| Commit message (Collapse) | Author | Age |
|\
| |
| |
| |
| |
| |
| | |
* commit '2143948381c8118bdc2f50bd4079520b9885bd54':
Drop unnecessary av_unused attributes.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| | |
|
| |
| |
| |
| |
| | |
Fixes CID1087088
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit 'e0c16e4e3259cf50b5bac4c23bb6e517f397c74b':
mpegvideo: move mpegvideo formats-related defines to mpegutils.h
Conflicts:
libavcodec/h264_cabac.c
libavcodec/h264_cavlc.c
libavcodec/h264_mvpred.h
libavcodec/svq1enc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| | |
|
| |
| |
| |
| | |
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| | |
closer to where the issue is
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| | |
The newly added optimizations do not work with the unchecked reader
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit 'fb0c9d41d685abb58575c5482ca33b8cd457c5ec':
avutil: remove timer.h include from internal.h
Conflicts:
libavcodec/ffv1dec.c
libavutil/internal.h
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
Added libavutil/timer.h include to all files with {START,STOP}_TIMER.
|
| |
| |
| |
| |
| | |
Found-by: Dale Curtis <dalecurtis@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| | |
* qatar/master:
h264_cabac: Mark functions calling decode_cabac_residual_internal as noinline
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| | |
This ensures that decode_cabac_residual_internal actually does get inlined,
which it otherwise does not, even though it is marked as always_inline.
|
| | |
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit '23e85be58fc64b2e804e68b0034a08a6d257e523':
h264: add a parameter to the CHROMA444 macro.
h264: add a parameter to the CHROMA422 macro.
Conflicts:
libavcodec/h264.c
libavcodec/h264.h
libavcodec/h264_cavlc.c
libavcodec/h264_loopfilter.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit '6d2b6f21eb45ffbda1103c772060303648714832':
h264: add a parameter to the CABAC macro.
h264: add a parameter to the FIELD_OR_MBAFF_PICTURE macro.
Conflicts:
libavcodec/h264.c
libavcodec/h264_cabac.c
libavcodec/h264_cavlc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit '7fa00653a550c0d24b3951c0f9fed6350ecf5ce4':
h264: add a parameter to the FIELD_PICTURE macro.
h264: add a parameter to the FRAME_MBAFF macro.
Conflicts:
libavcodec/h264.c
libavcodec/h264_loopfilter.c
libavcodec/h264_refs.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit 'da6be8fcec16a94d8084bda8bb8a0a411a96bcf7':
h264: add a parameter to the MB_FIELD macro.
h264: add a parameter to the MB_MBAFF macro.
Conflicts:
libavcodec/h264.c
libavcodec/h264_cabac.c
libavcodec/h264_cavlc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
| |
| |
| |
| | |
This way it does not look like a constant.
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit '759001c534287a96dc96d1e274665feb7059145d':
lavc decoders: work with refcounted frames.
Anton Khirnov (1):
lavc decoders: work with refcounted frames.
Clément Bœsch (47):
lavc/ansi: reset file
lavc/ansi: re-do refcounted frame changes from Anton
fraps: reset file
lavc/fraps: switch to refcounted frames
gifdec: reset file
lavc/gifdec: switch to refcounted frames
dsicinav: resolve conflicts
smc: resolve conflicts
zmbv: resolve conflicts
rpza: resolve conflicts
vble: resolve conflicts
xxan: resolve conflicts
targa: resolve conflicts
vmnc: resolve conflicts
utvideodec: resolve conflicts
tscc: resolve conflicts
ulti: resolve conflicts
ffv1dec: resolve conflicts
dnxhddec: resolve conflicts
v210dec: resolve conflicts
vp3: resolve conflicts
vcr1: resolve conflicts
v210x: resolve conflicts
wavpack: resolve conflicts
pngdec: fix compilation
roqvideodec: resolve conflicts
pictordec: resolve conflicts
mdec: resolve conflicts
tiertexseqv: resolve conflicts
smacker: resolve conflicts
vb: resolve conflicts
vqavideo: resolve conflicts
xl: resolve conflicts
tmv: resolve conflicts
vmdav: resolve conflicts
truemotion1: resolve conflicts
truemotion2: resolve conflicts
lcldec: fix compilation
libcelt_dec: fix compilation
qdrw: fix compilation
r210dec: fix compilation
rl2: fix compilation
wnv1: fix compilation
yop: fix compilation
tiff: resolve conflicts
interplayvideo: fix compilation
qpeg: resolve conflicts (FIXME/TESTME).
Hendrik Leppkes (33):
012v: convert to refcounted frames
8bps: fix compilation
8svx: resolve conflicts
4xm: resolve conflicts
aasc: resolve conflicts
bfi: fix compilation
aura: fix compilation
alsdec: resolve conflicts
avrndec: convert to refcounted frames
avuidec: convert to refcounted frames
bintext: convert to refcounted frames
cavsdec: resolve conflicts
brender_pix: convert to refcounted frames
cinepak: resolve conflicts
cinepak: avoid using AVFrame struct directly in private context
cljr: fix compilation
cpia: convert to refcounted frames
cscd: resolve conflicts
iff: resolve conflicts and do proper conversion to refcounted frames
4xm: fix reference frame handling
cyuv: fix compilation
dxa: fix compilation
eacmv: fix compilation
eamad: fix compilation
eatgv: fix compilation
escape124: remove unused variable.
escape130: convert to refcounted frames
evrcdec: convert to refcounted frames
exr: convert to refcounted frames
mvcdec: convert to refcounted frames
paf: properly free the frame data on decode close
sgirle: convert to refcounted frames
lavfi/moviesrc: use refcounted frames
Michael Niedermayer (56):
Merge commit '759001c534287a96dc96d1e274665feb7059145d'
resolve conflicts in headers
motion_est: resolve conflict
mpeg4videodec: fix conflicts
dpcm conflict fix
dpx: fix conflicts
indeo3: resolve confilcts
kmvc: resolve conflicts
kmvc: resolve conflicts
h264: resolve conflicts
utils: resolve conflicts
rawdec: resolve conflcits
mpegvideo: resolve conflicts
svq1enc: resolve conflicts
mpegvideo: dont clear data, fix assertion failure on fate vsynth1 with threads
pthreads: resolve conflicts
frame_thread_encoder: simple compilefix not yet tested
snow: update to buffer refs
crytsalhd: fix compile
dirac: switch to new API
sonic: update to new API
svq1: resolve conflict, update to new API
ffwavesynth: update to new buffer API
g729: update to new API
indeo5: fix compile
j2kdec: update to new buffer API
linopencore-amr: fix compile
libvorbisdec: update to new API
loco: fix compile
paf: update to new API
proresdec: update to new API
vp56: update to new api / resolve conflicts
xface: convert to refcounted frames
xan: fix compile&fate
v408: update to ref counted buffers
v308: update to ref counted buffers
yuv4dec: update to ref counted buffers
y41p: update to ref counted frames
xbm: update to refcounted frames
targa_y216: update to refcounted buffers
qpeg: fix fate/crash
cdxl: fix fate
tscc: fix reget buffer useage
targa_y216dec: fix style
msmpeg4: fix fate
h264: ref_picture() copy fields that have been lost too
update_frame_pool: use channel field
h264: Put code that prevents deadlocks back
mpegvideo: dont allow last == current
wmalossless: fix buffer ref messup
ff_alloc_picture: free tables in case of dimension mismatches
h264: fix null pointer dereference and assertion failure
frame_thread_encoder: update to bufrefs
ec: fix used arrays
snowdec: fix off by 1 error in dimensions check
h264: disallow single unpaired fields as references of frames
Paul B Mahol (2):
lavc/vima: convert to refcounted frames
sanm: convert to refcounted frames
Conflicts:
libavcodec/4xm.c
libavcodec/8bps.c
libavcodec/8svx.c
libavcodec/aasc.c
libavcodec/alsdec.c
libavcodec/anm.c
libavcodec/ansi.c
libavcodec/avs.c
libavcodec/bethsoftvideo.c
libavcodec/bfi.c
libavcodec/c93.c
libavcodec/cavsdec.c
libavcodec/cdgraphics.c
libavcodec/cinepak.c
libavcodec/cljr.c
libavcodec/cscd.c
libavcodec/dnxhddec.c
libavcodec/dpcm.c
libavcodec/dpx.c
libavcodec/dsicinav.c
libavcodec/dvdec.c
libavcodec/dxa.c
libavcodec/eacmv.c
libavcodec/eamad.c
libavcodec/eatgq.c
libavcodec/eatgv.c
libavcodec/eatqi.c
libavcodec/error_resilience.c
libavcodec/escape124.c
libavcodec/ffv1.h
libavcodec/ffv1dec.c
libavcodec/flicvideo.c
libavcodec/fraps.c
libavcodec/frwu.c
libavcodec/g723_1.c
libavcodec/gifdec.c
libavcodec/h264.c
libavcodec/h264.h
libavcodec/h264_direct.c
libavcodec/h264_loopfilter.c
libavcodec/h264_refs.c
libavcodec/huffyuvdec.c
libavcodec/idcinvideo.c
libavcodec/iff.c
libavcodec/indeo2.c
libavcodec/indeo3.c
libavcodec/internal.h
libavcodec/interplayvideo.c
libavcodec/ivi_common.c
libavcodec/jvdec.c
libavcodec/kgv1dec.c
libavcodec/kmvc.c
libavcodec/lagarith.c
libavcodec/libopenjpegdec.c
libavcodec/mdec.c
libavcodec/mimic.c
libavcodec/mjpegbdec.c
libavcodec/mjpegdec.c
libavcodec/mmvideo.c
libavcodec/motion_est.c
libavcodec/motionpixels.c
libavcodec/mpc7.c
libavcodec/mpeg12.c
libavcodec/mpeg4videodec.c
libavcodec/mpegvideo.c
libavcodec/mpegvideo.h
libavcodec/msrle.c
libavcodec/msvideo1.c
libavcodec/nuv.c
libavcodec/options_table.h
libavcodec/pcx.c
libavcodec/pictordec.c
libavcodec/pngdec.c
libavcodec/pnmdec.c
libavcodec/pthread.c
libavcodec/qpeg.c
libavcodec/qtrle.c
libavcodec/r210dec.c
libavcodec/rawdec.c
libavcodec/roqvideodec.c
libavcodec/rpza.c
libavcodec/smacker.c
libavcodec/smc.c
libavcodec/svq1dec.c
libavcodec/svq1enc.c
libavcodec/targa.c
libavcodec/tiertexseqv.c
libavcodec/tiff.c
libavcodec/tmv.c
libavcodec/truemotion1.c
libavcodec/truemotion2.c
libavcodec/tscc.c
libavcodec/ulti.c
libavcodec/utils.c
libavcodec/utvideodec.c
libavcodec/v210dec.c
libavcodec/v210x.c
libavcodec/vb.c
libavcodec/vble.c
libavcodec/vcr1.c
libavcodec/vmdav.c
libavcodec/vmnc.c
libavcodec/vp3.c
libavcodec/vp56.c
libavcodec/vp56.h
libavcodec/vp6.c
libavcodec/vqavideo.c
libavcodec/wavpack.c
libavcodec/xl.c
libavcodec/xxan.c
libavcodec/zmbv.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| | |
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit 'c242bbd8b6939507a1a6fb64101b0553d92d303f':
Remove unnecessary dsputil.h #includes
Conflicts:
libavcodec/ffv1.c
libavcodec/h261dec.c
libavcodec/h261enc.c
libavcodec/h264pred.c
libavcodec/lpc.h
libavcodec/mjpegdec.c
libavcodec/rectangle.h
libavcodec/x86/idct_sse2_xvid.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Instead, keep them in the bitstream buffer until we read them verbatim,
this saves a memcpy() and a subsequent clearing of the target buffer.
decode_cabac+decode_mb for a sample file (CAPM3_Sony_D.jsv) goes from
6121.4 to 6095.5 cycles, i.e. 26 cycles faster.
Signed-off-by: Martin Storsjö <martin@martin.st>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Instead, keep them in the bitstream buffer until we read them verbatim,
this saves a memcpy() and a subsequent clearing of the target buffer.
decode_cabac+decode_mb for a sample file (CAPM3_Sony_D.jsv) goes from
6121.4 to 6095.5 cycles, i.e. 26 cycles faster.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* qatar/master:
h264: deMpegEncContextize
Conflicts:
libavcodec/dxva2_h264.c
libavcodec/h264.c
libavcodec/h264.h
libavcodec/h264_cabac.c
libavcodec/h264_cavlc.c
libavcodec/h264_loopfilter.c
libavcodec/h264_mb_template.c
libavcodec/h264_parser.c
libavcodec/h264_ps.c
libavcodec/h264_refs.c
libavcodec/h264_sei.c
libavcodec/svq3.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Most of the changes are just trivial are just trivial replacements of
fields from MpegEncContext with equivalent fields in H264Context.
Everything in h264* other than h264.c are those trivial changes.
The nontrivial parts are:
1) extracting a simplified version of the frame management code from
mpegvideo.c. We don't need last/next_picture anymore, since h264 uses
its own more complex system already and those were set only to appease
the mpegvideo parts.
2) some tables that need to be allocated/freed in appropriate places.
3) hwaccels -- mostly trivial replacements.
for dxva, the draw_horiz_band() call is moved from
ff_dxva2_common_end_frame() to per-codec end_frame() callbacks,
because it's now different for h264 and MpegEncContext-based
decoders.
4) svq3 -- it does not use h264 complex reference system, so I just
added some very simplistic frame management instead and dropped the
use of ff_h264_frame_start(). Because of this I also had to move some
initialization code to svq3.
Additional fixes for chroma format and bit depth changes by
Janne Grunau <janne-libav@jannau.net>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
|
| |
| |
| |
| |
| |
| |
| | |
fix out of array read
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* commit '88bd7fdc821aaa0cbcf44cf075c62aaa42121e3f':
Drop DCTELEM typedef
Conflicts:
libavcodec/alpha/dsputil_alpha.h
libavcodec/alpha/motion_est_alpha.c
libavcodec/arm/dsputil_init_armv6.c
libavcodec/bfin/dsputil_bfin.h
libavcodec/bfin/pixels_bfin.S
libavcodec/cavs.c
libavcodec/cavsdec.c
libavcodec/dct-test.c
libavcodec/dnxhdenc.c
libavcodec/dsputil.c
libavcodec/dsputil.h
libavcodec/dsputil_template.c
libavcodec/eamad.c
libavcodec/h264_cavlc.c
libavcodec/h264idct_template.c
libavcodec/mpeg12.c
libavcodec/mpegvideo.c
libavcodec/mpegvideo.h
libavcodec/mpegvideo_enc.c
libavcodec/ppc/dsputil_altivec.c
libavcodec/proresdsp.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| | |
It does not help as an abstraction and adds dsputil dependencies.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The variable is copied to subsequent threads at the same time, so this
may cause wrong ref_count[] values to be copied to subsequent threads.
This bug was found using TSAN.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
|
| |
| |
| |
| |
| | |
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The variable is copied to subsequent threads at the same time, so this
may cause wrong ref_count[] values to be copied to subsequent threads.
This bug was found using TSAN and Helgrind.
Original patch by Ronald, adapted with a local_ref_count by Clément,
following the suggestion of Michael Niedermayer.
Signed-off-by: Clément Bœsch <clement.boesch@smartjog.com>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* qatar/master:
avcodec: Convert some commented-out printf/av_log instances to av_dlog
avcodec: Drop silly and/or broken printf debug output
avcodec: Drop some silly commented-out av_log() invocations
avformat: Convert some commented-out printf/av_log instances to av_dlog
avformat: Remove non-compiling and/or silly commented-out printf/av_log statements
Remove some silly disabled code.
ac3dec: ensure get_buffer() gets a buffer for the correct number of channels
Conflicts:
libavcodec/dnxhddec.c
libavcodec/ffv1.c
libavcodec/h264.c
libavcodec/h264_parser.c
libavcodec/mjpegdec.c
libavcodec/motion_est_template.c
libavcodec/mpegaudiodec.c
libavcodec/mpegvideo_enc.c
libavcodec/put_bits.h
libavcodec/ratecontrol.c
libavcodec/wmaenc.c
libavdevice/timefilter.c
libavformat/asfdec.c
libavformat/avidec.c
libavformat/avienc.c
libavformat/flvenc.c
libavformat/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| | |
|
| | |
|
| |
| |
| |
| | |
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* qatar/master:
log: Only include unistd.h if configure found it
ape: create audio stream before reading tags.
mov: make a length variable larger.
image2: Add "start_number" private option to the demuxer
image2: Add "start_number" private option to the muxer
avconv: remove a forgotten debugging printf.
avconv: use more descriptive names for hardcoded filters.
avconv: remove redundant handling of async.
doc/filters: fix typo.
h264: use asm cabac reader under a generic condition
Conflicts:
ffmpeg.c
libavformat/img2dec.c
libavformat/img2enc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This removes a dependency on implementation details from generic
code and allows easy addition of the equivalent optimisation for
other architectures than x86.
Signed-off-by: Mans Rullgard <mans@mansr.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This adds a hand-optimized assembly version for get_cabac much like the
existing one, but it works if the table offsets are RIP-relative.
Compared to the non-RIP-relative version this adds 2 lea instructions
and it needs one extra register. get_cabac() gets about 40% faster, for
an overall speedup of about 5%.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The reason is this is easier for PIC code (in particular on darwin...).
Keep the old names as pointers (static in cabac_functions.h so gcc
knows these are just immediate offsets) so the c code can nicely stay the same
(alternatively could use offsets directly in the functions needing the
tables). This should produce the same code as before with non-pic and better
code (confirmed) with pic.
The assembly uses the new table but still won't work for PIC case.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This adds a hand-optimized assembly version for get_cabac much like the
existing one, but it works if the table offsets are RIP-relative.
Compared to the non-RIP-relative version this adds 2 lea instructions
and it needs one extra register.
There is a surprisingly large performance improvement over the c version (more
so than the generated assembly seems to suggest) just in get_cabac, I measured
roughly 40% faster for get_cabac on a K8. However, overall the difference is
not that big, I measured roughly 5% on a test clip on a K8 and a Core2.
Hopefully it still compiles on x86 32bit...
Now that only one table is used, there's some chance even darwin as compiles
this (apparently the label arithmetic used previously doesn't work if it
involves symbols defined in a different file, thanks to Ronald S. Bultje for
helping me with this).
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The reason is this is easier for PIC code (in particular on darwin...).
Keep the old names as pointers (static in cabac_functions.h so gcc
knows these are just immediate offsets) so the c code can nicely stay the same
(alternatively could use offsets directly in the functions needing the
tables). This should produce the same code as before with non-pic and better
code (confirmed) with pic.
The assembly uses the new table but still won't work for PIC case.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
|