DocumentLoader::isCommittedButEmpty is timing-dependent.

DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty

Committed: https://crrev.com/dbb678baafd71e3af774d34cbac20b860b8f674e
Cr-Commit-Position: refs/heads/master@{#408277}

[Nate Chapin, 2015-10-22 22:47:10]

japhet@chromium.org changed reviewers:
+ dcheng@chromium.org

[Nate Chapin, 2015-10-22 22:47:10]

This is a post-merge followup to https://codereview.chromium.org/1357763002/.

I never got it fully passing bots before. This drops the enough value for data
received and replaces it wit a bit, which is more sane than trying to have
separate enum values for "not done but empty" and "done but empty".

[dcheng, 2015-10-23 00:19:21]

lgtm

[Nate Chapin, 2016-06-30 19:11:53]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[Nate Chapin, 2016-06-30 19:12:28]

Description was changed from

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389
==========

to

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty
==========

[commit-bot: I haz the power, 2016-06-30 19:12:43]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-06-30 20:18:21]

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

[commit-bot: I haz the power, 2016-06-30 20:18:22]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[Nate Chapin, 2016-06-30 23:40:06]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-30 23:40:32]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-01 00:55:31]

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

[commit-bot: I haz the power, 2016-07-01 00:55:31]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Nate Chapin, 2016-07-18 17:41:05]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-18 17:41:58]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Nate Chapin, 2016-07-18 18:54:54]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-18 18:55:38]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Nate Chapin, 2016-07-19 19:10:34]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-19 19:11:08]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-19 19:23:18]

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

[commit-bot: I haz the power, 2016-07-19 19:23:19]

Dry run: Try jobs failed on following builders:
  android_compile_dbg on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_comp...)
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[Nate Chapin, 2016-07-19 20:02:34]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-19 20:03:40]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Nate Chapin, 2016-07-20 18:38:29]

japhet@chromium.org changed reviewers:
+ sgurun@chromium.org

[Nate Chapin, 2016-07-20 18:38:30]

sgurun, would you mind reviewing this test change?

I'm fixing a bug in blink that incorrectly reports empty documents as non-empty,
which in turn leads us to show a blank page when we ought to show an error page.
However, this bugfix breaks
AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.

I believe this is because it expects an onReceievedError callback, which it then
expects to trigger an onPageFinished callback. However, this onPageFinished is
suppressed for error pages, per
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
and the test times out waiting for it.

I think it's sufficient for this test to just verify that it gets the
onReceivedError callback, but I'm far enough out of my area of expertise that
I'm not at all confident. Does that seem right to you?

[sgurun-gerrit only, 2016-07-20 18:46:15]

On 2016/07/20 18:38:30, Nate Chapin wrote:
> sgurun, would you mind reviewing this test change?
> 
> I'm fixing a bug in blink that incorrectly reports empty documents as
non-empty,
> which in turn leads us to show a blank page when we ought to show an error
page.
> However, this bugfix breaks
> AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> 
> I believe this is because it expects an onReceievedError callback, which it
then
> expects to trigger an onPageFinished callback. However, this onPageFinished is
> suppressed for error pages, per
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> and the test times out waiting for it.
> 
> I think it's sufficient for this test to just verify that it gets the
> onReceivedError callback, but I'm far enough out of my area of expertise that
> I'm not at all confident. Does that seem right to you?

himm, actually this is a pretty important piece of callback that is used by
webview apps, so require a close investigation. I will check.

[sgurun-gerrit only, 2016-07-21 23:59:43]

On 2016/07/20 18:46:15, sgurun wrote:
> On 2016/07/20 18:38:30, Nate Chapin wrote:
> > sgurun, would you mind reviewing this test change?
> > 
> > I'm fixing a bug in blink that incorrectly reports empty documents as
> non-empty,
> > which in turn leads us to show a blank page when we ought to show an error
> page.
> > However, this bugfix breaks
> > AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> > 
> > I believe this is because it expects an onReceievedError callback, which it
> then
> > expects to trigger an onPageFinished callback. However, this onPageFinished
is
> > suppressed for error pages, per
> >
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> > and the test times out waiting for it.
> > 
> > I think it's sufficient for this test to just verify that it gets the
> > onReceivedError callback, but I'm far enough out of my area of expertise
that
> > I'm not at all confident. Does that seem right to you?
> 
> himm, actually this is a pretty important piece of callback that is used by
> webview apps, so require a close investigation. I will check.

comparing to the previous patch it seems we don't get onreceivederror anymore
after this change (because it gave an exception in line 685). But then your
final patch to fix is to remove onpagefinished, and add a onreceivederror. I am
kind a confused, I wonder if it became flaky and passed just by chance.

[sgurun-gerrit only, 2016-07-22 00:00:28]

sgurun@chromium.org changed reviewers:
+ gsennton@chromium.org

[sgurun-gerrit only, 2016-07-22 00:01:05]

On 2016/07/21 23:59:43, sgurun wrote:
> On 2016/07/20 18:46:15, sgurun wrote:
> > On 2016/07/20 18:38:30, Nate Chapin wrote:
> > > sgurun, would you mind reviewing this test change?
> > > 
> > > I'm fixing a bug in blink that incorrectly reports empty documents as
> > non-empty,
> > > which in turn leads us to show a blank page when we ought to show an error
> > page.
> > > However, this bugfix breaks
> > > AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> > > 
> > > I believe this is because it expects an onReceievedError callback, which
it
> > then
> > > expects to trigger an onPageFinished callback. However, this
onPageFinished
> is
> > > suppressed for error pages, per
> > >
> >
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> > > and the test times out waiting for it.
> > > 
> > > I think it's sufficient for this test to just verify that it gets the
> > > onReceivedError callback, but I'm far enough out of my area of expertise
> that
> > > I'm not at all confident. Does that seem right to you?
> > 
> > himm, actually this is a pretty important piece of callback that is used by
> > webview apps, so require a close investigation. I will check.
> 
> comparing to the previous patch it seems we don't get onreceivederror anymore
> after this change (because it gave an exception in line 685). But then your
> final patch to fix is to remove onpagefinished, and add a onreceivederror. I
am
> kind a confused, I wonder if it became flaky and passed just by chance.


Gustav, can you take a look at this change? You dealt with these for a while
during shouldoverrideurlloading work.

[gsennton, 2016-07-26 15:34:58]

On 2016/07/20 18:38:30, Nate Chapin wrote:
> sgurun, would you mind reviewing this test change?
> 
> I'm fixing a bug in blink that incorrectly reports empty documents as
non-empty,
> which in turn leads us to show a blank page when we ought to show an error
page.
> However, this bugfix breaks
> AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> 
> I believe this is because it expects an onReceievedError callback, which it
then
> expects to trigger an onPageFinished callback. However, this onPageFinished is
> suppressed for error pages, per
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> and the test times out waiting for it.
> 
> I think it's sufficient for this test to just verify that it gets the
> onReceivedError callback, but I'm far enough out of my area of expertise that
> I'm not at all confident. Does that seem right to you?

To ensure we are on the same page here: by "empty documents" you mean a http
response with an empty body?
It looks like the only reason we are calling onReceivedError in this case is
https://chromiumcodereview.appspot.com/12620004

I have no idea whether or not we will break apps by no longer calling
onPageFinished for empty responses, but I guess we could just post
onPageFinished from shouldInterceptRequest (just like
https://chromiumcodereview.appspot.com/12620004 posts onReceivedError) if we do
end up breaking apps?

[gsennton, 2016-07-26 15:35:19]

https://codereview.chromium.org/1364863002/diff/430001/android_webview/javate...
File
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java
(right):

https://codereview.chromium.org/1364863002/diff/430001/android_webview/javate...
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java:684:
onReceivedErrorHelper.waitForCallback(onReceivedErrorHelperCallCount, 1);
This call is now unnecessary given that you pass onReceivedErrorHelper to
loadUrlSync.

[Nate Chapin, 2016-07-26 18:39:01]

https://codereview.chromium.org/1364863002/diff/430001/android_webview/javate...
File
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java
(right):

https://codereview.chromium.org/1364863002/diff/430001/android_webview/javate...
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java:684:
onReceivedErrorHelper.waitForCallback(onReceivedErrorHelperCallCount, 1);
On 2016/07/26 15:35:19, gsennton wrote:
> This call is now unnecessary given that you pass onReceivedErrorHelper to
> loadUrlSync.

Done.

[Nate Chapin, 2016-07-26 18:40:38]

On 2016/07/26 15:34:58, gsennton wrote:
> On 2016/07/20 18:38:30, Nate Chapin wrote:
> > sgurun, would you mind reviewing this test change?
> > 
> > I'm fixing a bug in blink that incorrectly reports empty documents as
> non-empty,
> > which in turn leads us to show a blank page when we ought to show an error
> page.
> > However, this bugfix breaks
> > AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> > 
> > I believe this is because it expects an onReceievedError callback, which it
> then
> > expects to trigger an onPageFinished callback. However, this onPageFinished
is
> > suppressed for error pages, per
> >
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> > and the test times out waiting for it.
> > 
> > I think it's sufficient for this test to just verify that it gets the
> > onReceivedError callback, but I'm far enough out of my area of expertise
that
> > I'm not at all confident. Does that seem right to you?
> 
> To ensure we are on the same page here: by "empty documents" you mean a http
> response with an empty body?
> It looks like the only reason we are calling onReceivedError in this case is
> https://chromiumcodereview.appspot.com/12620004

Correct. The bug I'm fixing is specific to an empty body on a HTTP 4xx/5xx
response,
but I don't think the status code matters for the android_webview test breakage?

> 
> I have no idea whether or not we will break apps by no longer calling
> onPageFinished for empty responses, but I guess we could just post
> onPageFinished from shouldInterceptRequest (just like
> https://chromiumcodereview.appspot.com/12620004 posts onReceivedError) if we
do
> end up breaking apps?

I'd believe that. Would you like me to emulate the onPageFinished in this CL, or
wait to see if not calling onPageFinished actually breaks something?

[Nate Chapin, 2016-07-26 18:44:17]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-26 18:44:43]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-26 19:42:21]

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

[commit-bot: I haz the power, 2016-07-26 19:42:22]

Dry run: Try jobs failed on following builders:
  android_clang_dbg_recipe on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)

[Nate Chapin, 2016-07-26 19:44:12]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-26 19:44:55]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-26 21:18:45]

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

[commit-bot: I haz the power, 2016-07-26 21:18:47]

Dry run: Try jobs failed on following builders:
  linux_chromium_asan_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[sgurun-gerrit only, 2016-07-27 02:10:10]

On 2016/07/26 18:40:38, Nate Chapin wrote:
> On 2016/07/26 15:34:58, gsennton wrote:
> > On 2016/07/20 18:38:30, Nate Chapin wrote:
> > > sgurun, would you mind reviewing this test change?
> > > 
> > > I'm fixing a bug in blink that incorrectly reports empty documents as
> > non-empty,
> > > which in turn leads us to show a blank page when we ought to show an error
> > page.
> > > However, this bugfix breaks
> > > AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> > > 
> > > I believe this is because it expects an onReceievedError callback, which
it
> > then
> > > expects to trigger an onPageFinished callback. However, this
onPageFinished
> is
> > > suppressed for error pages, per
> > >
> >
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> > > and the test times out waiting for it.
> > > 
> > > I think it's sufficient for this test to just verify that it gets the
> > > onReceivedError callback, but I'm far enough out of my area of expertise
> that
> > > I'm not at all confident. Does that seem right to you?
> > 
> > To ensure we are on the same page here: by "empty documents" you mean a http
> > response with an empty body?
> > It looks like the only reason we are calling onReceivedError in this case is
> > https://chromiumcodereview.appspot.com/12620004
> 
> Correct. The bug I'm fixing is specific to an empty body on a HTTP 4xx/5xx
> response,
> but I don't think the status code matters for the android_webview test
breakage?
> 
> > 
> > I have no idea whether or not we will break apps by no longer calling
> > onPageFinished for empty responses, but I guess we could just post
> > onPageFinished from shouldInterceptRequest (just like
> > https://chromiumcodereview.appspot.com/12620004 posts onReceivedError) if we
> do
> > end up breaking apps?
> 
> I'd believe that. Would you like me to emulate the onPageFinished in this CL,
or
> wait to see if not calling onPageFinished actually breaks something?

I prefer calling onPageFinished to prevent behavioral change. No matter how
subtle the change is, it is not possible for us to monitor behavior of apps (we
don't have UMA). So let's not break if possible.

[sgurun-gerrit only, 2016-07-27 02:13:01]

On 2016/07/27 02:10:10, sgurun wrote:
> On 2016/07/26 18:40:38, Nate Chapin wrote:
> > On 2016/07/26 15:34:58, gsennton wrote:
> > > On 2016/07/20 18:38:30, Nate Chapin wrote:
> > > > sgurun, would you mind reviewing this test change?
> > > > 
> > > > I'm fixing a bug in blink that incorrectly reports empty documents as
> > > non-empty,
> > > > which in turn leads us to show a blank page when we ought to show an
error
> > > page.
> > > > However, this bugfix breaks
> > > > AwContentsClientShouldInterceptRequestTest#testOnReceivedErrorCallback.
> > > > 
> > > > I believe this is because it expects an onReceievedError callback, which
> it
> > > then
> > > > expects to trigger an onPageFinished callback. However, this
> onPageFinished
> > is
> > > > suppressed for error pages, per
> > > >
> > >
> >
>
https://codereview.chromium.org/1001003004/diff/140001/android_webview/java/s...,
> > > > and the test times out waiting for it.
> > > > 
> > > > I think it's sufficient for this test to just verify that it gets the
> > > > onReceivedError callback, but I'm far enough out of my area of expertise
> > that
> > > > I'm not at all confident. Does that seem right to you?
> > > 
> > > To ensure we are on the same page here: by "empty documents" you mean a
http
> > > response with an empty body?
> > > It looks like the only reason we are calling onReceivedError in this case
is
> > > https://chromiumcodereview.appspot.com/12620004
> > 
> > Correct. The bug I'm fixing is specific to an empty body on a HTTP 4xx/5xx
> > response,
> > but I don't think the status code matters for the android_webview test
> breakage?
> > 
> > > 
> > > I have no idea whether or not we will break apps by no longer calling
> > > onPageFinished for empty responses, but I guess we could just post
> > > onPageFinished from shouldInterceptRequest (just like
> > > https://chromiumcodereview.appspot.com/12620004 posts onReceivedError) if
we
> > do
> > > end up breaking apps?
> > 
> > I'd believe that. Would you like me to emulate the onPageFinished in this
CL,
> or
> > wait to see if not calling onPageFinished actually breaks something?
> 
> I prefer calling onPageFinished to prevent behavioral change. No matter how
> subtle the change is, it is not possible for us to monitor behavior of apps
(we
> don't have UMA). So let's not break if possible.

I think the callback should be here:
https://cs.chromium.org/chromium/src/android_webview/java/src/org/chromium/an...

[gsennton, 2016-07-27 09:38:09]

Alright, yeah, I agree with Selim here, let's just emulate the onPageFinished in
this CL.
Could you also change the corresponding comment to indicate that an empty
response won't trigger neither onReceivedError nor onPageFinished, and therefore
we synthesize both callbacks (currently the comment only indicates that we need
to emulate onReceivedError)?

https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
File
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java
(right):

https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java:679:
public void testOnReceivedErrorCallback() throws Throwable {
Reminder: now that we want to emulate the onPageFinished call ourselves this
test should be left unchanged (so that we wait both for onPageFinished and
onReceivedError).

[Nate Chapin, 2016-07-27 21:44:54]

The CQ bit was checked by japhet@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-27 21:45:21]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Nate Chapin, 2016-07-27 22:21:27]

OK, now emulating onPageFinished, PTAL.

https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
File
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java
(right):

https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java:679:
public void testOnReceivedErrorCallback() throws Throwable {
On 2016/07/27 09:38:09, gsennton wrote:
> Reminder: now that we want to emulate the onPageFinished call ourselves this
> test should be left unchanged (so that we wait both for onPageFinished and
> onReceivedError).

Done.

[sgurun-gerrit only, 2016-07-27 22:35:26]

On 2016/07/27 22:21:27, Nate Chapin wrote:
> OK, now emulating onPageFinished, PTAL.
> 
>
https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
> File
>
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java
> (right):
> 
>
https://codereview.chromium.org/1364863002/diff/470001/android_webview/javate...
>
android_webview/javatests/src/org/chromium/android_webview/test/AwContentsClientShouldInterceptRequestTest.java:679:
> public void testOnReceivedErrorCallback() throws Throwable {
> On 2016/07/27 09:38:09, gsennton wrote:
> > Reminder: now that we want to emulate the onPageFinished call ourselves this
> > test should be left unchanged (so that we wait both for onPageFinished and
> > onReceivedError).
> 
> Done.

lgtm, hopefully the tests will pass :)

[commit-bot: I haz the power, 2016-07-27 23:30:32]

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

[commit-bot: I haz the power, 2016-07-27 23:30:34]

Dry run: This issue passed the CQ dry run.

[Nate Chapin, 2016-07-27 23:33:55]

The CQ bit was checked by japhet@chromium.org

[Nate Chapin, 2016-07-27 23:33:55]

The patchset sent to the CQ was uploaded after l-g-t-m from dcheng@chromium.org
Link to the patchset: https://codereview.chromium.org/1364863002/#ps490001
(title: "Emulate onPageFinished")

[commit-bot: I haz the power, 2016-07-27 23:34:18]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-27 23:41:12]

Description was changed from

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty
==========

to

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty
==========

[commit-bot: I haz the power, 2016-07-27 23:41:13]

Committed patchset #26 (id:490001)

[commit-bot: I haz the power, 2016-07-27 23:44:42]

Description was changed from

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty
==========

to

==========
DocumentLoader::isCommittedButEmpty is timing-dependent.

It's possible for it to be checked after the DocumentLoader has finished, at
which point it will incorrectly return false.

BUG=520389, 331745
TEST=DocumentLoaderTest.isCommittedButEmpty

Committed: https://crrev.com/dbb678baafd71e3af774d34cbac20b860b8f674e
Cr-Commit-Position: refs/heads/master@{#408277}
==========

[commit-bot: I haz the power, 2016-07-27 23:44:43]

Patchset 26 (id:??) landed as
https://crrev.com/dbb678baafd71e3af774d34cbac20b860b8f674e
Cr-Commit-Position: refs/heads/master@{#408277}