The canvas only needs to be saved once, per local SkSVGRenderContext.
Add a helper (saveOnce) to implement this optimization.
Change-Id: I0c21fa78ad9fd5d3d11de0a29f8441620488d676
Reviewed-on: https://skia-review.googlesource.com/58340
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Robert Phillips <robertphillips@google.com>
This does not actually add any additional functionality to the generators.
Once this lands I will enable the generators one at a time to more easily
monitor the effects of each one.
Bug: skia:
Change-Id: I382a1acfaebcbf9ad44c9873b87cdbbe02a13602
Reviewed-on: https://skia-review.googlesource.com/57083
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
211bff3f6c..852fe87341
$ git log 211bff3f6..852fe8734 --date=short --no-merges --format='%ad %ae %s'
2017-10-10 oetuaho Fix HLSL for switch statements that don't end in a branch
Created with:
roll-dep skia/third_party/externals/angle2
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Perf-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Test-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Test-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Test-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Build-Debian9-GCC-x86_64-Release-ANGLE
TBR=reed@google.com
Change-Id: I55ec97c9ff596c2c5c7072e66497b58cb69efb2f
Reviewed-on: https://skia-review.googlesource.com/58365
Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
This patch returns error and exit from loop when there is error in
comparing bitmap test.
Bug: skia:None
Change-Id: If7fe9d8cf6b936bf1fb84de69e73c453a6c253de
Reviewed-on: https://skia-review.googlesource.com/57600
Reviewed-by: Leon Scroggins <scroggo@google.com>
Commit-Queue: Leon Scroggins <scroggo@google.com>
The method is used for clip path resolution, and is supposed to transform the path
into parent node coords -- so it should use the forward matrix, not the inverse.
Change-Id: Id6eb9fbaf024b083e19f56eddb1c293becea48fc
Reviewed-on: https://skia-review.googlesource.com/58280
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
I'm hoping that recent fixes to timing logic will prevent these problems.
Bug: skia:6534
Change-Id: Ia735e91887164c7a4882d5d3dd67df8c6115bfe9
Reviewed-on: https://skia-review.googlesource.com/54660
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
78507c6ee5..128e536ab6
$ git log 78507c6ee..128e536ab --date=short --no-merges --format='%ad %ae %s'
2017-10-11 yunchao.he Fix a small build issue.
Created with:
roll-dep skia/third_party/externals/angle2
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Perf-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Test-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Test-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Test-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Build-Debian9-GCC-x86_64-Release-ANGLE
TBR=reed@google.com
Change-Id: Iee3bad5e3f4f140b26654df65a66f0754de1fc97
Reviewed-on: https://skia-review.googlesource.com/58221
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
We've been copying the text first, then the transforms.
That's a good way to get the transforms out of alignment.
This CL swaps the order of the two.
Bug: skia:7133
Change-Id: If8cd402b9ffba1787345dc4b5ffd4ee6abb14f33
Reviewed-on: https://skia-review.googlesource.com/57941
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Add mulHi to base SkNx, and specialize implementations for Sk4u for
neon and sse.
Add casts for converting from uint8_t by 4 to uint32_t by 4.
Cq-Include-Trybots: skia.primary:Test-Debian9-Clang-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD
Change-Id: I29a32e2ad9812a47fff841ceca334e562362836f
Reviewed-on: https://skia-review.googlesource.com/57960
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Herb Derby <herb@google.com>
Previously, we were reporting numbers that were far too low, because
we were getting way ahead of the GPU, and then spending all of our time
in finish (which isn't timed). That led to us picking very high loop
counts, so our wall clock time to run nanobench was very high, and our
reported times were very low. This fixes all of that, and removes all
the spam about not having fence support.
Change-Id: Ib9dfc043da82bf8ee6645b8627cfade66eb9864e
Reviewed-on: https://skia-review.googlesource.com/58001
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
On ANGLE, at least, this frame gives us much more consistent results.
Bug: skia:
Change-Id: Ifdecc8451ef51490c08057645214738180b1a366
Reviewed-on: https://skia-review.googlesource.com/57884
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Bug: skia:
Change-Id: I4f35413995cf73c6f130476d6b36e530120aa7ed
Reviewed-on: https://skia-review.googlesource.com/57901
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Reed <reed@google.com>
Currently we use 'fill-rule' when emitting clip paths. This is wrong:
per spec [1], clip paths observe 'clip-rule', not 'fill-rule'.
[1] https://www.w3.org/TR/SVG/masking.html#ClipRuleProperty
Change-Id: Idf81de05e9601663c8dbc9856900ffa679daf4a5
Reviewed-on: https://skia-review.googlesource.com/57661
Reviewed-by: Robert Phillips <robertphillips@google.com>
Reviewed-by: Stephan Altmueller <stephana@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
- Actually request extension version of fence sync functions
- Fix incorrect usage of dlopen/dlsym
- Also fixed same bugs in Mac code, although we never hit that
code path.
Should fix iOS devices, giving more accurate (and less spammy)
results from nanobench.
Bug: skia:
Change-Id: I3456b301ef9b0b6559160d1d21c77bd93139d39a
Reviewed-on: https://skia-review.googlesource.com/57740
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
1fce3283f0..af5070f72d
$ git log 1fce3283f..af5070f72 --date=short --no-merges --format='%ad %ae %s'
2017-10-10 oetuaho Hide RemoveSwitchFallThrough implementation in the .cpp file
2017-09-27 oetuaho Add support for arrays of arrays to VariableLocation
2017-10-10 jiajia.qin Remove some unused codes
Created with:
roll-dep skia/third_party/externals/angle2
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Perf-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Test-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Test-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Test-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Build-Debian9-GCC-x86_64-Release-ANGLE
TBR=reed@google.com
Change-Id: Id947fe40368713be2158246b9e62540b676c0fdb
Reviewed-on: https://skia-review.googlesource.com/57781
Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
The fuzzer has built a single test case that causes out-of-bounds
reads on both the src and dst tables. I'm glad we have it.
Next follow ups may include:
- have byte_tables_rgb do its own clamping
- replace table_{r,g,b} here with a single stage analogous
to byte_tables_rgb that looks up the three float tables
safely
- maybe replace byte_tables_rgb with that. I'm not really
sure why src->XYZ tables are floats but XYZ->dst are bytes.
I'm going to try some of these out before attempting to reland
in Chrome. As usual, rebaselining Blink makes things a pain.
Bug: chromium:772684, skia:7114
Change-Id: Id8759e766330e1c7689c0847bf2cd35d422ebbcd
Reviewed-on: https://skia-review.googlesource.com/57760
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
8886f0fc8d..1fce3283f0
$ git log 8886f0fc8..1fce3283f --date=short --no-merges --format='%ad %ae %s'
2017-10-09 jmadill Samples: Add command line flag to select renderer.
2017-10-09 brianosman Fix several bugs in the timeout logic of clientWait
Created with:
roll-dep skia/third_party/externals/angle2
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Perf-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Test-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Test-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Test-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Build-Debian9-GCC-x86_64-Release-ANGLE
TBR=reed@google.com
Change-Id: I8f9c178dc0485f299e5bfda18402414a92c9bfd9
Reviewed-on: https://skia-review.googlesource.com/57700
Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
W/o this fix the strict constraint on the specialImage draws can sometimes be removed. This CL ensures that the temporary SkImage always has the actual dimensions of the special Image.
Note: the replacement of full screen clears w/ draws revealed this bug b.c. the image filters were only drawing a rect for the content area of the special images. Thus when the final DAG result was drawn to the canvas some of the bleed through (from the removal of the strict constraint) was showing.
Bug: 755871, 768134
skbug.com/7122
Change-Id: Id67b7143225c24b716e260c973f3bbf68f20188a
Reviewed-on: https://skia-review.googlesource.com/57660
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Adds it for Vulkan and non-Vulkan
Bug: skia:7080
NOTRY=true
Change-Id: I5c4b30b91b96ad6ce9694c42550c4e909de4ba41
Reviewed-on: https://skia-review.googlesource.com/57560
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Ravi Mistry <rmistry@google.com>
https://www.w3.org/TR/SVG/linking.html#AElement
The Skia SVG DOM is not a user agent, so link semantics are not
particularly interesting. But since <a> can wrap actual content, it is
important to not drop it on the floor.
The simplest thing to do is turn it into a <g> node and run with it.
Change-Id: I4e103553354746ceb49bc1038680a1c477a1320a
Reviewed-on: https://skia-review.googlesource.com/57620
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
359487c822..d63d0007c1
$ git log 359487c82..d63d0007c --date=short --no-merges --format='%ad %ae %s'
2017-10-06 kainino dontInitializeUninitializedLocals on Qualcomm only
Created with:
roll-dep skia/third_party/externals/angle2
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Perf-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Perf-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Perf-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Test-Win10-MSVC-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-ANGLE,Test-Win10-MSVC-Golo-GPU-QuadroP400-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-ANGLE,Test-Win10-MSVC-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-ANGLE,Test-Win10-MSVC-ShuttleC-GPU-GTX960-x86_64-Debug-ANGLE,Build-Debian9-GCC-x86_64-Release-ANGLE
TBR=reed@google.com
Change-Id: If5b0c9b3517a789931cb896da8f7ece996114181
Reviewed-on: https://skia-review.googlesource.com/57422
Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
This reverts commit b6abb9b4e0.
Reason for revert: It doesn't make sense that this CL would affect the tests implicated in the perf regression in skia:7143, and the revert had no effect on the perf of those tests. Seems like the perf alert was either noise or due to a different CL.
Original change's description:
> Revert "Fix a couple float-cast-overflow in SkScan*."
>
> This reverts commit 3cd0bef0fd.
>
> Reason for revert: https://bugs.chromium.org/p/skia/issues/detail?id=7143
>
> Original change's description:
> > Fix a couple float-cast-overflow in SkScan*.
> >
> > Bug: skia:5060
> > Change-Id: I60a48993c77631aaad9354bb86b13204dc618bf4
> > Reviewed-on: https://skia-review.googlesource.com/47422
> > Commit-Queue: Ben Wagner <benjaminwagner@google.com>
> > Reviewed-by: Mike Reed <reed@google.com>
>
> TBR=benjaminwagner@google.com,reed@google.com
>
> # Not skipping CQ checks because original CL landed > 1 day ago.
>
> Bug: skia:7143
> Change-Id: I0f19720a7d8344789a375bbb6b9e28bf4f4e55ae
> Reviewed-on: https://skia-review.googlesource.com/57240
> Commit-Queue: Ben Wagner <benjaminwagner@google.com>
> Reviewed-by: Ben Wagner <benjaminwagner@google.com>
TBR=benjaminwagner@google.com,reed@google.com
Change-Id: I29ac47d6665e2e52ee2a6500488dc407c8d2af1c
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:7143
Reviewed-on: https://skia-review.googlesource.com/57440
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>
Newer clang compiler produces the diagnostic
../../tests/PathOpsDebug.cpp:1349:44:
error: comparison of unsigned enum expression >= 0 is always true
[-Werror,-Wtautological-unsigned-enum-zero-compare]
if (!firstOp && c->operand() && op >= 0) {
The check seems to be defending against bad values of SkPathOp, but
this now seems unnecessary. This also adds 'rdiff' to kPathOpStr for
future debugging.
Change-Id: I08e5ba77c56e519ce4d9ae89491f7ccd5eb7f944
Reviewed-on: https://skia-review.googlesource.com/57104
Reviewed-by: Cary Clark <caryclark@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Also adds a presubmit to prevent adding trailing whitespace to source
code in the future.
Change-Id: I41a4df81487f6f00aa19b188f0cac6a3377efde6
Reviewed-on: https://skia-review.googlesource.com/57380
Reviewed-by: Ravi Mistry <rmistry@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Also add a presubmit so they don't get added to source code.
Change-Id: I6a85c6a934b1068a63646a0dcc0d3a08baa96ced
Reviewed-on: https://skia-review.googlesource.com/57110
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
