MemoryPressure: sends the signal to the renderer.

MemoryPressure: sends the signal to the renderer.

BUG=251723
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=209686

[bulach, 2013-06-21 13:58:55]

palmer: OWNERS for the new IPC
jam: OWNERS for content
benm: following discussion from the bug, I think this will solve the case you
described.

[benm (inactive), 2013-06-21 14:02:01]

mnaganov, joth fyi

[benm (inactive), 2013-06-21 15:02:06]

This looks great Marcus, and will certainly help us in the WebView keep better
control of memory usage. Thanks!

[joth, 2013-06-21 21:52:13]

https://codereview.chromium.org/17351006/diff/1/content/browser/renderer_host...
File content/browser/renderer_host/render_widget_host_impl.h (right):

https://codereview.chromium.org/17351006/diff/1/content/browser/renderer_host...
content/browser/renderer_host/render_widget_host_impl.h:722:
base::MemoryPressureListener::MemoryPressureLevel level);
I think you want to do this at the RenderProcess, not RenderWidget level.
otherwise we will send N duplicate messages to each renderer for the N widgets
it hosts.

Also, conceptually do we need this to go to other sandbox child processes? Not
sure if we have common IPC code across process types, but if we do we could
share the message.


Android specific question: couldn't the renderers each listen to
ComponentCallbacks2.onTrimMemory themselves? It is quite possible (and likely)
that a render process will get put into the background kill list before the
browser is.

[palmer, 2013-06-21 21:58:04]

IPC security LGTM

[bulach, 2013-06-28 01:15:19]

great idea joth, certainly simplify things for android. ptal.

I'll talk to other people on other platforms about their plans, once we cross
that bridge they may be able to reuse some of the previous patchset.

[joth, 2013-07-01 22:01:06]

lgtm

[commit-bot: I haz the power, 2013-07-02 08:48:00]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/bulach@chromium.org/17351006/8001

[commit-bot: I haz the power, 2013-07-02 11:44:46]

Retried try job too often on linux_aura for step(s) browser_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_aura...

[commit-bot: I haz the power, 2013-07-02 12:00:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/bulach@chromium.org/17351006/8001

[commit-bot: I haz the power, 2013-07-02 12:48:23]

Change committed as 209686