https://bugzilla.novell.com/show_bug.cgi?id=382172
User hpj@novell.com added comment
https://bugzilla.novell.com/show_bug.cgi?id=382172#c19
--- Comment #19 from Hans Petter Jansson 2008-05-11 20:27:36 MST ---
I'm seeing a similar thing re. gdmsetup. However, I couldn't get it to work
consistently between reboots until I mucked about some more (not sure what
exactly triggered it; possibly growing history files offsetting some race).
I saved snapshots of the home dir before and after it "started working", and
diffed those, but didn't find much interesting. Also tried to remove
~/.nautilus and various other files, which didn't make the problem come back.
The D-Bus and GVFS process list appears the same when it works and when it
doesn't, except that when it _is_ working, gvfsd-trash's VM is consistently
twice as large (20MB vs 10MB, no significant change in RSS) - probably a thread
gets spawned when the monitoring works.
Next I tried to gdb gvfsd-trash since it's possible that the monitor code is
blocking on a pipe or something (it's bizarre that starting gdmsetup makes it
start working, so maybe there's a D-Bus pipe that gets a push?).
Couldn't gdb anything today however, since I can't get all the debuginfo I need
due to bug 387148 coming back again.
Also looked at gdmsetup code, and there's no obvious interaction with GVFS or
even D-Bus there. Maybe it writes to a file that makes the monitoring code wake
up and also perform other queued tasks.
Took a brief look at gvfsd-trash code and couldn't find anything wrong there -
but it's a lot of code and I only looked for about 15 minutes or so. I don't
even know what I'm looking for yet.
Will wait for debuginfo availability and try the gdb route.
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.