This adds a way to test/demonstrate how an external client
using Typescript can make use of CanvasKit in a browser
(currently a JS library with Typescript bindings supplied).
See: modules/canvaskit/external_test/typescript_browser/
This also adds a way that we hope CanvasKit might be used
in a future release, that is, via ES6 modules.
See: modules/canvaskit/external_test/typescript_browser_es6/
These TS files can be built into the JS files needed by the
respective index.html files using the appropriate target
in modules/canvaskit/external_test/Makefile. Then, `make serve`
will bring up a HTTP Server that will make the folders
accessible on localhost:8000
Change-Id: Ie4ddc2588b4bdccd4a727f11b540289cf4f85795
Bug: skia:11077, skia:12539
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468214
Reviewed-by: Leandro Lovisolo <lovisolo@google.com>
When the device crashes we get no data from the run so skipping tests is preferable.
Bug: skia:12617
Change-Id: I3f69cb91cdf4e1e9659c81f5a6c717e0d911c2fa
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468278
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This will be used to populate a trace buffer for the SkSL debugger.
See http://go/sksl-tracing for details and rationale.
Change-Id: I4c218c65ff01c339cf460e97e41566860a694720
Bug: skia:12614
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468436
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Also updates MtlRenderPipeline to be MtlGraphicsPipeline to match the
type it extends, although we could choose to have it stay matched with
the backend object (MTLRenderPipeline) that it wraps instead.
Bug: skia:12466
Change-Id: Ida118e68a93d737d21edca15a59f5e64e17b5fd0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467780
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Change-Id: I937e0594b8fa3a1cc5ac2e2b80d1f0f406772817
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468338
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Heather Miller <hcm@google.com>
Also removes the path-specific functions now that it only maps points.
Bug: skia:12524
Change-Id: I63c0707b8b7814c93f472c33f108319774192cbd
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468038
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Change-Id: Ic110ea129cf902d8d1a87c6133e71a1fdeb366f6
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468376
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Julia Lavrova <jlavrova@google.com>
This reverts commit b61804e94c.
Reason for revert: chromium perf regression
Original change's description:
> Add ConvertPixels versions of PremulAlphaRoundTrip
>
> Prior to the force_highp trick, the GrConvertPixels version failed, just
> like the GPU would do if we disabled the canvas2D fast path. With the
> highp trick, all tests pass.
>
> Bug: skia:12592
> Change-Id: I63ad2fd3b67863b6a736316e7c7b3b9bd2ee8970
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467516
> Commit-Queue: Brian Osman <brianosman@google.com>
> Reviewed-by: Brian Salomon <bsalomon@google.com>
Bug: skia:12592
Bug: chromium:1267108
Change-Id: Ic2b9d6cc029e1519be928615585a0b22c7037b56
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468276
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Commit-Queue: Brian Osman <brianosman@google.com>
When this is enabled, the SkVM code generator will emit trace
instructions for debugging purposes. (The trace instructions are a work
in progress, so at present, the flag doesn't do anything meaningful.)
Change-Id: Ia7d66840d915b1a7e531a3069e641c840bb9c0eb
Bug: skia:12614
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467764
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
... all because you can't forward declare an inner class.
Change-Id: I62685ca2482caa636ae54c622d6b358f37159339
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467763
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
This is a reland of 74105c5d09
Original change's description:
> Remove GPU-related calls from SkSurface.h when !SK_SUPPORT_GPU
>
> Change-Id: Idca02c40bd8f540919702f09ba2a809acc377e67
> Bug: skia:12584
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464295
> Reviewed-by: Brian Osman <brianosman@google.com>
Bug: skia:12584
Change-Id: Id679bd61eddb341598e149a7a87e3ba9f0dc8943
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464976
Reviewed-by: Brian Osman <brianosman@google.com>
It doesn't allow VAO 0.
Bug: b/197500792
Change-Id: Ife82553a2a81bbe02299693b0706d87b8032abc0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/465556
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
This is the pre-iterator behavior of MiddleOutPolygonTriangulator.
It's slower, but may lead to better triangulations.
We may or may not keep it, depending on Perf's opinion.
Bug: skia:12620
Change-Id: I46f39b551b32af4eebfe8221cbb48a4332db83a8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468096
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Moves the earlier, non-recursive chopping code from the wedge
tessellator to a shared location and shares it with the curve
tessellator. Creates a 'PatchAttribs' enum to control which data gets
written out with tessellation patches, including a 'kFanPoint' for
wedges.
Bug: skia:12524
Change-Id: I79b84bf603e80aa1e8732939c60a53c3604cd679
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/468016
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
SkTHashTable::Slot required a default constructor on T, even though the
default constructed T was never used. Model SkTHashTable::Slot on
std::optional to not require the default constructor.
Change-Id: Ie8912f132ba0f375c80a73fe7ae9022762b4794a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467083
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
This effect does not yet support pinned edges or taper
also sneak in a filename change (CCToner) to keep all effect names standard
Change-Id: I17f2b5463408556775bc12a972358abd4d8d8690
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467319
Commit-Queue: Jorge Betancourt <jmbetancourt@google.com>
Reviewed-by: Florin Malita <fmalita@google.com>
Use SkVMBlitter::DebugName for the string. This is easier to read, and
has some actually useful information about the blitter.
Added magenta highlight of the hovered element, similar to the GPU.
Change-Id: Ic9b5a0f61e092c8aa555f375d5d2f2de22cc45fe
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467977
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
It's vestigial.
Change-Id: I63791bd78e565faf203eefe1537699e924446d85
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467978
Commit-Queue: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Much like http://review.skia.org/467759, this CL defensively guards
against programs which consume more space than is reasonable. Globals
exist outside of functions, so they wouldn't be caught by the stack size
checks.
Change-Id: I035f27d57bc329508820a729a1e367ecaadfe156
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467760
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Functions that declare variables totaling more than 100,000 slots will
now generate an error.
This is only a partial mitigation to the problem, as a sophisticated
attack could still chain/nest multiple functions together to consume
extremely large amounts of stack. However, this mitigation is still more
sophisticated than our peers; both WebGL and glslang are susceptible to
similar problems, and in the general case (ES3+ with full flow control)
it's intractable.
Change-Id: I153c75267c017a23f59fe9e59f6e391197ee6101
Bug: oss-fuzz:40304, oss-fuzz:40694
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467759
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Bug: skia:12466
Change-Id: Icd23abfa9ab1197a03e5d0f177747ea77b6b77e8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/465391
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Caveats:
- You want to run viewer with `--skvm` to see much/anything
- The cache is LRU, and doesn't get flushed automatically. Hitting
'Clear' will flush it, so it will pick up only the blitters used
on the current slide. (Otherwise they accumulate as you navigate).
- No way to determine which blitter is for which primitive, yet.
I'd like to do the GPU-style magenta highlight, but that may be
tricky, because we need to preserve the semantics of the original
blitter (including destination color type, most importantly).
Change-Id: I2df763fdb697d87471ca0816a3b7087ffb4fc4e4
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467783
Reviewed-by: Herb Derby <herb@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
The skstd::optional already does this correctly, but it appears to be
one of the guarantees of this class which isn't tested.
Change-Id: Iedd9004496f10c1ec56da4ebc5e246880c9a457f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467781
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
COLR fonts can use a magic palette index value 0xFFFF to
indicate the foreground color of the drawing context. Support this by
defining a flag plus saving the foreground palette value on the
SkScalerContextRec when creating the SkStrikeCache entry. Then the
FreeType backend can use the palette value when performing the COLR
drawing runs.
Fixed: skia:12576
Change-Id: I573c67482d9c53dade04eba60449e8ed1b8528fd
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463656
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Making the strike-to-source scale part of the SkStrikeSpec needs
to be handled more carefully now that the "SkMatrix is always the
same" invariant is broken.
Start with MakeCanonicalized focusing mainly on SkFont.
Change-Id: I8e3ab6120751ce280fea03b87be200155d736130
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467080
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>
When we failed to commit, we were keeping the old command buffer around
which could lead to issues if we tried to use it later.
Change-Id: I7397735931c49d5383490bbc213059bd42d2ea84
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467777
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Bug: b/197827241
Change-Id: I39cda3f3dd46f9a340c319aaf9463253db2a7705
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467357
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Prior to the force_highp trick, the GrConvertPixels version failed, just
like the GPU would do if we disabled the canvas2D fast path. With the
highp trick, all tests pass.
Bug: skia:12592
Change-Id: I63ad2fd3b67863b6a736316e7c7b3b9bd2ee8970
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/467516
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
