platform/upstream/libSkiaSharp.git
9 years agoStreamline Thumb config.
mtklein [Thu, 2 Apr 2015 17:47:57 +0000 (10:47 -0700)]
Streamline Thumb config.

Enable Thumb2 iff ARM v7.  (We don't turn it on today for ARM <v7, and ARM v8 doesn't support it.)

BUG=skia:

Review URL: https://codereview.chromium.org/1054993002

9 years agoMake Daisy ARMv7.
mtklein [Thu, 2 Apr 2015 17:24:23 +0000 (10:24 -0700)]
Make Daisy ARMv7.

This should be a ~noop as far as gold.skia.org goes.
After this, I'll try out NEON.

BUG=skia:1843

Review URL: https://codereview.chromium.org/1056793004

9 years agotidy up chromeos_setup.sh
mtklein [Thu, 2 Apr 2015 16:27:57 +0000 (09:27 -0700)]
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/+/3c2809bc612f4a265770914f860d214c9665dc4a

CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-Arm7-Debug-CrOS_Daisy-Trybot

Review URL: https://codereview.chromium.org/1051253002

9 years agoRename GrBitmapTextContextB to GrAtlasTextContext
joshualitt [Thu, 2 Apr 2015 16:19:04 +0000 (09:19 -0700)]
Rename GrBitmapTextContextB to GrAtlasTextContext

BUG=skia:

Review URL: https://codereview.chromium.org/1050173002

9 years agoRevert of tidy up chromeos_setup.sh (patchset #1 id:1 of https://codereview.chromium...
mtklein [Thu, 2 Apr 2015 16:02:12 +0000 (09:02 -0700)]
Revert of tidy up chromeos_setup.sh (patchset #1 id:1 of 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/+/3c2809bc612f4a265770914f860d214c9665dc4a

TBR=borenet@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:1843

Review URL: https://codereview.chromium.org/1059443002

9 years agotidy up chromeos_setup.sh
mtklein [Thu, 2 Apr 2015 15:52:36 +0000 (08:52 -0700)]
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

Review URL: https://codereview.chromium.org/1051253002

9 years agoRevert of Add constant color GrFP. (patchset #10 id:180001 of https://codereview...
bsalomon [Thu, 2 Apr 2015 15:33:54 +0000 (08:33 -0700)]
Revert of Add constant color GrFP. (patchset #10 id:180001 of https://codereview.chromium.org/978713002/)

Reason for revert:
Revert while investigating assertions.

Original issue's description:
> Add constant color GrFP.
>
> Committed: https://skia.googlesource.com/skia/+/dfbbec436cbcacc3270d4b28357c8393e67d6494

TBR=egdaniel@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1055023002

9 years agoSkip desk_baidu skp on Valgrind bot
borenet [Thu, 2 Apr 2015 14:24:03 +0000 (07:24 -0700)]
Skip desk_baidu skp on Valgrind bot

This has been running for 16 hours:
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Valgrind/builds/12/

BUG=skia:3506

Review URL: https://codereview.chromium.org/1056883002

9 years agopass legal premul values to bitmap -- do we still need this GM?
reed [Thu, 2 Apr 2015 12:43:09 +0000 (05:43 -0700)]
pass legal premul values to bitmap -- do we still need this GM?

BUG=skia:

Review URL: https://codereview.chromium.org/1052083004

9 years agoImplement approx-match support in image filter saveLayer() offscreen.
senorblanco [Thu, 2 Apr 2015 11:54:56 +0000 (04:54 -0700)]
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/+/46112cf2a7c7307f1c9eebb5f881cbda15aa460c

Review URL: https://codereview.chromium.org/1034733002

9 years ago[SkDebugger] Flatten drawPicture ops
fmalita [Thu, 2 Apr 2015 03:58:37 +0000 (20:58 -0700)]
[SkDebugger] Flatten drawPicture ops

Add two drawPicture bracketing ops (BeginDrawPicture, EndDrawPicture) to
replace the current DrawPicture op, and flatten picture contents.

Review URL: https://codereview.chromium.org/1048383002

9 years agoRevert of Implement approx-match support in image filter saveLayer() offscreen. ...
rmistry [Wed, 1 Apr 2015 22:53:13 +0000 (15:53 -0700)]
Revert of Implement approx-match support in image filter saveLayer() offscreen. (patchset #31 id:590001 of https://codereview.chromium.org/1034733002/)

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/51
https://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/+/46112cf2a7c7307f1c9eebb5f881cbda15aa460c

TBR=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

9 years agoRevert of Fix GLSL error on Android. (patchset #2 id:20001 of https://codereview...
rmistry [Wed, 1 Apr 2015 22:51:33 +0000 (15:51 -0700)]
Revert of Fix GLSL error on Android. (patchset #2 id:20001 of https://codereview.chromium.org/1053873002/)

Reason for revert:
Spoke to Stephan 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/51
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-ShuttleA-GPU-GTX660-x86_64-Debug/builds/54

Original issue's description:
> Fix GLSL error on Android.
>
> BUG=skia:
> TBR=bsalomon
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/f90cd8e0e39af02c3826c80366efa3c06e88f642

TBR=bsalomon@google.com,senorblanco@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1056713002

9 years agoFix GLSL error on Android.
senorblanco [Wed, 1 Apr 2015 22:13:29 +0000 (15:13 -0700)]
Fix GLSL error on Android.

BUG=skia:
TBR=bsalomon
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1053873002

9 years agoAdd constant color GrFP.
bsalomon [Wed, 1 Apr 2015 21:54:57 +0000 (14:54 -0700)]
Add constant color GrFP.

Review URL: https://codereview.chromium.org/978713002

9 years agonanobench does not need to handle failed rewind.
scroggo [Wed, 1 Apr 2015 21:34:40 +0000 (14:34 -0700)]
nanobench does not need to handle failed rewind.

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

9 years agoImplement approx-match support in image filter saveLayer() offscreen.
senorblanco [Wed, 1 Apr 2015 21:17:14 +0000 (14:17 -0700)]
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

Review URL: https://codereview.chromium.org/1034733002

9 years agoThat last CL subverted the purpose of the test. Put it back how it was.
mtklein [Wed, 1 Apr 2015 20:36:23 +0000 (13:36 -0700)]
That last CL subverted the purpose of the test.  Put it back how it was.

BUG=skia:

Review URL: https://codereview.chromium.org/1055633003

9 years agoDM: JSON output adds process_max_resident_set_size_MB
halcanary [Wed, 1 Apr 2015 20:31:34 +0000 (13:31 -0700)]
DM: JSON output adds process_max_resident_set_size_MB

Review URL: https://codereview.chromium.org/917943002

9 years agoSkPDF: SkPDFGraphicState Lookup hashtabled
halcanary [Wed, 1 Apr 2015 20:31:19 +0000 (13:31 -0700)]
SkPDF: SkPDFGraphicState Lookup hashtabled

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

9 years agoImplicit constructors for SkFunction are much more readable.
mtklein [Wed, 1 Apr 2015 20:08:50 +0000 (13:08 -0700)]
Implicit constructors for SkFunction are much more readable.

BUG=skia:

Review URL: https://codereview.chromium.org/1052663004

9 years agoEmulate distance field gamma fix by making glyphs thicker or thinner
jvanverth [Wed, 1 Apr 2015 19:43:48 +0000 (12:43 -0700)]
Emulate distance field gamma fix by making glyphs thicker or thinner

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

9 years agomove Atlas Text Context to its own file
joshualitt [Wed, 1 Apr 2015 19:40:54 +0000 (12:40 -0700)]
move Atlas Text Context to its own file

BUG=skia:

Review URL: https://codereview.chromium.org/1045723010

9 years agoUpdate the Android scripts to refresh device names and capabilities.
djsollen [Wed, 1 Apr 2015 19:15:39 +0000 (12:15 -0700)]
Update the Android scripts to refresh device names and capabilities.

Review URL: https://codereview.chromium.org/1054633002

9 years agoIco test with embedded png
msarett [Wed, 1 Apr 2015 19:09:21 +0000 (12:09 -0700)]
Ico test with embedded png

BUG=skia:3257

Review URL: https://codereview.chromium.org/1054673002

9 years agoAdd timing SkCodec to nanobench.
scroggo [Wed, 1 Apr 2015 19:09:17 +0000 (12:09 -0700)]
Add timing SkCodec to nanobench.

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

9 years agoSwitch to one single bitmap text blob cache allocation
joshualitt [Wed, 1 Apr 2015 18:30:27 +0000 (11:30 -0700)]
Switch to one single bitmap text blob cache allocation

BUG=skia:

Review URL: https://codereview.chromium.org/1041953002

9 years agoConstructor and call argument forwarding for SkFunction.
mtklein [Wed, 1 Apr 2015 18:26:31 +0000 (11:26 -0700)]
Constructor and call argument forwarding for SkFunction.

BUG=skia:

Review URL: https://codereview.chromium.org/1050113003

9 years agoSkCodec::onGetScanlineDecoder must call rewind.
scroggo [Wed, 1 Apr 2015 18:25:20 +0000 (11:25 -0700)]
SkCodec::onGetScanlineDecoder must call rewind.

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

9 years agoAdd SkTHashSet::find()
mtklein [Wed, 1 Apr 2015 18:21:27 +0000 (11:21 -0700)]
Add SkTHashSet::find()

BUG=skia:

Review URL: https://codereview.chromium.org/1058553002

9 years agoRevert of Enable both static and dynamically linked libpng (patchset #4 id:60001...
djsollen [Wed, 1 Apr 2015 18:20:12 +0000 (11:20 -0700)]
Revert of Enable both static and dynamically linked libpng (patchset #4 id:60001 of https://codereview.chromium.org/1032253003/)

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/+/2469c999518e7b0063d35e9e2eb074a0477c21ac

TBR=scroggo@google.com,msarett@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1050183002

9 years agoEnable both static and dynamically linked libpng
djsollen [Wed, 1 Apr 2015 17:33:55 +0000 (10:33 -0700)]
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.

Review URL: https://codereview.chromium.org/1032253003

9 years agoAcknowledge that SkIcoCodec can rewind and test it.
scroggo [Wed, 1 Apr 2015 17:24:37 +0000 (10:24 -0700)]
Acknowledge that SkIcoCodec can rewind and test it.

Since SkIcoCodec has an SkCodec for its encoded images, backed by
SkMemoryStreams, the SkMemoryStream can always rewind, and will be
rewound by the sub-codec if necessary (now that SkBmpCodec and
SkPngCodec support rewinding).

Depends on https://codereview.chromium.org/1057483003/ and
https://codereview.chromium.org/1048423003/ (DIFFERENT ISSUES).

BUG=skia:3257

Review URL: https://codereview.chromium.org/1054603002

9 years agoMake SkPngCodec support rewinding properly.
scroggo [Wed, 1 Apr 2015 16:33:23 +0000 (09:33 -0700)]
Make SkPngCodec support rewinding properly.

Separate out the code for reading the header, and use it to reinitialize
fPng_ptr and fInfo_ptr after a rewind.

Use common code to clean up fPng_ptr and fInfo_ptr, and set them to
NULL and treat them as NULL as appropriate.

Update the test to expect SkPngCodec to succeed.

BUG=skia:3257

Review URL: https://codereview.chromium.org/1048423003

9 years agoFind the HASHTAGS file even if you are not at the checkout root.
rmistry [Wed, 1 Apr 2015 16:12:51 +0000 (09:12 -0700)]
Find the HASHTAGS file even if you are not at the checkout root.
The post upload hook was failing if you were not at the root, this fixes it.

NOTRY=true

Review URL: https://codereview.chromium.org/1058483002

9 years agosmall-object optimization for SkFunction
mtklein [Wed, 1 Apr 2015 15:11:16 +0000 (08:11 -0700)]
small-object optimization for SkFunction

Anything <= sizeof(void*) will be inlined, avoiding heap allocation.

BUG=skia:

Review URL: https://codereview.chromium.org/1048243002

9 years agoHandle rewinds in SkBmpCodec.
scroggo [Wed, 1 Apr 2015 14:39:40 +0000 (07:39 -0700)]
Handle rewinds in SkBmpCodec.

Factor our BMP code for reading the header, and call it after a rewind.

BUG=skia:3257

Review URL: https://codereview.chromium.org/1057483003

9 years agoCreating a new wrapper for gif decoder
msarett [Wed, 1 Apr 2015 13:58:48 +0000 (06:58 -0700)]
Creating a new wrapper for gif decoder

BUG=skia:3257
BUG=skia:3534

Review URL: https://codereview.chromium.org/1022673011

9 years agoGrGLInterface: Add support for NV_framebuffer_mixed_samples
vbuzinov [Wed, 1 Apr 2015 13:29:49 +0000 (06:29 -0700)]
GrGLInterface: Add support for NV_framebuffer_mixed_samples

Import glCoverageModulation if NV_framebuffer_mixed samples
is available

BUG=skia:3177

Review URL: https://codereview.chromium.org/993363002

9 years agoRevert of Implement approx-match support in image filter saveLayer() offscreen. ...
rmistry [Wed, 1 Apr 2015 11:19:44 +0000 (04:19 -0700)]
Revert of Implement approx-match support in image filter saveLayer() offscreen. (patchset #27 id:510001 of https://codereview.chromium.org/1034733002/)

Reason for revert:
Looks like this change is causing layout test failures which is blocking Skia's DEPS roll into Chromium:
https://codereview.chromium.org/1050563002/
https://codereview.chromium.org/1043133005/
https://codereview.chromium.org/1048273002/

Reverting to see if this fixes the DEPS roll.

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) 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.
>
> 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

TBR=bsalomon@google.com,reed@chromium.org,senorblanco@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:3532

Review URL: https://codereview.chromium.org/1057443003

9 years agoSketch SkFunction
mtklein [Tue, 31 Mar 2015 21:24:27 +0000 (14:24 -0700)]
Sketch SkFunction

Let's start with baby steps in case some bot can't handle this.
I have left many TODOs, most of which I know how to do if this
looks feasible and useful.

BUG=skia:

Review URL: https://codereview.chromium.org/1049223003

9 years agoBitmapTextBatch and BitmapTextBlob
joshualitt [Tue, 31 Mar 2015 20:32:05 +0000 (13:32 -0700)]
BitmapTextBatch and BitmapTextBlob

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/eed1dae04932483579b02c10f0706127d3f5d984

Review URL: https://codereview.chromium.org/1011403004

9 years agoSkPDF: style nit
halcanary [Tue, 31 Mar 2015 19:44:47 +0000 (12:44 -0700)]
SkPDF: style nit

TBR=mtklein@google.com

Review URL: https://codereview.chromium.org/1039313004

9 years agoSkPDF: Factor SkPDFCatalog into SkPDFObjNumMap and SkPDFSubstituteMap
halcanary [Tue, 31 Mar 2015 19:30:20 +0000 (12:30 -0700)]
SkPDF: Factor SkPDFCatalog into SkPDFObjNumMap and SkPDFSubstituteMap

Motivation: Keep separate features separate.  Also, future
linearization work will need to have several objNumMap
objects share a substituteMap.  Also "catalog" has a
specific meaning in PDF.  This catalog did not map to that
catalog.

-   Modify SkPDFObject::emitObject and SkPDFObject::addResources
    interface to requiore SkPDFObjNumMap and SkPDFSubstituteMap.
-   SkPDFObjNumMap const in SkPDFObject::emitObject.
-   Remove SkPDFCatalog.cpp/.h
-   Modify SkDocument_PDF.cpp to use new functions
-   Fold in SkPDFStream::populate
-   Fold in SkPDFBitmap::emitDict
-   Move SkPDFObjNumMap and SkPDFSubstituteMap to SkPDFTypes.h
-   Note (via assert) that SkPDFArray & SkPDFDict don't need to
    check substitutes.
-   Remove extra space from SkPDFDict serialization.
-   SkPDFBitmap SkPDFType0Font SkPDFGraphicState SkPDFStream
    updated to new interface.
-   PDFPrimitivesTest updated for new interface.

BUG=skia:3585

Review URL: https://codereview.chromium.org/1049753002

9 years agoskia: Fix text subpixel half sample default for GPU rasterization
hendrikw [Tue, 31 Mar 2015 19:02:48 +0000 (12:02 -0700)]
skia: Fix text subpixel half sample default for GPU rasterization

In software, SkDraw1Glyph::Proc SkDraw1Glyph::init initializes
fHalfSampleX, fHalfSampleY to SkGlyph::kSubpixelRound.  In hardware,
however, these values were initialized to zero.

Fix, match software's implementation

BUG=471919

Review URL: https://codereview.chromium.org/1047803004

9 years agoRevert of BitmapTextBatch and BitmapTextBlob (patchset #18 id:360001 of https://coder...
joshualitt [Tue, 31 Mar 2015 18:33:08 +0000 (11:33 -0700)]
Revert of BitmapTextBatch and BitmapTextBlob (patchset #18 id:360001 of https://codereview.chromium.org/1011403004/)

Reason for revert:
Breaks a unit test on mac

Original issue's description:
> BitmapTextBatch and BitmapTextBlob
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/eed1dae04932483579b02c10f0706127d3f5d984

TBR=fmalita@chromium.org,reed@google.com,jvanverth@google.com,robertphillips@google.com,bsalomon@google.com,jvanverth@chromium.org,joshualitt@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1050633002

9 years agoBitmapTextBatch and BitmapTextBlob
joshualitt [Tue, 31 Mar 2015 18:04:53 +0000 (11:04 -0700)]
BitmapTextBatch and BitmapTextBlob

BUG=skia:

Review URL: https://codereview.chromium.org/1011403004

9 years agoTurn on Chromium's top 25 page set
rmistry [Tue, 31 Mar 2015 17:59:15 +0000 (10:59 -0700)]
Turn on Chromium's top 25 page set

 #crskps

BUG=skia:
BUG=skia:3574
NOTRY=true
TBR=borenet

Review URL: https://codereview.chromium.org/1045153004

9 years agoImplement approx-match support in image filter saveLayer() offscreen.
senorblanco [Tue, 31 Mar 2015 15:35:15 +0000 (08:35 -0700)]
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) 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.

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

Review URL: https://codereview.chromium.org/1034733002

9 years agoSkPDF: remove SK_NO_FLATE & dead code in SkPDFStream
halcanary [Tue, 31 Mar 2015 15:22:01 +0000 (08:22 -0700)]
SkPDF: remove SK_NO_FLATE & dead code in SkPDFStream

    SkPDFStream copy constructor
    SkPDFStream Substitute mechanism
    SkPDFStream::setData(NULL);
    SkPDFStream SK_NO_FLATE logic

BUG=skia:3585
TBR=bsalomon@google.com,reed@google.com

Review URL: https://codereview.chromium.org/1041183002

9 years agoback to Sk4f for SkPMColor
mtklein [Tue, 31 Mar 2015 15:17:00 +0000 (08:17 -0700)]
back to Sk4f for SkPMColor
 #floats

BUG=skia:
BUG=skia:3592

Review URL: https://codereview.chromium.org/1047823002

9 years agoFix cr SKPs naming
rmistry [Tue, 31 Mar 2015 15:03:02 +0000 (08:03 -0700)]
Fix cr SKPs naming

 #crskps

BUG=skia:
BUG=skia:3574

NOTRY=true

TBR=borenet

Review URL: https://codereview.chromium.org/1044193003

9 years agoTurn on key mobile sites cr page set and delete dashed desktop
rmistry [Tue, 31 Mar 2015 13:22:55 +0000 (06:22 -0700)]
Turn on key mobile sites cr page set and delete dashed desktop

 #crskps
BUG=skia:3574
BUG=skia:3605

NOTRY=true

Review URL: https://codereview.chromium.org/1040233004

9 years agoFix gs.upload_dir_contents arguments mismatch
rmistry [Tue, 31 Mar 2015 12:20:12 +0000 (05:20 -0700)]
Fix gs.upload_dir_contents arguments mismatch

Fixing bug introduced by https://codereview.chromium.org/737123005

BUG=skia:
NOTRY=true
TBR=borenet

Review URL: https://codereview.chromium.org/1041003003

9 years agoAdd ability to capture SKPs from Chromium's pagesets
rmistry [Tue, 31 Mar 2015 12:04:29 +0000 (05:04 -0700)]
Add ability to capture SKPs from Chromium's pagesets

Also added support for pagesets with multiple pages in them.

 #crskps

BUG=skia:1723
BUG=skia:3574

NOTRY=true

Review URL: https://codereview.chromium.org/1047703003

9 years agowebpages_playback.py: Support storing to local directories
kkinnunen [Tue, 31 Mar 2015 07:18:26 +0000 (00:18 -0700)]
webpages_playback.py: Support storing to local directories

Make webpages_playback support storing to local directories as an
alternative to Google Storage.

Rename argument --dest_gsbase to --data_store=<location>. Make this parameter
accept either gs:// url or directory for local file system storage.

Rename argument --upload_to_gs to --upload.

Implement file existence check, download and upload for the file
operation.
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/737123005

9 years agoclamp matrix-translate before converting to pmcolor
reed [Mon, 30 Mar 2015 20:40:30 +0000 (13:40 -0700)]
clamp matrix-translate before converting to pmcolor

.. thanks to https://codereview.chromium.org/1032593003/ layout failures

BUG=skia:

Review URL: https://codereview.chromium.org/1041203004

9 years agoAdd option to embed font data into executable.
bungeman [Mon, 30 Mar 2015 19:53:48 +0000 (12:53 -0700)]
Add option to embed font data into executable.

Some tools would like to be built with all resources embedded.
This change makes it possible to build a font manager which
uses font data embedded into the executable.

Review URL: https://codereview.chromium.org/1015723004

9 years agoSilence giflib warning on mac
msarett [Mon, 30 Mar 2015 18:59:36 +0000 (11:59 -0700)]
Silence giflib warning on mac

BUG=skia:3601

Review URL: https://codereview.chromium.org/1046923004

9 years agoRefactor Sk2x<T> + Sk4x<T> into SkNf<N,T> and SkNi<N,T>
mtklein [Mon, 30 Mar 2015 17:50:27 +0000 (10:50 -0700)]
Refactor Sk2x<T> + Sk4x<T> into SkNf<N,T> and SkNi<N,T>

The primary feature this delivers is SkNf and SkNd for arbitrary power-of-two N.  Non-specialized types or types larger than 128 bits should now Just Work (and we can drop in a specialization to make them faster).  Sk4s is now just a typedef for SkNf<4, SkScalar>; Sk4d is SkNf<4, double>, Sk2f SkNf<2, float>, etc.

This also makes implementing new specializations easier and more encapsulated.  We're now using template specialization, which means the specialized versions don't have to leak out so much from SkNx_sse.h  and SkNx_neon.h.

This design leaves us room to grow up, e.g to SkNf<8, SkScalar> == Sk8s, and to grown down too, to things like SkNi<8, uint16_t> == Sk8h.

To simplify things, I've stripped away most APIs (swizzles, casts, reinterpret_casts) that no one's using yet.  I will happily add them back if they seem useful.

You shouldn't feel bad about using any of the typedef Sk4s, Sk4f, Sk4d, Sk2s, Sk2f, Sk2d, Sk4i, etc.  Here's how you should feel:
  - Sk4f, Sk4s, Sk2d: feel awesome
  - Sk2f, Sk2s, Sk4d: feel pretty good

No public API changes.
TBR=reed@google.com

BUG=skia:3592

Review URL: https://codereview.chromium.org/1048593002

9 years agoSmall change to move GrProcessor and GrBatch pools over to SkSpinlock
joshualitt [Mon, 30 Mar 2015 16:53:47 +0000 (09:53 -0700)]
Small change to move GrProcessor and GrBatch pools over to SkSpinlock

BUG=skia:

Review URL: https://codereview.chromium.org/1040133002

9 years agoExtract the spinlock from SkOnce as SkSpinlock.
mtklein [Mon, 30 Mar 2015 15:13:33 +0000 (08:13 -0700)]
Extract the spinlock from SkOnce as SkSpinlock.

This uses slightly newer APIs from SkAtomics.h to make it a bit more efficient.

No public API changes.
TBR=reed@google.com

BUG=skia:

Review URL: https://codereview.chromium.org/1039323002

9 years agoRevert of Remove SkClipStack's manual rounding of BW clip rects (patchset #3 id:80001...
robertphillips [Mon, 30 Mar 2015 15:09:58 +0000 (08:09 -0700)]
Revert of Remove SkClipStack's manual rounding of BW clip rects (patchset #3 id:80001 of https://codereview.chromium.org/1033453003/)

Reason for revert:
Reverting due to performance regression: https://code.google.com/p/skia/issues/detail?id=3597

Original issue's description:
> Remove SkClipStack's manual rounding of BW clip rects
>
> The full fix for this bug is nudging the image in device space. That is going to be a large change. This CL should address the immediate problem.
>
> This CL will alter the following GMs:
> clipdrawdraw
> convex_poly_clip
> complexclip_bw_*
> filltypespersp
> complexclip3_simple
>
>
>
> BUG=423834
>
> Committed: https://skia.googlesource.com/skia/+/e523d4f90c3368c555282a98b41ca5ee2045103e

TBR=bsalomon@google.com,joshualitt@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=423834

Review URL: https://codereview.chromium.org/1045853002

9 years agoFix variable names in xfermodes2 gm test
cdalton [Mon, 30 Mar 2015 14:52:57 +0000 (07:52 -0700)]
Fix variable names in xfermodes2 gm test

BUG=skia:

Review URL: https://codereview.chromium.org/1041913002

9 years agoWIP: Added support for giflib, updated jpeg and png
msarett [Mon, 30 Mar 2015 14:52:52 +0000 (07:52 -0700)]
WIP: Added support for giflib, updated jpeg and png

BUG=skia:3257

Committed: https://skia.googlesource.com/skia/+/255dcd11992ebe74eb54202c48cf5394d33a8ce6

Review URL: https://codereview.chromium.org/1038863003

9 years agoAdd hashtag for the new skps bug
rmistry [Mon, 30 Mar 2015 13:27:21 +0000 (06:27 -0700)]
Add hashtag for the new skps bug

 #crskps

BUG=skia:3574
NOTRY=true
NOTREECHECKS=true
TBR=borenet

Review URL: https://codereview.chromium.org/1042873002

9 years agoRevert of WIP: Added support for giflib, updated jpeg and png (patchset #8 id:260001...
borenet [Mon, 30 Mar 2015 12:51:49 +0000 (05:51 -0700)]
Revert of WIP: Added support for giflib, updated jpeg and png (patchset #8 id:260001 of https://codereview.chromium.org/1038863003/)

Reason for revert:
Trying out revert to see if it fixes Android bots.

Original issue's description:
> WIP: Added support for giflib, updated jpeg and png
>
> BUG=skia:3257
>
> Committed: https://skia.googlesource.com/skia/+/255dcd11992ebe74eb54202c48cf5394d33a8ce6

TBR=djsollen@google.com,scroggo@google.com,msarett@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:3257

Review URL: https://codereview.chromium.org/1048713003

9 years agowebpages_playback.py: Avoid crash when retrying to capture skp
kkinnunen [Mon, 30 Mar 2015 05:33:16 +0000 (22:33 -0700)]
webpages_playback.py: Avoid crash when retrying to capture skp

Move the .wpr files after recording, in the recording retry loop.
Previously they were moved in the skp capture retry loop.

Review URL: https://codereview.chromium.org/1033223005

9 years agouse Sk4f for matrix math
reed [Sun, 29 Mar 2015 18:58:48 +0000 (11:58 -0700)]
use Sk4f for matrix math

Need to land SK_SUPPORT_LEGACY_SCALAR_MAPPOINTS in chrome to suppress Affine
version which causes slight differences (which will need to be rebaselined)

BUG=skia:

Review URL: https://codereview.chromium.org/1045493002

9 years agoUpdate SKP version
skia.buildbots [Sun, 29 Mar 2015 08:19:04 +0000 (01:19 -0700)]
Update SKP version

Automatic commit by the RecreateSKPs bot.

TBR=

Review URL: https://codereview.chromium.org/1046623002

9 years agoRevert "Implement approx-match support in image filter saveLayer() offscreen."
senorblanco [Sat, 28 Mar 2015 21:50:05 +0000 (14:50 -0700)]
Revert "Implement approx-match support in image filter saveLayer() offscreen."

This reverts commit b97dafefe63ea0a1bbce8e8b209f4920983fb8b9.

SkLightingImageFilter boundaries are incorrect (see GM:lighting).

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1048583002

9 years agoImplement approx-match support in image filter saveLayer() offscreen.
senorblanco [Sat, 28 Mar 2015 20:43:14 +0000 (13:43 -0700)]
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, more TBD) whose GPU
implementation depends on 0..1 texture coordinates.
3) Remove the exception for GPU-side image filters in
SkCanvas::internalSaveLayer().

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

Review URL: https://codereview.chromium.org/1034733002

9 years agouse table of procs (and unrolling) to speed up mapPts
reed [Fri, 27 Mar 2015 21:00:41 +0000 (14:00 -0700)]
use table of procs (and unrolling) to speed up mapPts

BUG=skia:

Review URL: https://codereview.chromium.org/1040783002

9 years agoRemove SkMemoryStream::peek()
scroggo [Fri, 27 Mar 2015 20:27:51 +0000 (13:27 -0700)]
Remove SkMemoryStream::peek()

I'd like to add a new API to SkStream for peeking - i.e. reading some
bytes without advancing the stream. This will be implemented for the
streams where it makes sense. I think the function should look
something like the following:

size_t peek(void* buffer, size_t bytesToRead) {
     return this->onPeek(buffer, bytesToRead);
}

virtual size_t onPeek(void* buffer, size_t bytesToRead) {
    return 0;   // unimplemented base class.
}

In order to avoid confusion, I'd like to remove SkMemoryStream::peek(),
which is not currently used internally, by Chrome, or by Android as far
as I can tell. There is also another function does the same thing:
getPosition().

BUG=skia:3257

Review URL: https://codereview.chromium.org/1039373002

9 years agoUse a black background in imagefiltersgraph so 8888 and 565 look similar.
mtklein [Fri, 27 Mar 2015 19:33:46 +0000 (12:33 -0700)]
Use a black background in imagefiltersgraph so 8888 and 565 look similar.

BUG=skia:

Review URL: https://codereview.chromium.org/1041773003

9 years agoHousekeeping to rename GrGL compressed texture formats to match GL specs
egdaniel [Fri, 27 Mar 2015 19:22:10 +0000 (12:22 -0700)]
Housekeeping to rename GrGL compressed texture formats to match GL specs

BUG=skia:

Review URL: https://codereview.chromium.org/1035243002

9 years agoMove HWUI boilerplate into utils/android
tomhudson [Fri, 27 Mar 2015 19:22:01 +0000 (12:22 -0700)]
Move HWUI boilerplate into utils/android

Duplicate code from the HWUI backends for DM and nanobench
moves into a single place, saving a hundred lines or more of
cut-and-paste.

There's some indication that this increases the incidence of
SkCanvas "Unable to find device for layer." warnings, but no
clear degradation in test results.

R=djsollen@google.com,mtklein@google.com
BUG=skia:3589

Review URL: https://codereview.chromium.org/1036303002

9 years agoWIP: Added support for giflib, updated jpeg and png
msarett [Fri, 27 Mar 2015 19:17:00 +0000 (12:17 -0700)]
WIP: Added support for giflib, updated jpeg and png

BUG=skia:3257

Review URL: https://codereview.chromium.org/1038863003

9 years agoSkCodec: add wbmp class
halcanary [Fri, 27 Mar 2015 19:16:53 +0000 (12:16 -0700)]
SkCodec: add wbmp class

Review URL: https://codereview.chromium.org/1006583005

9 years agotools: add sync-and-gyp script
halcanary [Fri, 27 Mar 2015 19:11:49 +0000 (12:11 -0700)]
tools:  add sync-and-gyp script

NOTRY=true

Review URL: https://codereview.chromium.org/1035003004

9 years agoMinor cleanup in nanobench
tomhudson [Fri, 27 Mar 2015 19:11:44 +0000 (12:11 -0700)]
Minor cleanup in nanobench

Simplify time() by removing conditionals; reduce the amount of
parameter passing.
Add a convenience function to Target.

R=mtklein@google.com
BUG=skia:3595

Review URL: https://codereview.chromium.org/1039253002

9 years agoWhitespace change to see if Android bots are fixed
borenet [Fri, 27 Mar 2015 16:46:25 +0000 (09:46 -0700)]
Whitespace change to see if Android bots are fixed

TBR=rmistry
NOTREECHECKS=true
BUG=skia:2073

Review URL: https://codereview.chromium.org/1039023003

9 years agoUpdate [dm|nanobench]_flags for CPU/GPU split
borenet [Fri, 27 Mar 2015 12:42:18 +0000 (05:42 -0700)]
Update [dm|nanobench]_flags for CPU/GPU split

TBR=mtklein
NOTREECHECKS=true
BUG=skia:2073

Review URL: https://codereview.chromium.org/1036223002

9 years agoimpl SkConvertQuadToCubic w/ Sk2s
reed [Fri, 27 Mar 2015 03:22:33 +0000 (20:22 -0700)]
impl SkConvertQuadToCubic w/ Sk2s

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1041573002

9 years agoAdd matrix constructing helpers to SkMatrix
robertphillips [Fri, 27 Mar 2015 02:57:08 +0000 (19:57 -0700)]
Add matrix constructing helpers to SkMatrix

Review URL: https://codereview.chromium.org/1034273002

9 years agoSkPDF fix object counting error
halcanary [Thu, 26 Mar 2015 22:35:18 +0000 (15:35 -0700)]
SkPDF fix object counting error

Review URL: https://codereview.chromium.org/1012483003

9 years agoRemove some validation and asserts from tessellating path renderer.
senorblanco [Thu, 26 Mar 2015 21:52:45 +0000 (14:52 -0700)]
Remove some validation and asserts from tessellating path renderer.

In some cases, resolving an intersection can cause a vertex to go
slightly out-of-order with edges which have already been processed.
This doesn't cause any algorithmic errors, but it's difficult to detect
without impacting performance significantly.

Also, the GPU infrastructure fires asserts when attempting
to allocate 0-length vertex buffers. Early-out instead,
since there's nothing to draw.

Review URL: https://codereview.chromium.org/1032253005

9 years agoFix DiscardableMemoryPool::free race condition
halcanary [Thu, 26 Mar 2015 21:08:56 +0000 (14:08 -0700)]
Fix DiscardableMemoryPool::free race condition

BUG=skia:3596

Review URL: https://codereview.chromium.org/1017943003

9 years agoRevert of Make the canvas draw looper setup update the canvas save count (patchset...
reed [Thu, 26 Mar 2015 20:40:09 +0000 (13:40 -0700)]
Revert of Make the canvas draw looper setup update the canvas save count (patchset #1 id:1 of https://codereview.chromium.org/1034033004/)

Reason for revert:
makes internalSave and internalSaveLayer inconsistent. Need to find a different solution.

Original issue's description:
> Make the canvas draw looper setup update the canvas save count
>
> Image filter in a paint would leave save count in wrong state
> for normal draws. This could be observed through the canvas
> references during the draw call. An example of this is
> inspecting the canvas during a draw looper.
>
> patch from issue 993863002 at patchset 20001 (http://crrev.com/993863002#ps20001)
>
> BUG=skia:
> TBR=kkinnunen@nvidia.com
>
> Committed: https://skia.googlesource.com/skia/+/fd3a91e1fc4de69611b5297f624a1cd65db4ced1

TBR=kkinnunen@nvidia.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1037653004

9 years agoMake the canvas draw looper setup update the canvas save count
reed [Thu, 26 Mar 2015 20:29:56 +0000 (13:29 -0700)]
Make the canvas draw looper setup update the canvas save count

Image filter in a paint would leave save count in wrong state
for normal draws. This could be observed through the canvas
references during the draw call. An example of this is
inspecting the canvas during a draw looper.

patch from issue 993863002 at patchset 20001 (http://crrev.com/993863002#ps20001)

BUG=skia:
TBR=kkinnunen@nvidia.com

Review URL: https://codereview.chromium.org/1034033004

9 years agoSkPMFloat::trunc()
mtklein [Thu, 26 Mar 2015 19:32:29 +0000 (12:32 -0700)]
SkPMFloat::trunc()

Add and test trunc(), which is what get() used to be before rounding.
Using trunc() is a ~40% speedup on our linear gradient bench.

#neon #floats
BUG=skia:3592
#n5
#n9
CQ_INCLUDE_TRYBOTS=client.skia.android:Test-Android-Nexus5-Adreno330-Arm7-Debug-Trybot;client.skia.android:Test-Android-Nexus9-TegraK1-Arm64-Release-Trybot

Review URL: https://codereview.chromium.org/1032243002

9 years agoAndroid HWUI backend Nanobench
tomhudson [Thu, 26 Mar 2015 18:28:06 +0000 (11:28 -0700)]
Android HWUI backend Nanobench

Uses filtering canvas from utils/android, shared with DM.
Follow-up plans in https://skbug.com/3589, https://skbug.com/3595

R=djsollen@google.com

Review URL: https://codereview.chromium.org/1029423010

9 years agosmall fix for nanobench segfault when not running any tests
joshualitt [Thu, 26 Mar 2015 17:41:02 +0000 (10:41 -0700)]
small fix for nanobench segfault when not running any tests

BUG=skia:

Review URL: https://codereview.chromium.org/1030353004

9 years agoRemove unused HWAA flag and uniqueID field from GrDrawTargetCaps.
egdaniel [Thu, 26 Mar 2015 17:13:05 +0000 (10:13 -0700)]
Remove unused HWAA flag and uniqueID field from GrDrawTargetCaps.

BUG=skia:

Review URL: https://codereview.chromium.org/1019303005

9 years agoDebugger: remove dead feature (SkPicture offset display) & fix bug (unbalanced indents)
robertphillips [Thu, 26 Mar 2015 17:08:04 +0000 (10:08 -0700)]
Debugger: remove dead feature (SkPicture offset display) & fix bug (unbalanced indents)

Displaying the offset into an SkPicture hasn't worked for a while so this CL deletes the feature.
When "Save Layer" was renamed to "SaveLayer" the code that computes the indent in the list view was broken. This CL patches the problem.

Review URL: https://codereview.chromium.org/1034733004

9 years agoSilence PNG warnings from SkPngCodec.
scroggo [Thu, 26 Mar 2015 17:07:56 +0000 (10:07 -0700)]
Silence PNG warnings from SkPngCodec.

Review URL: https://codereview.chromium.org/1009633005

9 years agouse new faster/vector impl for chopping conics
reed [Thu, 26 Mar 2015 16:10:22 +0000 (09:10 -0700)]
use new faster/vector impl for chopping conics

BUG=skia:

Review URL: https://codereview.chromium.org/1035943002

9 years agoAdd support for using alternative backends (like DirectX) when creating a GrGpu.
egdaniel [Thu, 26 Mar 2015 16:09:41 +0000 (09:09 -0700)]
Add support for using alternative backends (like DirectX) when creating a GrGpu.

BUG=skia:

Review URL: https://codereview.chromium.org/1038643002

9 years agoadd legacy pathop enums; fix uninitialized warning
caryclark [Thu, 26 Mar 2015 16:05:12 +0000 (09:05 -0700)]
add legacy pathop enums; fix uninitialized warning

R=reed@google.com
BUG=skia:3588
NOTREECHECKS=true
NOTRY=true
NOPRESUBMIT=true

Review URL: https://codereview.chromium.org/1034073004

9 years agocumulative pathops patch
caryclark [Thu, 26 Mar 2015 14:52:43 +0000 (07:52 -0700)]
cumulative pathops patch

Replace the implicit curve intersection with a geometric curve intersection. The implicit intersection proved mathematically unstable and took a long time to zero in on an answer.

Use pointers instead of indices to refer to parts of curves. Indices required awkward renumbering.

Unify t and point values so that small intervals can be eliminated in one pass.

Break cubics up front to eliminate loops and cusps.

Make the Simplify and Op code more regular and eliminate arbitrary differences.

Add a builder that takes an array of paths and operators.

Delete unused code.

BUG=skia:3588
R=reed@google.com

Review URL: https://codereview.chromium.org/1037573004