Re: [RFC/PATCH 2/2] driver core: power management debugging

[Nigel Cunningham <nigel@nigel.suspend2.net>]

[-- Attachment #1.1: Type: text/plain, Size: 2687 bytes --]

Hi.

On Sat, 2007-04-28 at 10:42 -0400, Alan Stern wrote:
> On Sat, 28 Apr 2007, Nigel Cunningham wrote:
> 
> > Hi Alan.
> 
> > Sorry. I thought you were wrong for a minute, but then I looked again at
> > the messages in my dmesg...
> > 
> > [   33.944214] Device driver usbdev1.1_ep00 lacks bus and class support for being resumed.
> > [   34.051765] Device driver usbdev1.1_ep81 lacks bus and class support for being resumed.
> > [   34.113740] Device driver usbdev2.1_ep00 lacks bus and class support for being resumed.
> > [   34.221541] Device driver usbdev2.1_ep81 lacks bus and class support for being resumed.
> > [   34.251562] Device driver usbdev3.1_ep00 lacks bus and class support for being resumed.
> > [   34.361345] Device driver usbdev3.1_ep81 lacks bus and class support for being resumed.
> > 
> > They're coming from the other printk, of course.
> > 
> > > Now perhaps you would prefer to check the USB interface drivers -- there 
> > > are many of them, and quite a few don't have suspend or resume methods.  
> > > You would need to modify usb_register_driver() instead of 
> > > usb_register_device_driver().
> > 
> > Would they be the ones covered above?
> 
> No.  As Greg pointed out, these usbdevXX_epYY "devices" are nothing but 
> placeholders at the moment.  They don't actually do anything and they have 
> no need for power management.  (But they do manage to clutter up the 
> system log with lots of extraneous warnings from the PM core...)

Ok, so they could have the pm_safe flag set to suppress the message.

> > > On the other hand, the drivers' maintainers are probably quite aware of 
> > > the missing PM support, so it's not clear that printing out warning 
> > > messages will actually help anybody.
> > 
> > It can help the user, when they're looking for possibilities as to why things aren't working.
> 
> Maybe.  But the warnings will occur when the driver is registered, which 
> is often long before the problem shows up.  The user may not make the 
> connection.

True, but if we make them show up earlier, they at least have a chance
to see what might cause problems before the problems occur. If we do it
at the time, they might have zero chance to see messages like this.

> On a completely different topic: Nigel, now's your big chance!  If you 
> hurry, you can rename Suspend2 to Hibernate -- beating out Pavel, who will 
> then be forced to rename swsusp to Hibernate2!  :-)

LOL. I don't care about names, or about beating Pavel at anything. I
just want to get better hibernation support in the kernel. It would be
nice to get rid of that awful swsusp name though! :)

Nigel

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]