On Wed, Jul 06, 2005 at 03:34:54AM +0400, Ivan Kokshaysky wrote:

On Tue, Jul 05, 2005 at 10:46:20PM +0100, Russell King wrote:

quoted

Rather than reimplementing the internals of pci_update_resource() it
may be worth splitting the common stuff out so it gets fixed for both
pci_update_resource() and pci_enable_device().

Just use pci_update_resource().

 
Problem: pci_update_resource doesn't exist for sparc64.

John, I'd also suggest following changes to the patch:
- move the code to pci_set_power_state(), where it belongs to;
- explicitly check for D3hot->D0 transition *and* for the
  No_Soft_Reset bit, to avoid unnecessary config space accesses;
- add a quote from PCI spec (as a comment) explaining why is it needed.

I have reformulated the patch to account for these comments, but I am
not currently using pci_update_resource for the reason stated above.
I'll go ahead and post the new patch for comment.  If we can resolve
the pci_update_resource issue, I'll post another (either alternative
or additional) patch to cover that.  Patch to follow...

Thanks!

John
-- 
John W. Linville
linville@tuxdriver.com