Attempt to fix hanging WebRequestBlocking.

Attempt to fix hanging WebRequestBlocking.

I found that it hangs while waiting for the tab to navigate after
complexLoadCancelled. I'm guessing that it could be a timing issue of
navigating after cancelling a load, so the fix is to use a fresh tab for the remaining tests.

BUG=91715
TEST=no

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=97152

[Matt Perry, 2011-08-11 22:54:58]

Nevermind, this fails the try bots. It must be something else.

[Matt Perry, 2011-08-17 00:20:35]

OK, this is ready for a look-see again. This seems to make the trybots happy.
We'll see how it goes on the main builders.

[jochen (gone - plz use gerrit), 2011-08-17 11:21:43]

On 2011/08/17 00:20:35, Matt Perry wrote:
> OK, this is ready for a look-see again. This seems to make the trybots happy.
> We'll see how it goes on the main builders.

are you sure this isn't hiding a bug? Like all requests being blocked for that
specific tab?

[Matt Perry, 2011-08-17 17:09:51]

On 2011/08/17 11:21:43, jochen wrote:
> On 2011/08/17 00:20:35, Matt Perry wrote:
> > OK, this is ready for a look-see again. This seems to make the trybots
happy.
> > We'll see how it goes on the main builders.
> 
> are you sure this isn't hiding a bug? Like all requests being blocked for that
> specific tab?

Navigation is a fickle beast, and I wouldn't be surprised if there were some
race condition here that we're hiding. On the other hand, I've spent enough time
looking at this that I've given up, and just want to reenable the test so we
have at least some coverage here.

[Matt Perry, 2011-08-17 17:11:17]

On 2011/08/17 17:09:51, Matt Perry wrote:
> On 2011/08/17 11:21:43, jochen wrote:
> > On 2011/08/17 00:20:35, Matt Perry wrote:
> > > OK, this is ready for a look-see again. This seems to make the trybots
> happy.
> > > We'll see how it goes on the main builders.
> > 
> > are you sure this isn't hiding a bug? Like all requests being blocked for
that
> > specific tab?
> 
> Navigation is a fickle beast, and I wouldn't be surprised if there were some
> race condition here that we're hiding. On the other hand, I've spent enough
time
> looking at this that I've given up, and just want to reenable the test so we
> have at least some coverage here.

BTW, the problem only shows up on the buildbots. It works fine on my machine
100% of the time. So it's most likely a timing issue.

[jochen (gone - plz use gerrit), 2011-08-17 17:23:08]

LGTM then

Can you add this to the spreadsheet so we don't lose track of this?

[commit-bot: I haz the power, 2011-08-17 17:27:49]

Presubmit check for 7635003-3001 failed and returned exit status 1.

Running presubmit commit checks ...

** Presubmit Warnings **
License must match:
.*? Copyright \(c\) 2011 The Chromium Authors\. All rights reserved\.\n.*? Use
of this source code is governed by a BSD-style license that can be\n.*? found in
the LICENSE file\.\n
Found a bad license header in these files:
  chrome/test/data/extensions/api_test/webrequest/framework.js

Was the presubmit check useful? Please send feedback & hate mail to
maruel@chromium.org!