Re: Filesystem tests crashing host

[Mindaugas Rasiukevicius <rmind%netbsd.org@localhost>]

Andreas Gustafsson <gson%gson.org@localhost> wrote:
> Jukka Ruohonen wrote:
> > On Thu, Apr 14, 2011 at 06:27:19AM -0700, Paul Goyette wrote:
> > > The fault is coming from the following code at lines 817-821 of 
> > > src/sys/kern/kern_descrip.c (rev. 1.212, in which christos@ touched
> > > the close-on-exec stuff)
> > > 
> > >   if (fp->f_ops != NULL) {
> > >           error = (*fp->f_ops->fo_close)(fp);
> > >   } else {
> > >           error = 0;
> > >   }
> > 
> > Now that this is diagnosed, any idea how to fix this regression that
> > broke the whole test infrastructure?
> 
> I still say revert rmind's changes of 2011.04.11.22.31.43, because
> that's when the failures started.  My logs show six test runs between
> christos' change to kern_descrip.c (at 2011.04.10.15.45.33) and rmind's
> changes, and none of those test runs paniced; after rmind's changes,
> every single test run has paniced.

Problem is not diagnosed.  It cannot be reproduced on real hardware,
and I do not see how f_ops can become invalid when using semaphore.
Even if we assume that it can - the semaphore code should actually be
*used* in the first place.  However, it seems that neither failing
ATF tests, nor ATF itself are using semaphores.  Can somebody prove
me wrong on this?

Perhaps a simple printf("f_type = %d\n", fp->f_type) would hint what
type of descriptor is actually failing.  Also, a wild guess - can one
reproduce the problem with the following changes reverted:

http://mail-index.netbsd.org/source-changes/2011/04/10/msg020849.html

-- 
Mindaugas