create and thread batch tracker object

create and thread batch tracker object

I remove the factory on the next CL

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/c3a6eb23483e5d28073b509a5f637f41660de294

Committed: https://skia.googlesource.com/skia/+/87f48d997ec29e5eeaa7567355775e93465dd60d

[joshualitt, 2014-12-02 22:07:54]

joshualitt@google.com changed reviewers:
+ bsalomon@google.com, joshualitt@google.com

[joshualitt, 2014-12-02 22:07:55]

This CL simply creates the BatchTracker object and threads it into GPs.

There are factories in this CL, but I remove them in the next CL.  This CL was
nearly finished when we decided to remove the factories

[bsalomon, 2014-12-03 14:38:20]

mostly looks good. Some questions and request for comments.

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
File src/gpu/GrGeometryProcessor.h (right):

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
src/gpu/GrGeometryProcessor.h:28: // kPreAllocSize is actually more like a
maxsize.
huh?

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
src/gpu/GrGeometryProcessor.h:96: struct InitBT {
comments?

We're getting rid of fColor and fCoverage, right?

[joshualitt, 2014-12-03 14:49:24]

On 2014/12/03 14:38:20, bsalomon wrote:
> mostly looks good. Some questions and request for comments.
> 
>
https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
> File src/gpu/GrGeometryProcessor.h (right):
> 
>
https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
> src/gpu/GrGeometryProcessor.h:28: // kPreAllocSize is actually more like a
> maxsize.
> huh?
Well, we don't really 'push back' onto the batch tracker, we just kind of cast
it.  There is actually no way to push back onto the array.  I could add such a
means, but I really want it to live on the stack. Thoughts?
> 
>
https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
> src/gpu/GrGeometryProcessor.h:96: struct InitBT {
> comments?
> 
> We're getting rid of fColor and fCoverage, right?
I think this has to wait until we have the bundles and deferred geometry system
in place, right?  Eventually, I *think* initBT takes a bundle and a batch
tracker.  Until the whole deferred system is in place though, the GP is going to
have to figure its identity out via interactions with the optstate.

[bsalomon, 2014-12-03 14:58:52]

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
File src/gpu/GrGeometryProcessor.h (right):

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
src/gpu/GrGeometryProcessor.h:16: * A struct for tracking batching decisions. 
While this lives on optstate, it is managed
Can you make optstate be GrOptState? I've noticed that when we don't use actual
class names in our comments, we wind up with stale comments because they get
missed by search/replace.

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
src/gpu/GrGeometryProcessor.h:28: // kPreAllocSize is actually more like a
maxsize.
If it's a max size and it's preallocated then SkSTArray isn't the right thing.
Just stick a uint8_t fData[32] in there.

How does the client know how big this is? Seems like there should be a public
facing enum or something.

https://codereview.chromium.org/772513002/diff/40001/src/gpu/GrGeometryProces...
src/gpu/GrGeometryProcessor.h:96: struct InitBT {
What I meant is that GrGP shouldn't have an input color and coverage and
GrOptState shouldn't have a constant color/coverage. The GP is (or should be)
output only.

[joshualitt, 2014-12-03 19:29:18]

feedback incorporated

[bsalomon, 2014-12-03 19:38:32]

lgtm

https://codereview.chromium.org/772513002/diff/80001/src/gpu/gl/GrGLGeometryP...
File src/gpu/gl/GrGLGeometryProcessor.h (right):

https://codereview.chromium.org/772513002/diff/80001/src/gpu/gl/GrGLGeometryP...
src/gpu/gl/GrGLGeometryProcessor.h:56: to have an identical effect key as the
one that created this GrGLGeometryProcessor.  */
identical processor key?

https://codereview.chromium.org/772513002/diff/80001/src/gpu/gl/GrGLProcessor.h
File src/gpu/gl/GrGLProcessor.h (right):

https://codereview.chromium.org/772513002/diff/80001/src/gpu/gl/GrGLProcessor...
src/gpu/gl/GrGLProcessor.h:123: virtual void setData(const
GrGLProgramDataManager&, const GrProcessor&) {}
Why can't it be fp?

effect->processor

[joshualitt, 2014-12-03 22:45:43]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2014-12-03 22:46:20]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/772513002/100001

[commit-bot: I haz the power, 2014-12-03 22:46:24]

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

[commit-bot: I haz the power, 2014-12-03 22:46:25]

Failed to apply patch for src/gpu/GrOptDrawState.cpp:
While running git apply --index -3 -p1;
  error: patch failed: src/gpu/GrOptDrawState.cpp:122
  Falling back to three-way merge...
  Applied patch to 'src/gpu/GrOptDrawState.cpp' with conflicts.
  U src/gpu/GrOptDrawState.cpp

Patch:       src/gpu/GrOptDrawState.cpp
Index: src/gpu/GrOptDrawState.cpp
diff --git a/src/gpu/GrOptDrawState.cpp b/src/gpu/GrOptDrawState.cpp
index
6f6f0af88821741649829ad680d488e15bdac12a..0d976c786700e45ba93cff7e7869bbcebee33afa
100644
--- a/src/gpu/GrOptDrawState.cpp
+++ b/src/gpu/GrOptDrawState.cpp
@@ -122,6 +122,16 @@ GrOptDrawState::GrOptDrawState(const GrDrawState&
drawState,
 
     // now create a key
     gpu->buildProgramDesc(*this, descInfo, drawType, &fDesc);
+
+    // let the GP init the batch tracker
+    if (drawState.hasGeometryProcessor()) {
+        GrGeometryProcessor::InitBT init;
+        init.fOutputColor = descInfo.fInputColorIsUsed;
+        init.fOutputCoverage = descInfo.fInputCoverageIsUsed;
+        init.fColor = this->getColor();
+        init.fCoverage = this->getCoverage();
+        fGeometryProcessor->initBatchTracker(&fBatchTracker, init);
+    }
 };
 
 void GrOptDrawState::setOutputStateInfo(const GrDrawState& ds,

[joshualitt, 2014-12-03 22:50:08]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2014-12-03 22:50:31]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/772513002/140001

[commit-bot: I haz the power, 2014-12-03 22:54:03]

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

[commit-bot: I haz the power, 2014-12-03 22:54:03]

Try jobs failed on following builders:
  Build-Win-VS2013-x86-Debug-Trybot on client.skia.compile
(http://build.chromium.org/p/client.skia.compile/builders/Build-Win-VS2013-x86...)

[joshualitt, 2014-12-03 23:04:08]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2014-12-03 23:04:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/772513002/140001

[commit-bot: I haz the power, 2014-12-03 23:08:12]

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

[commit-bot: I haz the power, 2014-12-03 23:08:14]

Try jobs failed on following builders:
  Build-Win-VS2013-x86-Debug-Trybot on client.skia.compile
(http://build.chromium.org/p/client.skia.compile/builders/Build-Win-VS2013-x86...)

[joshualitt, 2014-12-03 23:24:49]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2014-12-03 23:25:52]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/772513002/160001

[commit-bot: I haz the power, 2014-12-03 23:34:32]

Committed patchset #9 (id:160001) as
https://skia.googlesource.com/skia/+/c3a6eb23483e5d28073b509a5f637f41660de294

[joshualitt, 2014-12-04 13:59:21]

A revert of this CL (patchset #9 id:160001) has been created in
https://codereview.chromium.org/774133003/ by joshualitt@google.com.

The reason for reverting is: Breaking linux layout test.

[joshualitt, 2014-12-04 18:27:04]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2014-12-04 18:27:26]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/772513002/180001

[commit-bot: I haz the power, 2014-12-04 18:41:44]

Committed patchset #10 (id:180001) as
https://skia.googlesource.com/skia/+/87f48d997ec29e5eeaa7567355775e93465dd60d