From: Brian Maly <bmaly@redhat.com>
Date: Wed, 19 Dec 2007 15:07:38 -0500
Subject: [acpi] improve reporting of power states
Message-id: 47697A0A.8090009@redhat.com
O-Subject: [RHEL5 patch] ACPI: improve reporting of power states
Bugzilla: 210716

resolves BZ 210716

This patch does not really fix a bug, rather just improves reporting of
power states to ease in debugging ACPI bugs.
In the case of this BZ, the kernel appears to be trying to set a power
state higher than the parent because thats the power state it got from
the ACPI tables, so either hardware is powered off completely or we have
BIOS issue. With this patch you will be able to see if the power state
is D0-D3 and this will help a bit in debugging the issue. It saves a
step anyway, and these sorts of problem (power state conflicts) are not
uncommon, i.e. bound to be seen again.

Brian

diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c
index 766332e..75404bb 100644
--- a/drivers/acpi/bus.c
+++ b/drivers/acpi/bus.c
@@ -219,6 +219,10 @@ int acpi_bus_set_power(acpi_handle handle, int state)
 		printk(KERN_WARNING PREFIX
 			      "Cannot set device to a higher-powered"
 			      " state than parent\n");
+		printk(KERN_WARNING PREFIX
+			      "Device requested power state D%d but"
+			      " parents power state is at D%d\n", state, device->parent->power.state);
+
 		return -ENODEV;
 	}