These types are ref-counted, but don't otherwise need a vtable.
This makes them good candidates for SkNVRefCnt.
Destruction can be a little more direct, and if nothing else,
sizeof(T) will get a little smaller by dropping the vptr.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2232433002
Review-Url: https://codereview.chromium.org/2232433002
About 9x faster than Murmur3 for long inputs.
Most of this is a mechanical change from SkChecksum::Murmur3(...) to SkOpts::hash(...).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2208903002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot;master.client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac-Clang-x86_64-Release-CMake-Trybot
Review-Url: https://codereview.chromium.org/2208903002
The code protected by this flag is no longer used. Remove the flag and
code. This also removes SK_SUPPORT_LEGACY_TYPEFACE_PTR from Android, as
it is no longer needed.
TBR=reed
Only removes already guarded API.
Review-Url: https://codereview.chromium.org/2223933002
SkLiteRecorder, a new SkCanvas, fills out SkLiteDL, a new SkDrawable.
This SkDrawable is a display list similar to SkRecord and SkBigPicture / SkRecordedDrawable, but with a few new design points inspired by Android and slimming paint:
1) SkLiteDL is structured as one big contiguous array rather than the two layer structure of SkRecord. This trades away flexibility and large-op-count performance for better data locality for small to medium size pictures.
2) We keep a global freelist of SkLiteDLs, both reusing the SkLiteDL struct itself and its contiguous byte array. This keeps the expected number of mallocs per display list allocation <1 (really, ~0) for cyclical use cases.
These two together mean recording is faster. Measuring against the code we use at head, SkLiteRecorder trends about ~3x faster across various size pictures, matching speed at 0 draws and beating the special-case 1-draw pictures we have today. (I.e. we won't need those special case implementations anymore, because they're slower than this new generic code.) This new strategy records 10 drawRects() in about the same time the old strategy took for 2.
This strategy stays the winner until at least 500 drawRect()s on my laptop, where I stopped checking.
A simpler alternative to freelisting is also possible (but not implemented here), where we allow the client to manually reset() an SkLiteDL for reuse when its refcnt is 1. That's essentially what we're doing with the freelist, except tracking what's available for reuse globally instead of making the client do it.
This code is not fully capable yet, but most of the key design points are there. The internal structure of SkLiteDL is the area I expect to be most volatile (anything involving Op), but its interface and the whole of SkLiteRecorder ought to be just about done.
You can run nanobench --match picture_overhead as a demo. Everything it exercises is fully fleshed out, so what it tests is an apples-to-apples comparison as far as recording costs go. I have not yet compared playback performance.
It should be simple to wrap this into an SkPicture subclass if we want.
I won't start proposing we replace anything old with anything new quite yet until I have more ducks in a row, but this does look pretty promising (similar to the SkRecord over old SkPicture change a couple years ago) and I'd like to land, experiment, iterate, especially with an eye toward Android.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2213333002
Review-Url: https://codereview.chromium.org/2213333002
In Firefox, we use SkCanvas::saveLayer in combination with a backdrop that initializes the layer to the background. When this is blended back onto background using transparency, where the source and destination pixel colors are the same, the resulting color after the blend is not preserved due to the lost precision mentioned above. In cases where this operation is repeatedly performed, this causes substantially noticeable differences in color as evidenced in this downstream Firefox bug report: https://bugzilla.mozilla.org/show_bug.cgi?id=1200684
In the test-case in the downstream report, essentially it does blend(src=0xFF2E3338, dst=0xFF2E3338, scale=217), which gives the result 0xFF2E3237, while we would expect to get back 0xFF2E3338.
This problem goes away if the blend is instead reformulated to effectively do (src*src_scale + dst*dst_scale)>>8, which keeps the intermediate precision during the addition before shifting it off.
This modifies the blending operations thusly. The performance should remain mostly unchanged, or possibly improve slightly, so there should be no real downside to doing this, with the benefit of making the results more accurate. Without this, it is currently unsafe for Firefox to blend a layer back onto itself that was initialized with a copy of its background.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2097883002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
[mtklein adds...]
No public API changes.
TBR=reed@google.com
Review-Url: https://codereview.chromium.org/2097883002
Change SkDataTable::NewXXX to SkDataTable::MakeXXX and return sk_sp.
This updates users of SkDataTable to sk_sp as well.
There do not appear to be any external users of these methods.
Review-Url: https://codereview.chromium.org/2211143002
Lessons learned
1. ImageShader (correctly) always compresses (typically via PNG) during serialization. This has the surprise results of
- if the image was marked opaque, but has some non-opaque pixels (i.e. bug in blitter or caller), then compressing may "fix" those pixels, making the deserialized version draw differently. bug filed.
- 565 compressess/decompresses to 8888 (at least on Mac), which draws differently (esp. under some filters). bug filed.
2. BitmapShader did not enforce a copy for mutable bitmaps, but ImageShader does (since it creates an Image). Thus the former would see subsequent changes to the pixels after shader creation, while the latter does not, hence the change to the BlitRow test to avoid this modify-after-create pattern. I sure hope this prev. behavior was a bug/undefined-behavior, since this CL changes that.
BUG=skia:5595
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2195893002
Review-Url: https://codereview.chromium.org/2195893002
No one appears to be using the default 2.2.
This makes SK_GAMMA_SRGB the default / a no-op, which seems more realistic.
Chrome's explicitly setting SK_GAMMA_EXPONENT or SK_GAMMA_SRGB.
We set Android platform explicitly to 1.4.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2198073002
Review-Url: https://codereview.chromium.org/2198073002
Reason for revert:
UBSAN says we're reading a bad bool here:
bool usesDistanceVectorField() const { return fUsesDistanceVectorField; }
../../../include/gpu/GrPaint.h:83:51: runtime error: load of value 239, which is not a valid value for type 'bool'
SUMMARY: AddressSanitizer: undefined-behavior ../../../include/gpu/GrPaint.h:83:51 in
Seems likely also the root of Valgrind failure:
https://luci-milo.appspot.com/swarming/task/30522e4f2241cb10
Original issue's description:
> GrFP can express distance vector field req., program builder declares variable for it
>
> This update allows fragment processors to require a field of vectors to the nearest edge. This requirement propagates:
>
> - from child FPs to their parent
> - from parent FPs to the GrPaint
> - from GrPaint through the PipelineBuilder into GrPipeline
> - acessed from GrPipeline by GrGLSLProgramBuilder
>
> GrGLSL generates a variable for the distance vector and passes it down to the GeometryProcessor->emitCode() method.
>
> This CL's base is the CL for adding the BevelNormalSource API: https://codereview.chromium.org/2080993002
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2114993002
>
> Committed: https://skia.googlesource.com/skia/+/4ef6dfa7089c092c67b0d5ec34e89c1e319af196TBR=egdaniel@google.com,robertphillips@google.com,bsalomon@google.com,dvonbeck@google.com
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=skia:
Review-Url: https://codereview.chromium.org/2201613002
Adds the ability for GLInstancedRendering to use
glDrawElementsInstanced when glDrawElementsIndirect is not supported.
The only remaining 3.1 dependency now is EXT_texture_buffer.
Also moves the cap for glDraw*Instanced out of GrCaps and into
GrGLCaps.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2193303002
Review-Url: https://codereview.chromium.org/2193303002
In order to remove this define it will first need to be defined in
Chromium. However, doing so causes redefinition warnings as errors.
Only define this macro if it is not already defined to avoid this.
TBR=reed
This doesn't change any API.
Review-Url: https://codereview.chromium.org/2198453002
This update allows fragment processors to require a field of vectors to the nearest edge. This requirement propagates:
- from child FPs to their parent
- from parent FPs to the GrPaint
- from GrPaint through the PipelineBuilder into GrPipeline
- acessed from GrPipeline by GrGLSLProgramBuilder
GrGLSL generates a variable for the distance vector and passes it down to the GeometryProcessor->emitCode() method.
This CL's base is the CL for adding the BevelNormalSource API: https://codereview.chromium.org/2080993002
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2114993002
Review-Url: https://codereview.chromium.org/2114993002
New API mirrors the form of similar APIs in SkRegion,
SkMatrix, etc.
This also fixes a bug:
SkImageInfo appears in a object that Chrome stores in
discardable memory. So when sk_sp<SkColorSpace> was added
to SkImageInfo a leak was introduced. We'll use this new
method and deserialize to store the SkColorSpace in the
discardable object.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2192903002
Review-Url: https://codereview.chromium.org/2192903002
