| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
icon dimensions, text positioning are same as IE infobars. Buttons now implement user selected theme.
BUG=None
TEST=None
Review URL: http://codereview.chromium.org/6321025
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78311 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
works correctly
in debug builds without firing any DCHECKs. Registered the proxy dictionary pref which tripped up
a recent change to add a system level URLRequestContext.
BUG=none
TEST=chrome frame net tests should run in debug builds without firing DCHECKs
Review URL: http://codereview.chromium.org/6697029
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78306 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=76112
Review URL: http://codereview.chromium.org/6673026
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78255 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
make the
ChromeContentBrowserClient pointer which is passed to the ContentClient object a scoped_ptr
We were passing a stack pointer which caused the net tests to crash at exit.
BUG=none
TEST=chrome frame net tests should run to completion.
TBR=amit
Review URL: http://codereview.chromium.org/6696003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78114 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=tsepez
Review URL: http://codereview.chromium.org/6682033
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78099 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
78037 which
adds a ContentBrowserClient interface which serves as a delegate to the content code.
Fix is to initialize the ContentBrowserClient interface as part of the initial bootstapping
code in the chrome frame net tests suite.
BUG=none
TEST=chrome frame net tests should run to completion.
Review URL: http://codereview.chromium.org/6685060
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@78073 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
their own file.
TBR=jianli
Review URL: http://codereview.chromium.org/6681028
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77988 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
anymore.
BUG=75842
TEST=
TBR=jshin@chromium.org
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77916 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=avi
Review URL: http://codereview.chromium.org/6684015
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77903 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
There are probably a few more bits and pieces that can be removed, but I think this is the majority of it.
BUG=51934
Review URL: http://codereview.chromium.org/6576020
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77888 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
browser. We issue
the NavigateWithBindCtx interface call to issue a navigation such that it reuses the existing
bind context. This basically results in a cross apartment deadlock at times when we are in the
context of a OnStopBinding call in our bind status callback implementation. The Navigate call
is marshaled to a different thread which tries to grab a urlmon critical section which is
held in the context of the OnStopBinding call by the current thread.
Fix is to use PostMessage to ensure that we unwind the current call chain.
BUG=73985
TEST=As described in the bug.
Review URL: http://codereview.chromium.org/6677004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77732 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is to avoid having to change a bunch of build/release configuration files
everytime ICU version is changed.
This will bring in the following two CLs in ICU and googleurl:
http://codereview.chromium.org/6625032
http://codereview.chromium.org/6623060
BUG=74076
TEST=Chrome can be built and a binary built runs fine with an accompanying Chrome CL in place.
Review URL: http://codereview.chromium.org/6624060
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77684 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
Fourth part of the remaining files. The last one will be deleting the stub files.
BUG=None
TEST=None
Review URL: http://codereview.chromium.org/6657003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77511 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=mark
Review URL: http://codereview.chromium.org/6612037
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77485 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
user IE8 builder.
The flakiness occurs because IE starts up before chrome_frame_helper which results in a failure
to register a BHO for the IE window.
Fix is to make the ChromeFrame registrar object a member of the chrome frame net test suite.
Rearranged some code which ensures that we can now run chrome_frame_net_tests in debug builds
without annoying DCHECKs
BUG=none
TEST=chrome frame net tests.
Review URL: http://codereview.chromium.org/6650003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77362 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NTLM authentication.
The bug was that the POST request would be issued to the server with no data.
Reason: The cbstgmedData member in the BINDINFO structure passed into our
IBindStatusCallback::GetBindInfo implementation needs to be initialized to the post
data length. Not setting this causes urlmon to treat this as a chunked upload thus
causing it to not set the Content-Length header in the outgoing POST request.
Many thanks to Igor Tandetnik for pointing this out.
Removed the code to set the Content-Length header in our BeginningTransaction implementation
as this is no longer needed.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=62687
BUG=62687
TEST=manually at this point with an IIS web server configured for NTLM authentication.
Review URL: http://codereview.chromium.org/6651003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77348 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Enable ChromeFrame net tests. These tests regressed with the change in the chunked upload
code in net to move the formatting of individual chunks to the http network stack.
This caused the URLRequestTestHTTP.TestPostChunkedDataBeforeStart test to hang in ChromeFrame.
Fix is to ensure that the individual chunks are properly formatted while marshaling them over
IPC.
BUG=none
TEST=chrome frame net tests should pass.
TBR=amit
Review URL: http://codereview.chromium.org/6621024
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77010 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
change in the chunked upload
code in net to move the formatting of individual chunks to the http network stack.
This caused the URLRequestTestHTTP.TestPostChunkedDataBeforeStart test to hang in ChromeFrame.
Fix is to ensure that the individual chunks are properly formatted while marshaling them over
IPC.
BUG=none
TEST=chrome frame net tests should pass.
Review URL: http://codereview.chromium.org/6626035
TBR=ananta@chromium.org
Review URL: http://codereview.chromium.org/6628029
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77004 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
chunked upload
code in net to move the formatting of individual chunks to the http network stack.
This caused the URLRequestTestHTTP.TestPostChunkedDataBeforeStart test to hang in ChromeFrame.
Fix is to ensure that the individual chunks are properly formatted while marshaling them over
IPC.
BUG=none
TEST=chrome frame net tests should pass.
Review URL: http://codereview.chromium.org/6626035
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@77002 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=carnitas
TEST=compiles
Review URL: http://codereview.chromium.org/6609008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76962 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
get triggered running locally and the tests appear to fail on related changes.
Also, they are not instantiating an AtExitManager early enough, so the call to chrome::RegisterPathProvider() may fail. This might have been reordered into a working state in some builds.
Also, fix a call into profile code before registering certain needed preferences which causes a DCHECK.
BUG=None
TEST=None
Review URL: http://codereview.chromium.org/6627009
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76927 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
the webserver requested NTLM
authentication. This is due to a bug in urlmon on IE6 and IE7 which manifests itself when
the post data is passed to urlmon as an IStream.
Fix is to pass in the uploaded data as a HGLOBAL. We always pass in a copy of the HGLOBAL
which points to the posted data to urlmon. This is to have it accessible for reissuing
navigation requests which target downloads.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=62687
BUG=62687
TEST=manually at this point. As we need a server which supports NTLM authentication like IIS.
Review URL: http://codereview.chromium.org/6603006
TBR=ananta@chromium.org
Review URL: http://codereview.chromium.org/6626008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76886 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
requested NTLM
authentication. This is due to a bug in urlmon on IE6 and IE7 which manifests itself when
the post data is passed to urlmon as an IStream.
Fix is to pass in the uploaded data as a HGLOBAL. We always pass in a copy of the HGLOBAL
which points to the posted data to urlmon. This is to have it accessible for reissuing
navigation requests which target downloads.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=62687
BUG=62687
TEST=manually at this point. As we need a server which supports NTLM authentication like IIS.
Review URL: http://codereview.chromium.org/6603006
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76880 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=avi
Review URL: http://codereview.chromium.org/6596103
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76552 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=carnitas
TEST=compiles; existing unit tests.
Review URL: http://codereview.chromium.org/6602049
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76467 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
tests. It appears that
we don't receive the OnQuit notification reliably after invoking IWebBrowser2::Quit. Changing
the corresponding expectation to a LOG.
BUG=none
TEST=ChromeFrame tests.
TBR=amit
Review URL: http://codereview.chromium.org/6591061
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76402 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
| |
TBR=avi
Review URL: http://codereview.chromium.org/6591066
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76390 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fix a bug in the DllRedirector module loading code.
Expand the alternate version generator code to provide a function for generating new versions of single modules.
Remove the LI SACL from the shared memory lock for Win XP.
BUG=NONE
TEST=NONE
Review URL: http://codereview.chromium.org/6600001
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76338 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
browser_thread.h for another change, since 450 files include it.
TBR=avi
Review URL: http://codereview.chromium.org/6596060
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76326 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
which fails on
IE6 after the change to target downloads in a new window. Testing if this fixes it.
BUG=none
TEST=CF_DownloadFileFromPost passes on all CF builders.
TBR=amit
Review URL: http://codereview.chromium.org/6588070
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76310 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
need to pass
the post data as well while reissuing the navigation. This ensures that a POST request is
correctly sent out at all times instead of a GET request. While the assumption is that reusing
the bind context would achieve this behavior, it does not appear to be true at all times.
While fixing this I also found that reissuing a navigation via a NavigateWithBindCtx call
causes the current chrome frame document to receive an Unload call at times, leading to
the page being rendered useless after the request finishes. We should reissue the navigation
on a new window to work around this problem.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=73985
BUG=73985
TEST=As described in the bug.
Review URL: http://codereview.chromium.org/6598016
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@76282 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
when multiple
ChromeFrame versions are loaded in IE. We have code in ChromeFrame which attempts to fallback
to the older version in case it is already loaded. This is done by creating a named shared memory
block which contains the version of the loaded dll. Newer versions look at this block and fallback
to the older version if any. However due to a small oversight this never worked as the shared
memory block was created unnamed. Thus the fallbacks never occurred leading to multiple
chrome frame binaries attempting to patch urlmon, etc.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=74047
BUG=74047
Review URL: http://codereview.chromium.org/6581045
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75969 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A bunch of hard-coded references to icudt42.dll are updated to
refer to icudt46.dll.
See also http://codereview.chromium.org/6575042/ (and there's one more accompanying CL - internal).
I'll follow up this CL with a change to rename icudtXX.dll to icudt.dll
so that I don't have to do this again next time.
BUG=73991,61514
TEST=daily archived builds have icudt46.dll
Review URL: http://codereview.chromium.org/6580046
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75961 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These failures occurred
because of new expectations in these tests which match the server ip address and port being
connected to. We need to return that information from ChromeFrame.
Fix is to send over the corresponding HostPortPair structure in the AutomationMsg_RequestStarted
IPC message. A major part of this CL is basically the groundwork to pass this information over
via automation.
BUG=none
TEST=chrome frame net tests should pass.
Review URL: http://codereview.chromium.org/6575027
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75942 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
for ChromeFrame
as it uses the host (IE's) network stack.
BUG=none
TEST=chrome frame net tests.
TBR=amit
Review URL: http://codereview.chromium.org/6580015
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75819 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
append the anchor
portion of the url retrieved from the navigation manager to the url retrieved from the
protocol sink wrapper. This is needed as the protocol sink wrapper sees the url without
the anchor portion.
BUG=71860
TEST=ChromeFrame anchor url tests.
TBR=amit
Review URL: http://codereview.chromium.org/6546081
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75704 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
navigate chrome to
the correct URL to ensure that scripts on the page see the correct URL. This functionality
regressed when the protocol sink patch landed.
Fix is to read the URL from the ProtData object which wraps the underlying protocol sink,
when our active document loads.
Removed stale code from the Chromeframe moniker patch which eventually needs to go.
Fixes bug http://code.google.com/p/chromium/issues/detail?can=2&q=71860
BUG=71860
TEST=As described in the bug.
Review URL: http://codereview.chromium.org/6469103
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75656 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
BUG=73425
TEST=NONE
Review URL: http://codereview.chromium.org/6549022
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75635 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
active document. This
fixes parts of bug http://code.google.com/p/chromium/issues/detail?id=55764
On IE9 a border is still drawn around all mshtml windows, which does not happen with ChromeFrame.
This needs more investigation.
BUG=55764
TEST=Launch IE9 and navigate to a ChromeFrame enabled page. The ChromeFrame page should not
have a client edge border.
Review URL: http://codereview.chromium.org/6531022
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75427 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(Third attempt to land http://codereview.chromium.org/6409040/ -- now without memory leaks)
- cloud_policy.proto autogenerated from policy_templates.json
- C++ method decoding the protobuf also autogenerated from policy_templates.json
- changed policy fetching mechanism to fetch new-style policy protobufs
BUG=68309, chromium-os:11253, chromium-os:11255
TEST=CloudPolicyCacheTest.*; also manual test against python testserver
Review URL: http://codereview.chromium.org/6532019
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75259 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
Merged correctly this time.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/6523067
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75165 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
a WinInet API
InternetSetOption.
Fixes bug http://code.google.com/p/chromium/issues/detail?id=50328
BUG=50328
TEST=none
Review URL: http://codereview.chromium.org/6480092
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75162 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The bug was that the ProfilePolicyContext was getting initialized before the ProfileImplIOData::Handle was initialized.
BUG=67237
TEST=none
Review URL: http://codereview.chromium.org/6526029
TBR=willchan@chromium.org
Review URL: http://codereview.chromium.org/6480120
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75152 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
The bug was that the ProfilePolicyContext was getting initialized before the ProfileImplIOData::Handle was initialized.
BUG=67237
TEST=none
Review URL: http://codereview.chromium.org/6526029
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75143 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(Second attempt to land http://codereview.chromium.org/6409040/)
- cloud_policy.proto autogenerated from policy_templats.json
- C++ method decoding the protobuf also autogenerated from policy_templates.json
- changed policy fetching mechanism to fetch new-style policy protobufs
BUG=68309, chromium-os:11253, chromium-os:11255
TEST=CloudPolicyCacheTest.*; also manual test against python testserver
Review URL: http://codereview.chromium.org/6523058
TBR=jkummerow@chromium.org
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75115 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(Second attempt to land http://codereview.chromium.org/6409040/)
- cloud_policy.proto autogenerated from policy_templats.json
- C++ method decoding the protobuf also autogenerated from policy_templates.json
- changed policy fetching mechanism to fetch new-style policy protobufs
BUG=68309, chromium-os:11253, chromium-os:11255
TEST=CloudPolicyCacheTest.*; also manual test against python testserver
Review URL: http://codereview.chromium.org/6523058
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75106 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- cloud_policy.proto autogenerated from policy_templates.json
- C++ method decoding the protobuf autogenerated from policy_templates.json
- changed policy fetching mechanism to fetch new-style policy protobufs
BUG=68309, chromium-os:11253, chromium-os:11255
TEST=CloudPolicyCacheTest.*; also manual test against python testserver
Review URL: http://codereview.chromium.org/6409040
TBR=jkummerow@chromium.org
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75084 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- cloud_policy.proto autogenerated from policy_templates.json
- C++ method decoding the protobuf autogenerated from policy_templates.json
- changed policy fetching mechanism to fetch new-style policy protobufs
BUG=68309, chromium-os:11253, chromium-os:11255
TEST=CloudPolicyCacheTest.*; also manual test against python testserver
Review URL: http://codereview.chromium.org/6409040
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@75082 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=71130
Review URL: http://codereview.chromium.org/6523009
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@74966 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The crash occurs while processing an accelerator reflected from chrome and based on
the crash dump occurs while dereferencing a NULL IOleInPlaceSite member.
Fix is to check for the same.
BUG=72957
TEST=none at this point.
Review URL: http://codereview.chromium.org/6524003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@74883 0039d316-1c4b-4281-b951-d872f2087c98
|