[3267] in linux-scsi channel archive
Re: devfs - SCSI - ZIP test
daemon@ATHENA.MIT.EDU (Eric Youngdale)
Sun Feb 8 21:02:10 1998
Date: Sun, 8 Feb 1998 21:23:44 -0500 (EST)
From: Eric Youngdale <eric@andante.jic.com>
To: Richard Gooch <rgooch@atnf.CSIRO.AU>
cc: "Leonard N. Zubkoff" <lnz@dandelion.com>, linux-scsi@vger.rutgers.edu
In-Reply-To: <199802090132.MAA32000@vindaloo.atnf.CSIRO.AU>
I am not 100% sure myself - there is a comment in hosts.c that
says we shouldn't do it, but I cannot think of any reason why myself. I
would suggest asking Dario Ballabio, since it is his comment.
-Eric
"The woods are lovely, dark and deep. But I have promises to keep,
And lines to code before I sleep, And lines to code before I sleep."
On Mon, 9 Feb 1998, Richard Gooch wrote:
> Leonard N. Zubkoff writes:
> > Date: Sun, 8 Feb 1998 11:25:23 +1100
> > From: Richard Gooch <rgooch@atnf.CSIRO.AU>
> >
> > How about a simple patch to re-use the last SCSI host number if the
> > driver which allocated it was unloaded? Leonard?
> >
> > Back when this feature was first introduced, I wanted something similar and I
> > vaguely recall there was a reason that modules behave the way they do. I
> > personally don't see anything wrong with reusing the number if the
> > load/unbloads nest, but there may be something I don't know about. Why not
> > propose this on linux-scsi and see if anyone sees a reason not to do this.
>
> OK, done.
>
> Just to give background for people on linux-scsi: the original problem
> which prompted this was that when a SCSI driver is loaded/unloaded
> many times, the SCSI host number increases without each time (until it
> wraps). This is particularly unhelpful with devfs.
>
> To solve this, a simple patch to decrement the host counter when a
> module with the last allocated SCSI number is unloaded would do
> nicely, I think.
>
> Regards,
>
> Richard....
>
