linux-stable.git/drivers/crypto, branch v7.1.12 Linux kernel stable tree crypto: tegra - fix rctx->cryptlen calculation in tegra_gcm_do_one_req() 2026-08-23T12:29:29+00:00 Vladislav Dronov vdronov@redhat.com 2026-07-20T22:33:31+00:00 c6237834d9994de209cb90c7a2c461247bce8e90 [ Upstream commit 360f2974fcea49c61f6d6f81554741a9eeee7168 ] Perform rctx->cryptlen calculation in tegra_gcm_do_one_req() the same way it is done in tegra_ccm_crypt_init(). The current formulae may lead to a crash if a caller does not call tegra_gcm_setauthsize() and so ctx->authsize remains zero. Then a decrypt operation with incorrect rctx->cryptlen will lead to a write beyound rctx->dst_sg buffer. As a follow-up cleanup delete struct tegra_aead_ctx->authsize field since it appears to be completely unused. Also simplify tegra_ccm_setauthsize() and tegra_gcm_setauthsize() functions respectively. Fixes: 0880bb3b00c8 ("crypto: tegra - Add Tegra Security Engine driver") Signed-off-by: Vladislav Dronov <vdronov@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 360f2974fcea49c61f6d6f81554741a9eeee7168 ]

Perform rctx->cryptlen calculation in tegra_gcm_do_one_req() the same way
it is done in tegra_ccm_crypt_init(). The current formulae may lead to a
crash if a caller does not call tegra_gcm_setauthsize() and so ctx->authsize
remains zero. Then a decrypt operation with incorrect rctx->cryptlen will
lead to a write beyound rctx->dst_sg buffer.

As a follow-up cleanup delete struct tegra_aead_ctx->authsize field since
it appears to be completely unused. Also simplify tegra_ccm_setauthsize()
and tegra_gcm_setauthsize() functions respectively.

Fixes: 0880bb3b00c8 ("crypto: tegra - Add Tegra Security Engine driver")
Signed-off-by: Vladislav Dronov <vdronov@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: qce - fix error path in devm_qce_register_algs 2026-08-23T12:29:21+00:00 Thorsten Blum thorsten.blum@linux.dev 2026-07-24T08:15:38+00:00 4e88b4fda48282f3fd504b4d4f5d2d4f996b76ce commit 9c75402286409f5e1a75e4a445555c84066f89db upstream. If ops->register_algs() fails, the error path repeatedly calls the same ops->unregister_algs() from the failed registration. Use the loop index to unregister the previously registered algorithms instead. Fixes: e80cf84b6087 ("crypto: qce - unregister previously registered algos in error path") Cc: stable@vger.kernel.org Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev> Reviewed-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 9c75402286409f5e1a75e4a445555c84066f89db upstream.

If ops->register_algs() fails, the error path repeatedly calls the same
ops->unregister_algs() from the failed registration. Use the loop index
to unregister the previously registered algorithms instead.

Fixes: e80cf84b6087 ("crypto: qce - unregister previously registered algos in error path")
Cc: stable@vger.kernel.org
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Reviewed-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
crypto: starfive - use scatterlist length before DMA mapping 2026-08-23T12:29:21+00:00 Thorsten Blum thorsten.blum@linux.dev 2026-07-25T09:06:10+00:00 61b20fba32d65591e9a6d7717cf1a6002c5c407c commit 6b36f13891ab4709b7d60023005176cdd5c368cf upstream. Using sg_dma_len() is only valid after mapping a scatterlist with dma_map_sg(). However, starfive_aes_aead_do_one_req() uses it before mapping the scatterlist. Use the original scatterlist length because the DMA length has not been populated yet when CONFIG_NEED_SG_DMA_LENGTH=y. Fixes: 7467147ef9bf ("crypto: starfive - Use dma for aes requests") Cc: stable@vger.kernel.org Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 6b36f13891ab4709b7d60023005176cdd5c368cf upstream.

Using sg_dma_len() is only valid after mapping a scatterlist with
dma_map_sg(). However, starfive_aes_aead_do_one_req() uses it before
mapping the scatterlist.

Use the original scatterlist length because the DMA length has not been
populated yet when CONFIG_NEED_SG_DMA_LENGTH=y.

Fixes: 7467147ef9bf ("crypto: starfive - Use dma for aes requests")
Cc: stable@vger.kernel.org
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
crypto: tegra - Don't touch bo refcount in host1x bo pin/unpin 2026-08-03T09:25:16+00:00 Mikko Perttunen mperttunen@nvidia.com 2026-05-15T02:34:52+00:00 9382304e25c63d9533670ea3313367a2036bba23 [ Upstream commit f8c9c57d750346abd213ffed2ae3cacb0268e9f1 ] Since commit "gpu: host1x: Allow entries in BO caches to be freed", host1x_bo_pin() and host1x_bo_unpin() handle the bo's refcount themselves. .pin/.unpin callbacks should not adjust it. Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit f8c9c57d750346abd213ffed2ae3cacb0268e9f1 ]

Since commit "gpu: host1x: Allow entries in BO caches to be freed",
host1x_bo_pin() and host1x_bo_unpin() handle the bo's refcount
themselves. .pin/.unpin callbacks should not adjust it.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: xilinx-trng - Remove crypto_rng interface 2026-07-24T14:21:26+00:00 Eric Biggers ebiggers@kernel.org 2026-05-31T19:17:35+00:00 9634db561e1594c151923f4950c012161c545c93 commit 32b4d29280ed2a991dc196d5845b892acedc63b8 upstream. Implementing the crypto_rng interface has no purpose, as it isn't used in practice. It's being removed from other drivers too. Just remove it. This leaves hwrng, which is actually used. Tagging with 'Cc stable' due to the bugs that this removes: - xtrng_trng_generate() sometimes returned success even when it didn't fill in all the bytes. - It was possible for xtrng_trng_generate() and xtrng_hwrng_trng_read() to run concurrently and interfere with each other, as the locking code in xtrng_hwrng_trng_read() was broken. Fixes: 8979744aca80 ("crypto: xilinx - Add TRNG driver for Versal") Cc: stable@vger.kernel.org Signed-off-by: Eric Biggers <ebiggers@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 32b4d29280ed2a991dc196d5845b892acedc63b8 upstream.

Implementing the crypto_rng interface has no purpose, as it isn't used
in practice.  It's being removed from other drivers too.  Just remove
it.  This leaves hwrng, which is actually used.

Tagging with 'Cc stable' due to the bugs that this removes:

  - xtrng_trng_generate() sometimes returned success even when it didn't
    fill in all the bytes.

  - It was possible for xtrng_trng_generate() and
    xtrng_hwrng_trng_read() to run concurrently and interfere with each
    other, as the locking code in xtrng_hwrng_trng_read() was broken.

Fixes: 8979744aca80 ("crypto: xilinx - Add TRNG driver for Versal")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
crypto: cavium/cpt - fix DMA cleanup using wrong loop index 2026-07-24T14:19:50+00:00 Felix Gu ustc.gu@gmail.com 2026-06-02T14:55:35+00:00 d319b83b97b3585550d9ae592dfa78c755b6129e [ Upstream commit 9dbf173bd32d5f81b005008b682bfb50aa093455 ] The sg_cleanup error path used list[i] instead of list[j] when unmapping DMA buffers, leaking successfully mapped entries and repeatedly unmapping the failed one. Fixes: c694b233295b ("crypto: cavium - Add the Virtual Function driver for CPT") Signed-off-by: Felix Gu <ustc.gu@gmail.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 9dbf173bd32d5f81b005008b682bfb50aa093455 ]

The sg_cleanup error path used list[i] instead of list[j] when unmapping
DMA buffers, leaking successfully mapped entries and repeatedly unmapping
the failed one.

Fixes: c694b233295b ("crypto: cavium - Add the Virtual Function driver for CPT")
Signed-off-by: Felix Gu <ustc.gu@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: marvell/octeontx - fix DMA cleanup using wrong loop index 2026-07-24T14:19:50+00:00 Felix Gu ustc.gu@gmail.com 2026-06-02T14:38:26+00:00 acff30cfc0d72465b51b0bfdf019f1cb54e15314 [ Upstream commit 7891c64c0520519782470ba29bac8a5761e295d8 ] The sg_cleanup path used list[i] instead of list[j] when unmapping DMA buffers, leaking successfully mapped entries and repeatedly unmapping the failed one. Fixes: 10b4f09491bf ("crypto: marvell - add the Virtual Function driver for CPT") Signed-off-by: Felix Gu <ustc.gu@gmail.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 7891c64c0520519782470ba29bac8a5761e295d8 ]

The sg_cleanup path used list[i] instead of list[j] when unmapping DMA
buffers, leaking successfully mapped entries and repeatedly unmapping
the failed one.

Fixes: 10b4f09491bf ("crypto: marvell - add the Virtual Function driver for CPT")
Signed-off-by: Felix Gu <ustc.gu@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: ccp/tsm - Enable the root port after the endpoint 2026-07-24T14:19:24+00:00 Alexey Kardashevskiy aik@amd.com 2026-05-21T07:43:01+00:00 6945ad6db8d939c668f50e8ba2af24e50bcc4fe7 [ Upstream commit 50e506201bab875545c7a9443bd7e1c71804e553 ] The PCIe r7.0, chapter "6.33.8 Other IDE Rules" mandates if selective IDE is enabled for config requersts, a stream must be enabled on the endpoint before enabling it on the rootport: === For Selective IDE, the Stream must not be used until it has been enabled in both Partner Ports. For cases where one of the Partner Ports is a Root Port and Selective IDE for Configuration Requests is enabled, the other Partner Port must be enabled prior to the Root Port. For other scenarios, the mechanisms to satisfy this requirement are implementation-specific. === Do what the spec says. Fixes: 4be423572da1 ("crypto/ccp: Implement SEV-TIO PCIe IDE (phase1)") Signed-off-by: Alexey Kardashevskiy <aik@amd.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 50e506201bab875545c7a9443bd7e1c71804e553 ]

The PCIe r7.0, chapter "6.33.8 Other IDE Rules" mandates if selective IDE
is enabled for config requersts, a stream must be enabled on the endpoint
before enabling it on the rootport:

===
For Selective IDE, the Stream must not be used until it has been enabled in
both Partner Ports. For cases where one of the Partner Ports is a Root Port
and Selective IDE for Configuration Requests is enabled, the other
Partner Port must be enabled prior to the Root Port. For other scenarios,
the mechanisms to satisfy this requirement are implementation-specific.
===

Do what the spec says.

Fixes: 4be423572da1 ("crypto/ccp: Implement SEV-TIO PCIe IDE (phase1)")
Signed-off-by: Alexey Kardashevskiy <aik@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: tegra - Return ENOMEM when input buffer allocation fails for ccm 2026-07-24T14:19:24+00:00 Herbert Xu herbert@gondor.apana.org.au 2026-05-20T02:51:14+00:00 20360e125a8dd411b0cc2660cb70f58590740818 [ Upstream commit 690a5f9e5c972a580565ce544ed1627ccf1e84de ] Ensure the ENOMEM error value is set when the input buffer allocation fails in tegra_ccm_do_one_req. Fixes: 1e245948ca0c ("crypto: tegra - finalize crypto req on error") Reported-by: Vladislav Dronov <vdronov@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Reviewed-by: Vladislav Dronov <vdronov@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 690a5f9e5c972a580565ce544ed1627ccf1e84de ]

Ensure the ENOMEM error value is set when the input buffer allocation
fails in tegra_ccm_do_one_req.

Fixes: 1e245948ca0c ("crypto: tegra - finalize crypto req on error")
Reported-by: Vladislav Dronov <vdronov@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Reviewed-by: Vladislav Dronov <vdronov@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
crypto: tegra - Fix dma_free_coherent size error 2026-07-24T14:19:24+00:00 Herbert Xu herbert@gondor.apana.org.au 2026-05-19T04:22:18+00:00 ec49878f2780a985b9efbd68121979c493f4afbd [ Upstream commit 03215b8457784540acc741e6331e355b62c6c8ab ] When freeing a coherent DMA buffer, the size must match the value that was used during the allocation. Unfortunately the size field in the tegra driver gets overwritten by this point so it no longer matches and creates a warning. Fix this by saving a copy of the size on the stack. Note that the ccm function actually mixes up the inbuf and outbuf sizes, but it doesn't matter because the two sizes are actually equal. Fixes: 1cb328da4e8f ("crypto: tegra - Do not use fixed size buffers") Reporeted-by: Patrick Talbert <ptalbert@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Reviewed-by: Vladislav Dronov <vdronov@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 03215b8457784540acc741e6331e355b62c6c8ab ]

When freeing a coherent DMA buffer, the size must match the value
that was used during the allocation.

Unfortunately the size field in the tegra driver gets overwritten
by this point so it no longer matches and creates a warning.

Fix this by saving a copy of the size on the stack.

Note that the ccm function actually mixes up the inbuf and outbuf
sizes, but it doesn't matter because the two sizes are actually
equal.

Fixes: 1cb328da4e8f ("crypto: tegra - Do not use fixed size buffers")
Reporeted-by: Patrick Talbert <ptalbert@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Reviewed-by: Vladislav Dronov <vdronov@redhat.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>