The colorfilter is applied to a single (paint's) color, so the bench does not
measure the filter at all, but simply the blit of a color.
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1055383002
It's looking like the previous CL did not fix the Daisy bot GMs,
even though that's still the only bit of code I can find that was
ignoring color order. Puzzled. Reverting arm_version=7 for now.
BUG=skia:1843
Review URL: https://codereview.chromium.org/1051423002
Add a virtual method on SkStream which will do a "peek" some bytes, so
that those bytes are read, but the next call to read will be
unaffected.
Implement peek for SkMemoryStream, where the implementation is simple
and obvious.
Implement peek on SkFrontBufferedStream.
Add tests.
Motivated by decoding streams which cannot be rewound.
TBR=reed@google.com
BUG=skia:3257
Review URL: https://codereview.chromium.org/1044953002
I don't see any color-order handling logic in the 32-bit code.
BUG=skia:1843
CQ_EXCLUDE_TRYBOTS=client.skia.compile:Build-Win-MSVC-x86-Debug-Trybot,Build-Win-MSVC-x86_64-Debug-Trybot
R=mtklein@google.com
Review URL: https://codereview.chromium.org/1051683003
This font contains two glyphs, one for 'a' and one for 'A'. The em size
is 128, and the 'a' fits in this. The big 'A', however, is ~3000 in
in each direction.
Review URL: https://codereview.chromium.org/1016153002
Each of these conversion functions now only asserts is output is valid.
For SkPMColor -> SkPMFloat, we assert isValid().
For SkPMFloat -> SkPMColor, we SkPMColorAssert.
#floats
BUG=skia:
BUG=skia:3592
Review URL: https://codereview.chromium.org/1055093002
- remove unused alex
- streamline Link's config
- remove misleading Daisy config:
1) armv7=1 does nothing. We meant to type arm_version=7 here.
2) arm_neon=1 does nothing unless arm_version == 7.
3) arm_thumb=0 is the default when arm_version <= 7.
4) skia_arch_width=32 is the default when skia_arch_type=arm.
I'd just fix this to make Daisy arm_version=7 and arm_neon=1 (and
arm_thumb=1, which I'm going to separately make the default for
arm_version=7), but there are known color-order bugs with our
NEON procs that would make Daisy start pushing bad images to
Gold. Going to take baby steps here...
BUG=skia:1843
Committed: https://skia.googlesource.com/skia/+/3c2809bc612f4a265770914f860d214c9665dc4a
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-Arm7-Debug-CrOS_Daisy-Trybot
Review URL: https://codereview.chromium.org/1051253002
Reason for revert:
arm_thumb not defined
Original issue's description:
> tidy up chromeos_setup.sh
>
> - remove unused alex
> - streamline Link's config
> - remove misleading Daisy config:
> 1) armv7=1 does nothing. We meant to type arm_version=7 here.
> 2) arm_neon=1 does nothing unless arm_version == 7.
> 3) arm_thumb=0 is the default when arm_version <= 7.
> 4) skia_arch_width=32 is the default when skia_arch_type=arm.
>
> I'd just fix this to make Daisy arm_version=7 and arm_neon=1 (and
> arm_thumb=1, which I'm going to separately make the default for
> arm_version=7), but there are known color-order bugs with our
> NEON procs that would make Daisy start pushing bad images to
> Gold. Going to take baby steps here...
>
> BUG=skia:1843
>
> Committed: https://skia.googlesource.com/skia/+/3c2809bc612f4a265770914f860d214c9665dc4aTBR=borenet@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:1843
Review URL: https://codereview.chromium.org/1059443002
- remove unused alex
- streamline Link's config
- remove misleading Daisy config:
1) armv7=1 does nothing. We meant to type arm_version=7 here.
2) arm_neon=1 does nothing unless arm_version == 7.
3) arm_thumb=0 is the default when arm_version <= 7.
4) skia_arch_width=32 is the default when skia_arch_type=arm.
I'd just fix this to make Daisy arm_version=7 and arm_neon=1 (and
arm_thumb=1, which I'm going to separately make the default for
arm_version=7), but there are known color-order bugs with our
NEON procs that would make Daisy start pushing bad images to
Gold. Going to take baby steps here...
BUG=skia:1843
Review URL: https://codereview.chromium.org/1051253002
Currently, the GPU-side image filter implementation creates
exact-match textures for the offscreen backing stores for
saveLayer(). This is because several filters have GPU
implementations which depend on the texture coordinates
being 0..1.
The fix is three-fold:
1) Store the actual requested size in the SkGpuDevice, so
that when wrapping it in an SkBitmap for passing to
filterImage(), we can give it the original size.
2) Fix the filters (SkMagnifierImageFilter,
SkLightingImageFilter, SkMatrixConvolutionImageFilter,
SkMatrixImageFilter) whose GPU implementation depends on
0..1 texture coordinates.
3) Remove the exception for GPU-side image filters in
SkCanvas::internalSaveLayer().
For the lighting filters, there were two bugs which were
cancelling each other out: the sobel filter matrix was
being computed upside down, but then we'd negate the
resulting normal. This worked fine in the exact-match case,
but in the approx-match case we'd sample garbage along
the edge pixels. Also, we never implemented the edge pixels
according to spec in the GPU case. It requires a
different fragment shader for each edge of the nine-patch,
which meant we couldn't use asFragmentProcessor(), and had
to implement the drawing via a filterImageGPU() override.
In order to avoid polluting the public API, I inserted a
new base class, SkLightingImageFilterInternal above
Sk[Diffuse|Specular]LightingImageFilter to handle the
implementation.
For the SkMatrixConvolutionImageFilter, it seems the
GLSL clamp() function occasionally returns values outside
the clamped range, resulting in access of garbage
texels even in GL_NEAREST. The fix here is to clamp to a
rect inset by half a texel. There was also a bug in
the unpremultiply step when fConvolveAlpha is false.
For SkMatrixImageFilter, the fix was to make the generic
draw path be more careful about when to use texture domain.
If the bitmap already has a texture, use texture domain
if the srcRect is smaller than the entire texture (not
the entire bitmap).
N.B.: this change will cause some minor pixel diffs in the
GPU results of the following GMs (and possibly more):
matriximagefilter, matrixconvolution, imagefiltersscaled,
lighting, imagemagnifier, filterfastbounds,
complexclip_aa_Layer_invert, complexclip_aa_layer,
complexclip_bw_layer_invert, complexclip_bw_layer.
BUG=skia:3532
Committed: https://skia.googlesource.com/skia/+/b97dafefe63ea0a1bbce8e8b209f4920983fb8b9
Committed: https://skia.googlesource.com/skia/+/f5f8518fe0bbd2703e4ffc1b11ad7b4312ff7641
Committed: https://skia.googlesource.com/skia/+/46112cf2a7c7307f1c9eebb5f881cbda15aa460c
Review URL: https://codereview.chromium.org/1034733002
Add two drawPicture bracketing ops (BeginDrawPicture, EndDrawPicture) to
replace the current DrawPicture op, and flatten picture contents.
Review URL: https://codereview.chromium.org/1048383002
Reason for revert:
Spoke to Stephen about this. Reverting because failing debug builds:
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Mac10.9-Clang-MacMini6.2-GPU-HD4000-x86_64-Debug/builds/51https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-ShuttleA-GPU-GTX660-x86_64-Debug/builds/54
Original issue's description:
> Implement approx-match support in image filter saveLayer() offscreen.
>
> Currently, the GPU-side image filter implementation creates
> exact-match textures for the offscreen backing stores for
> saveLayer(). This is because several filters have GPU
> implementations which depend on the texture coordinates
> being 0..1.
>
> The fix is three-fold:
>
> 1) Store the actual requested size in the SkGpuDevice, so
> that when wrapping it in an SkBitmap for passing to
> filterImage(), we can give it the original size.
> 2) Fix the filters (SkMagnifierImageFilter,
> SkLightingImageFilter, SkMatrixConvolutionImageFilter,
> SkMatrixImageFilter) whose GPU implementation depends on
> 0..1 texture coordinates.
> 3) Remove the exception for GPU-side image filters in
> SkCanvas::internalSaveLayer().
>
> For the lighting filters, there were two bugs which were
> cancelling each other out: the sobel filter matrix was
> being computed upside down, but then we'd negate the
> resulting normal. This worked fine in the exact-match case,
> but in the approx-match case we'd sample garbage along
> the edge pixels. Also, we never implemented the edge pixels
> according to spec in the GPU case. It requires a
> different fragment shader for each edge of the nine-patch,
> which meant we couldn't use asFragmentProcessor(), and had
> to implement the drawing via a filterImageGPU() override.
> In order to avoid polluting the public API, I inserted a
> new base class, SkLightingImageFilterInternal above
> Sk[Diffuse|Specular]LightingImageFilter to handle the
> implementation.
>
> For the SkMatrixConvolutionImageFilter, it seems the
> GLSL clamp() function occasionally returns values outside
> the clamped range, resulting in access of garbage
> texels even in GL_NEAREST. The fix here is to clamp to a
> rect inset by half a texel. There was also a bug in
> the unpremultiply step when fConvolveAlpha is false.
>
> For SkMatrixImageFilter, the fix was to make the generic
> draw path be more careful about when to use texture domain.
> If the bitmap already has a texture, use texture domain
> if the srcRect is smaller than the entire texture (not
> the entire bitmap).
>
> N.B.: this change will cause some minor pixel diffs in the
> GPU results of the following GMs (and possibly more):
> matriximagefilter, matrixconvolution, imagefiltersscaled,
> lighting, imagemagnifier, filterfastbounds,
> complexclip_aa_Layer_invert, complexclip_aa_layer,
> complexclip_bw_layer_invert, complexclip_bw_layer.
>
> BUG=skia:3532
>
> Committed: https://skia.googlesource.com/skia/+/b97dafefe63ea0a1bbce8e8b209f4920983fb8b9
>
> Committed: https://skia.googlesource.com/skia/+/f5f8518fe0bbd2703e4ffc1b11ad7b4312ff7641
>
> Committed: https://skia.googlesource.com/skia/+/46112cf2a7c7307f1c9eebb5f881cbda15aa460cTBR=bsalomon@google.com,reed@chromium.org,senorblanco@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:3532
Review URL: https://codereview.chromium.org/1057693002
Now that all SkCodecs can rewind (assuming the stream is rewindable),
we do not need to special case it.
Pointed out by Derek in the code review that added this.
TBR=djsollen
Review URL: https://codereview.chromium.org/1058633002
Currently, the GPU-side image filter implementation creates
exact-match textures for the offscreen backing stores for
saveLayer(). This is because several filters have GPU
implementations which depend on the texture coordinates
being 0..1.
The fix is three-fold:
1) Store the actual requested size in the SkGpuDevice, so
that when wrapping it in an SkBitmap for passing to
filterImage(), we can give it the original size.
2) Fix the filters (SkMagnifierImageFilter,
SkLightingImageFilter, SkMatrixConvolutionImageFilter,
SkMatrixImageFilter) whose GPU implementation depends on
0..1 texture coordinates.
3) Remove the exception for GPU-side image filters in
SkCanvas::internalSaveLayer().
For the lighting filters, there were two bugs which were
cancelling each other out: the sobel filter matrix was
being computed upside down, but then we'd negate the
resulting normal. This worked fine in the exact-match case,
but in the approx-match case we'd sample garbage along
the edge pixels. Also, we never implemented the edge pixels
according to spec in the GPU case. It requires a
different fragment shader for each edge of the nine-patch,
which meant we couldn't use asFragmentProcessor(), and had
to implement the drawing via a filterImageGPU() override.
In order to avoid polluting the public API, I inserted a
new base class, SkLightingImageFilterInternal above
Sk[Diffuse|Specular]LightingImageFilter to handle the
implementation.
For the SkMatrixConvolutionImageFilter, it seems the
GLSL clamp() function occasionally returns values outside
the clamped range, resulting in access of garbage
texels even in GL_NEAREST. The fix here is to clamp to a
rect inset by half a texel. There was also a bug in
the unpremultiply step when fConvolveAlpha is false.
For SkMatrixImageFilter, the fix was to make the generic
draw path be more careful about when to use texture domain.
If the bitmap already has a texture, use texture domain
if the srcRect is smaller than the entire texture (not
the entire bitmap).
N.B.: this change will cause some minor pixel diffs in the
GPU results of the following GMs (and possibly more):
matriximagefilter, matrixconvolution, imagefiltersscaled,
lighting, imagemagnifier, filterfastbounds,
complexclip_aa_Layer_invert, complexclip_aa_layer,
complexclip_bw_layer_invert, complexclip_bw_layer.
BUG=skia:3532
Committed: https://skia.googlesource.com/skia/+/b97dafefe63ea0a1bbce8e8b209f4920983fb8b9
Committed: https://skia.googlesource.com/skia/+/f5f8518fe0bbd2703e4ffc1b11ad7b4312ff7641
Review URL: https://codereview.chromium.org/1034733002
In Release, running `dm --src skp --config pdf`, I get a
speedup of about 1.2%.
SkPDFGraphicState class:
- Holds the subset of SkPaint that maps to a PDF Graphics
State
- These fields are easily comparable, making hashtable
comparisons easy.
SkPDFCanon:
- findGraphicState() takes a SkPDFGraphicState, not a SkPaint
- fGraphicStateRecords is a SkHashSet, not a SkTDArray
SkPDFGraphicState:
- mode_for_pdf() replaces logic inside equivalent(), but is
only called once per lookup.
- emitObject() no longer modifies the SkPDFGraphicState to
cache the SkPDFDict stucture. (Since it is de-duped,
this get no speedup).
- Static Functions that don't use the canon return a plain
SkPDFDict now. No need for fPopulated.
SkTHash.h
- SkHashSet::forall added
SkPDFDevice; SkPDFShader
- Updated for new SkPDFGraphicState interface.
BUG=skia:3585
Review URL: https://codereview.chromium.org/1046293002
The idea here is that we determine the 0.5 crossover for each
row in the gamma table, then invert the mapping to determine
which point that maps to in the original range [-.65,
.65]. That gives us a change in the apparent width of the
glyph that closely corresponds to the change produced by the gamma fix.
BUG=skia:2933
Review URL: https://codereview.chromium.org/1042373002
CodecBench:
Add new class for timing using SkCodec.
DecodingBench:
Include creating a decoder inside the loop. This is to have a better
comparison against SkCodec. SkCodec's factory function does not
necessarily read the same amount as SkImageDecoder's, so in order to
have a meaningful comparison, read the entire stream from the
beginning. Also for comparison, create a new SkStream from the
SkData each time.
Add a debugging check to make sure we have an SkImageDecoder.
Add include guards.
nanobench.cpp:
Decode using SkCodec.
When decoding using SkImageDecoder, exclude benches where we decoded
to a different color type than requested. SkImageDecoder may decide to
decode to a different type, in which case the name is misleading.
TODOs:
Now that we ignore color types that do not match the desired
color type, we should add Index8. This also means calling the more
complex version of getPixels so CodecBench can support kIndex8.
BUG=skia:3257
Review URL: https://codereview.chromium.org/1044363002
This mirrors the behavior in onGetPixels, and allows the implementation
to share code for handling calls to rewindIfNeeded.
This also fixes a bug where getScanlineDecoder was calling
rewindIfNeeded and treating the result as a bool.
In SkPngCodec, factor out the code to call rewindIfNeeded, and call it
in both onGetPixels and onGetScanlineDecoder.
Update the test to include testing the scanline decoder. Rename "gen"
to "codec" now that it must be an SkCodec.
BUG=skia:3257
Depends on https://codereview.chromium.org/1048423003/ (DIFFERENT ISSUE).
Review URL: https://codereview.chromium.org/1050893002
Reason for revert:
breaking the nexus_9 and ios builds.
Original issue's description:
> Enable both static and dynamically linked libpng
>
> All platforms except android are configured to use the statically linked copy of libpng. Android uses the system provided dynamic copy for SkImageDecoder and the static copy for SkCodec. The exception being android framework builds that currently use the dynamic copy everywhere.
>
> This CL also enables NEON optimizations for libpng.
>
> Committed: https://skia.googlesource.com/skia/+/2469c999518e7b0063d35e9e2eb074a0477c21acTBR=scroggo@google.com,msarett@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1050183002
All platforms except android are configured to use the statically linked copy of libpng. Android uses the system provided dynamic copy for SkImageDecoder and the static copy for SkCodec. The exception being android framework builds that currently use the dynamic copy everywhere.
This CL also enables NEON optimizations for libpng.
Review URL: https://codereview.chromium.org/1032253003
