Date: Mon, 9 Dec 2002 13:12:49 -0800 From: Marcel Moolenaar <marcel@xcllnt.net> To: yokase@lapis.plala.or.jp Cc: freebsd-ia64@FreeBSD.ORG Subject: Re: Console log (Re: 5.0-DP2 panics on i2000) Message-ID: <20021209131249.C43839@kayak.xcllnt.net> In-Reply-To: <20021209234907057.yokase@lapis.mail.plala.or.jp> References: <20021209234907057.yokase@lapis.mail.plala.or.jp>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Dec 09, 2002 at 11:49:21PM +0900, yokase@lapis.plala.or.jp wrote:
>
> Attachment to this mail
> is the records of console(redirection).
> Please check this.
On the one hand the log is promising. The machine appears to be the same
as the prototypes we have been using. The bad thing about the log is that
it tells us that the trap happens way too early. Even before we can see
what the chipset is or if the box is a multi-processor machine. Then again,
it does rule out a lot and the most likely cause seems to be ACPI.
See below...
> Entering /kernel at 0xe000000000512000...
> Copyright (c) 1992-2002 The FreeBSD Project.
> Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
> The Regents of the University of California. All rights reserved.
> FreeBSD 5.0-DP2 #0: Wed Nov 20 00:39:41 GMT 2002
> root@itanium.pn.xcllnt.net:/usr/obj/usr/src/sys/BOOTMFS
> Preloaded elf kernel "/kernel" at 0xe000000001464000.
> Preloaded mfs_root "/mfsroot" at 0xe0000000014640b0.
> CPU: Itanium (733.47-Mhz)
> Origin = "GenuineIntel" Model = 0 Revision = 6
> Features = 0x0
> real memory = 1056210944 (1007 MB)
> avail memory = 1010417664 (963 MB)
> FPSWA Revision = 0x10006, Entry = 0xe00000003f13a050
>
> fatal kernel trap (cpu 0):
>
> trap vector = 0x4 (Alternate Data TLB)
> cr.iip = 0xe00000003fef00f0
> cr.ipsr = 0x101008022010 (mfl,ic,dt,rt,cpl=0,it,ri=0,bn)
> cr.isr = 0x400000000 (code=0,vector=0,r,ei=0)
> cr.ifa = 0x0
> cr.iim = 0x6
> curthread = 0xe000000000bb7b68
> pid = 0, comm =
My guess is that ACPI is the problem. Normally after FPSWA we get
something like:
Table 'FACP' at 0xe00000007ffd8070
Table 'APIC' at 0xe00000007ffda000
Local APIC entry
:
It looks to me that we're expecting that certain pointers are valid
and end up dereferencing NULL pointers.
There are a couple of things we can try (if you like, of course :-)
1. Download 5.0-RC1. We released it this weekend and it should gradually
appear on FreeBSD mirrors. It would be good to know if the problem
still exists in RC1 so that we can try to fix it before we official
release 5.0 (scheduled Januari 10). If I remember correctly, we have
a new ACPI code drop that may fix this...
2. Boot Red Hat or Windows XP and try to get as detailed a dump of
ACPI as possible. I'm not sure either Red Hat or Windows has any
tools to dump the ACPI tables, but it would definitely help.
In the mean time, I'll try to get my hands on the i2000 specs. If all
else fails, I can install Red Hat on my machine and create an ACPI
dumping tool that you can then use under Red Hat.
Thank you so far. It's much appreciated!
--
Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ia64" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021209131249.C43839>