Bug 7100 - Can't start Winbindd process => crash
Summary: Can't start Winbindd process => crash
Status: RESOLVED FIXED
Alias: None
Product: Samba 3.5
Classification: Unclassified
Component: Winbind (show other bugs)
Version: 3.5.0rc1
Hardware: Other Linux
: P3 normal
Target Milestone: ---
Assignee: Michael Adam
QA Contact: Samba QA Contact
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-02-04 15:31 UTC by Matthieu Patou
Modified: 2010-02-09 10:26 UTC (History)
0 users

See Also:

Attachments
configuration file (1.25 KB, application/octet-stream)
2010-02-04 15:34 UTC, Matthieu Patou 		no flags Details
Log of windbindd -d 10 (43.27 KB, application/octet-stream)
2010-02-04 15:36 UTC, Matthieu Patou 		no flags Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Matthieu Patou 2010-02-04 15:31:07 UTC 
Starting the process with ./bin/winbindd -d xx
See attached files for more info.
Comment 1 Matthieu Patou 2010-02-04 15:34:54 UTC 
Created attachment 5274 [details]
configuration file
Comment 2 Matthieu Patou 2010-02-04 15:36:39 UTC 
Created attachment 5275 [details]
Log of windbindd -d 10
Comment 3 Matthieu Patou 2010-02-04 15:38:40 UTC 
Note: the log is done with tdb backend for idmap.
Ie. using

idmap uid = 50000-60000
idmap gid = 50000-60000

instead of 

idmap config SAMBA4: default = yes
idmap config SAMBA4: backend = rid
idmap config SAMBA4: range = 10000-20000

Both idmap configuaration produce the crash
Comment 4 Jeremy Allison 2010-02-04 16:50:14 UTC 
Can you test this with either RC2, or with the latest code pulled out of the 3-5-stable git tree ? There were some winbindd crash fixes that went in recently and I don't want to track this down if we've already fixed it.

Thanks,

Jeremy.
Comment 5 Matthieu Patou 2010-02-05 03:20:42 UTC 
It's most probably fixed in 026b23062eeee23dffeb627800b3cb6f55d89ad6 s3: Fix a winbind segfault in "trusted_domains"
Comment 6 Matthieu Patou 2010-02-09 10:26:05 UTC 
Jeremy,

it seems indeed to be fixed in current s3.5git.