Scott LaVarnway [Thu, 28 Apr 2011 14:53:59 +0000 (10:53 -0400)]
Consolidated build inter predictors
Code cleanup.
Change-Id: Ic8b0167851116c64ddf08e8a3d302fb09ab61146
Scott LaVarnway [Wed, 27 Apr 2011 19:25:32 +0000 (15:25 -0400)]
Removed unused code in reconinter
The skip flag is never set by the encoder for SPLITMV.
Change-Id: I5ae6457edb3a1193cb5b05a6d61772c13b1dc506
John Koleszar [Wed, 27 Apr 2011 19:02:55 +0000 (12:02 -0700)]
Merge "SSE2/SSSE3 optimizations for build_predictors_mbuv{,_s}()."
Ronald S. Bultje [Wed, 27 Apr 2011 17:05:10 +0000 (10:05 -0700)]
SSE2/SSSE3 optimizations for build_predictors_mbuv{,_s}().
decoding
before
10.425
10.432
10.423
=10.426
after:
10.405
10.416
10.398
=10.406, 0.2% faster
encoding
before
14.252
14.331
14.250
14.223
14.241
14.220
14.221
=14.248
after
14.095
14.090
14.085
14.095
14.064
14.081
14.089
=14.086, 1.1% faster
Change-Id: I483d3d8f0deda8ad434cea76e16028380722aee2
Fritz Koenig [Wed, 27 Apr 2011 18:27:59 +0000 (11:27 -0700)]
vpxenc: remove duplicate --fps from vpxenc usage message
Fixes issue #323
Change-Id: I41c297df37afe186a8425ed2e2a95032069dcb9a
John Koleszar [Wed, 27 Apr 2011 16:04:48 +0000 (12:04 -0400)]
vpxdec: test for frame corruption
This change simply exercises the VP8D_GET_FRAME_CORRUPTED control,
outputting a warning message at the end if the bit was set for any
frames. Should never produce any output for good input.
Change-Id: Idaf6ba8f53660f47763cd563fa1485938580a37d
John Koleszar [Wed, 27 Apr 2011 16:03:45 +0000 (09:03 -0700)]
Merge "Speed up VP8DX_BOOL_DECODER_FILL"
John Koleszar [Wed, 27 Apr 2011 16:03:24 +0000 (09:03 -0700)]
Merge "Update VP8DX_BOOL_DECODER_FILL to better detect EOS"
John Koleszar [Tue, 26 Apr 2011 16:52:17 +0000 (12:52 -0400)]
Speed up VP8DX_BOOL_DECODER_FILL
The end-of-buffer check is hoisted out of the inner loop. Gives
about 0.5% improvement on x86_64.
Change-Id: I8e3ed08af7d33468c5c749af36c2dfa19677f971
John Koleszar [Tue, 26 Apr 2011 16:36:03 +0000 (12:36 -0400)]
Update VP8DX_BOOL_DECODER_FILL to better detect EOS
Allow more reliable detection of truncated bitstreams by being more
precise with the count of "virtual" bits in the value buffer.
Specifically, the VP8_LOTS_OF_BITS value is accumulated into count,
rather than being assigned, which was losing the prior value,
increasing the required tolerance when testing for the error condition.
Change-Id: Ib5172eaa57323b939c439fff8a8ab5fa38da9b69
John Koleszar [Tue, 26 Apr 2011 20:45:30 +0000 (16:45 -0400)]
Refactor calc_iframe_target_size
Combine calc_iframe_target_size, previously only used for forced
keyframes, with calc_auto_iframe_target_size, which handled most
keyframes.
Change-Id: I227051361cf46727caa5cd2b155752d2c9789364
John Koleszar [Mon, 25 Apr 2011 19:02:54 +0000 (15:02 -0400)]
Move pick_frame_size() to ratectrl.c
This is a first step in cleaning up the redundancies between
vp8_calc_{auto_,}iframe_target_size. The pick_frame_size() function is
moved to ratectrl.c, and made to be the primary interface. This means
that the various calc_*_target_size functions can be made private.
Change-Id: I66a9a62a5f9c23c818015e03f92f3757bf3bb5c8
Scott LaVarnway [Tue, 26 Apr 2011 17:12:37 +0000 (10:12 -0700)]
Merge "Test vector mismatch fix"
Scott LaVarnway [Tue, 26 Apr 2011 13:37:19 +0000 (09:37 -0400)]
Test vector mismatch fix
Fixed test vector mismatch that was introduced
in the "Removed dc_diff from MB_MODE_INFO"
(Ie2b9cdf9e0f4e8b932bbd36e0878c05bffd28931)
Change-Id: I98fa509b418e757b5cdc4baa71202f4168dc14ec
Johann [Mon, 25 Apr 2011 21:51:07 +0000 (14:51 -0700)]
Merge "remove simpler_lpf"
Johann [Mon, 25 Apr 2011 21:37:41 +0000 (17:37 -0400)]
remove simpler_lpf
the decision to run the regular or simple loopfilter is made outside the
function and managed with pointers
stop tracking the option in two places. use filter_type exclusively
Change-Id: I39d7b5d1352885efc632c0a94aaf56b72cc2fe15
John Koleszar [Mon, 25 Apr 2011 18:30:57 +0000 (14:30 -0400)]
Fix duplicate vp8_compute_frame_size_bounds
Likely introduced by a bad automatic merge from gerrit.
Change-Id: I0c6dd6ec18809cf9492f524d283fa4a3a8f4088b
John Koleszar [Mon, 25 Apr 2011 18:05:00 +0000 (11:05 -0700)]
Merge "Remove unused functions"
John Koleszar [Mon, 25 Apr 2011 17:54:20 +0000 (13:54 -0400)]
Remove unused functions
Remove estimate_min_frame_size() and calc_low_ss_err(), as they are
never referenced.
Change-Id: I3293363c14ef70b79c4678ca27aa65b345077726
John Koleszar [Mon, 25 Apr 2011 17:49:32 +0000 (10:49 -0700)]
Merge "Change rc undershoot/overshoot semantics"
John Koleszar [Mon, 25 Apr 2011 17:48:44 +0000 (10:48 -0700)]
Merge "Stereo 3D format support for vpxenc"
John Koleszar [Mon, 25 Apr 2011 17:48:13 +0000 (10:48 -0700)]
Merge "Limit size of initial keyframe in one-pass."
John Koleszar [Mon, 25 Apr 2011 17:47:57 +0000 (10:47 -0700)]
Merge "Add rc_max_intra_bitrate_pct control"
John Koleszar [Tue, 19 Apr 2011 18:05:27 +0000 (14:05 -0400)]
Limit size of initial keyframe in one-pass.
Rather than using a default size of 1/2 or 3/2 seconds for the first
frame, use a fraction of the initial buffer level to give the
application some control.
This will likely undergo further refinement as size limits on key
frames are currently under discussion on codec-devel@, but this gives
much better behavior for small buffer sizes as a starting point.
Change-Id: Ieba55b86517b81e51e6f0a9fe27aabba295acab0
John Koleszar [Mon, 25 Apr 2011 15:44:50 +0000 (11:44 -0400)]
Add rc_max_intra_bitrate_pct control
Adds a control to limit the maximum size of a keyframe, as a function of
the per-frame bitrate. See this thread[1] for more detailed discussion:
[1]: http://groups.google.com/a/webmproject.org/group/codec-devel/browse_thread/thread/
271b944a5e47ca38
Change-Id: I7337707642eb8041d1e593efc2edfdf66db02a94
John Koleszar [Mon, 25 Apr 2011 16:35:12 +0000 (09:35 -0700)]
Merge "bug fix possible keyframe context divide by zero"
James Berry [Fri, 22 Apr 2011 15:54:18 +0000 (11:54 -0400)]
bug fix possible keyframe context divide by zero
vp8_adjust_key_frame_context() divides by
estimate_keyframe_frequency() which can
return 0 in the case where --kf-max-dist=0.
Change-Id: Idfc59653478a0073187cd2aa420e98a321103daa
Alok Ahuja [Thu, 21 Apr 2011 07:50:07 +0000 (00:50 -0700)]
Stereo 3D format support for vpxenc
Create a new input parameter to allow specifying
the packed frame stereo 3d format. A default value
of mono will be written in the absence of user
specified input
Change-Id: I576d9952ab5d7e2076fbf1b282016a9a1baaa103
Johann [Mon, 25 Apr 2011 13:52:38 +0000 (06:52 -0700)]
Merge "keep values in registers during quantization"
Scott LaVarnway [Mon, 25 Apr 2011 13:45:43 +0000 (06:45 -0700)]
Merge "Removed unnecessary frame type checks"
Scott LaVarnway [Mon, 25 Apr 2011 13:45:32 +0000 (06:45 -0700)]
Merge "Removed dc_diff from MB_MODE_INFO"
Yaowu Xu [Fri, 22 Apr 2011 21:08:08 +0000 (14:08 -0700)]
Merge "make two compiler options explicit for Visual Studio projects"
Ronald S. Bultje [Thu, 21 Apr 2011 20:35:02 +0000 (16:35 -0400)]
Fix overflow in temporal_filter_apply_sse2().
The accumulator array is an integer array, so use paddd instead of paddw
to add values to it. Fixes overflows when using large --arnr-maxframes
(>8) values.
Change-Id: Iad83794caa02400a65f3ab5760f2517e082d66ae
John Koleszar [Thu, 21 Apr 2011 23:54:14 +0000 (16:54 -0700)]
Merge "Remove unused kf rate variables"
Adrian Grange [Thu, 21 Apr 2011 22:45:57 +0000 (15:45 -0700)]
Corrected format specifiers in debug print statements
The arguments to these fprintfs are int not long int so
the format specifier should be "%d" and not "%ld". This
was writing garbage in the linux build.
Change-Id: I3d2aa8a448d52e6dc08858d825bf394929b47cf3
Yaowu Xu [Thu, 21 Apr 2011 20:27:42 +0000 (13:27 -0700)]
make two compiler options explicit for Visual Studio projects
This patch changes the release configuration of MS VS projects to
explicitly use two compiler options "Maximize Speed (/O2)" and
"Favor fast code(/Ot)".
Change-Id: I0bf8343d9ca195851332b91ec69c69ee4e31ce2a
Johann [Wed, 13 Apr 2011 20:38:02 +0000 (16:38 -0400)]
keep values in registers during quantization
add an sse4 quantizer so we can use pinsrw/pextrw and keep values in xmm
registers instead of proxying through the stack. and as long as we're
bumping up, use some ssse3 instructions in the EOB detection (see ssse3
fast quantizer)
pick up about a percent on 32bit and about two on 64bit.
Change-Id: If15abba0e8b037a1d231c0edf33501545c9d9363
Scott LaVarnway [Thu, 21 Apr 2011 18:59:42 +0000 (14:59 -0400)]
Removed unnecessary frame type checks
ref_frame is set to INTRA_FRAME for keyframes. The B_PRED
mode is only used in intra frames.
Change-Id: I9bac8bec7c736300d47994f3cb570329edf11ec0
Scott LaVarnway [Thu, 21 Apr 2011 18:38:36 +0000 (14:38 -0400)]
Removed dc_diff from MB_MODE_INFO
The dc_diff flag is used to skip loopfiltering. Instead
of setting this flag in the decoder/encoder, we now check
for this condition in the loopfilter.
Change-Id: Ie2b9cdf9e0f4e8b932bbd36e0878c05bffd28931
Scott LaVarnway [Wed, 20 Apr 2011 19:45:12 +0000 (15:45 -0400)]
Removed force_no_skip
force_no_skip is always set to zero.
Change-Id: I89b61c5e0bee34627a9c07c05f3517e1db76af77
Scott LaVarnway [Tue, 19 Apr 2011 15:42:15 +0000 (11:42 -0400)]
Removed redundant checks of the mode_info_context flags
Code cleanup. The build inter predictor functions are
redundantly checking the mode_info_context for either
INTRA_FRAME or SPLITMV.
Change-Id: I4d58c3a5192a4c2cec5c24ab1caf608bf13aebfb
John Koleszar [Tue, 19 Apr 2011 20:08:45 +0000 (16:08 -0400)]
Remove unused kf rate variables
Remove tot_key_frame_bits and prior_key_frame_size[] as they were
tracked but never used. Remove intra_frame_target, as it was only
used to initialize prior_key_frame_size.
Refactor vp8_adjust_key_frame_context() some to remove unnecessary
calculations.
Change-Id: Icbc2c83d2b90e184be03e6f9679e678f3a4bce8f
Johann [Fri, 15 Apr 2011 14:05:20 +0000 (10:05 -0400)]
modify SAVE_XMM for potential 64bit use
the win64 abi requires saving and restoring xmm6:xmm15. currently
SAVE_XMM and RESTORE XMM only allow for saving xmm6:xmm7. allow
specifying the highest register used and if the stack is unaligned.
Change-Id: Ica5699622ffe3346d3a486f48eef0206c51cf867
Johann [Tue, 19 Apr 2011 13:32:10 +0000 (06:32 -0700)]
Merge "Add save/restore xmm registers in x86 assembly code"
Johann [Thu, 7 Apr 2011 17:17:22 +0000 (13:17 -0400)]
Add save/restore xmm registers in x86 assembly code
Went through the code and fixed it. Verified on Windows.
Where possible, remove dependencies on xmm[67]
Current code relies on pushing rbp to the stack to get 16 byte
alignment. This broke when rbp wasn't pushed
(vp8/encoder/x86/sad_sse3.asm). Work around this by using unaligned
memory accesses. Revisit this and the offsets in
vp8/encoder/x86/sad_sse3.asm in another change to SAVE_XMM.
Change-Id: I5f940994d3ebfd977c3d68446cef20fd78b07877
Yunqing Wang [Mon, 18 Apr 2011 20:20:04 +0000 (13:20 -0700)]
Merge "Use sub-pixel search's SSE in mode selection"
Yunqing Wang [Mon, 18 Apr 2011 19:48:34 +0000 (15:48 -0400)]
Use sub-pixel search's SSE in mode selection
Passed SSE from sub-pixel search back to pick_inter_mode
function, which is compared with the encode_breakout to
see if we could skip evaluating the remaining modes.
Change-Id: I4a86442834f0d1b880a19e21ea52d17d505f941d
Yunqing Wang [Mon, 18 Apr 2011 17:11:41 +0000 (10:11 -0700)]
Merge "Handle long delay between video frames in multi-thread decoder(issue 312)"
Johann [Mon, 18 Apr 2011 17:03:40 +0000 (10:03 -0700)]
Merge "store quant_shift as an unsigned char"
Yaowu Xu [Mon, 18 Apr 2011 16:53:26 +0000 (09:53 -0700)]
Merge "Add spin-wait pause intrinsic for Windows x64 platform."
Yaowu Xu [Mon, 18 Apr 2011 14:44:33 +0000 (07:44 -0700)]
Merge "fixed an overflow in ssim calculation"
Scott LaVarnway [Fri, 15 Apr 2011 19:09:07 +0000 (15:09 -0400)]
Removed unused timers
Change-Id: I209803b9dbed2b2f6d02258fd7a3963a6645f4ab
John Koleszar [Mon, 18 Apr 2011 13:07:57 +0000 (06:07 -0700)]
Merge "added -fomit-frame-pointer flag for gcc builds"
Yunqing Wang [Fri, 15 Apr 2011 16:57:15 +0000 (12:57 -0400)]
Handle long delay between video frames in multi-thread decoder(issue 312)
This is reported by m...@hesotech.de (see issue 312):
"The decoder causes an access violation
when you decode the first frame, then make a pause of about
60 seconds and then decode further frames. But only if
vpx_codec_dec_cfg_t.threads> 1.
This is caused by a timeout of WaitForSingleObject.
When I change the definition of VPXINFINITE to INFINITE(0xFFFFFFFF),
the problem is solved."
Reproduced the crash and verified the changes on Windows platform.
This brings the behavior inline with the other platforms using sem_wait().
Change-Id: I27b32f90bce05846ef2684b50f7a88f292299da1
Johann [Fri, 15 Apr 2011 20:32:54 +0000 (13:32 -0700)]
Merge "remove dead code, add missing RESTORE_XMM"
Scott LaVarnway [Fri, 15 Apr 2011 19:59:21 +0000 (15:59 -0400)]
added -fomit-frame-pointer flag for gcc builds
According to the docs, this should have been enabled, but
the disassembled output shows otherwise. This improved
the encode/decode performance.
Change-Id: I45ad7e6d299b89ac3166d7ef7da75b74994344c6
Johann [Fri, 15 Apr 2011 17:43:24 +0000 (13:43 -0400)]
remove executable bit
source files are not executable
Change-Id: Id2c7294695a22217468426423979f68f02d82340
Adrian Grange [Fri, 15 Apr 2011 15:37:19 +0000 (08:37 -0700)]
Merge "Fix usage of value returned by vp8_pick_intra4x4mby_modes"
Yunqing Wang [Fri, 15 Apr 2011 15:17:03 +0000 (08:17 -0700)]
Merge "Reduce unnecessary distortion computation"
Johann [Fri, 15 Apr 2011 14:11:53 +0000 (10:11 -0400)]
remove dead code, add missing RESTORE_XMM
vp8_filter_block1d16_h4_ssse3 was never called
because UNSHADOW_ARGS moves the stack by 'mov rsp, rbp', the issue was
masked. however, if/when win64 used those registers for persistant data,
issues could/will arise.
Change-Id: I56d6effca0aeba1f86082689771cb10145d39651
John Koleszar [Fri, 15 Apr 2011 12:44:39 +0000 (08:44 -0400)]
Fix off-by-one in copy_and_extend_plane
Should only copy h lines, not h+1.
Change-Id: I802a85686635900459c6dc79596189033e5298d8
Yunqing Wang [Thu, 14 Apr 2011 19:53:33 +0000 (15:53 -0400)]
Reduce unnecessary distortion computation
In vp8_pick_inter_mode(), for NEWMV mode, use the error result got
from motion search as distortion. This helps performance in real-
time mode.
Change-Id: I398c4e46cc5381f7d874e748cf78827ef0e0860c
John Koleszar [Thu, 14 Apr 2011 19:35:01 +0000 (12:35 -0700)]
Merge "Refactor lookahead ring buffer"
Fritz Koenig [Thu, 14 Apr 2011 18:56:18 +0000 (11:56 -0700)]
Merge "Use consistent delimiters."
Adrian Grange [Thu, 14 Apr 2011 17:31:56 +0000 (10:31 -0700)]
Fix usage of value returned by vp8_pick_intra4x4mby_modes
The value of distortion2 returned by vp8_pick_intra4x4mby_modes
was being overwritten by the value returned by get16x16prederror
before it was tested.
Change-Id: If00e80332b272c5545c3a7e381c8041e8319b41a
Johann [Thu, 14 Apr 2011 15:55:22 +0000 (08:55 -0700)]
Merge "update configure for ios sdk 4.3"
Joshua Bleecher Snyder [Thu, 14 Apr 2011 13:28:49 +0000 (09:28 -0400)]
update configure for ios sdk 4.3
update for the latest version of the ios sdk. adding
usr/lib/system fixes a missing libcache.dylib issue
make isysroot path more DRY
Change-Id: Ib748ef3dac3cac2e4848fbffa1e9a0112eac826b
Fritz Koenig [Wed, 13 Apr 2011 22:06:17 +0000 (15:06 -0700)]
Use consistent delimiters.
opsnr.stt file was using \t for delimiters on everything
except between VPXSSIM and Time.
Change-Id: I6284c4e40c05ff642bf4b0170dca062c279a42df
Adrian Grange [Wed, 13 Apr 2011 19:56:46 +0000 (12:56 -0700)]
Fixed use of early breakout in vp8_pick_intra4x4mby_modes
Index i is used to detect early breakout from the first loop, but
its value is lost due to reuse in the second for loop. I moved
the position of the second loop and did some format cleanup.
Change-Id: I02780eae1bd89df4b6c000fb8a018b0837aac2e5
John Koleszar [Wed, 13 Apr 2011 18:00:18 +0000 (14:00 -0400)]
Refactor lookahead ring buffer
This patch cleans up the source buffer storage and copy mechanism to
allow access through a standard push/pop/peek interface. This approach
also avoids an extra copy in the case where the source is not a
multiple of 16, fixing issue #102.
Change-Id: I05808c39f5743625cb4c7af54cc841b9b10fdbd9
Johann [Wed, 13 Apr 2011 17:45:58 +0000 (13:45 -0400)]
store quant_shift as an unsigned char
in encodframe.c, quant_shift is set to 0 or 1 in vp8cx_invert_quant
only use 8 bits to store this, instead of 16. will allow saving an
xmm register in an updated version of the regular quantize
Change-Id: Ie88c47fe2aff5af0283dab1147fb2791e4b12f90
John Koleszar [Mon, 11 Apr 2011 15:29:23 +0000 (11:29 -0400)]
Change rc undershoot/overshoot semantics
This patch changes the rc_undershoot_pct and rc_overshoot_pct controls
to set the "aggressiveness" of rate adaptation, by limiting the
amount of difference between the target buffer level and the actual
buffer level which is applied to the target frame rate for this frame.
This patch was initially provided by arosenberg at logitech.com as
an attachment to issue #270. It was modified to separate these controls
from the other unrelated modifications in that patch, as well as to
use the pre-existing variables rather than introducing new ones.
Change-Id: Id542e3f5667dd92d857d5eabf29878f2fd730a62
John Koleszar [Tue, 12 Apr 2011 13:59:00 +0000 (06:59 -0700)]
Merge "Bugfix for error accumulator stats"
John Koleszar [Mon, 11 Apr 2011 17:05:08 +0000 (13:05 -0400)]
Bugfix for error accumulator stats
Previous to commit de4e9e3, there was an early return in the alt-ref
case that was inadvertantly removed when the function was refactored
to return void. This patch restores the prior behavior.
Change-Id: I783ffd594a4690297e2742f99526fd7ad67698b2
John Koleszar [Tue, 12 Apr 2011 12:34:12 +0000 (05:34 -0700)]
Merge "Fix encoder range check for frame width and height"
Attila Nagy [Tue, 12 Apr 2011 12:01:22 +0000 (15:01 +0300)]
Fix encoder range check for frame width and height
14 bits available in the bistream => valid range [1..16383]
Removed unused local vars.
Change-Id: Icf3385e47a9fa13af70053129c2248671f285583
Yunqing Wang [Mon, 11 Apr 2011 19:55:04 +0000 (15:55 -0400)]
Set cpu_used range to [-16, 16] in real-time mode
Remove encoding speed limitation in real-time mode.
Change-Id: Ib5e35d8bb522b2a25f3e4ad5cfe2788ebebb3617
Yunqing Wang [Thu, 7 Apr 2011 20:57:25 +0000 (16:57 -0400)]
Define RDCOST only once
Clean up the code.
Change-Id: I7db048efa4d972b528d553a7921bc45979621129
John Koleszar [Mon, 11 Apr 2011 14:34:40 +0000 (10:34 -0400)]
Remove unused files
Change-Id: I36ca3f2f4620358033da34daf764f0b388dacd08
Yunqing Wang [Fri, 8 Apr 2011 20:15:26 +0000 (16:15 -0400)]
Fix input MV for full search
Input MV needs to be modified to full-pixel precision.
Change-Id: Ic5d78e41bf27077e325024332b9fe89f76c44f0c
Johann Koenig [Fri, 8 Apr 2011 17:05:47 +0000 (10:05 -0700)]
Merge "use asm_offsets with vp8_fast_quantize_b_sse3"
John Koleszar [Fri, 8 Apr 2011 15:20:32 +0000 (08:20 -0700)]
Merge "Error accumulator stats bug."
Paul Wilkins [Fri, 8 Apr 2011 13:21:36 +0000 (14:21 +0100)]
Error accumulator stats bug.
The error accumulator stats values cpi->prediction_error and
cpi->intra_error were being populated with rd values not
distortion values.
These are only "currently" used in a limited way for RT compress
key frame detection.
Change-Id: I2702ba1cab6e49ab8dc096ba75b6b34ab3573021
Jim Bankoski [Mon, 28 Mar 2011 23:39:05 +0000 (16:39 -0700)]
fixed an overflow in ssim calculation
This commit fixed an overflow in ssim calculation, added register
save and restore to make sure assembly code working for x64 platform.
It also changed the sampling points to every 4x4 instead of 8x8 and
adjusted the constants in SSIM calculation to match the scale of
previous VPXSSIM.
Change-Id: Ia4dbb8c69eac55812f4662c88ab4653b6720537b
Johann Koenig [Thu, 7 Apr 2011 20:40:05 +0000 (16:40 -0400)]
use asm_offsets with vp8_fast_quantize_b_sse3
on the same order as the sse2 fast quantize change: ~2%
except for 32bit. only a slight improvment there.
Change-Id: Iff80e5f1ce7e646eebfdc8871405458ff911986b
James Berry [Thu, 7 Apr 2011 19:03:49 +0000 (15:03 -0400)]
Use correct 32 bit comparisons for SAD breakout.
Rax updated to eax to avoid uninitialized memory
usage.
Change-Id: Iedb953f104329ede2a786fc648a47f1be2f3798a
Johann [Wed, 6 Apr 2011 17:53:55 +0000 (10:53 -0700)]
Merge "use asm_offsets with vp8_fast_quantize_b_sse2"
Yunqing Wang [Wed, 6 Apr 2011 13:12:13 +0000 (06:12 -0700)]
Merge "Minor modification"
Yunqing Wang [Wed, 6 Apr 2011 13:08:47 +0000 (09:08 -0400)]
Minor modification
A small change.
Change-Id: I2e7726e58370a95d0319361f4f6ad231138d1328
Johann [Thu, 24 Mar 2011 17:31:10 +0000 (13:31 -0400)]
use asm_offsets with vp8_fast_quantize_b_sse2
on the same order as the regular quantize change: ~2%
Change-Id: I5c9eec18e89ae7345dd96945cb740e6f349cee86
Scott LaVarnway [Mon, 4 Apr 2011 18:18:31 +0000 (14:18 -0400)]
Fixed unused variable warnings for firstpass.c
Change-Id: I8378a9a541ade2f098359a7b20fa08e6c1596d80
John Koleszar [Mon, 4 Apr 2011 15:58:25 +0000 (08:58 -0700)]
Merge "Slightly simplify vp8_decode_mb_tokens."
Johann [Mon, 4 Apr 2011 15:56:25 +0000 (08:56 -0700)]
Merge "tweak vp8_regular_quantize_b_sse2"
Gaute Strokkenes [Mon, 4 Apr 2011 15:47:22 +0000 (16:47 +0100)]
Slightly simplify vp8_decode_mb_tokens.
Change-Id: I0058ba7dcfc50a3374b712197639ac337f8726be
Yunqing Wang [Mon, 4 Apr 2011 15:40:51 +0000 (08:40 -0700)]
Merge "Use full-pixel MV in mvsadcost calculation"
John Koleszar [Mon, 4 Apr 2011 15:29:50 +0000 (08:29 -0700)]
Merge "support obj_int_extract on cygwin"
Yunqing Wang [Fri, 1 Apr 2011 20:41:58 +0000 (16:41 -0400)]
Use full-pixel MV in mvsadcost calculation
MV sad cost error is only used in full-pixel motion search,
which only need full-pixel resolution instead of quarter-pixel
resolution. This change reduced mvsadcost table size, and
removed unneccessary pamameter passing since this table is
constant once it is generated.
Change-Id: I9f931e55f6abc3c99011321f1dfb2f3562e6f6b0
Johann [Thu, 31 Mar 2011 20:35:22 +0000 (16:35 -0400)]
support obj_int_extract on cygwin
cygwin doesn't support _sopen. drop down to the lowest common
denominator and merge main for all platforms. this also opens the door
for supporting multiple object formats with a single binary.
Change-Id: I7cd45091639d447434e6d5db2e19cfc9988f8630
John Koleszar [Fri, 1 Apr 2011 14:55:55 +0000 (07:55 -0700)]
Merge "vpxenc: die on realloc failures"
Johann [Fri, 1 Apr 2011 13:58:23 +0000 (09:58 -0400)]
tweak vp8_regular_quantize_b_sse2
rather than look up rc in the zig zag table, embed it in the macro. this
also allows us to shuffle some values in the macro and keep *d in rsi
gains of about the same order as the obj_int_extract implementation: ~2%
Change-Id: Ib7252dd10eee66e0af8b0e567426122781dc053d