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

Issue 744133 link

Starred by 2 users

Issue metadata

Status: Untriaged
Owner: ----
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: 2
Type: Bug



Sign in to add a comment

page update stalled

Reported by pdk...@gmail.com, Jul 16 2017

Issue description

UserAgent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.68 Safari/537.36

Example URL:

Steps to reproduce the problem:
This problem occurs relatively frequently. Navigating to a new page still show the old page for several seconds, even though the new page was fully loaded. Switching tabs back and forth shows the title of the new page in the tab.

It's noteworthy that the cursor usually appears to react correctly, in that if there's a link at its position on the (not rendered) new page, it changes accordingly.

It does not only happen on navigation, but also seemingly randomly.

What is the expected behavior?

What went wrong?
^

Does it occur on multiple sites: Yes

Is it a problem with a plugin? No 

Did this work before? Yes M57+

Does this work in other browsers? Yes

Chrome version: 60.0.3112.68  Channel: beta
OS Version: Ubuntu 14.04
Flash Version:
 
Labels: Needs-Triage-M60 Needs-Bisect
Cc: krajshree@chromium.org
Labels: Needs-Feedback
Unable to reproduce the issue on Windows 10 and Ubuntu 14.04 using chrome beta version #60.0.3112.66 and latest dev #61.0.3153.4.

Attached a screen cast for reference.

Following are the steps followed to reproduce the issue.
------------
1. Navigated to www.wikipedia.org
2. Navigated to another page by clicking on the link in that page. Repeated the step 3-4 times.
3. Observed that the new page got updated instantly and did not show the old page.

pdknsk@ - Could you please provide a sample url to test the issue from TE-end. This will help us in triaging the issue further.

Thanks...!!
		
744133.ogv
21.6 MB Download

Comment 3 by pdk...@gmail.com, Jul 19 2017

​It won't reproduce that easily.​
Project Member

Comment 4 by sheriffbot@chromium.org, Jul 19 2017

Labels: -Needs-Feedback
Thank you for providing more feedback. Adding requester "krajshree@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Comment 5 by pdk...@gmail.com, Jul 19 2017

chrome://gpu

GPU0	VENDOR = 0x8086, DEVICE= 0x0a16
Optimus	false
Optimus	false
AMD switchable	false
Driver vendor	Mesa
Driver version	11.0.2
Driver date	
Pixel shader version	3.30
Vertex shader version	3.30
Max. MSAA samples	8
Machine model name	
Machine model version	
GL_VENDOR	Intel Open Source Technology Center
GL_RENDERER	Mesa DRI Intel(R) Haswell Mobile
GL_VERSION	3.3 (Core Profile) Mesa 11.0.2
Components: -Blink Blink>Loader
Labels: Needs-Feedback
Are you able to provide a trace? (chrome://tracing) of the issue occurring? It might help reveal some useful data.

Comment 7 by pdk...@gmail.com, Jul 19 2017

​Which categories should I record?
Project Member

Comment 8 by sheriffbot@chromium.org, Jul 19 2017

Cc: dtapu...@chromium.org
Labels: -Needs-Feedback
Thank you for providing more feedback. Adding requester "dtapuska@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Comment 9 Deleted

Labels: -Needs-Bisect
As per comment #3, it seems that the issue doesn't reproduce consistently.
Hence, removing the Needs-Bisect label as of now.
Please feel free to add the same if the issue reproduces consistently.

Thanks...!!
dtapuska@ - Could you please respond to comment #7.

Thanks...!!
I'd start with a Javascript and Rendering trace.

But an all categories trace is usually best.
Labels: Needs-Feedback

Comment 14 by pdk...@gmail.com, Aug 10 2017

​If I interpret the trace correctly, it blocks on
LayerTreeHostImpl::SetVisible.​
Project Member

Comment 15 by sheriffbot@chromium.org, Aug 10 2017

Cc: yhirano@chromium.org
Labels: -Needs-Feedback
Thank you for providing more feedback. Adding requester "yhirano@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
Components: UI>Browser>Navigation
Labels: Needs-Feedback
I would like to ask for the trace log file, if possible.

Is the issue specific to certain sites?

Comment 17 by pdk...@gmail.com, Aug 17 2017

It happens totally intermittently, on any page. Sometimes within minutes of browsing, and sometimes not. Otherwise I'd have already bisected it.

I recorded a new trace on a fresh profile, and fortunately the problem occurred quickly. I noticed that it often (but not only) happens when opening a new tab. 

And on new incognito tabs it happens in two steps: first the dark background is rendered, including the small bullet points next to the text, but not the text itself. Now it stalls and blocks further rendering in the whole browser. When it resumes, usually within a few seconds to a minute, the text is rendered, as well as the incognito logo. In particular that it renders the bullet points is odd, I think.

I stopped recording a few seconds after rendering resumed. So the tab affected is the last tab opened.


trace_bug.json.gz
4.6 MB Download
Project Member

Comment 18 by sheriffbot@chromium.org, Aug 17 2017

Labels: -Needs-Feedback
Thank you for providing more feedback. Adding requester "yhirano@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Comment 19 by pdk...@gmail.com, Aug 17 2017

I tried to record it, but it didn't occur. I only managed to record a micro-stall on the incognito NTP.

Comment 20 by pdk...@gmail.com, Aug 17 2017

bug.webm
139 KB View Download
Labels: M-64
Status: Untriaged (was: Unconfirmed)
As per comment #17 and #20, the issue seems to repro on incognito window using latest stable #62.0.3202.75 and latest dev #64.0.3253.3. It doesn't occur on M59 as the incognito window doesn't have the bullet points.
It occurs inconsistently and doesn't occur on every page.
Hence, marking it as untriaged from M60, the build when the new incognito window with bullet points was introduced.
Requesting someone from dev team to investigate it further.

Thanks...!!

Comment 22 by pdk...@gmail.com, Nov 18 2017

Still in M63, but far less frequently.
Components: -Blink>Loader Blink>Paint
Labels: Needs-Feedback
Looks like this hadn't been re-triaged for a while.  Also from #17 it doesn't seem it's loading issue but maybe Paint or Compositing? Adding Compositing label to get help for further triage.

Reporter: would you be able to update the issue again if you are still seeing this in the latest release? Thanks for your input!!
Components: -Blink>Paint Internals>Compositing
To compositing based on comment #14, and the fact that painting does not control presentation of pages outside of invalidation (which does not seem implicated here).
Still have this in 68.0.3440.124.
Also in M70.

Sign in to add a comment