platform/upstream/libSkiaSharp.git
9 years agoAdd Load2/store2 to Sk4x (dumb impl for now)
reed [Wed, 18 Mar 2015 19:56:46 +0000 (12:56 -0700)]
Add Load2/store2 to Sk4x (dumb impl for now)

patch from issue 1001003002 at patchset 1 (http://crrev.com/1001003002#ps1)

BUG=skia:

Review URL: https://codereview.chromium.org/1001453006

9 years agoRemove uniqueID from all filter serialization.
senorblanco [Wed, 18 Mar 2015 19:11:33 +0000 (12:11 -0700)]
Remove uniqueID from all filter serialization.

(This is essentially a revert of https://codereview.chromium.org/503833002/.)

This was necessary back when SkPaint was flattened even for in-process use. Now that we only flatten SkPaint for cross-process use, there's no need to serialize UniqueIDs.

Note: SkDropShadowImageFilter is being constructed with a croprect and UniqueID (of 0) in Blink. I've made the uniqueID param default to 0 temporarily, until this rolls in and Blink can be changed. (Blink can't be changed first, since unlike the other filters, there's no constructor that takes a cropRect but not a uniqueID.)

BUG=skia:

Review URL: https://codereview.chromium.org/1019493002

9 years agoextend neon table as well for gray8
reed [Wed, 18 Mar 2015 18:52:35 +0000 (11:52 -0700)]
extend neon table as well for gray8

BUG=skia:

Review URL: https://codereview.chromium.org/1021583002

9 years agoSimple patch to have draw path call an internal drawPath
joshualitt [Wed, 18 Mar 2015 18:49:12 +0000 (11:49 -0700)]
Simple patch to have draw path call an internal drawPath

The purpose of this patch is to expose an internal draw path call which text contexts can use.

BUG=skia:

Review URL: https://codereview.chromium.org/1012143005

9 years agoLet's find out what's up with the TSAN bot test failures.
mtklein [Wed, 18 Mar 2015 18:32:21 +0000 (11:32 -0700)]
Let's find out what's up with the TSAN bot test failures.

BUG=skia:

Review URL: https://codereview.chromium.org/1020573002

9 years agoAdding swizzles for bmp:
msarett [Wed, 18 Mar 2015 18:11:19 +0000 (11:11 -0700)]
Adding swizzles for bmp:
    We now support kN32 and kRGB_565 color types.
    Additionally, we support premul, unpremul, and opaque alpha types.
    Unpremul is currently untested as we cannot currently draw to unpremul.

BUG=skia:

Review URL: https://codereview.chromium.org/1013743003

9 years agoUpdate to identically match the sources used on Android
djsollen [Wed, 18 Mar 2015 17:58:02 +0000 (10:58 -0700)]
Update to identically match the sources used on Android

Review URL: https://codereview.chromium.org/1020563002

9 years agoImplement SkCodec::getScaledDimensions.
scroggo [Wed, 18 Mar 2015 17:50:37 +0000 (10:50 -0700)]
Implement SkCodec::getScaledDimensions.

Left out of a prior CL since no one was calling it...

Review URL: https://codereview.chromium.org/995303004

9 years agoSkPMFloat: avoid loads and stores where possible.
mtklein [Wed, 18 Mar 2015 16:51:23 +0000 (09:51 -0700)]
SkPMFloat: avoid loads and stores where possible.

A store/load pair like this is a redundant no-op:
  store simd_register_a, memory_address
  load  memory_address, simd_register_a

Everyone seems to be good at removing those when using SSE, but GCC and Clang
are pretty terrible at this for NEON.  We end up issuing both redundant
commands, usually to and from the stack.  That's slow.  Let's not do that.

This CL unions in the native SIMD register type into SkPMFloat, so that we can
assign to and from it directly, which is generating a lot better NEON code.  On
my Nexus 5, the benchmarks improve from 36ns to 23ns.

SSE is just as fast either way, but I paralleled the NEON code for consistency.
It's a little terser.  And because it needed the platform headers anyway, I
moved all includes into SkPMFloat.h, again only for consistency.

I'd union in Sk4f too to make its conversion methods a little clearer,
but MSVC won't let me (it has a copy constructor... they're apparently not up
to speed with C++11 unrestricted unions).

BUG=skia:

Review URL: https://codereview.chromium.org/1015083004

9 years agoRestore GDI text size rounding.
bungeman [Wed, 18 Mar 2015 15:25:43 +0000 (08:25 -0700)]
Restore GDI text size rounding.

Add call to SkScalarRoundToScalar(). The old code calculated the scale
from the text size, but now the text size is calculated from the scale
(which is arguably the right way to think about it). However, the old
code always rounded the final resulting text size, while the new code
does not.

In the 'no hinting' case, the text size is already rounded to an integer
(so that the rest of the matrix is minimized). In the 'hinted' case, the
entire scale has been removed from the matrix, so the scale value is the
'real' residual size. The old code rounded this size, and the new code
should as well.

BUG=464784

Review URL: https://codereview.chromium.org/1014953002

9 years agoAdd comment describing SkColorTable constructor.
scroggo [Wed, 18 Mar 2015 15:00:27 +0000 (08:00 -0700)]
Add comment describing SkColorTable constructor.

Review URL: https://codereview.chromium.org/1011263002

9 years agoIndexed PNG decoding: Ensure color table is large enough that the bit depth of the...
dml [Wed, 18 Mar 2015 13:03:29 +0000 (06:03 -0700)]
Indexed PNG decoding: Ensure color table is large enough that the bit depth of the image will not allow reads beyond its end.

BUG=skia:3440
R=rmistry@google.com, scroggo@google.com

Committed: https://skia.googlesource.com/skia/+/493c1ce1cd406ef28683203146274154783452ce

Review URL: https://codereview.chromium.org/948163002

9 years agoWhitespace change to test new GCE compile bot
Ravi Mistry [Wed, 18 Mar 2015 12:50:30 +0000 (08:50 -0400)]
Whitespace change to test new GCE compile bot

9 years agoDisplay currently running tests as the keepalive message.
mtklein [Wed, 18 Mar 2015 12:27:14 +0000 (05:27 -0700)]
Display currently running tests as the keepalive message.

...
( 466MB   550) 31.9ms   gpu gm colorwheelnative
( 466MB   549) 50.1ms   gpu gm colorwheel
( 468MB   548) 512ms    gpu gm colortype_xfermodes
Currently running:
    unit test PathOpsRectsThreaded
    unit test PathOpsSimplifyDegeneratesThreaded
    unit test PathOpsSimplifyQuadralateralsThreaded
    gpu gm colortype

( 468MB   547) 90.8ms   gpu gm colortype
( 469MB   546) 2.03s    unit test PathOpsSimplifyDegeneratesThreaded
...

BUG=skia:

Review URL: https://codereview.chromium.org/1017903002

9 years agoadd kGray_8_SkColorType
reed [Wed, 18 Mar 2015 00:44:06 +0000 (17:44 -0700)]
add kGray_8_SkColorType

patch from issue 1014783003 at patchset 60001 (http://crrev.com/1014783003#ps60001)

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1010343002

9 years agoImprove tracking of bound FBOs in GrGLGpu.
bsalomon [Tue, 17 Mar 2015 22:55:42 +0000 (15:55 -0700)]
Improve tracking of bound FBOs in GrGLGpu.

Committed: https://skia.googlesource.com/skia/+/d2ad8eb5801e2c8c0fa544a6a776bb46eedde2a0

Committed: https://skia.googlesource.com/skia/+/b2af2d8b83ca4774c3b3bb1e49bc72605faa9589

Committed: https://skia.googlesource.com/skia/+/0b70b86a7e9fda52ee7ebc1b9897eeaa09b9abef

Committed: https://skia.googlesource.com/skia/+/6ba6fa15261be591f33cf0e5df7134e4fc6432ac

Committed: https://skia.googlesource.com/skia/+/dc963b9264908f53650c40a97cff414101dd3e88

Review URL: https://codereview.chromium.org/949263002

9 years agoUse sample count for temp renderbuffer when clearing stenci
Brian Salomon [Tue, 17 Mar 2015 20:03:19 +0000 (16:03 -0400)]
Use sample count for temp renderbuffer when clearing stenci

TBR=robertphillips@google.com

Review URL: https://codereview.chromium.org/1016903002

9 years agoDM: disable image src type on Windows.
mtklein [Tue, 17 Mar 2015 19:52:16 +0000 (12:52 -0700)]
DM: disable image src type on Windows.

BUG=skia:

Review URL: https://codereview.chromium.org/1012183002

9 years agoRevert of Improve tracking of bound FBOs in GrGLGpu. (patchset #16 id:300001 of https...
bsalomon [Tue, 17 Mar 2015 19:46:56 +0000 (12:46 -0700)]
Revert of Improve tracking of bound FBOs in GrGLGpu. (patchset #16 id:300001 of https://codereview.chromium.org/949263002/)

Reason for revert:
locally hit assert.

Original issue's description:
> Improve tracking of bound FBOs in GrGLGpu.
>
> Committed: https://skia.googlesource.com/skia/+/d2ad8eb5801e2c8c0fa544a6a776bb46eedde2a0
>
> Committed: https://skia.googlesource.com/skia/+/b2af2d8b83ca4774c3b3bb1e49bc72605faa9589
>
> Committed: https://skia.googlesource.com/skia/+/0b70b86a7e9fda52ee7ebc1b9897eeaa09b9abef
>
> Committed: https://skia.googlesource.com/skia/+/6ba6fa15261be591f33cf0e5df7134e4fc6432ac
>
> Committed: https://skia.googlesource.com/skia/+/dc963b9264908f53650c40a97cff414101dd3e88

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

Review URL: https://codereview.chromium.org/1017803002

9 years agoImprove tracking of bound FBOs in GrGLGpu.
bsalomon [Tue, 17 Mar 2015 19:33:33 +0000 (12:33 -0700)]
Improve tracking of bound FBOs in GrGLGpu.

Committed: https://skia.googlesource.com/skia/+/d2ad8eb5801e2c8c0fa544a6a776bb46eedde2a0

Committed: https://skia.googlesource.com/skia/+/b2af2d8b83ca4774c3b3bb1e49bc72605faa9589

Committed: https://skia.googlesource.com/skia/+/0b70b86a7e9fda52ee7ebc1b9897eeaa09b9abef

Committed: https://skia.googlesource.com/skia/+/6ba6fa15261be591f33cf0e5df7134e4fc6432ac

Review URL: https://codereview.chromium.org/949263002

9 years agoAdd stencil-only clear bug workaround back but for all devices.
bsalomon [Tue, 17 Mar 2015 19:06:19 +0000 (12:06 -0700)]
Add stencil-only clear bug workaround back but for all devices.

TBR=robertphillips@google.com

Review URL: https://codereview.chromium.org/1010253002

9 years agoCurrent RSS on linux and android too?
mtklein [Tue, 17 Mar 2015 17:24:49 +0000 (10:24 -0700)]
Current RSS on linux and android too?

BUG=skia:

Review URL: https://codereview.chromium.org/1009313003

9 years agoDoc: htmlify roles table to allow easier modification
halcanary [Tue, 17 Mar 2015 16:56:33 +0000 (09:56 -0700)]
Doc:  htmlify roles table to allow easier modification

TODO: make sure the table is up to date.
NOTRY=true
DOCS_PREVIEW= https://skia.org/?cl=1011753004

Review URL: https://codereview.chromium.org/1011753004

9 years agoimagefilters always need to disable lcd
reed [Tue, 17 Mar 2015 15:34:38 +0000 (08:34 -0700)]
imagefilters always need to disable lcd

This prev CL accidentally stopped forcing (indirectly) the pixel-geometry to unknown. https://codereview.chromium.org/1010503002

BUG=467882

Review URL: https://codereview.chromium.org/1012113002

9 years agoRemove SkLONGLONG.
bungeman [Tue, 17 Mar 2015 14:23:39 +0000 (07:23 -0700)]
Remove SkLONGLONG.

All users now define SkLONGLONG.
This fixes a long outstanding TODO now that int64_t is required.

BUG=skia:179

Review URL: https://codereview.chromium.org/1000933003

9 years agoRevert "Indexed PNG decoding: Ensure color table is large enough that the bit depth...
scroggo [Tue, 17 Mar 2015 12:25:54 +0000 (05:25 -0700)]
Revert "Indexed PNG decoding: Ensure color table is large enough that the bit depth of the image will not allow reads beyond its end."

This reverts commit 493c1ce1cd406ef28683203146274154783452ce.

NOTRY=true
NOTREECHECKS=true
TBR=egdaniel@google.com,dml@google.com
BUG=skia:

Review URL: https://codereview.chromium.org/1014553003

9 years agoIndexed PNG decoding: Ensure color table is large enough that the bit depth of the...
Leon Scroggins III [Tue, 17 Mar 2015 12:14:07 +0000 (08:14 -0400)]
Indexed PNG decoding: Ensure color table is large enough that the bit depth of the image will not allow reads beyond its end.

BUG=skia:3440
R=rmistry@google.com, scroggo@google.com

Review URL: https://codereview.chromium.org/948163002

9 years agoOption for SkCodec to treat dst as all zeroes.
scroggo [Tue, 17 Mar 2015 12:02:17 +0000 (05:02 -0700)]
Option for SkCodec to treat dst as all zeroes.

This recreates SkImageDecoder's feature to skip writing zeroes for
SkCodec.

Review URL: https://codereview.chromium.org/980903002

9 years agoDisable GL_MULTISAMPLE when it's not being used
cdalton [Tue, 17 Mar 2015 00:19:58 +0000 (17:19 -0700)]
Disable GL_MULTISAMPLE when it's not being used

Updates Gr to disable GL_MULTISAMPLE when drawing to an aliased render
target. This is an important step for NV_framebuffer_mixed_samples,
where just because the color buffer isn't multisampled doesn't mean
other stages in the graphics pipeline aren't. It also removes the need
for the "isLineDraw" workaround that was in GrGLGpu::flushHWAAState.

BUG=skia:

Review URL: https://codereview.chromium.org/1002013006

9 years agowiden leak supression. looks like start_keepalive is sometimes inlined.
mtklein [Mon, 16 Mar 2015 22:06:00 +0000 (15:06 -0700)]
widen leak supression.  looks like start_keepalive is sometimes inlined.

BUG=skia:

Review URL: https://codereview.chromium.org/1008873003

9 years agoFix build for UCLIBC platforms
smcgruer [Mon, 16 Mar 2015 21:45:00 +0000 (14:45 -0700)]
Fix build for UCLIBC platforms

malloc_usable_size does not exist in UCLIBC, so fall back to just
returning 0 for SkVarAlloc::heap_size().

BUG=skia:

Review URL: https://codereview.chromium.org/1006073003

9 years agoMove GrAutoLocaleSetter to new file and fix issue with null locale
bsalomon [Mon, 16 Mar 2015 21:00:52 +0000 (14:00 -0700)]
Move GrAutoLocaleSetter to new file and fix issue with null locale

TBR=egdaniel@google.com
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/1002623004

9 years agoremove dead code
reed [Mon, 16 Mar 2015 20:56:42 +0000 (13:56 -0700)]
remove dead code

BUG=skia:

Review URL: https://codereview.chromium.org/1016523002

9 years agoremove unneeded flags
reed [Mon, 16 Mar 2015 20:55:48 +0000 (13:55 -0700)]
remove unneeded flags

BUG=skia:

Review URL: https://codereview.chromium.org/1010063002

9 years agoFix up some issues introduced by https://codereview.chromium.org/1013773002/
jvanverth [Mon, 16 Mar 2015 19:58:43 +0000 (12:58 -0700)]
Fix up some issues introduced by https://codereview.chromium.org/1013773002/

Need to check for zero-length gradients when using the new correction.
Make sure we use the old text sizes for color emoji test in dftext GM.

BUG=skia:3540

Review URL: https://codereview.chromium.org/1009973004

9 years agoalways reset locale (not just in debug).
Brian Salomon [Mon, 16 Mar 2015 19:36:44 +0000 (15:36 -0400)]
always reset locale (not just in debug).

BUG=skia:

Review URL: https://codereview.chromium.org/1007283009

9 years agofix ios build wrt locale?
Brian Salomon [Mon, 16 Mar 2015 19:05:23 +0000 (15:05 -0400)]
fix ios build wrt locale?

BUG=skia:3330

Review URL: https://codereview.chromium.org/1012773004

9 years agoUse C locale for numerics when emitting shaders.
bsalomon [Mon, 16 Mar 2015 18:56:29 +0000 (11:56 -0700)]
Use C locale for numerics when emitting shaders.

BUG=skia:3330

Review URL: https://codereview.chromium.org/1012723002

9 years agoRevert "Revert of fix for invalid for loop syntax broke build (patchset #1 id:1 of...
msarett [Mon, 16 Mar 2015 18:55:18 +0000 (11:55 -0700)]
Revert "Revert of fix for invalid for loop syntax broke build (patchset #1 id:1 of https://codereview.chromium.org/1007373003/)"

This reverts commit d18475854ce232de71c5463e0654f459d4abfd43.

Revert "Revert "Implementation of image decoding for bmp files, in accordance with the new API.""

This reverts commit dfdec78a5d02e8690998741a9fe5b71a08ca3232.

BUG=skia:

TBR=

Review URL: https://codereview.chromium.org/1016443003

9 years agoFix SDF gradient calculation for non-uniform xforms
jvanverth [Mon, 16 Mar 2015 18:32:49 +0000 (11:32 -0700)]
Fix SDF gradient calculation for non-uniform xforms

Also adds two non-uniform test cases to the dftext GM.

BUG=skia:3540

Review URL: https://codereview.chromium.org/1013773002

9 years agoRevert "Implementation of image decoding for bmp files, in accordance with the new...
msarett [Mon, 16 Mar 2015 18:07:02 +0000 (11:07 -0700)]
Revert "Implementation of image decoding for bmp files, in accordance with the new API."

This reverts commit 3675874468de7228944ce21922e6ec863f5f2310.

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

Review URL: https://codereview.chromium.org/1012873002

9 years agoRevert of fix for invalid for loop syntax broke build (patchset #1 id:1 of https...
egdaniel [Mon, 16 Mar 2015 17:50:12 +0000 (10:50 -0700)]
Revert of fix for invalid for loop syntax broke build (patchset #1 id:1 of https://codereview.chromium.org/1007373003/)

Reason for revert:
needed to revert early patch

Original issue's description:
> fix for invalid for loop syntax broke build
> NOTREECHECKS=true
> NOTRY=true
> TBR=
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/fdc310855dbf04ffb33787f5c0d76fef90b303d4

TBR=msarett@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1015473002

9 years agoDM: allow SKPs to be smaller than 1000x1000
mtklein [Mon, 16 Mar 2015 17:38:07 +0000 (10:38 -0700)]
DM: allow SKPs to be smaller than 1000x1000

This sniffs the .skp dimensions and intersects them with our 1000x1000 viewport.

This fixes things like desk_carsvg.skp, which is only 902 pixels tall.  In 565 now,
the remaining 98 pixels draw as black, which looks funny and is confusing to triage.

No apparent affect on DM memory usage.  (We're about to map the file anyway.)

BUG=skia:

Review URL: https://codereview.chromium.org/986103002

9 years agoFilterLevel -> SkFilterQuality
reed [Mon, 16 Mar 2015 17:26:13 +0000 (10:26 -0700)]
FilterLevel -> SkFilterQuality

BUG=skia:
TBR=
NOTREECHECKS=True
NOTRY=True

Review URL: https://codereview.chromium.org/1010023002

9 years agoSkPaint::FilterLevel -> SkFilterQuality
reed [Mon, 16 Mar 2015 17:08:34 +0000 (10:08 -0700)]
SkPaint::FilterLevel -> SkFilterQuality

clone (+rebase) of https://codereview.chromium.org/1009183002/

BUG=skia:
TBR=scroggo@google.com

Review URL: https://codereview.chromium.org/1014533004

9 years agoAdd __STDC_LIMIT_MACROS to SkDraw.cpp.
bungeman [Mon, 16 Mar 2015 17:00:09 +0000 (13:00 -0400)]
Add __STDC_LIMIT_MACROS to SkDraw.cpp.

9 years agoTest for glyphs which straddle the edge of device space.
bungeman [Mon, 16 Mar 2015 16:49:54 +0000 (09:49 -0700)]
Test for glyphs which straddle the edge of device space.

If a glyph ends up with a left edge near the extreme right edge of
device space, the right edge may end up at the extreme left. The
current clip tests do not test for this situation and attempt to
draw the glyphs. This results in the blitters never reaching the
right edge of the glyph and so reading from other memory.

BUG=chromium:467011

Review URL: https://codereview.chromium.org/1012763002

9 years agofix for invalid for loop syntax broke build
msarett [Mon, 16 Mar 2015 15:47:55 +0000 (08:47 -0700)]
fix for invalid for loop syntax broke build
NOTREECHECKS=true
NOTRY=true
TBR=
BUG=skia:

Review URL: https://codereview.chromium.org/1007373003

9 years agoImplementation of image decoding for bmp files, in accordance with the new API.
msarett [Mon, 16 Mar 2015 15:27:53 +0000 (08:27 -0700)]
Implementation of image decoding for bmp files, in accordance with the new API.

Currently decodes to opaque and unpremul.

Tested on local test suite.

BUG=skia:3257

Review URL: https://codereview.chromium.org/947283002

9 years agocheck for null-pixelref
reed [Mon, 16 Mar 2015 14:28:12 +0000 (07:28 -0700)]
check for null-pixelref

BUG=skia:3497

Review URL: https://codereview.chromium.org/1004313002

9 years agoremove deprecated setFilterBitmap/isFilterBitmap
reed [Mon, 16 Mar 2015 14:24:05 +0000 (07:24 -0700)]
remove deprecated setFilterBitmap/isFilterBitmap

BUG=skia:

Review URL: https://codereview.chromium.org/1006213002

9 years agoSuppress keepalive thread leak in DM.
mtklein [Mon, 16 Mar 2015 13:59:18 +0000 (06:59 -0700)]
Suppress keepalive thread leak in DM.

BUG=skia:

Review URL: https://codereview.chromium.org/1006143002

9 years agomove kTileModeCount out of TileMode enum
reed [Sun, 15 Mar 2015 21:01:21 +0000 (14:01 -0700)]
move kTileModeCount out of TileMode enum

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1009193002

9 years agoUpdate SKP version
skia.buildbots [Sun, 15 Mar 2015 08:13:36 +0000 (01:13 -0700)]
Update SKP version

Automatic commit by the RecreateSKPs bot.

TBR=

Review URL: https://codereview.chromium.org/1005353003

9 years agoremove legacy onCreateCompatibleDevice
reed [Sun, 15 Mar 2015 00:51:34 +0000 (17:51 -0700)]
remove legacy onCreateCompatibleDevice

needs this to land in chrome first : https://codereview.chromium.org/1013453002/

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1006183002

9 years agoRevert of Improve tracking of bound FBOs in GrGLGpu. (patchset #12 id:220001 of https...
bsalomon [Sat, 14 Mar 2015 19:08:51 +0000 (12:08 -0700)]
Revert of Improve tracking of bound FBOs in GrGLGpu. (patchset #12 id:220001 of https://codereview.chromium.org/949263002/)

Reason for revert:
That one needs to be reverted to fix black boxes with GPU
raster, and GPU bot failures.

BUG=466690

Original issue's description:
> Improve tracking of bound FBOs in GrGLGpu.
>
> Committed: https://skia.googlesource.com/skia/+/d2ad8eb5801e2c8c0fa544a6a776bb46eedde2a0
>
> Committed: https://skia.googlesource.com/skia/+/b2af2d8b83ca4774c3b3bb1e49bc72605faa9589
>
> Committed: https://skia.googlesource.com/skia/+/0b70b86a7e9fda52ee7ebc1b9897eeaa09b9abef
>
> Committed: https://skia.googlesource.com/skia/+/6ba6fa15261be591f33cf0e5df7134e4fc6432ac

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

Review URL: https://codereview.chromium.org/1011493002

9 years agoRevert of Add terrible workaround to fix stencil clear crash on S3. (patchset #1...
ccameron [Sat, 14 Mar 2015 19:06:54 +0000 (12:06 -0700)]
Revert of Add terrible workaround to fix stencil clear crash on S3. (patchset #1 id:1 of https://codereview.chromium.org/1001693003/)

Reason for revert:
Needs to be reverted in order to cleanly revert
https://codereview.chromium.org/949263002

That one needs to be reverted to fix black boxes with GPU
raster, and GPU bot failures.

BUG=466690

Original issue's description:
> Add terrible workaround to fix stencil clear crash on S3.
>
> BUG=skia:3525
>
> Committed: https://skia.googlesource.com/skia/+/084340bb9bbb538baf02a11ac36e29f0d9cdc493

TBR=joshualitt@google.com,bsalomon@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:3525

Review URL: https://codereview.chromium.org/1005263002

9 years agoRevert[6] of Change device creation to see the (optional) layer-paint
reed [Sat, 14 Mar 2015 17:54:31 +0000 (10:54 -0700)]
Revert[6] of Change device creation to see the (optional) layer-paint

This reverts commit 173e5fe5f4be28272246e5676f5d2e5c4b1f9167.

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1010503002

9 years agoSkip absurdly slow PDF tests on Valgrind bot.
mtklein [Sat, 14 Mar 2015 14:24:29 +0000 (07:24 -0700)]
Skip absurdly slow PDF tests on Valgrind bot.

Our Valgrind-with-keepalive CPU bot is still running its first run as I write
this.  It's been going ~48 hours.  'pdf gm fontmgr_iter' finished after ~19
hours.  'pdf image PANO_20121023_214540.jpg' still seems to be running.

After this, the next slowest will be '565 gm fontmgr_iter' at about 37 minutes.

TBR=borenet@google.com

BUG=skia:

Review URL: https://codereview.chromium.org/1003423002

9 years agoadd onCreateDevice to eventually replace onCreateCompatibleDevice
reed [Sat, 14 Mar 2015 11:30:21 +0000 (04:30 -0700)]
add onCreateDevice to eventually replace onCreateCompatibleDevice

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/1006923004

9 years agoRevert of Revert of Revert of Revert of Revert of Change device creation to see the...
reed [Sat, 14 Mar 2015 03:05:18 +0000 (20:05 -0700)]
Revert of Revert of Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1005173004/)

Reason for revert:
arrrg. this is a staging nightmare. override required on the chrome side. must revert (again)

Original issue's description:
> Revert of Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1001423002/)
>
> Reason for revert:
> chrome now has the new virtual, so trying again
>
> Original issue's description:
> > Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1006923002/)
> >
> > Reason for revert:
> > platform_canvas tests failures
> >
> > skia_unittests (with patch) skia_unittests (with patch) PlatformCanvas.TranslateLayer failed 2
> > Flakiness dashboard
> >
> > failures:
> > PlatformCanvas.TranslateLayer
> > PlatformCanvas.FillLayer
> >
> > Original issue's description:
> > > Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1008863002/)
> > >
> > > Reason for revert:
> > > guard in chrome has landed
> > >
> > > Original issue's description:
> > > > Revert of Change device creation to see the (optional) layer-paint (patchset #9 id:160001 of https://codereview.chromium.org/988413003/)
> > > >
> > > > Reason for revert:
> > > > need to have chrome opt-in for the older API before this can land (in chrome)
> > > >
> > > > Original issue's description:
> > > > > Change device creation to see the (optional) layer-paint
> > > > >
> > > > > Motivation:
> > > > >
> > > > > PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.
> > > > >
> > > > > This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.
> > > > >
> > > > > Committed: https://skia.googlesource.com/skia/+/1182d9a96b80bd12183ee7c81325a979a51ee0c0
> > > >
> > > > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > > > NOPRESUBMIT=true
> > > > NOTREECHECKS=true
> > > > NOTRY=true
> > > >
> > > > Committed: https://skia.googlesource.com/skia/+/0e040f7da2fdfeb49aa60d24117306e3b1e6ea90
> > >
> > > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > > NOPRESUBMIT=true
> > > NOTREECHECKS=true
> > > NOTRY=true
> > >
> > > Committed: https://skia.googlesource.com/skia/+/f7076a13e2d4269903b34ef2780e1c84723e4477
> >
> > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > NOPRESUBMIT=true
> > NOTREECHECKS=true
> > NOTRY=true
> >
> > Committed: https://skia.googlesource.com/skia/+/8e14d660b2a434bc708a70180c84210883611683
>
> TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com,reed@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/307d1ed129ff75eb64137dea75df858f9e250b69

TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1005183003

9 years agoRevert of Revert of Revert of Revert of Change device creation to see the (optional...
reed [Sat, 14 Mar 2015 02:52:59 +0000 (19:52 -0700)]
Revert of Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1001423002/)

Reason for revert:
chrome now has the new virtual, so trying again

Original issue's description:
> Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1006923002/)
>
> Reason for revert:
> platform_canvas tests failures
>
> skia_unittests (with patch) skia_unittests (with patch) PlatformCanvas.TranslateLayer failed 2
> Flakiness dashboard
>
> failures:
> PlatformCanvas.TranslateLayer
> PlatformCanvas.FillLayer
>
> Original issue's description:
> > Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1008863002/)
> >
> > Reason for revert:
> > guard in chrome has landed
> >
> > Original issue's description:
> > > Revert of Change device creation to see the (optional) layer-paint (patchset #9 id:160001 of https://codereview.chromium.org/988413003/)
> > >
> > > Reason for revert:
> > > need to have chrome opt-in for the older API before this can land (in chrome)
> > >
> > > Original issue's description:
> > > > Change device creation to see the (optional) layer-paint
> > > >
> > > > Motivation:
> > > >
> > > > PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.
> > > >
> > > > This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.
> > > >
> > > > Committed: https://skia.googlesource.com/skia/+/1182d9a96b80bd12183ee7c81325a979a51ee0c0
> > >
> > > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > > NOPRESUBMIT=true
> > > NOTREECHECKS=true
> > > NOTRY=true
> > >
> > > Committed: https://skia.googlesource.com/skia/+/0e040f7da2fdfeb49aa60d24117306e3b1e6ea90
> >
> > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > NOPRESUBMIT=true
> > NOTREECHECKS=true
> > NOTRY=true
> >
> > Committed: https://skia.googlesource.com/skia/+/f7076a13e2d4269903b34ef2780e1c84723e4477
>
> TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/8e14d660b2a434bc708a70180c84210883611683

TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1005173004

9 years agoRevert of Revert of Revert of Change device creation to see the (optional) layer...
reed [Fri, 13 Mar 2015 19:34:32 +0000 (12:34 -0700)]
Revert of Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1006923002/)

Reason for revert:
platform_canvas tests failures

skia_unittests (with patch) skia_unittests (with patch) PlatformCanvas.TranslateLayer failed 2
Flakiness dashboard

failures:
PlatformCanvas.TranslateLayer
PlatformCanvas.FillLayer

Original issue's description:
> Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1008863002/)
>
> Reason for revert:
> guard in chrome has landed
>
> Original issue's description:
> > Revert of Change device creation to see the (optional) layer-paint (patchset #9 id:160001 of https://codereview.chromium.org/988413003/)
> >
> > Reason for revert:
> > need to have chrome opt-in for the older API before this can land (in chrome)
> >
> > Original issue's description:
> > > Change device creation to see the (optional) layer-paint
> > >
> > > Motivation:
> > >
> > > PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.
> > >
> > > This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.
> > >
> > > Committed: https://skia.googlesource.com/skia/+/1182d9a96b80bd12183ee7c81325a979a51ee0c0
> >
> > TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> > NOPRESUBMIT=true
> > NOTREECHECKS=true
> > NOTRY=true
> >
> > Committed: https://skia.googlesource.com/skia/+/0e040f7da2fdfeb49aa60d24117306e3b1e6ea90
>
> TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/f7076a13e2d4269903b34ef2780e1c84723e4477

TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1001423002

9 years agoCreation of GrBatchAtlas and Distancefieldpathrenderer batch
joshualitt [Fri, 13 Mar 2015 18:47:42 +0000 (11:47 -0700)]
Creation of GrBatchAtlas and Distancefieldpathrenderer batch

BUG=skia:

Review URL: https://codereview.chromium.org/975303005

9 years agoFix bug with very large round rects with large radii
robertphillips [Fri, 13 Mar 2015 16:53:01 +0000 (09:53 -0700)]
Fix bug with very large round rects with large radii

BUG=463920

Review URL: https://codereview.chromium.org/998903003

9 years agoHandle paths that do not report empty, but have no edges.
reed [Fri, 13 Mar 2015 15:48:26 +0000 (08:48 -0700)]
Handle paths that do not report empty, but have no edges.

patch from issue 999963005 at patchset 1 (http://crrev.com/999963005#ps1)

BUG=skia:3527

Review URL: https://codereview.chromium.org/1008883002

9 years agoRevert of Revert of Change device creation to see the (optional) layer-paint (patchse...
reed [Fri, 13 Mar 2015 15:46:12 +0000 (08:46 -0700)]
Revert of Revert of Change device creation to see the (optional) layer-paint (patchset #1 id:1 of https://codereview.chromium.org/1008863002/)

Reason for revert:
guard in chrome has landed

Original issue's description:
> Revert of Change device creation to see the (optional) layer-paint (patchset #9 id:160001 of https://codereview.chromium.org/988413003/)
>
> Reason for revert:
> need to have chrome opt-in for the older API before this can land (in chrome)
>
> Original issue's description:
> > Change device creation to see the (optional) layer-paint
> >
> > Motivation:
> >
> > PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.
> >
> > This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.
> >
> > Committed: https://skia.googlesource.com/skia/+/1182d9a96b80bd12183ee7c81325a979a51ee0c0
>
> TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/0e040f7da2fdfeb49aa60d24117306e3b1e6ea90

TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1006923002

9 years agoCall to retrieve SkDescriptor for SkScalarContext from SkPaint.
joshualitt [Fri, 13 Mar 2015 15:38:43 +0000 (08:38 -0700)]
Call to retrieve SkDescriptor for SkScalarContext from SkPaint.

Review URL: https://codereview.chromium.org/1005103003

9 years agoIco security issues fix
msarett [Fri, 13 Mar 2015 15:07:01 +0000 (08:07 -0700)]
Ico security issues fix

BUG=skia:3401
BUG=skia:3426
BUG=skia:3441

Review URL: https://codereview.chromium.org/996173005

9 years agoRevert of Change device creation to see the (optional) layer-paint (patchset #9 id...
reed [Fri, 13 Mar 2015 14:28:28 +0000 (07:28 -0700)]
Revert of Change device creation to see the (optional) layer-paint (patchset #9 id:160001 of https://codereview.chromium.org/988413003/)

Reason for revert:
need to have chrome opt-in for the older API before this can land (in chrome)

Original issue's description:
> Change device creation to see the (optional) layer-paint
>
> Motivation:
>
> PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.
>
> This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.
>
> Committed: https://skia.googlesource.com/skia/+/1182d9a96b80bd12183ee7c81325a979a51ee0c0

TBR=halcanary@google.com,senorblanco@google.com,robertphillips@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1008863002

9 years agoChange device creation to see the (optional) layer-paint
reed [Fri, 13 Mar 2015 14:16:09 +0000 (07:16 -0700)]
Change device creation to see the (optional) layer-paint

Motivation:

PDFDevice currently relies on 1) being told that the layer's paint has an imagefilter, and in the case, it creates a rasterdevice. It then relies on (2) canvas itself sniffing the layer's paint and offering to apply-the-imagefilter to call drawSprite instead of drawDevice.

This subtle interchange is fragile, and also does not support other unsupported PDF features like colorfilters. This CL is a step toward making this use-raster-instead-of-native approach to layers more completely in the subclass' hands.

Review URL: https://codereview.chromium.org/988413003

9 years agoLower SkPictureShader tile size limits
fmalita [Fri, 13 Mar 2015 13:50:44 +0000 (06:50 -0700)]
Lower SkPictureShader tile size limits

Let's drop this to 4M pixels - maybe it will help relieve DM pressure on
mobile devices.

R=reed@google.com,mtklein@google.com
BUG=462348

Review URL: https://codereview.chromium.org/998423004

9 years agosome utils for rect and matrix
reed [Fri, 13 Mar 2015 13:08:28 +0000 (06:08 -0700)]
some utils for rect and matrix

BUG=skia:

Review URL: https://codereview.chromium.org/1003813003

9 years agoHave DM and nanobench display times in minutes when they're above 60 seconds.
mtklein [Thu, 12 Mar 2015 17:29:32 +0000 (10:29 -0700)]
Have DM and nanobench display times in minutes when they're above 60 seconds.

Getting tired of trying to divide in my head when reading Valgrind logs.

BUG=skia:

Review URL: https://codereview.chromium.org/1004703002

9 years agoIgnore public API checks if COMMIT=false is in the description
rmistry [Thu, 12 Mar 2015 16:48:10 +0000 (09:48 -0700)]
Ignore public API checks if COMMIT=false is in the description

BUG=skia:
NOTRY=true

Committed: https://skia.googlesource.com/skia/+/d1783aefcc0da86c5ff1d124c54704252d817621

Review URL: https://codereview.chromium.org/1006473002

9 years agoRevert of Ignore public API checks if COMMIT=false is in the description (patchset...
mtklein [Thu, 12 Mar 2015 16:33:46 +0000 (09:33 -0700)]
Revert of Ignore public API checks if COMMIT=false is in the description (patchset #4 id:80001 of https://codereview.chromium.org/1006473002/)

Reason for revert:
I cannot upload patches:

~/skia (skskp) $ git clu --bypass-hooks
++ git log -1 --pretty=%B
+ git cl upload -t update
Using 50% similarity for rename/copy detection. Override with --similarity.
Loaded authentication cookies from /Users/mtklein/.codereview_upload_cookies
Running presubmit upload checks ...

** Presubmit ERRORS **
Since the CL is editing public API, you must have an LGTM from one of: ('reed@chromium.org', 'reed@google.com', 'bsalomon@chromium.org', 'bsalomon@google.com', 'djsollen@chromium.org', 'djsollen@google.com')

Original issue's description:
> Ignore public API checks if COMMIT=false is in the description
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/d1783aefcc0da86c5ff1d124c54704252d817621

TBR=borenet@google.com,rmistry@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1005583002

9 years agoIgnore public API checks if COMMIT=false is in the description
rmistry [Thu, 12 Mar 2015 15:43:22 +0000 (08:43 -0700)]
Ignore public API checks if COMMIT=false is in the description

BUG=skia:

Review URL: https://codereview.chromium.org/1006473002

9 years agoDM: display current memory usage (instead of peak) when available.
mtklein [Thu, 12 Mar 2015 15:24:21 +0000 (08:24 -0700)]
DM: display current memory usage (instead of peak) when available.

Seems strictly more useful.

This implements Mac and Windows, which seemed easy.  Don't know how to do this on Linux yet.

BUG=skia:

CQ_EXTRA_TRYBOTS=client.skia:Test-Mac10.9-MacMini6.2-HD4000-x86_64-Debug-Trybot

NOTREECHECKS=true
TBR=halcanary@google.com

Review URL: https://codereview.chromium.org/990723002

9 years agoMake presubmit check that #if or #ifdef does not come before includes.
rmistry [Thu, 12 Mar 2015 14:48:40 +0000 (07:48 -0700)]
Make presubmit check that #if or #ifdef does not come before includes.

BUG=skia:3362

Review URL: https://codereview.chromium.org/998273002

9 years agoRejigger keep-alive thread so TSAN and LSAN don't complain.
mtklein [Thu, 12 Mar 2015 14:16:56 +0000 (07:16 -0700)]
Rejigger keep-alive thread so TSAN and LSAN don't complain.

BUG=skia:

Review URL: https://codereview.chromium.org/1006443002

9 years agoAdd #include first in Sk4x_* files
rmistry [Thu, 12 Mar 2015 14:09:25 +0000 (07:09 -0700)]
Add #include first in Sk4x_* files

BUG=skia:3362

Review URL: https://codereview.chromium.org/1004693002

9 years agoDM: add keepalive for Valgrind bot.
mtklein [Thu, 12 Mar 2015 13:28:54 +0000 (06:28 -0700)]
DM: add keepalive for Valgrind bot.

BUG=skia:

Review URL: https://codereview.chromium.org/998263003

9 years agoClean up SkDynamicAnnotations.
mtklein [Thu, 12 Mar 2015 12:27:46 +0000 (05:27 -0700)]
Clean up SkDynamicAnnotations.

    Unprotected reads -> relaxed reads.
    Unprotected write -> relaxed write.

The only unprotected write we had was in SkTraceEvent, which it looks like we nabbed from Chrome at some point and changed only to silence TSAN.  Chrome's version uses AtomicWord / NoBarrier_Load / NoBarrier_Store, which boils down to the same as here, intptr_t / relaxed load / relaxed store.

This leaves one place where we're lying a bit to TSAN, in include/core/SkLazyPtr.h where we're doing a data-dependent consume load.  We're telling TSAN it's consume, but telling any other compiler to compile it as relaxed, given how they all upgrade consume to acquire.  This eliminates a barrier for us on ARM.  How do you guys deal with this?  Just use a consume memory order, take the hit, and hope compilers get smarter one day?

BUG=chromium:465721

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

Review URL: https://codereview.chromium.org/996763002

9 years agoFix HWUI Sink for DM
Derek Sollenberger [Thu, 12 Mar 2015 11:52:22 +0000 (07:52 -0400)]
Fix HWUI Sink for DM

Review URL: https://codereview.chromium.org/996413002

9 years agoAdd terrible workaround to fix stencil clear crash on S3.
bsalomon [Wed, 11 Mar 2015 22:49:54 +0000 (15:49 -0700)]
Add terrible workaround to fix stencil clear crash on S3.

BUG=skia:3525

Review URL: https://codereview.chromium.org/1001693003

9 years agoGlyph positions maintain 32 bit integer part.
bungeman [Wed, 11 Mar 2015 21:05:29 +0000 (14:05 -0700)]
Glyph positions maintain 32 bit integer part.

A glyph position when mapped from canvas space to device space may land
outside the bounds of the current 16 bit integer part of device space.
Device space is already limited to 32 bits for the integer part,
but for a short space in drawText and drawPosText it is currently
limited to 16 bits (SkFixed). Raise this limit by moving to 48.16.
This matches the current similar fix for measureText.

BUG=chromium:375322

Review URL: https://codereview.chromium.org/977623002

9 years agoFix Windows build.
scroggo [Wed, 11 Mar 2015 20:58:17 +0000 (13:58 -0700)]
Fix Windows build.

SK_BUILD_FOR_WIN is no longer a valid way to check for building on
Windows (go figure). Build everywhere.

Remove the REPORTER_ASSERT, which was the failing part. It also isn't
necessary for the test, which is just that we are not leaking an
SkColorTable.

Also fix indentation.

TBR=bungeman@google.com,mtklein@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

BUG=skia:3457

Review URL: https://codereview.chromium.org/1002583002

9 years agoSince SkDrawFilter has serious architectural issues and may be
tomhudson [Wed, 11 Mar 2015 19:21:08 +0000 (12:21 -0700)]
Since SkDrawFilter has serious architectural issues and may be
deprecated, use a proxy SkCanvas for the same end: in every draw
call, inspect the paint, changing it to fit within the capabilities
of the Android Java (HWUI) drawing API.

Verified that this allows us to reenable all our ColorFilter tests.

R=djsollen@google.com

Review URL: https://codereview.chromium.org/997183003

9 years agodelete dead code
mtklein [Wed, 11 Mar 2015 19:17:00 +0000 (12:17 -0700)]
delete dead code

BUG=skia:

Review URL: https://codereview.chromium.org/1003493002

9 years agoFor safety include SkTypes.h in SkFontHost_mac.cpp
mtklein [Wed, 11 Mar 2015 19:01:25 +0000 (12:01 -0700)]
For safety include SkTypes.h in SkFontHost_mac.cpp

BUG=skia:3362

Review URL: https://codereview.chromium.org/996113003

9 years agoClean up poppler code.
mtklein [Wed, 11 Mar 2015 18:55:13 +0000 (11:55 -0700)]
Clean up poppler code.

We're not actually using it.

BUG=skia:3362
DOCS_PREVIEW= https://skia.org/?cl=1002493002

Review URL: https://codereview.chromium.org/1002493002

9 years agoAdjust atlas sizes to fix Mali400 precision issues
jvanverth [Wed, 11 Mar 2015 18:18:11 +0000 (11:18 -0700)]
Adjust atlas sizes to fix Mali400 precision issues

The previous change to atlas sizes introduced an issue where texture
coordinates for glyphs stored on the right side of the atlas were being
computed in an imprecise manner on Mali400 chips.

The only reasonable fix appears to be to use power-of-two texture sizes.
This widens the glyph atlas to the next power of 2 (or 2048) for A8 glyphs
only, and widens each GrPlot by 2x as well. By doing this, we can fit 3-4 large
distance field glyphs into a single GrPlot, which gives us 100-128 total large
glyphs at one time.

The 565 and 8888 glyph atlases are kept in their original size to preserve
space.

BUG=skia:3523

Review URL: https://codereview.chromium.org/994303003

9 years agoFix debugger's overdraw color filter
robertphillips [Wed, 11 Mar 2015 18:18:00 +0000 (11:18 -0700)]
Fix debugger's overdraw color filter

The colors seen in the overdraw color filter are not guaranteed to remain the colors from its own table. Looping through the table to find the old overdraw level is error prone and expensive.

Review URL: https://codereview.chromium.org/997913002

9 years agoUse proper SkFDot6 instead of int
qiankun.miao [Wed, 11 Mar 2015 18:12:59 +0000 (11:12 -0700)]
Use proper SkFDot6 instead of int

BUG=skia:

Review URL: https://codereview.chromium.org/960353005

9 years agoFix a memory leak when decoding corrupted indexed PNGs.
dml [Wed, 11 Mar 2015 18:12:54 +0000 (11:12 -0700)]
Fix a memory leak when decoding corrupted indexed PNGs.
Commit to branch refs/heads/png-leak

BUG=skia:3457

Committed: https://skia.googlesource.com/skia/+/561a1ca9559a1ea7589ab93350124284fcef3315

Review URL: https://codereview.chromium.org/951663002

9 years agoSimplify clip inside a layer -- will be applied in restore
reed [Wed, 11 Mar 2015 17:39:04 +0000 (10:39 -0700)]
Simplify clip inside a layer -- will be applied in restore

This reverts commit 44124650895833378938b4b174302fbf64403fe2.

NOTREECHECKS=True
BUG=skia:

Review URL: https://codereview.chromium.org/996233002

9 years agofix for valgrind preAbandonGpuContext
joshualitt [Wed, 11 Mar 2015 16:11:19 +0000 (09:11 -0700)]
fix for valgrind preAbandonGpuContext

NOTREECHECKS=true

BUG=skia:

Review URL: https://codereview.chromium.org/997923002

9 years agooptimization/fix: dirty the clip-bounds when we mod the clip in savelayer
reed [Wed, 11 Mar 2015 15:47:12 +0000 (08:47 -0700)]
optimization/fix: dirty the clip-bounds when we mod the clip in savelayer

Before the fix, we could use a stale cache of the clipbounds in quickReject. Often this could return false negatives, meaning we would try to draw more than we should (it would eventually be really clipped). Occasionally this could also report false positives (if the layer were outside of the normal canvas bounds, e.g. a layer with an offset imagefilter).

BUG=skia:
NOTREECHECKS=True

Review URL: https://codereview.chromium.org/983243003

9 years agoSkip WritePixels on GalaxyS3
mtklein [Wed, 11 Mar 2015 15:43:43 +0000 (08:43 -0700)]
Skip WritePixels on GalaxyS3

TBR=bsalomon@google.com

NOTREECHECKS=true

BUG=skia:

Review URL: https://codereview.chromium.org/994343003

9 years agoDM: show skipped and --blacklist notes only in verbose mode.
mtklein [Wed, 11 Mar 2015 14:19:15 +0000 (07:19 -0700)]
DM: show skipped and --blacklist notes only in verbose mode.

In non-verbose mode, these notes will spin away too fast to read anyway,
unless they're so long they end up leaving junk on the terminal.

NOTREECHECKS=true

BUG=skia:

Review URL: https://codereview.chromium.org/989083002