Re: [RFC][PATCH -mm 3/3] PM: Disable _request_firmware before hibernation/suspend

["Rafael J. Wysocki" <rjw@sisk.pl>]

On Sunday, 27 May 2007 22:45, Michael-Luke Jones wrote:
> On 27 May 2007, at 21:31, Rafael J. Wysocki wrote:
> 
> > From: Rafael J. Wysocki <rjw@sisk.pl>
> >
> > Use a hibernation and suspend notifier to disable the firmware  
> > requesting
> > mechanism before a hibernation/suspend and enable it after the  
> > operation.
> >
> > Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
> >
> >  drivers/base/firmware_class.c |   36 ++++++++++++++++++++++++++++++ 
> > ++++++
> >  1 file changed, 36 insertions(+)
> 
> I don't like this approach, as I feel that the firmware loading  
> interface should be able to detect if a firmware load request is not  
> being handled, due to absence of userspace / hotplug handler presence.

In principle, I agree.  In practice, though, I don't know how to make this
happen.

> Other circumstances in which this can be a problem is during bootup  
> when request_firmware() calls can be made before userspace is up and  
> init has run (even in the presence of an initramfs).
> 
> (Slightly OT: A particularly nasty race is when an initramfs  
> userspace is present, but firmware loading cannot occur because init  
> has not run, so proc hasn't been mounted, so a hotplug event handler  
> cannot be registered, despite the fact that the firmware is sitting  
> on the ramdisk mounted correctly...)
> 
> In short, a more general solution would be preferred, and preferably  
> one which allows firmware loading to *actually* occur once userspace  
> has actually turned up and registered a handler :)

I agree, but well ... ;-)

Greetings,
Rafael
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/