Re: [PATCH] remove dead drivers/ide/ppc/swarm.c

[Maciej W. Rozycki]

On Wed, 5 May 2004, Bartlomiej Zolnierkiewicz wrote:

> > so barring the completely bogus addresses reported (the interface is
> > decoded at 0x100b0000 and that's MMIO, unrelated to the PCI I/O space -- I
> > guess fixing it would require a significant update to ide-probe.c and
> > elsewhere), it looks OK.  This is in a little-endian configuration -- I'll
> 
> I suppose these are the addresses after ioremap() so no problem here.

 Not even that -- the exact calculation leading to the base address
reported is:

ioremap_nocache(0x100b0000) - ioremap(0xdc000000) + (0x1f0 << 5)

and is really a workaround for no better way of expressing an address of
an MMIO IDE interface (0xdc000000 is the base address of the PCI I/O
space, unswapped).  I see no reason to use virtual addresses for report 
here.

> > recheck for big endianness once I have appropriate userland installed
> > (which is likely not very soon).  It shouldn't matter much, though -- I've
> > checked documentation and I've found out that while the generic bus (the
> > IDE connector is wired to) is big-endian externally, the internal logic
> > does swapping if the system is configured for little-endian operation.
> 
> IDE bus must be always LE independently of system being LE or BE.
> 
> If system is LE this is OK as the internal logic does swapping but in case
> when system is BE swapping is also needed.

 Well, I suppose they have simply swapped bytes at the connector. ;-)  
This is a guess, though, as I lack docs in this area.  The important bit
is the endiannes of the generic bus used here never changes.

  Maciej

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@xxxxxxxxxxxxx, PGP key available        +
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/