Commit Graph

44 Commits

Author SHA1 Message Date
James Yu
eed005b076 VP8 encoder for ARMv8 by using NEON intrinsics 6
Add shortfdct_neon.c
- vp8_short_fdct4x4_neon
- vp8_short_fdct8x4_neon

Change-Id: I90152c803b484f5fab839473d632c50af0524e68
Signed-off-by: James Yu <james.yu@linaro.org>
2014-08-20 09:25:29 -07:00
James Yu
6d6fdd9c3d VP8 encoder for ARMv8 by using NEON intrinsics 3
Add subtract_neon.c
- vp8_subtract_b_neon
- vp8_subtract_mby_neon
- vp8_subtract_mbuv_neon

Change-Id: If9a17a093478552e3e3276eeaa3f098b9021d08c
Signed-off-by: James Yu <james.yu@linaro.org>
2014-08-20 09:20:55 -07:00
Scott LaVarnway
8013aaa10b VP8 encoder for ARMv8 by using NEON intrinsics 2
Add vp8_shortwalsh4x4_neon.c
- vp8_short_walsh4x4_neon

Change-Id: Ica5f584be608c9e636f62db14f563757e94be09b
Signed-off-by: James Yu <james.yu@linaro.org>
2014-08-20 09:19:23 -07:00
Johann
ce23931a3f Only build neon assembly for armv7 targets
Allow selectively building just the intrinsics for armv8

Change-Id: I2f29b2e4508b8b8e5649c2906b3159ad1d4ec477
2014-05-12 08:52:02 -07:00
Christian Duvivier
b52db6b7e8 ARM NEON version of denoiser.
Change-Id: I951abd4ad0078f78949f3cb79453ac334fb82a7e
2014-01-02 10:51:05 -08:00
Johann
7c7801e34d Each make file is responsible for its own directory
Change-Id: Ia8a59ce79045af209e49c68810bae44616422db5
2012-11-15 16:41:03 -08:00
Johann
e50f96a4a3 Move SAD and variance functions to common
The MFQE function of the postprocessor depends on these

Change-Id: I256a37c6de079fe92ce744b1f11e16526d06b50a
2012-03-05 16:50:33 -08:00
John Koleszar
8aae246089 RTCD: finalize removal of old RTCD system
This is the final commit in the series converting to the new RTCD
system. It removes the encoder csystemdependent files and the remaining
global function pointers that didn't conform to the old RTCD system.

Change-Id: I9649706f1bb89f0cbf431ab0e3e7552d37be4d8e
2012-01-30 12:10:48 -08:00
John Koleszar
be8af188d0 RTCD: add block subtraction functions
This commit continues the process of converting to the new RTCD
system.

Change-Id: Id8a287fdd4bd050ea4452e1582ad85520f3081be
2012-01-30 12:10:47 -08:00
John Koleszar
61311e6103 RTCD: add quantizer functions
This commit continues the process of converting to the new RTCD
system.

Change-Id: Iba9df4c03a508e51c37201c621be43523fae87d9
2012-01-30 12:10:46 -08:00
John Koleszar
510e0ab467 RTCD: add FDCT functions
This commit continues the process of converting to the new RTCD
system.

Change-Id: I3f9c07db65eb206f6363d21bdb80e871570da767
2012-01-30 12:10:42 -08:00
John Koleszar
83a91e789c RTCD: add variance functions
This commit continues the process of converting to the new RTCD
system.

Change-Id: Ie5c1aa480637e98dc3918fb562ff45c37a66c538
2012-01-30 12:08:30 -08:00
Fritz Koenig
892102842a Disconnect ARM tgt_isa from dsp extensions
A processor with ARMv7 instructions does not
necessarily have NEON dsp extensions.  This CL
has the added side effect of allowing the ability
to enable/disable the dsp extensions cleanly.

Change-Id: Ie1e879b8fe131885bc3d4138a0acc9ffe73a36df
2012-01-20 10:38:15 -08:00
Fritz Koenig
bd0c3409a8 Move neon only arm functions under arm/neon.
These files don't contain generic arm code, so should
only be compiled by neon.

Change-Id: Ie712823aa04d4235e7cfe7a3b725e73ee4c3e564
2011-09-20 10:51:06 -07:00
Johann
6829e62718 Merge "NEON FDCT updated to match current C code" 2011-09-20 09:51:05 -07:00
Tero Rintaluoma
0c2529a812 NEON FDCT updated to match current C code
- Removed fast_fdct4x4_neon and fast_fdct8x4_neon
- Uses now short_fdct4x4 and short_fdct8x4
- Gives ~1-2% speed-up on Cortex-A8/A9

Change-Id: Ib62f2cb2080ae719f8fa1d518a3a5e71278a41ec
2011-09-20 10:20:55 +03:00
Tero Rintaluoma
4c3ad66b7f Updated ARMv6 forward transforms to match C
- Updated walsh transform to match C
  (based on Change Id24f3392)
- Changed fast_fdct4x4 and 8x4 to short_fdct4x4 and 8x4
  correspondingly

Change-Id: I704e862f40e315b0a79997633c7bd9c347166a8e
2011-09-19 10:26:59 +03:00
Johann
3e8c6d3d35 include the arm header files in make dist
Change-Id: Ibcf5b4b14153f65ce1b53c3bfba87ad2feb17bbd
2011-08-01 17:20:21 -04:00
Attila Nagy
6f23f24afe configuration, support disabling any subset of ARM arch
Useful for leaving out any version specific asm files.

Change-Id: I233514410eb9d7ca88d2d2c839673122c507fa99
2011-06-21 10:39:01 +03:00
Tero Rintaluoma
61f0c090df neon fast quantize block pair
vp8_fast_quantize_b_pair_neon function added to quantize
two adjacent blocks at the same time to improve performance.
 - Additional 3-6% speedup compared to neon optimized fast
   quantizer (Tanya VGA@30fps, 1Mbps stream, cpu-used=-5..-16)

Change-Id: I3fcbf141e5d05e9118c38ca37310458afbabaa4e
2011-06-01 10:48:05 +03:00
Tero Rintaluoma
33fa7c4ebe neon fast quantizer updated
vp8_fast_quantize_b_neon function updated and further optimized.
 - match current C implementation of fast quantizer
 - updated to use asm_enc_offsets for structure members
 - updated ads2gas scripts to handle alignment issues

Change-Id: I5cbad9c460ad8ddb35d2970a8684cc620711c56d
2011-05-06 08:59:52 +03:00
Johann
ba11e24d47 Merge "Wrapper function removed from vp8_subtract_b_neon function call" 2011-04-01 05:47:21 -07:00
Tero Rintaluoma
cec76a36d6 Wrapper function removed from vp8_subtract_b_neon function call
Address calculations moved from encodemb_arm.c file to neon
optimized assembly function to save cycles in function calls.
 - vp8_subtract_b_neon_func replaced with vp8_subtract_b_neon
   that contains all needed address calculations
 - unnecessary file encodemb_arm.c removed
 - consistent with ARMv6 optimized version

Change-Id: I6cbc1a2670b56c2077f59995fcf8f70786b4990b
2011-04-01 10:06:44 +03:00
Johann
9d138379a2 Merge "ARMv6 optimized subtract functions" 2011-03-31 08:40:10 -07:00
Johann
0e43668546 Merge "Half pixel variance further optimized for ARMv6" 2011-03-29 12:14:54 -07:00
Tero Rintaluoma
6fdc9aa79f ARMv6 optimized subtract functions
Adds following ARMv6 optimized functions to encoder:
  - vp8_subtract_b_armv6
  - vp8_subtract_mby_armv6
  - vp8_subtract_mbuv_armv6

Gives 1-5% speed-up depending on input sequence and encoding
parameters. Functions have one stall cycle inside the loop body
on Cortex pipeline.

Change-Id: I19cca5408b9861b96f378e818eefeb3855238639
2011-03-29 16:52:00 +03:00
Johann
4be062bbc3 add asm_enc_offsets.c for all targets
now that we need asm_enc_offsets.c for x86 and arm and it is
harmless to build it for other targets, add it unconditionally

Change-Id: I320c5220afd94fee2b98bda9ff4e5e34c67062f3
2011-03-28 10:43:47 -04:00
Tero Rintaluoma
f5e433464b Half pixel variance further optimized for ARMv6
Half pixel interpolations optimized in variance calculations. Separate
function calls to vp8_filter_block2d_bil_x_pass_armv6 are avoided.On
average, performance improvement is 6-7% for VGA@30fps sequences.

Change-Id: Idb5f118a9d51548e824719d2cfe5be0fa6996628
2011-03-28 09:51:51 +03:00
Tero Rintaluoma
a61785b6a1 ARMv6 optimized fdct4x4
Optimized fdct4x4 (8x4) for ARMv6 instruction set.
  - No interlocks in Cortex-A8 pipeline
  - One interlock cycle in ARM11 pipeline
  - About 2.16 times faster than current C-code compiled with -O3

Change-Id: I60484ecd144365da45bb68a960d30196b59952b8
2011-03-21 13:33:45 +02:00
Attila Nagy
71bcd9f1af Add vp8_variance8x8_armv6 and vp8_sub_pixel_variance8x8_armv6 functions
Change-Id: I08edaffc62514907fa5e90e1689269e467c857f5
2011-03-15 15:50:44 +02:00
Johann
d0ec28b3d3 Merge "Add vp8_mse16x16_armv6 function" 2011-03-14 12:47:42 -07:00
Attila Nagy
e54dcfe88d Add vp8_mse16x16_armv6 function
Change-Id: I77e9f2f521a71089228f96e2db72524189364ffb
2011-03-14 14:38:31 +02:00
Tero Rintaluoma
7ab08e1fee ARMv6 optimized quantization
Adds new ARMv6 optimized function vp8_fast_quantize_b_armv6
to the encoder.

Change-Id: I40277ec8f82e8a6cbc453cf295a0cc9b2504b21e
2011-03-11 10:48:42 +02:00
Tero Rintaluoma
1ef86980b9 ARMv6 optimized sad16x16
Adds a new ARMv6 optimized function vp8_sad16x16_armv6 to encoder.

Change-Id: Ibbd7edb8b25cb7a5b522d391b1e9a690fe150e57
2011-02-11 11:14:07 +02:00
Tero Rintaluoma
cb14764fab Adds armv6 optimized variance calculation
Adds vp8_sub_pixel_variance16x16_armv6 function to encoder. Integrates
ARMv6 optimized bilinear interpolations from vp8/common/arm/armv6
and adds new assembly file for variance16x16 calculation.
 - vp8_filter_block2d_bil_first_pass_armv6   (integrated)
 - vp8_filter_block2d_bil_second_pass_armv6  (integrated)
 - vp8_variance16x16_armv6 (new)
 - bilinearfilter_arm.h (new)
Change-Id: I18a8331ce7d031ceedd6cd415ecacb0c8f3392db
2011-02-09 10:23:43 -05:00
Johann
40dcae9c2e clarify *_offsets.asm differences
it's difficult to mux the *_offsets.c files because of header conflicts.
make three instead, name them consistently and partititon the contents
to allow building them as required.

Change-Id: I8f9768c09279f934f44b6c5b0ec363f7943bb796
2011-02-08 16:35:43 -05:00
Attila Nagy
0def48b60f Adds vpx_vp8_enc_asm_offsets.c.o to OBJS-yes list
Change-Id: Ibd6e3bc82471839904b1086b499efc55f7c5cbaf
2011-01-26 17:06:09 +02:00
Johann
2168a94495 move new neon subpixel function
previously wasn't guarded with ifdef ARMV7, causing a link error with
ARMV6

Change-Id: I0526858be0b5f49b2bf11e9090180b2a6c48926d
2011-01-25 15:48:37 -05:00
John Koleszar
209d82ad72 Add half-pixel variance RTCD functions
NEON has optimized 16x16 half-pixel variance functions, but they
were not part of the RTCD framework. Add these functions to RTCD,
so that other platforms can make use of this optimization in the
future and special-case ARM code can be removed.

A number of functions were taking two variance functions as
parameters. These functions were changed to take a single
parameter, a pointer to a struct containing all the variance
functions for that block size. This provides additional flexibility
for calling additional variance functions (the half-pixel special
case, for example) and by initializing the table for all block sizes,
we don't have to construct this function pointer table for each
macroblock.

Change-Id: I78289ff36b2715f9a7aa04d5f6fbe3d23acdc29c
2010-10-26 20:00:56 -07:00
Timothy B. Terriberry
b71962fdc9 Add runtime CPU detection support for ARM.
The primary goal is to allow a binary to be built which supports
 NEON, but can fall back to non-NEON routines, since some Android
 devices do not have NEON, even if they are otherwise ARMv7 (e.g.,
 Tegra).
The configure-generated flags HAVE_ARMV7, etc., are used to decide
 which versions of each function to build, and when
 CONFIG_RUNTIME_CPU_DETECT is enabled, the correct version is chosen
 at run time.
In order for this to work, the CFLAGS must be set to something
 appropriate (e.g., without -mfpu=neon for ARMv7, and with
 appropriate -march and -mcpu for even earlier configurations), or
 the native C code will not be able to run.
The ASFLAGS must remain set for the most advanced instruction set
 required at build time, since the ARM assembler will refuse to emit
 them otherwise.
I have not attempted to make any changes to configure to do this
 automatically.
Doing so will probably require the addition of new configure options.

Many of the hooks for RTCD on ARM were already there, but a lot of
 the code had bit-rotted, and a good deal of the ARM-specific code
 is not integrated into the RTCD structs at all.
I did not try to resolve the latter, merely to add the minimal amount
 of protection around them to allow RTCD to work.
Those functions that were called based on an ifdef at the calling
 site were expanded to check the RTCD flags at that site, but they
 should be added to an RTCD struct somewhere in the future.
The functions invoked with global function pointers still are, but
 these should be moved into an RTCD struct for thread safety (I
 believe every platform currently supported has atomic pointer
 stores, but this is not guaranteed).

The encoder's boolhuff functions did not even have _c and armv7
 suffixes, and the correct version was resolved at link time.
The token packing functions did have appropriate suffixes, but the
 version was selected with a define, with no associated RTCD struct.
However, for both of these, the only armv7 instruction they actually
 used was rbit, and this was completely superfluous, so I reworked
 them to avoid it.
The only non-ARMv4 instruction remaining in them is clz, which is
 ARMv5 (not even ARMv5TE is required).
Considering that there are no ARM-specific configs which are not at
 least ARMv5TE, I did not try to detect these at runtime, and simply
 enable them for ARMv5 and above.

Finally, the NEON register saving code was completely non-reentrant,
 since it saved the registers to a global, static variable.
I moved the storage for this onto the stack.
A single binary built with this code was tested on an ARM11 (ARMv6)
 and a Cortex A8 (ARMv7 w/NEON), for both the encoder and decoder,
 and produced identical output, while using the correct accelerated
 functions on each.
I did not test on any earlier processors.

Change-Id: I45cbd63a614f4554c3b325c45d46c0806f009eaa
2010-10-25 09:23:29 -04:00
John Koleszar
c2140b8af1 Use WebM in copyright notice for consistency
Changes 'The VP8 project' to 'The WebM project', for consistency
with other webmproject.org repositories.

Fixes issue #97.

Change-Id: I37c13ed5fbdb9d334ceef71c6350e9febed9bbba
2010-09-09 10:01:21 -04:00
John Koleszar
94c52e4da8 cosmetics: trim trailing whitespace
When the license headers were updated, they accidentally contained
trailing whitespace, so unfortunately we have to touch all the files
again.

Change-Id: I236c05fade06589e417179c0444cb39b09e4200d
2010-06-18 13:06:11 -04:00
John Koleszar
09202d8071 LICENSE: update with latest text
Change-Id: Ieebea089095d9073b3a94932791099f614ce120c
2010-06-04 16:19:40 -04:00
John Koleszar
0ea50ce9cb Initial WebM release 2010-05-18 11:58:33 -04:00