Re: [ACPI] ARB_DIS

[Ducrot B. <du...@po...>]

On Tue, Apr 08, 2003 at 01:26:29PM +0200, Ducrot Bruno wrote:
> Hi, Andy.
> 
> On Mon, Apr 07, 2003 at 11:40:00AM -0700, Grover, Andrew wrote:
> > Hi Eric,
> > 
> > So in a patch applied 6 months ago, you added code to restore the
> > ARB_DIS bit on resume from S1. In looking more closely, it appears that
> > this is a duplicate of code already in acpi_leave_sleep_state. Given
> > that the OS should be calling that function after resuming S1, I am
> > considering removing that line from acpi_enter_sleep_state, but wanted
> > to check with you before I did. Any reason why I should stay?
> > 
> 
> Really, you should track more closely authors of patches...
> 
> Anyway, since you have applied ARB_DIS before calling AML methods
> in acpi_leave_sleep_state(), this is OK for me.

Actually, I supposed you applied this one.  Please apply.


--- linux-2.5.68/drivers/acpi/hardware/hwsleep.c	2003/04/29 17:53:16	1.1
+++ linux-2.5.68/drivers/acpi/hardware/hwsleep.c	2003/04/29 17:53:45
@@ -401,6 +401,10 @@
 
 	/* Ensure enter_sleep_state_prep -> enter_sleep_state ordering */
 
+	/* Disable BM arbitration */
+	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
+
+
 	acpi_gbl_sleep_type_a = ACPI_SLEEP_TYPE_INVALID;
 
 	/* Setup parameter object */
@@ -429,9 +433,6 @@
 	if (ACPI_FAILURE (status)) {
 		return_ACPI_STATUS (status);
 	}
-
-	/* Disable BM arbitration */
-	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
 
 	return_ACPI_STATUS (status);
 }



Cheers,

-- 
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.