On Monday 28 April 2003 07:47 am, Kelledin wrote:
> On Monday 28 April 2003 07:36 am, Keith Owens wrote:
> > On Mon, 28 Apr 2003 07:27:09 -0500,
> >
> > Kelledin <kelledin+XFS@xxxxxxxxxxxxxxxxxxx> wrote:
> > >As I mentioned in my last post,
> > >if the deadlock gets hit, allocating or unlinking an inode
> > > on the same fs as the deadlocked process breaks out of the
> > > deadlock.
> >
> > Does a sync command break out as well? IOW, do you have to
> > change the hung filesystem or is it enough just to kick the
> > disk?
>
> Appending to a file doesn't help, nor does reading a file. I
> apparently have to either create or remove a file before the
> deadlock breaks. Haven't tried a sync though.
Well...FWIW I just downgraded to the official 1.2 release,
applied against 2.4.19 along with the ptrace patch. The same
problem crops up. Again, turning on XFS debugging makes it go
away.
I'll see what i can do about cobbling together a spare x86 box,
and see if I can reproduce it there. That's going to take a
while though.
--
Kelledin
"If a server crashes in a server farm and no one pings it, does
it still cost four figures to fix?"
|