Make nanobench zoom animation time based

Make nanobench zoom animation time based

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/63a82855b1f0b83952b65fca330954c50ebe7a4b

[Chris Dalton, 2015-06-26 05:34:12]

cdalton@nvidia.com changed reviewers:
+ joshualitt@google.com

[bsalomon, 2015-06-26 13:02:50]

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

[bsalomon, 2015-06-26 13:02:50]

https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp
File bench/SKPAnimationBench.cpp (right):

https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp...
bench/SKPAnimationBench.cpp:50: , fPhaseShift(static_cast<double>(rand()) /
RAND_MAX) {
Why a random shift?

If it's needed, we use SkRandom for platform invariance.

[Chris Dalton, 2015-06-26 20:38:19]

https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp
File bench/SKPAnimationBench.cpp (right):

https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp...
bench/SKPAnimationBench.cpp:50: , fPhaseShift(static_cast<double>(rand()) /
RAND_MAX) {
On 2015/06/26 13:02:50, bsalomon wrote:
> Why a random shift?
> 
> If it's needed, we use SkRandom for platform invariance.

The idea of the shift was to try and randomize what zoom levels it got drawn at.
I thought without a random shift, the first sample would always be about the
same zoom level, which may have influenced the result if not many samples were
drawn.

But come to think of it, we prime the pipeline with a couple draws before we
start timing so it might not matter much at all / might not be worth the extra
complexity?

I looked at SkRandom but it seemed difficult to get a one-off random number
since you need to initialize the object with a seed. (Unless we reused a global
one.)

[bsalomon, 2015-06-26 20:56:20]

On 2015/06/26 20:38:19, Chris Dalton wrote:
> https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp
> File bench/SKPAnimationBench.cpp (right):
> 
>
https://codereview.chromium.org/1211253003/diff/1/bench/SKPAnimationBench.cpp...
> bench/SKPAnimationBench.cpp:50: , fPhaseShift(static_cast<double>(rand()) /
> RAND_MAX) {
> On 2015/06/26 13:02:50, bsalomon wrote:
> > Why a random shift?
> > 
> > If it's needed, we use SkRandom for platform invariance.
> 
> The idea of the shift was to try and randomize what zoom levels it got drawn
at.
> I thought without a random shift, the first sample would always be about the
> same zoom level, which may have influenced the result if not many samples were
> drawn.
> 
> But come to think of it, we prime the pipeline with a couple draws before we
> start timing so it might not matter much at all / might not be worth the extra
> complexity?

I think that's right. For reproducibility I'm wondering if the zoom level should
be reset once the benchmarking begins. I'm worried that if you ran different
params to nanobench you might get different numbers for the same tests.

[bsalomon, 2015-06-26 21:28:25]

lgtm

[Chris Dalton, 2015-06-26 21:31:39]

The CQ bit was checked by cdalton@nvidia.com

[commit-bot: I haz the power, 2015-06-26 21:31:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1211253003/20001

[commit-bot: I haz the power, 2015-06-26 21:31:52]

The author cdalton@nvidia.com has not signed Google Contributor License
Agreement. Please visit https://cla.developers.google.com to sign and manage
CLA.

[commit-bot: I haz the power, 2015-06-26 21:35:17]

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

[commit-bot: I haz the power, 2015-06-26 21:35:18]

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

[Chris Dalton, 2015-06-29 20:49:09]

The CQ bit was checked by cdalton@nvidia.com

[Chris Dalton, 2015-06-29 20:49:09]

The patchset sent to the CQ was uploaded after l-g-t-m from bsalomon@google.com
Link to the patchset: https://codereview.chromium.org/1211253003/#ps40001
(title: "fix msvc")

[commit-bot: I haz the power, 2015-06-29 20:49:21]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1211253003/40001

[commit-bot: I haz the power, 2015-06-29 20:49:23]

The author cdalton@nvidia.com has not signed Google Contributor License
Agreement. Please visit https://cla.developers.google.com to sign and manage
CLA.

[commit-bot: I haz the power, 2015-06-29 21:01:53]

The author cdalton@nvidia.com has not signed Google Contributor License
Agreement. Please visit https://cla.developers.google.com to sign and manage
CLA.

[commit-bot: I haz the power, 2015-06-29 21:06:12]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/63a82855b1f0b83952b65fca330954c50ebe7a4b