Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 19 Jan 1998 12:09:45 +0100 (CET)
From:      Sebastian Strollo <seb@erix.ericsson.se>
To:        freebsd-scsi@FreeBSD.ORG, freebsd-hardware@FreeBSD.ORG
Subject:   Re: running 2.2.5-RELEASE on a ASUS P2L97-S
Message-ID:  <199801191109.MAA17451@scotch.du.etx.ericsson.se>

next in thread | raw e-mail | index | archive | help
Hi again

Earlier I wrote:
> I am having trouble with a newly purchased machine that I have put
> together. Below is the specs. At first the system would completely
> lock up (i.e. only respond to the Mighty Big Reset Button) when
> putting load on the SCSI disk.
>... [lengthy description of the machine and the problem omitted]

Thank you for all your suggestions! I tried a couple of new
combinations of memory chips and cards, but ended up replacing the
motherboard, and whaddayanow, now the machine runs like a charm!
Hardware errors like these are scary. ;-(

David Greenman <dg@root.com>, wrote:
> ...
> >  1 (one) 64MB SDRAM 100MHz module
> 
>    Probably a defective or incompatible SDRAM module. I've heard a lot of
> reports about problems with SDRAM (as well as having my own bad experiance).

and Steve Passe <smp@csn.net> wrote:
> 
> ASUS specifies the "low chip count" memory parts, your DIMMs may have too
> many chips on them to drive properly...  When I put 64MB DIMMs on my L97
> I had to pay more for the DIMMs with the high-density chips to avoid this.
> Before you ask, I don't know the max number, and couldn't find it in
> the manual.  I trusted my vendor to spec the right parts...

Which got me to suspect the 64MB module, which is one of these "double
height" thingies with a lot of chips on them. (A FOAF also suggested
that these were actually non standard, since the leads to the chips
got too long.) However I also had two 32MB SDRAM modules, and when I
used those the machine froze as well.

The Classiest Man Alive <ksmm@cybercom.net>, wrote:
> 
> Based on the limited information you provided, the only thing that I find
> particularly suspect is the IRQ conflict with the ethernet card and the
> SCSI controller.  Have you tried this test without the ethernet card
> present?

No, I hadn't at first. The conflict got me suspicious as well, and
after reading the motherboard manual a bit more carefully I found in a
small footnote: "PCI slot 4 share the same interrupt number as the
onboard SCSI so PCI slot 4 card must be able to share an INT# or make
sure that it does not use an INT# at all". This is the slot I had the
ethernet card in, now I don't. However I got curious when we replaced
the motherboard so I did put the ethernet card in there to try it out,
and it seems to work ok to share the interrupt. I ran a couple of
"bonnie" (on the SCSI disks) at the same time as transferring the
contents of a CDROM (also on SCSI) to another host on the network,
without any crashes or lockups.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199801191109.MAA17451>