Linux MIPS Architecture development
 help / color / mirror / Atom feed
* pci-ip27 memory ranges
@ 2004-04-24  9:00 Stanislaw Skowronek
  2004-04-24 10:30 ` Christoph Hellwig
  2004-04-24 18:21 ` Ralf Baechle
  0 siblings, 2 replies; 3+ messages in thread
From: Stanislaw Skowronek @ 2004-04-24  9:00 UTC (permalink / raw)
  To: linux-mips

Why does the pci-ip27 driver register all memory resources (0UL - ~0UL)? I
had to change it to the appropriate Xtalk window to get the kernel to
recognize PCI at all.

Now, the IOC3 makes a kernel panic. Will see...

Stanislaw Skowronek

--<=>--
  Paranoid: one who is truly in touch with reality.

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: pci-ip27 memory ranges
  2004-04-24  9:00 pci-ip27 memory ranges Stanislaw Skowronek
@ 2004-04-24 10:30 ` Christoph Hellwig
  2004-04-24 18:21 ` Ralf Baechle
  1 sibling, 0 replies; 3+ messages in thread
From: Christoph Hellwig @ 2004-04-24 10:30 UTC (permalink / raw)
  To: Stanislaw Skowronek; +Cc: linux-mips

On Sat, Apr 24, 2004 at 11:00:44AM +0200, Stanislaw Skowronek wrote:
> Why does the pci-ip27 driver register all memory resources (0UL - ~0UL)?

The current PCI support for IP27 is more or less a huge hack.  I plan to
hack it up to do the right thing based on the IRIX code for it released
by SGI, but it'll take a bit because that code isn't exactly readable
and my time is rather limited.

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: pci-ip27 memory ranges
  2004-04-24  9:00 pci-ip27 memory ranges Stanislaw Skowronek
  2004-04-24 10:30 ` Christoph Hellwig
@ 2004-04-24 18:21 ` Ralf Baechle
  1 sibling, 0 replies; 3+ messages in thread
From: Ralf Baechle @ 2004-04-24 18:21 UTC (permalink / raw)
  To: Stanislaw Skowronek; +Cc: linux-mips

On Sat, Apr 24, 2004 at 11:00:44AM +0200, Stanislaw Skowronek wrote:

> Why does the pci-ip27 driver register all memory resources (0UL - ~0UL)? I
> had to change it to the appropriate Xtalk window to get the kernel to
> recognize PCI at all.

The IP27 PCI code as it is in CVS is incomplete or even broken - however
fixing isn't entirely trivial.  The current workaround which I use - and
which due to being just an evil hack is not in CVS - to only scan the PCI
code.  The problem is the firmware does a very simple setup which hurts
performance very badly, so eventually Linux will have to do a full PCI
configuration for all PCI busses.

> Now, the IOC3 makes a kernel panic. Will see...

  Ralf

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2004-04-24 18:22 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-04-24  9:00 pci-ip27 memory ranges Stanislaw Skowronek
2004-04-24 10:30 ` Christoph Hellwig
2004-04-24 18:21 ` Ralf Baechle

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox