Re: [PATCH] Switch ide parameters to new-style and make them unique.

["Jeremy Jackson" <jerj@coplanar.net>]

Cool stuff.

As far as making the parameters easy to parse, I think you would want to
have a single static tag before the = (equal) sign.  The kernel command line
parsing stuff provides parsing up to that point and dispatches to each
subsystem (or at least it used to), so:

ata.dev_noprobe=hda

should be

ata=dev_noprobe:hda,if_io_irq:0,0x1f0,7

or some such to use the generic code that's already there.

and possibly instead of io/irq pass PCI bus:device:fn and make passing raw
ioports a special case

Cheers,

Jeremy
----- Original Message -----
From: "Bartlomiej Zolnierkiewicz" <B.Zolnierkiewicz@elka.pw.edu.pl>
To: "Jeremy Jackson" <jerj@coplanar.net>
Cc: "Rusty Russell" <rusty@rustcorp.com.au>; "Alan Cox"
<alan@lxorguk.ukuu.org.uk>; <linux-kernel@vger.kernel.org>
Sent: Sunday, May 11, 2003 11:32 AM
Subject: Re: [PATCH] Switch ide parameters to new-style and make them
unique.


>
> On Sun, 11 May 2003, Jeremy Jackson wrote:
>
> > Haven't tested, but I have a few comments.
> >
> > First, I think this is a great step in the right
> > direction for the ide driver.
>
> Thanks.
>
> > I think at some point, the kernel command line parameters should be
> > consolidated behind a single ata=hda,noprobe or ata=if0,io0x1f0,irq7
type
> > parameter, instead of the hda= and ide0=.  Taking that one step furthur,
a
> > new syntax is needed, and having it go into 2.6 might pave the way for
> > removing the old cruft in 2.8?
>
> Yes, I was thinking about it.
> I prefer as easily parsable parameters as possible :-)
> fe. ata.dev_noprobe=hda and ata.if_io_irq=0,0x1f0,7.
>
> It is a question of shorter command line vs shorter parsing code.
>
> About a 2.6:
> I want to mark old command line plus HDIO_DRIVE_CMD and HDIO_DRIVE_TASK
> ioctls as obsoleted for 2.6.x and remove these cruft early in 2.7.x.
> I hope I will manage to do it, but it depends on Linus.
>
> > That would seem necessary, as I see it, to remove static ide_hwifs and
> > eventually support better hotswap.  (But even if it doesn't it would
still
> > clean up the ide parameters)
>
> FYI I have just done dynamic ide_hwifs allocations, patch needs
>     finishing (pdc4030 special case), polishing and testing.
>
> Regards,
> --
> Bartlomiej
>