On Thu, Aug 24, 2006 at 01:58:48PM -0700, I?aky P?rez-Gonz?lez wrote:
>
> Okay, got feedback from everybody and rehashed it to current gregkh's queue.
>
> wusb: basic device infrastructure modifications [take 3]
>
> This patch does a bunch of small modifications to the USB core so that WUSB
> d
On Thu, 24 Aug 2006, Iñaky Pérez-González wrote:
> > That's not so easy either. The fact is, hdev_to_hub() _isn't_ safe unless
> > you possess some extra prior knowledge.
>
> which priori knowledge? I had assumed that as long as it has registered,
> we were safe.
It's not enough for it to be re
On Thursday 24 August 2006 11:17, Alan Stern wrote:
> On Thu, 24 Aug 2006, Perez-Gonzalez, Inaky wrote:
> > Nothing that exporting something like a safe_hdev_to_hub() to the
> > host API cannot do.
>
> That's not so easy either. The fact is, hdev_to_hub() _isn't_ safe unless
> you possess some ex
On Thu, 24 Aug 2006, Perez-Gonzalez, Inaky wrote:
> >This won't be easy. For one thing, hcd_root_hub_lock is private to
> hcd.c
> >and hdev_to_hub is private to hub.c.
>
> Nothing that exporting something like a safe_hdev_to_hub() to the
> host API cannot do.
That's not so easy either. The fac
>From: Alan Stern [mailto:[EMAIL PROTECTED]
>
>> I was thinking about something like the following in my hc_probe():
>>
>> ...
>
>This won't be easy. For one thing, hcd_root_hub_lock is private to
hcd.c
>and hdev_to_hub is private to hub.c.
Nothing that exporting something like a safe_hdev_to_hub
On Wed, 23 Aug 2006, Iñaky Pérez-González wrote:
> I was thinking about something like the following in my hc_probe():
>
> ...
> usb_add_hcd(hcd);
> if (wait_event_timeout(&wq, hcd->rh_registered != 0, TIMEOUT) == 0) {
> result = -ETIMEDOUT;
> goto error...;
> }
> spin_lock_irqsave (&
On Wednesday 23 August 2006 19:23, Alan Stern wrote:
> On Wed, 23 Aug 2006, Iñaky Pérez-González wrote:
> > Ok, I am officially dumb: hdev_to_hub(). Dah!
> >
> > I'll change the code tomorrow to set a flag variable in struct usb_hub
> > and leave port-status unmodified.
> >
> > Thanks for the nudge
On Wed, 23 Aug 2006, Iñaky Pérez-González wrote:
> Ok, I am officially dumb: hdev_to_hub(). Dah!
>
> I'll change the code tomorrow to set a flag variable in struct usb_hub
> and leave port-status unmodified.
>
> Thanks for the nudge, Alan
Just remember my warning. You'll get an oops if you try
On Wednesday 23 August 2006 2:27 pm, Greg KH wrote:
> What happens in 4 years when they come out with a use for that bit?
Unlikely, but that's the reason to avoid using bits at the low end;
they seem to count up not down.
- Dave
pgppSAnqTO2zJ.pgp
Description: PGP signature
---
On Wednesday 23 August 2006 15:10, Alan Stern wrote:
>
> It isn't trivial, but it's also not impossible. There are a couple of
> places where the kernel does it. For example, look at
> usb_hcd_resume_root_hub() in hcd.c, which calls usb_resume_root_hub() in
> hub.c. Or look at usb_hc_died() in h
On Wed, 23 Aug 2006, Iñaky Pérez-González wrote:
> I am not happy at all with that approach, I know it could be a future
> maintenance mess. And I would have much more preferred to set a bit in
> struct usb_hub.
>
> But as I said above, going from struct usb_hcd to it's root hub
> 'struct usb_
On Wednesday 23 August 2006 14:27, Greg KH wrote:
> On Wed, Aug 23, 2006 at 12:57:48PM -0700, I?aky P?rez-Gonz?lez wrote:
> > 2. Introduce USB_PORT_STAT_WIRELESS; we take one of the reserved bits
> >in the USB port status (bit 15) to report that we are a wireless
> >(fake port).
> >
> >
On Wed, Aug 23, 2006 at 12:57:48PM -0700, I?aky P?rez-Gonz?lez wrote:
>
> 2. Introduce USB_PORT_STAT_WIRELESS; we take one of the reserved bits
>in the USB port status (bit 15) to report that we are a wireless
>(fake port).
>
>Yes, bad bad, it is taken by the spec. The rationale for d
Ok, repost changing the reserved bit to 15 from 5 and after trying to use
a 'struct hub'->wireless bit (rationale below). This is against's gregkh's
quilt tree.
WUSB: basic device infrastructure modifications
This patch does a bunch of small modifications to the USB core so that WUSB
devices are
14 matches
Mail list logo
