ICC needs at least -O1 to link so add this when optimisations are
otherwise disabled.
Originally committed as revision 23013 to svn://svn.ffmpeg.org/ffmpeg/trunk
Feel free to revert if you can specify a concrete case where this actually
is necessary.
Originally committed as revision 23006 to svn://svn.ffmpeg.org/ffmpeg/trunk
of failure return the stored value rather than the current errno,
which may be overwritten by a following call to close().
Originally committed as revision 23001 to svn://svn.ffmpeg.org/ffmpeg/trunk
Using macro templates allows the vp[56]_adjust functions to be
inlined instead of called through function pointers. The new
function pointers enable optimised implementations of the filters.
4% faster VP6 decoding on Cortex-A8.
Originally committed as revision 22992 to svn://svn.ffmpeg.org/ffmpeg/trunk
The normal byteswap functions might use inline asm which is suboptimal
with constants (and cannot be used in static initialisers), so special
macros for constants only is needed.
We should not rely on the gcc __builtin_constant_p() test since it is
not always available.
Originally committed as revision 22990 to svn://svn.ffmpeg.org/ffmpeg/trunk
function. Avoids the following warning:
libavdevice/v4l2.c: In function ‘v4l2_read_header’:
libavdevice/v4l2.c:586: warning: ‘codec_id’ may be used uninitialized in this function
Originally committed as revision 22986 to svn://svn.ffmpeg.org/ffmpeg/trunk
sample_aspect_ratio is known.
Avoid to show bogus values, which may confuse both the human and the
machine reader.
Based on a patch by Robert Krüger $(echo lsvfhfs@tjhobm7.ef | tr "b-za" "a-z").
Originally committed as revision 22983 to svn://svn.ffmpeg.org/ffmpeg/trunk
instead which is semantically equivalent.
See the thread:
Subject: [FFmpeg-devel] [RFC] The meaning of AVERROR_NOTSUPP
Date: Sat, 27 Mar 2010 00:51:13 +0100
Originally committed as revision 22981 to svn://svn.ffmpeg.org/ffmpeg/trunk
fields of common media header chunks (these can have different sizes depending
on the type of IFF file you read), better handle odd sizes (like RIFF, every
field is padded to word) and handle headerchunks after the BODY chunk.
Patch by Sebastian Vater <cdgs.basty googlemail com>.
Originally committed as revision 22977 to svn://svn.ffmpeg.org/ffmpeg/trunk
thus performance) slightly.
Patch by Sebastian Vater <cdgs.basty googlemail com>.
Originally committed as revision 22975 to svn://svn.ffmpeg.org/ffmpeg/trunk
compiler to optimize some math from mul/div to shr/shl). Also add a cast to
uint32_t when calling decodeplane32(), this silences a compiler warning.
Lastly, in decodeplane8/32(), flatten a double-loop into a single-loop and
calculate the length once before entering the loop instead of during every
iteration (since it doesn't change).
Patch by Sebastian Vater <cdgs.basty googlemail com>.
Originally committed as revision 22974 to svn://svn.ffmpeg.org/ffmpeg/trunk
Move check on frame size after the device is opened and after
device_try_init() is attempted. If the provided size value is 0x0,
perform a VIDIOC_G_FMT ioctl() on the device, which sets size to the
current settings.
Originally committed as revision 22971 to svn://svn.ffmpeg.org/ffmpeg/trunk
they are easier to understand. Also give the add a 'u' postfix to silence
a pre-c99 compiler warning.
Originally committed as revision 22965 to svn://svn.ffmpeg.org/ffmpeg/trunk
