Change-Id: I8ac756a088c1dddf475bf96f9a4a84ec2cb500be
Reviewed-on: https://skia-review.googlesource.com/c/169763
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
This is a reland of b07aba4214
Original change's description:
> Move remove ptr args to MakeRecAndEffects
>
> Move this conversion out through the transitive closure of calls. As you
> move up the stack, everything becomes refs instread of pointers.
>
> Reorder args of MakeRecAndEffects and setupCache to match the majority of other
> calls.
>
> Change-Id: I72baf457cd9140f76ee5f7122493284c4be5bcd0
> Reviewed-on: https://skia-review.googlesource.com/c/169765
> Reviewed-by: Ben Wagner <bungeman@google.com>
> Commit-Queue: Herb Derby <herb@google.com>
Change-Id: I2ff7f218ecc7b18ae6a2b293cecdb059eea77562
Reviewed-on: https://skia-review.googlesource.com/c/170222
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
This reverts commit b07aba4214.
Reason for revert: Changes GMs
Original change's description:
> Move remove ptr args to MakeRecAndEffects
>
> Move this conversion out through the transitive closure of calls. As you
> move up the stack, everything becomes refs instread of pointers.
>
> Reorder args of MakeRecAndEffects and setupCache to match the majority of other
> calls.
>
> Change-Id: I72baf457cd9140f76ee5f7122493284c4be5bcd0
> Reviewed-on: https://skia-review.googlesource.com/c/169765
> Reviewed-by: Ben Wagner <bungeman@google.com>
> Commit-Queue: Herb Derby <herb@google.com>
TBR=bungeman@google.com,herb@google.com
Change-Id: I29e54cea5b7b5709dd6db6816cb6556d73e003e0
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/170221
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Move this conversion out through the transitive closure of calls. As you
move up the stack, everything becomes refs instread of pointers.
Reorder args of MakeRecAndEffects and setupCache to match the majority of other
calls.
Change-Id: I72baf457cd9140f76ee5f7122493284c4be5bcd0
Reviewed-on: https://skia-review.googlesource.com/c/169765
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
SSE2 and NEON are common baseline instruction sets now,
so there's no need to runtime detect support for these routines.
I simplified the SSE and portable implementations while moving them.
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I34e96851735c8d7ad90198f3ac4bf86ff508f17c
Reviewed-on: https://skia-review.googlesource.com/c/170220
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Bug: skia:7903
Change-Id: I29957babe65e9e1c505166d01490233fe542a559
Reviewed-on: https://skia-review.googlesource.com/c/169824
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
They're only specialized up to SSE2 or NEON,
both of which are typical baseline builds now.
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel
Change-Id: If2b2bbd5b002038c68c0064ee78d75911a33b988
Reviewed-on: https://skia-review.googlesource.com/c/170064
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This was originally a reland of "Fix div-by-zero loophole in gradient factory func", c34dd6c526, but:
The change caused blink layout tests when encountering very small or zero radii. The original patch switched the order of checking if the radii are equal and if the start radius was 0. In the case where both radii are 0, the original code created an actual radial gradient of radius 0 and the new code rejected the shader. A radial gradient with radius of 0 properly renders the last border color as a fill.
This made me realize that the case when the center positions and the radii are the same can be handled more correctly than just always returning an empty shader, so the fix now applies simplifications to the gradient definition depending on the tile mode and should not trigger any blink tests. I added a row to the gradient edge cases GM to make sure it degrades gracefully.
Original change's description:
> Fix div-by-zero loophole in gradient factory func
>
> Bug: oss-fuzz:10373
> Change-Id: I4277fb63e3186ee34feaf09ecf6aeddeb532f9c1
> Reviewed-on: https://skia-review.googlesource.com/c/168269
> Reviewed-by: Kevin Lubick <kjlubick@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Docs-Preview: https://skia.org/?cl=168487
Bug: oss-fuzz:10373
Change-Id: Ib0a6e7f807560a5dcf24d1c8e0146817af2d9606
Reviewed-on: https://skia-review.googlesource.com/c/168487
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
SkBlend::RowFactory is only ever used to pass these function
pointers from SkBlitMask_D32.cpp to SkBlitMask_ARGB32.cpp,
so let's eliminate the middleman.
Change-Id: If74af775bb3cdc3eec9dc4ebeb180ac42b184a54
Reviewed-on: https://skia-review.googlesource.com/c/170062
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
- some refactoring
- remove SK_RESTRICT
- try to remove SK_SUPPORT_LEGACY_A8_MASKBLITTER
Change-Id: I3a270fa2423a66f6e49e5f4f89593a27a9a26e9f
Reviewed-on: https://skia-review.googlesource.com/c/170061
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This reverts commit 51b1c12bbc.
Reason for revert: reverting till flutter gets to 1.1 to fix build issues.
Original change's description:
> Have a GrBackendFormat be stored on gpu proxies.
>
> Bug: skia:
> Change-Id: Iaf1fb24ab29a61d44e5fa59a5e0867ed02dcda90
> Reviewed-on: https://skia-review.googlesource.com/c/168021
> Reviewed-by: Brian Osman <brianosman@google.com>
> Commit-Queue: Greg Daniel <egdaniel@google.com>
TBR=egdaniel@google.com,bsalomon@google.com,brianosman@google.com
Change-Id: I574fdc084ef5994596c51fb0d60423b5dc01b885
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: chromium:903701 chromium:903756
Reviewed-on: https://skia-review.googlesource.com/c/169835
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Change-Id: Ifa2b32b50f0ac04c15c4c9bb1f89272a4689a8f6
Reviewed-on: https://skia-review.googlesource.com/c/170063
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Didn't notice that this was on both SkISize and SkSize the first time
around.
Change-Id: I20d7da0f653aeb4810ab67c9991646b16670497a
Reviewed-on: https://skia-review.googlesource.com/c/169889
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
It always returns nullptr.
Change-Id: I8d24e0f771f292faba5192cbdd311544153d1f3a
Reviewed-on: https://skia-review.googlesource.com/c/170060
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
d856ca48f4..5d2dfa46c0
git log d856ca48f425..5d2dfa46c05c --date=short --no-merges --format='%ad %ae %s'
2018-11-09 jiajia.qin@intel.com ES31: Add array of arrays support in SSBO
2018-11-09 yizhou.jiang@intel.com Enable texture multisample extension on ES3 D3D
Created with:
gclient setdep -r third_party/externals/angle2@5d2dfa46c05c
The AutoRoll server is located here: https://autoroll.skia.org/r/angle-skia-autoroll
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.
CQ_INCLUDE_TRYBOTS=skia.primary:Build-Debian9-Clang-x86_64-Release-ANGLE;skia.primary:Perf-Win10-Clang-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-ShuttleC-GPU-GTX960-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-ShuttleC-GPU-GTX960-x86_64-Debug-All-ANGLE
TBR=allanmac@google.com
Change-Id: I8a38243255fc0ce269c0707b495d9709d4cac364
Reviewed-on: https://skia-review.googlesource.com/c/170002
Reviewed-by: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Some references to the old hint enum
remained, so switch them over.
This fixes bookmaker nightly.
TBR=reed@google.com
NOTRY=true
Docs-Preview: https://skia.org/?cl=169825
Bug: skia:
Change-Id: I2d13ae85e6e14e3445a03e361a50e91bb3bdedd2
Reviewed-on: https://skia-review.googlesource.com/c/169825
Auto-Submit: Cary Clark <caryclark@skia.org>
Commit-Queue: Cary Clark <caryclark@skia.org>
Reviewed-by: Cary Clark <caryclark@skia.org>
This is a little more complicated than other builds because we need
the webassembly binaries out fo skia-wasm-release when we do the webpack
step and we can't run docker (e.g. docker run skia-wasm-release cp foo.wasm)
inside of the skia-public/infra:prod image.
To get around this limitation, we add a step that extracts the binaries to
a /workspace/wasm-products folder (which persists between steps) and mount
that folder as a volume for when we make the jsfiddle output. The only tricky
thing is that we must manually mount the /workspace/__jsfiddle_staging folder
as a volume too, or it doesn't persist to the build step.
Something similar will be needed for skottie.skia.org after
https://skia-review.googlesource.com/c/buildbot/+/169700 lands.
Bug: skia:
Change-Id: Iec05fff7231a28ecfa209489e5c75a9621b334d2
Reviewed-on: https://skia-review.googlesource.com/c/169821
Reviewed-by: Joe Gregorio <jcgregorio@google.com>
Change-Id: Ie66aca5a4d8859b633c18d289495d1af3fd147d4
Reviewed-on: https://skia-review.googlesource.com/c/169764
Commit-Queue: Mike Klein <mtklein@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Many of them will start failing or crashing
once we set SK_DISABLE_READ_BUFFER there.
Alternatively, we could see if we can get away without
SK_DISABLE_READ_BUFFER. It's small value compared to
the other flags.
Change-Id: I447cbd09f68afed81b4b1db8cc234abd22389d46
Reviewed-on: https://skia-review.googlesource.com/c/169762
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Bug: skia:7901
Change-Id: I4d42ad752b18cc1cfbbb5d41eafc499c0541858b
Reviewed-on: https://skia-review.googlesource.com/c/169242
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
This resolves many scoping issues, which complicate this code.
Change-Id: Ib0d283d95ff8a563176a333c8d388947650df4d2
Reviewed-on: https://skia-review.googlesource.com/c/169761
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Bug: skia:
Change-Id: Iaf1fb24ab29a61d44e5fa59a5e0867ed02dcda90
Reviewed-on: https://skia-review.googlesource.com/c/168021
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
more strict bookmaker rules are now in effect;
all methods require documentation, and all documentation
is checked for accuracy.
In this case, the multiple temporary copies of SkPaint::getHinting
and SkPaint::setHinting are checked.
In an upcoming change, bookmaker will ignore symbols marked
as deprecated, experimental, or private; making this bookmaker's
strictness a little more palatable.
TBR=reed@google.com
Docs-Preview: https://skia.org/?cl=169822
Bug: skia:
Change-Id: If1bc079a13828a5ca435911338b9c75acbe8ad6f
Reviewed-on: https://skia-review.googlesource.com/c/169822
Commit-Queue: Cary Clark <caryclark@skia.org>
Commit-Queue: Mike Reed <reed@google.com>
Auto-Submit: Cary Clark <caryclark@skia.org>
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Cary Clark <caryclark@skia.org>
shadeSpanInternal() converts the count to float -> for ridiculously
wide devices, it may not be represented exactly -> we end up potentially
overflowing the dest buffer.
Refactor to stay in int -- this works because SkScalarTruncToInt now
saturates, and it also appears to be marginally faster.
Bug: skia:8536
Change-Id: Ia0b39543ef347332f545af11eabae84c63813fcd
Reviewed-on: https://skia-review.googlesource.com/c/169740
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
The attached Chromium bug got me looking at profiles of .skps that
render very quickly, mostly with memset, and I noticed a few little
microoptimizations to make.
This prebakes a bit more of the blitter memset logic, trimming redundant
calls to SkImageInfo::shiftPerPixel(). This also reduces the number of
lookups of the various SkOpts::memsetNN function pointers from being
once per row to once per rectangle being memset.
Bug: chromium:903195
Change-Id: Ifdee266ecd28be6266ea640fec754399035f083e
Reviewed-on: https://skia-review.googlesource.com/c/169681
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
To measure the performance impact, I patched in
https://skia-review.googlesource.com/c/skia/+/162980 "Add
bench/DecodeBench.cpp" and took the best time (the "min" column):
gcc 7.3, x86_64 Broadwell
Other Before After
11.6ms 7.58ms 5.01ms Decode_GIF_test640x479.gif
21.0ms 15.8ms 11.9ms Decode_GIF_flightAnim.gif
79.8µs 60.1µs 46.6µs Decode_GIF_color_wheel.gif
As ratios:
Other Before After
2.31x 1.51x 1.00x Decode_GIF_test640x479.gif
1.76x 1.33x 1.00x Decode_GIF_flightAnim.gif
1.71x 1.29x 1.00x Decode_GIF_color_wheel.gif
Other is with skia_use_wuffs=false
Before is with skia_use_wuffs=true, before this CL
After is with skia_use_wuffs=true, after this CL
Bug: skia:8235
Change-Id: I7a1df172d172ccd668dfcbeb9728ee053fe831b1
Reviewed-on: https://skia-review.googlesource.com/c/165640
Reviewed-by: Leon Scroggins <scroggo@google.com>
Commit-Queue: Leon Scroggins <scroggo@google.com>
905ee08219..d856ca48f4
git log 905ee08219fa..d856ca48f425 --date=short --no-merges --format='%ad %ae %s'
2018-11-08 syoussefi@chromium.org Vulkan: add clear test for emulated stencil or depth formats
2018-11-08 jiajia.qin@intel.com ES31: Implement MAX_SHADER_STORAGE_BLOCK_SIZE
Created with:
gclient setdep -r third_party/externals/angle2@d856ca48f425
The AutoRoll server is located here: https://autoroll.skia.org/r/angle-skia-autoroll
Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md
If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.
CQ_INCLUDE_TRYBOTS=skia.primary:Build-Debian9-Clang-x86_64-Release-ANGLE;skia.primary:Perf-Win10-Clang-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUC5i7RYH-GPU-IntelIris6100-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-All-ANGLE;skia.primary:Perf-Win10-Clang-ShuttleC-GPU-GTX960-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-AlphaR2-GPU-RadeonR9M470X-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-NUC6i5SYK-GPU-IntelIris540-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-NUCD34010WYKH-GPU-IntelHD4400-x86_64-Debug-All-ANGLE;skia.primary:Test-Win10-Clang-ShuttleC-GPU-GTX960-x86_64-Debug-All-ANGLE
TBR=allanmac@google.com
Change-Id: I239643ec7b2e92ae48f9bb8800767e18b5446ff3
Reviewed-on: https://skia-review.googlesource.com/c/169800
Reviewed-by: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
This is a utility to make vertex generation code in ops simpler and more
consistent. Some existing ops use tricks that may not translate well,
but most will. It enforces writing in-order, leverages templates to
handle sizing and variable numbers of arguments.
For now, I used it to update three ops, showing a relatively simple
case, and a more thorough refactor (of the CircleOp). I think this makes
a stronger case for the GP changes that make stride fast enough for
non-debug use -- most ops won't even need a struct anymore, they can ask
the GP for stride, then use the WriteVertexData utility to write out
data matching the attribute layout.
Bug: skia:
Change-Id: Iec649803181ec5064b4fb9692ae68e6424758944
Reviewed-on: https://skia-review.googlesource.com/c/169340
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
