10.4.Fox crash on LinkedIn.com logging
Hi!
Under MacOsX Leopard (PowerBook 1.64GHz) few days ago 10.4.Fox begin crashing on LinkedIn.com logging. I cleared cache, cookies, I cleared 10.4.Fox rand all it's reminiscents from my PowerBook & reinstalled it, bit the problem persists.
Keyboard shortcuts
Generic
| ? | Show this help |
|---|---|
| ESC | Blurs the current field |
Comment Form
| r | Focus the comment reply box |
|---|---|
| ^ + ↩ | Submit the comment |
You can use Command ⌘ instead of Control ^ on Mac
1 Posted by Dombi on 02 Sep, 2020 05:27 PM
Correction:
I cleared 10.4.Fox and all it's reminiscents from my PowerBook & reinstalled it, but the problem persists.
Support Staff 2 Posted by Chris (chtrusch... on 04 Sep, 2020 07:42 PM
Dombi, please go to Help>Troubleshooting Information>Refresh TenFourFox… After the browser is 'refreshed', quit TenFourFox, then go to Macintosh HD>users>[your name]>Library>Caches in the Finder. Delete the Firefox folder. Restart TenFourFox.
3 Posted by kby+tenfourfox on 05 Sep, 2020 03:42 AM
I also had this crash upon just going to the LinkedIn home page (before logging in. I have not cleared cache in this particular case (will try that and the refresh later when I am ready to deal with another crash), however I almost never go there (but happened to today) and I'm sure I've cleared cache since the last time I was there, I have attached the crash report, where I tried it in safe mode to make sure it wasn't any of my extensions.
Support Staff 4 Posted by Chris (chtrusch... on 05 Sep, 2020 07:36 AM
I cannot reproduce this crash on the homepage linkedin.com (without signing in).
However after signing in the browser reliably crashes sooner or later. I see the same items as in the 10.5 Server crash log.
All with fresh profiles:
10.5 G4 TFF 26: crash
10.4 G3 TFF 26: crash
10.5 G4 TFF 26 ion & baseline disabled: crash
10.5 G4 TFF 24: crash
10.5 G4 TFF 45.9: no crash, but doesn't load anything after login
Cameron, do you have an idea, or should I try to find a regression window?
5 Posted by Dombi on 06 Sep, 2020 07:31 AM
Hi!
Thanks a lot for the siggestions. After completing all suggested steps is worst: it crached even at accessing the LinkedIn page!
I attached the crash report.
Support Staff 6 Posted by Chris (chtrusch... on 06 Sep, 2020 08:09 AM
Thank you, your crash report is consistent with what we're seeing. We do have a crash bug here, and contrary to my first assumption, 'refreshing' the browser is not the solution.
Support Staff 7 Posted by Cameron Kaiser on 08 Sep, 2020 01:26 AM
I can, with some effort, reproduce this in a debug build. However, I don't know what's going wrong except that it's in JavaScript and (as Chris has already determined) it is not a problem with the JIT. This has been added to the worklist as issue 621: https://github.com/classilla/tenfourfox/issues/621
Because I don't have a good understanding of this bug right now, there is a decent chance I will not have sufficient time to fix it for FPR27.
8 Posted by marianconstr on 08 Sep, 2020 04:18 PM
Hi!
Thanks for your effort with the problem. I am waiting
with patency (not just because I have nothing to do, but for the general
quality of 10.4.Fox).
Attila
Tue, 08 Sep 2020 01:26:50
+0000 időpontban Cameron Kaiser írta:
> // Please reply above this
line
Support Staff 9 Posted by Cameron Kaiser on 09 Sep, 2020 05:21 AM
Chris, if you're able to find a regression window, that would be helpful since I don't have any obvious clues as to what's going on. The fact 45.9 doesn't crash does allege this was in one of the FPR changes.
Support Staff 10 Posted by Chris (chtrusch... on 09 Sep, 2020 03:39 PM
Regression window:
FPR 21 no crash
FPR 22b1 crash
In all cases I tested:
'no crash' means: doesn't load anything after log-in except what's in the screenshot.
'crash' means: crashes more or less immediately after you click 'sign in' at
https://www.linkedin.com/login
Support Staff 11 Posted by Cameron Kaiser on 09 Sep, 2020 08:43 PM
Noted. So we must have either enabled some code that crashes, or the code we added wasn't complete. Thanks for doing that, that gives me a place to start. Still no guarantees on timeframe. If nothing else, the crash will be wallpapered (no crash, but no load, as before).