Of note, the perf results reported are not directly
comparable CPU->GPU because the GPU ones are likely
emulated (no real gpu in the Docker container on an
GCE VM).
Bug: skia:
Change-Id: I52259085f4d9e22c45b67f3e8ce1211a5c6c6d3e
Reviewed-on: https://skia-review.googlesource.com/c/163126
Reviewed-by: Kevin Lubick <kjlubick@google.com>
A fuzzer has produced one, and hit the existing 4x4 == 3x3 assert.
Bug: chromium:895763
Change-Id: Id24b982464166fe5b79ab26c1a3a0e931b76dc4c
Reviewed-on: https://skia-review.googlesource.com/c/163166
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
All of Flutter's non-test drawing is done with Ganesh,
and the few shaders that are legacy-only are not used by Flutter.
This should cut a bunch of essentially-dead code from Flutter, though
some of the Flutter test images may change because those test images are
software rasterized. Hopefully nothing should look terribly different.
Looks like this cuts ~27K from Flutter.
Change-Id: I4c8078db17fe9d24b15ed9d8c6adb528a382f02a
Reviewed-on: https://skia-review.googlesource.com/c/163165
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This reverts commit 63bc5891e4.
Reason for revert: Chrome test has been fixed.
Original change's description:
> Revert "Remove ancient support for kUnknown_SkColorType in SkBitmapDevice"
>
> This reverts commit e3a5b1960d.
>
> Reason for revert: possibly breaking chrome roll
>
> Original change's description:
> > Remove ancient support for kUnknown_SkColorType in SkBitmapDevice
> >
> > Bug: skia:
> > Change-Id: I7559544f46b49be5ac6b38add8d1b61c010d3d6f
> > Reviewed-on: https://skia-review.googlesource.com/c/162284
> > Reviewed-by: Mike Reed <reed@google.com>
> > Commit-Queue: Brian Osman <brianosman@google.com>
>
> TBR=mtklein@google.com,brianosman@google.com,reed@google.com
>
> Change-Id: I95d502f80fc3c336484b306b50c9b5fe16f26a82
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: skia:
> Reviewed-on: https://skia-review.googlesource.com/c/163040
> Reviewed-by: Greg Daniel <egdaniel@google.com>
> Commit-Queue: Greg Daniel <egdaniel@google.com>
TBR=egdaniel@google.com,mtklein@google.com,brianosman@google.com,reed@google.com
Change-Id: I37aa32558cd75ddca8c3ba150e3b90d09033372b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/163127
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Current savings : 96K on intel mac
Breaks -- these subclass don't have a pipeline
- emboss
- perlinnoise
For now, shadercontext is still on -- clients must opt-out with SK_DISABLE_LEGACY_SHADERCONTEXT
Bug: skia:
Change-Id: Idd14cae4013e8de92f16b0707f8b8d43bb6d11dd
Reviewed-on: https://skia-review.googlesource.com/c/162362
Auto-Submit: Mike Reed <reed@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>
1d8a783c9d..c20ead32ad
git log 1d8a783c9db2..c20ead32ad6d --date=short --no-merges --format='%ad %ae %s'
2018-10-17 tikuta@chromium.org Remove unused lambda capture
2018-10-17 jmadill@chromium.org Vulkan: Fix deleting in-use descriptor sets.
Created with:
gclient setdep -r third_party/externals/angle2@c20ead32ad6d
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=egdaniel@google.com
Change-Id: Ibe81a4c380c942ffe6e43edc99ad1201d8e56216
Reviewed-on: https://skia-review.googlesource.com/c/163180
Reviewed-by: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
The serialization showed up as a significant cost on profiling,
primarily because of the lock acquired for query the glyph count of a
typeface and the SkData allocation each time. Avoid this by caching the
data for each typeface.
Note that since the typeface is already cached on the client side, we
shouldn't need to resend the typeface data each time. But serializing a
typeface doesn't guarentee that it will be sent, because of which we
can not make any assumptions about client state after serialize is
called.
Bug: 894200
Change-Id: I45e43e297d817826646ab08bd4937b2625603858
Reviewed-on: https://skia-review.googlesource.com/c/162841
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Khusal Sagar <khushalsagar@chromium.org>
Auto-Submit: Khusal Sagar <khushalsagar@chromium.org>
Change-Id: I2f08a3d26f411ef34d964a8f324696e97e8970ed
Reviewed-on: https://skia-review.googlesource.com/c/163162
Commit-Queue: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Auto-Submit: Herb Derby <herb@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Bug: skia:
Change-Id: If7f5eaf19fd268613ac883b268b424ca84b9be00
Reviewed-on: https://skia-review.googlesource.com/c/162660
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
public before private
some c++14 constexprification
remove unused functions
Change-Id: Idb49c476020c7fe9be624fc21c1371e99bf22bb1
Reviewed-on: https://skia-review.googlesource.com/c/162821
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
This reverts commit af119e58dc.
Reason for revert: Chrome reverted the CL that removed this - somehow things haven't exploded, but I'm putting this back for now.
Original change's description:
> Remove SK_LEGACY_XFORM_CANVAS_IN_PICTURE_IMAGES
>
> Chrome no longer uses this.
>
> Bug: skia:
> Change-Id: I245c592a040cd113e9bd3ddd1e582d27070baee1
> Reviewed-on: https://skia-review.googlesource.com/c/162740
> Auto-Submit: Brian Osman <brianosman@google.com>
> Commit-Queue: Mike Reed <reed@google.com>
> Reviewed-by: Mike Reed <reed@google.com>
TBR=mtklein@google.com,brianosman@google.com,reed@google.com
Change-Id: I681a17763815d6e16e8797789a006d72b84b4478
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/163120
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Basically no hoops to jump through - the same binary
that works in the browser works in Node.
Tested locally with Node 8.9.3.
This aligns the GPU and CPU APIs (that is, makeSurface)
and breaks out the GPU/CPU js interface parts into
their own files. We only need one of them and we know
which at compile time.
Bug: skia:
Change-Id: I6d141387403a792d2374cf904872c6dbc999abfb
Reviewed-on: https://skia-review.googlesource.com/c/162746
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Joe Gregorio <jcgregorio@google.com>
(fixing msan/asan/ubsan failure)
Pathops used templates for curve intersection.
Since only one template is required if curves share
an abstract base, remove the template altogether.
This makes the code easier to read, and incidentally
makes it slightly smaller and much faster.
This also removes debugging code specific to templates,
and removes Simplify code which isn't covered by tests
or fuzz.
This shaves the execution time of
pathops_unittest -V -x from 6m to 3m23s.
R=kjlubick@google.com
Bug: skia:
Change-Id: I3392df98244083d0327ce9c787dfe24d326ef4ed
Reviewed-on: https://skia-review.googlesource.com/c/162742
Commit-Queue: Cary Clark <caryclark@skia.org>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Fuzz SkAndroidCodec to help to catch errors in both incrementalDecode
and scanlineDecode. Try a variety of sample sizes, but cap it at 64.
Though sometimes larger sample sizes are used, the lower ones tend to
more common. Also draw the resulting bitmap to verify that we
initialized all pixels.
Independently test incrementalDecode to ensure that it initializes
rowsDecoded.
Change-Id: I20d8a408cd280262fdc62f902a6f04f0f57f5ad2
Reviewed-on: https://skia-review.googlesource.com/c/162025
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Auto-Submit: Leon Scroggins <scroggo@google.com>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
d633b1569a..1d8a783c9d
git log d633b1569a21..1d8a783c9db2 --date=short --no-merges --format='%ad %ae %s'
2018-10-17 jiajia.qin@intel.com Re-land: "ES31: Add vector and matrix support in SSBO for D3D"
2018-10-16 jmadill@chromium.org Vulkan: Update Android expectations.
Created with:
gclient setdep -r third_party/externals/angle2@1d8a783c9db2
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=egdaniel@google.com
Change-Id: I7ee6f36110e8cecd0ba296e63dbfd73d959faa1d
Reviewed-on: https://skia-review.googlesource.com/c/163080
Reviewed-by: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
This reverts commit e3a5b1960d.
Reason for revert: possibly breaking chrome roll
Original change's description:
> Remove ancient support for kUnknown_SkColorType in SkBitmapDevice
>
> Bug: skia:
> Change-Id: I7559544f46b49be5ac6b38add8d1b61c010d3d6f
> Reviewed-on: https://skia-review.googlesource.com/c/162284
> Reviewed-by: Mike Reed <reed@google.com>
> Commit-Queue: Brian Osman <brianosman@google.com>
TBR=mtklein@google.com,brianosman@google.com,reed@google.com
Change-Id: I95d502f80fc3c336484b306b50c9b5fe16f26a82
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/163040
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
If the GrTextBlob does not contain any subpixel position text,
then any translation is ok. This is because all positions and
SkGlyphCache entries will be aligned to pixel boundaries.
Change-Id: Ie2bfd72710295b98c65052a857b6bd33a4858b65
Reviewed-on: https://skia-review.googlesource.com/c/162860
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
6eba3c6104..d633b1569a
git log 6eba3c6104f2..d633b1569a21 --date=short --no-merges --format='%ad %ae %s'
2018-10-16 fjhenigman@chromium.org Get rid of VertexFormatType.
Created with:
gclient setdep -r third_party/externals/angle2@d633b1569a21
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=egdaniel@google.com
Change-Id: I850e9491cc6ef5886a44655d595dc305d308cc73
Reviewed-on: https://skia-review.googlesource.com/c/162840
Reviewed-by: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: skia-autoroll <skia-autoroll@skia-public.iam.gserviceaccount.com>
function now takes a constant ref, as needed.
re-written to make it easier to think about state.
Change-Id: I5b9935cd88b4f8b1f35d4df622adcfb61655d42b
Reviewed-on: https://skia-review.googlesource.com/c/160383
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Hal Canary <halcanary@google.com>
This is a reland of 4b7b2ceb4a
Original change's description:
> Have GrAtlasManager use legacy behavior
>
> The old code would generate a 1024 x 2048 x 4 atlas given a
> max size of 4MiB. And, the plot sizes would be 256 x 256. This CL
> mimics that behavior.
>
> Change-Id: Iae741d5409daa034c6c3086c8a09476fb343353d
> Reviewed-on: https://skia-review.googlesource.com/c/162321
> Commit-Queue: Herb Derby <herb@google.com>
> Reviewed-by: Mike Klein <mtklein@google.com>
Change-Id: I86cc8d2fb07f5317897f094c6fc172c1ca4c8a1c
Reviewed-on: https://skia-review.googlesource.com/c/162720
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Herb Derby <herb@google.com>
All uses have been converted to SkPMColor4f (or similar).
Bug: skia:
Change-Id: I220bd5eaf6c35b17321c1e8bc92ace7ff92908c4
Reviewed-on: https://skia-review.googlesource.com/c/162749
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
- Initialize rowsDecoded in SkSampledCodec. Otherwise,
fillIncompleteImage may be called with an uninitialized
value. This change was originally uploaded to AOSP as
https://android-review.googlesource.com/c/platform/external/skia/+/785816
- If SkPngCodec hits an error, still transform from the
interlace buffer (if needed) and set rowsDecoded properly.
- Do not copy uninitialized memory from the interlace buffer.
- Make BRD treat kErrorInInput like kIncompleteInput. The two errors
are different for the purposes of incremental decode. For a direct
decode, they're essentially the same - part was decoded, but then
the decode was interrupted. This allows testing images with
errors on the bots without reporting a failure.
- Add some images to test incomplete pngs
- errorInInput.png: color_wheel.png truncated with extra 0s at the
end to trigger kErrorInInput
- incInterlaced.png: basi3p02.png (from PNGSuite) similarly modified.
This triggers kIncompleteInput. We have several incomplete PNGs that
we are testing on the bots, but this is the first interlaced
- errorInInputInterlaced.png: This is the same as incInterlaced with
more zeroes at the end. The incorrect CRC (all zeroes) results in
kErrorInInput
Change-Id: I25e1ea917e1c1ca212f104cb9d70290265a51b2d
Reviewed-on: https://skia-review.googlesource.com/c/161822
Commit-Queue: Leon Scroggins <scroggo@google.com>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
This is the most misleading use-case, because these values are either
premul or unpremul depending on the gradient's flags. It turns out that
we only need to "lie" and alias in one spot, though, so I think this is
better than adding an entirely new kUnknown_SkAlphaType specialization
of SkRGBA4f.
Bug: skia:
Change-Id: I4b7e280355219d439f50a7fe54230ea17715abc3
Reviewed-on: https://skia-review.googlesource.com/c/162745
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Bug: skia:
Change-Id: I7559544f46b49be5ac6b38add8d1b61c010d3d6f
Reviewed-on: https://skia-review.googlesource.com/c/162284
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Chrome no longer uses this.
Bug: skia:
Change-Id: I245c592a040cd113e9bd3ddd1e582d27070baee1
Reviewed-on: https://skia-review.googlesource.com/c/162740
Auto-Submit: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Reed <reed@google.com>
Reviewed-by: Mike Reed <reed@google.com>
This adds SkPMColor4f conversions to/from RGBA bytes (ie GrColor).
I had previously made some free functions that did the same thing.
I'm ambivalent about which option is nicer, but wanted to have one
method, so I converted everything to use the new versions.
Bug: skia:
Change-Id: I4194c44b5bd12228075fd1932a14cf31c8d6a3c1
Reviewed-on: https://skia-review.googlesource.com/c/162560
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This reverts commit 521f1ed0b6.
Reason for revert: msan ubsan errors
Original change's description:
> remove pathop template
>
> Pathops used templates for curve intersection.
> Since only one template is required if curves share
> an abstract base, remove the template altogether.
>
> This makes the code easier to read, and incidentally
> makes it slightly smaller and much faster.
>
> This also removes debugging code specific to templates,
> and removes Simplify code which isn't covered by tests
> or fuzz.
>
> This shaves the execution time of
> pathops_unittest -V -x from 6m to 3m23s.
>
> R=kjlubick@google.com
>
> Bug: skia:
> Change-Id: I00c08210e47efed83295276ae89ad64e7ec07ade
> Reviewed-on: https://skia-review.googlesource.com/c/162021
> Commit-Queue: Cary Clark <caryclark@google.com>
> Reviewed-by: Kevin Lubick <kjlubick@google.com>
> Reviewed-by: Cary Clark <caryclark@google.com>
TBR=kjlubick@google.com,caryclark@google.com,caryclark@skia.org
Change-Id: Ic5828f7affb7df96ed4ca79f037cdbcfaea24384
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/162643
Reviewed-by: Cary Clark <caryclark@google.com>
Commit-Queue: Cary Clark <caryclark@google.com>
Pathops used templates for curve intersection.
Since only one template is required if curves share
an abstract base, remove the template altogether.
This makes the code easier to read, and incidentally
makes it slightly smaller and much faster.
This also removes debugging code specific to templates,
and removes Simplify code which isn't covered by tests
or fuzz.
This shaves the execution time of
pathops_unittest -V -x from 6m to 3m23s.
R=kjlubick@google.com
Bug: skia:
Change-Id: I00c08210e47efed83295276ae89ad64e7ec07ade
Reviewed-on: https://skia-review.googlesource.com/c/162021
Commit-Queue: Cary Clark <caryclark@google.com>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Cary Clark <caryclark@google.com>
