Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 17 Jun 2000 19:51:32 -0700 (PDT)
From:      John Baldwin <jhb@FreeBSD.ORG>
To:        Josh Paetzel <jpaetzel@hutchtel.net>
Cc:        FreeBSD-questions@FreeBSD.ORG, Eric Kozowski <eric@svjava.com>
Subject:   RE: 4.0-R install problem
Message-ID:  <200006180251.TAA04358@john.baldwin.cx>
In-Reply-To: <NEBBIJCLELPGBFNNJOFHIEHGCCAA.jpaetzel@hutchtel.net>

next in thread | previous in thread | raw e-mail | index | archive | help

On 17-Jun-00 Josh Paetzel wrote:
>> i installed 4.0-R via boot floppies and ftp on the following system:
>>
>> - tyan trinity
>> - 256m ram (single sdram dimm)
>> - celeron 533 (no over clock)
>> - buslogic bt-948 scsi controller
>> - netgear fa-310 tx ethernet
>> - diamond fire gl 1000 agp video
>> - ide disabled in bios
>> - serial 2 disabled
>>
>> install went just fine.  rebooted after install and got the following:
>>
>> /boot.config: -P
>> Keyboard: yes
>>
>> BTX loader 1.00  BTX version is 1.01
>> Console: internal video/keyboard
>> BIOS drive A: is disk 0
>> BIOS drive C: is disk 1
>> BIOS 639kN/261129kB available memory
>>
>> FreeBSD/i386 bootstrap loader, Revision 0.7
>> (root@monster.cdrom.com, Mon Mar 20 21:05:31 GMT 2000)
>>
>> int=0000000d  err=0000579c  efl=00010207  eip=0000b444
>> eax=0002da9c  ebx=00002710  ecx=00000000  edx=0002dafc
>> esi=00000000  edi=00000000  ebp=00094a48  esp=00094a0c
>> cs=002b  ds=0033  es=0033    fs=0033  gs=0033  ss=0033
>> cs:eip=1f f8 98 5f c0 78 b5 43-87 7a 3e bf 79 57 96 89
>> ss:esp=9f 57 00 00 64 37 02 00-00 18 00 00 48 4a 09 00
>> System halted
>>
>>
>> a poke of the reset button generates the exact same results.
>>
>> anyone know what might be causing this?

Well..  it could be bad hardware (memory or CPU, have you tried
others?) since it doesn't happen on most systems.  FWIW, the
%cs:%eip indicates that it is running code in the actual
loader itself, and it's on the following instruction stream:

00000000  1F                pop ds
00000001  F8                clc
00000002  98                cwde
00000003  5F                pop edi
00000004  C078B543          sar byte [eax-0x4b],0x43
00000008  797A              jns 0x84
0000000A  3EBF79579689      ds mov edi,0x89965779

The error is that it attempts to pop the value 0x579f into %ds,
which isn't a valid selector under BTX and causes a General
Protection Fault.  However, this code looks rather bogus anyways,
as 'cwde' is somewhat fishy, the 'sar' instruction which shifts an
8-bit byte to the right 0x43 times is very fishy, and the 'mov edi'
instruction with an extraneous 'ds' segment override prefix is
very fishy.  It could be a problem where the stack is getting
corrupted somehow, causing the loader to return into lala land and
start executing randomly until it hits a fault.  This could either
be a hardware problem or a nasty, subtle bug.  Please try out
different memory and/or CPU to see if it fixes it.

-- 

John Baldwin <jhb@FreeBSD.org> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.cslab.vt.edu/~jobaldwi/pgpkey.asc
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200006180251.TAA04358>