HighSierra: At times 'gmail.com' is not loading. |
|||||||||||||||||
Issue descriptionChrome: 61.0.3150.0 Canary OS X: Mac OS X 10.13 Beta (17A291m) 1. Launch Chrome browser 2. Try to access gmail.com by clicking 'Enter' at the end We are seeing blank page instead of gmail login screen. This is not consistently reproducible, however i do see this issue couple of times a day. Attaching the screen-cast for your reference. Thank you!
,
Jul 10 2017
It's not a VM and i've local machine with the set-up!
,
Jul 10 2017
Thank you. I upgraded my Touch Bar Mac to 17A291 but cannot reproduce. ccameron@ - any thoughts?
,
Jul 13 2017
ccameron@ - gentle ping. What are your thoughts on what's happening here.
,
Jul 13 2017
I don't see anything suspicious off the top of my head. That said, it looks like you're opening a gmail authentication window, which sometimes causes an iGPU->dGPU switch (shouldn't, but does) -- does this reproduce if you force usage of the iGPU via gfxCardStatus? Try also with the flags individually. --disable-remote-core-animation --disable-gpu
,
Jul 13 2017
Oh and one more set of flags --disable-mac-overlays
,
Jul 13 2017
,
Jul 18 2017
Reg c#5: Unable to reproduce this issue with above specified flags, however i experienced the similar behavior while browsing Canary#61.0.3159.0 today. Here is the attached video: https://drive.google.com/a/google.com/file/d/0B-vMgxWkGDE2TEo1ZVY0TmU0S0k/view?usp=sharing Thank you!
,
Jul 19 2017
> Reg c#5: Unable to reproduce this issue with above specified flags, however i experienced the similar behavior while browsing Canary#61.0.3159.0 today. Do you mean you experienced the issue in 61.0.3159.0 while running with the flags?
,
Jul 19 2017
Do you mean you experienced the issue in 61.0.3159.0 while running with the flags? Nope, i was seeing this issue without any flags.
,
Jul 21 2017
[mac triage] this does seem like ccameron's bailiwick - some kind of GPU interaction. Possibly something Apple need to iron out. But we can't do much without a reliable repro. I couldn't repro it. But if we possibly *can* repro with any of --disable-remote-core-animation or --disable-gpu or --disable-mac-overlays individually, then that could rule out the above and suggest a different issue.
,
Jul 21 2017
It looks like manoranjanr@ is in MTV. Perhaps ccameron@ you can sync up with manoranjanr@ when you're in MTV next week?
,
Jul 21 2017
There's a similar issue filed in issue 745872 . If we don't have a reliable repro for this, it will be nearly impossible to make progress on. I haven't changed anything in compositing in almost a year at this point.
,
Jul 21 2017
It sound like manoranjanr@ can repro it reliably.
,
Jul 24 2017
I'm not going to be in MTV for the next >1 month -- maybe someone in MTV can take a look?
,
Jul 28 2017
This also sounds similar to issue 749438 , which has managed to be reproduced.
,
Jul 28 2017
hmm.. one more scenario attached here. The strange thing here is repro is not consistent and based on my 2 cents it happens on certain newly opened websites which never have been browsed on this machine (preferably after OS upgrade?). Thank you!
,
Jul 28 2017
ccameron@, would it be possible for sharing latest chromium build to get the exact trace w/ '--show-mac-overlay-borders' flag? I am seeing some close similarity with https://bugs.chromium.org/p/chromium/issues/detail?id=583805 though not 100% sure. Thank you!
,
Jul 31 2017
Does the diagnosis in https://bugs.chromium.org/p/chromium/issues/detail?id=749438#c12 fit for this bug too? You can run Canary with --show-mac-overlay-borders at the command line -- it shouldn't need a Chromium build.
,
Jul 31 2017
Re c#19. I think yes! Since i've used the same laptop for reproducing both these issues.
,
Aug 3 2017
|
|||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||
Comment 1 by shrike@chromium.org
, Jul 10 2017