> On Wed, 21 Oct 1998, Gerard Roudier wrote:
>
> > The attached patch is against linux-2.1.125. It contains several fixes
> > and some diffs for Sparc (All this is detailed in ChangeLog.ncr53c8xx).
> >
> > If you could give this patch a try, especially on Alpha and Sparc, and
> > report problems if any, this will help me a lot.
>
> I have tried this and the previous patch on my problem Alpha with its
> problem Maxtor SCSI disk. No luck. Eventually, I can always trigger the
> endless command timeouts - even with the default queueing at 3 and queue
> depth at 4 (per your suggestion). The only way to achieve reliable
> operation is to boot with 'ncr53c8xx=tags:0' or to build the kernel with
> the 2.5f driver (in which case queue depth of 8 works perfectly).
>
> I'll grant that this is very frustrating, and very likely that it's a bug
You shouldn't be frustrated, IMO. I have had a look at maxtor web site and
ftp site. Given the very fiew informations on their SCSI disks I have
found, I may think they do not really care of user's problems, at least
for SCSI products (or perhaps their SCSI products were quite perfect ;-).
> in the Maxtor firmware. Nonetheless, I thank you for your continued
> efforts!
Trying a different SCSI disk on your system and seeing what happens should
be interesting.
> If it's any help, the problem is far more likely to occur when the machine
> is under heavy network load, e.g. building the kernel on an NFS-mounted
> filesystem while I'm logged in from a remote X server.
Network activity means PCI traffic and system load. May be the real cause
of the problem is not due to SCSI.
Regards,
Gerard.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/