Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 4 Mar 1998 18:09:15 -0800 (PST)
From:      Manfred Antar <root@mantar.slip.netcom.com>
To:        Tor Egge <Tor.Egge@idi.ntnu.no>
Cc:        smp@FreeBSD.ORG
Subject:   Re: Current SMP kernel panics on booting
Message-ID:  <Pine.BSF.3.96.980304180639.1308A-100000@mantar.slip.netcom.com>
In-Reply-To: <199803041544.QAA24636@pat.idi.ntnu.no>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 4 Mar 1998, Tor Egge wrote:

> > There were some smp file changes today and i built a new
> > a kernel. It drops into debugger when booting at the place where it used
> > to say CPU1 launched
> > here is what i get:
> > 
> > mp_lock=00000002; cpuid=0; lapic.id=00000000
> > instruction pointer = 0x8 : 0x0f012f89e
> > stack pointer = 0x10 : 0xf02a9f90
> > frame pointer = 0x10 : 0xf02a9f94
> > code segment = base 0x0 limit 0fffff, type 0x1b
> > 	     = DPL0, pres 1, def321, gran1
> > processor flags = interrupt enabled, IOPL = 0
> > current process = 0 (swapper)
> > interrupt mask = <-- SMP:XXX
> > kernel : type 29 trap, code = 0
> > stopped at _mbinit + 0xe movl $0,_mclfree
> 
> This probably means that the CPU got an interrupt for which it had no
> reasonable interrupt handler.  All vectors initially points to rsvd,
> which gives a T_RESERVED (29) trap.
> 
> This looks like the result of an incomplete kernel rebuild.  The
> config program no longer removes the old kernel compile directory by
> default.
> 
> I suggest a complete kernel rebuild, i.e. 
> 
> 	make clean && make depend && make
> 
> 
> The message shown by the kernel is not very informative.  Vectors
> 32..254 should probably initially point to a different handler.  On
> systems where APIC_IO is defined, that handler could examine the ISR
> register in the local apic.  A more informative message could be then
> be printed (e.g.  unexpected interrupt (vector 0x50) recieved.) in the
> panic message.
> 
> - Tor Egge
> 

I tried that and it didn't work. I'm not too sharp on using the debugger
I wish I could be of more help.
	I just did a fresh make world and suped the new smptest.h file
no problems now
Thanks Manfred


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.980304180639.1308A-100000>