Panic on 6.0 building ports

[Joe Marcus Clarke]

On Mon, 2004-08-30 at 14:35, Doug White wrote:
> On Sun, 29 Aug 2004, Joe Marcus Clarke wrote:
> 
> > On my Tinderbox machine (similar in operation to pointyhat), I was
> > building packages for 4.10-RELEASE and 5.2.1-RELEASE, and the machine
> > panicked.  This is a single-CPU Pentium 4 with 2 GB of RAM.  The drive
> > being used for builds is a Maxtor SATA drive connected to a Promise SATA
> > controller.
> 
> I'm guessing you didn't get a crashdump.  Can you try loading up your
> kernel.debug with kgdb and get line numbers for the first few lines in the
> trace?  l *vfs_vmio_release+0x1b should work.

Sorry, I thought I had included that.  The result was a little funny. 
It said that the address pointed to vfs_vmio_release in atomic.h line
154.

I tried to ``panic'' the system to get the core dump (I have dumps
configured), but it locked up, and not even a ``continue'' would work. 
Since then, I'm on a slightly newer kernel, and so far, no crash.  If it
dies again, I'll send more details, and hopefully get a core dump. 
Thanks for the follow up.

Joe

> 
> >
> > Here is the manually transcribed panic and trace:
> >
> > Fatal trap 12: page fault while in kernel mode
> > cpuid = 0; apic id = 00
> > fault virtual address	= 0x1c
> > fault code		= supervisor write, page not present
> > instruction pointer	= 0x8:0xc0533d07
> > stack pointer		= 0x10:0xf5f30a4c
> > frame pointer		= 0x10:0xf5f30a58
> > code segment		= base 0x0, limit 0xfffff, type 0x1b
> > 			= DPL 0, pres 1, def32 1, gran 1
> > processor eflags	= interrupt enabled, resume, IOPL = 0
> > current process		= 27441 (cpp0)
> > Stopped at  vfs_vmio_release+0x1b: lock cmpxchgl %ecx,0x1c(%edx)
> >
> > vfs_vmio_release(dc1fee68) at vfs_vmio_release+0x1b
> > getnewbuf(0,0,4000,4000,1c000) at getnewbuf+0x2b6
> > getblk(c8103210,7,0,4000,0) at getblk+0x400
> > cluster_read(c8103210,7,0,4000,0) at cluster_read+0xde
> > ffs_read(f5f30c14) at ffs_read+0x25e
> > vm_read(c3834e58,f5f30c88,c484f880,0,c5877160) at vn_read+0x178
> > dofileread(c5877160,c3834e58,7,814f000,2f2ed) at dofileread+0x95
> > read(c5877160,f5f30d14,3,4,293) at read+0x3b
> > syscall(2f,2f,2f,814f000,2f2ed) at syscall+0x287
> > Xint0x80_syscall() at Xint0x80_syscall+0x1f
> > --- syscall (3, FreeBSD ELF32, read) eip = 0x8064924, esp = 0xbfbfd394, ebp = 0xbfbfd3c0 ---
> >
> > The machine is running:
> >
> > FreeBSD fugu.marcuscom.com 6.0-CURRENT FreeBSD 6.0-CURRENT #0: Sun Aug
> > 29 01:02:18 EDT 2004
> > root at new-fugu.marcuscom.com:/space2/obj/usr/src/sys/FUGU  i386
> >
> > Joe
> >
> >
-- 
PGP Key : http://www.marcuscom.com/pgp.asc


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: This is a digitally signed message part
Url : http://lists.freebsd.org/pipermail/freebsd-current/attachments/20040830/c987facb/attachment.bin