New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 600284 link

Starred by 4 users

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Jun 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug-Regression



Sign in to add a comment

Regression: PDF page appears black with white patches after switching tab.

Reported by rk...@etouch.net, Apr 4 2016

Issue description

Chrome Version: 51.0.2699.0 Revision ced319c78e3f9acc95394d81cc6763ed9a9e8371-refs/heads/master@{#384828}(32/64 bit)
OS: Windows(Aero Enabled)

URL: https://chrome.google.com/webstore/detail/pdf-viewer/oemmndcbldboiebfnladdacbdfmadadm?hl=en

What steps will reproduce the problem?
(1) Launch chrome,navigate to above url and click on 'ADD TO CHROME' button.
(2) Open any pdf file,then open NTP and right click on bookmark bar and select 'Show bookmarks bar' option.
(3) Back navigate to previous tab and observe.

PDF page appears black with white patches after switching tab.

PDF page should appear properly.

This is a regression issue,broken in 'M-50' ,below is bisect info:

Good Build: 50.0.2640.0
Bad Build: 50.0.2643.0

Narrow Bisect:
https://chromium.googlesource.com/chromium/src/+log/8471a0a9e42cde08421463086aa1219548e5451a..c3faa5d9e87d0e746ba2ccfdbe6d886221ca47d0?pretty=fuller&n=100

Suspecting: r373599 ?

Note: Issue is not seen on Mac and Linux OS.

@danakj: please help me to reassign this issue,if your change is not cause for it. 

 
 
Actual_PDF.mp4
544 KB Download
Expected_PDF.mp4
456 KB Download

Comment 1 by rk...@etouch.net, Apr 4 2016

Summary: Regression: PDF page appears black with white patches after switching tab. (was: PDF page appears black with white patches after switching tab.)
Labels: ReleaseBlock-Stable
adding RB-label as this is recent regression, please change if required
Kinda suspicious of the bisect, nothing in that range looks related. My change is renaming functions.

Is this issue flaky? Any chance the bisect is wrong?
Cc: tkonch...@chromium.org
Labels: Needs-Feedback
Unable to reproduce the issue on win8.1, win7 chrome version 51.0.2699.0

rkote@, Could you please recheck the issue and if it still exists please provide a re bisect as per comment #3

Comment 5 by rk...@etouch.net, Apr 5 2016

Labels: -Needs-Feedback
Owner: brettw@chromium.org
With respect to comment 3:

Re bisected again for the above issue, getting below narrow bisect range:

Narrow Bisect:
https://chromium.googlesource.com/chromium/src/+log/7da8bd05a6def3df489ab3f237ce510f219ced6a..f246b3224a5371fbca304c15ab663351a76bb836?pretty=fuller&n=100

Suspecting: r373105

Could you please help me to reassign this issue,if your change is not cause for it?
Labels: Needs-Bisect
The above change (r373105) has already been reverted, so it wont be the culprit and also the above steps causing browser crash for builds 50.0.2640.0 & older. So this narrow bisect may not be the correct one.
Components: -Internals>Plugins>PDF
Owner: ----
Status: Untriaged (was: Assigned)
Please remember Chrome extensions that implement PDF viewing should not be filed for the internal PDF viewer component.

From the blame list, maybe r373095?
Cc: junov@chromium.org
Looping Junov@ for the above change per c#7.

Comment 9 by rk...@etouch.net, Apr 7 2016

Labels: -Needs-Bisect
With respect to comment 6:

Rechecked again and have considered tab crash as bad behavior, got appropriate Good build behavior (i.e  just before the tab crash build),with same narrow bisect mentioned in comment #5
rkote@, thank you for the update.
Cc: ranjitkan@chromium.org
Labels: Needs-triage
Just to update, still able to reproduce the issue on chrome version 51.0.2705.0 on Windows 8.1. Attached screen shot for the same. Could some one please take a look into it.
Black patch.png
200 KB View Download
Cc: -junov@chromium.org
Labels: -Needs-triage
Owner: junov@chromium.org
Status: Assigned (was: Untriaged)
junov@, could you please look into this change: https://chromium.googlesource.com/chromium/src/+/6b9c2b5d5fa1451e1e5a11a627534f5d3859d77b if possible?

Many thanks!
Cc: rnimmagadda@chromium.org
Just to update:

Still able to repro this issue on Windows 7 for Google Chrome Dev Version - 51.0.2704.7.

Screen-recording is attached.

Note: Issue fixed on Google Chrome Canary Version - 52.0.2708.0 


600284.mp4
5.7 MB Download
Just to update:

Still able to repro this issue on Windows 7 for Google Chrome Dev Version - 51.0.2704.19
junov@, Gentle ping! Could you please take a look.

Comment 16 by junov@chromium.org, Apr 25 2016

I cannot reproduce this bug.  It could be specific to certain graphics configurations. Could you attach the contents of your "chrome://gpu" page?

Comment 17 by rk...@etouch.net, Apr 26 2016

with respect to comment 16:

Kindly refer the attached chrome://gpu page
Final_GPU.png
843 KB View Download

Comment 18 by junov@chromium.org, Apr 26 2016

rkote@: thanks this is helpful

Could you try starting chrome with the command line option --disable-d3d11
Then try to reproduce the bug.

Then, try again with the command line option --use-gl=desktop

Then, try again with the command line option --disable-accelerated-2d-canvas

Thanks.

Comment 19 by rk...@etouch.net, Apr 27 2016

With respect to command 18:
 
Issue is reproducible on --disable-d3d11.

Issue is not reproducible on --disable-accelerated-2d-canvas and --use-gl=desktop

Thank you.

Comment 20 by junov@chromium.org, Apr 27 2016

Components: Internals>GPU>ANGLE
Status: Available (was: Assigned)
Thanks that is very helpful. Sounds like this is probably a bug in ANGLE.

ANGLE team: help!
Labels: -Pri-1 Pri-2
Can't repro on an Intel machine with D3D9 or 11. rkote@etouch.net can you try both updating your drivers to the latest and with Canary?
gpu.html
96.4 KB View Download
Labels: Needs-Feedback
rkote@, Could you please recheck the issue as per comment #21

Comment 23 by rk...@etouch.net, May 3 2016

Labels: -Needs-Feedback
With respect to comment 21 and 22:

Tested on Intel Graphics card with available driver version 9.17.10.4229 and latest canary, above issue is reproducible on it.
Without being able to reproduce it's very hard to diagnose, or determine where the bug is.
I am able to reproduce the issue on my Windows 7 machine, below are the steps which I followed 

Steps followed :
1. Install the "PDF Viewer" extension from Chrome Web store  https://chrome.google.com/webstore/detail/pdf-viewer/oemmndcbldboiebfnladdacbdfmadadm?hl=en
2. Once Extension is installed, open a sample PDF and before the pdf renders open a NTP and from bookmark bar right click and select 'Show bookmarks bar' option(This should be fast).


Note : Without the extension I wasn't able to reproduce the issue if I disable or remove the extension, When extension was present I am able to reproduce almost 100% of the time.
forgot to mention that I checked this on latest Chrome Beta i.e.,51.0.2704.36 
A friendly reminder that M51 Stable is launching soon! Your bug is labelled as Stable ReleaseBlock, pls make sure to land the fix and get it merged into the release branch by May 17. All changes MUST be merged into the release branch by 5pm on May 20 to make into the desktop Stable final build cut. Thanks!
Labels: -ReleaseBlock-Stable
Been traveling and haven't had a chance to investigate, but I do not think this is severe enough to block stable.
Just to update.

Observed this issue on Windows 7 for Google Chrome Canary Version - 52.0.2738.0 

Thank you.
600284.mp4
1.3 MB Download
Cc: junov@chromium.org
Owner: jmad...@chromium.org
Status: Started (was: Available)
Looking.
I found a machine I can repro and bisect this on. Will investigate tomorrow.
Cc: kbr@chromium.org jmad...@chromium.org
Components: -Internals>GPU>ANGLE
Owner: junov@chromium.org
There are two regressions here.

The first regression is working -> aw snap.

Narrow bisect:

https://chromium.googlesource.com/chromium/src/+log/7da8bd05a6def3df489ab3f237ce510f219ced6a..f246b3224a5371fbca304c15ab663351a76bb836

The second regression is aw snap -> incorrect rendering.

Narrow bisect:

https://chromium.googlesource.com/chromium/src/+log/8471a0a9e42cde08421463086aa1219548e5451a..c3faa5d9e87d0e746ba2ccfdbe6d886221ca47d0

I believe the regression range is correct in the description. I think the likely cause of the second regression is r663be1f7fadbbd62eecfa8709df952c78e1638c8 because of the change from crash -> incorrect rendering. I'm not sure what caused the first regression.

Since there are no ANGLE rolls in either regression range I'm unassigning myself and removing the ANGLE label. Justin you might be familiar with this stuff, can you take a look at your CL with this sample? You may need an Intel machine to repro.
Also, this should repro on most any Windows machine with --use-angle=d3d9 (also --disable-d3d11)

Comment 34 by ajha@chromium.org, May 24 2016

Is there an update on this issue?

Note: Issue is still reproducible on the latest canary(53.0.2346.0) on Windows-7.
Friendly Ping! Request someone to update on the issue?

I really appreciate the help.

Thank you!
Status: Assigned (was: Started)
Just to update:

This issue is still seen on win8.1 chrome version canary #53.0.2762.0 
Status: Started (was: Assigned)
Status: Fixed (was: Started)
It appears this got fixed by this other patch that landed yesterday:
https://codereview.chromium.org/2051513002/

The fix is in 53.0.2763.0

Would like a second opinion: Can someone verify in the Canary channel?
Labels: Merge-TBD
[Auto-generated comment by a script] We noticed that this issue is targeted for M-51; it appears the fix may have landed after branch point, meaning a merge might be required. Please confirm if a merge is required here - if so add Merge-Request-51 label, otherwise remove Merge-TBD label. Thanks.
Labels: -Merge-TBD
[Bulk edit]

Our blockerbot script was offline; it was recently brought back online, and thus labeled many old issues (including this one) erroneously.  Removing Merge-TBD label since all milestones for this issue are already completed; no further work should be done.

Sign in to add a comment