GNOME Bugzilla – Bug 789656
When I sign in in Preferences > Sync, the browser Segfaults.
Last modified: 2017-11-06 14:00:13 UTC
When I sign in in Preferences > Sync, the browser Segfaults. After I enter my password it momentarily says signin completed, and the poof... the browser crashes. Returning to Preferences > Sync, it then says "Sign in to continue to Firefox Sync" again.
Thanks for taking the time to report this. Without a stack trace from the crash it's very hard to determine what caused it. Can you get us a stack trace? Please see https://wiki.gnome.org/Community/GettingInTouch/Bugzilla/GettingTraces for more information on how to do so. When pasting a stack trace in this bug report, please reset the status of this bug report from NEEDINFO to its previous status. Thanks in advance!
Created attachment 363032 [details] trace
Happens every time on my i386 computer. Not on my amd64 computers.
From the stack trace:
+ Trace 238140
Thread 1 (Thread 0xaf9de980 (LWP 10398))
Since you're experiencing this only on 32-bit architecture, I bet it's related to the printing format for long unsigned in ephy_sync_crypto_create_assertion(). I think replacing %lu with PRIu64 should fix the problem.
Created attachment 363034 [details] [review] sync-crypto: Use PRIu64 macro to print unsigned longs Prevent failure for platforms where longs are not 8 bytes.
Maybe you can try the above patch and see whether it fixes the crash.
Thanks. I'll wait for a .deb . epiphany-browser: Installed: 3.26.1-1 Candidate: 3.26.1-1 Version table: *** 3.26.1-1 500 500 http://free.nchc.org.tw/debian unstable/main i386 Packages
We're not going to build a .deb for you to test this.
Review of attachment 363034 [details] [review]: Yes, this is correct and I'm confident it will fix the crash. Please push to gnome-3-26 as well as master. Did you look through the rest of the sync code to make sure you haven't made this mistake anywhere else? I don't see any other issues in ephy-sync-crypto.c, but I haven't checked other files.
(In reply to Michael Catanzaro from comment #9) > Did you look through the rest of the sync code to make sure you haven't made > this mistake anywhere else? I don't see any other issues in > ephy-sync-crypto.c, but I haven't checked other files. Yes, I've checked. This was the only one.
Attachment 363034 [details] pushed as 1bde93a - sync-crypto: Use PRIu64 macro to print unsigned longs