[GIT,PULL] Crypto Fixes for 6.1

Message ID Y1thZ/+Gh/ONyf7x@gondor.apana.org.au
State New
Headers
Series [GIT,PULL] Crypto Fixes for 6.1 |

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6.git tags/v6.1-p3

Message

Herbert Xu Oct. 28, 2022, 4:58 a.m. UTC
  Hi Linus:

The following changes since commit 9abf2313adc1ca1b6180c508c25f22f9395cc780:

  Linux 6.1-rc1 (2022-10-16 15:36:24 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6.git tags/v6.1-p3

for you to fetch changes up to 9f6035af06b526e678808d492fc0830aef6cfbd8:

  crypto: x86/polyval - Fix crashes when keys are not 16-byte aligned (2022-10-21 19:05:05 +0800)

----------------------------------------------------------------
This push fixes an alignment crash in x86/polyval.
----------------------------------------------------------------

Nathan Huckleberry (1):
      crypto: x86/polyval - Fix crashes when keys are not 16-byte aligned

 arch/x86/crypto/polyval-clmulni_glue.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

Thanks,
  

Comments

Linus Torvalds Oct. 28, 2022, 5 p.m. UTC | #1
On Thu, Oct 27, 2022 at 9:58 PM Herbert Xu <herbert@gondor.apana.org.au> wrote:
>
> This push fixes an alignment crash in x86/polyval.

I'm surprised that there isn't a cra_ctxalignment field. Instead there
is crypto_tfm_ctx_alignment(), but that is just an odd way to write
CRYPTO_MINALIGN.

            Linus
  
pr-tracker-bot@kernel.org Oct. 28, 2022, 5:02 p.m. UTC | #2
The pull request you sent on Fri, 28 Oct 2022 12:58:15 +0800:

> git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6.git tags/v6.1-p3

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/05c31d25cc9678cc173cf12e259d638e8a641f66

Thank you!
  
Herbert Xu Nov. 2, 2022, 9:49 a.m. UTC | #3
On Fri, Oct 28, 2022 at 10:00:34AM -0700, Linus Torvalds wrote:
> 
> I'm surprised that there isn't a cra_ctxalignment field. Instead there
> is crypto_tfm_ctx_alignment(), but that is just an odd way to write
> CRYPTO_MINALIGN.

It's a micro-optimisation to elide the align operaton because
most algorithms don't need it.

I've been meaning to add a helper for it though:

https://lore.kernel.org/all/E1noNhu-00BzV4-4N@fornost.hmeau.com/

But I got side tracked by the vacillating DMA discussion.

Cheers,