linux-mips
[Top] [All Lists]

Re: another 4kc machine check.

To: Ralf Baechle <ralf@linux-mips.org>
Subject: Re: another 4kc machine check.
From: Greg Weeks <greg.weeks@timesys.com>
Date: Fri, 08 Apr 2005 12:45:24 -0400
Cc: linux-mips@linux-mips.org
In-reply-to: <20050408161357.GB19166@linux-mips.org>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <42553E49.7080004@timesys.com> <4256991C.4020601@timesys.com> <20050408161357.GB19166@linux-mips.org>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mozilla Thunderbird 1.0 (X11/20041206)
Ralf Baechle wrote:

On Fri, Apr 08, 2005 at 10:45:48AM -0400, Greg Weeks wrote:

This is the malta branch with the tlb patch from Ralf applied. The LTP test:

Credits for this patch go to Maciej who actually went through the pain
of debugging this problem that did show it's ugly head long, long after
the affected revision has been obsoleted by newer revisions.

-bash-2.05b# shmem_test_06
This test does some mucking with shmat at fixed addresses that might not be correct for mips. I still wouldn't expect to see a machine check when it goes to free the shared memory areas. I'm going to try it on a 24K malta as soon as I get yamon on it and can get it booted.

I've been able to reproduce it on on two 4Kc boards of different revisions
so this seems to be something else.  It does not hit 5Kc and 24K.
Have you tried a 4kec? That's the only other mips board I have available.

Greg Weeks

<Prev in Thread] Current Thread [Next in Thread>