diff options
| author | Vitor Sessak <vitor1001@gmail.com> | 2007-12-01 22:21:04 +0000 |
|---|---|---|
| committer | Vitor Sessak <vitor1001@gmail.com> | 2007-12-01 22:21:04 +0000 |
| commit | 52b541ad798c2388e92a1a876550ff381f9b2346 (patch) | |
| tree | ed4fa9fbaf0558825c666dbc40eb78fbb981a546 /libavcodec/i386/mpegvideo_mmx.c | |
| parent | 329851e03251b58b43c0b55c2852a69dcbb79459 (diff) | |
spelling
Originally committed as revision 11122 to svn://svn.ffmpeg.org/ffmpeg/trunk
Diffstat (limited to 'libavcodec/i386/mpegvideo_mmx.c')
| -rw-r--r-- | libavcodec/i386/mpegvideo_mmx.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/libavcodec/i386/mpegvideo_mmx.c b/libavcodec/i386/mpegvideo_mmx.c index bc91c56ac3..90b553aa2b 100644 --- a/libavcodec/i386/mpegvideo_mmx.c +++ b/libavcodec/i386/mpegvideo_mmx.c @@ -177,7 +177,7 @@ asm volatile( if (level < -2048 || level > 2047) fprintf(stderr, "unquant error %d %d\n", i, level); #endif - We can suppose that result of two multiplications can't be greate of 0xFFFF + We can suppose that result of two multiplications can't be greater than 0xFFFF i.e. is 16-bit, so we use here only PMULLW instruction and can avoid a complex multiplication. ===================================================== |