smbfs crashes since approx. 10.1-RELEASE

[Rick Macklem]

Christian Kratzer wrote:
> Hi Rick,
> 
> On Sun, 18 Oct 2015, Rick Macklem wrote:
> 
> > Christian Kratzer wrote:
> >> Hi Rick,
> >>
> >> looks like your latest patch nailed the issue. The box has been up for 3
> >> days:
> >>
> >>      ck at noc3:~ % uptime
> >>      12:22PM  up 3 days,  4:11, 1 user, load averages: 0.07, 0.10, 0.08
> >>      ck at noc3:~ %
> >>
> >> If it does not crash over the weekend this seems to be it:
> >>
> > When I took a closer look, it appears that PR 172942 was a different crash
> > and
> > it appears that one was fixed via r264600.
> >
> > Your problem does not appear to be in the bugs database. (I will commit the
> > patch in mid-November anyhow, but creating a PR for this might be useful
> > for
> > others.)
> >
> > Btw, I think the attached patch (which includes this change) also fixes a
> > problem that caused a crash during mounting, reported via PR 201912.
> > (If you`d like to test this one that would be appreciated. It should be
> > applied to code not already patched with the one below, since the below
> > patch is included in it.)
> >
> > Thanks for your help with this, rick
> <snipp/>
> 
> I'll put your patch on the VM in question. Btw. it has been up for 6 days now
> without a crash.
> 
> Before I do that I would like to see that it really addresses PR 201912.
> 
> Do you have any idea how I could provoke that one ?
> 
Not really, I'm afraid. The patch deals with the failure cases in smb_vc_create(),
which I think was what caused the crash, given the backtrace in the PR.

You can look at smb_vc_create() and see there is a bunch of "goto fail;" cases,
but I don't know how to specifically tickle any of these?

You could "fake it" by putting a "goto fail;" at line#428, just after
the "error = ENOMEM;". This will break smb_vc_create() big time, but I
think it will generate the same crash as PR 201912.

> Ideally I would like to do the stuff that forces the panic, then apply
> the patch and see that the system stays stable despite me doing the
> silly moves again.
> 
That would be nice, but so long as I know that the patch doesn't cause a
regression, I am comfortable committing it. (This refers to the other 2
changes in the patch. It seems clear that the fix in smbiod2.patch is ok
to commit.)

Thanks for all your help with this, rick

> Greetings
> Christian
> 
> --
> Christian Kratzer                   CK Software GmbH
> Email:   ck at cksoft.de               Wildberger Weg 24/2
> Phone:   +49 7032 893 997 - 0       D-71126 Gaeufelden
> Fax:     +49 7032 893 997 - 9       HRB 245288, Amtsgericht Stuttgart
> Mobile:  +49 171 1947 843           Geschaeftsfuehrer: Christian Kratzer
> Web:     http://www.cksoft.de/
> _______________________________________________
> freebsd-stable at freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"
>