trackerd crashed with SIGSEGV in do_save_ondisk_email_message_for_imap

Bug #151637 reported by Anders Magnusson
48
Affects Status Importance Assigned to Milestone
tracker (Ubuntu)
Fix Released
Medium
Unassigned

Bug Description

Binary package hint: tracker

trackerd keeps crashing. The last message that I get before it crashes is:

saving email service 12 with uri "email://1138733807.10188.16@xxxx/INBOX/xxxx/xxxx/xxxx/xxxx;uid=14" and subject "häftpistol på rymmen !!" from "<email address hidden>"

GDB backtrace:

#0 do_save_ondisk_email_message_for_imap (db_con=0x8102d80, mail_msg=0x820c180) at tracker-email-evolution.c:2290
#1 0x08070929 in index_mail_messages_by_summary_file (db_con=0x8102d80, mail_type=MAIL_TYPE_IMAP, summary_file_path=<value optimized out>,
    load_meta_header=0x8071120 <load_summary_file_meta_header_for_imap>, load_mail=0x8072dd0 <load_mail_message_for_imap>,
    skip_mail=0x80725a0 <skip_mail_message_for_imap>, save_ondisk_mail=0x806f5d0 <save_ondisk_email_message_for_imap>) at tracker-email-evolution.c:1343
#2 0x08070e83 in evolution_index_file (db_con=0x8102d80, info=0x8633170) at tracker-email-evolution.c:464
#3 0x0806cee4 in tracker_email_index_file (db_con=0x8102d80, info=0x8633170, service=0x853a2e8 "EvolutionEmails") at tracker-email.c:115
#4 0x0805295a in process_files_thread () at trackerd.c:799
#5 0xb7d7a5af in ?? () from /usr/lib/libglib-2.0.so.0
#6 0x00000000 in ?? ()

I saw that there was a report for another crash in do_save_ondisk_email_message_for_imap (#130854), but that was supposed to be fixed in 0.6.1. I'm using version:
Tracker version 0.6.3 Copyright (c) 2005-2007 by Jamie McCracken (<email address hidden>)

Please let me know if you need anything else. I have the core dump if you'll need that.

Revision history for this message
Jamie McCracken (jamiemcc-blueyonder) wrote :

pls send core dump to <email address hidden>

Revision history for this message
Michal Pryc (migi) (michal-pryc) wrote :

I was trying to reproduce this bug on:
- Ubuntu 7.10 with Tracker version 0.6.3
- Solaris Nevada build 76 (x86) with Tracker version 0.6.3 and also never build from trunk-980

but I couldn't reproduce it.

Can you please write more about conditions on which you have hit this problem?
Also can you try to re-index e-mails and tell if you have still this problem?

Revision history for this message
Anders Magnusson (d98.ama) wrote :

I did a re-index (trackerd -v 2 -s 1 -R), but it still crashes on the same e-mail message as before.

I took a look at this specific e-mail (since it crashes on the same e-mail every time), and I noticed that it has 1117(!) recipients in the CC filed. Maybe that has something to do with it?

Other than that I couldn't find anything strange.

Revision history for this message
Jamie McCracken (jamiemcc-blueyonder) wrote :

Yes thats it!

we limit ours to 255 - I guess we need to make sure we dont exceed that limit

Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

Yes, a check looks like the way to go.

Changed in tracker:
importance: Undecided → Medium
status: New → Triaged
Revision history for this message
Jamie McCracken (jamiemcc-blueyonder) wrote :

this should be fixed in 0.6.4

Can someone confirm?

Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

All the duplicates are reported with 0.6.3.

I don't use Evolution, so I can't check this. I'll leave it open for a while and close it if we don't get any report with 0.6.4.

Changed in tracker:
status: Triaged → Fix Committed
Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

Closing as there has been no report about this in more than 3 weeks.

Changed in tracker:
status: Fix Committed → Fix Released
Revision history for this message
Ernst Persson (ernstp) wrote :

Got a crash in the same function today, probably dupe so now it happened again:
Bug #261537

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.