diff options
| author | Eric Biggers <ebiggers@google.com> | 2024-04-11 18:23:59 +0200 |
|---|---|---|
| committer | Herbert Xu <herbert@gondor.apana.org.au> | 2024-04-19 12:54:18 +0200 |
| commit | 7daba20cc72d4b3aa047cc3868b96e8e45d5eeca (patch) | |
| tree | 67deeda45ef115a307721e1d1812e97f868e7754 /arch/x86/crypto | |
| parent | crypto: x86/sha256-ni - optimize code size (diff) | |
| download | linux-7daba20cc72d4b3aa047cc3868b96e8e45d5eeca.tar.xz linux-7daba20cc72d4b3aa047cc3868b96e8e45d5eeca.zip | |
crypto: x86/sha256-ni - simplify do_4rounds
Instead of loading the message words into both MSG and \m0 and then
adding the round constants to MSG, load the message words into \m0 and
the round constants into MSG and then add \m0 to MSG. This shortens the
source code slightly. It changes the instructions slightly, but it
doesn't affect binary code size and doesn't seem to affect performance.
Suggested-by: Stefan Kanthak <stefan.kanthak@nexgo.de>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Diffstat (limited to 'arch/x86/crypto')
| -rw-r--r-- | arch/x86/crypto/sha256_ni_asm.S | 10 |
1 files changed, 4 insertions, 6 deletions
diff --git a/arch/x86/crypto/sha256_ni_asm.S b/arch/x86/crypto/sha256_ni_asm.S index ffc9f1c75c15..d515a55a3bc1 100644 --- a/arch/x86/crypto/sha256_ni_asm.S +++ b/arch/x86/crypto/sha256_ni_asm.S @@ -78,13 +78,11 @@ .macro do_4rounds i, m0, m1, m2, m3 .if \i < 16 - movdqu \i*4(DATA_PTR), MSG - pshufb SHUF_MASK, MSG - movdqa MSG, \m0 -.else - movdqa \m0, MSG + movdqu \i*4(DATA_PTR), \m0 + pshufb SHUF_MASK, \m0 .endif - paddd (\i-32)*4(SHA256CONSTANTS), MSG + movdqa (\i-32)*4(SHA256CONSTANTS), MSG + paddd \m0, MSG sha256rnds2 STATE0, STATE1 .if \i >= 12 && \i < 60 movdqa \m0, TMP |