part of the spec and causes problems otherwise.
Patch by Zhentan Feng <spyfeng gmail com>.
Originally committed as revision 24362 to svn://svn.ffmpeg.org/ffmpeg/trunk
most servers. Also remove a case where we manually aligned to 8 bytes,
since this is now no longer needed.
Patch by Zhentan Feng <spyfeng gmail com>.
Originally committed as revision 24360 to svn://svn.ffmpeg.org/ffmpeg/trunk
format ‘%d’ expects type ‘int’, but argument 3 has type ‘uint64_t’
Patch by Zhentan Feng <spyfeng gmail com>.
Originally committed as revision 24359 to svn://svn.ffmpeg.org/ffmpeg/trunk
The diff may provide useful information even if the command was
unsuccessful. The test is still treated as failed in this case.
Originally committed as revision 24353 to svn://svn.ffmpeg.org/ffmpeg/trunk
This fixes a leak introduced in rev 23942, since we write padding to the
buffer unconditionally.
Originally committed as revision 24342 to svn://svn.ffmpeg.org/ffmpeg/trunk
regular MMX code. Examples of this are the Core1 CPU. Instead, set a new flag,
FF_MM_SSE2/3SLOW, which can be checked for particular SSE2/3 functions that
have been checked specifically on such CPUs and are actually faster than
their MMX counterparts.
In addition, use this flag to enable particular VP8 and LPC SSE2 functions
that are faster than their MMX counterparts.
Based on a patch by Loren Merritt <lorenm AT u washington edu>.
Originally committed as revision 24340 to svn://svn.ffmpeg.org/ffmpeg/trunk
so that it does both U and V planes at the same time. This will have speed
advantages when using SSE2 (or higher) optimizations, since we can do both
the U and V rows together in a single xmm register.
This also renames filter16 to filter16y and filter8 to filter8uv so that it's
more obvious what each function is used for.
Originally committed as revision 24337 to svn://svn.ffmpeg.org/ffmpeg/trunk
This test verifies the pixdesc code by comparing the output with and
without a filter which should have no effect on the image. Since the
available pixel formats depend on the byte order of the machine, a
simple reference checksum is not possible.
The test originally tried to solve this by generating a reference file
on the fly. The problem with this is that the test framework expects
the reference file in the source tree, and writing to the source tree
is not allowed.
To avoid complicating the test framework, we instead provide two
reference files and select which to use based on the byte order.
Originally committed as revision 24330 to svn://svn.ffmpeg.org/ffmpeg/trunk
Don't change paramater passing, but instead use casts.
Shouldn't affect asm output on anything other than win64.
libswscale should work on win64 now.
The rest of ffmpeg still isn't win64 compatible due to the issue of xmm
clobbers, but swscale doesn't use any SSE.
Patch by Anton Mitrofanov <BugMaster AT narod DOT ru>.
Originally committed as revision 31751 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
All tests use the provided helper functions so prepending $target_exec
and using eval is no longer required.
Originally committed as revision 24317 to svn://svn.ffmpeg.org/ffmpeg/trunk
Zhentan Feng
Vitor Sessak
Carl Eugen Hoyos
Rafaël Carré
Martin Storsjö
Måns Rullgård
Ramiro Polla
Ronald S. Bultje
Alex Converse
Josh Allmann
Nathan Caldwell
Aurelien Jacobs
Peter Ross
Diego Biurrun
Stefano Sabatini
Anton Mitrofanov