aw: Release hardware onTrimMemory(MODERATE)

aw: Release hardware onTrimMemory(MODERATE)

Frameworks destroys the underlying EGLContext on after
certain levels of onTrimMemory, so webview must also release
its hardware draw resources to avoid using stale references.

Factor out release hardware resource code in aw_contents.cc
and browser_view_renderer.cc, and call it in onTrimMemory.

BUG=395657
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284591

[boliu, 2014-07-21 18:09:27]

ptal

[hush (inactive), 2014-07-21 21:14:04]

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
File android_webview/native/aw_contents.cc (right):

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
android_webview/native/aw_contents.cc:866:
shared_renderer_state_.SetInsideHardwareRelease(true);
how about base::AutoReset on a boolean member variable that belongs to
AwContents? (so that we keep less states in SharedRendererState)

[boliu, 2014-07-21 22:34:32]

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
File android_webview/native/aw_contents.cc (right):

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
android_webview/native/aw_contents.cc:866:
shared_renderer_state_.SetInsideHardwareRelease(true);
On 2014/07/21 21:14:04, hush wrote:
> how about base::AutoReset on a boolean member variable that belongs to
> AwContents? (so that we keep less states in SharedRendererState)

That's not thread safe, although something like autoreset would be better...

[hush (inactive), 2014-07-21 22:54:37]

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
File android_webview/native/aw_contents.cc (right):

https://codereview.chromium.org/408803002/diff/20001/android_webview/native/a...
android_webview/native/aw_contents.cc:866:
shared_renderer_state_.SetInsideHardwareRelease(true);
Okay. lgtm
On 2014/07/21 22:34:32, boliu wrote:
> On 2014/07/21 21:14:04, hush wrote:
> > how about base::AutoReset on a boolean member variable that belongs to
> > AwContents? (so that we keep less states in SharedRendererState)
> 
> That's not thread safe, although something like autoreset would be better...

[boliu, 2014-07-21 22:57:53]

And I just implemented the auto reset thing...ptal

[hush (inactive), 2014-07-21 23:16:14]

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
File android_webview/browser/shared_renderer_state.h (right):

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
android_webview/browser/shared_renderer_state.h:62: friend class
InsideHardwareReleaseReset;
You don't need friend class here because you have public setter and getter
already

[boliu, 2014-07-21 23:17:16]

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
File android_webview/browser/shared_renderer_state.h (right):

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
android_webview/browser/shared_renderer_state.h:62: friend class
InsideHardwareReleaseReset;
On 2014/07/21 23:16:14, hush wrote:
> You don't need friend class here because you have public setter and getter
> already

I moved the setter to private?

[hush (inactive), 2014-07-21 23:20:07]

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
File android_webview/browser/shared_renderer_state.h (right):

https://codereview.chromium.org/408803002/diff/40001/android_webview/browser/...
android_webview/browser/shared_renderer_state.h:62: friend class
InsideHardwareReleaseReset;
Ahh.. yes.
Thanks!
lgtm
On 2014/07/21 23:17:16, boliu wrote:
> On 2014/07/21 23:16:14, hush wrote:
> > You don't need friend class here because you have public setter and getter
> > already
> 
> I moved the setter to private?

[boliu, 2014-07-21 23:20:27]

The CQ bit was checked by boliu@chromium.org

[commit-bot: I haz the power, 2014-07-21 23:22:45]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/boliu@chromium.org/408803002/40001

[commit-bot: I haz the power, 2014-07-22 02:09:56]

Change committed as 284591