On Wed, 2007-07-11 at 21:21 +0200, Anders Johansson wrote:
> On Wednesday 11 July 2007 21:09:37 Dave Howorth wrote:
> > I pressed various buttons in the checks dialogs until I got to a point
> > where I had a checkbox that said something like 'ignore this conflict
> > just here', then told YaST to go ahead. Now it says I have gtk2
> > 2.10.6-24.2 (as before) and I have gtk2-debuginfo 2.10.6-13. BUT, when I
> > go into the Online Update module, there are no patches for gtk modules.
> > The closest I can see is for wxGTK.
> >
> > What am I missing?
>
> Good question, I'm not sure at the moment.
Well at least I don't feel such an idiot then! Let's hope somebody else
can tell us.
> But I still think a simple strace will tell you where it's hanging.
An strace shows that it's hanging on a futex:
write(6, "\0", 1) = 1
write(6, "AUTH EXTERNAL 30\r\n", 18) = 18
poll([{fd=6, events=POLLIN, revents=POLLIN}], 1, -1) = 1
read(6, "OK edc29446c3e132dcf7020f13cb28b"..., 2048) = 37
poll([{fd=6, events=POLLOUT, revents=POLLOUT}], 1, -1) = 1
write(6, "BEGIN\r\n", 7) = 7
poll([{fd=6, events=POLLIN|POLLOUT, revents=POLLIN|POLLOUT|POLLHUP}], 1,
-1) = 1
read(6, "", 2048) = 0
close(6) = 0
futex(0x6707d0, FUTEX_WAIT, 2, NULL
But we already knew that from the backtrace from gdb. I suspect the
problem is finding which module is calling that and/or not setting up
conditions so it will terminate. It's a large multithreaded codebase, so
I don't think it's a trivial problem - that's why they're asking for the
symbolic trace.
And since I can't seem to get the source either, I can't look for
myself!
Frustrating.
Cheers, Dave
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
