This patch implements basics for Xfermode SSE optimization. Based on
these basics, SSE2 implementation of multiply_modeproc is provided. SSE2
implementation for other modes will come in future. With this patch
performance of Xfermode_Multiply will improve about 45%. Here are the
data on desktop i7-3770.
before:
Xfermode_Multiply 8888: cmsecs = 33.30 565: cmsecs = 45.65
after:
Xfermode_Multiply 8888: cmsecs = 17.18 565: cmsecs = 24.87
BUG=
R=mtklein@google.com
Author: qiankun.miao@intel.com
Review URL: https://codereview.chromium.org/202903004
git-svn-id: http://skia.googlecode.com/svn/trunk@14006 2bbb7eff-a529-9590-31e7-b0007b416f81
Reason for revert:
GYP's failing on most (all?) bots.
Original issue's description:
> ARM Skia NEON patches - 35 - First AArch64 support
>
> Aarch64 support
>
> This change contains the necessary modifications to have Skia build and
> run properly on an ARMv8 processor in aarch64 execution state.
>
> Here's a list of the changes:
>
> - add an arm64 target to the build system + SK_CPU_ARM64 flag
>
> - MatrixTest was failing when built in Release mode. Fused MAC
> instructions were generated which made some intermediate results
> more accurate. As the test relies on result comparison, the more
> precise results when compared to others led to a gap bigger than
> what was tolerated. As I don't know if some actual skia code relies
> on results being comparable, I've disabled fused MAC instruction
> with -ffp-contract=off for arm64.
>
> - Modify include/core/SkOnce.h to have barriers work.
>
> - SK_CPU_ARM64 implies SK_ARM_NEON_MODE_ALWAYS.
>
> - use existing Xfermode optimisations with modifications that can be
> removed in the future when toolchains are ready. Also save a few
> instructions is two Xfermodes (will apply to ARM too).
>
> - use existing SkBoxBlur and SkMorphology optimisations.
>
> - use existing SkBlitMask optimisations
>
> - use existing BitmapProcState and Convolution optimisations.
>
> Future changes will include:
>
> - Blitters (only partialy merged upstream)
>
> - SkUtils (there's little value in sending asm optimisations without
> having them benchmarked on real hardware).
>
> Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
>
> BUG=skia:
>
> Committed: http://code.google.com/p/skia/source/detail?r=13980R=djsollen@google.com, reed@google.com, halcanary@google.com, kevin.petit@arm.comTBR=djsollen@google.com, halcanary@google.com, kevin.petit@arm.com, reed@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Author: mtklein@google.com
Review URL: https://codereview.chromium.org/216113005
git-svn-id: http://skia.googlecode.com/svn/trunk@13983 2bbb7eff-a529-9590-31e7-b0007b416f81
Aarch64 support
This change contains the necessary modifications to have Skia build and
run properly on an ARMv8 processor in aarch64 execution state.
Here's a list of the changes:
- add an arm64 target to the build system + SK_CPU_ARM64 flag
- MatrixTest was failing when built in Release mode. Fused MAC
instructions were generated which made some intermediate results
more accurate. As the test relies on result comparison, the more
precise results when compared to others led to a gap bigger than
what was tolerated. As I don't know if some actual skia code relies
on results being comparable, I've disabled fused MAC instruction
with -ffp-contract=off for arm64.
- Modify include/core/SkOnce.h to have barriers work.
- SK_CPU_ARM64 implies SK_ARM_NEON_MODE_ALWAYS.
- use existing Xfermode optimisations with modifications that can be
removed in the future when toolchains are ready. Also save a few
instructions is two Xfermodes (will apply to ARM too).
- use existing SkBoxBlur and SkMorphology optimisations.
- use existing SkBlitMask optimisations
- use existing BitmapProcState and Convolution optimisations.
Future changes will include:
- Blitters (only partialy merged upstream)
- SkUtils (there's little value in sending asm optimisations without
having them benchmarked on real hardware).
Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
BUG=skia:
R=djsollen@google.com, reed@google.com, mtklein@google.com, halcanary@google.com
Author: kevin.petit@arm.com
Review URL: https://codereview.chromium.org/143423004
git-svn-id: http://skia.googlecode.com/svn/trunk@13980 2bbb7eff-a529-9590-31e7-b0007b416f81
Generate SkUserConfig.
Include arm64 as another build flavor.
Add tests.
gyp/common_conditions.gypi:
Add conditions for Android framework. These will get written into the generated SkUserConfig.
include/core/SkUserConfig.h:
Generated version that will ultimately be checked into Android (but not here).
platform_tools/android/bin/gyp_to_android.py:
Generate SkUserConfig.
Add arm64 (note that arm64 is not currently respected by our gyp files, so it results in use _none.cpp for the various opts).
Reset the common defines, which are now passed to the generated SkUserConfig.
platform_tools/android/gyp_gen/generate_user_config.py:
New script to generate SkUserConfig.h.
platform_tools/android/gyp_gen/gypd_parser.py:
Fix a lint error (unused import).
platform_tools/android/gyp_gen/makefile_writer.py:
Append any remaining DEFINES to LOCAL_CFLAGS (previously this was done during parsing).
Add a warning for arm64 (corresponds to downstream Android.mk).
platform_tools/android/gyp_gen/vars_dict_lib.py:
Add OrderedSet.reset().
Add DEFINES to VarsDict.
platform_tools/android/tests/expectations/:
Add and update expectations files.
platform_tools/android/tests/generate_user_config_tests.py:
New test for generate_user_config.py
platform_tools/android/tests/inputs/SkUserConfig.h:
Input to the new test, so we don't have to update the expectations each time the real SkUserConfig.h changes.
platform_tools/android/tests/makefile_writer_tests.py:
Add a way to rebaseline test_write_local_vars, which has changed.
Refactor EXPECTATIONS_DIR and compare_files into a separate file for sharing with generate_user_config_tests.py.
platform_tools/android/tests/utils.py:
Common code for tests.
platform_tools/android/tests/var_dict_tests.py:
Use a for loop to test the new key (DEFINES) and future proof this test to test any new keys in the future.
BUG=skia:1975
R=djsollen@google.com, halcanary@google.com
Author: scroggo@google.com
Review URL: https://codereview.chromium.org/198063002
git-svn-id: http://skia.googlecode.com/svn/trunk@13975 2bbb7eff-a529-9590-31e7-b0007b416f81
Valgrind's whining about m.setConcat(m, I). That special cases into m =
m, which compiles into memcpy(&m, &m, sizeof(m)), and then Valgrind
decides that's not kosher because &m and &m overlap. (memmove would
have been fine, but this is compiler generated code; not much we can do.)
preConcat gets us the same thing with an early can-we-noop check for I,
sidestepping the whole problem.
BUG=skia:
R=bsalomon@google.com, mtklein@google.com
Author: mtklein@chromium.org
Review URL: https://codereview.chromium.org/214823003
git-svn-id: http://skia.googlecode.com/svn/trunk@13972 2bbb7eff-a529-9590-31e7-b0007b416f81
Use path rendering to render the text from outlines if supported by the
GPU. Implement this in GrStencilAndCoverTextContext by copying large
chunks of code from GrBitmapTextContext (drawText) and
GrDistanceFieldTextContext (drawPosText).
The drawing is implemented with "instanced" path drawing
functions.
Moves the creation of the "main" text context from SkGpuDevice to the
GrContext::createTextContext. This is done because the decision of which
text renderer is optimal can be made only with the internal
implementation-specific information of the context.
R=jvanverth@google.com, bsalomon@google.com
Author: kkinnunen@nvidia.com
Review URL: https://codereview.chromium.org/196133014
git-svn-id: http://skia.googlecode.com/svn/trunk@13962 2bbb7eff-a529-9590-31e7-b0007b416f81
