remove unused cachedBitmap from ImageBufferSurface

remove unused cachedBitmap from ImageBufferSurface

BUG=

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=198029

[reed1, 2015-06-29 21:08:59]

The CQ bit was checked by reed@google.com to run a CQ dry run

[commit-bot: I haz the power, 2015-06-29 21:09:11]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1217893004/1

[reed1, 2015-06-29 21:10:01]

The CQ bit was unchecked by reed@google.com

[reed1, 2015-06-29 21:48:23]

reed@google.com changed reviewers:
+ bsalomon@chromium.org, junov@chromium.org, kbr@chromium.org

[reed1, 2015-06-29 21:48:23]

ptal

[Ken Russell (switch to Gerrit), 2015-06-30 01:07:05]

Nice catch. LGTM

[reed2, 2015-06-30 01:19:16]

The CQ bit was checked by reed@chromium.org

[commit-bot: I haz the power, 2015-06-30 01:19:34]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1217893004/1

[commit-bot: I haz the power, 2015-06-30 01:23:24]

Committed patchset #1 (id:1) as
https://src.chromium.org/viewvc/blink?view=rev&revision=198029

[Justin Novosad, 2015-06-30 01:59:40]

Hmmm, this is indeed unused, but it is not supposed to be. This code was part of
an optimization to retain CPU-side copies of GPU-backed canvases to avoid
repetitive readbacks when drawing from an accelerated canvas to a
non-accelerated canvas. It looks like that optimization was accidentally
destroyed by an earlier refactoring.

[bsalomon, 2015-06-30 13:36:50]

On 2015/06/30 01:59:40, Justin Novosad wrote:
> Hmmm, this is indeed unused, but it is not supposed to be. This code was part
of
> an optimization to retain CPU-side copies of GPU-backed canvases to avoid
> repetitive readbacks when drawing from an accelerated canvas to a
> non-accelerated canvas. It looks like that optimization was accidentally
> destroyed by an earlier refactoring.

Doesn't Skia do this caching now in the SkResourceCache?

[reed1, 2015-06-30 13:54:46]

We cache the read-back from an SkImage (that is texture-backed), so using
texture-SkImages should be efficient if drawn to a raster target (or encoded for
DataURL)