On Wed, 4 May 2005, David Brownell wrote:
> On Sunday 01 May 2005 7:14 pm, Alan Stern wrote:
> > David:
> >
> > I found part of the source for the trouble I've had with root-hub
> > suspend/resume on OHCI. It's these two lines near the end of
> > ohci-hub.c:ohci_hub_suspend():
> >
> > if (status == 0)
> > hcd->state = HC_STATE_SUSPENDED;
>
> I think I remember why that's there. It pairs the earlier line
> in the same function to set hcd->state to QUIESCING. And the reason
> for that is because hcd->state is the only hook we have for, erm,
> quiescing the traffic going into the HCD.
More testing showed that this causes a problem at the end of ohci_irq(),
where the code says:
if (HC_IS_RUNNING(hcd->state)) {
ohci_writel (ohci, ints, ®s->intrstatus);
ohci_writel (ohci, OHCI_INTR_MIE, ®s->intrenable);
// flush those writes
(void) ohci_readl (ohci, &ohci->regs->control);
}
The HC_IS_RUNNING test fails after the state has been set to
HC_STATE_SUSPENDED, so the interrupt handling isn't completed. If I
replace that test with "if (1) {" then my system no longer hangs. It
still doesn't work correctly, in that the root hub doesn't get resumed as
one would expect, but that's a separate issue...
So one of the two places appears to be wrong. Either ohci_hub_suspend()
shouldn't set hcd->state to HC_STATE_SUSPENDED or ohci_irq() shouldn't
contain this test for HC_IS_RUNNING(hcd->state). Could this be an example
of trying to use hcd->state to mean too many different things?
Alan Stern
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4
opportunities to win big! Highest score wins.NEC IT Guy Games. Play to
win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
