Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 28 Feb 2005 16:14:45 +0000 (GMT)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        "J.R. Oldroyd" <fbsd@opal.com>
Cc:        current@freebsd.org
Subject:   Re: Fatal trap 12 in kernload()
Message-ID:  <Pine.NEB.3.96L.1050228161401.5974I-100000@fledge.watson.org>
In-Reply-To: <20050228153331.GA847@linwhf.opal.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 28 Feb 2005, J.R. Oldroyd wrote:

> I am seeing similar problem, using -current from 2/22 with atamk3l.
> Running on hyperthreaded i386 here.
> <snip>
> I have encountered this crash about five or six times since 2/22,
> but I've been otherwise occupied and not able to look at it.  Crash
> is infrequent, as in system's OK for many hours, then just freezes.
> Need to press the button to get it back.
> 
> Config is generic with IPFIREWALL, IPDIVERT and IPV6FIREWALL added
> and debugging turned off.
> 
> Since I have no other info at this time, I just built and installed a
> kernel with debugging on, so that next time I may have more details. 
> Don't yet have any repeatable way to reproduce this. 

If you set "debug.mpsafevfs=0" in loader.conf and reboot, does it go away?

Robert N M Watson


> 
> Previous kernel, 1/31 without atamk3, was stable.
> 
> 	-jr
> 
> 
> 
> On Feb 25, 21:21, Kris Kennaway wrote:
> > On Fri, Feb 25, 2005 at 04:47:14PM -0800, Doug White wrote:
> > > On Thu, 24 Feb 2005, Kris Kennaway wrote:
> > > 
> > > > > fault virtual address   = 0x7562676b
> > > 
> > > I agree with Dan, this is bogus. kernload() is the offset from kernbase
> > > where the ELF headers get stuck. I suspect ddb is resolving it like it
> > > resolves end -- its beyond the beginning of the kernel so it picks the
> > > next best match, like end shows up beyond the end of the symbol table.
> > > (FYI end usually indicates calls into a KLD.)
> > > 
> > > > > current process         = 52613 (getty)
> > > > > Tracing pid 52613 tid 100360 td 0xd2d3a000
> > > > > kernload(cd533500,3,2000,d2d3a000,3) at 0x7562676b
> > > > > devfs_open(f8225a4c,c072025a,1e6,c07205ff,d235f134) at devfs_open+0x291
> > > 
> > > Can you get an addr2line on this devfs_open call?  It appears to have
> > > tried to open an incompletely initialized tty device.  Which one would be
> > > nice to know :-)
> > 
> > It was ttyd0.  phk couldn't figure out how this happened, so I'm
> > trying to recreate it.
> > 
> > Kris
> > 
> > > > > VOP_OPEN_APV(c07340a0,f8225a4c,3,c076d398,1) at VOP_OPEN_APV+0x9e
> > > > > vn_open_cred(f8225bbc,f8225cbc,860,cd33e180,1) at vn_open_cred+0x45b
> > > > > vn_open(f8225bbc,f8225cbc,860,1,d2d3a000) at vn_open+0x33
> > > > > kern_open(d2d3a000,804f860,0,3,804f860) at kern_open+0xca
> > > > > open(d2d3a000,f8225d14,3a6,c071c691,d2d3a000) at open+0x36
> > > > > syscall(2f,2f,2f,2,804f860) at syscall+0x2c4
> > > > > Xint0x80_syscall() at Xint0x80_syscall+0x1f
> > > > > --- syscall (5, FreeBSD ELF32, open), eip = 0x280ca2cb, esp = 0xbfbfedfc, ebp = 0xbfbfee28 ---
> > > > > db>
> > > >
> > > > Kris
> > > >
> > > 
> > > -- 
> > > Doug White                    |  FreeBSD: The Power to Serve
> > > dwhite@gumbysoft.com          |  www.FreeBSD.org
> > > _______________________________________________
> > > freebsd-current@freebsd.org mailing list
> > > http://lists.freebsd.org/mailman/listinfo/freebsd-current
> > > To unsubscribe, send any mail to "freebsd-current-unsubscribe@freebsd.org"
> > > 
> 
> 
> 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.NEB.3.96L.1050228161401.5974I-100000>