_mask_len = SEC_SQE_MASK_LEN; ret = hisi_qm_debug_init(qm); if (ret) goto failed_to_create; @@ -675,8 +683,6 @@ static void sec_log_hw_error(struct hisi_qm *qm, u32 err_sts) } errs++; } - - writel(err_sts, qm->io_base + SEC_CORE_INT_SOURCE); } static u32 sec_get_hw_err_status(struct hisi_qm *qm) @@ -684,17 +690,36 @@ static u32 sec_get_hw_err_status(struct hisi_qm *qm) return readl(qm->io_base + SEC_CORE_INT_STATUS); } +static void sec_clear_hw_err_status(struct hisi_qm *qm, u32 err_sts) +{ + writel(err_sts, qm->io_base + SEC_CORE_INT_SOURCE); +} + +static void sec_open_axi_master_ooo(struct hisi_qm *qm) +{ + u32 val; + + val = readl(SEC_ADDR(qm, SEC_CONTROL_REG)); + writel(val & SEC_AXI_SHUTDOWN_DISABLE, SEC_ADDR(qm, SEC_CONTROL_REG)); + writel(val | SEC_AXI_SHUTDOWN_ENABLE, SEC_ADDR(qm, SEC_CONTROL_REG)); +} + static const struct hisi_qm_err_ini sec_err_ini = { + .hw_init = sec_set_user_domain_and_cache, .hw_err_enable = sec_hw_error_enable, .hw_err_disable = sec_hw_error_disable, .get_dev_hw_err_status = sec_get_hw_err_status, + .clear_dev_hw_err_status = sec_clear_hw_err_status, .log_dev_hw_err = sec_log_hw_error, + .open_axi_master_ooo = sec_open_axi_master_ooo, .err_info = { .ce = QM_BASE_CE, .nfe = QM_BASE_NFE | QM_ACC_DO_TASK_TIMEOUT | QM_ACC_WB_NOT_READY_TIMEOUT, .fe = 0, - .msi = QM_DB_RANDOM_INVALID, + .ecc_2bits_mask = SEC_CORE_INT_STATUS_M_ECC, + .msi_wr_port = BIT(0), + .acpi_rst = "SRST", } }; @@ -703,22 +728,14 @@ static int sec_pf_probe_init(struct sec_dev *sec) struct hisi_qm *qm = &sec->qm; int ret; - switch (qm->ver) { - case QM_HW_V1: + if (qm->ver == QM_HW_V1) qm->ctrl_qp_num = SEC_QUEUE_NUM_V1; - break; - - case QM_HW_V2: + else qm->ctrl_qp_num = SEC_QUEUE_NUM_V2; - break; - - default: - return -EINVAL; - } qm->err_ini = &sec_err_ini; - ret = sec_set_user_domain_and_cache(sec); + ret = sec_set_user_domain_and_cache(qm); if (ret) return ret; @@ -730,32 +747,30 @@ static int sec_pf_probe_init(struct sec_dev *sec) static int sec_qm_init(struct hisi_qm *qm, struct pci_dev *pdev) { - enum qm_hw_ver rev_id; - - rev_id = hisi_qm_get_hw_version(pdev); - if (rev_id == QM_HW_UNKNOWN) - return -ENODEV; + int ret; qm->pdev = pdev; - qm->ver = rev_id; - + qm->ver = pdev->revision; qm->sqe_size = SEC_SQE_SIZE; qm->dev_name = sec_name; + qm->fun_type = (pdev->device == SEC_PF_PCI_DEVICE_ID) ? QM_HW_PF : QM_HW_VF; - qm->use_dma_api = true; - - return hisi_qm_init(qm); -} - -static void sec_qm_uninit(struct hisi_qm *qm) -{ - hisi_qm_uninit(qm); -} - -static int sec_probe_init(struct hisi_qm *qm, struct sec_dev *sec) -{ - int ret; + if (qm->fun_type == QM_HW_PF) { + qm->qp_base = SEC_PF_DEF_Q_BASE; + qm->qp_num = pf_q_num; + qm->debug.curr_qm_qp_num = pf_q_num; + qm->qm_list = &sec_devices; + } else if (qm->fun_type == QM_HW_VF && qm->ver == QM_HW_V1) { + /* + * have no way to get qm configure in VM in v1 hardware, + * so currently force PF to uses SEC_PF_DEF_Q_NUM, and force + * to trigger only one VF in v1 hardware. + * v2 hardware has no such problem. + */ + qm->qp_base = SEC_PF_DEF_Q_NUM; + qm->qp_num = SEC_QUEUE_NUM_V1 - SEC_PF_DEF_Q_NUM; + } /* * WQ_HIGHPRI: SEC request must be low delayed, @@ -763,47 +778,38 @@ static int sec_probe_init(struct hisi_qm *qm, struct sec_dev *sec) * WQ_UNBOUND: SEC task is likely with long * running CPU intensive workloads. */ - qm->wq = alloc_workqueue("%s", WQ_HIGHPRI | - WQ_MEM_RECLAIM | WQ_UNBOUND, num_online_cpus(), - pci_name(qm->pdev)); + qm->wq = alloc_workqueue("%s", WQ_HIGHPRI | WQ_MEM_RECLAIM | + WQ_UNBOUND, num_online_cpus(), + pci_name(qm->pdev)); if (!qm->wq) { pci_err(qm->pdev, "fail to alloc workqueue\n"); return -ENOMEM; } - if (qm->fun_type == QM_HW_PF) { - qm->qp_base = SEC_PF_DEF_Q_BASE; - qm->qp_num = pf_q_num; - qm->debug.curr_qm_qp_num = pf_q_num; + ret = hisi_qm_init(qm); + if (ret) + destroy_workqueue(qm->wq); + + return ret; +} +static void sec_qm_uninit(struct hisi_qm *qm) +{ + hisi_qm_uninit(qm); +} + +static int sec_probe_init(struct sec_dev *sec) +{ + struct hisi_qm *qm = &sec->qm; + int ret; + + if (qm->fun_type == QM_HW_PF) { ret = sec_pf_probe_init(sec); if (ret) - goto err_probe_uninit; - } else if (qm->fun_type == QM_HW_VF) { - /* - * have no way to get qm configure in VM in v1 hardware, - * so currently force PF to uses SEC_PF_DEF_Q_NUM, and force - * to trigger only one VF in v1 hardware. - * v2 hardware has no such problem. - */ - if (qm->ver == QM_HW_V1) { - qm->qp_base = SEC_PF_DEF_Q_NUM; - qm->qp_num = SEC_QUEUE_NUM_V1 - SEC_PF_DEF_Q_NUM; - } else if (qm->ver == QM_HW_V2) { - /* v2 starts to support get vft by mailbox */ - ret = hisi_qm_get_vft(qm, &qm->qp_base, &qm->qp_num); - if (ret) - goto err_probe_uninit; - } - } else { - ret = -ENODEV; - goto err_probe_uninit; + return ret; } return 0; -err_probe_uninit: - destroy_workqueue(qm->wq); - return ret; } static void sec_probe_uninit(struct hisi_qm *qm) @@ -840,20 +846,17 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (!sec) return -ENOMEM; - pci_set_drvdata(pdev, sec); - - sec->ctx_q_num = ctx_q_num; - sec_iommu_used_check(sec); - qm = &sec->qm; - ret = sec_qm_init(qm, pdev); if (ret) { - pci_err(pdev, "Failed to pre init qm!\n"); + pci_err(pdev, "Failed to init SEC QM (%d)!\n", ret); return ret; } - ret = sec_probe_init(qm, sec); + sec->ctx_q_num = ctx_q_num; + sec_iommu_used_check(sec); + + ret = sec_probe_init(sec); if (ret) { pci_err(pdev, "Failed to probe!\n"); goto err_qm_uninit; @@ -877,8 +880,17 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto err_remove_from_list; } + if (qm->fun_type == QM_HW_PF && vfs_num) { + ret = hisi_qm_sriov_enable(pdev, vfs_num); + if (ret < 0) + goto err_crypto_unregister; + } + return 0; +err_crypto_unregister: + sec_unregister_from_crypto(); + err_remove_from_list: hisi_qm_del_from_list(qm, &sec_devices); sec_debugfs_exit(sec); @@ -893,110 +905,6 @@ err_qm_uninit: return ret; } -/* now we only support equal assignment */ -static int sec_vf_q_assign(struct sec_dev *sec, u32 num_vfs) -{ - struct hisi_qm *qm = &sec->qm; - u32 qp_num = qm->qp_num; - u32 q_base = qp_num; - u32 q_num, remain_q_num; - int i, j, ret; - - if (!num_vfs) - return -EINVAL; - - remain_q_num = qm->ctrl_qp_num - qp_num; - q_num = remain_q_num / num_vfs; - - for (i = 1; i <= num_vfs; i++) { - if (i == num_vfs) - q_num += remain_q_num % num_vfs; - ret = hisi_qm_set_vft(qm, i, q_base, q_num); - if (ret) { - for (j = i; j > 0; j--) - hisi_qm_set_vft(qm, j, 0, 0); - return ret; - } - q_base += q_num; - } - - return 0; -} - -static int sec_clear_vft_config(struct sec_dev *sec) -{ - struct hisi_qm *qm = &sec->qm; - u32 num_vfs = sec->num_vfs; - int ret; - u32 i; - - for (i = 1; i <= num_vfs; i++) { - ret = hisi_qm_set_vft(qm, i, 0, 0); - if (ret) - return ret; - } - - sec->num_vfs = 0; - - return 0; -} - -static int sec_sriov_enable(struct pci_dev *pdev, int max_vfs) -{ - struct sec_dev *sec = pci_get_drvdata(pdev); - int pre_existing_vfs, ret; - u32 num_vfs; - - pre_existing_vfs = pci_num_vf(pdev); - - if (pre_existing_vfs) { - pci_err(pdev, "Can't enable VF. Please disable at first!\n"); - return 0; - } - - num_vfs = min_t(u32, max_vfs, SEC_VF_NUM); - - ret = sec_vf_q_assign(sec, num_vfs); - if (ret) { - pci_err(pdev, "Can't assign queues for VF!\n"); - return ret; - } - - sec->num_vfs = num_vfs; - - ret = pci_enable_sriov(pdev, num_vfs); - if (ret) { - pci_err(pdev, "Can't enable VF!\n"); - sec_clear_vft_config(sec); - return ret; - } - - return num_vfs; -} - -static int sec_sriov_disable(struct pci_dev *pdev) -{ - struct sec_dev *sec = pci_get_drvdata(pdev); - - if (pci_vfs_assigned(pdev)) { - pci_err(pdev, "Can't disable VFs while VFs are assigned!\n"); - return -EPERM; - } - - /* remove in sec_pci_driver will be called to free VF resources */ - pci_disable_sriov(pdev); - - return sec_clear_vft_config(sec); -} - -static int sec_sriov_configure(struct pci_dev *pdev, int num_vfs) -{ - if (num_vfs) - return sec_sriov_enable(pdev, num_vfs); - else - return sec_sriov_disable(pdev); -} - static void sec_remove(struct pci_dev *pdev) { struct sec_dev *sec = pci_get_drvdata(pdev); @@ -1006,8 +914,8 @@ static void sec_remove(struct pci_dev *pdev) hisi_qm_del_from_list(qm, &sec_devices); - if (qm->fun_type == QM_HW_PF && sec->num_vfs) - (void)sec_sriov_disable(pdev); + if (qm->fun_type == QM_HW_PF && qm->vfs_num) + hisi_qm_sriov_disable(pdev); sec_debugfs_exit(sec); @@ -1023,6 +931,9 @@ static void sec_remove(struct pci_dev *pdev) static const struct pci_error_handlers sec_err_handler = { .error_detected = hisi_qm_dev_err_detected, + .slot_reset = hisi_qm_dev_slot_reset, + .reset_prepare = hisi_qm_reset_prepare, + .reset_done = hisi_qm_reset_done, }; static struct pci_driver sec_pci_driver = { @@ -1031,7 +942,7 @@ static struct pci_driver sec_pci_driver = { .probe = sec_probe, .remove = sec_remove, .err_handler = &sec_err_handler, - .sriov_configure = sec_sriov_configure, + .sriov_configure = hisi_qm_sriov_configure, }; static void sec_register_debugfs(void) diff --git a/drivers/crypto/hisilicon/zip/zip.h b/drivers/crypto/hisilicon/zip/zip.h index 82dc6f867171..f3ed4c0e5493 100644 --- a/drivers/crypto/hisilicon/zip/zip.h +++ b/drivers/crypto/hisilicon/zip/zip.h @@ -28,12 +28,20 @@ enum hisi_zip_error_type { HZIP_NC_ERR = 0x0d, }; +struct hisi_zip_dfx { + atomic64_t send_cnt; + atomic64_t recv_cnt; + atomic64_t send_busy_cnt; + atomic64_t err_bd_cnt; +}; + struct hisi_zip_ctrl; struct hisi_zip { struct hisi_qm qm; struct list_head list; struct hisi_zip_ctrl *ctrl; + struct hisi_zip_dfx dfx; }; struct hisi_zip_sqe { diff --git a/drivers/crypto/hisilicon/zip/zip_crypto.c b/drivers/crypto/hisilicon/zip/zip_crypto.c index 369ec3220574..c73707c2e539 100644 --- a/drivers/crypto/hisilicon/zip/zip_crypto.c +++ b/drivers/crypto/hisilicon/zip/zip_crypto.c @@ -64,7 +64,6 @@ struct hisi_zip_req_q { struct hisi_zip_qp_ctx { struct hisi_qp *qp; - struct hisi_zip_sqe zip_sqe; struct hisi_zip_req_q req_q; struct hisi_acc_sgl_pool *sgl_pool; struct hisi_zip *zip_dev; @@ -333,6 +332,7 @@ static void hisi_zip_acomp_cb(struct hisi_qp *qp, void *data) { struct hisi_zip_sqe *sqe = data; struct hisi_zip_qp_ctx *qp_ctx = qp->qp_ctx; + struct hisi_zip_dfx *dfx = &qp_ctx->zip_dev->dfx; struct hisi_zip_req_q *req_q = &qp_ctx->req_q; struct hisi_zip_req *req = req_q->q + sqe->tag; struct acomp_req *acomp_req = req->req; @@ -340,12 +340,14 @@ static void hisi_zip_acomp_cb(struct hisi_qp *qp, void *data) u32 status, dlen, head_size; int err = 0; + atomic64_inc(&dfx->recv_cnt); status = sqe->dw3 & HZIP_BD_STATUS_M; if (status != 0 && status != HZIP_NC_ERR) { dev_err(dev, "%scompress fail in qp%u: %u, output: %u\n", (qp->alg_type == 0) ? "" : "de", qp->qp_id, status, sqe->produced); + atomic64_inc(&dfx->err_bd_cnt); err = -EIO; } dlen = sqe->produced; @@ -484,11 +486,12 @@ static struct hisi_zip_req *hisi_zip_create_req(struct acomp_req *req, static int hisi_zip_do_work(struct hisi_zip_req *req, struct hisi_zip_qp_ctx *qp_ctx) { - struct hisi_zip_sqe *zip_sqe = &qp_ctx->zip_sqe; struct acomp_req *a_req = req->req; struct hisi_qp *qp = qp_ctx->qp; struct device *dev = &qp->qm->pdev->dev; struct hisi_acc_sgl_pool *pool = qp_ctx->sgl_pool; + struct hisi_zip_dfx *dfx = &qp_ctx->zip_dev->dfx; + struct hisi_zip_sqe zip_sqe; dma_addr_t input; dma_addr_t output; int ret; @@ -511,15 +514,18 @@ static int hisi_zip_do_work(struct hisi_zip_req *req, } req->dma_dst = output; - hisi_zip_fill_sqe(zip_sqe, qp->req_type, input, output, a_req->slen, + hisi_zip_fill_sqe(&zip_sqe, qp->req_type, input, output, a_req->slen, a_req->dlen, req->sskip, req->dskip); - hisi_zip_config_buf_type(zip_sqe, HZIP_SGL); - hisi_zip_config_tag(zip_sqe, req->req_id); + hisi_zip_config_buf_type(&zip_sqe, HZIP_SGL); + hisi_zip_config_tag(&zip_sqe, req->req_id); /* send command to start a task */ - ret = hisi_qp_send(qp, zip_sqe); - if (ret < 0) + atomic64_inc(&dfx->send_cnt); + ret = hisi_qp_send(qp, &zip_sqe); + if (ret < 0) { + atomic64_inc(&dfx->send_busy_cnt); goto err_unmap_output; + } return -EINPROGRESS; diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index fcc85d2dbd07..2229a21ae7c8 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -62,6 +62,7 @@ #define HZIP_CORE_INT_SOURCE 0x3010A0 #define HZIP_CORE_INT_MASK_REG 0x3010A4 +#define HZIP_CORE_INT_SET 0x3010A8 #define HZIP_CORE_INT_STATUS 0x3010AC #define HZIP_CORE_INT_STATUS_M_ECC BIT(1) #define HZIP_CORE_SRAM_ECC_ERR_INFO 0x301148 @@ -83,8 +84,13 @@ #define HZIP_SOFT_CTRL_CNT_CLR_CE 0x301000 #define SOFT_CTRL_CNT_CLR_CE_BIT BIT(0) +#define HZIP_SOFT_CTRL_ZIP_CONTROL 0x30100C +#define HZIP_AXI_SHUTDOWN_ENABLE BIT(14) +#define HZIP_WR_PORT BIT(11) #define HZIP_BUF_SIZE 22 +#define HZIP_SQE_MASK_OFFSET 64 +#define HZIP_SQE_MASK_LEN 48 static const char hisi_zip_name[] = "hisi_zip"; static struct dentry *hzip_debugfs_root; @@ -95,6 +101,18 @@ struct hisi_zip_hw_error { const char *msg; }; +struct zip_dfx_item { + const char *name; + u32 offset; +}; + +static struct zip_dfx_item zip_dfx_files[] = { + {"send_cnt", offsetof(struct hisi_zip_dfx, send_cnt)}, + {"recv_cnt", offsetof(struct hisi_zip_dfx, recv_cnt)}, + {"send_busy_cnt", offsetof(struct hisi_zip_dfx, send_busy_cnt)}, + {"err_bd_cnt", offsetof(struct hisi_zip_dfx, err_bd_cnt)}, +}; + static const struct hisi_zip_hw_error zip_hw_error[] = { { .int_msk = BIT(0), .msg = "zip_ecc_1bitt_err" }, { .int_msk = BIT(1), .msg = "zip_ecc_2bit_err" }, @@ -134,7 +152,6 @@ struct ctrl_debug_file { * Just relevant for PF. */ struct hisi_zip_ctrl { - u32 num_vfs; struct hisi_zip *hisi_zip; struct dentry *debug_root; struct ctrl_debug_file files[HZIP_DEBUG_FILE_NUM]; @@ -162,7 +179,7 @@ static const u64 core_offsets[] = { [HZIP_DECOMP_CORE5] = 0x309000, }; -static struct debugfs_reg32 hzip_dfx_regs[] = { +static const struct debugfs_reg32 hzip_dfx_regs[] = { {"HZIP_GET_BD_NUM ", 0x00ull}, {"HZIP_GET_RIGHT_BD ", 0x04ull}, {"HZIP_GET_ERROR_BD ", 0x08ull}, @@ -189,38 +206,7 @@ static struct debugfs_reg32 hzip_dfx_regs[] = { static int pf_q_num_set(const char *val, const struct kernel_param *kp) { - struct pci_dev *pdev = pci_get_device(PCI_VENDOR_ID_HUAWEI, - PCI_DEVICE_ID_ZIP_PF, NULL); - u32 n, q_num; - u8 rev_id; - int ret; - - if (!val) - return -EINVAL; - - if (!pdev) { - q_num = min_t(u32, HZIP_QUEUE_NUM_V1, HZIP_QUEUE_NUM_V2); - pr_info("No device found currently, suppose queue number is %d\n", - q_num); - } else { - rev_id = pdev->revision; - switch (rev_id) { - case QM_HW_V1: - q_num = HZIP_QUEUE_NUM_V1; - break; - case QM_HW_V2: - q_num = HZIP_QUEUE_NUM_V2; - break; - default: - return -EINVAL; - } - } - - ret = kstrtou32(val, 10, &n); - if (ret != 0 || n > q_num || n == 0) - return -EINVAL; - - return param_set_int(val, kp); + return q_num_set(val, kp, PCI_DEVICE_ID_ZIP_PF); } static const struct kernel_param_ops pf_q_num_ops = { @@ -232,9 +218,14 @@ static u32 pf_q_num = HZIP_PF_DEF_Q_NUM; module_param_cb(pf_q_num, &pf_q_num_ops, &pf_q_num, 0444); MODULE_PARM_DESC(pf_q_num, "Number of queues in PF(v1 1-4096, v2 1-1024)"); +static const struct kernel_param_ops vfs_num_ops = { + .set = vfs_num_set, + .get = param_get_int, +}; + static u32 vfs_num; -module_param(vfs_num, uint, 0444); -MODULE_PARM_DESC(vfs_num, "Number of VFs to enable(1-63)"); +module_param_cb(vfs_num, &vfs_num_ops, &vfs_num, 0444); +MODULE_PARM_DESC(vfs_num, "Number of VFs to enable(1-63), 0(default)"); static const struct pci_device_id hisi_zip_dev_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI, PCI_DEVICE_ID_ZIP_PF) }, @@ -250,9 +241,9 @@ int zip_create_qps(struct hisi_qp **qps, int qp_num) return hisi_qm_alloc_qps_node(&zip_devices, qp_num, 0, node, qps); } -static void hisi_zip_set_user_domain_and_cache(struct hisi_zip *hisi_zip) +static int hisi_zip_set_user_domain_and_cache(struct hisi_qm *qm) { - void __iomem *base = hisi_zip->qm.io_base; + void __iomem *base = qm->io_base; /* qm user domain */ writel(AXUSER_BASE, base + QM_ARUSER_M_CFG_1); @@ -279,7 +270,7 @@ static void hisi_zip_set_user_domain_and_cache(struct hisi_zip *hisi_zip) writel(AXUSER_BASE, base + HZIP_SGL_RUSER_32_63); writel(AXUSER_BASE, base + HZIP_BD_WUSER_32_63); - if (hisi_zip->qm.use_sva) { + if (qm->use_sva) { writel(AXUSER_BASE | AXUSER_SSV, base + HZIP_DATA_RUSER_32_63); writel(AXUSER_BASE | AXUSER_SSV, base + HZIP_DATA_WUSER_32_63); } else { @@ -295,10 +286,14 @@ static void hisi_zip_set_user_domain_and_cache(struct hisi_zip *hisi_zip) writel(SQC_CACHE_ENABLE | CQC_CACHE_ENABLE | SQC_CACHE_WB_ENABLE | CQC_CACHE_WB_ENABLE | FIELD_PREP(SQC_CACHE_WB_THRD, 1) | FIELD_PREP(CQC_CACHE_WB_THRD, 1), base + QM_CACHE_CTL); + + return 0; } static void hisi_zip_hw_error_enable(struct hisi_qm *qm) { + u32 val; + if (qm->ver == QM_HW_V1) { writel(HZIP_CORE_INT_MASK_ALL, qm->io_base + HZIP_CORE_INT_MASK_REG); @@ -317,12 +312,24 @@ static void hisi_zip_hw_error_enable(struct hisi_qm *qm) /* enable ZIP hw error interrupts */ writel(0, qm->io_base + HZIP_CORE_INT_MASK_REG); + + /* enable ZIP block master OOO when m-bit error occur */ + val = readl(qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); + val = val | HZIP_AXI_SHUTDOWN_ENABLE; + writel(val, qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); } static void hisi_zip_hw_error_disable(struct hisi_qm *qm) { + u32 val; + /* disable ZIP hw error interrupts */ writel(HZIP_CORE_INT_MASK_ALL, qm->io_base + HZIP_CORE_INT_MASK_REG); + + /* disable ZIP block master OOO when m-bit error occur */ + val = readl(qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); + val = val & ~HZIP_AXI_SHUTDOWN_ENABLE; + writel(val, qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); } static inline struct hisi_qm *file_to_qm(struct ctrl_debug_file *file) @@ -342,21 +349,20 @@ static u32 current_qm_read(struct ctrl_debug_file *file) static int current_qm_write(struct ctrl_debug_file *file, u32 val) { struct hisi_qm *qm = file_to_qm(file); - struct hisi_zip_ctrl *ctrl = file->ctrl; u32 vfq_num; u32 tmp; - if (val > ctrl->num_vfs) + if (val > qm->vfs_num) return -EINVAL; /* Calculate curr_qm_qp_num and store */ if (val == 0) { qm->debug.curr_qm_qp_num = qm->qp_num; } else { - vfq_num = (qm->ctrl_qp_num - qm->qp_num) / ctrl->num_vfs; - if (val == ctrl->num_vfs) + vfq_num = (qm->ctrl_qp_num - qm->qp_num) / qm->vfs_num; + if (val == qm->vfs_num) qm->debug.curr_qm_qp_num = qm->ctrl_qp_num - - qm->qp_num - (ctrl->num_vfs - 1) * vfq_num; + qm->qp_num - (qm->vfs_num - 1) * vfq_num; else qm->debug.curr_qm_qp_num = vfq_num; } @@ -477,6 +483,27 @@ static const struct file_operations ctrl_debug_fops = { .write = ctrl_debug_write, }; + +static int zip_debugfs_atomic64_set(void *data, u64 val) +{ + if (val) + return -EINVAL; + + atomic64_set((atomic64_t *)data, 0); + + return 0; +} + +static int zip_debugfs_atomic64_get(void *data, u64 *val) +{ + *val = atomic64_read((atomic64_t *)data); + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(zip_atomic64_ops, zip_debugfs_atomic64_get, + zip_debugfs_atomic64_set, "%llu\n"); + static int hisi_zip_core_debug_init(struct hisi_zip_ctrl *ctrl) { struct hisi_zip *hisi_zip = ctrl->hisi_zip; @@ -508,6 +535,25 @@ static int hisi_zip_core_debug_init(struct hisi_zip_ctrl *ctrl) return 0; } +static void hisi_zip_dfx_debug_init(struct hisi_qm *qm) +{ + struct hisi_zip *zip = container_of(qm, struct hisi_zip, qm); + struct hisi_zip_dfx *dfx = &zip->dfx; + struct dentry *tmp_dir; + void *data; + int i; + + tmp_dir = debugfs_create_dir("zip_dfx", qm->debug.debug_root); + for (i = 0; i < ARRAY_SIZE(zip_dfx_files); i++) { + data = (atomic64_t *)((uintptr_t)dfx + zip_dfx_files[i].offset); + debugfs_create_file(zip_dfx_files[i].name, + 0644, + tmp_dir, + data, + &zip_atomic64_ops); + } +} + static int hisi_zip_ctrl_debug_init(struct hisi_zip_ctrl *ctrl) { int i; @@ -534,6 +580,8 @@ static int hisi_zip_debugfs_init(struct hisi_zip *hisi_zip) dev_d = debugfs_create_dir(dev_name(dev), hzip_debugfs_root); + qm->debug.sqe_mask_offset = HZIP_SQE_MASK_OFFSET; + qm->debug.sqe_mask_len = HZIP_SQE_MASK_LEN; qm->debug.debug_root = dev_d; ret = hisi_qm_debug_init(qm); if (ret) @@ -546,6 +594,8 @@ static int hisi_zip_debugfs_init(struct hisi_zip *hisi_zip) goto failed_to_create; } + hisi_zip_dfx_debug_init(qm); + return 0; failed_to_create: @@ -598,8 +648,6 @@ static void hisi_zip_log_hw_error(struct hisi_qm *qm, u32 err_sts) } err++; } - - writel(err_sts, qm->io_base + HZIP_CORE_INT_SOURCE); } static u32 hisi_zip_get_hw_err_status(struct hisi_qm *qm) @@ -607,17 +655,55 @@ static u32 hisi_zip_get_hw_err_status(struct hisi_qm *qm) return readl(qm->io_base + HZIP_CORE_INT_STATUS); } +static void hisi_zip_clear_hw_err_status(struct hisi_qm *qm, u32 err_sts) +{ + writel(err_sts, qm->io_base + HZIP_CORE_INT_SOURCE); +} + +static void hisi_zip_open_axi_master_ooo(struct hisi_qm *qm) +{ + u32 val; + + val = readl(qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); + + writel(val & ~HZIP_AXI_SHUTDOWN_ENABLE, + qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); + + writel(val | HZIP_AXI_SHUTDOWN_ENABLE, + qm->io_base + HZIP_SOFT_CTRL_ZIP_CONTROL); +} + +static void hisi_zip_close_axi_master_ooo(struct hisi_qm *qm) +{ + u32 nfe_enb; + + /* Disable ECC Mbit error report. */ + nfe_enb = readl(qm->io_base + HZIP_CORE_INT_RAS_NFE_ENB); + writel(nfe_enb & ~HZIP_CORE_INT_STATUS_M_ECC, + qm->io_base + HZIP_CORE_INT_RAS_NFE_ENB); + + /* Inject zip ECC Mbit error to block master ooo. */ + writel(HZIP_CORE_INT_STATUS_M_ECC, + qm->io_base + HZIP_CORE_INT_SET); +} + static const struct hisi_qm_err_ini hisi_zip_err_ini = { + .hw_init = hisi_zip_set_user_domain_and_cache, .hw_err_enable = hisi_zip_hw_error_enable, .hw_err_disable = hisi_zip_hw_error_disable, .get_dev_hw_err_status = hisi_zip_get_hw_err_status, + .clear_dev_hw_err_status = hisi_zip_clear_hw_err_status, .log_dev_hw_err = hisi_zip_log_hw_error, + .open_axi_master_ooo = hisi_zip_open_axi_master_ooo, + .close_axi_master_ooo = hisi_zip_close_axi_master_ooo, .err_info = { .ce = QM_BASE_CE, .nfe = QM_BASE_NFE | QM_ACC_WB_NOT_READY_TIMEOUT, .fe = 0, - .msi = QM_DB_RANDOM_INVALID, + .ecc_2bits_mask = HZIP_CORE_INT_STATUS_M_ECC, + .msi_wr_port = HZIP_WR_PORT, + .acpi_rst = "ZRST", } }; @@ -633,177 +719,85 @@ static int hisi_zip_pf_probe_init(struct hisi_zip *hisi_zip) hisi_zip->ctrl = ctrl; ctrl->hisi_zip = hisi_zip; - switch (qm->ver) { - case QM_HW_V1: + if (qm->ver == QM_HW_V1) qm->ctrl_qp_num = HZIP_QUEUE_NUM_V1; - break; - - case QM_HW_V2: + else qm->ctrl_qp_num = HZIP_QUEUE_NUM_V2; - break; - - default: - return -EINVAL; - } qm->err_ini = &hisi_zip_err_ini; - hisi_zip_set_user_domain_and_cache(hisi_zip); + hisi_zip_set_user_domain_and_cache(qm); hisi_qm_dev_err_init(qm); hisi_zip_debug_regs_clear(hisi_zip); return 0; } -/* Currently we only support equal assignment */ -static int hisi_zip_vf_q_assign(struct hisi_zip *hisi_zip, int num_vfs) +static int hisi_zip_qm_init(struct hisi_qm *qm, struct pci_dev *pdev) { - struct hisi_qm *qm = &hisi_zip->qm; - u32 qp_num = qm->qp_num; - u32 q_base = qp_num; - u32 q_num, remain_q_num, i; - int ret; - - if (!num_vfs) - return -EINVAL; - - remain_q_num = qm->ctrl_qp_num - qp_num; - if (remain_q_num < num_vfs) - return -EINVAL; + qm->pdev = pdev; + qm->ver = pdev->revision; + qm->algs = "zlib\ngzip"; + qm->sqe_size = HZIP_SQE_SIZE; + qm->dev_name = hisi_zip_name; - q_num = remain_q_num / num_vfs; - for (i = 1; i <= num_vfs; i++) { - if (i == num_vfs) - q_num += remain_q_num % num_vfs; - ret = hisi_qm_set_vft(qm, i, q_base, q_num); - if (ret) - return ret; - q_base += q_num; + qm->fun_type = (pdev->device == PCI_DEVICE_ID_ZIP_PF) ? + QM_HW_PF : QM_HW_VF; + if (qm->fun_type == QM_HW_PF) { + qm->qp_base = HZIP_PF_DEF_Q_BASE; + qm->qp_num = pf_q_num; + qm->qm_list = &zip_devices; + } else if (qm->fun_type == QM_HW_VF && qm->ver == QM_HW_V1) { + /* + * have no way to get qm configure in VM in v1 hardware, + * so currently force PF to uses HZIP_PF_DEF_Q_NUM, and force + * to trigger only one VF in v1 hardware. + * + * v2 hardware has no such problem. + */ + qm->qp_base = HZIP_PF_DEF_Q_NUM; + qm->qp_num = HZIP_QUEUE_NUM_V1 - HZIP_PF_DEF_Q_NUM; } - return 0; + return hisi_qm_init(qm); } -static int hisi_zip_clear_vft_config(struct hisi_zip *hisi_zip) +static int hisi_zip_probe_init(struct hisi_zip *hisi_zip) { - struct hisi_zip_ctrl *ctrl = hisi_zip->ctrl; struct hisi_qm *qm = &hisi_zip->qm; - u32 i, num_vfs = ctrl->num_vfs; int ret; - for (i = 1; i <= num_vfs; i++) { - ret = hisi_qm_set_vft(qm, i, 0, 0); + if (qm->fun_type == QM_HW_PF) { + ret = hisi_zip_pf_probe_init(hisi_zip); if (ret) return ret; } - ctrl->num_vfs = 0; - return 0; } -static int hisi_zip_sriov_enable(struct pci_dev *pdev, int max_vfs) -{ - struct hisi_zip *hisi_zip = pci_get_drvdata(pdev); - int pre_existing_vfs, num_vfs, ret; - - pre_existing_vfs = pci_num_vf(pdev); - - if (pre_existing_vfs) { - dev_err(&pdev->dev, - "Can't enable VF. Please disable pre-enabled VFs!\n"); - return 0; - } - - num_vfs = min_t(int, max_vfs, HZIP_VF_NUM); - - ret = hisi_zip_vf_q_assign(hisi_zip, num_vfs); - if (ret) { - dev_err(&pdev->dev, "Can't assign queues for VF!\n"); - return ret; - } - - hisi_zip->ctrl->num_vfs = num_vfs; - - ret = pci_enable_sriov(pdev, num_vfs); - if (ret) { - dev_err(&pdev->dev, "Can't enable VF!\n"); - hisi_zip_clear_vft_config(hisi_zip); - return ret; - } - - return num_vfs; -} - -static int hisi_zip_sriov_disable(struct pci_dev *pdev) -{ - struct hisi_zip *hisi_zip = pci_get_drvdata(pdev); - - if (pci_vfs_assigned(pdev)) { - dev_err(&pdev->dev, - "Can't disable VFs while VFs are assigned!\n"); - return -EPERM; - } - - /* remove in hisi_zip_pci_driver will be called to free VF resources */ - pci_disable_sriov(pdev); - - return hisi_zip_clear_vft_config(hisi_zip); -} - static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct hisi_zip *hisi_zip; - enum qm_hw_ver rev_id; struct hisi_qm *qm; int ret; - rev_id = hisi_qm_get_hw_version(pdev); - if (rev_id == QM_HW_UNKNOWN) - return -EINVAL; - hisi_zip = devm_kzalloc(&pdev->dev, sizeof(*hisi_zip), GFP_KERNEL); if (!hisi_zip) return -ENOMEM; - pci_set_drvdata(pdev, hisi_zip); qm = &hisi_zip->qm; - qm->use_dma_api = true; - qm->pdev = pdev; - qm->ver = rev_id; - qm->algs = "zlib\ngzip"; - qm->sqe_size = HZIP_SQE_SIZE; - qm->dev_name = hisi_zip_name; - qm->fun_type = (pdev->device == PCI_DEVICE_ID_ZIP_PF) ? QM_HW_PF : - QM_HW_VF; - ret = hisi_qm_init(qm); + ret = hisi_zip_qm_init(qm, pdev); if (ret) { - dev_err(&pdev->dev, "Failed to init qm!\n"); + pci_err(pdev, "Failed to init ZIP QM (%d)!\n", ret); return ret; } - if (qm->fun_type == QM_HW_PF) { - ret = hisi_zip_pf_probe_init(hisi_zip); - if (ret) - return ret; - - qm->qp_base = HZIP_PF_DEF_Q_BASE; - qm->qp_num = pf_q_num; - } else if (qm->fun_type == QM_HW_VF) { - /* - * have no way to get qm configure in VM in v1 hardware, - * so currently force PF to uses HZIP_PF_DEF_Q_NUM, and force - * to trigger only one VF in v1 hardware. - * - * v2 hardware has no such problem. - */ - if (qm->ver == QM_HW_V1) { - qm->qp_base = HZIP_PF_DEF_Q_NUM; - qm->qp_num = HZIP_QUEUE_NUM_V1 - HZIP_PF_DEF_Q_NUM; - } else if (qm->ver == QM_HW_V2) - /* v2 starts to support get vft by mailbox */ - hisi_qm_get_vft(qm, &qm->qp_base, &qm->qp_num); + ret = hisi_zip_probe_init(hisi_zip); + if (ret) { + pci_err(pdev, "Failed to probe (%d)!\n", ret); + goto err_qm_uninit; } ret = hisi_qm_start(qm); @@ -823,7 +817,7 @@ static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) } if (qm->fun_type == QM_HW_PF && vfs_num > 0) { - ret = hisi_zip_sriov_enable(pdev, vfs_num); + ret = hisi_qm_sriov_enable(pdev, vfs_num); if (ret < 0) goto err_remove_from_list; } @@ -836,15 +830,8 @@ err_remove_from_list: hisi_qm_stop(qm); err_qm_uninit: hisi_qm_uninit(qm); - return ret; -} -static int hisi_zip_sriov_configure(struct pci_dev *pdev, int num_vfs) -{ - if (num_vfs == 0) - return hisi_zip_sriov_disable(pdev); - else - return hisi_zip_sriov_enable(pdev, num_vfs); + return ret; } static void hisi_zip_remove(struct pci_dev *pdev) @@ -852,8 +839,8 @@ static void hisi_zip_remove(struct pci_dev *pdev) struct hisi_zip *hisi_zip = pci_get_drvdata(pdev); struct hisi_qm *qm = &hisi_zip->qm; - if (qm->fun_type == QM_HW_PF && hisi_zip->ctrl->num_vfs != 0) - hisi_zip_sriov_disable(pdev); + if (qm->fun_type == QM_HW_PF && qm->vfs_num) + hisi_qm_sriov_disable(pdev); hisi_zip_debugfs_exit(hisi_zip); hisi_qm_stop(qm); @@ -865,6 +852,9 @@ static void hisi_zip_remove(struct pci_dev *pdev) static const struct pci_error_handlers hisi_zip_err_handler = { .error_detected = hisi_qm_dev_err_detected, + .slot_reset = hisi_qm_dev_slot_reset, + .reset_prepare = hisi_qm_reset_prepare, + .reset_done = hisi_qm_reset_done, }; static struct pci_driver hisi_zip_pci_driver = { @@ -873,7 +863,7 @@ static struct pci_driver hisi_zip_pci_driver = { .probe = hisi_zip_probe, .remove = hisi_zip_remove, .sriov_configure = IS_ENABLED(CONFIG_PCI_IOV) ? - hisi_zip_sriov_configure : NULL, + hisi_qm_sriov_configure : NULL, .err_handler = &hisi_zip_err_handler, }; diff --git a/drivers/crypto/marvell/octeontx/otx_cptpf_main.c b/drivers/crypto/marvell/octeontx/otx_cptpf_main.c index 200fb3303db0..34bb3063eb70 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptpf_main.c +++ b/drivers/crypto/marvell/octeontx/otx_cptpf_main.c @@ -79,13 +79,13 @@ static int otx_cpt_device_init(struct otx_cpt_device *cpt) /* Check BIST status */ bist = (u64)otx_cpt_check_bist_status(cpt); if (bist) { - dev_err(dev, "RAM BIST failed with code 0x%llx", bist); + dev_err(dev, "RAM BIST failed with code 0x%llx\n", bist); return -ENODEV; } bist = otx_cpt_check_exe_bist_status(cpt); if (bist) { - dev_err(dev, "Engine BIST failed with code 0x%llx", bist); + dev_err(dev, "Engine BIST failed with code 0x%llx\n", bist); return -ENODEV; } diff --git a/drivers/crypto/marvell/octeontx/otx_cptpf_mbox.c b/drivers/crypto/marvell/octeontx/otx_cptpf_mbox.c index a6774232e9a3..a9e3de65875a 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptpf_mbox.c +++ b/drivers/crypto/marvell/octeontx/otx_cptpf_mbox.c @@ -63,11 +63,11 @@ static void dump_mbox_msg(struct otx_cpt_mbox *mbox_msg, int vf_id) hex_dump_to_buffer(mbox_msg, sizeof(struct otx_cpt_mbox), 16, 8, raw_data_str, OTX_CPT_MAX_MBOX_DATA_STR_SIZE, false); if (vf_id >= 0) - pr_debug("MBOX opcode %s received from VF%d raw_data %s", + pr_debug("MBOX opcode %s received from VF%d raw_data %s\n", get_mbox_opcode_str(mbox_msg->msg), vf_id, raw_data_str); else - pr_debug("MBOX opcode %s received from PF raw_data %s", + pr_debug("MBOX opcode %s received from PF raw_data %s\n", get_mbox_opcode_str(mbox_msg->msg), raw_data_str); } @@ -140,20 +140,20 @@ static int otx_cpt_bind_vq_to_grp(struct otx_cpt_device *cpt, u8 q, u8 grp) struct otx_cpt_ucode *ucode; if (q >= cpt->max_vfs) { - dev_err(dev, "Requested queue %d is > than maximum avail %d", + dev_err(dev, "Requested queue %d is > than maximum avail %d\n", q, cpt->max_vfs); return -EINVAL; } if (grp >= OTX_CPT_MAX_ENGINE_GROUPS) { - dev_err(dev, "Requested group %d is > than maximum avail %d", + dev_err(dev, "Requested group %d is > than maximum avail %d\n", grp, OTX_CPT_MAX_ENGINE_GROUPS); return -EINVAL; } eng_grp = &cpt->eng_grps.grp[grp]; if (!eng_grp->is_enabled) { - dev_err(dev, "Requested engine group %d is disabled", grp); + dev_err(dev, "Requested engine group %d is disabled\n", grp); return -EINVAL; } @@ -212,7 +212,7 @@ static void otx_cpt_handle_mbox_intr(struct otx_cpt_device *cpt, int vf) vftype = otx_cpt_bind_vq_to_grp(cpt, vf, (u8)mbx.data); if ((vftype != OTX_CPT_AE_TYPES) && (vftype != OTX_CPT_SE_TYPES)) { - dev_err(dev, "VF%d binding to eng group %llu failed", + dev_err(dev, "VF%d binding to eng group %llu failed\n", vf, mbx.data); otx_cptpf_mbox_send_nack(cpt, vf, &mbx); } else { diff --git a/drivers/crypto/marvell/octeontx/otx_cptpf_ucode.c b/drivers/crypto/marvell/octeontx/otx_cptpf_ucode.c index d04baa319592..fec8f3b9b112 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptpf_ucode.c +++ b/drivers/crypto/marvell/octeontx/otx_cptpf_ucode.c @@ -62,7 +62,7 @@ static struct otx_cpt_bitmap get_cores_bmap(struct device *dev, int i; if (eng_grp->g->engs_num > OTX_CPT_MAX_ENGINES) { - dev_err(dev, "unsupported number of engines %d on octeontx", + dev_err(dev, "unsupported number of engines %d on octeontx\n", eng_grp->g->engs_num); return bmap; } @@ -78,7 +78,7 @@ static struct otx_cpt_bitmap get_cores_bmap(struct device *dev, } if (!found) - dev_err(dev, "No engines reserved for engine group %d", + dev_err(dev, "No engines reserved for engine group %d\n", eng_grp->idx); return bmap; } @@ -306,7 +306,7 @@ static int process_tar_file(struct device *dev, ucode_size = ntohl(ucode_hdr->code_length) * 2; if (!ucode_size || (size < round_up(ucode_size, 16) + sizeof(struct otx_cpt_ucode_hdr) + OTX_CPT_UCODE_SIGN_LEN)) { - dev_err(dev, "Ucode %s invalid size", filename); + dev_err(dev, "Ucode %s invalid size\n", filename); return -EINVAL; } @@ -379,18 +379,18 @@ static void print_tar_dbg_info(struct tar_arch_info_t *tar_arch, { struct tar_ucode_info_t *curr; - pr_debug("Tar archive filename %s", tar_filename); - pr_debug("Tar archive pointer %p, size %ld", tar_arch->fw->data, + pr_debug("Tar archive filename %s\n", tar_filename); + pr_debug("Tar archive pointer %p, size %ld\n", tar_arch->fw->data, tar_arch->fw->size); list_for_each_entry(curr, &tar_arch->ucodes, list) { - pr_debug("Ucode filename %s", curr->ucode.filename); - pr_debug("Ucode version string %s", curr->ucode.ver_str); - pr_debug("Ucode version %d.%d.%d.%d", + pr_debug("Ucode filename %s\n", curr->ucode.filename); + pr_debug("Ucode version string %s\n", curr->ucode.ver_str); + pr_debug("Ucode version %d.%d.%d.%d\n", curr->ucode.ver_num.nn, curr->ucode.ver_num.xx, curr->ucode.ver_num.yy, curr->ucode.ver_num.zz); - pr_debug("Ucode type (%d) %s", curr->ucode.type, + pr_debug("Ucode type (%d) %s\n", curr->ucode.type, get_ucode_type_str(curr->ucode.type)); - pr_debug("Ucode size %d", curr->ucode.size); + pr_debug("Ucode size %d\n", curr->ucode.size); pr_debug("Ucode ptr %p\n", curr->ucode_ptr); } } @@ -417,14 +417,14 @@ static struct tar_arch_info_t *load_tar_archive(struct device *dev, goto release_tar_arch; if (tar_arch->fw->size < TAR_BLOCK_LEN) { - dev_err(dev, "Invalid tar archive %s ", tar_filename); + dev_err(dev, "Invalid tar archive %s\n", tar_filename); goto release_tar_arch; } tar_size = tar_arch->fw->size; tar_blk = (struct tar_blk_t *) tar_arch->fw->data; if (strncmp(tar_blk->hdr.magic, TAR_MAGIC, TAR_MAGIC_LEN - 1)) { - dev_err(dev, "Unsupported format of tar archive %s", + dev_err(dev, "Unsupported format of tar archive %s\n", tar_filename); goto release_tar_arch; } @@ -437,7 +437,7 @@ static struct tar_arch_info_t *load_tar_archive(struct device *dev, if (tar_offs + cur_size > tar_size || tar_offs + 2*TAR_BLOCK_LEN > tar_size) { - dev_err(dev, "Invalid tar archive %s ", tar_filename); + dev_err(dev, "Invalid tar archive %s\n", tar_filename); goto release_tar_arch; } @@ -458,7 +458,7 @@ static struct tar_arch_info_t *load_tar_archive(struct device *dev, /* Check for the end of the archive */ if (tar_offs + 2*TAR_BLOCK_LEN > tar_size) { - dev_err(dev, "Invalid tar archive %s ", tar_filename); + dev_err(dev, "Invalid tar archive %s\n", tar_filename); goto release_tar_arch; } @@ -563,13 +563,13 @@ static void print_engs_info(struct otx_cpt_eng_grp_info *eng_grp, static void print_ucode_dbg_info(struct otx_cpt_ucode *ucode) { - pr_debug("Ucode info"); - pr_debug("Ucode version string %s", ucode->ver_str); - pr_debug("Ucode version %d.%d.%d.%d", ucode->ver_num.nn, + pr_debug("Ucode info\n"); + pr_debug("Ucode version string %s\n", ucode->ver_str); + pr_debug("Ucode version %d.%d.%d.%d\n", ucode->ver_num.nn, ucode->ver_num.xx, ucode->ver_num.yy, ucode->ver_num.zz); - pr_debug("Ucode type %s", get_ucode_type_str(ucode->type)); - pr_debug("Ucode size %d", ucode->size); - pr_debug("Ucode virt address %16.16llx", (u64)ucode->align_va); + pr_debug("Ucode type %s\n", get_ucode_type_str(ucode->type)); + pr_debug("Ucode size %d\n", ucode->size); + pr_debug("Ucode virt address %16.16llx\n", (u64)ucode->align_va); pr_debug("Ucode phys address %16.16llx\n", ucode->align_dma); } @@ -600,19 +600,19 @@ static void print_dbg_info(struct device *dev, u32 mask[4]; int i, j; - pr_debug("Engine groups global info"); - pr_debug("max SE %d, max AE %d", + pr_debug("Engine groups global info\n"); + pr_debug("max SE %d, max AE %d\n", eng_grps->avail.max_se_cnt, eng_grps->avail.max_ae_cnt); - pr_debug("free SE %d", eng_grps->avail.se_cnt); - pr_debug("free AE %d", eng_grps->avail.ae_cnt); + pr_debug("free SE %d\n", eng_grps->avail.se_cnt); + pr_debug("free AE %d\n", eng_grps->avail.ae_cnt); for (i = 0; i < OTX_CPT_MAX_ENGINE_GROUPS; i++) { grp = &eng_grps->grp[i]; - pr_debug("engine_group%d, state %s", i, grp->is_enabled ? + pr_debug("engine_group%d, state %s\n", i, grp->is_enabled ? "enabled" : "disabled"); if (grp->is_enabled) { mirrored_grp = &eng_grps->grp[grp->mirror.idx]; - pr_debug("Ucode0 filename %s, version %s", + pr_debug("Ucode0 filename %s, version %s\n", grp->mirror.is_ena ? mirrored_grp->ucode[0].filename : grp->ucode[0].filename, @@ -626,18 +626,18 @@ static void print_dbg_info(struct device *dev, if (engs->type) { print_engs_info(grp, engs_info, 2*OTX_CPT_UCODE_NAME_LENGTH, j); - pr_debug("Slot%d: %s", j, engs_info); + pr_debug("Slot%d: %s\n", j, engs_info); bitmap_to_arr32(mask, engs->bmap, eng_grps->engs_num); - pr_debug("Mask: %8.8x %8.8x %8.8x %8.8x", + pr_debug("Mask: %8.8x %8.8x %8.8x %8.8x\n", mask[3], mask[2], mask[1], mask[0]); } else - pr_debug("Slot%d not used", j); + pr_debug("Slot%d not used\n", j); } if (grp->is_enabled) { cpt_print_engines_mask(grp, dev, engs_mask, OTX_CPT_UCODE_NAME_LENGTH); - pr_debug("Cmask: %s", engs_mask); + pr_debug("Cmask: %s\n", engs_mask); } } } @@ -766,7 +766,7 @@ static int check_engines_availability(struct device *dev, if (avail_cnt < req_eng->count) { dev_err(dev, - "Error available %s engines %d < than requested %d", + "Error available %s engines %d < than requested %d\n", get_eng_type_str(req_eng->type), avail_cnt, req_eng->count); return -EBUSY; @@ -867,7 +867,7 @@ static int copy_ucode_to_dma_mem(struct device *dev, OTX_CPT_UCODE_ALIGNMENT, &ucode->dma, GFP_KERNEL); if (!ucode->va) { - dev_err(dev, "Unable to allocate space for microcode"); + dev_err(dev, "Unable to allocate space for microcode\n"); return -ENOMEM; } ucode->align_va = PTR_ALIGN(ucode->va, OTX_CPT_UCODE_ALIGNMENT); @@ -905,15 +905,15 @@ static int ucode_load(struct device *dev, struct otx_cpt_ucode *ucode, ucode->size = ntohl(ucode_hdr->code_length) * 2; if (!ucode->size || (fw->size < round_up(ucode->size, 16) + sizeof(struct otx_cpt_ucode_hdr) + OTX_CPT_UCODE_SIGN_LEN)) { - dev_err(dev, "Ucode %s invalid size", ucode_filename); + dev_err(dev, "Ucode %s invalid size\n", ucode_filename); ret = -EINVAL; goto release_fw; } ret = get_ucode_type(ucode_hdr, &ucode->type); if (ret) { - dev_err(dev, "Microcode %s unknown type 0x%x", ucode->filename, - ucode->type); + dev_err(dev, "Microcode %s unknown type 0x%x\n", + ucode->filename, ucode->type); goto release_fw; } @@ -1083,7 +1083,7 @@ static int eng_grp_update_masks(struct device *dev, break; default: - dev_err(dev, "Invalid engine type %d", engs->type); + dev_err(dev, "Invalid engine type %d\n", engs->type); return -EINVAL; } @@ -1142,13 +1142,14 @@ static int delete_engine_group(struct device *dev, return -EINVAL; if (eng_grp->mirror.ref_count) { - dev_err(dev, "Can't delete engine_group%d as it is used by:", + dev_err(dev, "Can't delete engine_group%d as it is used by engine_group(s):", eng_grp->idx); for (i = 0; i < OTX_CPT_MAX_ENGINE_GROUPS; i++) { if (eng_grp->g->grp[i].mirror.is_ena && eng_grp->g->grp[i].mirror.idx == eng_grp->idx) - dev_err(dev, "engine_group%d", i); + pr_cont(" %d", i); } + pr_cont("\n"); return -EINVAL; } @@ -1182,7 +1183,7 @@ static int validate_1_ucode_scenario(struct device *dev, if (!otx_cpt_uc_supports_eng_type(&eng_grp->ucode[0], engs[i].type)) { dev_err(dev, - "Microcode %s does not support %s engines", + "Microcode %s does not support %s engines\n", eng_grp->ucode[0].filename, get_eng_type_str(engs[i].type)); return -EINVAL; @@ -1220,7 +1221,7 @@ static int create_engine_group(struct device *dev, /* Validate if requested engine types are supported by this device */ for (i = 0; i < engs_cnt; i++) if (!dev_supports_eng_type(eng_grps, engs[i].type)) { - dev_err(dev, "Device does not support %s engines", + dev_err(dev, "Device does not support %s engines\n", get_eng_type_str(engs[i].type)); return -EPERM; } @@ -1228,7 +1229,7 @@ static int create_engine_group(struct device *dev, /* Find engine group which is not used */ eng_grp = find_unused_eng_grp(eng_grps); if (!eng_grp) { - dev_err(dev, "Error all engine groups are being used"); + dev_err(dev, "Error all engine groups are being used\n"); return -ENOSPC; } @@ -1298,11 +1299,11 @@ static int create_engine_group(struct device *dev, eng_grp->is_enabled = true; if (eng_grp->mirror.is_ena) dev_info(dev, - "Engine_group%d: reuse microcode %s from group %d", + "Engine_group%d: reuse microcode %s from group %d\n", eng_grp->idx, mirrored_eng_grp->ucode[0].ver_str, mirrored_eng_grp->idx); else - dev_info(dev, "Engine_group%d: microcode loaded %s", + dev_info(dev, "Engine_group%d: microcode loaded %s\n", eng_grp->idx, eng_grp->ucode[0].ver_str); return 0; @@ -1412,14 +1413,14 @@ static ssize_t ucode_load_store(struct device *dev, } else { if (del_grp_idx < 0 || del_grp_idx >= OTX_CPT_MAX_ENGINE_GROUPS) { - dev_err(dev, "Invalid engine group index %d", + dev_err(dev, "Invalid engine group index %d\n", del_grp_idx); ret = -EINVAL; return ret; } if (!eng_grps->grp[del_grp_idx].is_enabled) { - dev_err(dev, "Error engine_group%d is not configured", + dev_err(dev, "Error engine_group%d is not configured\n", del_grp_idx); ret = -EINVAL; return ret; @@ -1568,7 +1569,7 @@ void otx_cpt_disable_all_cores(struct otx_cpt_device *cpt) udelay(CSR_DELAY); reg = readq(cpt->reg_base + OTX_CPT_PF_EXEC_BUSY); if (timeout--) { - dev_warn(&cpt->pdev->dev, "Cores still busy"); + dev_warn(&cpt->pdev->dev, "Cores still busy\n"); break; } } @@ -1626,7 +1627,7 @@ int otx_cpt_init_eng_grps(struct pci_dev *pdev, eng_grps->avail.max_ae_cnt; if (eng_grps->engs_num > OTX_CPT_MAX_ENGINES) { dev_err(&pdev->dev, - "Number of engines %d > than max supported %d", + "Number of engines %d > than max supported %d\n", eng_grps->engs_num, OTX_CPT_MAX_ENGINES); ret = -EINVAL; goto err; diff --git a/drivers/crypto/marvell/octeontx/otx_cptvf_algs.c b/drivers/crypto/marvell/octeontx/otx_cptvf_algs.c index 06202bcffb33..60e744f680d3 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptvf_algs.c +++ b/drivers/crypto/marvell/octeontx/otx_cptvf_algs.c @@ -1660,7 +1660,7 @@ int otx_cpt_crypto_init(struct pci_dev *pdev, struct module *mod, case OTX_CPT_SE_TYPES: count = atomic_read(&se_devices.count); if (count >= CPT_MAX_VF_NUM) { - dev_err(&pdev->dev, "No space to add a new device"); + dev_err(&pdev->dev, "No space to add a new device\n"); ret = -ENOSPC; goto err; } @@ -1687,7 +1687,7 @@ int otx_cpt_crypto_init(struct pci_dev *pdev, struct module *mod, case OTX_CPT_AE_TYPES: count = atomic_read(&ae_devices.count); if (count >= CPT_MAX_VF_NUM) { - dev_err(&pdev->dev, "No space to a add new device"); + dev_err(&pdev->dev, "No space to a add new device\n"); ret = -ENOSPC; goto err; } @@ -1728,7 +1728,7 @@ void otx_cpt_crypto_exit(struct pci_dev *pdev, struct module *mod, } if (!dev_found) { - dev_err(&pdev->dev, "%s device not found", __func__); + dev_err(&pdev->dev, "%s device not found\n", __func__); goto exit; } diff --git a/drivers/crypto/marvell/octeontx/otx_cptvf_main.c b/drivers/crypto/marvell/octeontx/otx_cptvf_main.c index a91860b5dc77..ce3168327a39 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptvf_main.c +++ b/drivers/crypto/marvell/octeontx/otx_cptvf_main.c @@ -584,7 +584,7 @@ static irqreturn_t cptvf_done_intr_handler(int __always_unused irq, cptvf_write_vq_done_ack(cptvf, intr); wqe = get_cptvf_vq_wqe(cptvf, 0); if (unlikely(!wqe)) { - dev_err(&pdev->dev, "No work to schedule for VF (%d)", + dev_err(&pdev->dev, "No work to schedule for VF (%d)\n", cptvf->vfid); return IRQ_NONE; } @@ -602,7 +602,7 @@ static void cptvf_set_irq_affinity(struct otx_cptvf *cptvf, int vec) if (!zalloc_cpumask_var(&cptvf->affinity_mask[vec], GFP_KERNEL)) { dev_err(&pdev->dev, - "Allocation failed for affinity_mask for VF %d", + "Allocation failed for affinity_mask for VF %d\n", cptvf->vfid); return; } @@ -691,7 +691,7 @@ static ssize_t vf_engine_group_store(struct device *dev, return -EINVAL; if (val >= OTX_CPT_MAX_ENGINE_GROUPS) { - dev_err(dev, "Engine group >= than max available groups %d", + dev_err(dev, "Engine group >= than max available groups %d\n", OTX_CPT_MAX_ENGINE_GROUPS); return -EINVAL; } @@ -837,7 +837,7 @@ static int otx_cptvf_probe(struct pci_dev *pdev, cptvf_misc_intr_handler, 0, "CPT VF misc intr", cptvf); if (err) { - dev_err(dev, "Failed to request misc irq"); + dev_err(dev, "Failed to request misc irq\n"); goto free_vectors; } @@ -854,7 +854,7 @@ static int otx_cptvf_probe(struct pci_dev *pdev, cptvf->cqinfo.qchunksize = OTX_CPT_CMD_QCHUNK_SIZE; err = cptvf_sw_init(cptvf, OTX_CPT_CMD_QLEN, OTX_CPT_NUM_QS_PER_VF); if (err) { - dev_err(dev, "cptvf_sw_init() failed"); + dev_err(dev, "cptvf_sw_init() failed\n"); goto free_misc_irq; } /* Convey VQ LEN to PF */ @@ -946,7 +946,7 @@ static void otx_cptvf_remove(struct pci_dev *pdev) /* Convey DOWN to PF */ if (otx_cptvf_send_vf_down(cptvf)) { - dev_err(&pdev->dev, "PF not responding to DOWN msg"); + dev_err(&pdev->dev, "PF not responding to DOWN msg\n"); } else { sysfs_remove_group(&pdev->dev.kobj, &otx_cptvf_sysfs_group); otx_cpt_crypto_exit(pdev, THIS_MODULE, cptvf->vftype); diff --git a/drivers/crypto/marvell/octeontx/otx_cptvf_reqmgr.c b/drivers/crypto/marvell/octeontx/otx_cptvf_reqmgr.c index df839b880354..239195cccf93 100644 --- a/drivers/crypto/marvell/octeontx/otx_cptvf_reqmgr.c +++ b/drivers/crypto/marvell/octeontx/otx_cptvf_reqmgr.c @@ -314,7 +314,7 @@ static int process_request(struct pci_dev *pdev, struct otx_cpt_req_info *req, GFP_ATOMIC; ret = setup_sgio_list(pdev, &info, req, gfp); if (unlikely(ret)) { - dev_err(&pdev->dev, "Setting up SG list failed"); + dev_err(&pdev->dev, "Setting up SG list failed\n"); goto request_cleanup; } cpt_req->dlen = info->dlen; @@ -410,17 +410,17 @@ int otx_cpt_do_request(struct pci_dev *pdev, struct otx_cpt_req_info *req, struct otx_cptvf *cptvf = pci_get_drvdata(pdev); if (!otx_cpt_device_ready(cptvf)) { - dev_err(&pdev->dev, "CPT Device is not ready"); + dev_err(&pdev->dev, "CPT Device is not ready\n"); return -ENODEV; } if ((cptvf->vftype == OTX_CPT_SE_TYPES) && (!req->ctrl.s.se_req)) { - dev_err(&pdev->dev, "CPTVF-%d of SE TYPE got AE request", + dev_err(&pdev->dev, "CPTVF-%d of SE TYPE got AE request\n", cptvf->vfid); return -EINVAL; } else if ((cptvf->vftype == OTX_CPT_AE_TYPES) && (req->ctrl.s.se_req)) { - dev_err(&pdev->dev, "CPTVF-%d of AE TYPE got SE request", + dev_err(&pdev->dev, "CPTVF-%d of AE TYPE got SE request\n", cptvf->vfid); return -EINVAL; } @@ -461,7 +461,7 @@ static int cpt_process_ccode(struct pci_dev *pdev, /* check for timeout */ if (time_after_eq(jiffies, cpt_info->time_in + OTX_CPT_COMMAND_TIMEOUT * HZ)) - dev_warn(&pdev->dev, "Request timed out 0x%p", req); + dev_warn(&pdev->dev, "Request timed out 0x%p\n", req); else if (cpt_info->extra_time < OTX_CPT_TIME_IN_RESET_COUNT) { cpt_info->time_in = jiffies; cpt_info->extra_time++; diff --git a/drivers/crypto/mediatek/mtk-sha.c b/drivers/crypto/mediatek/mtk-sha.c index bd6309e57ab8..da3f0b8814aa 100644 --- a/drivers/crypto/mediatek/mtk-sha.c +++ b/drivers/crypto/mediatek/mtk-sha.c @@ -805,12 +805,9 @@ static int mtk_sha_setkey(struct crypto_ahash *tfm, const u8 *key, size_t ds = crypto_shash_digestsize(bctx->shash); int err, i; - SHASH_DESC_ON_STACK(shash, bctx->shash); - - shash->tfm = bctx->shash; - if (keylen > bs) { - err = crypto_shash_digest(shash, key, keylen, bctx->ipad); + err = crypto_shash_tfm_digest(bctx->shash, key, keylen, + bctx->ipad); if (err) return err; keylen = ds; diff --git a/drivers/crypto/n2_core.c b/drivers/crypto/n2_core.c index f5c468f2cc82..6a828bbecea4 100644 --- a/drivers/crypto/n2_core.c +++ b/drivers/crypto/n2_core.c @@ -462,7 +462,6 @@ static int n2_hmac_async_setkey(struct crypto_ahash *tfm, const u8 *key, struct n2_hmac_ctx *ctx = crypto_ahash_ctx(tfm); struct crypto_shash *child_shash = ctx->child_shash; struct crypto_ahash *fallback_tfm; - SHASH_DESC_ON_STACK(shash, child_shash); int err, bs, ds; fallback_tfm = ctx->base.fallback_tfm; @@ -470,14 +469,12 @@ static int n2_hmac_async_setkey(struct crypto_ahash *tfm, const u8 *key, if (err) return err; - shash->tfm = child_shash; - bs = crypto_shash_blocksize(child_shash); ds = crypto_shash_digestsize(child_shash); BUG_ON(ds > N2_HASH_KEY_MAX); if (keylen > bs) { - err = crypto_shash_digest(shash, key, keylen, - ctx->hash_key); + err = crypto_shash_tfm_digest(child_shash, key, keylen, + ctx->hash_key); if (err) return err; keylen = ds; diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c index e4072cd38585..063ad5d03f33 100644 --- a/drivers/crypto/omap-sham.c +++ b/drivers/crypto/omap-sham.c @@ -33,7 +33,6 @@ #include <linux/of_irq.h> #include <linux/delay.h> #include <linux/crypto.h> -#include <linux/cryptohash.h> #include <crypto/scatterwalk.h> #include <crypto/algapi.h> #include <crypto/sha.h> @@ -1245,16 +1244,6 @@ static int omap_sham_update(struct ahash_request *req) return omap_sham_enqueue(req, OP_UPDATE); } -static int omap_sham_shash_digest(struct crypto_shash *tfm, u32 flags, - const u8 *data, unsigned int len, u8 *out) -{ - SHASH_DESC_ON_STACK(shash, tfm); - - shash->tfm = tfm; - - return crypto_shash_digest(shash, data, len, out); -} - static int omap_sham_final_shash(struct ahash_request *req) { struct omap_sham_ctx *tctx = crypto_tfm_ctx(req->base.tfm); @@ -1270,9 +1259,8 @@ static int omap_sham_final_shash(struct ahash_request *req) !test_bit(FLAGS_AUTO_XOR, &ctx->dd->flags)) offset = get_block_size(ctx); - return omap_sham_shash_digest(tctx->fallback, req->base.flags, - ctx->buffer + offset, - ctx->bufcnt - offset, req->result); + return crypto_shash_tfm_digest(tctx->fallback, ctx->buffer + offset, + ctx->bufcnt - offset, req->result); } static int omap_sham_final(struct ahash_request *req) @@ -1351,9 +1339,8 @@ static int omap_sham_setkey(struct crypto_ahash *tfm, const u8 *key, return err; if (keylen > bs) { - err = omap_sham_shash_digest(bctx->shash, - crypto_shash_get_flags(bctx->shash), - key, keylen, bctx->ipad); + err = crypto_shash_tfm_digest(bctx->shash, key, keylen, + bctx->ipad); if (err) return err; keylen = ds; diff --git a/drivers/crypto/s5p-sss.c b/drivers/crypto/s5p-sss.c index 2a16800d2579..341433fbcc4a 100644 --- a/drivers/crypto/s5p-sss.c +++ b/drivers/crypto/s5p-sss.c @@ -1521,37 +1521,6 @@ static int s5p_hash_update(struct ahash_request *req) } /** - * s5p_hash_shash_digest() - calculate shash digest - * @tfm: crypto transformation - * @flags: tfm flags - * @data: input data - * @len: length of data - * @out: output buffer - */ -static int s5p_hash_shash_digest(struct crypto_shash *tfm, u32 flags, - const u8 *data, unsigned int len, u8 *out) -{ - SHASH_DESC_ON_STACK(shash, tfm); - - shash->tfm = tfm; - - return crypto_shash_digest(shash, data, len, out); -} - -/** - * s5p_hash_final_shash() - calculate shash digest - * @req: AHASH request - */ -static int s5p_hash_final_shash(struct ahash_request *req) -{ - struct s5p_hash_ctx *tctx = crypto_tfm_ctx(req->base.tfm); - struct s5p_hash_reqctx *ctx = ahash_request_ctx(req); - - return s5p_hash_shash_digest(tctx->fallback, req->base.flags, - ctx->buffer, ctx->bufcnt, req->result); -} - -/** * s5p_hash_final() - close up hash and calculate digest * @req: AHASH request * @@ -1582,8 +1551,12 @@ static int s5p_hash_final(struct ahash_request *req) if (ctx->error) return -EINVAL; /* uncompleted hash is not needed */ - if (!ctx->digcnt && ctx->bufcnt < BUFLEN) - return s5p_hash_final_shash(req); + if (!ctx->digcnt && ctx->bufcnt < BUFLEN) { + struct s5p_hash_ctx *tctx = crypto_tfm_ctx(req->base.tfm); + + return crypto_shash_tfm_digest(tctx->fallback, ctx->buffer, + ctx->bufcnt, req->result); + } return s5p_hash_enqueue(req, false); /* HASH_OP_FINAL */ } diff --git a/drivers/crypto/stm32/stm32-crc32.c b/drivers/crypto/stm32/stm32-crc32.c index 8e92e4ac79f1..3ba41148c2a4 100644 --- a/drivers/crypto/stm32/stm32-crc32.c +++ b/drivers/crypto/stm32/stm32-crc32.c @@ -28,18 +28,23 @@ /* Registers values */ #define CRC_CR_RESET BIT(0) -#define CRC_CR_REVERSE (BIT(7) | BIT(6) | BIT(5)) -#define CRC_INIT_DEFAULT 0xFFFFFFFF +#define CRC_CR_REV_IN_WORD (BIT(6) | BIT(5)) +#define CRC_CR_REV_IN_BYTE BIT(5) +#define CRC_CR_REV_OUT BIT(7) +#define CRC32C_INIT_DEFAULT 0xFFFFFFFF #define CRC_AUTOSUSPEND_DELAY 50 +static unsigned int burst_size; +module_param(burst_size, uint, 0644); +MODULE_PARM_DESC(burst_size, "Select burst byte size (0 unlimited)"); + struct stm32_crc { struct list_head list; struct device *dev; void __iomem *regs; struct clk *clk; - u8 pending_data[sizeof(u32)]; - size_t nb_pending_bytes; + spinlock_t lock; }; struct stm32_crc_list { @@ -59,14 +64,13 @@ struct stm32_crc_ctx { struct stm32_crc_desc_ctx { u32 partial; /* crc32c: partial in first 4 bytes of that struct */ - struct stm32_crc *crc; }; static int stm32_crc32_cra_init(struct crypto_tfm *tfm) { struct stm32_crc_ctx *mctx = crypto_tfm_ctx(tfm); - mctx->key = CRC_INIT_DEFAULT; + mctx->key = 0; mctx->poly = CRC32_POLY_LE; return 0; } @@ -75,7 +79,7 @@ static int stm32_crc32c_cra_init(struct crypto_tfm *tfm) { struct stm32_crc_ctx *mctx = crypto_tfm_ctx(tfm); - mctx->key = CRC_INIT_DEFAULT; + mctx->key = CRC32C_INIT_DEFAULT; mctx->poly = CRC32C_POLY_LE; return 0; } @@ -92,87 +96,135 @@ static int stm32_crc_setkey(struct crypto_shash *tfm, const u8 *key, return 0; } +static struct stm32_crc *stm32_crc_get_next_crc(void) +{ + struct stm32_crc *crc; + + spin_lock_bh(&crc_list.lock); + crc = list_first_entry(&crc_list.dev_list, struct stm32_crc, list); + if (crc) + list_move_tail(&crc->list, &crc_list.dev_list); + spin_unlock_bh(&crc_list.lock); + + return crc; +} + static int stm32_crc_init(struct shash_desc *desc) { struct stm32_crc_desc_ctx *ctx = shash_desc_ctx(desc); struct stm32_crc_ctx *mctx = crypto_shash_ctx(desc->tfm); struct stm32_crc *crc; + unsigned long flags; - spin_lock_bh(&crc_list.lock); - list_for_each_entry(crc, &crc_list.dev_list, list) { - ctx->crc = crc; - break; - } - spin_unlock_bh(&crc_list.lock); + crc = stm32_crc_get_next_crc(); + if (!crc) + return -ENODEV; + + pm_runtime_get_sync(crc->dev); - pm_runtime_get_sync(ctx->crc->dev); + spin_lock_irqsave(&crc->lock, flags); /* Reset, set key, poly and configure in bit reverse mode */ - writel_relaxed(bitrev32(mctx->key), ctx->crc->regs + CRC_INIT); - writel_relaxed(bitrev32(mctx->poly), ctx->crc->regs + CRC_POL); - writel_relaxed(CRC_CR_RESET | CRC_CR_REVERSE, ctx->crc->regs + CRC_CR); + writel_relaxed(bitrev32(mctx->key), crc->regs + CRC_INIT); + writel_relaxed(bitrev32(mctx->poly), crc->regs + CRC_POL); + writel_relaxed(CRC_CR_RESET | CRC_CR_REV_IN_WORD | CRC_CR_REV_OUT, + crc->regs + CRC_CR); /* Store partial result */ - ctx->partial = readl_relaxed(ctx->crc->regs + CRC_DR); - ctx->crc->nb_pending_bytes = 0; + ctx->partial = readl_relaxed(crc->regs + CRC_DR); - pm_runtime_mark_last_busy(ctx->crc->dev); - pm_runtime_put_autosuspend(ctx->crc->dev); + spin_unlock_irqrestore(&crc->lock, flags); + + pm_runtime_mark_last_busy(crc->dev); + pm_runtime_put_autosuspend(crc->dev); return 0; } -static int stm32_crc_update(struct shash_desc *desc, const u8 *d8, - unsigned int length) +static int burst_update(struct shash_desc *desc, const u8 *d8, + size_t length) { struct stm32_crc_desc_ctx *ctx = shash_desc_ctx(desc); - struct stm32_crc *crc = ctx->crc; - u32 *d32; - unsigned int i; + struct stm32_crc_ctx *mctx = crypto_shash_ctx(desc->tfm); + struct stm32_crc *crc; + unsigned long flags; + + crc = stm32_crc_get_next_crc(); + if (!crc) + return -ENODEV; pm_runtime_get_sync(crc->dev); - if (unlikely(crc->nb_pending_bytes)) { - while (crc->nb_pending_bytes != sizeof(u32) && length) { - /* Fill in pending data */ - crc->pending_data[crc->nb_pending_bytes++] = *(d8++); + spin_lock_irqsave(&crc->lock, flags); + + /* + * Restore previously calculated CRC for this context as init value + * Restore polynomial configuration + * Configure in register for word input data, + * Configure out register in reversed bit mode data. + */ + writel_relaxed(bitrev32(ctx->partial), crc->regs + CRC_INIT); + writel_relaxed(bitrev32(mctx->poly), crc->regs + CRC_POL); + writel_relaxed(CRC_CR_RESET | CRC_CR_REV_IN_WORD | CRC_CR_REV_OUT, + crc->regs + CRC_CR); + + if (d8 != PTR_ALIGN(d8, sizeof(u32))) { + /* Configure for byte data */ + writel_relaxed(CRC_CR_REV_IN_BYTE | CRC_CR_REV_OUT, + crc->regs + CRC_CR); + while (d8 != PTR_ALIGN(d8, sizeof(u32)) && length) { + writeb_relaxed(*d8++, crc->regs + CRC_DR); length--; } - - if (crc->nb_pending_bytes == sizeof(u32)) { - /* Process completed pending data */ - writel_relaxed(*(u32 *)crc->pending_data, - crc->regs + CRC_DR); - crc->nb_pending_bytes = 0; - } + /* Configure for word data */ + writel_relaxed(CRC_CR_REV_IN_WORD | CRC_CR_REV_OUT, + crc->regs + CRC_CR); } - d32 = (u32 *)d8; - for (i = 0; i < length >> 2; i++) - /* Process 32 bits data */ - writel_relaxed(*(d32++), crc->regs + CRC_DR); + for (; length >= sizeof(u32); d8 += sizeof(u32), length -= sizeof(u32)) + writel_relaxed(*((u32 *)d8), crc->regs + CRC_DR); + + if (length) { + /* Configure for byte data */ + writel_relaxed(CRC_CR_REV_IN_BYTE | CRC_CR_REV_OUT, + crc->regs + CRC_CR); + while (length--) + writeb_relaxed(*d8++, crc->regs + CRC_DR); + } /* Store partial result */ ctx->partial = readl_relaxed(crc->regs + CRC_DR); + spin_unlock_irqrestore(&crc->lock, flags); + pm_runtime_mark_last_busy(crc->dev); pm_runtime_put_autosuspend(crc->dev); - /* Check for pending data (non 32 bits) */ - length &= 3; - if (likely(!length)) - return 0; + return 0; +} - if ((crc->nb_pending_bytes + length) >= sizeof(u32)) { - /* Shall not happen */ - dev_err(crc->dev, "Pending data overflow\n"); - return -EINVAL; - } +static int stm32_crc_update(struct shash_desc *desc, const u8 *d8, + unsigned int length) +{ + const unsigned int burst_sz = burst_size; + unsigned int rem_sz; + const u8 *cur; + size_t size; + int ret; - d8 = (const u8 *)d32; - for (i = 0; i < length; i++) - /* Store pending data */ - crc->pending_data[crc->nb_pending_bytes++] = *(d8++); + if (!burst_sz) + return burst_update(desc, d8, length); + + /* Digest first bytes not 32bit aligned at first pass in the loop */ + size = min(length, + burst_sz + (unsigned int)d8 - ALIGN_DOWN((unsigned int)d8, + sizeof(u32))); + for (rem_sz = length, cur = d8; rem_sz; + rem_sz -= size, cur += size, size = min(rem_sz, burst_sz)) { + ret = burst_update(desc, cur, size); + if (ret) + return ret; + } return 0; } @@ -202,6 +254,8 @@ static int stm32_crc_digest(struct shash_desc *desc, const u8 *data, return stm32_crc_init(desc) ?: stm32_crc_finup(desc, data, length, out); } +static unsigned int refcnt; +static DEFINE_MUTEX(refcnt_lock); static struct shash_alg algs[] = { /* CRC-32 */ { @@ -284,20 +338,29 @@ static int stm32_crc_probe(struct platform_device *pdev) pm_runtime_get_noresume(dev); pm_runtime_set_active(dev); + pm_runtime_irq_safe(dev); pm_runtime_enable(dev); + spin_lock_init(&crc->lock); + platform_set_drvdata(pdev, crc); spin_lock(&crc_list.lock); list_add(&crc->list, &crc_list.dev_list); spin_unlock(&crc_list.lock); - ret = crypto_register_shashes(algs, ARRAY_SIZE(algs)); - if (ret) { - dev_err(dev, "Failed to register\n"); - clk_disable_unprepare(crc->clk); - return ret; + mutex_lock(&refcnt_lock); + if (!refcnt) { + ret = crypto_register_shashes(algs, ARRAY_SIZE(algs)); + if (ret) { + mutex_unlock(&refcnt_lock); + dev_err(dev, "Failed to register\n"); + clk_disable_unprepare(crc->clk); + return ret; + } } + refcnt++; + mutex_unlock(&refcnt_lock); dev_info(dev, "Initialized\n"); @@ -318,7 +381,10 @@ static int stm32_crc_remove(struct platform_device *pdev) list_del(&crc->list); spin_unlock(&crc_list.lock); - crypto_unregister_shashes(algs, ARRAY_SIZE(algs)); + mutex_lock(&refcnt_lock); + if (!--refcnt) + crypto_unregister_shashes(algs, ARRAY_SIZE(algs)); + mutex_unlock(&refcnt_lock); pm_runtime_disable(crc->dev); pm_runtime_put_noidle(crc->dev); @@ -328,34 +394,60 @@ static int stm32_crc_remove(struct platform_device *pdev) return 0; } -#ifdef CONFIG_PM -static int stm32_crc_runtime_suspend(struct device *dev) +static int __maybe_unused stm32_crc_suspend(struct device *dev) { struct stm32_crc *crc = dev_get_drvdata(dev); + int ret; - clk_disable_unprepare(crc->clk); + ret = pm_runtime_force_suspend(dev); + if (ret) + return ret; + + clk_unprepare(crc->clk); return 0; } -static int stm32_crc_runtime_resume(struct device *dev) +static int __maybe_unused stm32_crc_resume(struct device *dev) { struct stm32_crc *crc = dev_get_drvdata(dev); int ret; - ret = clk_prepare_enable(crc->clk); + ret = clk_prepare(crc->clk); if (ret) { - dev_err(crc->dev, "Failed to prepare_enable clock\n"); + dev_err(crc->dev, "Failed to prepare clock\n"); + return ret; + } + + return pm_runtime_force_resume(dev); +} + +static int __maybe_unused stm32_crc_runtime_suspend(struct device *dev) +{ + struct stm32_crc *crc = dev_get_drvdata(dev); + + clk_disable(crc->clk); + + return 0; +} + +static int __maybe_unused stm32_crc_runtime_resume(struct device *dev) +{ + struct stm32_crc *crc = dev_get_drvdata(dev); + int ret; + + ret = clk_enable(crc->clk); + if (ret) { + dev_err(crc->dev, "Failed to enable clock\n"); return ret; } return 0; } -#endif static const struct dev_pm_ops stm32_crc_pm_ops = { - SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend, - pm_runtime_force_resume) + SET_SYSTEM_SLEEP_PM_OPS(stm32_crc_suspend, + stm32_crc_resume) SET_RUNTIME_PM_OPS(stm32_crc_runtime_suspend, stm32_crc_runtime_resume, NULL) }; diff --git a/drivers/crypto/stm32/stm32-hash.c b/drivers/crypto/stm32/stm32-hash.c index 167b80eec437..03c5e6683805 100644 --- a/drivers/crypto/stm32/stm32-hash.c +++ b/drivers/crypto/stm32/stm32-hash.c @@ -507,6 +507,7 @@ static int stm32_hash_hmac_dma_send(struct stm32_hash_dev *hdev) static int stm32_hash_dma_init(struct stm32_hash_dev *hdev) { struct dma_slave_config dma_conf; + struct dma_chan *chan; int err; memset(&dma_conf, 0, sizeof(dma_conf)); @@ -518,11 +519,11 @@ static int stm32_hash_dma_init(struct stm32_hash_dev *hdev) dma_conf.dst_maxburst = hdev->dma_maxburst; dma_conf.device_fc = false; - hdev->dma_lch = dma_request_chan(hdev->dev, "in"); - if (IS_ERR(hdev->dma_lch)) { - dev_err(hdev->dev, "Couldn't acquire a slave DMA channel.\n"); - return PTR_ERR(hdev->dma_lch); - } + chan = dma_request_chan(hdev->dev, "in"); + if (IS_ERR(chan)) + return PTR_ERR(chan); + + hdev->dma_lch = chan; err = dmaengine_slave_config(hdev->dma_lch, &dma_conf); if (err) { @@ -1463,8 +1464,11 @@ static int stm32_hash_probe(struct platform_device *pdev) hdev->clk = devm_clk_get(&pdev->dev, NULL); if (IS_ERR(hdev->clk)) { - dev_err(dev, "failed to get clock for hash (%lu)\n", - PTR_ERR(hdev->clk)); + if (PTR_ERR(hdev->clk) != -EPROBE_DEFER) { + dev_err(dev, "failed to get clock for hash (%lu)\n", + PTR_ERR(hdev->clk)); + } + return PTR_ERR(hdev->clk); } @@ -1482,7 +1486,12 @@ static int stm32_hash_probe(struct platform_device *pdev) pm_runtime_enable(dev); hdev->rst = devm_reset_control_get(&pdev->dev, NULL); - if (!IS_ERR(hdev->rst)) { + if (IS_ERR(hdev->rst)) { + if (PTR_ERR(hdev->rst) == -EPROBE_DEFER) { + ret = -EPROBE_DEFER; + goto err_reset; + } + } else { reset_control_assert(hdev->rst); udelay(2); reset_control_deassert(hdev->rst); @@ -1493,8 +1502,15 @@ static int stm32_hash_probe(struct platform_device *pdev) platform_set_drvdata(pdev, hdev); ret = stm32_hash_dma_init(hdev); - if (ret) + switch (ret) { + case 0: + break; + case -ENOENT: dev_dbg(dev, "DMA mode not available\n"); + break; + default: + goto err_dma; + } spin_lock(&stm32_hash.lock); list_add_tail(&hdev->list, &stm32_hash.dev_list); @@ -1532,10 +1548,10 @@ err_engine: spin_lock(&stm32_hash.lock); list_del(&hdev->list); spin_unlock(&stm32_hash.lock); - +err_dma: if (hdev->dma_lch) dma_release_channel(hdev->dma_lch); - +err_reset: pm_runtime_disable(dev); pm_runtime_put_noidle(dev); diff --git a/drivers/nfc/s3fwrn5/firmware.c b/drivers/nfc/s3fwrn5/firmware.c index de613c623a2c..69857f080704 100644 --- a/drivers/nfc/s3fwrn5/firmware.c +++ b/drivers/nfc/s3fwrn5/firmware.c @@ -434,15 +434,7 @@ int s3fwrn5_fw_download(struct s3fwrn5_fw_info *fw_info) goto out; } - { - SHASH_DESC_ON_STACK(desc, tfm); - - desc->tfm = tfm; - - ret = crypto_shash_digest(desc, fw->image, image_size, - hash_data); - shash_desc_zero(desc); - } + ret = crypto_shash_tfm_digest(tfm, fw->image, image_size, hash_data); crypto_free_shash(tfm); if (ret) { diff --git a/fs/crypto/fname.c b/fs/crypto/fname.c index 4c212442a8f7..5c9fb013e3f7 100644 --- a/fs/crypto/fname.c +++ b/fs/crypto/fname.c @@ -83,13 +83,8 @@ static int fscrypt_do_sha256(const u8 *data, unsigned int data_len, u8 *result) tfm = prev_tfm; } } - { - SHASH_DESC_ON_STACK(desc, tfm); - desc->tfm = tfm; - - return crypto_shash_digest(desc, data, data_len, result); - } + return crypto_shash_tfm_digest(tfm, data, data_len, result); } static inline bool fscrypt_is_dot_dotdot(const struct qstr *str) diff --git a/fs/crypto/hkdf.c b/fs/crypto/hkdf.c index efb95bd19a89..0cba7928446d 100644 --- a/fs/crypto/hkdf.c +++ b/fs/crypto/hkdf.c @@ -44,17 +44,13 @@ static int hkdf_extract(struct crypto_shash *hmac_tfm, const u8 *ikm, unsigned int ikmlen, u8 prk[HKDF_HASHLEN]) { static const u8 default_salt[HKDF_HASHLEN]; - SHASH_DESC_ON_STACK(desc, hmac_tfm); int err; err = crypto_shash_setkey(hmac_tfm, default_salt, HKDF_HASHLEN); if (err) return err; - desc->tfm = hmac_tfm; - err = crypto_shash_digest(desc, ikm, ikmlen, prk); - shash_desc_zero(desc); - return err; + return crypto_shash_tfm_digest(hmac_tfm, ikm, ikmlen, prk); } /* diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 2c449aed1b92..0681540c48d9 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -48,18 +48,6 @@ void ecryptfs_from_hex(char *dst, char *src, int dst_size) } } -static int ecryptfs_hash_digest(struct crypto_shash *tfm, - char *src, int len, char *dst) -{ - SHASH_DESC_ON_STACK(desc, tfm); - int err; - - desc->tfm = tfm; - err = crypto_shash_digest(desc, src, len, dst); - shash_desc_zero(desc); - return err; -} - /** * ecryptfs_calculate_md5 - calculates the md5 of @src * @dst: Pointer to 16 bytes of allocated memory @@ -74,11 +62,8 @@ static int ecryptfs_calculate_md5(char *dst, struct ecryptfs_crypt_stat *crypt_stat, char *src, int len) { - struct crypto_shash *tfm; - int rc = 0; + int rc = crypto_shash_tfm_digest(crypt_stat->hash_tfm, src, len, dst); - tfm = crypt_stat->hash_tfm; - rc = ecryptfs_hash_digest(tfm, src, len, dst); if (rc) { printk(KERN_ERR "%s: Error computing crypto hash; rc = [%d]\n", diff --git a/fs/f2fs/hash.c b/fs/f2fs/hash.c index 5bc4dcd8fc03..8c4ea5003ef8 100644 --- a/fs/f2fs/hash.c +++ b/fs/f2fs/hash.c @@ -12,7 +12,6 @@ #include <linux/types.h> #include <linux/fs.h> #include <linux/f2fs_fs.h> -#include <linux/cryptohash.h> #include <linux/pagemap.h> #include <linux/unicode.h> diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index a8fb18609146..9e40dfecf1b1 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -127,16 +127,8 @@ nfs4_make_rec_clidname(char *dname, const struct xdr_netobj *clname) goto out; } - { - SHASH_DESC_ON_STACK(desc, tfm); - - desc->tfm = tfm; - - status = crypto_shash_digest(desc, clname->data, clname->len, - cksum.data); - shash_desc_zero(desc); - } - + status = crypto_shash_tfm_digest(tfm, clname->data, clname->len, + cksum.data); if (status) goto out; @@ -1148,7 +1140,6 @@ nfsd4_cld_create_v2(struct nfs4_client *clp) struct crypto_shash *tfm = cn->cn_tfm; struct xdr_netobj cksum; char *principal = NULL; - SHASH_DESC_ON_STACK(desc, tfm); /* Don't upcall if it's already stored */ if (test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags)) @@ -1170,16 +1161,14 @@ nfsd4_cld_create_v2(struct nfs4_client *clp) else if (clp->cl_cred.cr_principal) principal = clp->cl_cred.cr_principal; if (principal) { - desc->tfm = tfm; cksum.len = crypto_shash_digestsize(tfm); cksum.data = kmalloc(cksum.len, GFP_KERNEL); if (cksum.data == NULL) { ret = -ENOMEM; goto out; } - ret = crypto_shash_digest(desc, principal, strlen(principal), - cksum.data); - shash_desc_zero(desc); + ret = crypto_shash_tfm_digest(tfm, principal, strlen(principal), + cksum.data); if (ret) { kfree(cksum.data); goto out; @@ -1343,7 +1332,6 @@ nfsd4_cld_check_v2(struct nfs4_client *clp) struct crypto_shash *tfm = cn->cn_tfm; struct xdr_netobj cksum; char *principal = NULL; - SHASH_DESC_ON_STACK(desc, tfm); /* did we already find that this client is stable? */ if (test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags)) @@ -1381,14 +1369,12 @@ found: principal = clp->cl_cred.cr_principal; if (principal == NULL) return -ENOENT; - desc->tfm = tfm; cksum.len = crypto_shash_digestsize(tfm); cksum.data = kmalloc(cksum.len, GFP_KERNEL); if (cksum.data == NULL) return -ENOENT; - status = crypto_shash_digest(desc, principal, strlen(principal), - cksum.data); - shash_desc_zero(desc); + status = crypto_shash_tfm_digest(tfm, principal, + strlen(principal), cksum.data); if (status) { kfree(cksum.data); return -ENOENT; diff --git a/fs/ubifs/auth.c b/fs/ubifs/auth.c index f985a3fbbb36..cc5c0abfd536 100644 --- a/fs/ubifs/auth.c +++ b/fs/ubifs/auth.c @@ -31,15 +31,9 @@ int __ubifs_node_calc_hash(const struct ubifs_info *c, const void *node, u8 *hash) { const struct ubifs_ch *ch = node; - SHASH_DESC_ON_STACK(shash, c->hash_tfm); - int err; - - shash->tfm = c->hash_tfm; - err = crypto_shash_digest(shash, node, le32_to_cpu(ch->len), hash); - if (err < 0) - return err; - return 0; + return crypto_shash_tfm_digest(c->hash_tfm, node, le32_to_cpu(ch->len), + hash); } /** @@ -53,15 +47,7 @@ int __ubifs_node_calc_hash(const struct ubifs_info *c, const void *node, static int ubifs_hash_calc_hmac(const struct ubifs_info *c, const u8 *hash, u8 *hmac) { - SHASH_DESC_ON_STACK(shash, c->hmac_tfm); - int err; - - shash->tfm = c->hmac_tfm; - - err = crypto_shash_digest(shash, hash, c->hash_len, hmac); - if (err < 0) - return err; - return 0; + return crypto_shash_tfm_digest(c->hmac_tfm, hash, c->hash_len, hmac); } /** diff --git a/fs/ubifs/master.c b/fs/ubifs/master.c index 52a85c01397e..911d0555b9f2 100644 --- a/fs/ubifs/master.c +++ b/fs/ubifs/master.c @@ -68,12 +68,9 @@ static int mst_node_check_hash(const struct ubifs_info *c, u8 calc[UBIFS_MAX_HASH_LEN]; const void *node = mst; - SHASH_DESC_ON_STACK(shash, c->hash_tfm); - - shash->tfm = c->hash_tfm; - - crypto_shash_digest(shash, node + sizeof(struct ubifs_ch), - UBIFS_MST_NODE_SZ - sizeof(struct ubifs_ch), calc); + crypto_shash_tfm_digest(c->hash_tfm, node + sizeof(struct ubifs_ch), + UBIFS_MST_NODE_SZ - sizeof(struct ubifs_ch), + calc); if (ubifs_check_hash(c, expected, calc)) return -EPERM; diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index 01fcf7975047..b69ffac7e415 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -558,7 +558,7 @@ static int is_last_bud(struct ubifs_info *c, struct ubifs_bud *bud) return data == 0xFFFFFFFF; } -/* authenticate_sleb_hash and authenticate_sleb_hmac are split out for stack usage */ +/* authenticate_sleb_hash is split out for stack usage */ static int authenticate_sleb_hash(struct ubifs_info *c, struct shash_desc *log_hash, u8 *hash) { SHASH_DESC_ON_STACK(hash_desc, c->hash_tfm); @@ -569,15 +569,6 @@ static int authenticate_sleb_hash(struct ubifs_info *c, struct shash_desc *log_h return crypto_shash_final(hash_desc, hash); } -static int authenticate_sleb_hmac(struct ubifs_info *c, u8 *hash, u8 *hmac) -{ - SHASH_DESC_ON_STACK(hmac_desc, c->hmac_tfm); - - hmac_desc->tfm = c->hmac_tfm; - - return crypto_shash_digest(hmac_desc, hash, c->hash_len, hmac); -} - /** * authenticate_sleb - authenticate one scan LEB * @c: UBIFS file-system description object @@ -618,7 +609,8 @@ static int authenticate_sleb(struct ubifs_info *c, struct ubifs_scan_leb *sleb, if (err) goto out; - err = authenticate_sleb_hmac(c, hash, hmac); + err = crypto_shash_tfm_digest(c->hmac_tfm, hash, + c->hash_len, hmac); if (err) goto out; diff --git a/include/crypto/acompress.h b/include/crypto/acompress.h index d873f999b334..2b4d2b06ccbd 100644 --- a/include/crypto/acompress.h +++ b/include/crypto/acompress.h @@ -157,7 +157,7 @@ static inline int crypto_has_acomp(const char *alg_name, u32 type, u32 mask) { type &= ~CRYPTO_ALG_TYPE_MASK; type |= CRYPTO_ALG_TYPE_ACOMPRESS; - mask |= CRYPTO_ALG_TYPE_MASK; + mask |= CRYPTO_ALG_TYPE_ACOMPRESS_MASK; return crypto_has_alg(alg_name, type, mask); } diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h index e115f9215ed5..00a9cf98debe 100644 --- a/include/crypto/algapi.h +++ b/include/crypto/algapi.h @@ -125,6 +125,8 @@ int crypto_inst_setname(struct crypto_instance *inst, const char *name, void crypto_init_queue(struct crypto_queue *queue, unsigned int max_qlen); int crypto_enqueue_request(struct crypto_queue *queue, struct crypto_async_request *request); +void crypto_enqueue_request_head(struct crypto_queue *queue, + struct crypto_async_request *request); struct crypto_async_request *crypto_dequeue_request(struct crypto_queue *queue); static inline unsigned int crypto_queue_len(struct crypto_queue *queue) { diff --git a/include/crypto/drbg.h b/include/crypto/drbg.h index 8c9af21efce1..c4165126937e 100644 --- a/include/crypto/drbg.h +++ b/include/crypto/drbg.h @@ -184,11 +184,7 @@ static inline size_t drbg_max_addtl(struct drbg_state *drbg) static inline size_t drbg_max_requests(struct drbg_state *drbg) { /* SP800-90A requires 2**48 maximum requests before reseeding */ -#if (__BITS_PER_LONG == 32) - return SIZE_MAX; -#else - return (1UL<<48); -#endif + return (1<<20); } /* diff --git a/include/crypto/engine.h b/include/crypto/engine.h index e29cd67f93c7..3f06e40d063a 100644 --- a/include/crypto/engine.h +++ b/include/crypto/engine.h @@ -24,7 +24,9 @@ * @idling: the engine is entering idle state * @busy: request pump is busy * @running: the engine is on working - * @cur_req_prepared: current request is prepared + * @retry_support: indication that the hardware allows re-execution + * of a failed backlog request + * crypto-engine, in head position to keep order * @list: link with the global crypto engine list * @queue_lock: spinlock to syncronise access to request queue * @queue: the crypto queue of the engine @@ -35,6 +37,8 @@ * @unprepare_crypt_hardware: there are currently no more requests on the * queue so the subsystem notifies the driver that it may relax the * hardware by issuing this call + * @do_batch_requests: execute a batch of requests. Depends on multiple + * requests support. * @kworker: kthread worker struct for request pump * @pump_requests: work struct for scheduling work to the request pump * @priv_data: the engine private data @@ -45,7 +49,8 @@ struct crypto_engine { bool idling; bool busy; bool running; - bool cur_req_prepared; + + bool retry_support; struct list_head list; spinlock_t queue_lock; @@ -56,6 +61,8 @@ struct crypto_engine { int (*prepare_crypt_hardware)(struct crypto_engine *engine); int (*unprepare_crypt_hardware)(struct crypto_engine *engine); + int (*do_batch_requests)(struct crypto_engine *engine); + struct kthread_worker *kworker; struct kthread_work pump_requests; @@ -102,6 +109,10 @@ void crypto_finalize_skcipher_request(struct crypto_engine *engine, int crypto_engine_start(struct crypto_engine *engine); int crypto_engine_stop(struct crypto_engine *engine); struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt); +struct crypto_engine *crypto_engine_alloc_init_and_set(struct device *dev, + bool retry_support, + int (*cbk_do_batch)(struct crypto_engine *engine), + bool rt, int qlen); int crypto_engine_exit(struct crypto_engine *engine); #endif /* _CRYPTO_ENGINE_H */ diff --git a/include/crypto/hash.h b/include/crypto/hash.h index cee446c59497..4829d2367eda 100644 --- a/include/crypto/hash.h +++ b/include/crypto/hash.h @@ -856,6 +856,25 @@ int crypto_shash_digest(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out); /** + * crypto_shash_tfm_digest() - calculate message digest for buffer + * @tfm: hash transformation object + * @data: see crypto_shash_update() + * @len: see crypto_shash_update() + * @out: see crypto_shash_final() + * + * This is a simplified version of crypto_shash_digest() for users who don't + * want to allocate their own hash descriptor (shash_desc). Instead, + * crypto_shash_tfm_digest() takes a hash transformation object (crypto_shash) + * directly, and it allocates a hash descriptor on the stack internally. + * Note that this stack allocation may be fairly large. + * + * Context: Any context. + * Return: 0 on success; < 0 if an error occurred. + */ +int crypto_shash_tfm_digest(struct crypto_shash *tfm, const u8 *data, + unsigned int len, u8 *out); + +/** * crypto_shash_export() - extract operational state for message digest * @desc: reference to the operational state handle whose state is exported * @out: output buffer of sufficient size that can hold the hash state diff --git a/include/crypto/sha.h b/include/crypto/sha.h index 5c2132c71900..10753ff71d46 100644 --- a/include/crypto/sha.h +++ b/include/crypto/sha.h @@ -114,6 +114,16 @@ extern int crypto_sha512_finup(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *hash); /* + * An implementation of SHA-1's compression function. Don't use in new code! + * You shouldn't be using SHA-1, and even if you *have* to use SHA-1, this isn't + * the correct way to hash something with SHA-1 (use crypto_shash instead). + */ +#define SHA1_DIGEST_WORDS (SHA1_DIGEST_SIZE / 4) +#define SHA1_WORKSPACE_WORDS 16 +void sha1_init(__u32 *buf); +void sha1_transform(__u32 *digest, const char *data, __u32 *W); + +/* * Stand-alone implementation of the SHA256 algorithm. It is designed to * have as little dependencies as possible so it can be used in the * kexec_file purgatory. In other cases you should generally use the @@ -123,7 +133,7 @@ extern int crypto_sha512_finup(struct shash_desc *desc, const u8 *data, * For details see lib/crypto/sha256.c */ -static inline int sha256_init(struct sha256_state *sctx) +static inline void sha256_init(struct sha256_state *sctx) { sctx->state[0] = SHA256_H0; sctx->state[1] = SHA256_H1; @@ -134,14 +144,11 @@ static inline int sha256_init(struct sha256_state *sctx) sctx->state[6] = SHA256_H6; sctx->state[7] = SHA256_H7; sctx->count = 0; - - return 0; } -extern int sha256_update(struct sha256_state *sctx, const u8 *input, - unsigned int length); -extern int sha256_final(struct sha256_state *sctx, u8 *hash); +void sha256_update(struct sha256_state *sctx, const u8 *data, unsigned int len); +void sha256_final(struct sha256_state *sctx, u8 *out); -static inline int sha224_init(struct sha256_state *sctx) +static inline void sha224_init(struct sha256_state *sctx) { sctx->state[0] = SHA224_H0; sctx->state[1] = SHA224_H1; @@ -152,11 +159,8 @@ static inline int sha224_init(struct sha256_state *sctx) sctx->state[6] = SHA224_H6; sctx->state[7] = SHA224_H7; sctx->count = 0; - - return 0; } -extern int sha224_update(struct sha256_state *sctx, const u8 *input, - unsigned int length); -extern int sha224_final(struct sha256_state *sctx, u8 *hash); +void sha224_update(struct sha256_state *sctx, const u8 *data, unsigned int len); +void sha224_final(struct sha256_state *sctx, u8 *out); #endif diff --git a/include/crypto/sha256_base.h b/include/crypto/sha256_base.h index cea60cff80bd..6ded110783ae 100644 --- a/include/crypto/sha256_base.h +++ b/include/crypto/sha256_base.h @@ -22,14 +22,16 @@ static inline int sha224_base_init(struct shash_desc *desc) { struct sha256_state *sctx = shash_desc_ctx(desc); - return sha224_init(sctx); + sha224_init(sctx); + return 0; } static inline int sha256_base_init(struct shash_desc *desc) { struct sha256_state *sctx = shash_desc_ctx(desc); - return sha256_init(sctx); + sha256_init(sctx); + return 0; } static inline int sha256_base_do_update(struct shash_desc *desc, diff --git a/include/linux/cryptohash.h b/include/linux/cryptohash.h deleted file mode 100644 index f6ba4c3e60d7..000000000000 --- a/include/linux/cryptohash.h +++ /dev/null @@ -1,14 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __CRYPTOHASH_H -#define __CRYPTOHASH_H - -#include <uapi/linux/types.h> - -#define SHA_DIGEST_WORDS 5 -#define SHA_MESSAGE_BYTES (512 /*bits*/ / 8) -#define SHA_WORKSPACE_WORDS 16 - -void sha_init(__u32 *buf); -void sha_transform(__u32 *digest, const char *data, __u32 *W); - -#endif diff --git a/include/linux/filter.h b/include/linux/filter.h index 9b5aa5c483cc..ec45fd7992c9 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -16,11 +16,11 @@ #include <linux/workqueue.h> #include <linux/sched.h> #include <linux/capability.h> -#include <linux/cryptohash.h> #include <linux/set_memory.h> #include <linux/kallsyms.h> #include <linux/if_vlan.h> #include <linux/vmalloc.h> +#include <crypto/sha.h> #include <net/sch_generic.h> @@ -746,7 +746,7 @@ static inline u32 bpf_prog_insn_size(const struct bpf_prog *prog) static inline u32 bpf_prog_tag_scratch_size(const struct bpf_prog *prog) { return round_up(bpf_prog_insn_size(prog) + - sizeof(__be64) + 1, SHA_MESSAGE_BYTES); + sizeof(__be64) + 1, SHA1_BLOCK_SIZE); } static inline unsigned int bpf_prog_size(unsigned int proglen) diff --git a/include/linux/padata.h b/include/linux/padata.h index a0d8b41850b2..693cae9bfe66 100644 --- a/include/linux/padata.h +++ b/include/linux/padata.h @@ -139,7 +139,8 @@ struct padata_shell { /** * struct padata_instance - The overall control structure. * - * @node: Used by CPU hotplug. + * @cpu_online_node: Linkage for CPU online callback. + * @cpu_dead_node: Linkage for CPU offline callback. * @parallel_wq: The workqueue used for parallel work. * @serial_wq: The workqueue used for serial work. * @pslist: List of padata_shell objects attached to this instance. @@ -150,7 +151,8 @@ struct padata_shell { * @flags: padata flags. */ struct padata_instance { - struct hlist_node node; + struct hlist_node cpu_online_node; + struct hlist_node cpu_dead_node; struct workqueue_struct *parallel_wq; struct workqueue_struct *serial_wq; struct list_head pslist; diff --git a/include/linux/psp-sev.h b/include/linux/psp-sev.h index 5167bf2bfc75..7fbc8679145c 100644 --- a/include/linux/psp-sev.h +++ b/include/linux/psp-sev.h @@ -100,6 +100,8 @@ struct sev_data_init { u32 tmr_len; /* In */ } __packed; +#define SEV_INIT_FLAGS_SEV_ES 0x01 + /** * struct sev_data_pek_csr - PEK_CSR command parameters * diff --git a/include/net/tcp.h b/include/net/tcp.h index 6f8e60c6fbc7..cf5649a2e795 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -23,7 +23,6 @@ #include <linux/cache.h> #include <linux/percpu.h> #include <linux/skbuff.h> -#include <linux/cryptohash.h> #include <linux/kref.h> #include <linux/ktime.h> diff --git a/include/uapi/linux/psp-sev.h b/include/uapi/linux/psp-sev.h index 0549a5c622bf..91b4c63d5cbf 100644 --- a/include/uapi/linux/psp-sev.h +++ b/include/uapi/linux/psp-sev.h @@ -83,6 +83,8 @@ struct sev_user_data_status { __u32 guest_count; /* Out */ } __packed; +#define SEV_STATUS_FLAGS_CONFIG_ES 0x0100 + /** * struct sev_user_data_pek_csr - PEK_CSR command parameters * diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 916f5132a984..14aa1f74dd10 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -262,10 +262,10 @@ void __bpf_prog_free(struct bpf_prog *fp) int bpf_prog_calc_tag(struct bpf_prog *fp) { - const u32 bits_offset = SHA_MESSAGE_BYTES - sizeof(__be64); + const u32 bits_offset = SHA1_BLOCK_SIZE - sizeof(__be64); u32 raw_size = bpf_prog_tag_scratch_size(fp); - u32 digest[SHA_DIGEST_WORDS]; - u32 ws[SHA_WORKSPACE_WORDS]; + u32 digest[SHA1_DIGEST_WORDS]; + u32 ws[SHA1_WORKSPACE_WORDS]; u32 i, bsize, psize, blocks; struct bpf_insn *dst; bool was_ld_map; @@ -277,7 +277,7 @@ int bpf_prog_calc_tag(struct bpf_prog *fp) if (!raw) return -ENOMEM; - sha_init(digest); + sha1_init(digest); memset(ws, 0, sizeof(ws)); /* We need to take out the map fd for the digest calculation @@ -308,8 +308,8 @@ int bpf_prog_calc_tag(struct bpf_prog *fp) memset(&raw[psize], 0, raw_size - psize); raw[psize++] = 0x80; - bsize = round_up(psize, SHA_MESSAGE_BYTES); - blocks = bsize / SHA_MESSAGE_BYTES; + bsize = round_up(psize, SHA1_BLOCK_SIZE); + blocks = bsize / SHA1_BLOCK_SIZE; todo = raw; if (bsize - psize >= sizeof(__be64)) { bits = (__be64 *)(todo + bsize - sizeof(__be64)); @@ -320,12 +320,12 @@ int bpf_prog_calc_tag(struct bpf_prog *fp) *bits = cpu_to_be64((psize - 1) << 3); while (blocks--) { - sha_transform(digest, todo, ws); - todo += SHA_MESSAGE_BYTES; + sha1_transform(digest, todo, ws); + todo += SHA1_BLOCK_SIZE; } result = (__force __be32 *)digest; - for (i = 0; i < SHA_DIGEST_WORDS; i++) + for (i = 0; i < SHA1_DIGEST_WORDS; i++) result[i] = cpu_to_be32(digest[i]); memcpy(fp->tag, result, sizeof(fp->tag)); diff --git a/kernel/padata.c b/kernel/padata.c index a6afa12fb75e..aae789896616 100644 --- a/kernel/padata.c +++ b/kernel/padata.c @@ -703,7 +703,7 @@ static int padata_cpu_online(unsigned int cpu, struct hlist_node *node) struct padata_instance *pinst; int ret; - pinst = hlist_entry_safe(node, struct padata_instance, node); + pinst = hlist_entry_safe(node, struct padata_instance, cpu_online_node); if (!pinst_has_cpu(pinst, cpu)) return 0; @@ -718,7 +718,7 @@ static int padata_cpu_dead(unsigned int cpu, struct hlist_node *node) struct padata_instance *pinst; int ret; - pinst = hlist_entry_safe(node, struct padata_instance, node); + pinst = hlist_entry_safe(node, struct padata_instance, cpu_dead_node); if (!pinst_has_cpu(pinst, cpu)) return 0; @@ -734,8 +734,9 @@ static enum cpuhp_state hp_online; static void __padata_free(struct padata_instance *pinst) { #ifdef CONFIG_HOTPLUG_CPU - cpuhp_state_remove_instance_nocalls(CPUHP_PADATA_DEAD, &pinst->node); - cpuhp_state_remove_instance_nocalls(hp_online, &pinst->node); + cpuhp_state_remove_instance_nocalls(CPUHP_PADATA_DEAD, + &pinst->cpu_dead_node); + cpuhp_state_remove_instance_nocalls(hp_online, &pinst->cpu_online_node); #endif WARN_ON(!list_empty(&pinst->pslist)); @@ -939,9 +940,10 @@ static struct padata_instance *padata_alloc(const char *name, mutex_init(&pinst->lock); #ifdef CONFIG_HOTPLUG_CPU - cpuhp_state_add_instance_nocalls_cpuslocked(hp_online, &pinst->node); + cpuhp_state_add_instance_nocalls_cpuslocked(hp_online, + &pinst->cpu_online_node); cpuhp_state_add_instance_nocalls_cpuslocked(CPUHP_PADATA_DEAD, - &pinst->node); + &pinst->cpu_dead_node); #endif put_online_cpus(); diff --git a/lib/crypto/chacha.c b/lib/crypto/chacha.c index 65ead6b0c7e0..4ccbec442469 100644 --- a/lib/crypto/chacha.c +++ b/lib/crypto/chacha.c @@ -10,7 +10,6 @@ #include <linux/export.h> #include <linux/bitops.h> #include <linux/string.h> -#include <linux/cryptohash.h> #include <asm/unaligned.h> #include <crypto/chacha.h> diff --git a/lib/crypto/sha256.c b/lib/crypto/sha256.c index 66cb04b0cf4e..2e621697c5c3 100644 --- a/lib/crypto/sha256.c +++ b/lib/crypto/sha256.c @@ -206,7 +206,7 @@ static void sha256_transform(u32 *state, const u8 *input) memzero_explicit(W, 64 * sizeof(u32)); } -int sha256_update(struct sha256_state *sctx, const u8 *data, unsigned int len) +void sha256_update(struct sha256_state *sctx, const u8 *data, unsigned int len) { unsigned int partial, done; const u8 *src; @@ -232,18 +232,16 @@ int sha256_update(struct sha256_state *sctx, const u8 *data, unsigned int len) partial = 0; } memcpy(sctx->buf + partial, src, len - done); - - return 0; } EXPORT_SYMBOL(sha256_update); -int sha224_update(struct sha256_state *sctx, const u8 *data, unsigned int len) +void sha224_update(struct sha256_state *sctx, const u8 *data, unsigned int len) { - return sha256_update(sctx, data, len); + sha256_update(sctx, data, len); } EXPORT_SYMBOL(sha224_update); -static int __sha256_final(struct sha256_state *sctx, u8 *out, int digest_words) +static void __sha256_final(struct sha256_state *sctx, u8 *out, int digest_words) { __be32 *dst = (__be32 *)out; __be64 bits; @@ -268,19 +266,17 @@ static int __sha256_final(struct sha256_state *sctx, u8 *out, int digest_words) /* Zeroize sensitive information. */ memset(sctx, 0, sizeof(*sctx)); - - return 0; } -int sha256_final(struct sha256_state *sctx, u8 *out) +void sha256_final(struct sha256_state *sctx, u8 *out) { - return __sha256_final(sctx, out, 8); + __sha256_final(sctx, out, 8); } EXPORT_SYMBOL(sha256_final); -int sha224_final(struct sha256_state *sctx, u8 *out) +void sha224_final(struct sha256_state *sctx, u8 *out) { - return __sha256_final(sctx, out, 7); + __sha256_final(sctx, out, 7); } EXPORT_SYMBOL(sha224_final); diff --git a/lib/mpi/longlong.h b/lib/mpi/longlong.h index 891e1c3549c4..afbd99987cf8 100644 --- a/lib/mpi/longlong.h +++ b/lib/mpi/longlong.h @@ -653,7 +653,7 @@ do { \ ************** MIPS/64 ************** ***************************************/ #if (defined(__mips) && __mips >= 3) && W_TYPE_SIZE == 64 -#if defined(__mips_isa_rev) && __mips_isa_rev >= 6 +#if defined(__mips_isa_rev) && __mips_isa_rev >= 6 && defined(CONFIG_CC_IS_GCC) /* * GCC ends up emitting a __multi3 intrinsic call for MIPS64r6 with the plain C * code below, so we special case MIPS64r6 until the compiler can do better. diff --git a/lib/sha1.c b/lib/sha1.c index 1d96d2c02b82..49257a915bb6 100644 --- a/lib/sha1.c +++ b/lib/sha1.c @@ -9,7 +9,7 @@ #include <linux/kernel.h> #include <linux/export.h> #include <linux/bitops.h> -#include <linux/cryptohash.h> +#include <crypto/sha.h> #include <asm/unaligned.h> /* @@ -64,22 +64,24 @@ #define T_60_79(t, A, B, C, D, E) SHA_ROUND(t, SHA_MIX, (B^C^D) , 0xca62c1d6, A, B, C, D, E ) /** - * sha_transform - single block SHA1 transform + * sha1_transform - single block SHA1 transform (deprecated) * * @digest: 160 bit digest to update * @data: 512 bits of data to hash * @array: 16 words of workspace (see note) * - * This function generates a SHA1 digest for a single 512-bit block. - * Be warned, it does not handle padding and message digest, do not - * confuse it with the full FIPS 180-1 digest algorithm for variable - * length messages. + * This function executes SHA-1's internal compression function. It updates the + * 160-bit internal state (@digest) with a single 512-bit data block (@data). + * + * Don't use this function. SHA-1 is no longer considered secure. And even if + * you do have to use SHA-1, this isn't the correct way to hash something with + * SHA-1 as this doesn't handle padding and finalization. * * Note: If the hash is security sensitive, the caller should be sure * to clear the workspace. This is left to the caller to avoid * unnecessary clears between chained hashing operations. */ -void sha_transform(__u32 *digest, const char *data, __u32 *array) +void sha1_transform(__u32 *digest, const char *data, __u32 *array) { __u32 A, B, C, D, E; @@ -185,13 +187,13 @@ void sha_transform(__u32 *digest, const char *data, __u32 *array) digest[3] += D; digest[4] += E; } -EXPORT_SYMBOL(sha_transform); +EXPORT_SYMBOL(sha1_transform); /** - * sha_init - initialize the vectors for a SHA1 digest + * sha1_init - initialize the vectors for a SHA1 digest * @buf: vector to initialize */ -void sha_init(__u32 *buf) +void sha1_init(__u32 *buf) { buf[0] = 0x67452301; buf[1] = 0xefcdab89; @@ -199,4 +201,4 @@ void sha_init(__u32 *buf) buf[3] = 0x10325476; buf[4] = 0xc3d2e1f0; } -EXPORT_SYMBOL(sha_init); +EXPORT_SYMBOL(sha1_init); diff --git a/net/bluetooth/smp.c b/net/bluetooth/smp.c index 1476a91ce935..d022f126eb02 100644 --- a/net/bluetooth/smp.c +++ b/net/bluetooth/smp.c @@ -170,7 +170,6 @@ static int aes_cmac(struct crypto_shash *tfm, const u8 k[16], const u8 *m, size_t len, u8 mac[16]) { uint8_t tmp[16], mac_msb[16], msg_msb[CMAC_MSG_MAX]; - SHASH_DESC_ON_STACK(desc, tfm); int err; if (len > CMAC_MSG_MAX) @@ -181,8 +180,6 @@ static int aes_cmac(struct crypto_shash *tfm, const u8 k[16], const u8 *m, return -EINVAL; } - desc->tfm = tfm; - /* Swap key and message from LSB to MSB */ swap_buf(k, tmp, 16); swap_buf(m, msg_msb, len); @@ -196,8 +193,7 @@ static int aes_cmac(struct crypto_shash *tfm, const u8 k[16], const u8 *m, return err; } - err = crypto_shash_digest(desc, msg_msb, len, mac_msb); - shash_desc_zero(desc); + err = crypto_shash_tfm_digest(tfm, msg_msb, len, mac_msb); if (err) { BT_ERR("Hash computation error %d", err); return err; diff --git a/net/core/secure_seq.c b/net/core/secure_seq.c index 7b6b1d2c3d10..b5bc680d4755 100644 --- a/net/core/secure_seq.c +++ b/net/core/secure_seq.c @@ -5,7 +5,6 @@ #include <linux/kernel.h> #include <linux/init.h> -#include <linux/cryptohash.h> #include <linux/module.h> #include <linux/cache.h> #include <linux/random.h> diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index 24e319dfb510..f131cedf5ba6 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -3222,11 +3222,11 @@ static int ipv6_generate_stable_address(struct in6_addr *address, const struct inet6_dev *idev) { static DEFINE_SPINLOCK(lock); - static __u32 digest[SHA_DIGEST_WORDS]; - static __u32 workspace[SHA_WORKSPACE_WORDS]; + static __u32 digest[SHA1_DIGEST_WORDS]; + static __u32 workspace[SHA1_WORKSPACE_WORDS]; static union { - char __data[SHA_MESSAGE_BYTES]; + char __data[SHA1_BLOCK_SIZE]; struct { struct in6_addr secret; __be32 prefix[2]; @@ -3251,7 +3251,7 @@ static int ipv6_generate_stable_address(struct in6_addr *address, retry: spin_lock_bh(&lock); - sha_init(digest); + sha1_init(digest); memset(&data, 0, sizeof(data)); memset(workspace, 0, sizeof(workspace)); memcpy(data.hwaddr, idev->dev->perm_addr, idev->dev->addr_len); @@ -3260,7 +3260,7 @@ retry: data.secret = secret; data.dad_count = dad_count; - sha_transform(digest, data.__data, workspace); + sha1_transform(digest, data.__data, workspace); temp = *address; temp.s6_addr32[2] = (__force __be32)digest[0]; diff --git a/net/ipv6/seg6_hmac.c b/net/ipv6/seg6_hmac.c index ffcfcd2b128f..85dddfe3a2c6 100644 --- a/net/ipv6/seg6_hmac.c +++ b/net/ipv6/seg6_hmac.c @@ -34,7 +34,6 @@ #include <net/addrconf.h> #include <net/xfrm.h> -#include <linux/cryptohash.h> #include <crypto/hash.h> #include <crypto/sha.h> #include <net/seg6.h> diff --git a/net/mptcp/crypto.c b/net/mptcp/crypto.c index 0f5a414a9366..3d980713a9e2 100644 --- a/net/mptcp/crypto.c +++ b/net/mptcp/crypto.c @@ -59,7 +59,7 @@ void mptcp_crypto_hmac_sha(u64 key1, u64 key2, u8 *msg, int len, void *hmac) put_unaligned_be64(key2, key2be); /* Generate key xored with ipad */ - memset(input, 0x36, SHA_MESSAGE_BYTES); + memset(input, 0x36, SHA256_BLOCK_SIZE); for (i = 0; i < 8; i++) input[i] ^= key1be[i]; for (i = 0; i < 8; i++) @@ -76,7 +76,7 @@ void mptcp_crypto_hmac_sha(u64 key1, u64 key2, u8 *msg, int len, void *hmac) sha256_final(&state, &input[SHA256_BLOCK_SIZE]); /* Prepare second part of hmac */ - memset(input, 0x5C, SHA_MESSAGE_BYTES); + memset(input, 0x5C, SHA256_BLOCK_SIZE); for (i = 0; i < 8; i++) input[i] ^= key1be[i]; for (i = 0; i < 8; i++) diff --git a/net/sctp/auth.c b/net/sctp/auth.c index 4278764d82b8..83e97e8892e0 100644 --- a/net/sctp/auth.c +++ b/net/sctp/auth.c @@ -741,14 +741,8 @@ void sctp_auth_calculate_hmac(const struct sctp_association *asoc, if (crypto_shash_setkey(tfm, &asoc_key->data[0], asoc_key->len)) goto free; - { - SHASH_DESC_ON_STACK(desc, tfm); - - desc->tfm = tfm; - crypto_shash_digest(desc, (u8 *)auth, - end - (unsigned char *)auth, digest); - shash_desc_zero(desc); - } + crypto_shash_tfm_digest(tfm, (u8 *)auth, end - (unsigned char *)auth, + digest); free: if (free_key) diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index f7cb0b7faec2..47910470e532 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -1670,17 +1670,14 @@ static struct sctp_cookie_param *sctp_pack_cookie( ntohs(init_chunk->chunk_hdr->length), raw_addrs, addrs_len); if (sctp_sk(ep->base.sk)->hmac) { - SHASH_DESC_ON_STACK(desc, sctp_sk(ep->base.sk)->hmac); + struct crypto_shash *tfm = sctp_sk(ep->base.sk)->hmac; int err; /* Sign the message. */ - desc->tfm = sctp_sk(ep->base.sk)->hmac; - - err = crypto_shash_setkey(desc->tfm, ep->secret_key, + err = crypto_shash_setkey(tfm, ep->secret_key, sizeof(ep->secret_key)) ?: - crypto_shash_digest(desc, (u8 *)&cookie->c, bodysize, - cookie->signature); - shash_desc_zero(desc); + crypto_shash_tfm_digest(tfm, (u8 *)&cookie->c, bodysize, + cookie->signature); if (err) goto free_cookie; } @@ -1741,17 +1738,13 @@ struct sctp_association *sctp_unpack_cookie( /* Check the signature. */ { - SHASH_DESC_ON_STACK(desc, sctp_sk(ep->base.sk)->hmac); + struct crypto_shash *tfm = sctp_sk(ep->base.sk)->hmac; int err; - desc->tfm = sctp_sk(ep->base.sk)->hmac; - - err = crypto_shash_setkey(desc->tfm, ep->secret_key, + err = crypto_shash_setkey(tfm, ep->secret_key, sizeof(ep->secret_key)) ?: - crypto_shash_digest(desc, (u8 *)bear_cookie, bodysize, - digest); - shash_desc_zero(desc); - + crypto_shash_tfm_digest(tfm, (u8 *)bear_cookie, bodysize, + digest); if (err) { *error = -SCTP_IERROR_NOMEM; goto fail; diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c index f6797ba44bf7..14cf81d1a30b 100644 --- a/security/keys/encrypted-keys/encrypted.c +++ b/security/keys/encrypted-keys/encrypted.c @@ -323,19 +323,6 @@ error: return ukey; } -static int calc_hash(struct crypto_shash *tfm, u8 *digest, - const u8 *buf, unsigned int buflen) -{ - SHASH_DESC_ON_STACK(desc, tfm); - int err; - - desc->tfm = tfm; - - err = crypto_shash_digest(desc, buf, buflen, digest); - shash_desc_zero(desc); - return err; -} - static int calc_hmac(u8 *digest, const u8 *key, unsigned int keylen, const u8 *buf, unsigned int buflen) { @@ -351,7 +338,7 @@ static int calc_hmac(u8 *digest, const u8 *key, unsigned int keylen, err = crypto_shash_setkey(tfm, key, keylen); if (!err) - err = calc_hash(tfm, digest, buf, buflen); + err = crypto_shash_tfm_digest(tfm, buf, buflen, digest); crypto_free_shash(tfm); return err; } @@ -381,7 +368,8 @@ static int get_derived_key(u8 *derived_key, enum derived_key_type key_type, memcpy(derived_buf + strlen(derived_buf) + 1, master_key, master_keylen); - ret = calc_hash(hash_tfm, derived_key, derived_buf, derived_buf_len); + ret = crypto_shash_tfm_digest(hash_tfm, derived_buf, derived_buf_len, + derived_key); kzfree(derived_buf); return ret; } |