platform/upstream/libSkiaSharp.git
9 years agoRevert of add ganesh checkbox to skfiddle interface (patchset #6 id:270001 of https...
humper [Mon, 13 Oct 2014 23:05:05 +0000 (16:05 -0700)]
Revert of add ganesh checkbox to skfiddle interface (patchset #6 id:270001 of https://codereview.chromium.org/649313002/)

Reason for revert:
adding -Dskia_mesa=1 broke production; don't yet know why and don't want to leave things in a broken state.

Original issue's description:
> add ganesh checkbox to skfiddle interface
>
> BUG=skia:
> R=mtklein@google.com, tfarina@chromium.org
>
> Committed: https://skia.googlesource.com/skia/+/ccb54895c614632fd01d22b64ab614921f980eb1

TBR=tfarina@chromium.org,jcgregorio@google.com,mtklein@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/655633003

9 years agoadd ganesh checkbox to skfiddle interface
Greg Humphreys [Mon, 13 Oct 2014 22:12:33 +0000 (18:12 -0400)]
add ganesh checkbox to skfiddle interface

BUG=skia:
R=mtklein@google.com, tfarina@chromium.org

Review URL: https://codereview.chromium.org/649313002

9 years agoRevert of No threadsafe statics. (patchset #1 id:1 of https://codereview.chromium...
mtklein [Mon, 13 Oct 2014 21:50:41 +0000 (14:50 -0700)]
Revert of No threadsafe statics. (patchset #1 id:1 of https://codereview.chromium.org/654663002/)

Reason for revert:
Want to think about how to do this only to skialib and not to tools (gms, tests, etc.)

Original issue's description:
> No threadsafe statics.
>
> Chrome disables these for speed and code size, so we need
> to disable them to make sure our code is safe when used
> this way.
>
>     int foo() {
>         static int32_t atomic_thing;
>         return sk_atomic_inc(&atomic_thing);
>     }
>
> is not safe in Chrome.  Making the static global is:
>
>     static int32_t atomic_thing;
>     int foo() {
>         return sk_atomic_inc(&atomic_thing);
>     }
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/cad5d3e264535c919b80e1e2a85407307961f221

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

Review URL: https://codereview.chromium.org/649343003

9 years agoRevert of Create a single command buffer for GrInOrderDrawBuffer (patchset #14 id...
mtklein [Mon, 13 Oct 2014 21:00:42 +0000 (14:00 -0700)]
Revert of Create a single command buffer for GrInOrderDrawBuffer (patchset #14 id:1050001 of https://codereview.chromium.org/628453002/)

Reason for revert:
New test failing on Android: http://build.chromium.org/p/client.skia.android/builders/Test-Android-Nexus7-Tegra3-Arm7-Release/builds/89/steps/dm/logs/stdio

Original issue's description:
> Adds a GrTBaseList class that GrInOrderDrawBuffer uses to allocate
> all its commands interleaved in contiguous memory. GrTBaseList also
> supports extra data associated with objects, so we can store arrays
> inline without having to call malloc().
>
> Committed: https://skia.googlesource.com/skia/+/47c844aaba81e5a29c773b660e1d6062c766d253

TBR=bsalomon@google.com,reed@google.com,cdalton@nvidia.com
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/652843002

9 years agoLink to skiadocs site, since that is the canonical location for documentation.
Thiago Farina [Mon, 13 Oct 2014 20:56:30 +0000 (17:56 -0300)]
Link to skiadocs site, since that is the canonical location for documentation.

BUG=None
TEST=None
R=borenet@google.com, rmistry@google.com

Review URL: https://codereview.chromium.org/646853002

9 years agoexperimental parser
reed [Mon, 13 Oct 2014 20:53:48 +0000 (13:53 -0700)]
experimental parser

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/651023002

9 years agoFix reference to non-existant 'tests' target.
Thiago Farina [Mon, 13 Oct 2014 20:51:57 +0000 (17:51 -0300)]
Fix reference to non-existant 'tests' target.

This target was previously folded into dm. Do not advertise it anymore.

BUG=None
TEST=None
R=borenet@google.com

Review URL: https://codereview.chromium.org/646843002

9 years agoRequire SK_DECLARE_STATIC_LAZY_PTR is used in global scope.
mtklein [Mon, 13 Oct 2014 20:17:56 +0000 (13:17 -0700)]
Require SK_DECLARE_STATIC_LAZY_PTR is used in global scope.

Function- or method- local scope isn't threadsafe; the pointer is generally
zero-initialized on first use in function scope (i.e. lazily... we have to go
deeper), but for globals we can be pretty sure the linker will do that for us.

BUG=skia:

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

Review URL: https://codereview.chromium.org/651723003

9 years agoArchive svndiff script.
tfarina [Mon, 13 Oct 2014 19:58:33 +0000 (12:58 -0700)]
Archive svndiff script.

Nobody is using this script anymore, so to prevent more changes to it,
let's remove it now.

BUG=None
TEST=None
R=borenet@google.com

Review URL: https://codereview.chromium.org/646043002

9 years agoCreate helper functions to use in computeInvariantOutput calls
egdaniel [Mon, 13 Oct 2014 19:53:46 +0000 (12:53 -0700)]
Create helper functions to use in computeInvariantOutput calls

BUG=skia:

Review URL: https://codereview.chromium.org/643743003

9 years agoNo threadsafe statics.
mtklein [Mon, 13 Oct 2014 19:53:27 +0000 (12:53 -0700)]
No threadsafe statics.

Chrome disables these for speed and code size, so we need
to disable them to make sure our code is safe when used
this way.

    int foo() {
        static int32_t atomic_thing;
        return sk_atomic_inc(&atomic_thing);
    }

is not safe in Chrome.  Making the static global is:

    static int32_t atomic_thing;
    int foo() {
        return sk_atomic_inc(&atomic_thing);
    }

BUG=skia:

Review URL: https://codereview.chromium.org/654663002

9 years agoFix race in supports_simd().
mtklein [Mon, 13 Oct 2014 19:48:16 +0000 (12:48 -0700)]
Fix race in supports_simd().

Local statics are not thread safe in Chrome.  Use an SkLazyPtr instead.

See https://code.google.com/p/chromium/issues/detail?id=418041

BUG=418041

Review URL: https://codereview.chromium.org/655573002

9 years agoEnable color emoji GM on Android
jvanverth [Mon, 13 Oct 2014 19:48:10 +0000 (12:48 -0700)]
Enable color emoji GM on Android

BUG=skia:2887

Review URL: https://codereview.chromium.org/656433002

9 years agoAdds a GrTBaseList class that GrInOrderDrawBuffer uses to allocate
cdalton [Mon, 13 Oct 2014 19:43:10 +0000 (12:43 -0700)]
Adds a GrTBaseList class that GrInOrderDrawBuffer uses to allocate
all its commands interleaved in contiguous memory. GrTBaseList also
supports extra data associated with objects, so we can store arrays
inline without having to call malloc().

Review URL: https://codereview.chromium.org/628453002

9 years agobug_chomper: Make server's port configurable.
tfarina [Mon, 13 Oct 2014 19:43:01 +0000 (12:43 -0700)]
bug_chomper: Make server's port configurable.

Make 'port' a flag so you can change it from the command line, making the
server more flexible and allowing us to change in which port it listen
to requests.

$ ./run_server.sh --port :8002

BUG=None
TEST=see above
R=borenet@google.com

Review URL: https://codereview.chromium.org/649663003

9 years agoadd pictures to lua
reed [Mon, 13 Oct 2014 19:38:04 +0000 (12:38 -0700)]
add pictures to lua

BUG=skia:

Review URL: https://codereview.chromium.org/650263002

9 years agoFix webtry setup once instructions in README doc.
tfarina [Mon, 13 Oct 2014 19:37:52 +0000 (12:37 -0700)]
Fix webtry setup once instructions in README doc.

BUG=None
TEST=NOne
R=humper@google.com
TBR=humper@google.com

Review URL: https://codereview.chromium.org/648563003

9 years agoDevirtualize read/write pixels on surface.
bsalomon [Mon, 13 Oct 2014 19:32:55 +0000 (12:32 -0700)]
Devirtualize read/write pixels on surface.

Consolidate read/write funcs in context.

Remove support for reading pixels from a surface that's not a target. It's currently broken and neither used nor tested.

Review URL: https://codereview.chromium.org/648863002

9 years agoAllow GCC to build Skia on Mac.
mtklein [Mon, 13 Oct 2014 19:32:34 +0000 (12:32 -0700)]
Allow GCC to build Skia on Mac.

GCC doesn't understand -fasm-blocks or -mpascal-strings, but we don't care
about them.

While looking around in Gyp, I noticed a better way to disable warnings about
offsetof so that it doesn't tell us "disabling this warning makes no sense in C"
for every C source file we compile.

BUG=skia:

Review URL: https://codereview.chromium.org/650553002

9 years agoadd support for skfiddle width/height options
Greg Humphreys [Mon, 13 Oct 2014 17:58:09 +0000 (13:58 -0400)]
add support for skfiddle width/height options

BUG=skia:
R=jcgregorio@google.com

Review URL: https://codereview.chromium.org/656463002

9 years agoWhitespace change to trigger builds
borenet [Mon, 13 Oct 2014 16:52:20 +0000 (09:52 -0700)]
Whitespace change to trigger builds

BUG=skia:761
TBR=rmistry
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/621353002

9 years agoWhitespace change to test the CQ's new trybots
rmistry [Mon, 13 Oct 2014 13:02:28 +0000 (06:02 -0700)]
Whitespace change to test the CQ's new trybots

BUG=skia:2973
TBR=
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/654463003

9 years agoadd fade_slide transition to lua slides
reed [Mon, 13 Oct 2014 02:05:52 +0000 (19:05 -0700)]
add fade_slide transition to lua slides

TBR=
BUG=skia:

Review URL: https://codereview.chromium.org/646613004

9 years agoadd canvas:clear to lua
reed [Sun, 12 Oct 2014 19:18:40 +0000 (12:18 -0700)]
add canvas:clear to lua

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/648083002

9 years agoadd surfaces to lua
reed [Sun, 12 Oct 2014 17:36:47 +0000 (10:36 -0700)]
add surfaces to lua

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/649973002

9 years agoFix leak in GLPrograms test.
bsalomon [Sun, 12 Oct 2014 12:40:00 +0000 (05:40 -0700)]
Fix leak in GLPrograms test.

TBR=joshualitt@google.com

Review URL: https://codereview.chromium.org/648063002

9 years agoadd key handlers to lua
reed [Sat, 11 Oct 2014 20:13:11 +0000 (13:13 -0700)]
add key handlers to lua

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/652473002

9 years agoallow for lua click handlers
reed [Sat, 11 Oct 2014 18:28:07 +0000 (11:28 -0700)]
allow for lua click handlers

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/649013002

10 years agomove test for lua into separate resource file
reed [Sat, 11 Oct 2014 03:31:24 +0000 (20:31 -0700)]
move test for lua into separate resource file

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/645283002

10 years agobug_chomper: Use parenthesized declaration style for const and vars.
tfarina [Sat, 11 Oct 2014 01:40:45 +0000 (18:40 -0700)]
bug_chomper: Use parenthesized declaration style for const and vars.

It seems cleaner that way and nicer to read, also avoids repeating
'const' and 'var' everytime you have to add more items.

BUG=None
TEST=./run_server.sh
R=borenet@google.com

Review URL: https://codereview.chromium.org/645803003

10 years agoOpt state takes a GP instead of a GeometryStage
joshualitt [Sat, 11 Oct 2014 00:47:00 +0000 (17:47 -0700)]
Opt state takes a GP instead of a GeometryStage

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/71856d520461ae025a0332aa0ce9735a096d9baf

Review URL: https://codereview.chromium.org/637003003

10 years agoadd drawImageRect to lua, update sample
reed [Fri, 10 Oct 2014 22:28:19 +0000 (15:28 -0700)]
add drawImageRect to lua, update sample

BUG=skia:
TBR=

Review URL: https://codereview.chromium.org/642353005

10 years agoRevert of Opt state takes a GP instead of a GeometryStage (patchset #18 id:1500001...
joshualitt [Fri, 10 Oct 2014 21:11:59 +0000 (14:11 -0700)]
Revert of Opt state takes a GP instead of a GeometryStage (patchset #18 id:1500001 of https://codereview.chromium.org/637003003/)

Reason for revert:
Revert because this may be causing a break in a chrome gpu test

Original issue's description:
> Opt state takes a GP instead of a GeometryStage
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/71856d520461ae025a0332aa0ce9735a096d9baf

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

Review URL: https://codereview.chromium.org/647183002

10 years agoRemove empty recordy.gypi file.
tfarina [Fri, 10 Oct 2014 20:34:25 +0000 (13:34 -0700)]
Remove empty recordy.gypi file.

Fixes the TODO in that file.

BUG=None
TEST=None
R=mtklein@google.com

Review URL: https://codereview.chromium.org/643543002

10 years agoUpdate coloremoji expectations
jvanverth [Fri, 10 Oct 2014 19:58:46 +0000 (12:58 -0700)]
Update coloremoji expectations

TBR=reed@google.com
BUG=skia:2887

Review URL: https://codereview.chromium.org/649753003

10 years agoFix bug in GrCachedLayer reuse
robertphillips [Fri, 10 Oct 2014 18:38:29 +0000 (11:38 -0700)]
Fix bug in GrCachedLayer reuse

In the new MultiPictureDraw tests a single hoisted layer is reused multiple times. The previous plot locking scheme allowed GrCachedLayer objects to be aggressively deleted prematurely leaving the reusing GrHoistedLayer objects with dangling pointers.

This CL changes adds a new pseudo-ref to GrCachedLayer. (It can't be a real ref since the cached layers aren't deleted when it goes to 0).

NOTRY=true

Committed: https://skia.googlesource.com/skia/+/5c481666c9678f43e039ad605457be3854cf8de3

Review URL: https://codereview.chromium.org/640323002

10 years agoFix color emoji.
jvanverth [Fri, 10 Oct 2014 18:36:12 +0000 (11:36 -0700)]
Fix color emoji.

Removes the GrMaskFormat and single atlas in GrTextStrike.
Replaces it by storing the GrMaskFormat in each GrGlyph, and
doing a lookup for the correct atlas based on that.

Disables color glyph rendering in GrDistanceFieldTextContext
for now.

BUG=skia:2887

Committed: https://skia.googlesource.com/skia/+/bc92163ddfe957ad6ffbb02ac40e0ba75ff82216

Review URL: https://codereview.chromium.org/636183005

10 years agoOpt state takes a GP instead of a GeometryStage
joshualitt [Fri, 10 Oct 2014 16:56:55 +0000 (09:56 -0700)]
Opt state takes a GP instead of a GeometryStage

BUG=skia:

Review URL: https://codereview.chromium.org/637003003

10 years agodetect --loops is < 0 and interpret that as running forever (mostly)
reed [Fri, 10 Oct 2014 16:09:52 +0000 (09:09 -0700)]
detect --loops is < 0 and interpret that as running forever (mostly)

BUG=skia:

Review URL: https://codereview.chromium.org/643143002

10 years agoRevert of Fix color emoji. (patchset #11 id:320001 of https://codereview.chromium...
jvanverth [Fri, 10 Oct 2014 15:52:03 +0000 (08:52 -0700)]
Revert of Fix color emoji. (patchset #11 id:320001 of https://codereview.chromium.org/636183005/)

Reason for revert:
Crashing the Windows bots.

Original issue's description:
> Fix color emoji.
>
> Removes the GrMaskFormat and single atlas in GrTextStrike.
> Replaces it by storing the GrMaskFormat in each GrGlyph, and
> doing a lookup for the correct atlas based on that.
>
> Disables color glyph rendering in GrDistanceFieldTextContext
> for now.
>
> BUG=skia:2887
>
> Committed: https://skia.googlesource.com/skia/+/bc92163ddfe957ad6ffbb02ac40e0ba75ff82216

TBR=robertphillips@google.com,bungeman@google.com,reed@google.com,bsalomon@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:2887

Review URL: https://codereview.chromium.org/640413004

10 years agoFix color emoji.
jvanverth [Fri, 10 Oct 2014 15:21:29 +0000 (08:21 -0700)]
Fix color emoji.

Removes the GrMaskFormat and single atlas in GrTextStrike.
Replaces it by storing the GrMaskFormat in each GrGlyph, and
doing a lookup for the correct atlas based on that.

Disables color glyph rendering in GrDistanceFieldTextContext
for now.

BUG=skia:2887

Review URL: https://codereview.chromium.org/636183005

10 years agoIgnore complexclip_path_aa on Win7-HD2000 for the moment.
jvanverth [Fri, 10 Oct 2014 14:14:55 +0000 (07:14 -0700)]
Ignore complexclip_path_aa on Win7-HD2000 for the moment.

TBR=reed@google.com
BUG=skia:3006

Review URL: https://codereview.chromium.org/643103002

10 years agoupdating expectations
joshualitt [Fri, 10 Oct 2014 13:48:48 +0000 (06:48 -0700)]
updating expectations

TBR=
BUG=skia:

Review URL: https://codereview.chromium.org/646963002

10 years agoTurn SkCanvasStateUtils into a class with static functions.
tfarina [Fri, 10 Oct 2014 13:19:09 +0000 (06:19 -0700)]
Turn SkCanvasStateUtils into a class with static functions.

That simplifies the way to declare it a friend, as needed in SkCanvas.

BUG=skia:2914
TEST=make most
R=reed@google.com

Review URL: https://codereview.chromium.org/645773002

10 years agofaster SkRect::sort
reed [Fri, 10 Oct 2014 12:50:15 +0000 (05:50 -0700)]
faster SkRect::sort

BUG=skia:

Review URL: https://codereview.chromium.org/646863002

10 years agoAdd template.cpp to gitignore.
tfarina [Fri, 10 Oct 2014 12:20:51 +0000 (05:20 -0700)]
Add template.cpp to gitignore.

It is generated experimental/webtry/build script, not checked-in the
source tree.

BUG=None
TEST=git status do not show it anymore as untracked file.
R=humper@google.com

Review URL: https://codereview.chromium.org/647633002

10 years agoUpdate SKP version to 122
borenet [Fri, 10 Oct 2014 04:26:14 +0000 (21:26 -0700)]
Update SKP version to 122

Automatic commit by the RecreateSKPs bot.

TBR=

Review URL: https://codereview.chromium.org/647663002

10 years agoUse the function recently added to common in svndiff.py
Thiago Farina [Fri, 10 Oct 2014 03:25:29 +0000 (00:25 -0300)]
Use the function recently added to common in svndiff.py

GetModifiedFiles() function was added to common repo in
https://skia.googlesource.com/common/+/af1de867422c42855ae54ed3838c62ea445ea6b4

BUG=None
TEST=None
R=borenet@google.com

Review URL: https://codereview.chromium.org/634313006

10 years agoUse BBH reserve hook to preallocate space for tiles.
mtklein [Fri, 10 Oct 2014 01:22:41 +0000 (18:22 -0700)]
Use BBH reserve hook to preallocate space for tiles.

Before getting too far into changing how SkTileGrid stores its tiles, I figured I'd
better see how much I can tweak out the existing format.  Cleverly, that way
any improvements I make by changing the format will look that much less
impressive.

This CL looks like it will be a 5-15% win in time spent recording, with no effect
on playback.

This CL also shrinks the tiles to fit exactly when we're done inserting,
using newly added SkTDArray::shrinkToFit().  It's quite cheap to run (maybe
taking back 1-2% from those 5-15% wins), and means we'll lug around about 15%
fewer bytes in the tile grids.  Note though this strategy temporarily uses up to
30% more memory while building the tile grid.  For our largest SKPs, that's
maybe 75-100K extra.

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/52455cbc02d7f480d988ae7cdacc11ad69078c2c

CQ_EXTRA_TRYBOTS=tryserver.skia:Canary-Chrome-Ubuntu13.10-Ninja-x86_64-ToT-Trybot

Review URL: https://codereview.chromium.org/639823005

10 years agoRemove unused SaveLayerInfo::fValid field
robertphillips [Thu, 9 Oct 2014 23:43:42 +0000 (16:43 -0700)]
Remove unused SaveLayerInfo::fValid field

Minor code cleanup

Review URL: https://codereview.chromium.org/643713002

10 years agoRefactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key
robertphillips [Thu, 9 Oct 2014 22:36:06 +0000 (15:36 -0700)]
Refactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key

NOTRY=true

Committed: https://skia.googlesource.com/skia/+/27415b71bd529456165945e19b5b7efbebf6fb51

Review URL: https://codereview.chromium.org/640373002

10 years agoFPs now use the correct builder types(just a rename)
joshualitt [Thu, 9 Oct 2014 22:04:05 +0000 (15:04 -0700)]
FPs now use the correct builder types(just a rename)

BUG=skia:

Review URL: https://codereview.chromium.org/648463003

10 years agocleanup and optimize rect intersect routines
reed [Thu, 9 Oct 2014 21:29:01 +0000 (14:29 -0700)]
cleanup and optimize rect intersect routines

BUG=skia:

Review URL: https://codereview.chromium.org/640723004

10 years agoRevert of Use BBH reserve hook to preallocate space for tiles. (patchset #2 id:80001...
mtklein [Thu, 9 Oct 2014 20:55:20 +0000 (13:55 -0700)]
Revert of Use BBH reserve hook to preallocate space for tiles. (patchset #2 id:80001 of https://codereview.chromium.org/639823005/)

Reason for revert:
failed assertion "fXTiles * fYTiles != 0"

Original issue's description:
> Use BBH reserve hook to preallocate space for tiles.
>
> Before getting too far into changing how SkTileGrid stores its tiles, I figured I'd
> better see how much I can tweak out the existing format.  Cleverly, that way
> any improvements I make by changing the format will look that much less
> impressive.
>
> This CL looks like it will be a 5-15% win in time spent recording, with no effect
> on playback.
>
> This CL also shrinks the tiles to fit exactly when we're done inserting,
> using newly added SkTDArray::shrinkToFit().  It's quite cheap to run (maybe
> taking back 1-2% from those 5-15% wins), and means we'll lug around about 15%
> fewer bytes in the tile grids.  Note though this strategy temporarily uses up to
> 30% more memory while building the tile grid.  For our largest SKPs, that's
> maybe 75-100K extra.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/52455cbc02d7f480d988ae7cdacc11ad69078c2c

TBR=reed@google.com,robertphillips@google.com,mtklein@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/642933002

10 years agoStore color and coverage stages in a single array in GrOptDrawState.
egdaniel [Thu, 9 Oct 2014 20:47:05 +0000 (13:47 -0700)]
Store color and coverage stages in a single array in GrOptDrawState.

BUG=skia:

Review URL: https://codereview.chromium.org/641223002

10 years agoForce linking of static member variables for GLPrograms
joshualitt [Thu, 9 Oct 2014 20:12:35 +0000 (13:12 -0700)]
Force linking of static member variables for GLPrograms

BUG=skia:

Review URL: https://codereview.chromium.org/614163002

10 years agoRevert of Refactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache...
robertphillips [Thu, 9 Oct 2014 19:47:08 +0000 (12:47 -0700)]
Revert of Refactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key (patchset #1 id:1 of https://codereview.chromium.org/640373002/)

Reason for revert:
Failing unit tests

Original issue's description:
> Refactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key
>
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/27415b71bd529456165945e19b5b7efbebf6fb51

TBR=jvanverth@google.com
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/644823003

10 years agoRevert of Fix bug in plot locking system (patchset #3 id:80001 of https://codereview...
robertphillips [Thu, 9 Oct 2014 19:47:01 +0000 (12:47 -0700)]
Revert of Fix bug in plot locking system (patchset #3 id:80001 of https://codereview.chromium.org/640323002/)

Reason for revert:
Failing unit tests

Original issue's description:
> Fix bug in plot locking system
>
> In the new MultiPictureDraw tests a single hoisted layer is reused multiple times. The previous plot locking scheme allowed GrCachedLayer objects to be aggressively deleted prematurely leaving the reusing GrHoistedLayer objects with dangling pointers.
>
> This CL changes the plot locking system to add a pseudo-ref for each GrHoistedLayer.
>
> NOTRY=true
>
> Committed: https://skia.googlesource.com/skia/+/5c481666c9678f43e039ad605457be3854cf8de3

TBR=jvanverth@google.com
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/643673004

10 years agoUpdate expectations for distance field path rendering.
jvanverth [Thu, 9 Oct 2014 19:37:35 +0000 (12:37 -0700)]
Update expectations for distance field path rendering.

TBR=reed@google.com

Review URL: https://codereview.chromium.org/641203002

10 years agoUse BBH reserve hook to preallocate space for tiles.
mtklein [Thu, 9 Oct 2014 19:36:48 +0000 (12:36 -0700)]
Use BBH reserve hook to preallocate space for tiles.

Before getting too far into changing how SkTileGrid stores its tiles, I figured I'd
better see how much I can tweak out the existing format.  Cleverly, that way
any improvements I make by changing the format will look that much less
impressive.

This CL looks like it will be a 5-15% win in time spent recording, with no effect
on playback.

This CL also shrinks the tiles to fit exactly when we're done inserting,
using newly added SkTDArray::shrinkToFit().  It's quite cheap to run (maybe
taking back 1-2% from those 5-15% wins), and means we'll lug around about 15%
fewer bytes in the tile grids.  Note though this strategy temporarily uses up to
30% more memory while building the tile grid.  For our largest SKPs, that's
maybe 75-100K extra.

BUG=skia:

Review URL: https://codereview.chromium.org/639823005

10 years agoFix bug in plot locking system
robertphillips [Thu, 9 Oct 2014 19:30:10 +0000 (12:30 -0700)]
Fix bug in plot locking system

In the new MultiPictureDraw tests a single hoisted layer is reused multiple times. The previous plot locking scheme allowed GrCachedLayer objects to be aggressively deleted prematurely leaving the reusing GrHoistedLayer objects with dangling pointers.

This CL changes the plot locking system to add a pseudo-ref for each GrHoistedLayer.

NOTRY=true

Review URL: https://codereview.chromium.org/640323002

10 years agoRefactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key
robertphillips [Thu, 9 Oct 2014 19:29:24 +0000 (12:29 -0700)]
Refactor GrLayerHoister::FindLayersToHoist to prep for adding clip to cache key

NOTRY=true

Review URL: https://codereview.chromium.org/640373002

10 years agoFix restore in AVMR to have correct coverage count.
egdaniel [Thu, 9 Oct 2014 19:07:20 +0000 (12:07 -0700)]
Fix restore in AVMR to have correct coverage count.

BUG=skia:

Review URL: https://codereview.chromium.org/638173003

10 years agoSmall improvements to SkBitSet:
mtklein [Thu, 9 Oct 2014 18:49:30 +0000 (11:49 -0700)]
Small improvements to SkBitSet:
  - implement O(1) operations in SkBitSet.h so they can inline away
  - use calloc to allocate empty bitsets instead of malloc then clear
  - little style things

BUG=skia:

Review URL: https://codereview.chromium.org/640243003

10 years agoImprove SkARGB32_A8_BlitMask_SSE2
jmuizelaar [Thu, 9 Oct 2014 18:43:02 +0000 (11:43 -0700)]
Improve SkARGB32_A8_BlitMask_SSE2

With clang this:

-       movzbl  -3(%rbx), %edx
-       pxor    %xmm5, %xmm5
-       pinsrw  $0, %edx, %xmm5
-       pinsrw  $1, %edx, %xmm5
-       movzbl  -2(%rbx), %edx
-       pinsrw  $2, %edx, %xmm5
-       pinsrw  $3, %edx, %xmm5
-       movzbl  -1(%rbx), %edx
-       pinsrw  $4, %edx, %xmm5
-       pinsrw  $5, %edx, %xmm5
-       movzbl  (%rbx), %edx
-       pinsrw  $6, %edx, %xmm5
-       pinsrw  $7, %edx, %xmm5

becomes:

+       movd    (%rbx), %xmm4
+       punpcklbw       %xmm9, %xmm4
+       punpcklwd       %xmm4, %xmm4

And clang already does better codegen than msvc 2013 on this.

BUG=skia:

Review URL: https://codereview.chromium.org/609823003

10 years agoAttempt at fixing color cube bench
sugoi [Thu, 9 Oct 2014 18:35:09 +0000 (11:35 -0700)]
Attempt at fixing color cube bench

The original bench was hitting the cache since it was using the same color filter for all loops. By creating a new color filter within the loop, at least this part of it is solved. I'm not 100% sure this is the right way, but at least the numbers are a bit more reasonable and are affected by the output resolution.

BUG=skia:

Review URL: https://codereview.chromium.org/648483002

10 years agoWhitespace change to test CQ's trybots
rmistry [Thu, 9 Oct 2014 18:10:33 +0000 (11:10 -0700)]
Whitespace change to test CQ's trybots

(SkipBuildbotRuns)
TBR=
BUG=skia:

Review URL: https://codereview.chromium.org/646663002

10 years agoRemove tab parameter from GrGLSLMulVarBy4f function
egdaniel [Thu, 9 Oct 2014 17:34:58 +0000 (10:34 -0700)]
Remove tab parameter from GrGLSLMulVarBy4f function

With pretty printing of shader code, there is no longer a need to explictily
have tabs in our code.

BUG=skia:

Review URL: https://codereview.chromium.org/648463002

10 years agoUse presence of a content key as non-scratch indicator
bsalomon [Thu, 9 Oct 2014 16:57:18 +0000 (09:57 -0700)]
Use presence of a content key as non-scratch indicator

BUG=skia:2889

Committed: https://skia.googlesource.com/skia/+/9eefe0851eeaa8ded05b4774ebcb38ed201d5dbf

Review URL: https://codereview.chromium.org/639873002

10 years agoClarify what DM means in the README file.
tfarina [Thu, 9 Oct 2014 16:00:29 +0000 (09:00 -0700)]
Clarify what DM means in the README file.

It took me a while to figure out what this acronym meant. ;)
Googling for it lead me to gyp/dm.gyp.

BUG=None
TEST=None
R=mtklein@google.com

Review URL: https://codereview.chromium.org/638083003

10 years agoRemove references to out/Debug/tests executable.
tfarina [Thu, 9 Oct 2014 15:56:55 +0000 (08:56 -0700)]
Remove references to out/Debug/tests executable.

This target does not exists anymore, it was folded into out/Debug/dm.
You can't do `ninja -C out/Debug tests` either anymore.

BUG=None
TEST=None
R=mtklein@google.com

Review URL: https://codereview.chromium.org/638913005

10 years agoLeak fixed
sugoi [Thu, 9 Oct 2014 15:09:04 +0000 (08:09 -0700)]
Leak fixed

BUG=skia:

Review URL: https://codereview.chromium.org/639423002

10 years agogl programs rewrite
joshualitt [Thu, 9 Oct 2014 14:25:36 +0000 (07:25 -0700)]
gl programs rewrite

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/07a255310aca9f3e83bf741dc663a58818ad681c

Review URL: https://codereview.chromium.org/628633003

10 years agorename command line flag to avoid conflict, use very verbose
caryclark [Thu, 9 Oct 2014 14:08:59 +0000 (07:08 -0700)]
rename command line flag to avoid conflict, use very verbose

TBR=
NOTRY=true
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/645573003

10 years agoAdd SkBBoxHierarchy::reserve() as an optional size hint.
mtklein [Thu, 9 Oct 2014 13:49:47 +0000 (06:49 -0700)]
Add SkBBoxHierarchy::reserve() as an optional size hint.

I want to play around with how SkTileGrid stores its tiles.  Having a
cap on the number of insert() calls can be pretty handy.

While I'm at it, I gave flush() a default empty impl.  Like reserve(),
it's really an optional hook for subclasses.

BUG=skia:

Review URL: https://codereview.chromium.org/639933003

10 years agotry again to fix ios build
caryclark [Thu, 9 Oct 2014 13:37:50 +0000 (06:37 -0700)]
try again to fix ios build

TBR=
NOTRY=true
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/638403002

10 years agofix ios build
caryclark [Thu, 9 Oct 2014 13:13:49 +0000 (06:13 -0700)]
fix ios build

NOTREECHECKS=true
NOTRY=true
TBR=

Review URL: https://codereview.chromium.org/641063002

10 years agoDraw more accurate thick-stroked Beziers (disabled)
caryclark [Thu, 9 Oct 2014 12:36:03 +0000 (05:36 -0700)]
Draw more accurate thick-stroked Beziers (disabled)

Draw thick-stroked Beziers by computing the outset quadratic, measuring the error, and subdividing until the error is within a predetermined limit.

To try this CL out, change src/core/SkStroke.h:18 to

  #define QUAD_STROKE_APPROXIMATION 1

or from the command line: CPPFLAGS="-D QUAD_STROKE_APPROXIMATION=1" ./gyp_skia

Here's what's in this CL:

bench/BezierBench.cpp : a microbench for examining where the time is going
gm/beziers.cpp        : random Beziers with various thicknesses
gm/smallarc.cpp       : a distillation of bug skia:2769
samplecode/SampleRotateCircles.cpp : controls added for error, limit, width
src/core/SkStroke.cpp : the new stroke implementation (disabled)
tests/StrokerTest.cpp : a stroke torture test that checks normal and extreme values

The new stroke algorithm has a tweakable parameter:

  stroker.setError(1);  (SkStrokeRec.cpp:112)

The stroke error is the allowable gap between the midpoint of the stroke quadratic and the center Bezier. As the projection from the quadratic approaches the endpoints, the error is decreased proportionally so that it is always inside the quadratic curve.

An overview of how this works:
- For a given T range of a Bezier, compute the perpendiculars and find the points outset and inset for some radius.
- Construct tangents for the quadratic stroke.
- If the tangent don't intersect between them (may happen with cubics), subdivide.
- If the quadratic stroke end points are close (again, may happen with cubics), draw a line between them.
- Compute the quadratic formed by the intersecting tangents.
- If the midpoint of the quadratic is close to the midpoint of the Bezier perpendicular, return the quadratic.
- If the end of the stroke at the Bezier midpoint doesn't intersect the quad's bounds, subdivide.
- Find where the Bezier midpoint ray intersects the quadratic.
- If the intersection is too close to the quad's endpoints, subdivide.
- If the error is large proportional to the intersection's distance to the quad's endpoints, subdivide.

BUG=skia:723,skia:2769

Review URL: https://codereview.chromium.org/558163005

10 years agouse real pdfdevice behind SkDocument
reed [Thu, 9 Oct 2014 12:30:10 +0000 (05:30 -0700)]
use real pdfdevice behind SkDocument

BUG=skia:
TBR=bungeman@google.com

Review URL: https://codereview.chromium.org/645563002

10 years agoAdding 3D lut color filter
sugoi [Thu, 9 Oct 2014 12:27:23 +0000 (05:27 -0700)]
Adding 3D lut color filter

Included in this cl is support for 3D textures.

BUG=skia:

Review URL: https://codereview.chromium.org/580863004

10 years agoMake the Sk GL context class an abstract base class
kkinnunen [Thu, 9 Oct 2014 12:24:15 +0000 (05:24 -0700)]
Make the Sk GL context class an abstract base class

Make the Sk GL context class, SkGLNativeContext, an abstract base class. Before,
it depended on ifdefs to implement the platform dependent polymorphism.  Move
the logic to subclasses of the various platform implementations.

This a step to enable Skia embedders to compile dm and bench_pictures. The
concrete goal is to support running these test apps with Chromium command buffer.

With this change, Chromium can implement its own version of SkGLNativeContext
that uses command buffer, and host the implementation in its own repository.

Implements the above by renaming the SkGLContextHelper to SkGLContext and
removing the unneeded SkGLNativeContext. Also removes
SkGLNativeContext::AutoRestoreContext functionality, it appeared to be unused:
no use in Skia code, and no tests.

BUG=skia:2992

Committed: https://skia.googlesource.com/skia/+/a90ed4e83897b45d6331ee4c54e1edd4054de9a8

Review URL: https://codereview.chromium.org/630843002

10 years agoUpdate old tools to allow MultiPictureDraw rendering
robertphillips [Thu, 9 Oct 2014 11:59:19 +0000 (04:59 -0700)]
Update old tools to allow MultiPictureDraw rendering

I'll post a separate patch for nanobench and dm

Review URL: https://codereview.chromium.org/639013003

10 years agoUpdate SKP version to 121
borenet [Thu, 9 Oct 2014 04:24:38 +0000 (21:24 -0700)]
Update SKP version to 121

Automatic commit by the RecreateSKPs bot.

TBR=

Review URL: https://codereview.chromium.org/644633002

10 years agowebtry: Simply move sys/ directory into setup/.
tfarina [Thu, 9 Oct 2014 03:05:13 +0000 (20:05 -0700)]
webtry: Simply move sys/ directory into setup/.

That seems to make things more tied and that seems to be where we have
the same scripts/configs in the perf directory:
https://skia.googlesource.com/buildbot/+/master/perf/setup/sys/

That way we keep all files used/related to setting up a new webtry server in one place.

BUG=None
TEST=None
R=humper@google.com

Review URL: https://codereview.chromium.org/644513003

10 years agocall derived ~ from GrIORef
bsalomon [Wed, 8 Oct 2014 19:36:54 +0000 (12:36 -0700)]
call derived ~ from GrIORef

TBR=mtklein@google.com
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/641813002

10 years agoRemove ~GrIORef since last remaining virtual and now all inline
bsalomon [Wed, 8 Oct 2014 18:42:10 +0000 (11:42 -0700)]
Remove ~GrIORef since last remaining virtual and now all inline

TBR=reed@google.com
NOTRY=true

Review URL: https://codereview.chromium.org/638003003

10 years agoRevert of Use presence of a content key as non-scratch indicator (patchset #5 id...
bsalomon [Wed, 8 Oct 2014 18:21:33 +0000 (11:21 -0700)]
Revert of Use presence of a content key as non-scratch indicator (patchset #5 id:80001 of https://codereview.chromium.org/639873002/)

Reason for revert:
breaking nanobench on ubuntu

Original issue's description:
> Use presence of a content key as non-scratch indicator
>
> BUG=skia:2889
>
> Committed: https://skia.googlesource.com/skia/+/9eefe0851eeaa8ded05b4774ebcb38ed201d5dbf

TBR=robertphillips@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:2889

Review URL: https://codereview.chromium.org/642493003

10 years agouse SkDocument to capture pdfs from SampleApp 'e'
reed [Wed, 8 Oct 2014 18:10:51 +0000 (11:10 -0700)]
use SkDocument to capture pdfs from SampleApp 'e'

BUG=skia:
TBR=
NOTRY=True

Review URL: https://codereview.chromium.org/634293004

10 years agoUse presence of a content key as non-scratch indicator
bsalomon [Wed, 8 Oct 2014 17:48:15 +0000 (10:48 -0700)]
Use presence of a content key as non-scratch indicator

BUG=skia:2889

Review URL: https://codereview.chromium.org/639873002

10 years agoAdd SK_API to GrGpuResource
bsalomon [Wed, 8 Oct 2014 17:24:07 +0000 (10:24 -0700)]
Add SK_API to GrGpuResource
TBR=reed@google.com

Review URL: https://codereview.chromium.org/638963003

10 years agoRebase Nexus 10 gm failure
egdaniel [Wed, 8 Oct 2014 16:38:30 +0000 (09:38 -0700)]
Rebase Nexus 10 gm failure

TBR=robertphillips@google.com

BUG=skia:

Review URL: https://codereview.chromium.org/638043002

10 years agopre-build the resources library for the fiddle jail
Greg Humphreys [Wed, 8 Oct 2014 16:36:42 +0000 (12:36 -0400)]
pre-build the resources library for the fiddle jail

BUG=skia:
TBR=jcgregorio

Review URL: https://codereview.chromium.org/637353003

10 years agorepair the webtry setup process
Greg Humphreys [Wed, 8 Oct 2014 16:33:31 +0000 (12:33 -0400)]
repair the webtry setup process

The previous version of the script didn't properly update the code for the webtry
user, so only changes to skia would really be pulled, not changes to the actual
webtry server itself.

This version of the setup script is 100% automated and brings everything up to date.

It's also a little more verbose about what it's doing, so it's easier to debug.

BUG=skia:
R=tfarina@chromium.org
TBR=jcgregorio, mtklein, tfarina

Review URL: https://codereview.chromium.org/636173003

10 years agoallow fiddles to use the futex system call; fixes fonts finally
Greg Humphreys [Wed, 8 Oct 2014 16:21:41 +0000 (12:21 -0400)]
allow fiddles to use the futex system call; fixes fonts finally

BUG=skia:

Review URL: https://codereview.chromium.org/639013002

10 years agoRearrange code in TextContexts to be more consistent and match style guide.
jvanverth [Wed, 8 Oct 2014 16:07:27 +0000 (09:07 -0700)]
Rearrange code in TextContexts to be more consistent and match style guide.

Review URL: https://codereview.chromium.org/641613003

10 years agoIn the generated fiddle, set the portable fonts flag first.
Greg Humphreys [Wed, 8 Oct 2014 16:04:20 +0000 (12:04 -0400)]
In the generated fiddle, set the portable fonts flag first.

BUG=skia:
TBR=jcgregorio

Review URL: https://codereview.chromium.org/635373002

10 years agoGrResourceCache2 manages scratch texture.
bsalomon [Wed, 8 Oct 2014 15:40:09 +0000 (08:40 -0700)]
GrResourceCache2 manages scratch texture.

BUG=skia:2889

Review URL: https://codereview.chromium.org/608883003

10 years agoMake GM default to reseting gpu contexts on android.
bsalomon [Wed, 8 Oct 2014 15:17:11 +0000 (08:17 -0700)]
Make GM default to reseting gpu contexts on android.

TBR=mtklein@google.com

Committed: https://skia.googlesource.com/skia/+/8dd35ce178e4319685a870faf544a0b1d7aef5bb

Review URL: https://codereview.chromium.org/635293003

10 years agopreliminary support for fiddle font use
Greg Humphreys [Wed, 8 Oct 2014 15:09:21 +0000 (11:09 -0400)]
preliminary support for fiddle font use

BUG=skia:2998
TBR=jcgregorio, tfarina

Review URL: https://codereview.chromium.org/639833003

10 years agoFix conditional check when copying coverage stages in ODS
egdaniel [Wed, 8 Oct 2014 14:39:19 +0000 (07:39 -0700)]
Fix conditional check when copying coverage stages in ODS

BUG=skia:

Review URL: https://codereview.chromium.org/639773004