Dear Alan Stern,
> On Tue, 15 May 2012, Marek Vasut wrote:
> > Dear Alan Stern,
> >
> > > On Tue, 15 May 2012, Marek Vasut wrote:
> > > > This comes handy on the i.MX28 CPU, where the host IRQ also
> > > > signalises that the disconnection detector shall be enabled/disabled
> > > > on the PHY.
> > > >
> > > > Signed-off-by: Marek Vasut <marex@xxxxxxx>
> > > > Cc: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
> > > > Cc: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
> > > > Cc: Felipe Balbi <balbi@xxxxxx>
> > > > Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> > > > Cc: Linux USB <linux-usb@xxxxxxxxxxxxxxx>
> > > > ---
> > > >
> > > > drivers/usb/chipidea/host.c | 4 ++++
> > > > include/linux/usb/chipidea.h | 1 +
> > > > 2 files changed, 5 insertions(+)
> > > >
> > > > diff --git a/drivers/usb/chipidea/host.c
> > > > b/drivers/usb/chipidea/host.c index 9eacd21..97bb167 100644
> > > > --- a/drivers/usb/chipidea/host.c
> > > > +++ b/drivers/usb/chipidea/host.c
> > > > @@ -94,6 +94,10 @@ static const struct hc_driver ci_ehci_hc_driver =
> > > > {
> > > >
> > > > static irqreturn_t host_irq(struct ci13xxx *ci)
> > > > {
> > > >
> > > > + if (ci->udc_driver->notify_event)
> > > > + ci->udc_driver->notify_event(ci,
> > > > + CI13XXX_CONTROLLER_HOST_IRQ_EVENT);
> > > > +
> > > >
> > > > return usb_hcd_irq(ci->irq, ci->hcd);
> > > >
> > > > }
> > >
> > > This is a rather roundabout way of doing things. Why add a new
> > > notifier mechanism when you can simply register another interrupt
> > > handler for the same IRQ?
> >
> > You certainly can, but how would you access the EHCI registers from PHY
> > driver?
>
> The same way you get informed about ci above: The information would get
> passed to the interrupt handler.
>
> > They have separate memory resources. Because the notifier passes data
> > figured out by reading the EHCI register (USBSTS, PCD bit) and tells PHY
> > to enable/disable the disconnection detector accordingly.
>
> I'm not sure I really understand the purpose of all this. Why does the
> PHY driver need to know about every IRQ the host driver receives?
I use the .notify_event in ci13xxx to filter out this kind of event, read the
EHCI host register and then notify the PHY about the state of USB PCD.
> Also, since the code above is part of a _host_ driver, why does it call
> through a pointer in ci->udc_driver -- does that even make sense in a
> host-only situation?
Actually, that's the pointer to the ci_hrdc platform data. So it's always set,
in both host and gadget cases. It's just misnamed.
> And why can't it call the PHY driver directly?
I don't want to pull any PHY code into the ci13xxx host driver, it's completely
separate hardware.
> Alan Stern
Best regards,
Marek Vasut
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
