Add support for blending of LCD for all blend modes.

Add support for blending of LCD for all blend modes.
BUG=skia:

Committed: https://skia.googlesource.com/skia/+/723b0501e22373bb3e6c306daaceae02cda8a124

[egdaniel, 2015-09-02 21:16:56]

egdaniel@google.com changed reviewers:
+ bsalomon@google.com, joshualitt@chromium.org, jvanverth@google.com,
robertphillips@google.com

[egdaniel, 2015-09-02 21:16:56]

So this CL adds LCD blending support for all porter duff and advanced blends. It
will cause some of our blending to become different from raster since raster
drops LCD if it is not src_over. However, on some local tests my src_over and
src with no alpha now look the same where raster does not.

[robertphillips, 2015-09-03 13:28:41]

lgtm (but I can't say I understand this very deeply)

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/GrTextContext.cpp
File src/gpu/GrTextContext.cpp (right):

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/GrTextContext.c...
src/gpu/GrTextContext.cpp:81: bool GrTextContext::ShouldDisableLCD(const
SkPaint& paint) {
Can't we just pass nullptr to AsMode ?

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
File src/gpu/effects/GrPorterDuffXferProcessor.cpp (right):

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
src/gpu/effects/GrPorterDuffXferProcessor.cpp:142: /**
coeef -> coeff ?

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
src/gpu/effects/GrPorterDuffXferProcessor.cpp:750: 
Can/should this go in a header now ?

[egdaniel, 2015-09-03 13:50:38]

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/GrTextContext.cpp
File src/gpu/GrTextContext.cpp (right):

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/GrTextContext.c...
src/gpu/GrTextContext.cpp:81: bool GrTextContext::ShouldDisableLCD(const
SkPaint& paint) {
On 2015/09/03 13:28:41, robertphillips wrote:
> Can't we just pass nullptr to AsMode ?

Oh yeah, i originally had null here, then changed it to also need to check for
coeff modes only, then deleted coeff check when I realized advanced blends work
and forgot to change back to null.

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
File src/gpu/effects/GrPorterDuffXferProcessor.cpp (right):

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
src/gpu/effects/GrPorterDuffXferProcessor.cpp:142: /**
On 2015/09/03 13:28:41, robertphillips wrote:
> coeef -> coeff ?

Done.

https://codereview.chromium.org/1313623002/diff/40001/src/gpu/effects/GrPorte...
src/gpu/effects/GrPorterDuffXferProcessor.cpp:750: 
On 2015/09/03 13:28:41, robertphillips wrote:
> Can/should this go in a header now ?

Done.

[egdaniel, 2015-09-11 13:39:06]

just pinging Brian for a final review, no real rush on this cl though

[bsalomon, 2015-09-11 13:47:53]

lgtm but do we have good GM coverage?

[egdaniel, 2015-09-11 13:57:30]

On 2015/09/11 13:47:53, bsalomon wrote:
> lgtm but do we have good GM coverage?

we have some but weak GM coverage. I can hold off and put together a new GM
before landing this

[egdaniel, 2015-09-14 13:58:18]

added a gm

[bsalomon, 2015-09-14 14:32:45]

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp
File gm/lcdblendmodes.cpp (right):

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:101: canvas->drawText(string.c_str(), string.size(), 0, y,
paint);
should we have a variation that uses variable shading (e.g. gradient)?

[robertphillips, 2015-09-14 16:04:14]

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp
File gm/lcdblendmodes.cpp (right):

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:8: 
This doesn't seem correct ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:13: #include "SkCanvas.h"
Do we need SkPicture*.h ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:22: LcdBlendGM() {
kPointSize ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:23: const int pointSize = 36;
fTextHeight ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:31: }
\n ?
override ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:37: 
override ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:39: 
override ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:40: virtual void onDraw(SkCanvas* canvas) {
this-> ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:88: for (size_t m = 0; m < SK_ARRAY_COUNT(gModes); m++) {
make this one line ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:99: 
Why make a SkString just to get the c_str out ?

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:106: private:
put this last ?

[egdaniel, 2015-09-15 14:26:11]

Cleaned up gm, and added a gradient variation as well

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp
File gm/lcdblendmodes.cpp (right):

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:8: 
On 2015/09/14 16:04:14, robertphillips wrote:
> This doesn't seem correct ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:13: #include "SkCanvas.h"
On 2015/09/14 16:04:14, robertphillips wrote:
> Do we need SkPicture*.h ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:22: LcdBlendGM() {
On 2015/09/14 16:04:14, robertphillips wrote:
> kPointSize ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:23: const int pointSize = 36;
On 2015/09/14 16:04:14, robertphillips wrote:
> fTextHeight ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:31: }
On 2015/09/14 16:04:14, robertphillips wrote:
> \n ?
> override ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:37: 
On 2015/09/14 16:04:14, robertphillips wrote:
> override ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:39: 
On 2015/09/14 16:04:14, robertphillips wrote:
> override ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:40: virtual void onDraw(SkCanvas* canvas) {
On 2015/09/14 16:04:14, robertphillips wrote:
> this-> ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:88: for (size_t m = 0; m < SK_ARRAY_COUNT(gModes); m++) {
On 2015/09/14 16:04:14, robertphillips wrote:
> make this one line ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:99: 
On 2015/09/14 16:04:14, robertphillips wrote:
> Why make a SkString just to get the c_str out ?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:101: canvas->drawText(string.c_str(), string.size(), 0, y,
paint);
On 2015/09/14 14:32:45, bsalomon wrote:
> should we have a variation that uses variable shading (e.g. gradient)?

Done.

https://codereview.chromium.org/1313623002/diff/80001/gm/lcdblendmodes.cpp#ne...
gm/lcdblendmodes.cpp:106: private:
On 2015/09/14 16:04:14, robertphillips wrote:
> put this last ?

Done.

[robertphillips, 2015-09-15 14:34:41]

I think you're good to go.

[bsalomon, 2015-09-15 14:47:47]

lgtm

[egdaniel, 2015-09-15 16:16:15]

The CQ bit was checked by egdaniel@google.com

[egdaniel, 2015-09-15 16:16:16]

The patchset sent to the CQ was uploaded after l-g-t-m from
robertphillips@google.com
Link to the patchset: https://codereview.chromium.org/1313623002/#ps120001
(title: " ")

[commit-bot: I haz the power, 2015-09-15 16:16:25]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1313623002/120001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1313623002/120001

[commit-bot: I haz the power, 2015-09-15 16:17:26]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-09-15 16:17:27]

Try jobs failed on following builders:
  Build-Ubuntu-GCC-x86_64-Release-Trybot on client.skia.compile (JOB_FAILED,
http://build.chromium.org/p/client.skia.compile/builders/Build-Ubuntu-GCC-x86...)

[egdaniel, 2015-09-15 16:25:33]

The CQ bit was checked by egdaniel@google.com

[egdaniel, 2015-09-15 16:25:33]

The patchset sent to the CQ was uploaded after l-g-t-m from bsalomon@google.com,
robertphillips@google.com
Link to the patchset: https://codereview.chromium.org/1313623002/#ps160001
(title: " ")

[commit-bot: I haz the power, 2015-09-15 16:25:49]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1313623002/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1313623002/160001

[commit-bot: I haz the power, 2015-09-15 16:31:43]

Committed patchset #9 (id:160001) as
https://skia.googlesource.com/skia/+/723b0501e22373bb3e6c306daaceae02cda8a124