Bug 215634
| Summary: | evolution: crash at start | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Need Real Name <lsof> | ||||||||||||||
| Component: | evolution | Assignee: | Matthew Barnes <mbarnes> | ||||||||||||||
| Status: | CLOSED CURRENTRELEASE | QA Contact: | |||||||||||||||
| Severity: | medium | Docs Contact: | |||||||||||||||
| Priority: | medium | ||||||||||||||||
| Version: | 6 | CC: | mike | ||||||||||||||
| Target Milestone: | --- | ||||||||||||||||
| Target Release: | --- | ||||||||||||||||
| Hardware: | All | ||||||||||||||||
| OS: | Linux | ||||||||||||||||
| Whiteboard: | |||||||||||||||||
| Fixed In Version: | evolution-2.10.2-2.fc7 | Doc Type: | Bug Fix | ||||||||||||||
| Doc Text: | Story Points: | --- | |||||||||||||||
| Clone Of: | Environment: | ||||||||||||||||
| Last Closed: | 2007-06-17 21:46:10 UTC | Type: | --- | ||||||||||||||
| Regression: | --- | Mount Type: | --- | ||||||||||||||
| Documentation: | --- | CRM: | |||||||||||||||
| Verified Versions: | Category: | --- | |||||||||||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||||||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||||||||||
| Embargoed: | |||||||||||||||||
| Bug Depends On: | |||||||||||||||||
| Bug Blocks: | 150225 | ||||||||||||||||
| Attachments: |
|
||||||||||||||||
|
Description
Need Real Name
2006-11-14 22:57:13 UTC
Created attachment 141299 [details]
latest stack.
I can almost reproduce it - it seems to happen almost exclusively after I log
in for the first time (but not if I log out then in again).
I've installed evolution-data-server-debuginfo.i386 and
evolution-debuginfo.i386 as you can see, let me know if you want anything else
installed..
I still encounter this bug daily. Thanks for the bug report. Unfortunately the backtraces don't reveal the source of the crash. They both show all threads in the program in a normal wait mode. Have you been able to reproduce the crash more reliably since reporting this? (In reply to comment #3) > Have you been able to reproduce the crash more reliably since reporting this? More reliable than daily? No, sorry. Please can you look at my other bugs, the maintainers there seem to thin it's a library problem (bug 215551 and bug 215377). Particularly, they mention the top-secret bug 212424. Oh and bug 210748. Is there a status on this yet as to a fix or the fix getting included in the FC version? No, not yet. Will investigate this as time permits. A backtrace showing the location of the crash would likely speed the process along. Created attachment 148286 [details]
latest stack
Thanks. This backtrace shows the location of the crash. > Thanks. This backtrace shows the location of the crash.
Great. Any ETA for a fix?
(In reply to comment #10) > Great. Any ETA for a fix? I'm examining the relevant source code but can't guarantee anything until I have a way to reproduce the crash. The crash appears to be in Thread 6 of the backtrace in comment #8: #6 0x43765a7a in camel_certdb_nss_cert_get (certdb=0x81bae88, cert=0xb4743c10) at camel-tcp-stream-ssl.c:742 fingerprint = 0x82e9b88 "\200$4\b\200\004?Lh\004?L\001" path = 0x8302d38 "ÿÿÿÿ" ccert = (CamelCert *) 0x81bc328 st = {st_dev = 5634, __pad1 = 0, st_ino = 1822049, st_mode = 33152, st_nlink = 1, st_uid = 500, st_gid = 500, st_rdev = 0, __pad2 = 0, st_size = 827, st_blksize = 4096, st_blocks = 16, st_atim = { tv_sec = 1171747892, tv_nsec = 0}, st_mtim = {tv_sec = 1171743673, tv_nsec = 0}, st_ctim = {tv_sec = 1171743673, tv_nsec = 0}, __unused4 = 0, __unused5 = 0} nread = 827 n = <value optimized out> fd = 56 Both warnings originate near the point of the crash: (evolution:3032): camel-WARNING **: rawcert != derCer (evolution:3032): camel-WARNING **: cert size read truncated X\xf7r\xb4: 0 != 827 It looks like Camel was reading data from a certificate file. Camel being Camel, of course the read and interrupt handling logic was written by hand (probably incorrectly), as opposed to using the facilities provided by GLib. I notice the same warnings appear in all of the backtraces attached here. It may be that the certificate files are somehow corrupted, and the read logic is not handling the error condition properly. This is a shot in the dark, but you might try deleting your SSL certificates: $ rm -rf ~/.camel_certs You'll have to accept new certificates for each of your mail accounts that use SSL authentication, but it may correct the problem for you. In the meantime I'll look into making the certificate file read logic in Camel more robust. _Seems_ to be working so far.. Created attachment 152492 [details]
bug buddy
Nope. It's crashing again, but less frequently.
Created attachment 152923 [details]
latest crash
Looks like the problem wasn't solved then.
(does evo assume one ssl cert per account?)
This is *STILL* happenning. Evolution has crashed three times this morning on start. Created attachment 154203 [details]
crash while running
Upgraded to FC7T4 (fresh install actually).
Evolution sometimes crashes while running.
I am really not experiencing any crashes at all for the most part. Has anyone that is tried creating a brand new user, logged in and started evo from scratch to see if it still does it? And for some of those, are they on upgrades to 7t4 or new installs? Well mine isn't for a new user, but it is a fresh install. Not only did it crash eight times after starting the first time (something about a charset conversion problem with utf7), but it often hangs when not in the foreground. I've had enough. I've used Epiphany for 2 years since it because useless, so I'll switch to something that works. Created attachment 154425 [details]
Proposed patch
This patch should make reading cert files more reliable.
With the Fedora 7 release so close, I'm hesitant to put this in Rawhide until I
have confirmation that it works. Anyone care to test this?
Sure, I'll test it if you provide an rpm. Filed upstream with patch: http://bugzilla.gnome.org/show_bug.cgi?id=437331 I haven't been able to reproduce this issue so I can't say for sure whether the patch fixes it, but it at least seems to not break things so I'll go ahead and put it in Rawhide since I don't really have a means of distributing RPMs out-of-band. Please try evolution-data-server-1.10.1-6.fc7. Still running evolution-data-server-1.10.1-1.fc7 here, evolution-data-server-1.10.1-6.fc7 hasn't come through yet. You fixed it! Thanks a lot! evolution-2.10.2-2.fc7 Good! Closing this then. Feel free to reopen this bug if you encounter the problem again. Bug 248292 filed, likely related to this one. I doubt it's related to this. More likely related to the patch here: http://bugzilla.gnome.org/show_bug.cgi?id=376991 |