Chrome stopped using them in January:
https://chromium-review.googlesource.com/c/chromium/src/+/832593
For them, it significantly accelerated goma builds. From my tests,
I think we'll see some small speedup. Note that our older build
system invoked through CMD, which limited us to 8k of command line.
GN/ninja don't do that, so we have the full 32k limit. However,
we do use CMD to do our 32-bit builds... But that doesn't matter,
becase AFAICT the longest command line we generate right now is
only about 2.5k long.
Note that (like Chromium) this continues to use rsp files to link.
Those command lines *can* become ridiculously long.
The original motivation, and a nice benefit of this change:
It makes the output of ninja's compdb tool much more useful.
That tool emits JSON in a standard format describing the commands
used to build each source file. Other tools (eg Visual Studio Code)
can parse that JSON, and deduce the correct defines and include
paths to help with symbol navigation.
Change-Id: I73124f13d5117a0c31445cf6ce1c506d2f73609f
Reviewed-on: https://skia-review.googlesource.com/142584
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Bug: skia:8136
Change-Id: I18c4ad549c52346ebfe23d172597d5da205e5c4d
Reviewed-on: https://skia-review.googlesource.com/142105
Commit-Queue: Eric Boren <borenet@google.com>
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Instead of using the text blob through the stack start
using the glyph run list.
This CL is similar to a portion of https://skia-review.googlesource.com/c/skia/+/137224
which was reverted.
Change-Id: I1f0619bd2d13523f9af1a68ab27fb26abd086add
Reviewed-on: https://skia-review.googlesource.com/141543
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Some builds set NTDDI_VERSION to a very old version of Windows to ensure
they run on those older versions of Windows. The dwrite_3.h header uses
NTDDI_VERSION in an odd manner, hiding immutable declarations of enums,
structs, and interfaces. NTDDI_VERSION was indended to hide functions
and extensions to structs. DWrite has one function (to create the
factory) and no structs which will be extended.
Skia is already tested to run on the oldest supported Windows platform
(Windows 7), so an external build building Skia with NTDDI_VERSION as a
build define should have no issues at runtime if Skia ignores it. If
Skia doesn't ignore NTDDI_VERSION in this case the DWrite backend will
be runtime limited to interfaces in very old versions of Windows,
instead of using newer interfaces which become available at runtime.
Change-Id: I3e9ac2c4116d75588bfff391928d3cd446d6363a
Reviewed-on: https://skia-review.googlesource.com/142324
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Rename to clarify that it is a "raster" cache we're talking about.
clean up -- motivated by looking to make the purge of a staleID synchronous
Bug: skia:
Change-Id: I87493690dd5a4f2ebd002e2635ebd0e44fe27320
Reviewed-on: https://skia-review.googlesource.com/142325
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Mike Reed <reed@google.com>
bookmaker bot checked in more md files
with the wrong fiddle hashes.
Ravi is landing a CL to preventing it
from checking in more errors, for now.
Sadly, creating this CL also corrupted
the md output -- luckily it was an error
that could be detected. But perhaps
other errors are going undetected.
TMI: The call to fiddlecli wrote the wrong
"text" string to SKIRect_isEmpty, first
time I've seen that. Running it again
generated the correct output.
TBR=rmistry@google.com,jcgregario@google.com
No-Try: true
Docs-Preview: https://skia.org/?cl=142329
Bug: skia:6818
Change-Id: I6b57a4a070991b76a1acc44fb024439defc2d977
Reviewed-on: https://skia-review.googlesource.com/142329
Commit-Queue: Cary Clark <caryclark@google.com>
Auto-Submit: Cary Clark <caryclark@google.com>
Reviewed-by: Cary Clark <caryclark@google.com>
6a5d98c4b5..da92a476b5
git log 6a5d98c4b528..da92a476b50a --date=short --no-merges --format='%ad %ae %s'
2018-07-19 jiawei.shao@intel.com Clean up VaryingPacking
Created with:
gclient setdep -r third_party/externals/angle2@da92a476b50a
The AutoRoll server is located here: https://angle-skia-roll.skia.org
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=bungeman@google.com
Change-Id: I9b9db6f672ff6bca56995e28e60585a6b45b5431
Reviewed-on: https://skia-review.googlesource.com/142384
Reviewed-by: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
The SKTextBlob serialization code that is tested by Skia is
unused by Chrome. The serialization code that is
used by Chrome is untested by Skia.
Remove the unused code; test the used code.
The code path introduced nearly a year ago,
likely for slimming paint, attempts to make
text blobs smarter by allowing the reuse of typefaces.
Maybe there needs to be a Chrome bug / feature
request to use this?
If if turns out there is no interest to do so,
This CL aligns used interfaces with tests.
R=reed@google.com,fmalita@google.com,bungeman@google.com
Bug: skia:6818
Change-Id: I9b3ec0c326495322986ba26f20f901bcb208be73
Reviewed-on: https://skia-review.googlesource.com/141542
Reviewed-by: Cary Clark <caryclark@skia.org>
Reviewed-by: Florin Malita <fmalita@chromium.org>
Reviewed-by: Ben Wagner <bungeman@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Cary Clark <caryclark@skia.org>
Auto-Submit: Cary Clark <caryclark@skia.org>
Fold its only subclass (SkDefaultBitmapController) into base.
Change-Id: I77f5c9f3b459d644d5749b8ec34722ec6c2228bf
Reviewed-on: https://skia-review.googlesource.com/141541
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
* Output fiddle.json
* Do not send to CQ while we investigate
Bug: skia:8151
Change-Id: I309c732cc6b7db72054fd119b8c68dd32a765d6f
Reviewed-on: https://skia-review.googlesource.com/142301
Commit-Queue: Ravi Mistry <rmistry@google.com>
Reviewed-by: Cary Clark <caryclark@google.com>
38971fd135..6a5d98c4b5
git log 38971fd13599..6a5d98c4b528 --date=short --no-merges --format='%ad %ae %s'
2018-07-19 jiawei.shao@intel.com ES31: Implement Vertex Attrib Binding on D3D11
2018-07-18 geofflang@chromium.org Vulkan: Disable the depth/stencil states when there are no depth/stencil buffers
2018-07-18 jmadill@chromium.org Vulkan: Pass Renderer to FramebufferVk init.
2018-07-18 tobine@google.com Roll Vulkan repos fwd as of 7/17
2018-07-18 jmadill@chromium.org Vulkan: Fix size_t warning in TextureVk.
Created with:
gclient setdep -r third_party/externals/angle2@6a5d98c4b528
The AutoRoll server is located here: https://angle-skia-roll.skia.org
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=bungeman@google.com
Change-Id: Ie9bbcef80e3672de8bce85201568a69d7554b061
Reviewed-on: https://skia-review.googlesource.com/142381
Reviewed-by: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
This reverts commit 946c37057f.
Reason for revert: strict weak ordering: ((__x LT __y) && (__y LT __x)) != false
Original change's description:
> Reland "Add some optimizations to PolyUtils"
>
> This is a reland of 8bb0db3d07
>
> Original change's description:
> > Add some optimizations to PolyUtils
> >
> > * Switch inset/offset code to use a linked list rather than an array
> > * Use std::set to store active edge list for IsSimplePolygon rather than array
> > * Pre-alloc the priority queue for IsSimplePolygon
> > * When adding radial curves, expand the array all at once rather than pushing
> > one at a time.
> >
> > Bug: skia:
> > Change-Id: I692f8c29c500c41ec1d1be39d924d8a752676bf4
> > Reviewed-on: https://skia-review.googlesource.com/140787
> > Reviewed-by: Robert Phillips <robertphillips@google.com>
> > Commit-Queue: Jim Van Verth <jvanverth@google.com>
>
> Bug: skia:
> Change-Id: I3f5d42cfb941deab2b28bed020b37ce199e91d3d
> Reviewed-on: https://skia-review.googlesource.com/142200
> Reviewed-by: Robert Phillips <robertphillips@google.com>
> Commit-Queue: Jim Van Verth <jvanverth@google.com>
TBR=jvanverth@google.com,bsalomon@google.com,robertphillips@google.com
Change-Id: Ie8cdf2375613c51dedaf0d11125d6d22d88821df
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/142281
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
This reverts commit 880b9d302f.
Reason for revert: Appears to be breaking serialization.
See https://ci.chromium.org/raw/build/logs.chromium.org/skia/20180718T201711.486335899Z_000000000077bbff/+/annotations
Original change's description:
> Add kAAHairline to GpuPathRenderers
>
> With the upcoming ccpr stroking, this will no longer be the only path
> renderer that can handle hairlines.
>
> Bug: skia:
> Change-Id: Idf2543f053c8603c80fcd647ca18ffd9860a85f3
> Reviewed-on: https://skia-review.googlesource.com/142246
> Reviewed-by: Brian Osman <brianosman@google.com>
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
TBR=brianosman@google.com,csmartdalton@google.com
Change-Id: I1967c2cf6b1f1307270bf62bb04130cdf84317b7
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/142280
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Mysterious failures on the part of the bookmaker
bot corrupt the md files.
No-Try: true
Docs-Preview: https://skia.org/?cl=142173
Bug: skia:6818,skia:8151
Change-Id: Iab32e1a47d493b22cc78866f08b79399d983e139
TBR=jcgregario@google.com,rmistry@google.com
Reviewed-on: https://skia-review.googlesource.com/142173
Reviewed-by: Cary Clark <caryclark@google.com>
Commit-Queue: Cary Clark <caryclark@google.com>
Auto-Submit: Cary Clark <caryclark@google.com>
e2c00841a5..38971fd135
git log e2c00841a55f..38971fd13599 --date=short --no-merges --format='%ad %ae %s'
2018-07-18 geofflang@chromium.org Vulkan: Support multiple depth stencil formats for backbuffers.
2018-07-18 jmadill@chromium.org Replace D24_UNORM with a packed version.
Created with:
gclient setdep -r third_party/externals/angle2@38971fd13599
The AutoRoll server is located here: https://angle-skia-roll.skia.org
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=bungeman@google.com
Change-Id: Ice3b6dc487803c377cf37a93025cf6ba8aa9baa0
Reviewed-on: https://skia-review.googlesource.com/142250
Reviewed-by: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: <angle-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
With the upcoming ccpr stroking, this will no longer be the only path
renderer that can handle hairlines.
Bug: skia:
Change-Id: Idf2543f053c8603c80fcd647ca18ffd9860a85f3
Reviewed-on: https://skia-review.googlesource.com/142246
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Convert some of bookmaker to use real json instead
of rolling its own. Also check to see if all
hashes are read.
TBR=jcgregario@google.com
Docs-Preview: https://skia.org/?cl=142166
Bug: skia:8151
Change-Id: Ib35ecd69648faec3522903e0b552d37b04b73f8b
Reviewed-on: https://skia-review.googlesource.com/142166
Commit-Queue: Cary Clark <caryclark@skia.org>
Auto-Submit: Cary Clark <caryclark@skia.org>
Reviewed-by: Cary Clark <caryclark@skia.org>
We only care about the direction of the tangent vector, not the the
magnitude, so a 3x scale is irrelevant.
Bug: skia:
Change-Id: Icd2d82faf2c700fc794f8d4a59f21b32398b64f0
Reviewed-on: https://skia-review.googlesource.com/142245
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Update SkUTF16_CountUnichars to return -1 on invalid input instead of
simply asserting in debug. This also removes the unneeded checks for
zero length input. This also updates SkPaintPriv::ValidCountText and its
documentation to reflect the reality that it returns -1 on invalid data.
Bug: skia:8156
Change-Id: Ief227b7dec9d1e823b93e9061558f8412791fc09
Reviewed-on: https://skia-review.googlesource.com/142168
Reviewed-by: Hal Canary <halcanary@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
This is a reland of 8bb0db3d07
Original change's description:
> Add some optimizations to PolyUtils
>
> * Switch inset/offset code to use a linked list rather than an array
> * Use std::set to store active edge list for IsSimplePolygon rather than array
> * Pre-alloc the priority queue for IsSimplePolygon
> * When adding radial curves, expand the array all at once rather than pushing
> one at a time.
>
> Bug: skia:
> Change-Id: I692f8c29c500c41ec1d1be39d924d8a752676bf4
> Reviewed-on: https://skia-review.googlesource.com/140787
> Reviewed-by: Robert Phillips <robertphillips@google.com>
> Commit-Queue: Jim Van Verth <jvanverth@google.com>
Bug: skia:
Change-Id: I3f5d42cfb941deab2b28bed020b37ce199e91d3d
Reviewed-on: https://skia-review.googlesource.com/142200
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Bug: skia:
Change-Id: I6b3a4b3405662c89464aef7c2f106ec906992241
Reviewed-on: https://skia-review.googlesource.com/142161
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Ruiqi Mao <ruiqimao@google.com>
By default, decode images to a common color space (sRGB) before comparing.
This lets us compare images rendered to different gamuts (eg sRGB or P3),
where the raw pixel values are wildly different, even though the colors
being described are actually the same (or very similar).
Add a "--nocolorspace" option to get the existing behavior of comparing
the raw pixel values.
Also removed the source code for skimagediff, which hasn't been built in
a long time.
Change-Id: Ic55724b523348421bf21e9764d155730b94fc40a
Reviewed-on: https://skia-review.googlesource.com/141962
Auto-Submit: Brian Osman <brianosman@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Change-Id: I71f36920a11cb36c58c0c9ff1b468b77ca356c1b
Reviewed-on: https://skia-review.googlesource.com/142160
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>
This is a reland of 2267a09235
Original change's description:
> SkDEBUGF: use __VA_ARGS__
>
> Change-Id: I42a47e821ff7a7f6cec65b38a8216cabbf0acfce
> Reviewed-on: https://skia-review.googlesource.com/139860
> Reviewed-by: Mike Klein <mtklein@google.com>
> Commit-Queue: Hal Canary <halcanary@google.com>
Change-Id: Ia06567e441a414f4dcdbe5663160082f889f9fef
Reviewed-on: https://skia-review.googlesource.com/141762
Reviewed-by: Hal Canary <halcanary@google.com>
Commit-Queue: Hal Canary <halcanary@google.com>
This should take care of all the issues like:
Conditional jump or move depends on uninitialised value(s)
SkRect::join(float, float, float, float) (SkRect.cpp:119)
join (SkRect.h:1370)
GrDrawVerticesOp::GrDrawVerticesOp(...) (GrDrawVerticesOp.cpp:105)
Change-Id: I520ea55c9e310017bd8d5e2a7c9c343e3ac8eb40
Reviewed-on: https://skia-review.googlesource.com/142101
Commit-Queue: Ben Wagner <bungeman@google.com>
Auto-Submit: Robert Phillips <robertphillips@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
missed an example reference to SkTransferFunctionBehavior
TBR=caryclark@google.com
NOTRY=true
Docs-Preview: https://skia.org/?cl=142103
Bug: skia:6818
Change-Id: I966d37c0a2ad8154f81f801b3f523e899fcb9fde
Reviewed-on: https://skia-review.googlesource.com/142103
Commit-Queue: Cary Clark <caryclark@skia.org>
Auto-Submit: Cary Clark <caryclark@skia.org>
Reviewed-by: Cary Clark <caryclark@skia.org>
When two outer boundary edges are collinear, they are part of an
overlap region. As of 77169c8fd6, we (correctly) leave these edges
connected, but we must also adjust the winding to magnitude 1 to
keep the winding rules correct.
Note: this adds a new test case to the concavepaths GM, affecting all platforms.
Bug: 863389
Change-Id: I7e3a06df537cd189101e7ad39a4815a78be8fbdd
Reviewed-on: https://skia-review.googlesource.com/141952
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
