brianosman [Mon, 20 Jun 2016 01:05:12 +0000 (18:05 -0700)]
Revert of Integers can now be passed as uniforms (patchset #1 id:1 of https://codereview.chromium.org/
2076143003/ )
Reason for revert:
Compile failures on Vulkan platforms (eg https://build.chromium.org/p/client.skia.compile/builders/Build-Ubuntu-GCC-Arm7-Debug-Android_Vulkan)
Original issue's description:
> Integers can now be passed as uniforms; needed for passing color count to fragment shader
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2076143003
>
> Committed: https://skia.googlesource.com/skia/+/
cc3a22b369e1a60fa2acf2987f2934baf7c4b198
TBR=tomhudson@google.com,bsalomon@google.com,fmenozzi@google.com
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=skia:
Review-Url: https://codereview.chromium.org/
2080133002
mtklein [Sun, 19 Jun 2016 12:28:33 +0000 (05:28 -0700)]
Simplify mask/clip intersection, making sure to explicitly check for an empty mask.
Previously we were only asserting the mask wasn't empty, which isn't necessarily true when we're given pathological float coordinates like +Inf or NaN.
A local run of nanobench --match text_ was not able to show this is faster or slower.
This patch fixed this first Chrome bug on my desktop, and the second is probably a dupe.
BUG=chromium:619378,chromium:613912
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2073873002
Review-Url: https://codereview.chromium.org/
2073873002
fmenozzi [Sat, 18 Jun 2016 16:31:58 +0000 (09:31 -0700)]
Integers can now be passed as uniforms; needed for passing color count to fragment shader
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2076143003
Review-Url: https://codereview.chromium.org/
2076143003
egdaniel [Fri, 17 Jun 2016 22:11:45 +0000 (15:11 -0700)]
Revert of More removal of SkColorProfileType... (patchset #2 id:20001 of https://codereview.chromium.org/
2071393002/ )
Reason for revert:
This was still used in chrome
Original issue's description:
> More removal of SkColorProfileType...
>
> Scrub GMs. Remove the gDefaultProfile thing (it's unused now), along with
> the command line flag that was setting it in DM and nanobench.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2071393002
>
> Committed: https://skia.googlesource.com/skia/+/
944876f2745a62a839e49275daf93a0329372e67
TBR=msarett@google.com,reed@google.com,bsalomon@google.com,brianosman@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2074103004
hstern [Fri, 17 Jun 2016 21:06:13 +0000 (14:06 -0700)]
Fix mixup where SamplePathOverstroke wasn't added
- My sampleapp for the stroking bugs was inflight or something when
this change got made, so it didn't get added to this file.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2076283002
Review-Url: https://codereview.chromium.org/
2076283002
mtklein [Fri, 17 Jun 2016 20:47:53 +0000 (13:47 -0700)]
Quick bandaid for chromium:611002.
We're somehow receiving non-premultiplied src inputs like 0x00ffffff to this
SrcOver blend. That's a bug I intend to follow up on. But for a quick
compatibility fix, go back to treating values like 0x00ffffff as transparent,
like we used to before crrev.com/
1820313002.
This will not affect the correctness of code paths using properly premultiplied colors.
This should not change performance in any meaningful way.
The SIMD code paths (handling strides of 16 pixels at a time) happen to treat
invalid colors like 0x00fffff as transparent already.
BUG=chromium:611002
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2075173002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/
2075173002
brianosman [Fri, 17 Jun 2016 20:43:27 +0000 (13:43 -0700)]
More removal of SkColorProfileType...
Scrub GMs. Remove the gDefaultProfile thing (it's unused now), along with
the command line flag that was setting it in DM and nanobench.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2071393002
Review-Url: https://codereview.chromium.org/
2071393002
cabanier [Fri, 17 Jun 2016 19:38:53 +0000 (12:38 -0700)]
SkPDF: Use type 2/3 shading for gradient shaders
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=
1925233003
Review-Url: https://codereview.chromium.org/
1925233003
mtklein [Fri, 17 Jun 2016 19:09:16 +0000 (12:09 -0700)]
port SkColorXform_opts to Sk4f
I have tested that this compiles.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2078913003
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/
2078913003
egdaniel [Fri, 17 Jun 2016 19:04:05 +0000 (12:04 -0700)]
Update GYP for linux Vulkan to support SDK 1.0.13
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2077043002
Review-Url: https://codereview.chromium.org/
2077043002
liyuqian [Fri, 17 Jun 2016 18:04:44 +0000 (11:04 -0700)]
Keep drawer open in xlarge-land and add float fps
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2071333002
Review-Url: https://codereview.chromium.org/
2071333002
jvanverth [Fri, 17 Jun 2016 16:29:14 +0000 (09:29 -0700)]
Enable viewer in non-Vulkan builds
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2067023002
Committed: https://skia.googlesource.com/skia/+/
746b792a9aad1230ab69158a51c49b2e5437d95b
Review-Url: https://codereview.chromium.org/
2067023002
reed [Fri, 17 Jun 2016 16:26:41 +0000 (09:26 -0700)]
remove colorprofiletype from imageinfo
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2075853002
Review-Url: https://codereview.chromium.org/
2075853002
brianosman [Fri, 17 Jun 2016 14:03:45 +0000 (07:03 -0700)]
Don't try to set gamma treatment if texture creation fails.
Guessing this will fix the valgrind failure.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2075133002
Review-Url: https://codereview.chromium.org/
2075133002
liyuqian [Fri, 17 Jun 2016 01:09:11 +0000 (18:09 -0700)]
Quickly fix the Compile Problem of Unused kkImageColorXformMetaData
TBR=msarett@google.com
I don't know why the compiler is complaining that since ImageSlide.cpp is still
using it.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2075873003
Review-Url: https://codereview.chromium.org/
2075873003
bungeman [Thu, 16 Jun 2016 22:31:42 +0000 (15:31 -0700)]
Move headers in include/utils/win to src/utils/win.
These headers are not intended for external use and are only used in src.
Review-Url: https://codereview.chromium.org/
2070983002
bungeman [Thu, 16 Jun 2016 21:41:53 +0000 (14:41 -0700)]
Revert of Support pixel antialising in DirectWrite. (patchset #4 id:60001 of https://codereview.chromium.org/
2065833002/ )
Reason for revert:
Moving to the new glyph run analysis changes things anyway, and the Chromium Win 10 bots are having issues with this. Revert until I have time to make the suppression wider.
Original issue's description:
> Support pixel antialising in DirectWrite.
>
> DirectWrite2 supports pixel antialiasing and rendering without hinting.
>
> BUG=skia:5416
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2065833002
>
> TBR=reed
> Will move SkTScopedComPtr into src.
>
> Committed: https://skia.googlesource.com/skia/+/
bd770d619553a88eeaa64ff29082f62db5c9b4d2
TBR=reed@google.com,mtklein@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:5416
Review-Url: https://codereview.chromium.org/
2075913002
liyuqian [Thu, 16 Jun 2016 21:10:34 +0000 (14:10 -0700)]
Use Offscreen Surface for Split Screen
A dashed line is also added to split the screen. Currently, drawing
images with sRGB seems to still have bugs.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069653002
Review-Url: https://codereview.chromium.org/
2069653002
jvanverth [Thu, 16 Jun 2016 21:05:09 +0000 (14:05 -0700)]
Check for some potential subheap allocation failures.
Need to make sure we can allocate areas larger than our specified
subheap sizes, and deal with potential fragmentation in the main
system heap.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2072763002
Review-Url: https://codereview.chromium.org/
2072763002
brianosman [Thu, 16 Jun 2016 20:32:51 +0000 (13:32 -0700)]
Add missing include
TBR=egdaniel@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2070953002
Review-Url: https://codereview.chromium.org/
2070953002
brianosman [Thu, 16 Jun 2016 20:03:24 +0000 (13:03 -0700)]
Lots of progress on switching to SkColorSpace rather than SkColorProfileType
Fixed a bunch of code in Ganesh, as well as usage of SkColorProfileType in most of our tools (DM, SampleApp, Viewer, nanobench, skiaserve, HelloWorld).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069173002
Committed: https://skia.googlesource.com/skia/+/
6a61a875467646f8dbc37cfecf49e12d1f475170
Review-Url: https://codereview.chromium.org/
2069173002
dvonbeck [Thu, 16 Jun 2016 19:39:25 +0000 (12:39 -0700)]
Refactoring of GPU NormalMap handling out into its own class.
The purpose of this change is to refactor the handling of normal maps out of SkLightingShader, laying the groundwork to eventually allow for multiple normal sources.
What this CL includes:
- Created a new 'NormalMapFP', out of the existing normal map reading behavior in LightingFP.
- Encapsulates this new fragment processor on a new class NormalMapSource.
- Created a NormalSource abstraction that will interface with SkLightingShader.
- Adapted SkLightingShader to use the normals from its NormalSource field ON THE GPU SIDE. No changes done to the CPU side yet.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2043393002
Committed: https://skia.googlesource.com/skia/+/
87b0dd00cf9409c5fc990f5d0bb7c0df837f08da
Committed: https://skia.googlesource.com/skia/+/
a7d1e2a57aef2aa4913d4380646d60bbab761318
Review-Url: https://codereview.chromium.org/
2043393002
mtklein [Thu, 16 Jun 2016 19:11:51 +0000 (12:11 -0700)]
Clean up two unlaunched SSE 4.1 8888 blits.
This code was running on our bots but never in Chrome.
That's a bad state to be in.
My plan here use to be to redesign how our 8888 blits worked in SSE 4.1, mainly
for perfect correctness but also for speed, then to spread what I learned there
to SSE2, AVX+, and NEON.
I have since lost interest in changing any aspect of how our legacy 8888 blits
work. There's not much point in making them a bit or two more correct when the
math is fundamentally wrong.
This will cause many diffs in Gold, none perceptible.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062853002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/
6e472093009bf2fc4a8e53010b51040efcb71213
Review-Url: https://codereview.chromium.org/
2062853002
msarett [Thu, 16 Jun 2016 18:42:48 +0000 (11:42 -0700)]
Fix Vulkan build after API change
TBR=brianosman@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2075713002
Review-Url: https://codereview.chromium.org/
2075713002
brianosman [Thu, 16 Jun 2016 18:41:44 +0000 (11:41 -0700)]
Revert of Lots of progress switching to SkColorSpace rather than SkColorProfileType (patchset #10 id:180001 of https://codereview.chromium.org/
2069173002/ )
Reason for revert:
Mac crashes in GrUploadPixmapToTexture
Original issue's description:
> Lots of progress on switching to SkColorSpace rather than SkColorProfileType
>
> Fixed a bunch of code in Ganesh, as well as usage of SkColorProfileType in most of our tools (DM, SampleApp, Viewer, nanobench, skiaserve, HelloWorld).
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069173002
>
> Committed: https://skia.googlesource.com/skia/+/
6a61a875467646f8dbc37cfecf49e12d1f475170
TBR=reed@google.com,herb@google.com,msarett@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2072813002
brianosman [Thu, 16 Jun 2016 18:10:03 +0000 (11:10 -0700)]
Lots of progress on switching to SkColorSpace rather than SkColorProfileType
Fixed a bunch of code in Ganesh, as well as usage of SkColorProfileType in most of our tools (DM, SampleApp, Viewer, nanobench, skiaserve, HelloWorld).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069173002
Review-Url: https://codereview.chromium.org/
2069173002
msarett [Thu, 16 Jun 2016 17:50:55 +0000 (10:50 -0700)]
Implement fast, correct gamma conversion for color xforms
201295.jpg on HP z620
(300x280, most common form of sRGB profile)
QCMS Xform 0.495 ms
Skia Old Xform 0.235 ms
Skia NEW Xform 0.423 ms
Vs Old Code 0.56x
Vs QCMS 1.17x
So to summarize, we are now much slower than before,
but still a bit faster than QCMS. And now we are also
far more accurate than QCMS :).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060823003
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/
2060823003
bungeman [Thu, 16 Jun 2016 17:42:23 +0000 (10:42 -0700)]
Support pixel antialising in DirectWrite.
DirectWrite2 supports pixel antialiasing and rendering without hinting.
BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2065833002
TBR=reed
Will move SkTScopedComPtr into src.
Review-Url: https://codereview.chromium.org/
2065833002
jvanverth [Thu, 16 Jun 2016 16:52:35 +0000 (09:52 -0700)]
Add samples to Viewer.
This adds support with animation, assuming the sample has
implemented onAnimate. Event handling has not been
implemented.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2056343004
Committed: https://skia.googlesource.com/skia/+/
76963e73704a42a18c29d6fbdcccb566e5c67658
Review-Url: https://codereview.chromium.org/
2056343004
msarett [Thu, 16 Jun 2016 14:37:41 +0000 (07:37 -0700)]
Differentiate between sRGBGamma and 2Dot2Gamma
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2067833003
Review-Url: https://codereview.chromium.org/
2067833003
msarett [Thu, 16 Jun 2016 12:33:31 +0000 (05:33 -0700)]
Keep SkColorSpace and SkColorProfileType in sync
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068743003
Committed: https://skia.googlesource.com/skia/+/
04d35bd80dde7cace866781037cbbdcab14ab683
Review-Url: https://codereview.chromium.org/
2068743003
jvanverth [Thu, 16 Jun 2016 11:42:30 +0000 (04:42 -0700)]
Add heap tests for Vulkan.
Also fixes some bugs that were found.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2072453002
Review-Url: https://codereview.chromium.org/
2072453002
egdaniel [Wed, 15 Jun 2016 22:39:13 +0000 (15:39 -0700)]
Revert of Keep SkColorSpace and SkColorProfileType in sync (patchset #2 id:20001 of https://codereview.chromium.org/
2068743003/ )
Reason for revert:
breaking roll.
../../third_party/skia/include/core/SkImageInfo.h:202: error: undefined reference to 'SkColorSpace::NewNamed(SkColorSpace::Named)'
Original issue's description:
> Keep SkColorSpace and SkColorProfileType in sync
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068743003
>
> Committed: https://skia.googlesource.com/skia/+/
04d35bd80dde7cace866781037cbbdcab14ab683
TBR=reed@google.com,brianosman@google.com,herb@google.com,msarett@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2073593002
egdaniel [Wed, 15 Jun 2016 21:28:17 +0000 (14:28 -0700)]
Revert of Refactoring of GPU NormalMap handling out into its own class (patchset #9 id:160001 of https://codereview.chromium.org/
2043393002/ )
Reason for revert:
break deps roll
Original issue's description:
> Refactoring of GPU NormalMap handling out into its own class.
>
> The purpose of this change is to refactor the handling of normal maps out of SkLightingShader, laying the groundwork to eventually allow for multiple normal sources.
>
> What this CL includes:
>
> - Created a new 'NormalMapFP', out of the existing normal map reading behavior in LightingFP.
>
> - Encapsulates this new fragment processor on a new class NormalMapSource.
>
> - Created a NormalSource abstraction that will interface with SkLightingShader.
>
> - Adapted SkLightingShader to use the normals from its NormalSource field ON THE GPU SIDE. No changes done to the CPU side yet.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2043393002
>
> Committed: https://skia.googlesource.com/skia/+/
87b0dd00cf9409c5fc990f5d0bb7c0df837f08da
> Committed: https://skia.googlesource.com/skia/+/
a7d1e2a57aef2aa4913d4380646d60bbab761318
TBR=reed@google.com,dvonbeck@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2068983005
egdaniel [Wed, 15 Jun 2016 21:23:40 +0000 (14:23 -0700)]
Revert of Enable viewer in non-Vulkan builds (patchset #2 id:20001 of https://codereview.chromium.org/
2067023002/ )
Reason for revert:
breaking compile bots
Original issue's description:
> Enable viewer in non-Vulkan builds
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2067023002
>
> Committed: https://skia.googlesource.com/skia/+/
746b792a9aad1230ab69158a51c49b2e5437d95b
TBR=djsollen@google.com,jvanverth@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2072563002
bsalomon [Wed, 15 Jun 2016 21:15:30 +0000 (14:15 -0700)]
Disable SkImage_drawAbandonedGpuImage on Vulkan.
It crashes issuing a clear batch.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060263005
TBR=brianosman@google.com
Review-Url: https://codereview.chromium.org/
2060263005
jvanverth [Wed, 15 Jun 2016 20:44:15 +0000 (13:44 -0700)]
Enable viewer in non-Vulkan builds
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2067023002
Review-Url: https://codereview.chromium.org/
2067023002
csmartdalton [Wed, 15 Jun 2016 19:42:13 +0000 (12:42 -0700)]
Workaround NVIDIA driver bug with NV_sample_mask_override_coverage
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2070133002
Review-Url: https://codereview.chromium.org/
2070133002
msarett [Wed, 15 Jun 2016 19:12:48 +0000 (12:12 -0700)]
Keep SkColorSpace and SkColorProfileType in sync
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068743003
Review-Url: https://codereview.chromium.org/
2068743003
dvonbeck [Wed, 15 Jun 2016 19:07:56 +0000 (12:07 -0700)]
Refactoring of GPU NormalMap handling out into its own class.
The purpose of this change is to refactor the handling of normal maps out of SkLightingShader, laying the groundwork to eventually allow for multiple normal sources.
What this CL includes:
- Created a new 'NormalMapFP', out of the existing normal map reading behavior in LightingFP.
- Encapsulates this new fragment processor on a new class NormalMapSource.
- Created a NormalSource abstraction that will interface with SkLightingShader.
- Adapted SkLightingShader to use the normals from its NormalSource field ON THE GPU SIDE. No changes done to the CPU side yet.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2043393002
Committed: https://skia.googlesource.com/skia/+/
87b0dd00cf9409c5fc990f5d0bb7c0df837f08da
Review-Url: https://codereview.chromium.org/
2043393002
borenet [Wed, 15 Jun 2016 19:07:42 +0000 (12:07 -0700)]
Add asset management scripts
These provide an easy way to create assets to be used by bots,
eg. Android SDK.
To create an asset:
$ infra/bots/assets/assets.py add android_sdk
(adds scripts in infra/bots/assets/android_sdk)
To upload a new version of an asset:
$ infra/bots/assets/android_sdk/upload.py -t $ANDROID_SDK_ROOT
(uploads Android SDK to GS, writes a version file)
$ git commit
$ git cl upload
To download the current version of the asset:
$ infra/bots/assets/android_sdk/download.py -t ../tmp
BUG=skia:5427
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069543002
Review-Url: https://codereview.chromium.org/
2069543002
dvonbeck [Wed, 15 Jun 2016 16:13:50 +0000 (09:13 -0700)]
Added warning for variable-length arrays to GYP, since they are unsupprted under Win targets
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068863002
Review-Url: https://codereview.chromium.org/
2068863002
hstern [Wed, 15 Jun 2016 16:13:34 +0000 (09:13 -0700)]
Add SampleApp that shows problems in stroking
- When the stroke width gets too big the raster algorithm produces
incorrect output
- This also causes the GPU path to produce incorrect output if the width
is larger than an arbitrary limit in GrAALinearizingConvexPathRenderer
(which is the renderer used for this specific geometry)
- Other non-convex geometries probably also produce strange results
BUG=skia:5405,skia:5406
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060903002
Review-Url: https://codereview.chromium.org/
2060903002
bsalomon [Wed, 15 Jun 2016 16:00:21 +0000 (09:00 -0700)]
Actually fix leak in deferred_texture_image_data
TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068863003
Review-Url: https://codereview.chromium.org/
2068863003
bsalomon [Wed, 15 Jun 2016 14:36:15 +0000 (07:36 -0700)]
Add test for rendering an already abandoned texture-backed SkImage.
BUG=chromium:618858
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2067133002
Review-Url: https://codereview.chromium.org/
2067133002
bsalomon [Tue, 14 Jun 2016 23:13:09 +0000 (16:13 -0700)]
Check for abandoned texture in GrTextureAdjuster::refTextureSafeForParams
BUG=chromium:618858
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2063313002
TBR=egdaniel@google.com
Review-Url: https://codereview.chromium.org/
2063313002
egdaniel [Tue, 14 Jun 2016 22:30:05 +0000 (15:30 -0700)]
Revert of Refactoring of GPU NormalMap handling out into its own class (patchset #7 id:120001 of https://codereview.chromium.org/
2043393002/ )
Reason for revert:
Breaking build and deps roll. Need to move include of SkBitmapProcShader in SkLightingShader.cpp from gpu include list to general list.
Original issue's description:
> Refactoring of GPU NormalMap handling out into its own class.
>
> The purpose of this change is to refactor the handling of normal maps out of SkLightingShader, laying the groundwork to eventually allow for multiple normal sources.
>
> What this CL includes:
>
> - Created a new 'NormalMapFP', out of the existing normal map reading behavior in LightingFP.
>
> - Encapsulates this new fragment processor on a new class NormalMapSource.
>
> - Created a NormalSource abstraction that will interface with SkLightingShader.
>
> - Adapted SkLightingShader to use the normals from its NormalSource field ON THE GPU SIDE. No changes done to the CPU side yet.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2043393002
>
> Committed: https://skia.googlesource.com/skia/+/
87b0dd00cf9409c5fc990f5d0bb7c0df837f08da
TBR=reed@google.com,dvonbeck@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2062133004
bsalomon [Tue, 14 Jun 2016 21:44:28 +0000 (14:44 -0700)]
Fix leak of GrContextThreadSafeProxy in deferred_texture_image_data
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069443004
Review-Url: https://codereview.chromium.org/
2069443004
dvonbeck [Tue, 14 Jun 2016 21:43:52 +0000 (14:43 -0700)]
Refactoring of GPU NormalMap handling out into its own class.
The purpose of this change is to refactor the handling of normal maps out of SkLightingShader, laying the groundwork to eventually allow for multiple normal sources.
What this CL includes:
- Created a new 'NormalMapFP', out of the existing normal map reading behavior in LightingFP.
- Encapsulates this new fragment processor on a new class NormalMapSource.
- Created a NormalSource abstraction that will interface with SkLightingShader.
- Adapted SkLightingShader to use the normals from its NormalSource field ON THE GPU SIDE. No changes done to the CPU side yet.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2043393002
Review-Url: https://codereview.chromium.org/
2043393002
bsalomon [Tue, 14 Jun 2016 21:37:21 +0000 (14:37 -0700)]
Fix GrShape to preserve inverseness of rrects for strokes but not dashes.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2051263003
Review-Url: https://codereview.chromium.org/
2051263003
bsalomon [Tue, 14 Jun 2016 21:37:12 +0000 (14:37 -0700)]
Bin circular blur profile textures by scale and blur to radius ratio.
This reduces the number of profile textures that will be generated throughout an animation and also caps the texture size. It could probably be better tuned.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062743003
Review-Url: https://codereview.chromium.org/
2062743003
halcanary [Tue, 14 Jun 2016 18:06:37 +0000 (11:06 -0700)]
SkMultiPictureDocument: don't rely on SkPicture::cullRect
SkPicture::cullRect() is not guaranteed to be the same as the
bounds passed to SkPictureRecorder::beginRecording().
Review-Url: https://codereview.chromium.org/
2067473003
bsalomon [Tue, 14 Jun 2016 16:43:52 +0000 (09:43 -0700)]
Remove command buffer sample app support.
This hasn't compiled for months. We are deemphasizing SampleApp in favor of viewer. Windowed command buffer is not currently something we currently plan to maintain.
TBR=reed@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2061233002
Review-Url: https://codereview.chromium.org/
2061233002
bsalomon [Tue, 14 Jun 2016 16:32:41 +0000 (09:32 -0700)]
Store GLSL max samplers using uint32_t rather than uint8_t
TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2061123003
Review-Url: https://codereview.chromium.org/
2061123003
jvanverth [Tue, 14 Jun 2016 16:01:33 +0000 (09:01 -0700)]
Another linux Vulkan fix
BUG=skia:
TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069443003
Review-Url: https://codereview.chromium.org/
2069443003
jvanverth [Tue, 14 Jun 2016 15:32:44 +0000 (08:32 -0700)]
Fix linux compile error
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2069633002
TBR=egdaniel@google.com
Review-Url: https://codereview.chromium.org/
2069633002
brianosman [Tue, 14 Jun 2016 15:23:48 +0000 (08:23 -0700)]
If we fail to create a window surface with sRGB, try again without.
Fixes crash when switching to sRGB mode on Nexus 9, among others.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2066743003
Review-Url: https://codereview.chromium.org/
2066743003
brianosman [Tue, 14 Jun 2016 11:53:09 +0000 (04:53 -0700)]
Fix compile error (uninitialized variable use)
switch is already handling all cases, but some compilers don't care.
This suppresses errors about potentially using vkLoadOp and vkStoreOp
uninitialized.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2068523002
Review-Url: https://codereview.chromium.org/
2068523002
liyuqian [Mon, 13 Jun 2016 21:38:35 +0000 (14:38 -0700)]
Quickly fix 0f to 0.0f
TBR=egdaniel@google.com
So Visual Studio won't complain.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2064863002
Review-Url: https://codereview.chromium.org/
2064863002
jvanverth [Mon, 13 Jun 2016 21:28:07 +0000 (14:28 -0700)]
Create free list heap for suballocation
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2029763002
Review-Url: https://codereview.chromium.org/
2029763002
liyuqian [Mon, 13 Jun 2016 20:25:52 +0000 (13:25 -0700)]
Quickly fix the double to SkScalar issue
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060823004
Review-Url: https://codereview.chromium.org/
2060823004
liyuqian [Mon, 13 Jun 2016 19:26:45 +0000 (12:26 -0700)]
Display JPGs in Viewer with Split Screen and ColorSpaceXform
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2058753002
Review-Url: https://codereview.chromium.org/
2058753002
bsalomon [Mon, 13 Jun 2016 19:13:03 +0000 (12:13 -0700)]
Move create_image_large in ImageTest.cpp inside #if SK_SUPPORT_GPU
TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062943002
Review-Url: https://codereview.chromium.org/
2062943002
mtklein [Mon, 13 Jun 2016 18:55:57 +0000 (11:55 -0700)]
Revert of Clean up two unlaunched SSE 4.1 8888 blits. (patchset #1 id:1 of https://codereview.chromium.org/
2062853002/ )
Reason for revert:
Breaks a couple Google3 goldens. I need to rebaseline google3 with -DSK_SUPPORT_LEGACY_X86_BLITS first, then reland this.
Original issue's description:
> Clean up two unlaunched SSE 4.1 8888 blits.
>
> This code was running on our bots but never in Chrome.
> That's a bad state to be in.
>
> My plan here use to be to redesign how our 8888 blits worked in SSE 4.1, mainly
> for perfect correctness but also for speed, then to spread what I learned there
> to SSE2, AVX+, and NEON.
>
> I have since lost interest in changing any aspect of how our legacy 8888 blits
> work. There's not much point in making them a bit or two more correct when the
> math is fundamentally wrong.
>
> This will cause many diffs in Gold, none perceptible.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062853002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/
6e472093009bf2fc4a8e53010b51040efcb71213
TBR=reed@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2066453003
egdaniel [Mon, 13 Jun 2016 18:30:10 +0000 (11:30 -0700)]
Revert of Add samples to Viewer. (patchset #4 id:60001 of https://codereview.chromium.org/
2056343004/ )
Reason for revert:
breaking on android compile
Original issue's description:
> Add samples to Viewer.
>
> This adds support with animation, assuming the sample has
> implemented onAnimate. Event handling has not been
> implemented.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2056343004
>
> Committed: https://skia.googlesource.com/skia/+/
76963e73704a42a18c29d6fbdcccb566e5c67658
TBR=bsalomon@google.com,jvanverth@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/
2063883002
mtklein [Mon, 13 Jun 2016 18:27:55 +0000 (11:27 -0700)]
Revert of Update typeface serialization style. (patchset #3 id:40001 of https://codereview.chromium.org/
2030683002/ )
Reason for revert:
serialize-8888 broken for typefacestyles on Mac. (Bold weighting goes missing.)
Original issue's description:
> Update typeface serialization style.
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2030683002
>
> Committed: https://skia.googlesource.com/skia/+/
7ae1c72e593823e6e300e4199558555765bcec17
TBR=reed@google.com,bungeman@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review-Url: https://codereview.chromium.org/
2067583002
msarett [Mon, 13 Jun 2016 18:25:03 +0000 (11:25 -0700)]
Fix SkImage::asLegacyBitmap() to respect color space info
BUG=skia:5428
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2061783002
Review-Url: https://codereview.chromium.org/
2061783002
ericrk [Mon, 13 Jun 2016 18:18:14 +0000 (11:18 -0700)]
Add prescale option to deferred params
Currently, Skia always uploads GPU textures at full resolution. This
change allows us to pass a pre-scale mip level to the
deferred texture image logic, which causes us to pre-scale the image
to the given mip level, and upload that mip level instead of the full
image.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2007113008
Review-Url: https://codereview.chromium.org/
2007113008
mtklein [Mon, 13 Jun 2016 17:48:06 +0000 (10:48 -0700)]
Clean up two unlaunched SSE 4.1 8888 blits.
This code was running on our bots but never in Chrome.
That's a bad state to be in.
My plan here use to be to redesign how our 8888 blits worked in SSE 4.1, mainly
for perfect correctness but also for speed, then to spread what I learned there
to SSE2, AVX+, and NEON.
I have since lost interest in changing any aspect of how our legacy 8888 blits
work. There's not much point in making them a bit or two more correct when the
math is fundamentally wrong.
This will cause many diffs in Gold, none perceptible.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062853002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/
2062853002
bungeman [Mon, 13 Jun 2016 17:35:05 +0000 (10:35 -0700)]
Update typeface serialization style.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2030683002
Review-Url: https://codereview.chromium.org/
2030683002
bsalomon [Mon, 13 Jun 2016 17:22:48 +0000 (10:22 -0700)]
Add control over whether lines are special cased in SkDashPath. Disable when called from GrShape.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2055253002
Review-Url: https://codereview.chromium.org/
2055253002
ethannicholas [Mon, 13 Jun 2016 17:20:52 +0000 (10:20 -0700)]
skiaserve no longer crashes when no X server is present
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2059263002
Review-Url: https://codereview.chromium.org/
2059263002
jvanverth [Mon, 13 Jun 2016 16:31:26 +0000 (09:31 -0700)]
Add samples to Viewer.
This adds support with animation, assuming the sample has
implemented onAnimate. Event handling has not been
implemented.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2056343004
Review-Url: https://codereview.chromium.org/
2056343004
brianosman [Mon, 13 Jun 2016 14:59:48 +0000 (07:59 -0700)]
Better (?) interface for controlling sRGB-ness of mipmaps on GrTexture
Switch to new SkSourceGammaTreatment type. Move to a separate setter,
rather than an optional argument to dirty. Set the flag correctly after
uploading CPU-built mips.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2058143002
Review-Url: https://codereview.chromium.org/
2058143002
brianosman [Mon, 13 Jun 2016 14:49:07 +0000 (07:49 -0700)]
Request an sRGB FBO0 on Android. Fixes sRGB GL mode.
Tried to comment my logic, but it's totally possible that we run on some
device where this doesn't work. From a strict reading of the relevant
specs, what I'm doing shouldn't work on the Nexus 5X, but it's necessary
to get correct results.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2054223002
Review-Url: https://codereview.chromium.org/
2054223002
update-skps [Mon, 13 Jun 2016 11:10:29 +0000 (04:10 -0700)]
Update SKP version
Automatic commit by the RecreateSKPs bot.
TBR=
NO_MERGE_BUILDS
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2065433003
Review-Url: https://codereview.chromium.org/
2065433003
reed [Sat, 11 Jun 2016 11:48:12 +0000 (04:48 -0700)]
tweak mipmap_srgb gm to use integer coordinates
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2061553002
TBR=
Review-Url: https://codereview.chromium.org/
2061553002
bsalomon [Sat, 11 Jun 2016 01:28:06 +0000 (18:28 -0700)]
Make large array of GrShapes in test code heap allocated
TBR=mtklein@chormium.org
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2058343002
Review-Url: https://codereview.chromium.org/
2058343002
csmartdalton [Fri, 10 Jun 2016 21:06:32 +0000 (14:06 -0700)]
Move multisampleDisableSupport into GrCaps
Non-platform code needs this information in order to make rendering
decisions.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2062443002
Review-Url: https://codereview.chromium.org/
2062443002
mtklein [Fri, 10 Jun 2016 20:56:35 +0000 (13:56 -0700)]
Type-erase SkAutoMutexAcquire and SkAutoExclusive.
This is purely for convenience, to not need to write the lock type
in the guard anymore. This should all inline away.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2055023003
Review-Url: https://codereview.chromium.org/
2055023003
reed [Fri, 10 Jun 2016 20:21:07 +0000 (13:21 -0700)]
fix non-square stroke rects (non-aa)
fiddle showing bug: https://fiddle.skia.org/c/
a21ecf5e33da84f52eba26e8c2f3d280
BUG=skia:5408
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060543002
Review-Url: https://codereview.chromium.org/
2060543002
herb [Fri, 10 Jun 2016 20:01:27 +0000 (13:01 -0700)]
Make SkBlitter hierarchy explicit about what needs to be implemented.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2053823002
Review-Url: https://codereview.chromium.org/
2053823002
kjlubick [Fri, 10 Jun 2016 19:54:37 +0000 (12:54 -0700)]
Remove shaderc cmake compiling
It is no longer needed as of https://skia.googlesource.com/skia/+/
b7b4293524e0c1d4c1635c4710aceafaca47b313
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2060503002
Review-Url: https://codereview.chromium.org/
2060503002
bungeman [Fri, 10 Jun 2016 19:43:17 +0000 (12:43 -0700)]
Implement SkFontMgr_Custom::onCreateFromFontData.
This fixes serialization issues. We could create a non-default gx
typeface, but not deserialize it correctly.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2057293002
Review-Url: https://codereview.chromium.org/
2057293002
csmartdalton [Fri, 10 Jun 2016 19:32:57 +0000 (12:32 -0700)]
Fix GrDrawContextPriv::stencilRect
Updates stencilRect to not call getFillRectBatch. This method can
return a batch that uses coverage AA, which isn't appropriate for
stencil draws.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2057613003
Review-Url: https://codereview.chromium.org/
2057613003
bsalomon [Fri, 10 Jun 2016 19:09:59 +0000 (12:09 -0700)]
Extend inverse_paths GM to test dashing
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2057213002
Review-Url: https://codereview.chromium.org/
2057213002
kjlubick [Fri, 10 Jun 2016 19:07:50 +0000 (12:07 -0700)]
Create .gyp rules for shaderc libraries
To build shaderc, gyp used to punt to cmake, which did not work well on the bots and I couldn't find any way to pipe the variables into cmake which would have made it happy.
So, I wrote a gyp file for shaderc and its dependencies. Nothing super complicated (thank goodness), except for the files that spirv-tools needed generated.
BUG=skia:5186
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2047833003
Review-Url: https://codereview.chromium.org/
2047833003
reed [Fri, 10 Jun 2016 18:41:47 +0000 (11:41 -0700)]
respect srgb gamma when building mips
Proposed policy:
- If the target is *legacy* (e.g. L32/PMColor) ignore gamma
- If the target is S32/F16 respect gamma
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2029373004
Review-Url: https://codereview.chromium.org/
2029373004
bsalomon [Fri, 10 Jun 2016 15:05:14 +0000 (08:05 -0700)]
Make GrShape capable of representing inverse filled rrects.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2056523002
Review-Url: https://codereview.chromium.org/
2056523002
caryclark [Fri, 10 Jun 2016 14:59:50 +0000 (07:59 -0700)]
add fail condition to addexpanded
If coincident pairs don't match,
give up rather than deref null.
R=fmalita@chromium.org
BUG=618991
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2058773002
Review-Url: https://codereview.chromium.org/
2058773002
djsollen [Fri, 10 Jun 2016 14:50:00 +0000 (07:50 -0700)]
Cleanup VulkanWindowContext for use on Android
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2058763002
Review-Url: https://codereview.chromium.org/
2058763002
bsalomon [Fri, 10 Jun 2016 13:58:16 +0000 (06:58 -0700)]
Add blurcircles to blacklist for macmini4.1
BUG=skia:5414
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2053373002
Review-Url: https://codereview.chromium.org/
2053373002
egdaniel [Fri, 10 Jun 2016 13:42:45 +0000 (06:42 -0700)]
Define LEAN_AND_MEAN and NOMINMAX for all windows builds
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2055223002
Review-Url: https://codereview.chromium.org/
2055223002
bungeman [Fri, 10 Jun 2016 11:14:51 +0000 (04:14 -0700)]
Implement all onCreateFromStream overloads.
The custom font manager does not implement all onCreateFromStream
overloads leading to warnings. Since this needs to be implemented
anyway, implement it now.
BUG=skia:5412
Review-Url: https://codereview.chromium.org/
2053903003
csmartdalton [Fri, 10 Jun 2016 02:42:38 +0000 (19:42 -0700)]
Fix glDraw*Indirect prototypes
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2050403002
Review-Url: https://codereview.chromium.org/
2050403002
mtklein [Fri, 10 Jun 2016 01:57:20 +0000 (18:57 -0700)]
Skip verylargebitmap serialization on Android too
I'm trying to figure out why we've been skipping serialize-8888 for these.
They don't draw wrong. Seems like they just eat up all memory while serializing,
maybe only in Debug mode.
This should fix the red Nexus9 bot.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2050423003
Review-Url: https://codereview.chromium.org/
2050423003
mtklein [Thu, 9 Jun 2016 20:40:56 +0000 (13:40 -0700)]
Move immintrin/arm_neon includes to where they are used.
On my Mac (so, immintrin), this improves compile time, both wall and cpu,
by about 16%. To test I ran this on an SSD with files hot in their caches:
$ env CC=/usr/bin/clang CXX=/usr/bin/clang++ ./gyp_skia && \
ninja -C out/Release -t clean && \
time ninja -C out/Release
Before: 159 wall / 3367 cpu
159 wall / 3368 cpu
After: 137 wall / 2860 cpu
136 wall / 2863 cpu
I also tried further refining immintrin down to emmintrin / tmmintrin / smmintrin etc.
That made no signficant difference, so I've kept immintrin for its simplicity.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2045633002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
TBR=reed@google.com
No public API changes.
Committed: https://skia.googlesource.com/skia/+/
12dfaaa53c23f3d03050bde8f64136ac1f44164a
Review-Url: https://codereview.chromium.org/
2045633002
jbauman [Thu, 9 Jun 2016 20:24:48 +0000 (13:24 -0700)]
Add NV12 texture conversion support.
If textures 1 and 2 passed into MakeFromYUVTexturesCopy are the same,
then treat the input as NV12 and sample from the g component of texture
2.
BUG=skia:5209
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2016593002
Review-Url: https://codereview.chromium.org/
2016593002
brianosman [Thu, 9 Jun 2016 20:11:08 +0000 (13:11 -0700)]
Add control of manual mipmapping to GrContextOptions
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2054623003
Committed: https://skia.googlesource.com/skia/+/
97e398d98928f9497063594ebe633efe2d0f4968
Review-Url: https://codereview.chromium.org/
2054623003
djsollen [Thu, 9 Jun 2016 20:07:13 +0000 (13:07 -0700)]
Cleanup needed to use WindowContext for Android
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=
2055693003
Review-Url: https://codereview.chromium.org/
2055693003