Mercurial > libavcodec.hg
diff mem.c @ 2979:bfabfdf9ce55 libavcodec
COSMETICS: tabs --> spaces, some prettyprinting
| author | diego |
|---|---|
| date | Thu, 22 Dec 2005 01:10:11 +0000 |
| parents | ef2149182f1c |
| children | e23d2bddf122 |
line wrap: on
line diff
--- a/mem.c Wed Dec 21 17:50:40 2005 +0000 +++ b/mem.c Thu Dec 22 01:10:11 2005 +0000 @@ -64,8 +64,8 @@ Indeed, we should align it: on 4 for 386 on 16 for 486 - on 32 for 586, PPro - k6-III - on 64 for K7 (maybe for P3 too). + on 32 for 586, PPro - k6-III + on 64 for K7 (maybe for P3 too). Because L1 and L2 caches are aligned on those values. But I don't want to code such logic here! */ @@ -76,13 +76,13 @@ Why not larger? because i didnt see a difference in benchmarks ... */ /* benchmarks with p3 - memalign(64)+1 3071,3051,3032 - memalign(64)+2 3051,3032,3041 - memalign(64)+4 2911,2896,2915 - memalign(64)+8 2545,2554,2550 - memalign(64)+16 2543,2572,2563 - memalign(64)+32 2546,2545,2571 - memalign(64)+64 2570,2533,2558 + memalign(64)+1 3071,3051,3032 + memalign(64)+2 3051,3032,3041 + memalign(64)+4 2911,2896,2915 + memalign(64)+8 2545,2554,2550 + memalign(64)+16 2543,2572,2563 + memalign(64)+32 2546,2545,2571 + memalign(64)+64 2570,2533,2558 btw, malloc seems to do 8 byte alignment by default here */