On Wednesday 14 June 2006 1:31 am, Rodolfo Giometti wrote:
> On Tue, Jun 13, 2006 at 09:00:51PM -0700, David Brownell wrote:
> > On Monday 29 May 2006 6:29 am, Rodolfo Giometti wrote:
> > > + .suspend = ohci_au1xxx_suspend,
> > > + .resume = ohci_au1xxx_resume,
> >
> > You shouldn't add both the PCI glue calls (those above) *AND*
> > the platform bus glue calls:
> >
> > > + .suspend = ohci_hcd_au1xxx_drv_suspend,
> > > + .resume = ohci_hcd_au1xxx_drv_resume,
> >
> > These platform bus glue calls should suffice ... nothing
> > should ever be calling the PCI bus glue calls on with a
> > platform bus device, in any case!!
>
> Mmm... I'm not sure to understand what you mean. Can you please
> explain better which is the problem or give me a URL where I can read
> something about?
See linux/drivers/usb/core/hcd.h where the first two
entries (in hc_driver) are described as PCI-only.
That driver uses platform_bus, not PCI.
The problem is exactly what I said: the ohci_au1xxx_{suspend,resume}
calls should not exist.
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
