Jingning Han [Thu, 8 Aug 2013 02:05:14 +0000 (19:05 -0700)]
Merge "Use low precision 32x32fdct for encodemb in speed1"
Dmitry Kovalev [Wed, 7 Aug 2013 23:32:09 +0000 (16:32 -0700)]
Merge "Using only one scale function in scale_factors struct."
Dmitry Kovalev [Wed, 7 Aug 2013 23:08:24 +0000 (16:08 -0700)]
Merge "Adding ss_size_lookup table."
Jingning Han [Wed, 7 Aug 2013 22:22:51 +0000 (15:22 -0700)]
Use low precision 32x32fdct for encodemb in speed1
The low precision 32x32 fdct has all the intermediate steps within
16-bit depth, hence allowing faster SSE2 implementation, at the
expense of larger round-trip error. It was used in the rate-distortion
optimization search loop only.
Using the low precision version, in replace of the high precision one,
affects the compression performance by about 0.7% (derf, stdhd) at
speed 0. For speed 1, it makes derf set down by only 0.017%.
Change-Id: I4e7d18fac5bea5317b91c8e7dabae143bc6b5c8b
Dmitry Kovalev [Wed, 7 Aug 2013 22:33:17 +0000 (15:33 -0700)]
Adding ss_size_lookup table.
Removing the old one bsize_from_dim_lookup. Now we have a way to determine
block size for plane using its subsampling values (ss_size_lookup). And
then we can find the number of pixels in the block (num_pels_log2_lookup).
Change-Id: I6fc981da2ae093de81741d3d78eaefed11015db9
Dmitry Kovalev [Wed, 7 Aug 2013 22:28:30 +0000 (15:28 -0700)]
Merge "Removing NMS_STATS defines."
Christian Duvivier [Mon, 5 Aug 2013 00:37:05 +0000 (17:37 -0700)]
Neon version of vp9_short_idct4x4_add.
Change-Id: Idec4cae0cb9b3a29835fd2750d354c1393d47aa4
Deb Mukherjee [Wed, 7 Aug 2013 00:28:48 +0000 (17:28 -0700)]
Merge "Clean ups of the subpel search functions"
Deb Mukherjee [Tue, 6 Aug 2013 22:53:35 +0000 (15:53 -0700)]
Clean ups of the subpel search functions
Removes some unused code and speed features, and organizes the
interfaces for fractional mv step functions for use in new speed
features to come.
In the process a new speed feature - number of iterations per
step during the subpel search - is exposed.
No change when this parameter is set as the original value of 3.
Results:
subpel_iters_per_step = 3: baseline
subpel_iters_per_step = 2: psnr -0.067%, 1% speedup
subpel_iters_per_step = 1: psnr -0.331%, 3-4% speedup
Change-Id: I2eba8a21f6461be8caf56af04a5337257a5693a8
Dmitry Kovalev [Tue, 6 Aug 2013 23:00:01 +0000 (16:00 -0700)]
Merge "Motion vector code cleanup."
Dmitry Kovalev [Tue, 6 Aug 2013 22:43:56 +0000 (15:43 -0700)]
Using only one scale function in scale_factors struct.
Functions scale_mv_q4 and scale_mv_q3_to_q4 were almost identical except
q3->q4 conversion in scale_mv_q3_to_q4. Now q3->q4 conversion happens
directly in vp9_build_inter_predictor.
Also adding useful constants: SUBPEL_BITS and SUBPEL_MASK.
Change-Id: Ia0a6ad2ac07c45fdf95a5139ece6286c035e9639
Jingning Han [Tue, 6 Aug 2013 21:47:30 +0000 (14:47 -0700)]
Merge "Place holder for high-precision 32x32 fdct"
Jim Bankoski [Tue, 6 Aug 2013 15:25:10 +0000 (08:25 -0700)]
variance x86inc guards
also fixed bug in sad calcs
Change-Id: I6571fcbe37556c16ae32be66dc0fd879852aac1d
Jim Bankoski [Tue, 6 Aug 2013 14:44:56 +0000 (07:44 -0700)]
sse3 intrapred x86inc protected
Change-Id: I4a3c83119cdf8a205920034c8019d855d5504605
Deb Mukherjee [Tue, 6 Aug 2013 21:03:27 +0000 (14:03 -0700)]
Merge "Flexible support for various pattern searches"
Jim Bankoski [Tue, 6 Aug 2013 14:31:54 +0000 (07:31 -0700)]
sad + miscellaneous updates
Enable use_x86inc as a commandline option. Fix Bug with sse2 when
x86inc is disabled. Adds Sad asm protection to x86inc protection
Change-Id: Iee0f9dd235ea10e8ace512eb362ba9bebe8c9df6
Dmitry Kovalev [Tue, 6 Aug 2013 18:57:45 +0000 (11:57 -0700)]
Merge "Inlining vp9_get_pred_probs_switchable_interp function."
Deb Mukherjee [Mon, 22 Jul 2013 21:47:57 +0000 (14:47 -0700)]
Flexible support for various pattern searches
Adds a few pattern searches to achieve various tradeoffs
between motion estimation complexity and performance.
The search framework is unified across these searches so that a
common pattern search function is used for all. Besides it will
be easier to experiment with various patterns or combinations
thereof at different scales in the future.
The new pattern search is multi-scale and is capable of using
different patterns at different scales.
The new hex search uses 8 points at the smallest scale
and 6 points at other scales.
Two other pattern searches - big-diamond and square are
also added. Big diamond uses 4 points at the smallest scale and
8 points in diamond shape at the larger scales.
Square is very similar conceptually to the default n-step search
but is somewhat faster since it keeps only one survivor across
all scales.
Psnr/speed-up results on derf300:
hex: -1.6% psnr%, 6-8% speed-up
big-diamond: -0.96% psnr, 4-5% speedup
square: -0.93% psnr, 4-5% speedup
Change-Id: I02a7ef5193f762601e0994e2c99399a3535a43d2
Jingning Han [Tue, 6 Aug 2013 18:10:12 +0000 (11:10 -0700)]
Place holder for high-precision 32x32 fdct
Resolve compile warnings on re-define FDCT32x32_2D template.
Change-Id: Idb3a54ef8d2710ce7245b726379a0e5c875f5cad
Dmitry Kovalev [Tue, 6 Aug 2013 18:04:31 +0000 (11:04 -0700)]
Inlining vp9_get_pred_probs_switchable_interp function.
There was no benefit having this function. For example, inside
read_switchable_filter_type switchable filter context was calculated twice.
Change-Id: I79cd5bf95cbc0f6d8bf91a2e32289e01b18dcff1
Jingning Han [Tue, 6 Aug 2013 17:46:41 +0000 (10:46 -0700)]
Merge "Move fdct32x32 SSE2 implementation in separate file."
Jim Bankoski [Tue, 6 Aug 2013 17:39:19 +0000 (10:39 -0700)]
Merge "intrapred x86inc guards"
Dmitry Kovalev [Mon, 5 Aug 2013 23:27:51 +0000 (16:27 -0700)]
Motion vector code cleanup.
Converting arguments of two functions (clamp_mv_ref, lower_mv_precision)
from int_mv* to MV*. Rewriting is_inside function to make it much shorter.
Change-Id: Ie4c4cf3eccd46707c7df099ec21fb1b61c72fc7a
Dmitry Kovalev [Tue, 6 Aug 2013 17:30:37 +0000 (10:30 -0700)]
Merge "Finally removing all old block size constants."
Dmitry Kovalev [Tue, 6 Aug 2013 17:30:26 +0000 (10:30 -0700)]
Merge "Changing the order switchable filter enum constants."
Dmitry Kovalev [Tue, 6 Aug 2013 17:29:57 +0000 (10:29 -0700)]
Merge "Removing unused functions."
Deb Mukherjee [Tue, 6 Aug 2013 17:19:15 +0000 (10:19 -0700)]
Merge "Add variance based mode/skipping"
Christian Duvivier [Mon, 5 Aug 2013 22:22:13 +0000 (15:22 -0700)]
Move fdct32x32 SSE2 implementation in separate file.
This is in preparation for the SSE2 version of the high-precision
32x32 forward DCT which will share a lot of code with the existing
low precision version used for rate-distortion search.
Change-Id: I7084b6bdfb480b1fabb8493fb14e3f7fcc7888c0
Jim Bankoski [Tue, 6 Aug 2013 16:39:30 +0000 (09:39 -0700)]
intrapred x86inc guards
Change-Id: If0399d8e11f4ebe75a5c91abb8d6a52a7709065b
Jim Bankoski [Tue, 6 Aug 2013 00:27:30 +0000 (17:27 -0700)]
block error / x86inc mods
Change-Id: Icb607745634e10b9bac5019d06661ece09fcdb40
Jim Bankoski [Tue, 6 Aug 2013 00:28:52 +0000 (17:28 -0700)]
reworked config for use_x86_inc
Support enabling it or disabling it. Moved read out to configure.sh
so that its done once instead of in make and in config.
Change-Id: I73a9190cf31de9f03e8a577f478fa522f8c01c8b
James Zern [Mon, 5 Aug 2013 23:07:09 +0000 (16:07 -0700)]
Merge changes I082959ab,Ib6932640
* changes:
vp9/decoder: threaded row-based loop filter
vp9/decoder: add thread worker
Dmitry Kovalev [Mon, 5 Aug 2013 22:23:49 +0000 (15:23 -0700)]
Finally removing all old block size constants.
Change-Id: I3aae21e88b876d53ecc955260479980ffe04ad8d
Jim Bankoski [Mon, 5 Aug 2013 21:48:20 +0000 (14:48 -0700)]
fixed script problem with config_force_x86_inc
Change-Id: I226e5094d216b09dc47fa5511a66e2d314608000
Jim Bankoski [Mon, 5 Aug 2013 21:17:49 +0000 (14:17 -0700)]
Merge "Begin to restrict x86inc.asm usage"
Deb Mukherjee [Wed, 31 Jul 2013 16:33:58 +0000 (09:33 -0700)]
Add variance based mode/skipping
Adds a speed feature to skip all intra modes other than
DC_PRED if the source variance is small. This feature is
made part of speed 1 and up.
Results on derf300: psnr -0.07%, speedup about 1-2%
Also uses the source variance to fine-tune the early
termination criteria when FLAG_EARLY_TERMINATE is on.
This feature is made part of speed 2 and up.
Results on derf300: psnr -0.52%, speedup about 5-7%
Change-Id: I59e38aa836557cfa5405ae706fc64815cbfe4232
Jim Bankoski [Mon, 5 Aug 2013 21:02:02 +0000 (14:02 -0700)]
Merge "cleanups after bw bh code"
James Zern [Wed, 31 Jul 2013 23:15:10 +0000 (16:15 -0700)]
vp9/decoder: threaded row-based loop filter
Currently the only threaded option for vp9 decode. Enabled when the
decoder config thread count is > 1.
Change-Id: I082959abac9e31aa4a38ed9fd68b94680e57f4df
James Zern [Wed, 31 Jul 2013 05:46:58 +0000 (22:46 -0700)]
vp9/decoder: add thread worker
vp9/decoder/vp9_thread.[hc]
Original source:
http://git.chromium.org/webm/libwebp.git
100644 blob
b1615d0fb8d311666b2fa4561076c62d72c2e3ff src/utils/thread.c
100644 blob
13a61a4c84194c3374080cbf03d881d3cd6af40d src/utils/thread.h
Local modifications:
- s/WebP/VP9/g
- camelcase functions -> lower with _'s
Change-Id: Ib6932640ee34f8b4782c6fbd15864a59d5d4c5fe
Dmitry Kovalev [Mon, 5 Aug 2013 19:26:15 +0000 (12:26 -0700)]
Changing the order switchable filter enum constants.
This changeset allows to remove vp9_switchable_interp and
vp9_switchable_interp_map arrays and make code much clear. Actually we
still have to use these mapping but only inside read_interp_filter_type and
write_interp_filter_type functions.
Change-Id: I4026c6f8c4acefba6c81421b7bacbaa52cc45f50
Jim Bankoski [Mon, 5 Aug 2013 19:15:52 +0000 (12:15 -0700)]
cleanups after bw bh code
Cons bw/bh parms that should have been const. Additional formatting.
Change-Id: Icd36a5c9dc17dadd7284315ac0d6fef1a565ca16
Paweł Hajdan [Fri, 2 Aug 2013 03:08:42 +0000 (20:08 -0700)]
Update README
- new date
- add VP9 to the title
- update list of available targets
Change-Id: I56263336db393020bac5da8e42fbac3a276ffb1f
Jim Bankoski [Mon, 5 Aug 2013 19:07:30 +0000 (12:07 -0700)]
Begin to restrict x86inc.asm usage
Chromium does not support 32bit builds for Mac which use x86inc.asm.
Make the files which include it work if 64bit or not PIC enabled
starting with vp9_copy_sse2.asm
Consolidate these targets in vp9_rtcd_defs.sh
Change-Id: If18f0b957a611efd085a3ee7d245cf1eb91e8248
Dmitry Kovalev [Fri, 2 Aug 2013 18:45:21 +0000 (11:45 -0700)]
Replacing long block size enum values with shorter ones (2).
Change-Id: I428c4d42212b757112e3acfe5b81314cfbb5fd6b
Dmitry Kovalev [Mon, 5 Aug 2013 08:52:11 +0000 (01:52 -0700)]
Merge "Cleaning up vp9_build_inter_predictor function."
Dmitry Kovalev [Sun, 4 Aug 2013 09:52:22 +0000 (02:52 -0700)]
Merge "Replacing "txfm" with "tx" in identifiers."
Jim Bankoski [Sun, 4 Aug 2013 02:51:56 +0000 (19:51 -0700)]
reworked find_mv_ref
This is an attempt at rewriting vp9_find_mv_refs_idx. I believe that it gains
about 1-2% decode speed
Change-Id: Ia5359c94ce9bb43b32652890e605e9a385485c1b
Dmitry Kovalev [Sat, 3 Aug 2013 00:28:23 +0000 (17:28 -0700)]
Replacing "txfm" with "tx" in identifiers.
Consistent names with TX_SIZE, TX_MODE, and TX_MODE.
Change-Id: I79592218bf5a40ace89197a34a06ee7de581ed8d
Dmitry Kovalev [Sat, 3 Aug 2013 00:10:15 +0000 (17:10 -0700)]
Removing NMS_STATS defines.
Change-Id: Iabab0e59042a33456df1d449c0d0f01debc00c7c
Dmitry Kovalev [Fri, 2 Aug 2013 23:54:32 +0000 (16:54 -0700)]
Merge "Adding is_inter_block function."
Dmitry Kovalev [Fri, 2 Aug 2013 23:53:18 +0000 (16:53 -0700)]
Cleaning up vp9_build_inter_predictor function.
Change-Id: I94f6b4272b95ac101de6d10f048116ba065788b0
Dmitry Kovalev [Fri, 2 Aug 2013 23:41:09 +0000 (16:41 -0700)]
Removing unused functions.
Removed functions:
model_rd_for_sb_y,
block_error_sby,
get_sb_variance
Change-Id: Iec458df180caf6f8eac3605773841a4121dd3a8f
Dmitry Kovalev [Fri, 2 Aug 2013 23:30:06 +0000 (16:30 -0700)]
Merge "Changing function arg type from int_mv* to MV*."
Dmitry Kovalev [Fri, 2 Aug 2013 23:27:05 +0000 (16:27 -0700)]
Merge "Cleanups around allow_high_precision_mv flag."
Dmitry Kovalev [Fri, 2 Aug 2013 23:25:33 +0000 (16:25 -0700)]
Adding is_inter_block function.
Using it instead of long unclear verbose check
"mbmi->ref_frame[0] != INTRA_FRAME".
Change-Id: I9c7b4b3797942fa962bf3ba7460fff3084beabe9
Dmitry Kovalev [Fri, 2 Aug 2013 23:22:50 +0000 (16:22 -0700)]
Merge "Cleaning up set_contexts_on_border function."
Yunqing Wang [Fri, 2 Aug 2013 22:55:09 +0000 (15:55 -0700)]
Merge "Add more checking to using_small_partition_info"
Dmitry Kovalev [Fri, 2 Aug 2013 22:52:26 +0000 (15:52 -0700)]
Cleaning up set_contexts_on_border function.
Change-Id: I8f21c18b29f54b277fb1c167f278f109d9f3b996
Dmitry Kovalev [Fri, 2 Aug 2013 22:26:32 +0000 (15:26 -0700)]
Changing function arg type from int_mv* to MV*.
Change-Id: Ic878d31df2ce783a2c9a8c4bc9ed301ec8ffe25e
Dmitry Kovalev [Fri, 2 Aug 2013 18:53:49 +0000 (11:53 -0700)]
Moving struct loop_filter_info from *.h to *.c file.
Change-Id: I3fe90eb40088a5b07bdc7d66d93ffe6ef99943d5
Adrian Grange [Fri, 2 Aug 2013 18:37:47 +0000 (11:37 -0700)]
Merge "Fixed typos and added a few explanatory comments"
Adrian Grange [Fri, 2 Aug 2013 18:36:46 +0000 (11:36 -0700)]
Merge "Changed name of rd_pick_intra4x4mby_modes"
Johann [Fri, 2 Aug 2013 18:28:31 +0000 (11:28 -0700)]
Merge "vp9: neon: convolve: replace some insns with simpler equivalents"
Johann [Fri, 2 Aug 2013 18:28:25 +0000 (11:28 -0700)]
Merge "vp9: neon: convolve: simplify branching to C fallbacks"
Johann [Fri, 2 Aug 2013 18:28:04 +0000 (11:28 -0700)]
Merge "vp9: neon: optimise loads in horiz convolve functions"
Johann [Fri, 2 Aug 2013 18:27:52 +0000 (11:27 -0700)]
Merge "vp9: neon: add vp9_mb_lpf_* functions"
Dmitry Kovalev [Fri, 2 Aug 2013 18:21:16 +0000 (11:21 -0700)]
Cleanups around allow_high_precision_mv flag.
Change-Id: Ic07f5f8ffeaedd5b7513b464871f83afc82dcd5c
Dmitry Kovalev [Fri, 2 Aug 2013 17:48:27 +0000 (10:48 -0700)]
Replacing long block size enum values with shorter ones.
Change-Id: I0e9329490828684a4fd46f540d89114cc68e8407
Yunqing Wang [Fri, 2 Aug 2013 16:58:46 +0000 (09:58 -0700)]
Merge "Comment out 2 unused speed features"
Mans Rullgard [Tue, 30 Jul 2013 17:08:17 +0000 (18:08 +0100)]
vp9: neon: convolve: replace some insns with simpler equivalents
Change-Id: I5d6906772e6e6adf68d7f0fd5b8b5207a64a3a37
Mans Rullgard [Tue, 30 Jul 2013 16:33:52 +0000 (17:33 +0100)]
vp9: neon: convolve: simplify branching to C fallbacks
Change-Id: Ic7cacd02d6dc9243ad8fc85082c5618a9d1e66dc
Mans Rullgard [Tue, 30 Jul 2013 17:11:06 +0000 (18:11 +0100)]
vp9: neon: optimise loads in horiz convolve functions
Loading to single lanes in multiple registers is expensive since
it requires a read and write of each register which saturates
the register file access. Loading to single registers followed
by a separate transpose reduces this pressure.
Change-Id: I4cc35887ddbca80e5e635b50d2b1d158de9668ee
Mans Rullgard [Thu, 4 Jul 2013 18:58:41 +0000 (14:58 -0400)]
vp9: neon: add vp9_mb_lpf_* functions
Change-Id: I13e0880df234f15abc4cc7c57fe84488d5d46a75
Dmitry Kovalev [Fri, 2 Aug 2013 06:17:17 +0000 (23:17 -0700)]
Merge "Cleanup: replacing xd->seg with seg, and xd->lf with lf."
Dmitry Kovalev [Fri, 2 Aug 2013 06:16:56 +0000 (23:16 -0700)]
Merge "Cleanup: reusing clamp_mv function."
Jingning Han [Thu, 1 Aug 2013 22:41:35 +0000 (15:41 -0700)]
Merge "Remove unused vp9_short_idct10_32x32_add"
Dmitry Kovalev [Thu, 1 Aug 2013 21:53:14 +0000 (14:53 -0700)]
Cleanup: replacing xd->seg with seg, and xd->lf with lf.
Change-Id: I73b59d7699a8e7e7acd3bf8041cb6c98ce9ba4bf
Dmitry Kovalev [Thu, 1 Aug 2013 22:07:12 +0000 (15:07 -0700)]
Merge "Cleanup: removing unused function arguments."
Dmitry Kovalev [Thu, 1 Aug 2013 22:06:34 +0000 (15:06 -0700)]
Cleanup: reusing clamp_mv function.
Change-Id: I8715f08a3554bdb557c5f935f1dfbd671f18e766
Dmitry Kovalev [Thu, 1 Aug 2013 21:50:14 +0000 (14:50 -0700)]
Merge "Nice looking motion vector clamping functions."
Deb Mukherjee [Thu, 1 Aug 2013 21:18:43 +0000 (14:18 -0700)]
Merge "Adds a source variance computation function"
Dmitry Kovalev [Thu, 1 Aug 2013 21:15:53 +0000 (14:15 -0700)]
Merge "vp9_get_pred_context_intra_inter cleanup."
Dmitry Kovalev [Wed, 31 Jul 2013 23:59:15 +0000 (16:59 -0700)]
Cleanup: removing unused function arguments.
Change-Id: I27471768980fc631916069f24bc7c482a5c9ca17
Dmitry Kovalev [Wed, 31 Jul 2013 23:11:03 +0000 (16:11 -0700)]
Nice looking motion vector clamping functions.
Removing assign_and_clamp_mv function, making implementation of clamp_mv
and clamp_mv2 more clear and consistent.
Change-Id: Iecd08e1c1bf0379f8314ebe01811f8253f4ade58
Deb Mukherjee [Thu, 1 Aug 2013 19:56:12 +0000 (12:56 -0700)]
Adds a source variance computation function
Adds a function to compute source variance for various
sb_types to be used for pruning mode and partition searches.
[The existing activity measure function is currently specialized
for only 16x16 MBs and needs to be updated].
Change-Id: I22a41e6f1430184201487326fdbebb9b47e6fc24
Jingning Han [Thu, 1 Aug 2013 19:45:16 +0000 (12:45 -0700)]
Remove unused vp9_short_idct10_32x32_add
The inverse 32x32 transform detects all zero entries and skips the
computations accordingly per 8 rows in the first 1-D operation. The
function vp9_short_idct10_32x32_add performs differently and is not
used anywhere, hence removed.
Change-Id: Ic4fad422debbde7b6b6ffed47c69fbd4268a906c
Jingning Han [Thu, 1 Aug 2013 18:53:39 +0000 (11:53 -0700)]
Merge "Optimize 32x32 2D inverse DCT for speed-up"
Yunqing Wang [Thu, 1 Aug 2013 18:37:41 +0000 (11:37 -0700)]
Add more checking to using_small_partition_info
If the partition is out of partition size range, we don't
need to process small partition information.
Change-Id: Ice9bfbbdebe1f2ef79271a3aee17de0ed4608376
Yunqing Wang [Thu, 1 Aug 2013 18:03:34 +0000 (11:03 -0700)]
Comment out 2 unused speed features
use_min_partition_size and use_max_partition_size are not used
currently, and could be added back if needed later.
Change-Id: Ib22a9c06b064567a7c1d6d5445567ed77e0d3acc
Dmitry Kovalev [Thu, 1 Aug 2013 17:19:51 +0000 (10:19 -0700)]
Merge "Adding missing const to vp9_extra_bits array."
Adrian Grange [Thu, 1 Aug 2013 16:04:15 +0000 (09:04 -0700)]
Fixed typos and added a few explanatory comments
Change-Id: Ib4e4b41094b54874ee34343dd77c0c131ceed9d2
Adrian Grange [Wed, 31 Jul 2013 19:58:19 +0000 (12:58 -0700)]
Changed name of rd_pick_intra4x4mby_modes
The function name rd_pick_intra4x4mby_modes is confusing, so
I changed it to rd_pick_intra_sub_8x8_y_modes to better
reflect what the function does. Also added const qualifiers
to some of the input parameters and removed camel-case.
Change-Id: I23d53d4c7af5d79ed8a471acd59a09bbb47add39
Dmitry Kovalev [Thu, 1 Aug 2013 01:51:18 +0000 (18:51 -0700)]
Adding missing const to vp9_extra_bits array.
Change-Id: Icd128ab58719e0b9066bdfa66a5d0d427a84d6df
Dmitry Kovalev [Thu, 1 Aug 2013 01:33:04 +0000 (18:33 -0700)]
vp9_get_pred_context_intra_inter cleanup.
Change-Id: I8beeee4c020425175f7d5ec83be86afa7b95da1a
Jingning Han [Wed, 31 Jul 2013 23:50:34 +0000 (16:50 -0700)]
Optimize 32x32 2D inverse DCT for speed-up
This commit exploits the sparsity of quantized coefficient matrix.
It detects each 32x8 array and skip the corresponding inverse
transformation if all entries are zero.
For ped1080p at 8000 kbps, this on average reduces the runtime of
32x32 inverse 2D-DCT SSE2 function from 6256 cycles -> 5200
cycles. It makes the overall encoding process about 2% faster at
speed 0. The speed-up is more pronounceable for the decoding process.
Change-Id: If20056c3566bd117642a76f8884c83e8bc8efbcf
Jingning Han [Thu, 1 Aug 2013 00:02:06 +0000 (17:02 -0700)]
Remove unnecessary arguments in rd_pick_ref_frame
This commit removes redundant arguments passing in the function of
rd_pick_reference_frame. This resolves the clang warnings about
potential use of uninitialized values.
Change-Id: Ic68f949a9f8fcd0a583786b0c75321104ea44739
Dmitry Kovalev [Wed, 31 Jul 2013 22:03:36 +0000 (15:03 -0700)]
vp9_decodemv.c cleanup.
Inlining VP9_NMV_UPDATE_PROB constant, consistent local variable names.
Change-Id: I01692501982568fa535882d6b320e3c692f88abb
Dmitry Kovalev [Wed, 31 Jul 2013 18:58:00 +0000 (11:58 -0700)]
Removing get_mi_{row, col} functions.
Passing mi_row and mi_col parameters to functions explicitly. Removing
unused xd argument from scale_mv function.
Change-Id: Icb4c495ec72d26fb066c14470d3ae0b741fbf18a
Dmitry Kovalev [Wed, 31 Jul 2013 19:03:04 +0000 (12:03 -0700)]
Merge "Removing unused "ishp" arguments."
Dmitry Kovalev [Wed, 31 Jul 2013 19:02:35 +0000 (12:02 -0700)]
Merge "Consistent update for inter_mode probabilities."