Relaxes window size DCHECKs in StatusBubbleMac.

Relaxes window size DCHECKs in StatusBubbleMac.

These DCHECKs are firing in telemetry tests on the GPU bots.  Removes the
DCHECKs and instead adds debug logging to help track down the root cause.

Since the window size invariants are not holding in some situations, this CL
also contains code to fixup the window size if it is wrong.

BUG=464754
TEST=None

Committed: https://crrev.com/014fd0a1313597e72bc99a46795cb778a8e40149
Cr-Commit-Position: refs/heads/master@{#321000}

[rohitrao (ping after 24h), 2015-03-17 18:10:52]

rohitrao@chromium.org changed reviewers:
+ erikchen@chromium.org, kbr@chromium.org

[erikchen, 2015-03-17 18:14:31]

https://codereview.chromium.org/1018693003/diff/1/chrome/browser/ui/cocoa/sta...
File chrome/browser/ui/cocoa/status_bubble_mac.mm (right):

https://codereview.chromium.org/1018693003/diff/1/chrome/browser/ui/cocoa/sta...
chrome/browser/ui/cocoa/status_bubble_mac.mm:214: if
(!CGSizeEqualToSize(frame.size, ui::kWindowSizeDeterminedLater.size)) {
You may want to wrap this logic so that it doesn't fire on release builds.

[erikchen, 2015-03-17 18:15:38]

https://codereview.chromium.org/1018693003/diff/1/chrome/browser/ui/cocoa/sta...
File chrome/browser/ui/cocoa/status_bubble_mac.mm (right):

https://codereview.chromium.org/1018693003/diff/1/chrome/browser/ui/cocoa/sta...
chrome/browser/ui/cocoa/status_bubble_mac.mm:214: if
(!CGSizeEqualToSize(frame.size, ui::kWindowSizeDeterminedLater.size)) {
On 2015/03/17 18:14:31, erikchen wrote:
> You may want to wrap this logic so that it doesn't fire on release builds. 

I'm referring to the logs messages and stack_traces.

[rohitrao (ping after 24h), 2015-03-17 18:17:21]

kbr@ can chime in, but I think the bot in question is running a release build?

My current plan is to rip the logging out as soon as I get a hit on the bots,
probably within a day or two.  The logging will go out to canary, but shouldn't
last for too long.

[erikchen, 2015-03-17 18:20:02]

On 2015/03/17 18:17:21, rohitrao wrote:
> kbr@ can chime in, but I think the bot in question is running a release build?
> 
> My current plan is to rip the logging out as soon as I get a hit on the bots,
> probably within a day or two.  The logging will go out to canary, but
shouldn't
> last for too long.

The stack traces might not be as helpful as you want, if this is a release
build...

File a crbug, indicating that the log statements are temporary and should be
removed in a couple of days, and include the crbug in your TODOs. lgtm.

[rohitrao (ping after 24h), 2015-03-17 19:10:23]

The CQ bit was checked by rohitrao@chromium.org

[rohitrao (ping after 24h), 2015-03-17 19:10:24]

The patchset sent to the CQ was uploaded after l-g-t-m from
erikchen@chromium.org
Link to the patchset: https://codereview.chromium.org/1018693003/#ps20001
(title: "Bug filed.")

[commit-bot: I haz the power, 2015-03-17 19:10:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1018693003/20001

[commit-bot: I haz the power, 2015-03-17 22:26:05]

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

[commit-bot: I haz the power, 2015-03-17 22:26:06]

Try jobs failed on following builders:
  linux_chromium_rel_ng on tryserver.chromium.linux (JOB_TIMED_OUT, no build
URL)
Timed out jobs are not retried to avoid causing additional load on the builders
with large pending queues.

[rohitrao (ping after 24h), 2015-03-17 22:28:43]

The CQ bit was checked by rohitrao@chromium.org

[commit-bot: I haz the power, 2015-03-17 22:29:08]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1018693003/20001

[commit-bot: I haz the power, 2015-03-17 22:34:46]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-03-17 22:35:24]

Patchset 2 (id:??) landed as
https://crrev.com/014fd0a1313597e72bc99a46795cb778a8e40149
Cr-Commit-Position: refs/heads/master@{#321000}

[Ken Russell (switch to Gerrit), 2015-03-17 22:36:23]

lgtm overall.

https://codereview.chromium.org/1018693003/diff/20001/chrome/browser/ui/cocoa...
File chrome/browser/ui/cocoa/status_bubble_mac.mm (right):

https://codereview.chromium.org/1018693003/diff/20001/chrome/browser/ui/cocoa...
chrome/browser/ui/cocoa/status_bubble_mac.mm:219: frame.size =
ui::kWindowSizeDeterminedLater.size;
Note: I'm not sure how well base::debug::StackTrace works in non-debug builds. I
might have suggested surrounding this logging code with:

#if DCHECK_IS_ON()

(The bots build Release, with dcheck_always_on=1.)