Re: [patch 3/3] clockevents: Fix resume logic - updated version

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

On Thursday, 10 May 2007 11:27, Thomas Gleixner wrote:
> On Thu, 2007-05-10 at 02:18 -0700, Andrew Morton wrote:
> > > > > If that patch makes the problem go away, then we should have a quite
> > > > > good hint what we need to look at.
> > > > 
> > > > No joy, sorry.  It still hangs at the last statement in acpi_evaluate_object().
> > > 
> > > Can you add "nolapic_timer" to the command line please ?
> > > 
> > 
> > That works.
> 
> Ok, that boils it down to the change, which affects the lapic timer
> resume. It does nothing else, than fiddling in some APIC registers, but
> I don't see how this affects the ACPI stuff. This smells extremly fishy.

Hmm, I wonder if it might be related to the execution of the _GTS ACPI control
method in acpi_enter_sleep_state_prep() in violation of the spec.

Andrew, could you please try to apply the following change on top of the
previous ones and see if the machine still hangs in the same place without
"nolapic_timer"?

---
 drivers/acpi/hardware/hwsleep.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: linux-2.6.21/drivers/acpi/hardware/hwsleep.c
===================================================================
--- linux-2.6.21.orig/drivers/acpi/hardware/hwsleep.c
+++ linux-2.6.21/drivers/acpi/hardware/hwsleep.c
@@ -200,10 +200,10 @@ acpi_status acpi_enter_sleep_state_prep(
 		return_ACPI_STATUS(status);
 	}
 
-	status = acpi_evaluate_object(NULL, METHOD_NAME__GTS, &arg_list, NULL);
+	/*status = acpi_evaluate_object(NULL, METHOD_NAME__GTS, &arg_list, NULL);
 	if (ACPI_FAILURE(status) && status != AE_NOT_FOUND) {
 		return_ACPI_STATUS(status);
-	}
+	}*/
 
 	/* Setup the argument to _SST */