Re: PCI not restored correctly after suspend to ram

[Pavel Machek <pavel-+ZI9xUNit7I@public.gmane.org>]

Hi!

> ...
> > I can get most devices devies to work by setting the pci registers with setpci 
> > after resume. I use the attached fixpci.sh script to do so. I can get my 
> > soundcard, the ethernet card(02:02.0) and the wireless(02:04.0) to work. I 
> > have not tested firewire(02:07.0). The secound Cardbus bride(02:06.1) seems 
> > to have more problems, it's not even listed correctly.
> > The file "lspci_after_suspend_fix" contains the output of lspci -vvx after 
> > running fixpci.sh
> > 
> > Restoring the pci registers is a workaround, but I'd like the kernel to handle 
> > this so I don't have to unload and reload all modules, especially the 
> > soundcard driver.
> > 
> > It this a acpi bug, a bug in the pci driver or a bug in the drivers for the 
> > devices?
> 
> Stefan,
> Bruno is correct, the PCI config space should be saved and restored by
> the driver's suspend and resume routines.  suspend/resume routines
> typically do other hardware dependent stuff, but your debugging suggests
> that at least for some of these devices, simply saving and restoring pci
> config space (pci_save_state()/pci_restore_state()) may be sufficient.
> 
> Time to take a look at the drivers for those devices...

Perhaps doing it once in PCI layer is better, then?

								Pavel

-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click