On Mon, 30 Oct 2006, Oliver Neukum wrote:
> Am Montag, 30. Oktober 2006 16:32 schrieb Alan Stern:
> > in its release() routine. The driver doesn't need to maintain its own
> > counter; that purpose is served by intf->pm_usage_cnt.
>
> It doesn't need to maintain a counter for this purpose. My
>
Am Montag, 30. Oktober 2006 16:32 schrieb Alan Stern:
> in its release() routine. The driver doesn't need to maintain its own
> counter; that purpose is served by intf->pm_usage_cnt.
It doesn't need to maintain a counter for this purpose. My
point is that the driver is likely to need to maintain
On Mon, 30 Oct 2006, Oliver Neukum wrote:
> > Take a look at usb-skeleton.c to see how they can be used. They are well
> > suited to the drivers which present a simple character-device interface to
> > userspace.
>
> It seems to me that those drivers use:
>
> if (desc->counter++) {
> ...
> }
>
Am Montag, 30. Oktober 2006 03:21 schrieb Alan Stern:
> On Sun, 29 Oct 2006, Oliver Neukum wrote:
>
> > Am Sonntag, 29. Oktober 2006 16:21 schrieben Sie:
> > > If you're only going to disable autopm on open and enable on close then
> > > you could have used usb_autopm_get_interface() and
> >
> >
On Sun, 29 Oct 2006, Oliver Neukum wrote:
> Am Sonntag, 29. Oktober 2006 16:21 schrieben Sie:
> > If you're only going to disable autopm on open and enable on close then
> > you could have used usb_autopm_get_interface() and
>
> Upon further thought, I fail to see the need for these interfaces.
>
On Sun, 29 Oct 2006, Oliver Neukum wrote:
> Hi,
>
> this patch makes hid devices do simple open/close powersave.
> This is helpful only in a minority of cases, but it works. Powersave
> with timeouts will need this as a foundation.
> Alan, I've made use of your suggestion and put the macros into
Am Sonntag, 29. Oktober 2006 16:21 schrieben Sie:
> If you're only going to disable autopm on open and enable on close then
> you could have used usb_autopm_get_interface() and
Upon further thought, I fail to see the need for these interfaces.
They do something that is simpler done within a driver
On Sun, 29 Oct 2006, Oliver Neukum wrote:
> > The usb_autpm_enable/disable things should go in a separate patch. Since
> > you find them useful, I'll add them into the usb_autopm_set_interface
> > patch and submit that on its own. It might be a good idea to make them
> > static inlines rather
Am Sonntag, 29. Oktober 2006 16:21 schrieb Alan Stern:
> P.S.: Did you ever figure out why ksuspend_usbd was using 97% of your CPU
> time?
Quite a coincidence. Yes, I've figured it out. My fault.
Regards
Oliver
Am Sonntag, 29. Oktober 2006 16:21 schrieb Alan Stern:
> On Sun, 29 Oct 2006, Oliver Neukum wrote:
>
> > Hi,
> >
> > this patch makes hid devices do simple open/close powersave.
> > This is helpful only in a minority of cases, but it works. Powersave
> > with timeouts will need this as a foundati
Hi,
this patch makes hid devices do simple open/close powersave.
This is helpful only in a minority of cases, but it works. Powersave
with timeouts will need this as a foundation.
Alan, I've made use of your suggestion and put the macros into
usb.h so they can be commonly used.
Regards
11 matches
Mail list logo
