svn commit: r285050 - in head: lib/libutil usr.sbin/pwd_mkdb

Don Lewis truckman at FreeBSD.org
Thu Oct 13 23:06:21 UTC 2016


On 13 Oct, Alan Somers wrote:
> On Thu, Jul 2, 2015 at 11:31 AM, Renato Botelho <garga at freebsd.org> wrote:
>> Author: garga (ports committer)
>> Date: Thu Jul  2 17:30:59 2015
>> New Revision: 285050
>> URL: https://svnweb.freebsd.org/changeset/base/285050
>>
>> Log:
>>   When passwd or group information is changed (by pw, vipw, chpass, ...)
>>   temporary file is created and then a rename() call move it to official file.
>>   This operation didn't have any check to make sure data was written to disk
>>   and if a power cycle happens system could end up with a 0 length passwd
>>   or group database.
>>
>>   There is a pfSense bug with more infor about it:
>>
>>   https://redmine.pfsense.org/issues/4523
>>
>>   The following changes were made to protect passwd and group operations:
>>
>>   * lib/libutil/gr_util.c:
>>    - Replace mkstemp() by mkostemp() with O_SYNC flag to create temp file
>>    - After rename(), fsync() call on directory for faster result
>>
>>   * lib/libutil/pw_util.c
>>    - Replace mkstemp() by mkostemp() with O_SYNC flag to create temp file
>>
>>   * usr.sbin/pwd_mkdb/pwd_mkdb.c
>>    - Added O_SYNC flag on dbopen() calls
>>    - After rename(), fsync() call on directory for faster result
>>
>>   * lib/libutil/pw_util.3
>>    - pw_lock() returns a file descriptor to master password file on success
>>
>>   Differential Revision:        https://reviews.freebsd.org/D2978
>>   Approved by:  bapt
>>   Sponsored by: Netgate
>>
>> Modified:
>>   head/lib/libutil/gr_util.c
>>   head/lib/libutil/pw_util.3
>>   head/lib/libutil/pw_util.c
>>   head/usr.sbin/pwd_mkdb/pwd_mkdb.c
> 
> This change is making certain pw operations very slow on ZFS root
> systems.  The problem is that when you open a file with O_SYNC, every
> single write(2) call turns into a zil_commit on ZFS, which is fairly
> expensive.  Did you consider using fsync(2) on the temporary files
> instead of opening them with O_SYNC?  I just tried that now, and I see
> a considerable speedup when running the tests in
> /usr/tests/usr.sbin/pw:
> 
> Using O_SYNC, as CURRENT does: 4 minutes 5.2 seconds
> No synchronous operations at all: 49.5 seconds
> Using fsync(2): 56.0 seconds

pwd_mkdb was fixed back in February with by switching to fsync() in
r295925.  It looks like libutil was not fixed, though.




More information about the svn-src-all mailing list