draft gpu support in nanobench

draft gpu support in nanobench

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/bb6a02823929584231c8e080ee69e7fb1178cbfb

[mtklein, 2014-06-26 01:39:30]

This should let us get a basic idea of the approach profiled.

I haven't done any testing on mobile, so I've ignored swapping buffers for now.

[bsalomon, 2014-06-30 19:58:12]

mostly lgtm.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:34: DEFINE_string(config, "565 8888 gpu nonrendering",
Does regular bench have the other gpu configs (msaa, nvprmsaa, amgle, ...).
We'll probably want them in nanobench.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:37: DEFINE_int32(gpuFrameLag, 10, "Overestimate of maximum
number of frames GPU allows to lag.");
I'd default to 5 or fewer.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:166: if (bench->isSuitableFor(Benchmark::kGPU_Backend)
style nit, don't we do && on prev line?

[mtklein_C, 2014-06-30 23:05:29]

Worth another look.  Adding more GPU modes made me reorganize a bit.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:34: DEFINE_string(config, "565 8888 gpu nonrendering",
On 2014/06/30 19:58:12, bsalomon wrote:
> Does regular bench have the other gpu configs (msaa, nvprmsaa, amgle, ...).
> We'll probably want them in nanobench.

Yup, done.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:37: DEFINE_int32(gpuFrameLag, 10, "Overestimate of maximum
number of frames GPU allows to lag.");
On 2014/06/30 19:58:12, bsalomon wrote:
> I'd default to 5 or fewer.

You got it.

https://codereview.chromium.org/359473004/diff/180001/bench/nanobench.cpp#new...
bench/nanobench.cpp:166: if (bench->isSuitableFor(Benchmark::kGPU_Backend)
On 2014/06/30 19:58:12, bsalomon wrote:
> style nit, don't we do && on prev line?

Dunno, but Done.

[bsalomon, 2014-07-01 13:23:49]

https://codereview.chromium.org/359473004/diff/260001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/260001/bench/nanobench.cpp#new...
bench/nanobench.cpp:124: SkGLContextHelper* gl =
gGrFactory.getGLContext(GrContextFactory::kNative_GLContextType);
The GL context type here needs to match that of the target config. (e.g. angle,
null, nvpr, ...)

[mtklein, 2014-07-01 14:02:43]

https://codereview.chromium.org/359473004/diff/260001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/260001/bench/nanobench.cpp#new...
bench/nanobench.cpp:124: SkGLContextHelper* gl =
gGrFactory.getGLContext(GrContextFactory::kNative_GLContextType);
On 2014/07/01 13:23:49, bsalomon wrote:
> The GL context type here needs to match that of the target config. (e.g.
angle,
> null, nvpr, ...)

Oh, duh.  Done.

[bsalomon, 2014-07-01 14:06:08]

lgtm

https://codereview.chromium.org/359473004/diff/320001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/320001/bench/nanobench.cpp#new...
bench/nanobench.cpp:8: // TODO: clean up before submitting
I submitted a change that changes the below references to this macro to simply 0

[mtklein_C, 2014-07-01 14:08:45]

https://codereview.chromium.org/359473004/diff/320001/bench/nanobench.cpp
File bench/nanobench.cpp (right):

https://codereview.chromium.org/359473004/diff/320001/bench/nanobench.cpp#new...
bench/nanobench.cpp:8: // TODO: clean up before submitting
On 2014/07/01 14:06:08, bsalomon wrote:
> I submitted a change that changes the below references to this macro to simply
0

Perfect.

[mtklein_C, 2014-07-01 14:08:48]

The CQ bit was checked by mtklein@chromium.org

[commit-bot: I haz the power, 2014-07-01 14:08:58]

CQ is trying da patch. Follow status at
 https://skia-tree-status.appspot.com/cq/mtklein@chromium.org/359473004/340001

[commit-bot: I haz the power, 2014-07-01 14:57:19]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  Build-Win7-VS2010-x86-Debug-Trybot on tryserver.skia
(http://108.170.220.76:10117/builders/Build-Win7-VS2010-x86-Debug-Trybot/build...)

[mtklein, 2014-07-01 14:59:02]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2014-07-01 14:59:09]

CQ is trying da patch. Follow status at
 https://skia-tree-status.appspot.com/cq/mtklein@chromium.org/359473004/360001

[mtklein, 2014-07-01 15:16:00]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2014-07-01 15:16:10]

CQ is trying da patch. Follow status at
 https://skia-tree-status.appspot.com/cq/mtklein@chromium.org/359473004/380001

[commit-bot: I haz the power, 2014-07-01 15:43:45]

Change committed as bb6a02823929584231c8e080ee69e7fb1178cbfb