linux-2.6-microblaze.git
3 years agorandom: initialize ChaCha20 constants with correct endianness
Eric Biggers [Mon, 22 Mar 2021 05:13:47 +0000 (22:13 -0700)]
random: initialize ChaCha20 constants with correct endianness

On big endian CPUs, the ChaCha20-based CRNG is using the wrong
endianness for the ChaCha20 constants.

This doesn't matter cryptographically, but technically it means it's not
ChaCha20 anymore.  Fix it to always use the standard constants.

Cc: linux-crypto@vger.kernel.org
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: rng - fix crypto_rng_reset() refcounting when !CRYPTO_STATS
Eric Biggers [Mon, 22 Mar 2021 05:07:48 +0000 (22:07 -0700)]
crypto: rng - fix crypto_rng_reset() refcounting when !CRYPTO_STATS

crypto_stats_get() is a no-op when the kernel is compiled without
CONFIG_CRYPTO_STATS, so pairing it with crypto_alg_put() unconditionally
(as crypto_rng_reset() does) is wrong.

Fix this by moving the call to crypto_stats_get() to just before the
actual algorithm operation which might need it.  This makes it always
paired with crypto_stats_rng_seed().

Fixes: eed74b3eba9e ("crypto: rng - Fix a refcounting bug in crypto_rng_reset()")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: nx - fix incorrect kernel-doc comment syntax in files
Aditya Srivastava [Sun, 21 Mar 2021 12:30:07 +0000 (18:00 +0530)]
crypto: nx - fix incorrect kernel-doc comment syntax in files

The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
There are certain files in drivers/crypto/nx, which follow this syntax,
but the content inside does not comply with kernel-doc.
Such lines were probably not meant for kernel-doc parsing, but are parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warnings from kernel-doc.

E.g., presence of kernel-doc like comment in the header lines for
drivers/crypto/nx/nx-sha256.c at header causes these warnings:
"warning: Function parameter or member 'tfm' not described in 'nx_crypto_ctx_sha256_init'"
"warning: expecting prototype for SHA(). Prototype was for nx_crypto_ctx_sha256_init() instead"

Similarly for other files too.

Provide a simple fix by replacing such occurrences with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.

Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ux500 - fix incorrect kernel-doc comment syntax
Aditya Srivastava [Sun, 21 Mar 2021 12:09:12 +0000 (17:39 +0530)]
crypto: ux500 - fix incorrect kernel-doc comment syntax

The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
There are certain files in drivers/crypto/ux500, which follow this syntax,
but the content inside does not comply with kernel-doc.
Such lines were probably not meant for kernel-doc parsing, but are parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warnings from kernel-doc.

E.g., presence of kernel-doc like comment in the header lines for
drivers/crypto/ux500/cryp/cryp.h at header causes this warning:

"warning: expecting prototype for ST(). Prototype was for _CRYP_H_() instead"

Similarly for other files too.

Provide a simple fix by replacing such occurrences with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.

Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: amcc - fix incorrect kernel-doc comment syntax in files
Aditya Srivastava [Sun, 21 Mar 2021 12:08:32 +0000 (17:38 +0530)]
crypto: amcc - fix incorrect kernel-doc comment syntax in files

The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
There are certain files in drivers/crypto/amcc, which follow this syntax,
but the content inside does not comply with kernel-doc.
Such lines were probably not meant for kernel-doc parsing, but are parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warnings from kernel-doc.

E.g., presence of kernel-doc like comment in
drivers/crypto/amcc/crypto4xx_alg.c at header, and some other lines,
causes these warnings by kernel-doc:

"warning: expecting prototype for AMCC SoC PPC4xx Crypto Driver(). Prototype was for set_dynamic_sa_command_0() instead"
"warning: Function parameter or member 'dir' not described in 'set_dynamic_sa_command_0'"
etc..

Provide a simple fix by replacing such occurrences with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.

Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: vmx - fix incorrect kernel-doc comment syntax in files
Aditya Srivastava [Sat, 20 Mar 2021 20:25:25 +0000 (01:55 +0530)]
crypto: vmx - fix incorrect kernel-doc comment syntax in files

The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
There are certain files in drivers/crypto/vmx, which follow this syntax,
but the content inside does not comply with kernel-doc.
Such lines were probably not meant for kernel-doc parsing, but are parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warnings from kernel-doc.

E.g., presence of kernel-doc like comment in the header line for
drivers/crypto/vmx/vmx.c causes this warning by kernel-doc:

"warning: expecting prototype for Routines supporting VMX instructions on the Power 8(). Prototype was for p8_init() instead"

Similarly for other files too.

Provide a simple fix by replacing such occurrences with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.

Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Reviewed-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/qm - add queue isolation support for Kunpeng930
Weili Qian [Sat, 20 Mar 2021 11:27:46 +0000 (19:27 +0800)]
crypto: hisilicon/qm - add queue isolation support for Kunpeng930

Kunpeng930 supports doorbell isolation to ensure that each queue
has an independent doorbell address space.

Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/qm - set the number of queues for function
Weili Qian [Sat, 20 Mar 2021 11:27:45 +0000 (19:27 +0800)]
crypto: hisilicon/qm - set the number of queues for function

Kunpeng930 supports queue doorbell isolation.
When doorbell isolation is enabled, it supports to obtain the
maximum number of queues of one function from hardware register.
Otherwise, the 'max_qp_num' is the total number of queues.

When assigning queues to VF, it is necessary to ensure that the number
of VF queues does not exceed 'max_qp_num'.

Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/qm - move 'CURRENT_QM' code to qm.c
Weili Qian [Sat, 20 Mar 2021 11:27:44 +0000 (19:27 +0800)]
crypto: hisilicon/qm - move 'CURRENT_QM' code to qm.c

Since the code related to 'CURRENT_QM' debugfs is exactly same in
sec/hpre/zip driver, move 'CURRENT_QM' to qm.c to reduce duplicate code.

Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/qm - set the total number of queues
Weili Qian [Sat, 20 Mar 2021 11:27:43 +0000 (19:27 +0800)]
crypto: hisilicon/qm - set the total number of queues

Move the configuration of the total number of queues 'ctrl_qp_num'
from sec2/hpre/zip to qm.c. And get the total number of queues
from the hardware register for Kunpeng930.

Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - fix Kconfig
Hui Tang [Fri, 19 Mar 2021 10:45:39 +0000 (18:45 +0800)]
crypto: hisilicon/hpre - fix Kconfig

hpre select 'CRYPTO_ECDH' and 'CRYPTO_CURVE25519'.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - fix "hpre_ctx_init" resource leak
Hui Tang [Fri, 19 Mar 2021 10:45:27 +0000 (18:45 +0800)]
crypto: hisilicon/hpre - fix "hpre_ctx_init" resource leak

When calling "hpre_ctx_set" fails, stop and put qp,
otherwise will leak qp resource.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon - fix the check on dma address
Hui Tang [Fri, 19 Mar 2021 10:45:05 +0000 (18:45 +0800)]
crypto: hisilicon - fix the check on dma address

System may be able to get physical address of zero if not reserved by
firmware.

The dma address obtained by 'dma_alloc_coherent' is valid, since already
checking cpu va before, so do not check again.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - optimise 'hpre_algs_register' error path
Hui Tang [Fri, 19 Mar 2021 10:44:19 +0000 (18:44 +0800)]
crypto: hisilicon/hpre - optimise 'hpre_algs_register' error path

There is redundant code especially when registing new algorithms
in the future.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - delete wrap of 'CONFIG_CRYPTO_DH'
Hui Tang [Fri, 19 Mar 2021 10:44:18 +0000 (18:44 +0800)]
crypto: hisilicon/hpre - delete wrap of 'CONFIG_CRYPTO_DH'

'CRYPTO_DH' has selected in 'Kconfig', so delete 'CONFIG_CRYPTO_DH'.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecc - Correct an error in the comments
Meng Yu [Fri, 19 Mar 2021 09:13:34 +0000 (17:13 +0800)]
crypto: ecc - Correct an error in the comments

Remove repeated word 'bit' in comments.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qat - ADF_STATUS_PF_RUNNING should be set after adf_dev_init
Tong Zhang [Fri, 19 Mar 2021 03:40:00 +0000 (23:40 -0400)]
crypto: qat - ADF_STATUS_PF_RUNNING should be set after adf_dev_init

ADF_STATUS_PF_RUNNING is (only) used and checked by adf_vf2pf_shutdown()
before calling adf_iov_putmsg()->mutex_lock(vf2pf_lock), however the
vf2pf_lock is initialized in adf_dev_init(), which can fail and when it
fail, the vf2pf_lock is either not initialized or destroyed, a subsequent
use of vf2pf_lock will cause issue.
To fix this issue, only set this flag if adf_dev_init() returns 0.

[    7.178404] BUG: KASAN: user-memory-access in __mutex_lock.isra.0+0x1ac/0x7c0
[    7.180345] Call Trace:
[    7.182576]  mutex_lock+0xc9/0xd0
[    7.183257]  adf_iov_putmsg+0x118/0x1a0 [intel_qat]
[    7.183541]  adf_vf2pf_shutdown+0x4d/0x7b [intel_qat]
[    7.183834]  adf_dev_shutdown+0x172/0x2b0 [intel_qat]
[    7.184127]  adf_probe+0x5e9/0x600 [qat_dh895xccvf]

Signed-off-by: Tong Zhang <ztong0001@gmail.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Fixes: 25c6ffb249f6 ("crypto: qat - check if PF is running")
Acked-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qat - don't release uninitialized resources
Tong Zhang [Fri, 19 Mar 2021 03:39:59 +0000 (23:39 -0400)]
crypto: qat - don't release uninitialized resources

adf_vf_isr_resource_alloc() is not unwinding correctly when error
happens and it want to release uninitialized resources.
To fix this, only release initialized resources.

[    1.792845] Trying to free already-free IRQ 11
[    1.793091] WARNING: CPU: 0 PID: 182 at kernel/irq/manage.c:1821 free_irq+0x202/0x380
[    1.801340] Call Trace:
[    1.801477]  adf_vf_isr_resource_free+0x32/0xb0 [intel_qat]
[    1.801785]  adf_vf_isr_resource_alloc+0x14d/0x150 [intel_qat]
[    1.802105]  adf_dev_init+0xba/0x140 [intel_qat]

Signed-off-by: Tong Zhang <ztong0001@gmail.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Fixes: dd0f368398ea ("crypto: qat - Add qat dh895xcc VF driver")
Acked-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: nitrox - Demote non-compliant kernel-doc headers
Lee Jones [Thu, 18 Mar 2021 12:44:22 +0000 (12:44 +0000)]
crypto: nitrox - Demote non-compliant kernel-doc headers

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/cavium/nitrox/nitrox_isr.c:17: warning: expecting prototype for One vector for each type of ring(). Prototype was for NR_RING_VECTORS() instead
 drivers/crypto/cavium/nitrox/nitrox_isr.c:224: warning: Function parameter or member 'irq' not described in 'nps_core_int_isr'
 drivers/crypto/cavium/nitrox/nitrox_isr.c:224: warning: Function parameter or member 'data' not described in 'nps_core_int_isr'

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: nx - Repair some kernel-doc problems
Lee Jones [Thu, 18 Mar 2021 12:44:21 +0000 (12:44 +0000)]
crypto: nx - Repair some kernel-doc problems

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/nx/nx-aes-cbc.c:24: warning: Function parameter or member 'tfm' not described in 'cbc_aes_nx_set_key'
 drivers/crypto/nx/nx-aes-cbc.c:24: warning: Function parameter or member 'in_key' not described in 'cbc_aes_nx_set_key'
 drivers/crypto/nx/nx-aes-cbc.c:24: warning: Function parameter or member 'key_len' not described in 'cbc_aes_nx_set_key'
 drivers/crypto/nx/nx-aes-cbc.c:24: warning: expecting prototype for Nest Accelerators driver(). Prototype was for cbc_aes_nx_set_key() instead
 drivers/crypto/nx/nx_debugfs.c:34: warning: Function parameter or member 'drv' not described in 'nx_debugfs_init'
 drivers/crypto/nx/nx_debugfs.c:34: warning: expecting prototype for Nest Accelerators driver(). Prototype was for nx_debugfs_init() instead
 drivers/crypto/nx/nx.c:31: warning: Incorrect use of kernel-doc format:  * nx_hcall_sync - make an H_COP_OP hcall for the passed in op structure
 drivers/crypto/nx/nx.c:43: warning: Function parameter or member 'nx_ctx' not described in 'nx_hcall_sync'
 drivers/crypto/nx/nx.c:43: warning: Function parameter or member 'op' not described in 'nx_hcall_sync'
 drivers/crypto/nx/nx.c:43: warning: Function parameter or member 'may_sleep' not described in 'nx_hcall_sync'
 drivers/crypto/nx/nx.c:43: warning: expecting prototype for Nest Accelerators driver(). Prototype was for nx_hcall_sync() instead
 drivers/crypto/nx/nx.c:209: warning: Function parameter or member 'nbytes' not described in 'trim_sg_list'

Cc: "Breno Leitão" <leitao@debian.org>
Cc: Nayna Jain <nayna@linux.ibm.com>
Cc: Paulo Flabiano Smorigo <pfsmorigo@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Kent Yoder <yoder1@us.ibm.com>
Cc: linux-crypto@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: vmx - Source headers are not good kernel-doc candidates
Lee Jones [Thu, 18 Mar 2021 12:44:20 +0000 (12:44 +0000)]
crypto: vmx - Source headers are not good kernel-doc candidates

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/vmx/vmx.c:23: warning: expecting prototype for Routines supporting VMX instructions on the Power 8(). Prototype was for p8_init() instead

Cc: "Breno Leitão" <leitao@debian.org>
Cc: Nayna Jain <nayna@linux.ibm.com>
Cc: Paulo Flabiano Smorigo <pfsmorigo@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Henrique Cerri <mhcerri@br.ibm.com>
Cc: linux-crypto@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: caam - Provide the name of the function and provide missing descriptions
Lee Jones [Thu, 18 Mar 2021 12:44:19 +0000 (12:44 +0000)]
crypto: caam - Provide the name of the function and provide missing descriptions

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/caam/caampkc.c:199: warning: expecting prototype for from a given scatterlist(). Prototype was for caam_rsa_count_leading_zeros() instead
 drivers/crypto/caam/caamalg_qi2.c:87: warning: Function parameter or member 'xts_key_fallback' not described in 'caam_ctx'
 drivers/crypto/caam/caamalg_qi2.c:87: warning: Function parameter or member 'fallback' not described in 'caam_ctx'

Cc: "Horia Geantă" <horia.geanta@nxp.com>
Cc: Aymen Sghaier <aymen.sghaier@nxp.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Reviewed-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: atmel-ecc - Struct headers need to start with keyword 'struct'
Lee Jones [Thu, 18 Mar 2021 12:44:18 +0000 (12:44 +0000)]
crypto: atmel-ecc - Struct headers need to start with keyword 'struct'

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/atmel-ecc.c:41: warning: cannot understand function prototype: 'struct atmel_ecdh_ctx '

Cc: Tudor Ambarus <tudor.ambarus@microchip.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Nicolas Ferre <nicolas.ferre@microchip.com>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: Ludovic Desroches <ludovic.desroches@microchip.com>
Cc: linux-crypto@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Reviewed-by: Tudor Ambarus <tudor.ambarus@microchip.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: keembay - Fix incorrectly named functions/structs
Lee Jones [Thu, 18 Mar 2021 12:44:17 +0000 (12:44 +0000)]
crypto: keembay - Fix incorrectly named functions/structs

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/keembay/ocs-hcu.c:107: warning: expecting prototype for struct ocs_hcu_dma_list. Prototype was for struct ocs_hcu_dma_entry instead
 drivers/crypto/keembay/ocs-hcu.c:127: warning: expecting prototype for struct ocs_dma_list. Prototype was for struct ocs_hcu_dma_list instead
 drivers/crypto/keembay/ocs-hcu.c:610: warning: expecting prototype for ocs_hcu_digest(). Prototype was for ocs_hcu_hash_update() instead
 drivers/crypto/keembay/ocs-hcu.c:648: warning: expecting prototype for ocs_hcu_hash_final(). Prototype was for ocs_hcu_hash_finup() instead

Cc: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Cc: Declan Murphy <declan.murphy@intel.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Reviewed-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ux500/hash - Fix worthy kernel-doc headers and remove others
Lee Jones [Thu, 18 Mar 2021 12:44:16 +0000 (12:44 +0000)]
crypto: ux500/hash - Fix worthy kernel-doc headers and remove others

The ones remove here not only fail to conform to kernel-doc, but also
provide no value, so let's remove them completely in this case.

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/ux500/hash/hash_core.c:368: warning: Function parameter or member 'ctx' not described in 'hash_get_device_data'
 drivers/crypto/ux500/hash/hash_core.c:368: warning: Excess function parameter 'hash_ctx' description in 'hash_get_device_data'
 drivers/crypto/ux500/hash/hash_core.c:551: warning: expecting prototype for hash_init(). Prototype was for ux500_hash_init() instead
 drivers/crypto/ux500/hash/hash_core.c:592: warning: Function parameter or member 'length' not described in 'hash_processblock'
 drivers/crypto/ux500/hash/hash_core.c:1302: warning: expecting prototype for hash_update(). Prototype was for ahash_update() instead
 drivers/crypto/ux500/hash/hash_core.c:1322: warning: expecting prototype for hash_final(). Prototype was for ahash_final() instead
 drivers/crypto/ux500/hash/hash_core.c:1622: warning: Function parameter or member 'device_data' not described in 'ahash_algs_register_all'
 drivers/crypto/ux500/hash/hash_core.c:1622: warning: expecting prototype for hash_algs_register_all(). Prototype was for ahash_algs_register_all() instead
 drivers/crypto/ux500/hash/hash_core.c:1647: warning: Function parameter or member 'device_data' not described in 'ahash_algs_unregister_all'
 drivers/crypto/ux500/hash/hash_core.c:1647: warning: expecting prototype for hash_algs_unregister_all(). Prototype was for ahash_algs_unregister_all() instead
 drivers/crypto/ux500/cryp/cryp.c:19: warning: Incorrect use of kernel-doc format:  * cryp_wait_until_done - wait until the device logic is not busy
 drivers/crypto/ux500/cryp/cryp.c:22: warning: Function parameter or member 'device_data' not described in 'cryp_wait_until_done'
 drivers/crypto/ux500/cryp/cryp.c:22: warning: expecting prototype for ST(). Prototype was for cryp_wait_until_done() instead
 drivers/crypto/ux500/cryp/cryp.c:292: warning: Function parameter or member 'cryp_mode' not described in 'cryp_save_device_context'
 drivers/crypto/ux500/cryp/cryp_irq.c:21: warning: Function parameter or member 'device_data' not described in 'cryp_enable_irq_src'
 drivers/crypto/ux500/cryp/cryp_irq.c:21: warning: Function parameter or member 'irq_src' not described in 'cryp_enable_irq_src'
 drivers/crypto/ux500/cryp/cryp_irq.c:21: warning: expecting prototype for ST(). Prototype was for cryp_enable_irq_src() instead
 drivers/crypto/ux500/cryp/cryp_core.c:42: warning: expecting prototype for ST(). Prototype was for CRYP_MAX_KEY_SIZE() instead
 drivers/crypto/ux500/cryp/cryp_core.c:91: warning: Function parameter or member 'key' not described in 'cryp_ctx'
 drivers/crypto/ux500/cryp/cryp_core.c:91: warning: Function parameter or member 'session_id' not described in 'cryp_ctx'

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Shujuan Chen <shujuan.chen@stericsson.com>
Cc: Joakim Bech <joakim.xx.bech@stericsson.com>
Cc: Berne Hebark <berne.herbark@stericsson.com>
Cc: Niklas Hernaeus <niklas.hernaeus@stericsson.com>
Cc: Jonas Linde <jonas.linde@stericsson.com>
Cc: Andreas Westin <andreas.westin@stericsson.com>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: chelsio - Fix some kernel-doc issues
Lee Jones [Thu, 18 Mar 2021 12:44:15 +0000 (12:44 +0000)]
crypto: chelsio - Fix some kernel-doc issues

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/chelsio/chcr_core.c:2: warning: wrong kernel-doc identifier on line:
 drivers/crypto/chelsio/chcr_algo.c:806: warning: Function parameter or member 'wrparam' not described in 'create_cipher_wr'
 drivers/crypto/chelsio/chcr_algo.c:806: warning: Excess function parameter 'req' description in 'create_cipher_wr'
 drivers/crypto/chelsio/chcr_algo.c:806: warning: Excess function parameter 'ctx' description in 'create_cipher_wr'
 drivers/crypto/chelsio/chcr_algo.c:806: warning: Excess function parameter 'qid' description in 'create_cipher_wr'
 drivers/crypto/chelsio/chcr_algo.c:806: warning: Excess function parameter 'op_type' description in 'create_cipher_wr'
 drivers/crypto/chelsio/chcr_algo.c:1566: warning: Function parameter or member 'req' not described in 'create_hash_wr'
 drivers/crypto/chelsio/chcr_algo.c:1566: warning: Function parameter or member 'param' not described in 'create_hash_wr'

Cc: Ayush Sawal <ayush.sawal@chelsio.com>
Cc: Vinay Kumar Yadav <vinay.yadav@chelsio.com>
Cc: Rohit Maheshwari <rohitm@chelsio.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Manoj Malviya <manojmalviya@chelsio.com>
Cc: Atul Gupta <atul.gupta@chelsio.com>
Cc: Jitendra Lulla <jlulla@chelsio.com>
Cc: M R Gowda <yeshaswi@chelsio.com>
Cc: Harsh Jain <harsh@chelsio.com>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: bcm - Fix a whole host of kernel-doc misdemeanours
Lee Jones [Thu, 18 Mar 2021 12:44:14 +0000 (12:44 +0000)]
crypto: bcm - Fix a whole host of kernel-doc misdemeanours

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/bcm/util.c:136: warning: Function parameter or member 'from_nents' not described in 'spu_msg_sg_add'
 drivers/crypto/bcm/util.c:136: warning: Function parameter or member 'length' not described in 'spu_msg_sg_add'
 drivers/crypto/bcm/spu.c:464: warning: Excess function parameter 'Return' description in 'spum_gcm_ccm_pad_len'
 drivers/crypto/bcm/spu.c:524: warning: Function parameter or member 'iv_len' not described in 'spum_aead_ivlen'
 drivers/crypto/bcm/spu.c:524: warning: expecting prototype for spu_aead_ivlen(). Prototype was for spum_aead_ivlen() instead
 drivers/crypto/bcm/spu.c:556: warning: Function parameter or member 'alg_digest_size' not described in 'spum_digest_size'
 drivers/crypto/bcm/spu.c:556: warning: Function parameter or member 'alg' not described in 'spum_digest_size'
 drivers/crypto/bcm/spu.c:556: warning: Function parameter or member 'htype' not described in 'spum_digest_size'
 drivers/crypto/bcm/spu.c:583: warning: bad line:
 drivers/crypto/bcm/spu.c:927: warning: Function parameter or member 'is_inbound' not described in 'spum_cipher_req_finish'
 drivers/crypto/bcm/spu.c:927: warning: Excess function parameter 'isInbound' description in 'spum_cipher_req_finish'
 drivers/crypto/bcm/spu2.c:557: warning: Function parameter or member 'fmd' not described in 'spu2_fmd_init'
 drivers/crypto/bcm/spu2.c:557: warning: Function parameter or member 'spu2_type' not described in 'spu2_fmd_init'
 drivers/crypto/bcm/spu2.c:557: warning: Excess function parameter 'spu2_cipher_type' description in 'spu2_fmd_init'
 drivers/crypto/bcm/spu2.c:615: warning: Function parameter or member 'auth_first' not described in 'spu2_fmd_ctrl0_write'
 drivers/crypto/bcm/spu2.c:615: warning: Excess function parameter 'authFirst' description in 'spu2_fmd_ctrl0_write'
 drivers/crypto/bcm/spu2.c:666: warning: Function parameter or member 'is_inbound' not described in 'spu2_fmd_ctrl1_write'
 drivers/crypto/bcm/spu2.c:803: warning: expecting prototype for spu_payload_length(). Prototype was for spu2_payload_length() instead
 drivers/crypto/bcm/spu2.c:825: warning: Function parameter or member 'is_hash' not described in 'spu2_response_hdr_len'
 drivers/crypto/bcm/spu2.c:825: warning: expecting prototype for spu_response_hdr_len(). Prototype was for spu2_response_hdr_len() instead
 drivers/crypto/bcm/spu2.c:843: warning: expecting prototype for spu_hash_pad_len(). Prototype was for spu2_hash_pad_len() instead
 drivers/crypto/bcm/spu2.c:855: warning: Function parameter or member 'cipher_mode' not described in 'spu2_gcm_ccm_pad_len'
 drivers/crypto/bcm/spu2.c:855: warning: Function parameter or member 'data_size' not described in 'spu2_gcm_ccm_pad_len'
 drivers/crypto/bcm/spu2.c:855: warning: expecting prototype for spu2_gcm_ccm_padlen(). Prototype was for spu2_gcm_ccm_pad_len() instead
 drivers/crypto/bcm/spu2.c:872: warning: expecting prototype for spu_assoc_resp_len(). Prototype was for spu2_assoc_resp_len() instead
 drivers/crypto/bcm/spu2.c:919: warning: Function parameter or member 'alg_digest_size' not described in 'spu2_digest_size'
 drivers/crypto/bcm/spu2.c:919: warning: Function parameter or member 'alg' not described in 'spu2_digest_size'
 drivers/crypto/bcm/spu2.c:919: warning: Function parameter or member 'htype' not described in 'spu2_digest_size'
 drivers/crypto/bcm/spu2.c:945: warning: expecting prototype for spu_create_request(). Prototype was for spu2_create_request() instead
 drivers/crypto/bcm/spu2.c:1122: warning: expecting prototype for spu_cipher_req_init(). Prototype was for spu2_cipher_req_init() instead
 drivers/crypto/bcm/spu2.c:1182: warning: Function parameter or member 'is_inbound' not described in 'spu2_cipher_req_finish'
 drivers/crypto/bcm/spu2.c:1182: warning: expecting prototype for spu_cipher_req_finish(). Prototype was for spu2_cipher_req_finish() instead
 drivers/crypto/bcm/spu2.c:1243: warning: expecting prototype for spu_request_pad(). Prototype was for spu2_request_pad() instead
 drivers/crypto/bcm/spu2.c:1321: warning: expecting prototype for spu_status_process(). Prototype was for spu2_status_process() instead
 drivers/crypto/bcm/cipher.c:1048: warning: Function parameter or member 'req' not described in 'spu_aead_rx_sg_create'
 drivers/crypto/bcm/cipher.c:2966: warning: Function parameter or member 'cipher' not described in 'rfc4543_gcm_esp_setkey'
 drivers/crypto/bcm/cipher.c:2966: warning: Function parameter or member 'key' not described in 'rfc4543_gcm_esp_setkey'
 drivers/crypto/bcm/cipher.c:2966: warning: Function parameter or member 'keylen' not described in 'rfc4543_gcm_esp_setkey'

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Rob Rice <rob.rice@broadcom.com>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/sec - Supply missing description for 'sec_queue_empty()'s 'queue...
Lee Jones [Thu, 18 Mar 2021 12:44:13 +0000 (12:44 +0000)]
crypto: hisilicon/sec - Supply missing description for 'sec_queue_empty()'s 'queue' param

Fixes the following W=1 kernel build warning(s):

 drivers/crypto/hisilicon/sec/sec_drv.c:843: warning: Function parameter or member 'queue' not described in 'sec_queue_empty'

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Zaibo Xu <xuzaibo@huawei.com>
Cc: Jonathan Cameron <jonathan.cameron@huawei.com>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: inside-secure - Minor typo fix in the file safexcel.c
Bhaskar Chowdhury [Wed, 17 Mar 2021 09:14:45 +0000 (14:44 +0530)]
crypto: inside-secure - Minor typo fix in the file safexcel.c

s/procesing/processing/

Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Acked-by: Antoine Tenart <atenart@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: jitterentropy - Put constants on the right side of the expression
Milan Djurovic [Wed, 17 Mar 2021 01:44:03 +0000 (18:44 -0700)]
crypto: jitterentropy - Put constants on the right side of the expression

This patch fixes the following checkpatch.pl warnings:

crypto/jitterentropy.c:600: WARNING: Comparisons should place the constant on the right side of the test
crypto/jitterentropy.c:681: WARNING: Comparisons should place the constant on the right side of the test
crypto/jitterentropy.c:772: WARNING: Comparisons should place the constant on the right side of the test
crypto/jitterentropy.c:829: WARNING: Comparisons should place the constant on the right side of the test

Signed-off-by: Milan Djurovic <mdjurovic@zohomail.com>
Reviewed-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: core - convert sysfs sprintf/snprintf family to sysfs_emit
Zihao Tang [Tue, 16 Mar 2021 12:34:12 +0000 (20:34 +0800)]
hwrng: core - convert sysfs sprintf/snprintf family to sysfs_emit

Fix the following coccicheck warning:

drivers/char/hw_random/core.c:399:8-16: WARNING: use scnprintf or sprintf.

Signed-off-by: Zihao Tang <tangzihao1@hisilicon.com>
Signed-off-by: Jay Fang <f.fangjian@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: allwinner - Fix the parameter of dma_unmap_sg()
Xiang Chen [Tue, 16 Mar 2021 01:55:26 +0000 (09:55 +0800)]
crypto: allwinner - Fix the parameter of dma_unmap_sg()

For function dma_unmap_sg(), the <nents> parameter should be number of
elements in the scatterlist prior to the mapping, not after the mapping.
So fix this usage.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Acked-by: Corentin LABBE <clabbe.montjoie@gmail.com>
Tested-by: Corentin LABBE <clabbe.montjoie@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ux500 - Fix the parameter of dma_unmap_sg()
Xiang Chen [Tue, 16 Mar 2021 01:55:25 +0000 (09:55 +0800)]
crypto: ux500 - Fix the parameter of dma_unmap_sg()

For function dma_unmap_sg(), the <nents> parameter should be number of
elements in the scatterlist prior to the mapping, not after the mapping.
So fix this usage.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: cavium - Fix the parameter of dma_unmap_sg()
Xiang Chen [Tue, 16 Mar 2021 01:55:24 +0000 (09:55 +0800)]
crypto: cavium - Fix the parameter of dma_unmap_sg()

For function dma_unmap_sg(), the <nents> parameter should be number of
elements in the scatterlist prior to the mapping, not after the mapping.
So fix this usage.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: amlogic - Fix the parameter of dma_unmap_sg()
Xiang Chen [Tue, 16 Mar 2021 01:55:23 +0000 (09:55 +0800)]
crypto: amlogic - Fix the parameter of dma_unmap_sg()

For function dma_unmap_sg(), the <nents> parameter should be number of
elements in the scatterlist prior to the mapping, not after the mapping.
So fix this usage.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ccp - fix command queuing to TEE ring buffer
Rijo Thomas [Mon, 15 Mar 2021 08:25:29 +0000 (13:55 +0530)]
crypto: ccp - fix command queuing to TEE ring buffer

Multiple threads or clients can submit a command to the TEE ring
buffer. This patch helps to synchronize command submission to the
ring.

One thread shall write a command to a TEE ring buffer entry only if:

 - Trusted OS has notified that the TEE command for the given entry
   has been processed and driver has copied the TEE response into
   client buffer.

 - The command entry is empty and can be written into.

After a command has been written to the TEE ring buffer, the global
wptr (mutex protected) shall be incremented for use by next client.

If PSP became unresponsive while processing TEE request from a
client, then further command submission to queue will be disabled.

Fixes: 33960acccfbd (crypto: ccp - add TEE support for Raven Ridge)
Reviewed-by: Devaraj Rangasamy <Devaraj.Rangasamy@amd.com>
Signed-off-by: Rijo Thomas <Rijo-john.Thomas@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ccp - reduce tee command status polling interval from 5ms to 1ms
Rijo Thomas [Mon, 15 Mar 2021 08:25:28 +0000 (13:55 +0530)]
crypto: ccp - reduce tee command status polling interval from 5ms to 1ms

The PSP TEE device driver polls the command status variable every
5ms to check for command completion. Reduce this time to 1ms so that
there is an improvement in driver response time to clients which submit
TEE commands.

Reviewed-by: Devaraj Rangasamy <Devaraj.Rangasamy@amd.com>
Signed-off-by: Rijo Thomas <Rijo-john.Thomas@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: intel - Fix included header from 'asm
Tian Tao [Mon, 15 Mar 2021 06:12:04 +0000 (14:12 +0800)]
hwrng: intel - Fix included header from 'asm

This commit fixes the checkpatch warning:
WARNING: Use #include <linux/io.h> instead of <asm/io.h>
34: FILE: drivers/char/hw_random/intel-rng.c:34:

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agoMerge branch 'ecc'
Herbert Xu [Fri, 26 Mar 2021 08:55:55 +0000 (19:55 +1100)]
Merge branch 'ecc'

This pulls in the NIST P384/256/192 x509 changes.

3 years agox509: Add OID for NIST P384 and extend parser for it
Saulo Alessandre [Tue, 16 Mar 2021 21:07:39 +0000 (17:07 -0400)]
x509: Add OID for NIST P384 and extend parser for it

Prepare the x509 parser to accept NIST P384 certificates and add the
OID for ansip384r1, which is the identifier for NIST P384.

Summary of changes:

* crypto/asymmetric_keys/x509_cert_parser.c
  - prepare x509 parser to load NIST P384

* include/linux/oid_registry.h
  - add OID_ansip384r1

Signed-off-by: Saulo Alessandre <saulo.alessandre@tse.jus.br>
Tested-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agoima: Support EC keys for signature verification
Stefan Berger [Tue, 16 Mar 2021 21:07:38 +0000 (17:07 -0400)]
ima: Support EC keys for signature verification

Add support for IMA signature verification for EC keys. Since SHA type
of hashes can be used by RSA and ECDSA signature schemes we need to
look at the key and derive from the key which signature scheme to use.
Since this can be applied to all types of keys, we change the selection
of the encoding type to be driven by the key's signature scheme rather
than by the hash type.

Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com>
Cc: linux-integrity@vger.kernel.org
Cc: David Howells <dhowells@redhat.com>
Cc: keyrings@vger.kernel.org
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Vitaly Chikunov <vt@altlinux.org>
Reviewed-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Acked-by: Mimi Zohar <zohar@linux.ibm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agox509: Add support for parsing x509 certs with ECDSA keys
Stefan Berger [Tue, 16 Mar 2021 21:07:37 +0000 (17:07 -0400)]
x509: Add support for parsing x509 certs with ECDSA keys

Add support for parsing of x509 certificates that contain ECDSA keys,
such as NIST P256, that have been signed by a CA using any of the
current SHA hash algorithms.

Cc: David Howells <dhowells@redhat.com>
Cc: keyrings@vger.kernel.org
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agox509: Detect sm2 keys by their parameters OID
Stefan Berger [Tue, 16 Mar 2021 21:07:36 +0000 (17:07 -0400)]
x509: Detect sm2 keys by their parameters OID

Detect whether a key is an sm2 type of key by its OID in the parameters
array rather than assuming that everything under OID_id_ecPublicKey
is sm2, which is not the case.

Cc: David Howells <dhowells@redhat.com>
Cc: keyrings@vger.kernel.org
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Tested-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecdsa - Register NIST P384 and extend test suite
Saulo Alessandre [Tue, 16 Mar 2021 21:07:35 +0000 (17:07 -0400)]
crypto: ecdsa - Register NIST P384 and extend test suite

Register NIST P384 as an akcipher and extend the testmgr with
NIST P384-specific test vectors.

Summary of changes:

* crypto/ecdsa.c
  - add ecdsa_nist_p384_init_tfm
  - register and unregister P384 tfm

* crypto/testmgr.c
  - add test vector for P384 on vector of tests

* crypto/testmgr.h
  - add test vector params for P384(sha1, sha224, sha256, sha384
    and sha512)

Signed-off-by: Saulo Alessandre <saulo.alessandre@tse.jus.br>
Tested-by: Stefan Berger <stefanb@linux.ibm.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecc - Add math to support fast NIST P384
Saulo Alessandre [Tue, 16 Mar 2021 21:07:34 +0000 (17:07 -0400)]
crypto: ecc - Add math to support fast NIST P384

Add the math needed for NIST P384 and adapt certain functions'
parameters so that the ecc_curve is passed to vli_mmod_fast. This
allows to identify the curve by its name prefix and the appropriate
function for fast mmod calculation can be used.

Summary of changes:

* crypto/ecc.c
  - add vli_mmod_fast_384
  - change some routines to pass ecc_curve forward until vli_mmod_fast

* crypto/ecc.h
  - add ECC_CURVE_NIST_P384_DIGITS
  - change ECC_MAX_DIGITS to P384 size

Signed-off-by: Saulo Alessandre <saulo.alessandre@tse.jus.br>
Tested-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecc - Add NIST P384 curve parameters
Saulo Alessandre [Tue, 16 Mar 2021 21:07:33 +0000 (17:07 -0400)]
crypto: ecc - Add NIST P384 curve parameters

Add the parameters for the NIST P384 curve and define a new curve ID
for it. Make the curve available in ecc_get_curve.

Summary of changes:

* crypto/ecc_curve_defs.h
  - add nist_p384 params

* include/crypto/ecdh.h
  - add ECC_CURVE_NIST_P384

* crypto/ecc.c
  - change ecc_get_curve to accept nist_p384

Signed-off-by: Saulo Alessandre <saulo.alessandre@tse.jus.br>
Tested-by: Stefan Berger <stefanb@linux.ibm.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecdsa - Add support for ECDSA signature verification
Stefan Berger [Tue, 16 Mar 2021 21:07:32 +0000 (17:07 -0400)]
crypto: ecdsa - Add support for ECDSA signature verification

Add support for parsing the parameters of a NIST P256 or NIST P192 key.
Enable signature verification using these keys. The new module is
enabled with CONFIG_ECDSA:
  Elliptic Curve Digital Signature Algorithm (NIST P192, P256 etc.)
  is A NIST cryptographic standard algorithm. Only signature verification
  is implemented.

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-crypto@vger.kernel.org
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agooid_registry: Add OIDs for ECDSA with SHA224/256/384/512
Stefan Berger [Tue, 16 Mar 2021 21:07:31 +0000 (17:07 -0400)]
oid_registry: Add OIDs for ECDSA with SHA224/256/384/512

Add OIDs for ECDSA with SHA224/256/384/512.

Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: cctrng - delete redundant printing of return value
Wang Qing [Sat, 13 Mar 2021 07:47:42 +0000 (15:47 +0800)]
hwrng: cctrng - delete redundant printing of return value

platform_get_irq() has already checked and printed the return value,
the printing here is nothing special, it is not necessary at all.

Signed-off-by: Wang Qing <wangqing@vivo.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/sec - fixes some driver coding style
Longfang Liu [Sat, 13 Mar 2021 07:28:25 +0000 (15:28 +0800)]
crypto: hisilicon/sec - fixes some driver coding style

cleanup static check errors for SEC

Signed-off-by: Longfang Liu <liulongfang@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/sec - fixes some coding style
Longfang Liu [Sat, 13 Mar 2021 07:28:24 +0000 (15:28 +0800)]
crypto: hisilicon/sec - fixes some coding style

1.delete the original complex method of obtaining the
current device and replace it with the initialized
device pointer.
2.fixes some coding style

Signed-off-by: Longfang Liu <liulongfang@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/sec - fixes a printing error
Longfang Liu [Sat, 13 Mar 2021 07:28:23 +0000 (15:28 +0800)]
crypto: hisilicon/sec - fixes a printing error

When the log is output here, the device has not
been initialized yet.

Signed-off-by: Longfang Liu <liulongfang@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: ba431 - use devm_platform_ioremap_resource() to simplify
Tian Tao [Sat, 13 Mar 2021 01:47:38 +0000 (09:47 +0800)]
hwrng: ba431 - use devm_platform_ioremap_resource() to simplify

Use devm_platform_ioremap_resource() to simplify the code.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: cctrng - use devm_platform_ioremap_resource() to simplify
Tian Tao [Sat, 13 Mar 2021 01:42:35 +0000 (09:42 +0800)]
hwrng: cctrng - use devm_platform_ioremap_resource() to simplify

Use devm_platform_ioremap_resource() to simplify the code.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: arm/chacha-scalar - switch to common rev_l macro
Ard Biesheuvel [Wed, 10 Mar 2021 10:14:21 +0000 (11:14 +0100)]
crypto: arm/chacha-scalar - switch to common rev_l macro

Drop the local definition of a byte swapping macro and use the common
one instead.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Nicolas Pitre <nico@fluxnic.net>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: arm/aes-scalar - switch to common rev_l/mov_l macros
Ard Biesheuvel [Wed, 10 Mar 2021 10:14:20 +0000 (11:14 +0100)]
crypto: arm/aes-scalar - switch to common rev_l/mov_l macros

The scalar AES implementation has some locally defined macros which
reimplement things that are now available in macros defined in
assembler.h. So let's switch to those.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Nicolas Pitre <nico@fluxnic.net>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: arm/blake2s - fix for big endian
Eric Biggers [Wed, 10 Mar 2021 07:27:26 +0000 (23:27 -0800)]
crypto: arm/blake2s - fix for big endian

The new ARM BLAKE2s code doesn't work correctly (fails the self-tests)
in big endian kernel builds because it doesn't swap the endianness of
the message words when loading them.  Fix this.

Fixes: 5172d322d34c ("crypto: arm/blake2s - add ARM scalar optimized BLAKE2s")
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: aegis128 - Move simd prototypes into aegis.h
Herbert Xu [Mon, 8 Mar 2021 05:41:32 +0000 (16:41 +1100)]
crypto: aegis128 - Move simd prototypes into aegis.h

This patch fixes missing prototype warnings in crypto/aegis128-neon.c.

Fixes: a4397635afea ("crypto: aegis128 - provide a SIMD...")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: sun4i-ss - simplify optional reset handling
Philipp Zabel [Fri, 5 Mar 2021 09:12:36 +0000 (10:12 +0100)]
crypto: sun4i-ss - simplify optional reset handling

As of commit bb475230b8e5 ("reset: make optional functions really
optional"), the reset framework API calls use NULL pointers to describe
optional, non-present reset controls.

This allows to unconditionally return errors from
devm_reset_control_get_optional_exclusive.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: bcm2835 - add reset support
Álvaro Fernández Rojas [Fri, 5 Mar 2021 07:01:32 +0000 (08:01 +0100)]
hwrng: bcm2835 - add reset support

BCM6368 devices need to reset the IPSEC controller in order to generate true
random numbers.

This is what BCM6368 produces without a reset:
root@OpenWrt:/# cat /dev/hwrng | rngtest -c 1000
rngtest 6.10
Copyright (c) 2004 by Henrique de Moraes Holschuh
This is free software; see the source for copying conditions.  There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

rngtest: starting FIPS tests...
rngtest: bits received from input: 20000032
rngtest: FIPS 140-2 successes: 0
rngtest: FIPS 140-2 failures: 1000
rngtest: FIPS 140-2(2001-10-10) Monobit: 2
rngtest: FIPS 140-2(2001-10-10) Poker: 1000
rngtest: FIPS 140-2(2001-10-10) Runs: 1000
rngtest: FIPS 140-2(2001-10-10) Long run: 30
rngtest: FIPS 140-2(2001-10-10) Continuous run: 0
rngtest: input channel speed: (min=37.253; avg=320.827; max=635.783)Mibits/s
rngtest: FIPS tests speed: (min=12.141; avg=15.034; max=16.428)Mibits/s
rngtest: Program run time: 1336176 microseconds

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agodt-bindings: rng: bcm2835: document reset support
Álvaro Fernández Rojas [Fri, 5 Mar 2021 07:01:31 +0000 (08:01 +0100)]
dt-bindings: rng: bcm2835: document reset support

brcm,bcm6368-rng controllers require resetting the IPSEC clock in order to get
a functional RNG.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agodt-bindings: rng: bcm2835: add clock constraints
Álvaro Fernández Rojas [Fri, 5 Mar 2021 07:01:30 +0000 (08:01 +0100)]
dt-bindings: rng: bcm2835: add clock constraints

brcm,bcm6368-rng controllers require enabling the IPSEC clock in order to get
a functional RNG.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qat - fix use of 'dma_map_single'
Hui Tang [Fri, 5 Mar 2021 06:35:02 +0000 (14:35 +0800)]
crypto: qat - fix use of 'dma_map_single'

DMA_TO_DEVICE synchronisation must be done after the last modification
of the memory region by the software and before it is handed off to
the device.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qat - fix unmap invalid dma address
Hui Tang [Fri, 5 Mar 2021 06:35:01 +0000 (14:35 +0800)]
crypto: qat - fix unmap invalid dma address

'dma_mapping_error' return a negative value if 'dma_addr' is equal to
'DMA_MAPPING_ERROR' not zero, so fix initialization of 'dma_addr'.

Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: sun8i-ce - fix error return code in sun8i_ce_prng_generate()
Jia-Ju Bai [Fri, 5 Mar 2021 01:57:17 +0000 (17:57 -0800)]
crypto: sun8i-ce - fix error return code in sun8i_ce_prng_generate()

When dma_mapping_error() returns an error, no error return code of
sun8i_ce_prng_generate() is assigned.
To fix this bug, err is assigned with -EFAULT as error return code.

Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Tested-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Acked-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - add 'CURVE25519' algorithm
Meng Yu [Thu, 4 Mar 2021 06:35:50 +0000 (14:35 +0800)]
crypto: hisilicon/hpre - add 'CURVE25519' algorithm

Enable 'CURVE25519' algorithm in Kunpeng 930.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecc - add curve25519 params and expose them
Meng Yu [Thu, 4 Mar 2021 06:35:49 +0000 (14:35 +0800)]
crypto: ecc - add curve25519 params and expose them

1. Add curve 25519 parameters in 'crypto/ecc_curve_defs.h';
2. Add curve25519 interface 'ecc_get_curve25519_param' in
   'include/crypto/ecc_curve.h', to make its parameters be
   exposed to everyone in kernel tree.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - add 'ECDH' algorithm
Meng Yu [Thu, 4 Mar 2021 06:35:48 +0000 (14:35 +0800)]
crypto: hisilicon/hpre - add 'ECDH' algorithm

1. Enable 'ECDH' algorithm in Kunpeng 930;
2. HPRE ECDH Support: ecdh-nist-p192, ecdh-nist-p256.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecc - expose ecc curves
Meng Yu [Thu, 4 Mar 2021 06:35:47 +0000 (14:35 +0800)]
crypto: ecc - expose ecc curves

Move 'ecc_get_curve' to 'include/crypto/ecc_curve.h', so everyone
in kernel tree can easily get ecc curve params;

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ecdh - move curve_id of ECDH from the key to algorithm name
Meng Yu [Thu, 4 Mar 2021 06:35:46 +0000 (14:35 +0800)]
crypto: ecdh - move curve_id of ECDH from the key to algorithm name

1. crypto and crypto/atmel-ecc:
   Move curve id of ECDH from the key into the algorithm name instead
   in crypto and atmel-ecc, so ECDH algorithm name change form 'ecdh'
   to 'ecdh-nist-pxxx', and we cannot use 'curve_id' in 'struct ecdh';
2. crypto/testmgr and net/bluetooth:
   Modify 'testmgr.c', 'testmgr.h' and 'net/bluetooth' to adapt
   the modification.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - add algorithm type
Meng Yu [Thu, 4 Mar 2021 06:35:45 +0000 (14:35 +0800)]
crypto: hisilicon/hpre - add algorithm type

Algorithm type is brought in to get hardware HPRE queue
to support different algorithms.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: hisilicon/hpre - add version adapt to new algorithms
Meng Yu [Thu, 4 Mar 2021 06:35:44 +0000 (14:35 +0800)]
crypto: hisilicon/hpre - add version adapt to new algorithms

A new generation of accelerator Kunpeng930 has appeared, and the
corresponding driver needs to be updated to support some new
algorithms of Kunpeng930. To be compatible with Kunpeng920, we
add parameter 'struct hisi_qm *qm' to sec_algs_(un)register to
identify the chip's version.

Signed-off-by: Meng Yu <yumeng18@huawei.com>
Reviewed-by: Zaibo Xu <xuzaibo@huawei.com>
Reviewed-by: Longfang Liu <liulongfang@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: ccp - Don't initialize SEV support without the SEV feature
Tom Lendacky [Wed, 3 Mar 2021 22:31:09 +0000 (16:31 -0600)]
crypto: ccp - Don't initialize SEV support without the SEV feature

If SEV has been disabled (e.g. through BIOS), the driver probe will still
issue SEV firmware commands. The SEV INIT firmware command will return an
error in this situation, but the error code is a general error code that
doesn't highlight the exact reason.

Add a check for X86_FEATURE_SEV in sev_dev_init() and emit a meaningful
message and skip attempting to initialize the SEV firmware if the feature
is not enabled. Since building the SEV code is dependent on X86_64, adding
the check won't cause any build problems.

Cc: John Allen <john.allen@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Reviewed-By: Brijesh Singh <brijesh.singh@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: api - check for ERR pointers in crypto_destroy_tfm()
Ard Biesheuvel [Tue, 2 Mar 2021 20:33:03 +0000 (21:33 +0100)]
crypto: api - check for ERR pointers in crypto_destroy_tfm()

Given that crypto_alloc_tfm() may return ERR pointers, and to avoid
crashes on obscure error paths where such pointers are presented to
crypto_destroy_tfm() (such as [0]), add an ERR_PTR check there
before dereferencing the second argument as a struct crypto_tfm
pointer.

[0] https://lore.kernel.org/linux-crypto/000000000000de949705bc59e0f6@google.com/

Reported-by: syzbot+12cf5fbfdeba210a89dd@syzkaller.appspotmail.com
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: sun8i-ss - Fix memory leak of object d when dma_iv fails to map
Colin Ian King [Tue, 2 Mar 2021 16:34:46 +0000 (16:34 +0000)]
crypto: sun8i-ss - Fix memory leak of object d when dma_iv fails to map

In the case where the dma_iv mapping fails, the return error path leaks
the memory allocated to object d.  Fix this by adding a new error return
label and jumping to this to ensure d is free'd before the return.

Addresses-Coverity: ("Resource leak")
Fixes: ac2614d721de ("crypto: sun8i-ss - Add support for the PRNG")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Tested-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Acked-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: cavium - remove unused including <linux/version.h>
Tian Tao [Mon, 1 Mar 2021 03:54:12 +0000 (11:54 +0800)]
crypto: cavium - remove unused including <linux/version.h>

Remove including <linux/version.h> that don't need it.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: pic32 - Use device-managed registration API
Tian Tao [Mon, 1 Mar 2021 03:24:53 +0000 (11:24 +0800)]
hwrng: pic32 - Use device-managed registration API

Use devm_hwrng_register to get rid of manual unregistration.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: cctrng - Use device-managed registration API
Tian Tao [Mon, 1 Mar 2021 02:43:48 +0000 (10:43 +0800)]
hwrng: cctrng - Use device-managed registration API

Use devm_hwrng_register to get rid of manual unregistration.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: nx - add missing call to of_node_put()
Yang Li [Fri, 26 Feb 2021 01:23:06 +0000 (09:23 +0800)]
crypto: nx - add missing call to of_node_put()

In one of the error paths of the for_each_child_of_node() loop,
add missing call to of_node_put().

Fix the following coccicheck warning:
./drivers/crypto/nx/nx-common-powernv.c:927:1-23: WARNING: Function
"for_each_child_of_node" should have of_node_put() before return around
line 936.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: testmgr - delete some redundant code
Kai Ye [Tue, 23 Feb 2021 03:42:04 +0000 (11:42 +0800)]
crypto: testmgr - delete some redundant code

Delete sg_data function, because sg_data function definition same as
sg_virt(), so need to delete it and use sg_virt() replace to sg_data().

Signed-off-by: Kai Ye <yekai13@huawei.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: omap - Fix included header from 'asm'
Tian Tao [Fri, 19 Feb 2021 11:19:18 +0000 (19:19 +0800)]
hwrng: omap - Fix included header from 'asm'

This commit fixes the checkpatch warning:
WARNING: Use #include <linux/io.h> instead of <asm/io.h>
drivers/char/hw_random/omap-rng.c:34

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: s5p-sss - initialize APB clock after the AXI bus clock for SlimSSS
Krzysztof Kozlowski [Fri, 12 Feb 2021 16:35:26 +0000 (17:35 +0100)]
crypto: s5p-sss - initialize APB clock after the AXI bus clock for SlimSSS

The driver for Slim Security Subsystem (SlimSSS) on Exynos5433 takes two
clocks - aclk (AXI/AHB clock) and pclk (APB/Advanced Peripheral Bus
clock).  The "aclk", as main high speed bus clock, is enabled first.  Then
the "pclk" is enabled.

However the driver assigned reversed names for lookup of these clocks
from devicetree, so effectively the "pclk" was enabled first.

Although it might not matter in reality, the correct order is to enable
first main/high speed bus clock - "aclk".  Also this was the intention
of the actual code.

Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: sun8i-ss - fix result memory leak on error path
Corentin Labbe [Fri, 12 Feb 2021 08:46:10 +0000 (09:46 +0100)]
crypto: sun8i-ss - fix result memory leak on error path

This patch fixes a memory leak on an error path.

Fixes: d9b45418a917 ("crypto: sun8i-ss - support hash algorithms")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Remove totallen and offset in qce_start
Thara Gopinath [Thu, 11 Feb 2021 20:01:28 +0000 (15:01 -0500)]
crypto: qce - Remove totallen and offset in qce_start

totallen is used to get the size of the data to be transformed.
This is also available via nbytes or cryptlen in the qce_sha_reqctx
and qce_cipher_ctx. Similarly offset convey nothing for the supported
encryption and authentication transformations and is always 0.
Remove these two redundant parameters in qce_start.

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Remover src_tbl from qce_cipher_reqctx
Thara Gopinath [Thu, 11 Feb 2021 20:01:27 +0000 (15:01 -0500)]
crypto: qce - Remover src_tbl from qce_cipher_reqctx

src_table is unused and hence remove it from struct qce_cipher_reqctx

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Set data unit size to message length for AES XTS transformation
Thara Gopinath [Thu, 11 Feb 2021 20:01:26 +0000 (15:01 -0500)]
crypto: qce - Set data unit size to message length for AES XTS transformation

Set the register REG_ENCR_XTS_DU_SIZE to cryptlen for AES XTS
transformation. Anything else causes the engine to return back
wrong results.

Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Improve the conditions for requesting AES fallback cipher
Thara Gopinath [Thu, 11 Feb 2021 20:01:25 +0000 (15:01 -0500)]
crypto: qce - Improve the conditions for requesting AES fallback cipher

The following are the conditions for requesting AES fallback cipher.
-  AES-192
- AES-XTS request with len <= 512 byte (Allow messages of length
  less than 512 bytes for all other AES encryption algorithms other
  than AES XTS)
- AES-XTS request with len > QCE_SECTOR_SIZE and is not a multiple
  of it

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Set ivsize to 0 for ecb(aes)
Thara Gopinath [Thu, 11 Feb 2021 20:01:24 +0000 (15:01 -0500)]
crypto: qce - Set ivsize to 0 for ecb(aes)

ECB transformations do not have an IV and hence set the ivsize to 0 for
ecb(aes).

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Return error for non-blocksize data(ECB/CBC algorithms)
Thara Gopinath [Thu, 11 Feb 2021 20:01:23 +0000 (15:01 -0500)]
crypto: qce - Return error for non-blocksize data(ECB/CBC algorithms)

ECB/CBC encryption/decryption requires the data to be blocksize aligned.
Crypto engine hangs on non-block sized operations for these algorithms.
Return invalid data if data size is not blocksize aligned for these
algorithms.

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Return error for zero length messages
Thara Gopinath [Thu, 11 Feb 2021 20:01:22 +0000 (15:01 -0500)]
crypto: qce - Return error for zero length messages

Crypto engine BAM dma does not support 0 length data. Return unsupported
if zero length messages are passed for transformation.

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Return unsupported if any three keys are same for DES3 algorithms
Thara Gopinath [Thu, 11 Feb 2021 20:01:21 +0000 (15:01 -0500)]
crypto: qce - Return unsupported if any three keys are same for DES3 algorithms

Return unsupported if any three keys are same for DES3 algorithms
since CE does not support this and the operation causes the engine to
hang.

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Return unsupported if key1 and key 2 are same for AES XTS algorithm
Thara Gopinath [Thu, 11 Feb 2021 20:01:20 +0000 (15:01 -0500)]
crypto: qce - Return unsupported if key1 and key 2 are same for AES XTS algorithm

Crypto engine does not support key1 = key2 for AES XTS algorithm; the
operation hangs the engines.  Return -EINVAL in case key1 and key2 are the
same.

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Hold back a block of data to be transferred as part of final
Thara Gopinath [Thu, 11 Feb 2021 20:01:19 +0000 (15:01 -0500)]
crypto: qce - Hold back a block of data to be transferred as part of final

If the available data to transfer is exactly a multiple of block size, save
the last block to be transferred in qce_ahash_final (with the last block
bit set) if this is indeed the end of data stream. If not this saved block
will be transferred as part of next update. If this block is not held back
and if this is indeed the end of data stream, the digest obtained will be
wrong since qce_ahash_final will see that rctx->buflen is 0 and return
doing nothing which in turn means that a digest will not be copied to the
destination result buffer.  qce_ahash_final cannot be made to alter this
behavior and allowed to proceed if rctx->buflen is 0 because the crypto
engine BAM does not allow for zero length transfers.

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qce - Restore/save ahash state with custom struct in export/import
Thara Gopinath [Thu, 11 Feb 2021 20:01:18 +0000 (15:01 -0500)]
crypto: qce - Restore/save ahash state with custom struct in export/import

Export and import interfaces save and restore partial transformation
states. The partial states were being stored and restored in struct
sha1_state for sha1/hmac(sha1) transformations and sha256_state for
sha256/hmac(sha256) transformations.This led to a bunch of corner cases
where improper state was being stored and restored. A few of the corner
cases that turned up during testing are:

- wrong byte_count restored if export/import is called twice without h/w
transaction in between
- wrong buflen restored back if the pending buffer
length is exactly the block size.
- wrong state restored if buffer length is 0.

To fix these issues, save and restore the partial transformation state
using the newly introduced qce_sha_saved_state struct. This ensures that
all the pieces required to properly restart the transformation is captured
and restored back

Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: keembay-ocs-aes - Fix error return code in kmb_ocs_aes_probe()
Wei Yongjun [Wed, 10 Feb 2021 07:45:27 +0000 (07:45 +0000)]
crypto: keembay-ocs-aes - Fix error return code in kmb_ocs_aes_probe()

Fix to return negative error code -ENOMEM from the error handling
case instead of 0, as done elsewhere in this function.

Fixes: 885743324513 ("crypto: keembay - Add support for Keem Bay OCS AES/SM4")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: keembay-ocs-hcu - Fix error return code in kmb_ocs_hcu_probe()
Wei Yongjun [Wed, 10 Feb 2021 07:43:50 +0000 (07:43 +0000)]
crypto: keembay-ocs-hcu - Fix error return code in kmb_ocs_hcu_probe()

Fix to return negative error code -ENOMEM from the error handling
case instead of 0, as done elsewhere in this function.

Fixes: 472b04444cd3 ("crypto: keembay - Add Keem Bay OCS HCU driver")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: serpent - Fix sparse byte order warnings
Herbert Xu [Wed, 10 Feb 2021 07:15:57 +0000 (18:15 +1100)]
crypto: serpent - Fix sparse byte order warnings

This patch fixes the byte order markings in serpent.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Tested-by: Ard Biesheuvel <ardb@kernel.org> # arm64 big-endian
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: amlogic - Fix unnecessary check in meson_crypto_probe()
Tang Bin [Wed, 10 Feb 2021 03:16:37 +0000 (11:16 +0800)]
crypto: amlogic - Fix unnecessary check in meson_crypto_probe()

The function meson_crypto_probe() is only called with an openfirmware
platform device. Therefore there is no need to check that the passed
in device is NULL.

Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agohwrng: bcm2835 - remove redundant null check
Tian Tao [Wed, 10 Feb 2021 01:14:58 +0000 (09:14 +0800)]
hwrng: bcm2835 - remove redundant null check

clk_prepare_enable() and clk_disable_unprepare() will check
NULL clock parameter, so It is not necessary to add additional checks.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
3 years agocrypto: qat - fix spelling mistake: "messge" -> "message"
Bhaskar Chowdhury [Tue, 9 Feb 2021 10:28:55 +0000 (10:28 +0000)]
crypto: qat - fix spelling mistake: "messge" -> "message"

Trivial fix to spelling mistake in adf_pf2vf_msg.c and adf_vf2pf_msg.c.
s/messge/message/

Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Signed-off-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>