Re: TSTILE, pmap, and bad analog numbers - progress?

[christos%astron.com@localhost (Christos Zoulas)]

In article <p06240800ce0a7c9665b9@[71.39.101.51]>,
Donald Lee  <MacPPC2%c.icompute.com@localhost> wrote:
>>> BTW - I tried to attach to httpd with gdb, and gdb locked up, too.
>>
>>gdb locked up as in tstiled unkillably, or locked up as in just wedged
>>and could be killed?  The latter would be unsurprising, the
>>former...well, I'd have to read the relevant code, but it would
>>surprise me a bit.
>>
>
>This was a production machine, so I didn't have the chance to play with it.
>I did a "/etc/rc.d/apache restart" and everything "got better".  The gdb
>"woke up" and it's still running, so I think the answer to your question is
>that it was "just wedged".

Looks like:

- one of the httpd threads in tstile tries to exit, but is stuck
- one is doing select
- one is doing nfsrcv

Are you running nfs on the machine? Can you make httpd not interact with
nfs and see if it still gets stuck?

christos