Re: iBook G4 PmDisk

[Michael Schmitz <schmitz@opal.biophys.uni-duesseldorf.de>]

> > > Mine is a PowerBook5,5 model. I'll narrow down the list of modules
> > > interfering with suspend (I'm not happy to lose the USB mouse, for
> > > instance). There's another annoying feature of suspend - the system clock
> > Well, the mouse does work again after resume, does it?
>
> Sure. And unloading uhci_hcd is absolutely required for resume :-(

ohci_hcd is what I meant, of course. There's suspend support but that
seems broken.

> therm_adt746x won't stop on suspend, that's something that might be fixed,
> though.

The following patch fixes suspend with the thermostat driver loaded for
me. Please test on other hardware. A similar fix (adding the if
(current->flags & PF_FREEZE) refrigerator(PF_FREEZE);) should be
sufficient for the therm_windtunnel module. May require conditionalizing
on CONFIG_PM_DISK as far as the refrigerator call is concerned...


--- drivers/macintosh/therm_adt746x.c.org	2004-09-02 15:18:04.000000000 +0200
+++ drivers/macintosh/therm_adt746x.c	2004-09-02 16:22:46.893592704 +0200
@@ -228,6 +228,18 @@
 	th->last_speed[fan] = speed;
 }

+void msleep_interruptible(unsigned int msecs)
+{
+	unsigned long timeout = msecs_to_jiffies(msecs);
+
+	while (timeout) {
+		set_current_state(TASK_INTERRUPTIBLE);
+		if (current->flags & PF_FREEZE)
+			refrigerator(PF_FREEZE);
+		timeout = schedule_timeout(timeout);
+	}
+}
+
 static int monitor_task(void *arg)
 {
 	struct thermostat* th = arg;
@@ -246,7 +258,8 @@

 	while(monitor_running)
 	{
-		msleep(2000);
+		/* msleep waits uninterruptibly! */
+		msleep_interruptible(2000);

 		/* Check status */
 		/* local   : chip */

Hope I'm not missing something obvious,

	Michael