[PATCH] Alexander Malysh's forgotten acpi_leave_sleep_state() IRQ patch

[Huw Rogers <count0-tC47gz4GrgtWk0Htik3J/w@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 607 bytes --]

Feb 14, Alexander Malysh wrote:
> We disable non wakeup gpes (and store those values) with interrupts disabled, 
> but enable those with interrupts enabled which were already overwritten while 
> handling of wakeup interrupt. So we must restore non wakeup gpes as long 
> interrupts are disabled.

The second half of his patch never made it into 2.6.5, and remains
necessary. Without this patch my N258SA0 is dead on resume, requiring
removal of battery and mains power cable to get it back. With this patch,
I get some improvement. Attached.
-- 
Huw Rogers <count0-tC47gz4GrgtWk0Htik3J/w@public.gmane.org>

[-- Attachment #2: acpi_leave_sleep_state_irq_fix.diff --]
[-- Type: application/octet-stream, Size: 706 bytes --]

--- linux-2.6.5/drivers/acpi/sleep/main.c.orig	2004-05-01 22:34:11.000000000 -0400
+++ linux-2.6.5/drivers/acpi/sleep/main.c	2004-05-01 22:34:52.000000000 -0400
@@ -107,7 +107,6 @@
 	default:
 		return -EINVAL;
 	}
-	local_irq_restore(flags);
 	printk(KERN_DEBUG "Back to C!\n");
 
 	/* restore processor state
@@ -118,6 +117,8 @@
 	if (state > PM_SUSPEND_STANDBY)
 		acpi_restore_state_mem();
 
+	acpi_leave_sleep_state(state);
+	local_irq_restore(flags);
 
 	return ACPI_SUCCESS(status) ? 0 : -EFAULT;
 }
@@ -133,8 +134,6 @@
 
 static int acpi_pm_finish(u32 state)
 {
-	acpi_leave_sleep_state(state);
-
 	/* reset firmware waking vector */
 	acpi_set_firmware_waking_vector((acpi_physical_address) 0);