Re: [PATCH 1/2] crypto: s(h)aving 40+ bytes off arch/x86/crypto/sha256_ni_asm.S

[Eric Biggers]

On Mon, Apr 08, 2024 at 11:26:52AM +0200, Stefan Kanthak wrote:
> @@ -315,11 +315,11 @@
>          jne             .Lloop0
>  
>          /* Write hash values back in the correct order */
> -        pshufd          $0x1B, STATE0,  STATE0          /* FEBA */
> -        pshufd          $0xB1, STATE1,  STATE1          /* DCHG */
>          movdqa          STATE0, MSGTMP4
> -        pblendw         $0xF0, STATE1,  STATE0          /* DCBA */
> -        palignr         $8, MSGTMP4, STATE1             /* HGFE */
> +        punpcklqdq      STATE1, STATE0                  /* GHEF */
> +        punpckhqdq      MSGTMP4, STATE1                 /* ABCD */
> +        pshufd          $0xB1, STATE0,  STATE0          /* HGFE */
> +        pshufd          $0x1B, STATE1,  STATE1          /* DCBA */
>  
>          movdqu          STATE0, 0*16(DIGEST_PTR)
>          movdqu          STATE1, 1*16(DIGEST_PTR)

Please make sure to run the crypto self-tests too.  The above is storing the two
halves of the state in the wrong order.

Thanks,

- Eric