Date: Mon, 29 Nov 1999 11:23:11 -0500 (EST) From: David Gilbert <dgilbert@velocet.ca> To: stable@freebsd.org Subject: curproc 0? Message-ID: <14402.43119.163568.35734@trooper.velocet.net>
index | next in thread | raw e-mail
I was going through the crash I just posted, and I was looking at the
tsleep funciton in kern_synch.c. line 382 is:
struct proc *p = curproc;
... now doing a 'p curproc' in gdb gave 0... so I'm wondering where
curproc comes from --- a static variable somewhere?
I've been experimenting with vinum here. I have 8 18Gig Atlas
IV's.... and I gather this has something to do with how busy I make
them.
If I 'dump -0af - /usr | team 1m 8 | restore rf -' ... then I get the
crash I just posted. If I remove team, then I don't get the crash.
Similarly, I was running bonnie processes in parallel --- to see how
the performance was affected by number of processes hitting the
drives. With 1, 2, 3, or 4 bonnie processes, everything ran fine.
With 8 bonnie processes, I got the crash.
I can repeat this at will and the system can be made available for
testing as required, too.
vinumlock.c calls tsleep in a function that's dealing with locks over
the range of stripes on the volume. Is the call to tsleep() improper
here because it can somehow occur when we are not within a process?
Dave.
--
============================================================================
|David Gilbert, Velocet Communications. | Two things can only be |
|Mail: dgilbert@velocet.net | equal if and only if they |
|http://www.velocet.net/~dgilbert | are precisely opposite. |
=========================================================GLO================
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
home |
help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14402.43119.163568.35734>