The GP will likely require this value when dealing with a geometry
shader. In the future we may wish to either switch to device-space
geometry shaders, or else put this value in an "sk_" builtin.
BUG=skia:
Change-Id: I8dff88fc219feef84d39fb7bbd08f3b5686f53d2
Reviewed-on: https://skia-review.googlesource.com/8362
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
When sanitizing contours, if the first and last vertices coincide,
continue with the previous vertex, not the next vertex, since we
may otherwise exit prematurely. Also, round the last vertex before
entering the loop, just in case it coincides with the first.
Add a test case to exercise the above, and another one which exercises
the intruding-vertex workaround.
BUG=691593
Change-Id: Ic28a9308a21164d185edef0ee6fbc29b40742149
Reviewed-on: https://skia-review.googlesource.com/8364
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
One more piece of https://skia-review.googlesource.com/c/8230/.
-mno-red-zone makes it safe for x86 stages to use the stack on Windows
(at the expense of an extra sub and add to the stack pointer on !Windows).
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug,Test-Win10-MSVC-Golo-GPU-GT610-x86_64-Release,Test-Mac-Clang-MacMini6.2-CPU-AVX-x86_64-Debug
Change-Id: I81f8220e790b201757a7e1e9752b2fe94520ccbb
Reviewed-on: https://skia-review.googlesource.com/8352
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
This reverts commit ff590a1244.
Reason for revert: This breaks the android roll because they are using
a private call. Updating android tests to use new api.
Original change's description:
> Always make SkImageShaders in heap.
>
> I made a couple of measurments, and it looks like any differences is
> well below the noise threshold.
>
> Just for the record run1: .9991 of baseline and run2 .9988 of baseline.
> I was using top25 .skps as workload.
>
> TBR=mtklein@google.com
>
> Change-Id: If4fa06e5d5df72fb67dbb4bbb99c926f05765897
> Reviewed-on: https://skia-review.googlesource.com/8341
> Reviewed-by: Herb Derby <herb@google.com>
> Commit-Queue: Herb Derby <herb@google.com>
>
TBR=mtklein@chromium.org,mtklein@google.com,herb@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Ibdaafc796702e250933b62e5f4abb5e2ce8d40c0
Reviewed-on: https://skia-review.googlesource.com/8393
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Herb Derby <herb@google.com>
This should land as a no-op, as nothing using the stack yet.
It will allow us to land the SSE2 code, which does spill in to_srgb.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug,Test-Win10-MSVC-Golo-GPU-GT610-x86_64-Release
Change-Id: Ie8000d50a85d56ad6229736d79f3a947f06a808c
Reviewed-on: https://skia-review.googlesource.com/8389
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
Change-Id: I0ab76132b9e21544ed3dfb87bd7adc91c4c4e656
Reviewed-on: https://skia-review.googlesource.com/8387
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Hal Canary <halcanary@google.com>
Also: SkPDFShader::State isi now zero-initilized.
No change in PDF tests.
BUG=chromium:690875
Change-Id: Ibc56cc9435362733adf50cbb51b11c9413572e7f
Reviewed-on: https://skia-review.googlesource.com/8355
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Hal Canary <halcanary@google.com>
Fixes out of range colors produced by matrix convolution and dither effects. Adds modulate optimization to matrix convolution.
Change-Id: I8424250a52e864f4b5feaf4474293695c26039d8
Reviewed-on: https://skia-review.googlesource.com/8351
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
I made a couple of measurments, and it looks like any differences is
well below the noise threshold.
Just for the record run1: .9991 of baseline and run2 .9988 of baseline.
I was using top25 .skps as workload.
TBR=mtklein@google.com
Change-Id: If4fa06e5d5df72fb67dbb4bbb99c926f05765897
Reviewed-on: https://skia-review.googlesource.com/8341
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Also fix such that the Nexus Players get kicked three times
not one and a half times.
BUG=skia:6045
Change-Id: Ie9142f809a979a46b17a8ad9a205279d6fcb9bd5
Reviewed-on: https://skia-review.googlesource.com/8266
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Change-Id: I960ded854e6bc7cdee029a7393cac2a686c41754
Reviewed-on: https://skia-review.googlesource.com/8308
Commit-Queue: Hal Canary <halcanary@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Ensure the data in the RGBA texture is premul.
Reenable the "modulation" optimization flag on GrAlphaThresholdFragmentProcessor.
Change-Id: I49ef215cee83ea74526c589bcee4bc215bc28ba4
Reviewed-on: https://skia-review.googlesource.com/8323
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Made it a separate ImGui window (rather than part of the debug window).
Bring it up with 'z'. Draggable/resizable. Variable zoom scale. Enjoy.
BUG=skia:
Change-Id: I949ab398126c892c8d353aaebcc8403765f42841
Reviewed-on: https://skia-review.googlesource.com/8357
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Before this, fNext always pointed at the next (prev) saveLayer in the stack.
Typically each layer is isolated, as it is defined to "consume" the current clip.
Android has an option to not affect/consume the clip, hence the support for looping
through this link-list in updateMC().
The current code always executes this loop, subtracting the current layer's clip from
the global clip, so typically the 2nd iteration will see an empty remaining clip and
draw nothing ... but we still pay for the subtract and the draw-overhead.
This change makes fNext point to the next layer ONLY if the current layer was marked
as non-consuming.
As a side-effect, this change also now detects the "last" restore by looking for
fMCRec == null, rather than fNext == nullptr. fMCRec was always null on the last
layer, so this change is safe, and could have landed independently.
BUG=skia:6214
Change-Id: I787574fa35c4869d3b884054aece925f457ad5bd
Reviewed-on: https://skia-review.googlesource.com/8348
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Mike Reed <reed@google.com>
This is another slice of https://skia-review.googlesource.com/c/8230/.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug,Test-Win10-MSVC-Golo-GPU-GT610-x86_64-Release
Change-Id: I9a95b319c9695874c975d17d230f1569a8996f83
Reviewed-on: https://skia-review.googlesource.com/8350
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This is one little slice of https://skia-review.googlesource.com/c/8230/.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug,Test-Win10-MSVC-Golo-GPU-GT610-x86_64-Release
Change-Id: I3a03246d28f34bf7692d4cb79425510c78a7f8c1
Reviewed-on: https://skia-review.googlesource.com/8349
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This reverts commit 1fe55dc9fa.
Reason for revert: breaking Win GPU bots?
Original change's description:
> Generate sse2/sse4.1 splices, use them.
>
> While we're at it, tidy up build_stages.py a bit.
> Redirecting stdout seems a lot easier than print >>f all over the place.
>
> TODO: non-VEX-encoded before_loop() and after_loop()
>
> CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug
>
> Change-Id: I3f38e55f081670dd598c6050435466d9f394e5be
> Reviewed-on: https://skia-review.googlesource.com/8230
> Commit-Queue: Mike Klein <mtklein@chromium.org>
> Reviewed-by: Herb Derby <herb@google.com>
>
TBR=mtklein@chromium.org,herb@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug
Change-Id: Iba1905c54cb2dc42a880b9e6a8093835ffd530a2
Reviewed-on: https://skia-review.googlesource.com/8347
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>
This reverts commit 2aeae78a2c.
Reason for revert: It looks like this is breaking the android roll
out/target/product/angler/obj/SHARED_LIBRARIES/libandroid_runtime_intermediates/android_graphics_Canvas.o frameworks/base/core/jni/android_graphics_Canvas.cpp"
frameworks/base/core/jni/android_graphics_Canvas.cpp:178:15: error: incomplete type 'SkRegion' named in nested name specifier
Original change's description:
> IWYU
>
> BUG=skia:
>
> Change-Id: Ib6b4d52841dbe3fa69a86ddb6b97d6a5d0f004ee
> Reviewed-on: https://skia-review.googlesource.com/8231
> Reviewed-by: Florin Malita <fmalita@chromium.org>
> Commit-Queue: Mike Reed <reed@google.com>
>
TBR=fmalita@chromium.org,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Change-Id: If979c3dd9dc3fe08ac450ced113b3d1e9a86f02a
Reviewed-on: https://skia-review.googlesource.com/8346
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
This reverts commit 368af4605d.
Reason for revert: checking to see if this is breaking the Chrome DEPS roll
Original change's description:
> More non-invertible shader local matrix guards
>
> Change-Id: Ida9a300420ff1d883e617cdaadb80e88c99ad226
> Reviewed-on: https://skia-review.googlesource.com/8333
> Reviewed-by: Mike Reed <reed@google.com>
> Commit-Queue: Florin Malita <fmalita@chromium.org>
>
TBR=brianosman@google.com,fmalita@chromium.org,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Id0a771342b9be27c85d91f511d814297b8c3e0b8
Reviewed-on: https://skia-review.googlesource.com/8345
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
BUG=skia:
Change-Id: I3a41b5affdfad830cea969a1e4932794d27f4f7a
Reviewed-on: https://skia-review.googlesource.com/8331
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>
BUG=skia:
Change-Id: I7f467c094fb0bcb983bd86d07cb9bd7be34666b3
Reviewed-on: https://skia-review.googlesource.com/8332
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
This reverts commit 6e83b13c22.
Reason for revert: Fractional path sizes are causing asserts on the bots.
Original change's description:
> Use SDF path miplevels based on the original path's size.
>
> Should produce sharper results than arbitrary fixed sizes.
> Adds a new test to pathfill GM.
>
> BUG=chromium:682918
>
> Change-Id: I5a394098665d01e995a244fde278236f1471e6c9
> Reviewed-on: https://skia-review.googlesource.com/8328
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Jim Van Verth <jvanverth@google.com>
>
TBR=jvanverth@google.com,bsalomon@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:682918
Change-Id: I4a52df808ef3f769d0e6f75785148d46936a6747
Reviewed-on: https://skia-review.googlesource.com/8342
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Should produce sharper results than arbitrary fixed sizes.
Adds a new test to pathfill GM.
BUG=chromium:682918
Change-Id: I5a394098665d01e995a244fde278236f1471e6c9
Reviewed-on: https://skia-review.googlesource.com/8328
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
While we're at it, tidy up build_stages.py a bit.
Redirecting stdout seems a lot easier than print >>f all over the place.
TODO: non-VEX-encoded before_loop() and after_loop()
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Win2k8-MSVC-GCE-CPU-AVX2-x86_64-Debug
Change-Id: I3f38e55f081670dd598c6050435466d9f394e5be
Reviewed-on: https://skia-review.googlesource.com/8230
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
Currently Skia tries hard to use symmetric rendering with DirectWrite
as often as possible. However, particularly on Windows 7 with CJK fonts,
thin horizontal strokes can be rendered without sufficient contrast
because the font was relying on the 6x1 oversampling for what is
effecitvely drop-out control. This change will only allow symmetric
rendering if the font allows it in the 'gasp' table.
BUG=chromium:645055
Change-Id: I45a9d5e4a0b49bb969c44fb20dc92528dfe9c48d
Reviewed-on: https://skia-review.googlesource.com/8268
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Ben Wagner <bungeman@google.com>
This doesn't yet delete the old virtuals.
It still uses the color and component flags model for the pipeline input and blended output but I'm planning to change those as well.
Change-Id: I64e2ec0fe9ed9fae3aabf1ca8c9bc0582fc7565a
Reviewed-on: https://skia-review.googlesource.com/7760
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Code and docs are at: https://github.com/ocornut/imgui
ImGui is an open source immediate mode GUI library that's
lightweight and fairly simply to integrate. Widget functions
return their state, and the library emits vertex and index
data to render everything. It's got a huge set of built-in
widgets and really robust layout control.
For the initial integration, I had to fix up event handling
in the viewer's app framework (to get mouse wheel and more
keys, etc...).
The new viewer 'Debug' window is toggled with the space bar.
For this change, I've added one feature to that window: the
slide picker. It's got a list of all slides, with filtering
support, and the ability to click to switch slides.
I also included the ImGui 'Demo' window (toggled with 'g').
This is nicely laid out, and includes examples of pretty
much everything the library can do. It also serves as good
documentation - find something that looks like what you want,
and then go look at the corresponding code (all of it is in
imgui_demo.cpp).
I have other CLs with other features (like directly editing
the primaries of the working color space), but I wanted to
land this chunk first, then start adding more features.
Other than adding new debugging features, there are few
more outstanding work items:
1) Raster doesn't render the GUI correctly, due to non-
invertible pos -> UV matrices. Florin is working on that.
2) Touch inputs aren't being routed yet, so the GUI isn't
usable on Android yet. Might also be tough to work with,
given the size.
3) ImGui has clipboard integration (that's why it wants
the C, X, and V keys), but we need to wire it up to the
OS' clipboard functions.
4) Draw commands can carry a void* payload to support
drawing images (using whatever mechanism the engine has).
I'd like to set that up (probably using SkImage*), which
makes it really easy to add visualization of off-screen
images in GMs, etc...
BUG=skia:
Change-Id: Iac2a63e37228d33141cb55b7e4d60bf11b7e9ae1
Reviewed-on: https://skia-review.googlesource.com/7702
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
