This lets us target older machines with SkSplicer.
SSE2 and SSE4.1 are the sweet spots if we're going to pick two more.
Nothing too interesting here except maybe the f16<->f32 code.
I rearranged a little to keep things consistent across platforms.
Next CL will get this into _generated.h and use it when appropriate.
Change-Id: Ibbdc61ea7a45d22b4f4058b01f75161ea74a7726
Reviewed-on: https://skia-review.googlesource.com/8193
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
We have a couple ways to detect CPU features on ARM:
- on ARMv8, getauxval(AT_HWCAP)
- on ARMv7, getauxval(AT_HWCAP) and cpu-features.h
This guards each of these methods with preprocessor guards to match
exactly when we can use them. Today they're sort of a mix of that and
higher level expectations about particular build and operating systems.
I'm looking into doing this directly by reading CPU registers,
much like we do for x86 further up the file.
None of this is super important right now, so as long as we don't decide
that we have these features when we don't, things will be fine. It's no
big deal for now if we fail to detect them.
Change-Id: I3b7768483086d0f3f4f6516b754c3ea5ec2d03e5
Reviewed-on: https://skia-review.googlesource.com/8182
Reviewed-by: Chinmay Garde <chinmaygarde@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
BUG=skia:6119
Change-Id: I5b5d4298ee21bdcc2cc9ce1fae0826cd56752c2f
Reviewed-on: https://skia-review.googlesource.com/8081
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
It appears that the top-level function named has switched to just "main"
Change-Id: I33a18a8d433867e759312d09e5b258f934f495a4
Reviewed-on: https://skia-review.googlesource.com/8194
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
WindowContext still supports color spaces, but not other color
types. Any off-screen rendering is the app's responsibility.
This change also adds (working) F16 support to viewer. Note that
the previous 10-bit and FP16 support in WindowContext was broken.
There was no code to push the off-screen canvas to the window.
If you ever made it to the unreachable off-screen code path in
createSurface, it would have simply stopped drawing.
The decision to limit the window's gamut to sRGB is mostly driven
by my desire to add real-time editing of gamut. This design lets
us do that, without tearing down and rebuilding the window for
every change. An application could still supply a different gamut
via setDisplayParams and render directly to the back buffer with
proper color correction.
BUG=skia:
Change-Id: I94df35c7a42faee396009acc83683e40bb3c284d
Reviewed-on: https://skia-review.googlesource.com/8153
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Since the SkArenaAlloc handles calling the dtor, it is not longer needed
in the test.
Change-Id: I70a09be7bd0e71bf1e3d55ef08b5e87742e0bd18
Reviewed-on: https://skia-review.googlesource.com/8191
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Build flag available for backends to begin testing their impl.
Need to formalize save/restore, and how to forward these to device but not on picture canvases.
BUG=skia:6214
Change-Id: Ic5c0afba3e8c84fcf124567e63fe2f5880b623e7
Reviewed-on: https://skia-review.googlesource.com/8183
Commit-Queue: Mike Reed <reed@google.com>
Reviewed-by: Florin Malita <fmalita@chromium.org>
Reviewed-by: Brian Salomon <bsalomon@google.com>
This keeps correct linear blending as the only option exericsed,
but it should be easy to see how to turn on blend-wrong mode.
Change-Id: I7d87ef8ed00e8990107bd36b826f8d229d930400
Reviewed-on: https://skia-review.googlesource.com/8125
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This reverts commit e88cf6b7aa.
Reason for revert: See if breaking DEPS roll
Original change's description:
> Remove asTextureRef from SkSpecialImage & update effects accordingly
>
> This CL also renames SkSpecialImage::asTextureProxy to asTextureProxyRef
>
> Change-Id: I5ed8e475bb9688453b825ae4500ed0e8d324b5ac
> Reviewed-on: https://skia-review.googlesource.com/7995
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Robert Phillips <robertphillips@google.com>
>
TBR=bsalomon@google.com,robertphillips@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: I227cbd5fcaf7e2f86e858331d9ec7ff7a5f203ca
Reviewed-on: https://skia-review.googlesource.com/8184
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
TBR=mtklein@google.com
Change-Id: Id7621548995b368164d74c817e288c34ef656bfb
Reviewed-on: https://skia-review.googlesource.com/8180
Reviewed-by: Herb Derby <herb@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Herb Derby <herb@google.com>
Currently create_descriptor will crash if handed a familyName which is
nullptr. Instead it should simply create a descriptor without a family name
and allow CoreText to resolve a default font. This further simplifies
default font handling on Mac.
BUG=skia:6196
Change-Id: I0a2d081240e4cb5bd51dd3516ea2595277055fe0
Reviewed-on: https://skia-review.googlesource.com/8071
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Ben Wagner <bungeman@google.com>
This CL also renames SkSpecialImage::asTextureProxy to asTextureProxyRef
Change-Id: I5ed8e475bb9688453b825ae4500ed0e8d324b5ac
Reviewed-on: https://skia-review.googlesource.com/7995
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
These bugs haven't been updated in over a year, so probably not worthwhile to continue to run the tests.
BUG=skia:4697,skia:4704,skia:4694,skia:4705
Change-Id: Ie77cfdf1a4a12a1eb61f90a3f93e479deb0349ab
Reviewed-on: https://skia-review.googlesource.com/5577
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>
This is just a warm up refactor to make the existing stages more
amenable to running in place without splicing. Hopefully the driver
interpreter loop for that will be my next CL.
Change-Id: Ie76868b59af8ff3a10e2cd64f03c3f8d8a5031dd
Reviewed-on: https://skia-review.googlesource.com/8142
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
- Added default implementation of onMakeContext to support use in android.
Searches for uses:
"public SkShader" package:^chromium$ -file:^src/third_party/skia
package:^aosp.* "public SkShader" -file:external/skia -file:.*third_party/skia
package:^android$ "public SkShader" -file:external/skia -file:.*third_party/skia
... shows that no subclass overrides onCreateContext.
TBR=reed@google.comTBR=mtklein@google.com
Change-Id: I8bd5f57a79534574e344b165d31dccee41c31767
Reviewed-on: https://skia-review.googlesource.com/8140
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Herb Derby <herb@google.com>
BUG=skia:6070
Change-Id: I2ccd5c5a773094da5cadcc934ec953b0bc9de757
Reviewed-on: https://skia-review.googlesource.com/8131
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Eric Boren <borenet@google.com>
Change-Id: I8755fd7a316d284627e25f88e02b69e2f67f9622
Reviewed-on: https://skia-review.googlesource.com/8126
Commit-Queue: Hal Canary <halcanary@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
BUG=skia:
Whitespace change in public API.
Change-Id: Iaab7b0ed6b157b1c246eae5f0f0440b0ae4d72ab
Reviewed-on: https://skia-review.googlesource.com/8130
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This reverts commit 2b57b7f7a7.
Reason for revert: Android compile failing
Original change's description:
> Use SkArenaAlloc instead of SkSmallAllocator in the SkAutoBlitterChoose code.
>
>
> TBR=reed@google.com
> Change-Id: Iefb044bf7657fbf982f23aa91a3f4d013ce2c626
> Reviewed-on: https://skia-review.googlesource.com/7786
> Reviewed-by: Mike Klein <mtklein@chromium.org>
> Reviewed-by: Herb Derby <herb@google.com>
> Commit-Queue: Herb Derby <herb@google.com>
>
TBR=mtklein@chromium.org,mtklein@google.com,herb@google.com,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Id09c35377dddae0811d998b7d0c34c422325a5bc
Reviewed-on: https://skia-review.googlesource.com/8129
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
BUG=skia:
Change-Id: I01c5e1874c9a034febc64e25b3aaafb5050393a6
Reviewed-on: https://skia-review.googlesource.com/8021
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>
Ensures that an image is GPU backed on the passed-in GrContxt. The new
version requires a destination color space (intended usage of the image),
so we can make a proper decision about decoded format.
This reverts commit d263413a2a.
BUG=skia:
Change-Id: Ibccddbafc301779559592045ed5a5fa9264e7432
Reviewed-on: https://skia-review.googlesource.com/8116
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
The 64-bit bots look fine. We're only testing GPU on iOS, so it's not
super critical to have 32-bit iOS testers. I'll keep the builders.
Change-Id: I405153fd9da63b541584d78bdb975b5800679391
Reviewed-on: https://skia-review.googlesource.com/8114
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
SkSplicer is an experiment, and _lowp SkSplicer is an experiment^2.
It's not that it's a bad idea... it just makes working on the main
SkSplicer experiment more complicated.
While not live at head, it'll live on in my heart.
Change-Id: Ib099370be25623e4cf1bbbca211634e3fe92b475
Reviewed-on: https://skia-review.googlesource.com/8112
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Change-Id: I0c2f050c70755523abfbe98c17e90a90ecbedff5
Reviewed-on: https://skia-review.googlesource.com/8113
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Hal Canary <halcanary@google.com>
TBR=reed@google.com
Change-Id: Iefb044bf7657fbf982f23aa91a3f4d013ce2c626
Reviewed-on: https://skia-review.googlesource.com/7786
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Copy-paste error.
BUG=skia:
Change-Id: Ifebe23ee7ac4a46e351f10ebe0f4c68490bfeb44
Reviewed-on: https://skia-review.googlesource.com/8110
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
BUG=skia:6070
Change-Id: I66ac3057f7863f822445d5163e54ca3b11755068
Reviewed-on: https://skia-review.googlesource.com/8107
Reviewed-by: Eric Boren <borenet@google.com>
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>
Change-Id: Id0db49ad45b341d39abf7ee532ecc0799a832192
Reviewed-on: https://skia-review.googlesource.com/8079
Commit-Queue: Stan Iliev <stani@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
An upcoming CL (https://skia-review.googlesource.com/c/7995/ (Remove asTextureRef from SkSpecialImage & update effects accordingly)) modifies some untested portions of the SkMagnifierImageFilter.
This adds a test to prevents regressions.
Change-Id: I9fa406f699e39fa393212e7f63a457b015b36edb
Reviewed-on: https://skia-review.googlesource.com/8023
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Because it's 2017.
CQ_INCLUDE_TRYBOTS=skia.primary:Perf-iOS-Clang-iPadMini4-GPU-GX6450-arm64-Debug,Perf-iOS-Clang-iPadMini4-GPU-GX6450-arm64-Release,Test-iOS-Clang-iPadMini4-GPU-GX6450-arm64-Debug,Test-iOS-Clang-iPadMini4-GPU-GX6450-arm64-Release
Change-Id: I8708585a3416962d2563c17e6cc357da48ad9519
Reviewed-on: https://skia-review.googlesource.com/8106
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Uses Sk2f to apply a translation-only matrix when the vertex attributes contain only positions and colors.
We should look at how to generalize this for other draw vertices cases and other ops.
Change-Id: I5eb692982dc216b1c0a71209c969672b0562143c
Reviewed-on: https://skia-review.googlesource.com/8103
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
We were snapping the top-left of the destination rectangle to integers,
but using the original (fractional) size. This led to us losing rows or
columns of pixels at the first tile boundary.
Also added a GM that demonstrates the bug (now renders correctly).
BUG=skia:
Change-Id: I50629dab9dd90fedad2c7e3393a1b1d1c7a8d45e
Reviewed-on: https://skia-review.googlesource.com/8102
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
(Long description, but actually quite a small change.)
This is the first change inspired by the "straight skeleton" algorithm.
This algorithm gives us a mental model to describe the edge-antialiasing
problem: consider the shape as walls of a house, the outer alpha
geometry is a 45-degree "roof" up to a flat top at a height of 1.0
(the filled interior). The faces of the sloping roof join at the
bisectors between the inner and outer points.
When the shape being drawn is sufficiently thin, there should be no flat
roof, and the sloping roof meets at an edge (the straight skeleton).
This patch detects cases where an edge inverts on stroking, which
indicates that the flat roof has turned inside out, and should be
reduced to a point instead. The model above describes what to do:
follow down the "roof" along the bisectors to their intersection.
This is the point to which an inverted edge should be collapsed.
Fortunately, the bisector edges are easy to compute: they're the
connector edges joining inner and outer points. Linearly interpolating
the distance from the top to the bottom point gives the alpha we
should use to approximate coverage.
Now that we are correctly handling inversions, bevelling outer edges
is no longer necesary, since pointy outer edges won't cause nasty
opaque artifacts.
A couple of other quality improvements: on intersection, always lerp
the alpha of connector edge, even if the opposite edge is an inner edge
(later, when these edges are collapsed, we need this value to compute
the correct alpha). Fix the case where an intruding outer vertex
intersects exactly with an inner edge by maxing its alpha with the
computed value in check_for_intersection(). Finally, we also no longer
round off the intersections produced by Line::intersect(), since it
introduces a loss of quality with no measurable performance benefit.
Change-Id: I6fd93df3a57fffc0895e8cb68adbdba626ded0f1
Reviewed-on: https://skia-review.googlesource.com/8028
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Stephan White <senorblanco@chromium.org>
