SampleBigBlur is super simple and now covered by gms and AnimBlur
sample.
SampleBigGradient now has a gm shallow_gradient_linear.
SampleConcavePaths has a more complete gm concavepaths.
SamplePoints now has an almost exact gm.
RasterAllocatorSample is made into a gm so that it will be run by d. It
appears to be the only test of SkRasterHandleAllocator, so it should
probably run.
Change-Id: Iad7b99d0f92898fc4b2fdccc5aae35d0277f2fff
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/219400
Reviewed-by: Hal Canary <halcanary@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Fixes a bug where even-odd fill rule was not being preserved on paths
that were reused from a stashed fp16 coverage count atlas.
Bug: skia:8782
Change-Id: I6698498a6f4c8df8eff10b19beb80e49663a577c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/218047
Reviewed-by: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
This patch improves the straight skeleton implementation used by
GrTessellator to collapse overlap regions in the alpha gradient border.
The resulting quality improvement can be seen in the
"thinconcavepaths" GM, for example, where the coverage values of
the "thin right angle", "thin rect and triangle" and "skinny snake"
are now lower and match the raster backend more closely. It also improves
correctness, such as on the linked Chromium bugs below.
Previously, the straight skeleton was performed using the same Vertex
and Edge classes used for tessellation, but this led to fragility in
maintaining the connectivity and ordering required by those classes.
Instead, that functionality has been moved to new SSEdge and SSVertex
classes. Their construction results in alternating SSVertex and
SSEdges around the boundary of the each overlap region, shrunk by one
SSEdge as the boundary collapses.
Applying events may now also create further events (chained events),
as intersections between newly-adjacent bisectors change the structure
of the skeleton. This is always calculated from the bisectors of
original boundary, not the shrunk boundary of the straight skeleton,
which is why each SSEdge points at its originating Edge. If the edges
are parallel or near-parallel, the bisector may be infinite. This is
handled by a new flavour of create_event(), which does a Line/Line
intersection (rather than an Edge/Edge intersection) to find the
intersection between the infinite bisector and an adjacent bisector.
Several ancillary bugs were fixed: the priority queue used to represent
edge events was sorting the inner edges incorrectly. These need to be
sorted in descending not ascending order of coverage. Its implementation
was moved from Skia's TDPQueue to std::priority_queue(), which is more
flexible in specifying a comparator.
check_for_intersection()'s partner synthesis code was moved into a new
function, compute_bisector(), also used by the chained skeleton events
code.
Degenerate edges are now removed during the simplify_boundary() pass.
They were previously detected but ignored, causing incorrect inner and
outer tangents to be computed in stroke_boundary().
An fSynthetic flag was added to Vertex, in order to detect vertices
which cannot be moved by an edge collapse event (e.g., intersections
with non-boundary edges, merged vertices, straight skeleton vertices).
More raw implementation notes:
Connect straight skeleton vertices as we find them, so we don't
have to use partnering (ss_connect()).
Only disconnect edges which are still alive after event application.
Add a check for near-parallel lines in compute_bisector().
Don't use edge type to determine direction to offset for bisectors.
The winding should already include this information.
Move Event ownership to SSEdge.
If we're down to the last two edges in a skeleton, don't check for events.
Add concave_arc_and_circle GM.
Add a collapsepaths GM.
Bug: 941429, 913349, 929915, 945853
Change-Id: Ib89e231d0e8611f8735fd3592db6391da096369d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/215094
Commit-Queue: Stephen White <senorblanco@chromium.org>
Reviewed-by: Robert Phillips <robertphillips@google.com>
This is implemented at backend-neutral level and so misses some
opportunities to reduce the number of passes in the GPU backend.
Filter quality is interpreted as:
none - single nearest neighbor resampling
low - chain of bilinear resamplings. 2x up/down except for one
step which may be smaller than 2x.
medium - same as low
high - when both scale factors are up then same as low but with bicubic
filtering rather than linear. Otherwise, same as low.
Bug: skia:8962
Change-Id: I4467636c14b802d6a0d9b5c363c1ad9e87a1a44b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/213831
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
This reverts commit 360dc593c0.
Reason for revert: Doesn't reproduce issue.
Original change's description:
> Large circle clipping gm.
>
> Seeing if this reproduces a reported problem on PowerVR devices.
>
> Bug: b/123437630
>
> Change-Id: I161cd04034cb7d217ff519a4b26e521bf36ed4b5
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/212727
> Reviewed-by: Brian Osman <brianosman@google.com>
> Commit-Queue: Brian Salomon <bsalomon@google.com>
TBR=bsalomon@google.com,brianosman@google.com
Change-Id: Ida6225513c24ba68d4744e7871cce38555dddb61
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: b/123437630
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/212964
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Seeing if this reproduces a reported problem on PowerVR devices.
Bug: b/123437630
Change-Id: I161cd04034cb7d217ff519a4b26e521bf36ed4b5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/212727
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Bug: chromium:955467
Change-Id: I1c4cae2499db926aa6b629e8d730fcc8cb45be6d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/212030
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Also fix the corners of clamp mode with alpha.
Also, add a GM.
Bug:chromium:957275
Change-Id: Icd288ff522e7ea70662380791f5ee2de628a5ef2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/211594
Auto-Submit: Hal Canary <halcanary@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Bug: skia:9028
Change-Id: I8e3d37050d3fce7602eee62ae911eae756e603a2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/211100
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
There is still a bit of manual mucking about with iwyu output to get
things nice, but the checker seems to be doing ok and the process is now
a bit easier. Will see how it goes.
This also pointed out the amount of code behind ifdefs should be
minimized by using the build system and 'constexpr if' when possible.
Change-Id: Ic63fa33c65e5ff40b58858e15fc51f27d862e20d
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/211349
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Bug: chromium:946965
Change-Id: Ia85e07d7fc76aa9e0b0a8fe0daf3ab80d517cbd9
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/209167
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
If all glyphs were culled (by being too small), we could end up with a
run that never called setHasBitmap. That would prevent us from looking
at the view matrix on a subsequent draw, when deciding if we needed to
regenerate.
Bug: skia:8955
Change-Id: Ic7a2539762527f91bdb50ab78bdf5801bfda0034
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/206266
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Bug: chromium:947055
Change-Id: If271112285aa413a71c094502c81b501c77a129e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/204742
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Auto-Submit: Michael Ludwig <michaelludwig@google.com>
These GMs are confusing to triage, and are checking behavior that I
don't think we care about any longer.
Bug: skia:6652
Change-Id: I331f9a51623a0e90d4a848c8209be93403bc90ee
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/202128
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This was another stop-gap color management "solution".
Bug: skia:
Change-Id: I7c0c362840dd35aad51ad8780f2dab591c42a7e2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/199720
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Bug: chromium:938592
Change-Id: I317c13c6f81c54989267325061bcb1c57428f478
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/200043
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Bug: skia:
Change-Id: I1f9e9013d20496554891b72749ed9b0844747566
Reviewed-on: https://skia-review.googlesource.com/c/191570
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Brings back GL ETC1 support and adds Vulkan support as well.
Bug: skia:8684
Change-Id: Ie65da0a3172793081f0e4072f161bfb9b14678bc
Reviewed-on: https://skia-review.googlesource.com/c/179724
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Bug: skia:8664
Change-Id: Ic1b4f5afb2847eabb2290cbba808c5692a876972
Reviewed-on: https://skia-review.googlesource.com/c/182152
Auto-Submit: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
There was an issue on Android where a client was drawing a small bit of text,
and then a large chunk of text below it. On the next frame they started fading
out the text and then swapped out all of the large chunk of text. We were not
marking the Plot containing the small bit of text as still in use in this case,
so it was getting overwritten by later uploads. That bug has been fixed,
but this GM tests that codepath.
Bug: b/118850208
Change-Id: I81b795d6b0ee5d5d0b8e380823a568a52118ed0c
Reviewed-on: https://skia-review.googlesource.com/c/178924
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
A stroked arc with round caps was batched with a filled circle. The circle op
code would choose a GeometryProcessor configuration that expected round cap
centers as vertex attributes. However, the tessellation code for the filled
circle would not put in dummy round cap centers and then didn't advance the
pointer into which vertex data was being written by the expected vertex
stride.
Bug: b/119394958
Change-Id: I6fe95b32d750599e775ed96e656757fe3087795a
Reviewed-on: https://skia-review.googlesource.com/c/177881
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Bug: skia:6886
Change-Id: Ic184ef7e5dd76615f626581b897a3606c6241aa7
Reviewed-on: https://skia-review.googlesource.com/c/177001
Reviewed-by: Yuqian Li <liyuqian@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Auto-Submit: Mike Klein <mtklein@google.com>
Added a GM that demonstrates the bug. Should draw a blur with
the center masked out, and a circular blurry shape that's
roughly the inverse. On raster, this was already the result.
On GPU, the blurred/eroded layer becomes a subset with an
origin other than (0,0), and that layer was shifted.
I *think* this is the correct fix - we are including 'offset'
in the texture matrices, but that's just based on the crop
rects and adjustments from each filter. We still need to adjust
the texture coords for the subsets themselves.
Bug: chromium:905548
Change-Id: I19c936adad90311aef243a9395a270d2e015df2f
Reviewed-on: https://skia-review.googlesource.com/c/173321
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
This reverts commit 1a2476d294.
Reason for revert: Fixes printf signatures and asserts.
Original change's description:
> Revert "Initial definition of fill rect op"
>
> This reverts commit d3c92d9a36.
>
> Reason for revert: printf build failure on gcc, assert failures on CQ
>
> Original change's description:
> > Initial definition of fill rect op
> >
> > Bug: skia:
> > Change-Id: Ie0c99eb5163501853d1adc885bd3841f90a71924
> > Reviewed-on: https://skia-review.googlesource.com/c/163486
> > Reviewed-by: Brian Salomon <bsalomon@google.com>
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
>
> TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com
>
> Change-Id: Ib32f91a39d91aeb87982a7b19719485e4a1bf8ae
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: skia:
> Reviewed-on: https://skia-review.googlesource.com/c/173233
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com
Change-Id: I415913a269ba5bcdebd169b5ebc3510673247bfd
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/173234
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
This reverts commit d3c92d9a36.
Reason for revert: <INSERT REASONING HERE>
Original change's description:
> Initial definition of fill rect op
>
> Bug: skia:
> Change-Id: Ie0c99eb5163501853d1adc885bd3841f90a71924
> Reviewed-on: https://skia-review.googlesource.com/c/163486
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
TBR=bsalomon@google.com,csmartdalton@google.com,michaelludwig@google.com
Change-Id: Ib32f91a39d91aeb87982a7b19719485e4a1bf8ae
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:
Reviewed-on: https://skia-review.googlesource.com/c/173233
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Bug: skia:
Change-Id: Ie0c99eb5163501853d1adc885bd3841f90a71924
Reviewed-on: https://skia-review.googlesource.com/c/163486
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
These aren't testing Skia core functionality
Bug: skia:7518
Change-Id: Ib564dc7b0aa8f137c2c40141fa5d7e9a1bfe4d64
Reviewed-on: https://skia-review.googlesource.com/c/172968
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This was based on old ideas about color management. We have better
mechanisms for testing this now.
Bug: skia:
Change-Id: If59b5039f31ab0ebbdbed4205c941dd9266f67c1
Reviewed-on: https://skia-review.googlesource.com/c/172860
Commit-Queue: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Auto-Submit: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Fixes an issue where negative scales caused an inset rather than an outset.
Add GM.
Bug: chromium:899512
Change-Id: I9164c76da479af80d4f5389b057ec52a946726fb
Reviewed-on: https://skia-review.googlesource.com/c/171641
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Adds fwidth() to SkSL, and adds a gm that draws an AA squircle to test
it.
Bug: skia:
Change-Id: Ida306cc535a1d4b4568d0ad5cc9a5f235098f4e8
Reviewed-on: https://skia-review.googlesource.com/c/170726
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
This reverts commit 95af4726bf.
Reason for revert: I think this may not have been the reason the Android roll was failing. We've rolled, so it's a good time to try again.
Original change's description:
> Revert "Improve degenerate 2pt conical gradient cases"
>
> This reverts commit 879dab87ab.
>
> Reason for revert: Android roll failed.
> https://sponge.corp.google.com/target?id=93bc6b8d-9b42-4805-b204-46ae62f1b005&target=x86+CtsGraphicsTestCases&searchFor=&show=FAILED&sortBy=STATUS
> A test VectorDrawableTest.testVectorDrawableGradient fails.
>
> Original change's description:
> > Improve degenerate 2pt conical gradient cases
> >
> > 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>
>
> TBR=caryclark@google.com,fmalita@chromium.org,fmalita@google.com,reed@google.com,michaelludwig@google.com
>
> Change-Id: I91b896c4a438c02206679b327a01b47f40993965
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: oss-fuzz:10373
> Reviewed-on: https://skia-review.googlesource.com/c/170272
> Reviewed-by: Stan Iliev <stani@google.com>
> Commit-Queue: Stan Iliev <stani@google.com>
TBR=caryclark@google.com,fmalita@chromium.org,fmalita@google.com,reed@google.com,stani@google.com,michaelludwig@google.com
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: oss-fuzz:10373
Change-Id: I7577fcea9eb8a875e94723ab2cca2fcc990b82b2
Reviewed-on: https://skia-review.googlesource.com/c/170279
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This reverts commit 879dab87ab.
Reason for revert: Android roll failed.
https://sponge.corp.google.com/target?id=93bc6b8d-9b42-4805-b204-46ae62f1b005&target=x86+CtsGraphicsTestCases&searchFor=&show=FAILED&sortBy=STATUS
A test VectorDrawableTest.testVectorDrawableGradient fails.
Original change's description:
> Improve degenerate 2pt conical gradient cases
>
> 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>
TBR=caryclark@google.com,fmalita@chromium.org,fmalita@google.com,reed@google.com,michaelludwig@google.com
Change-Id: I91b896c4a438c02206679b327a01b47f40993965
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: oss-fuzz:10373
Reviewed-on: https://skia-review.googlesource.com/c/170272
Reviewed-by: Stan Iliev <stani@google.com>
Commit-Queue: Stan Iliev <stani@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>
Change-Id: I3126fb8b055b58e45f1bd0d913413b4d4d38f032
Reviewed-on: https://skia-review.googlesource.com/c/166740
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
This reverts commit 0a0304c426.
Reason for revert: Breaking metal
Original change's description:
> Add experimental API to draw a set of SkImages in one SkCanvas call.
>
> The client provides a src and dst rect for each image as well as
> a bitfield that indicates whether each edge of the image should be
> antialiased. This per-edge AA is useful for tiled compositors.
>
> Rather than take a full SkPaint this API only takes an alpha, a filter
> quality (which is pinned to kLow), and a blend mode. This is a likely
> point of future evolution.
>
> Currently the API is only fully implemented for kSrcOver on the GPU
> backend. With other blend modes or on other backends AA will be ignored
> for images that do not have all four edge AA flags set.
>
> BUG: skia:8444
>
> Change-Id: I143998dda8ad6a25f64e18cd600392ba553030ac
> Reviewed-on: https://skia-review.googlesource.com/c/159062
> Commit-Queue: Brian Salomon <bsalomon@google.com>
> Reviewed-by: Mike Klein <mtklein@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>
TBR=mtklein@google.com,bsalomon@google.com,brianosman@google.com,reed@google.com
Change-Id: I815baaeee5de9c6722cf2b9d071a8e2f7c1b6a96
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/161622
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
The client provides a src and dst rect for each image as well as
a bitfield that indicates whether each edge of the image should be
antialiased. This per-edge AA is useful for tiled compositors.
Rather than take a full SkPaint this API only takes an alpha, a filter
quality (which is pinned to kLow), and a blend mode. This is a likely
point of future evolution.
Currently the API is only fully implemented for kSrcOver on the GPU
backend. With other blend modes or on other backends AA will be ignored
for images that do not have all four edge AA flags set.
BUG: skia:8444
Change-Id: I143998dda8ad6a25f64e18cd600392ba553030ac
Reviewed-on: https://skia-review.googlesource.com/c/159062
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
This is a reland of f065907ccc
3rd time's the charm:
The new analytic gradient shader was sporadically triggering violations of the coverage as alpha
compatibility optimization. Unfortunately, even when using the same device and random seed for the
test, the bots did not always reproduce the error. However, we identified the likely cause of the
violation.
The test requires that all output channels are less than the input alpha, which it uses to validate
whether or not the shader is modulating its values by the input alpha. This test does not pass if
the RGB values are greater than 1. The original version of the analytic gradient shader used half4s
for its scale and bias values. Given the threshold limit for hardstops of 0.00024 (SkNearlyZero),
a very small interval that is not treated as a hardstop can create a scale or bias of over 4000.
This moves into the very imprecise region of 16-bit floats, making it plausible that the gradient
outputs colors greater than 1, due to rounding. The kicker is that the random test generation for
stop locations does not use a uniform distribution, but is instead biased towards the remaining
interval, which increases the likelihood of generating a small interval that is not treated as a
hard stop. We are keeping this behavior since ill-conditioned gradients are useful in testing.
Original change's description:
> Reland "Implement an explicit binary search-based analytic gradient colorizer"
>
> This reverts commit 9461dcf130.
>
> Reason for revert: Fixes for ANGLE's incorrect shader behavior
>
> Original change's description:
> > Revert "Implement an explicit binary search-based analytic gradient colorizer"
> >
> > This reverts commit dcc85fc610.
> >
> > Reason for revert: ANGLE is frequently corrupted, particularly radial_gradient4 and mixershader
> >
> > Original change's description:
> > > Implement an explicit binary search-based analytic gradient colorizer
> > >
> > > Provides a reasonably flexible fragment processor that defines another
> > > colorizer implementation for gradients. It can support up to 8
> > > interpolation intervals (which is 16 colors if every stop is a hard stop
> > > or 9 colors if every stop is a smooth transition). It
> > > supports mixtures of hard and smooth stops. It is conditionally compiled
> > > into versions specific to the interval count (so it can produce up to
> > > 8 shader variants).
> > >
> > > The GrGradientShader controller does not remove the single and dual
> > > interval colorizers, which are useful specializations of this explicit
> > > binary search colorizer. Similarly, since it can only handle up to 8
> > > intervals, the texture colorizer is still used as a fallback.
> > >
> > > Currently it does not employ capabilities detection to determine if the
> > > hardware can support the number of required uniforms, which can become
> > > substantial for the larger gradient configurations.
> > >
> > > Bug: chromium:796479, chromium:729727, chromium:696603, chromium:543625, chromium:414254
> > > Change-Id: Ia1f735a5019766ae4796cc22964b2913db34b95b
> > > Reviewed-on: https://skia-review.googlesource.com/155080
> > > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
> > > Reviewed-by: Brian Osman <brianosman@google.com>
> >
> > TBR=bsalomon@google.com,brianosman@google.com,michaelludwig@google.com
> >
> > Change-Id: I351a387f0528e4c2db2d47ab2e5d6b336991fb98
> > No-Presubmit: true
> > No-Tree-Checks: true
> > No-Try: true
> > Bug: chromium:796479, chromium:729727, chromium:696603, chromium:543625, chromium:414254
> > Reviewed-on: https://skia-review.googlesource.com/156541
> > Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> > Commit-Queue: Michael Ludwig <michaelludwig@google.com>
>
> TBR=bsalomon@google.com,brianosman@google.com,michaelludwig@google.com
>
> Change-Id: I2aca36307d88c26905d860ec29417ec68c6037cc
> Bug: chromium:796479, chromium:729727, chromium:696603, chromium:543625, chromium:414254
> Reviewed-on: https://skia-review.googlesource.com/156542
> Reviewed-by: Michael Ludwig <michaelludwig@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Bug: chromium:796479, chromium:729727, chromium:696603, chromium:543625, chromium:414254
Change-Id: I2d050624781c77cdd160291cadbadac602b48bde
Reviewed-on: https://skia-review.googlesource.com/c/157569
Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Bug: skia:7903
Change-Id: I2399e4aaae6b91914d637ff0e921e04596b52836
Reviewed-on: https://skia-review.googlesource.com/159141
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
