Re: [FFmpeg-devel] [PATCH 06/10] lavu/aes: add x86 AESNI optimizations

2015-10-13 Thread Henrik Gramner
I changed the asm a bit and made it about 1 cycle faster on Haswell and slightly smaller (-48 bytes overall incl. alignment on 64-bit Linux). %macro AES_CRYPT 1 cglobal aes_%1rypt, 6,6,2 shl r3d, 4 add r5d, r5d add r0, 0x60 add r2, r3 add r1, r3

Re: [FFmpeg-devel] [PATCH 06/10] lavu/aes: add x86 AESNI optimizations

2015-10-13 Thread Henrik Gramner
On Tue, Oct 13, 2015 at 2:33 AM, Rodger Combs wrote: > +%macro AES_CRYPT 1 > +%if %1 == 1 > +%define CRYPT aesdec > +%define LAST aesdeclast > +cglobal aes_decrypt, 6,6,2 > +%else > +%define CRYPT aesenc > +%define LAST aesenclast > +cglobal aes_encrypt, 6,6,2 > +%endif > +pxor xm1, xm1 > +

[FFmpeg-devel] [PATCH 06/10] lavu/aes: add x86 AESNI optimizations

2015-10-12 Thread Rodger Combs
--- libavutil/aes.c | 4 ++ libavutil/aes_internal.h | 2 + libavutil/x86/Makefile | 4 +- libavutil/x86/aes.asm| 98 libavutil/x86/aes_init.c | 37 ++ 5 files changed, 144 insertions(+), 1 deletion(-) create mode