Switch non-MPD nanobench path to use a separate canvas per tile

Switch non-MPD nanobench path to use a separate canvas per tile


It is desirable that, when layer hoisting is disabled, the MPD and non-MPD timings be
roughly the same. Unfortunately, using a separate canvas for each tile (a requirement
for MPD) introduces its own discrepancy into the timing. Using a separate canvas for
each tile doesn't seem to make a difference for 8888 (see the non-MPD 8888 column below)
but slows down GPU rendering (see the non-MPD GPU column below). Since this is how
Chromium renders I propose switching to this regimen (even though it is "slowing down"
GPU rendering).


nanobench mean times (ms) with layer hoisting disabled (for desk_amazon.skp)


8888
                          MPD     non-MPD
1 canvas (old-style)     0.628    1.71
separate (new-style)     0.795    1.63

GPU
                          MPD     non-MPD
1 canvas (old-style)     2.34     1.69
separate (new-style)     2.32     2.66

Committed: https://skia.googlesource.com/skia/+/a3e52724ac8b9fa7b48507bff4fa8e558a213e49

[robertphillips, 2014-12-03 16:21:40]

robertphillips@google.com changed reviewers:
+ bsalomon@google.com, mtklein@google.com

[robertphillips, 2014-12-09 18:15:35]

The CQ bit was checked by robertphillips@google.com

[commit-bot: I haz the power, 2014-12-09 18:16:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/779643002/20001

[commit-bot: I haz the power, 2014-12-09 18:16:40]

Note for Reviewers:
The CQ is waiting for an approval.
If you believe that the CL is not ready yet, or if you would like to L-G-T-M
with comments then please uncheck the CQ checkbox.

Waiting for LGTM from valid reviewer(s) till 2014-12-10 00:16 UTC

[robertphillips, 2014-12-09 18:23:43]

The CQ bit was unchecked by robertphillips@google.com

[bsalomon, 2014-12-09 18:26:10]

lgtm

[robertphillips, 2014-12-09 18:27:05]

The CQ bit was checked by robertphillips@google.com

[commit-bot: I haz the power, 2014-12-09 18:27:45]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/779643002/20001

[commit-bot: I haz the power, 2014-12-09 18:27:58]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/a3e52724ac8b9fa7b48507bff4fa8e558a213e49

[robertphillips, 2014-12-09 20:52:29]

It appears that the additional expense when using multiple surfaces in the GPU
path is due to render target swapping. I've timed the old method (rendering into
a single canvas with clips) with and without transforms on the drawPicture call
(but keeping the overall content the same) and the times are roughly the same.