Save the return value of qat_uclo_map_auth_fw() function so that the
function qat_uclo_wr_mimage() could return the correct value.
This way, the procedure of adf_gen2_ae_fw_load() function could stop
and exit properly by checking the return value of qat_uclo_wr_mimage().
Signed-off-by: NJack Xu <jack.xu@intel.com>
Co-developed-by: NZhehui Xiang <zhehui.xiang@intel.com>
Signed-off-by: NZhehui Xiang <zhehui.xiang@intel.com>
Reviewed-by: NGiovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

If an error occurs after calling 'sp_get_irqs()', 'sp_free_irqs()' must be
called as already done in the error handling path.

Fixes: f4d18d65 ("crypto: ccp - Abstract interrupt registeration")
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: NJohn Allen <john.allen@amd.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Kunpeng930 could be able to close master ooo when NFE occurs, which will
disable memory accessing from device and execute tasks. This ensures that
errors do not spread.

This patch enables the hardware to close master ooo when an error occurs
by writing hardware registers, and ensures that the driver will not drain
qp because the hardware will empty the tasks automatically.
Signed-off-by: NWeili Qian <qianweili@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

In order to avoid reporting an exception but the error type is not
configured, the driver needs to configure the error type first, and then
enable the error interrupt. Before executing the task, hardware error
initialization is needed so that the hardware can detect the error in time.
Signed-off-by: NWeili Qian <qianweili@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Before device reset, the driver sets 'QM_RESETTING' flag, but after reset,
the wrong flag is cleared. This patch modifies the problem of inconsistent
flags.
Signed-off-by: NWeili Qian <qianweili@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The device needs to be initialized first, and then restart the queue to
execute tasks after PF reset.
Signed-off-by: NWeili Qian <qianweili@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The sa_ul_probe creates child devices using of_platform_populate(),
but these are not cleaned up in driver remove. Clean these up
by removing the child devices using of_platform_depopulate().
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Simplify the platform_get_resource() and devm_ioremap_resource()
calls with devm_platform_ioremap_resource(). Also add error checking
and move up this block to simplify the cleanup in sa_ul_probe().
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Simplify the probe function by using the of_device_get_match_data()
helper instead of open coding. The logic is also moved up to fix the
missing pm_runtime cleanup in case of a match failure.

Fixes: 0bc42311 ("crypto: sa2ul - Add support for AM64")
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The pm_runtime APIs added first in commit 7694b6ca ("crypto: sa2ul -
Add crypto driver") are not unwound properly and was fixed up partially
in commit 13343bad ("crypto: sa2ul - Fix PM reference leak in
sa_ul_probe()"). This fixed up the pm_runtime usage count but not the
state. Fix this properly.

Fixes: 13343bad ("crypto: sa2ul - Fix PM reference leak in sa_ul_probe()")
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The sa_dma_init() function doesn't release the requested dma channels
on all failure paths. Any failure in this function also ends up
leaking the dma pool created in sa_init_mem() in the sa_ul_probe()
function. Fix all of these issues.

Fixes: 7694b6ca ("crypto: sa2ul - Add crypto driver")
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The TI K3 family of SoCs have a SA2UL IP that contains a
SafeXcel IP-76 RNG block which is supported by the OMAP
RNG driver. Allow this driver to be built for TI K3
family as well.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Reviewed-by: NTero Kristo <kristo@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Return error immediately if it goto 'default' path.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Delete rudundant macro definition.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Use 'GENMASK' to generate mask value, just make the code clearer.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Delete rudundant variable initialization.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Remove complex macro of 'HPRE_DEV' and replace with the initialized
device pointer.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Functional macro lacks type checking, which is not as strict as function
call checking.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Only init a structure member each line, just to keep the code neat.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The macro 'HPRE_ADDR' is unnecessary, so expanding it.
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently, an invalid dma address may be unmapped when calling
'xx_data_clr_all' in error path, so check dma address of sqe in/out
if initialized before calling 'dma_free_coherent' or 'dma_unmap_single'.

Fixes: a9214b0b ("crypto: hisilicon - fix the check on dma address")
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently,'cra_driver_name' cannot be used to specify ecdh algorithm
with a special curve, so extending it with curve name.

Fixes: 6763f5ea ("crypto: ecdh - move curve_id of ECDH from ...")
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently, 'cra_driver_name' cannot be used to specify ecdh algorithm
with a special curve, so extending it with curve name.

Although using 'cra_name' can also specify a special curve, but ecdh
generic driver cannot be specified when vendor hardware accelerator
has registered.

Fixes: 6763f5ea ("crypto: ecdh - move curve_id of ECDH from ...")
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Actual data length of the 'secret' is not equal to the 'secret_size'.

Since the 'curve_id' has removed in the 'secret', the 'secret_size'
should subtract the length of the 'curve_id'.

Fixes: 6763f5ea ("crypto: ecdh - move curve_id of ECDH from ...")
Signed-off-by: NHui Tang <tanghui20@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

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

Fixes: 8a63b199 ("crypto: ux500 - Add driver for HASH hardware")
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NZhen Lei <thunder.leizhen@huawei.com>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch adds missing MODULE_DEVICE_TABLE definition which generates
correct modalias for automatic loading of this driver when it is built
as an external module.
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NBixuan Cui <cuibixuan@huawei.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Fix function name in nitrox_reqmgr.c kernel-doc comment
to remove a warning.

drivers/crypto/cavium/nitrox/nitrox_reqmgr.c:382: warning: expecting
prototype for nitrox_se_request(). Prototype was for
nitrox_process_se_request() instead
Reported-by: NAbaci Robot <abaci@linux.alibaba.com>
Signed-off-by: NYang Li <yang.lee@linux.alibaba.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Pointer 'sg' is being initialized however this value is never
read as 'sg' is assigned a same value in for_each_sg().
Remove the redundant assignment.

Cleans up clang warning:

drivers/crypto/cavium/nitrox/nitrox_reqmgr.c:161:22: warning: Value
stored to 'sg' during its initialization is never read
[clang-analyzer-deadcode.DeadStores]
Reported-by: NAbaci Robot <abaci@linux.alibaba.com>
Signed-off-by: NYang Li <yang.lee@linux.alibaba.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

No maintainer exists for ixp4xx_crypto, since I have access to a board
with it, I propose to maintain it.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

drivers/crypto/ixp4xx_crypto.c is missing in the IXP4XX arch file list.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch fixes all alignment issues reported by checkpatch.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

fixes all single statement issues reported by checkpatch
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch fixes all checkpatch report about static init.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Fixes all whitespace issues reported by checkpatch
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Convert all old printk to dev_xxx.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Fixes all issues reported by checkpatch about "unsigned", lets convert
them to unsigned int.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Testing ixp4xx_crypto lead to:
alg: skcipher: ecb(des)-ixp4xx encryption overran dst buffer on test vector 0, cfg="two even aligned splits"

The HW overwrites destination always when sg_nents() > 1.
The problem seems that the HW always write areq->cryptlen bytes on the
last SG.
A comment in driver's code seems to give a clue that multiple SG was not
planned "This was never tested by Intel for more than one dst buffer, I think".
So let's add a fallback for this situation.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Crypto selftests fail on ixp4xx since it do not update IV after skcipher
requests.

Fixes: 81bef015 ("crypto: ixp4xx - Hardware crypto support for IXP4xx CPUs")
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Testing ixp4xx_crypto with CONFIG_DMA_API_DEBUG lead to the following error:
DMA-API: platform ixp4xx_crypto.0: device driver tries to free DMA memory it has not allocated [device address=0x0000000000000000] [size=24 bytes]

This is due to dma_unmap using the wrong address.

Fixes: 0d44dc59 ("crypto: ixp4xx - Fix handling of chained sg buffers")
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

pm_runtime_resume_and_get() wraps around pm_runtime_get_sync() and
decrements the runtime PM usage counter in case the latter function
fails and keeps the counter balanced.
Signed-off-by: NŁukasz Stelmach <l.stelmach@samsung.com>
Reviewed-by: NKrzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>