I agree. But so what? The portable way is noted not to work with the
Future Domain hardware since 1996....
That hasn't changed.
So.... do we select a "portable" way to do it? Or do we want the
hardware to ACTUALLY WORK??? It doesn't do ANY GOOD AT ALL to have a
portable driver if it doesn't work with the hardware.
If you got a better way to do it that works, please, do it. Otherwise,
leave the loop for the PCI detection in the driver.
Brian
Martin Mares wrote:
>
> Hi,
>
> Scanning the PCI busses directly as you do in your patch is both ugly and
> non-portable. We should find a better solution, preferably debug my original
> probing code which makes use of the generic PCI layer.
>
> > + *** 19 Apr 1998: As James T. McKinley wrote above, there is a problem
> > + with the Future Domain hardware that prevents pcibios_find_device
> > + (now pci_find_device) from operating properly. Following is the
> > + device call properly implemented by Martin Mares:
> > +
> > + if ((pdev = pci_find_device(PCI_VENDOR_ID, PCI_DEVICE_ID_FD_36C70, pdev)) == NULL)
> > + return 0;
> > +
> > + Unfortunately, it always returns NULL on my Adaptec 2920 SCSI host
> > + adaptor. James reported to me that several people have written him
> > + about the problem with the new driver, so here is the original loop
> > + routine that he wrote to detect the PCI adaptors properly.
>
> Yes, it _must_ always return NULL as you've misspelled my misspelling fix I've
> sent you. The correct line is:
>
> if ((pdev = pci_find_device(PCI_VENDOR_ID_FD, PCI_DEVICE_ID_FD_36C70, pdev)) == NULL)
>
> and it should be present in 2.1.97. Please test that.
>
> Have a nice fortnight
> --
> Martin `MJ' Mares <mj@ucw.cz> http://atrey.karlin.mff.cuni.cz/~mj/
> Faculty of Math and Physics, Charles University, Prague, Czech Rep., Earth
> "In accord to UNIX philosophy, PERL gives you enough rope to hang yourself." -- Larry Wall
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu