Re: [markgross@thengar.org: [RFC] wake up notifications and suspendblocking (aka more wakelock stuff)]

[Alan Stern]

On Wed, 12 Oct 2011, mark gross wrote:

> > > > sometimes devices that are not wake up sources need critical sections
> > > > where suspend is a problem.
> > 
> > Mark, can you give some examples?  This isn't the same as the firmware 
> > update thing, is it?  That deserves to be handled by a completely 
> > separate mechanism.
> 
> The biggest example I know of is any usb gadget implementation while
> connected to a USB host.  If the device is connected to the USB bus then
> it needs to not suspend because it cannot respond to the USB commands
> for bus power.  i.e. if its connected to a laptop and the suspend the
> laptop the gadget needs to handle the USB protocol packets.
> 
> Also, when charging battery under OS control we don't want to suspend
> because it will be hard to respond to thermal events  or changing
> battery conditions if the system is in suspend.

This is like the firmware update problem -- it has nothing to do with 
wakeups.  It could be handled very easily either by abusing a wakeup 
source as Neil suggested or by adding a different mechanism for 
preventing system sleeps.

> > Although many of these problems could be solved by adopting a suitable
> > protocol in userspace, currently there is no such protocol.  Even if
> > one did exist, the process of getting all the relevant programs to
> > adopt it would take quite a while.  This is a case where a problem can
> > be solved either in the kernel or in userspace, and the in-kernel
> > solution may be simpler.
> 
> I do think a better solution would involve some kernel coordination.
> 
> I don't want to split hairs on if its 100% possible to implement a
> correct solution only in user mode.  but, I think there are corner cases
> that may be hard or impossible to get right with only user mode.  See
> gadget issue above.  Not all gadgets have a user mode daemon backing up
> the kernel part do they?

I don't know.  But if they don't have some sort of userspace component
for power management then they never go into suspend, because the
kernel doesn't initiate suspends by itself.

Alan Stern

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