On Mon, Oct 26, 2015 at 03:25:48AM -0500, Rodger Combs wrote: > crypto_bench comparison for AES-128-ECB: > > lavu_aesni AES-128-ECB size: 1048576 runs: 1024 time: 0.596 +- 0.081 > lavu_c AES-128-ECB size: 1048576 runs: 1024 time: 17.007 +- 2.131 > crypto AES-128-ECB size: 1048576 runs: 1024 time: 0.612 +- 1.857 > gcrypt AES-128-ECB size: 1048576 runs: 1024 time: 1.123 +- 0.224 > tomcrypt AES-128-ECB size: 1048576 runs: 1024 time: 9.038 +- 0.790 > > Improved-By: Henrik Gramner <hen...@gramner.com> > --- > libavutil/aes.c | 4 +++ > libavutil/aes_internal.h | 2 ++ > libavutil/x86/Makefile | 4 ++- > libavutil/x86/aes.asm | 92 > ++++++++++++++++++++++++++++++++++++++++++++++++ > libavutil/x86/aes_init.c | 37 +++++++++++++++++++ > 5 files changed, 138 insertions(+), 1 deletion(-) > create mode 100644 libavutil/x86/aes.asm > create mode 100644 libavutil/x86/aes_init.c > > diff --git a/libavutil/aes.c b/libavutil/aes.c > index 4b871a0..40db681 100644 > --- a/libavutil/aes.c > +++ b/libavutil/aes.c > @@ -161,6 +161,8 @@ static void aes_decrypt(AVAES *a, uint8_t *dst, const > uint8_t *src, > void av_aes_crypt(AVAES *a, uint8_t *dst, const uint8_t *src, > int count, uint8_t *iv, int decrypt) > { > + if (count <= 0) > + return;
if only the ASM version needs this check then it should only be done for the asm version [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB I have often repented speaking, but never of holding my tongue. -- Xenocrates
signature.asc
Description: Digital signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel