Date: Fri, 9 Aug 2002 13:51:32 -0700 (PDT) From: Nate Lawson <nate@root.org> To: Matthew Jacob <mjacob@feral.com> Cc: Andrew Gallatin <gallatin@cs.duke.edu>, Marcin Gryszkalis <mgryszkalis@cerint.pl>, freebsd-scsi@FreeBSD.ORG, freebsd-alpha <freebsd-alpha@FreeBSD.ORG> Subject: Re: da driver problem? Message-ID: <Pine.BSF.4.21.0208091330210.23638-100000@root.org> In-Reply-To: <Pine.BSF.4.21.0208091218000.67834-100000@beppo>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 9 Aug 2002, Matthew Jacob wrote:
> On Fri, 9 Aug 2002, Nate Lawson wrote:
>
> > To be more specific, the Qlogic isp* cards ALWAYS require you to load
> > firmware to give full access to the drive and do anything useful. In the
> > x86 world, the isp boards only have enough code included on board to
> > manage int 13h for bootstrapping the OS and attempting to use them without
> > loading firmware is not defined. With the Qlogic 2xxx boards in
> > particular, NetBSD and FreeBSD won't even boot without loading the
> > firmware (kernel hangs after probing the card).
>
> This isn't always the case. Also, if this is the case, where's the PR? Don't
> start rumours, please.
I never filed a PR because I assumed this behavior was what was documented
in ispfw(4). I don't currently have any isp hw so this may have been
fixed after I used the cards (May-Sept 2001).
ispfw(4):
... the isp(4)
driver will notice that firmware is available to be downloaded onto
Qlogic cards (to replace the usually out of date firmware on the
cards).
This will kick the f/w into getting unstuck.
The problem I was referring to was the fact that for my 2100 and 2200, the
card would get wedged after printing the probe message if a drive was
attached and I didn't load firmware. Causing a LIP didn't help. Loading
firmware did work.
[good description of qlogic flash deleted]
> > Perhaps a comment should go into GENERIC above the isp entry saying to
> > be sure to enable ispfw_load? Or perhaps the hardware notes that go in
> > the release?
>
> well, it would be better that:
>
> a) The install process puts ispfw on the 3rd floppy
>
> b) Only the f/w relevant to a card is pulled in
>
> c) freeing up and unloading modules and reclaiming memory actually worked
All great suggestions. If b) was part of the probe routine, then
ispfw_load could default to YES.
-Nate
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0208091330210.23638-100000>