Re: [PATCH] Correct autosuspend_delay for non-hubs.

On Sat, Sep 22, 2007 at 10:08:26AM -0400, Alan Stern wrote:
> On Fri, 21 Sep 2007, Sarah Sharp wrote:
> 
> > Without this patch, `cat /sys/bus/usb/devices/*/power/autosuspend` will return
> > 0 for any device that is not a hub.  This happens because of the integer divide
> > in show_autosuspend in sysfs.c
> > 
> > Signed-off-by: Sarah Sharp <[email protected]>
> > ---
> >  drivers/usb/core/quirks.c |    2 +-
> >  1 files changed, 1 insertions(+), 1 deletions(-)
> > 
> > diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c
> > index ebf3dc2..4a0e474 100644
> > --- a/drivers/usb/core/quirks.c
> > +++ b/drivers/usb/core/quirks.c
> > @@ -156,6 +156,6 @@ void usb_detect_quirks(struct usb_device *udev)
> >  	/* By default, disable autosuspend for all non-hubs */
> >  #ifdef	CONFIG_USB_SUSPEND
> >  	if (udev->descriptor.bDeviceClass != USB_CLASS_HUB)
> > -		udev->autosuspend_delay = -1;
> > +		udev->autosuspend_delay = -1 * HZ;
> >  #endif
> >  }
> 
> Amusingly, my patch as995 (submitted just a few minutes earlier) 
> already includes this change.

Yes, I was amused that they came in so close together. :)  Although I
think that setting autosuspend_disabled is a better solution.  Is there
any reason that autosuspend_delay should be set to -1?  The topic was
brought up before, but I never saw a response from Oliver.

Sarah

On Mon, Aug 20, 2007 at 12:28:13PM -0400, Alan Stern wrote:
> On Mon, 20 Aug 2007, Jean Delvare wrote:
> 
> > On Mon, 20 Aug 2007 10:00:26 -0400 (EDT), Alan Stern wrote:
> > > On Mon, 20 Aug 2007, Jean Delvare wrote:
> > > 
> > > > Hmm, maybe my question wasn't totally clear, let me rephrase. Shouldn't
> > > > sysfs reflect the fact that autosuspend is disabled for my device
> > > > somehow? When I read "2" from the autosuspend file, it gives me the
> > > > impression that autosuspend is enabled, and I think it's confusing.
> > > 
> > > There's a different sysfs attribute that gets set when autosuspend is 
> > > disabled.  The .../power/level attribute file can be set to "on", 
> > > "auto", or "suspend".
> > > 
> > > > (I thought that an autosuspend value of 0 meant that autosuspend was
> > > > disabled, but your reply make me understand that this isn't the case.
> > > > Sorry for the confusion.)
> > > 
> > > It was the case in 2.6.22, but in 2.6.23 0 means to autosuspend as soon 
> > > as possible.
> > 
> > Ah, OK, thanks for the clarification. But then isn't level=on redundant
> > with autosuspend=-1?
> 
> It is.  The intent was that the two attribute files could have 
> different permissions.  Oliver can explain it all.
> 
> Alan Stern

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFG9UR3RqNbFWK10kwRAnqPAKCWcG6L1LyALUQ4m07rh7NBBtYGYACfTsyV
e8WUwr1E50u++EOfMhFnK6Y=
=nqiK
-----END PGP SIGNATURE-----