Re: 2.6.21-rc1: known regressions (part 2)

From: Ingo Molnar <mingo@elte.hu>
Date: Mar 5, 2007, 16:41 UTC
Message-ID: <20070305164133.GA2441@elte.hu>
In-Reply-To: <20070305161450.GA4972@mellanox.co.il>

* Michael S. Tsirkin <mst@mellanox.co.il> wrote:

Show 6 quoted lines

> > Quoting Ingo Molnar <mingo@elte.hu>:
> >  git-bisect good      0539771d7236b425f285652f6f297cc7939c8f9a
> > 
> >  81450b73dde07f473a4a7208b209b4c8b7251d90 is first bad commit
> 
> I have confirmed these two on my system.

you could probably get quite a bit further in bisecting the other breakage, by using the following method:

manully apply the patch below to 81450b73dde and retest. It will most likely work. Then FIRST unapply the patch and mark the tree via 'git-bisect good' and continue the bisection. Then try to apply the patch again. If it's already included - ignore the rejected patch. Whenever git-bisect offers you a new commit, just try to apply the patch. Ok? This way you'll be able to avoid the known ACPI breakage, and zoom in on the unknown breakage.

	Ingo

---------------->
commit f3ccb06f3b8e0cf42b579db21f3ca7f17fcc3f38
Author: Alexey Starikovskiy <alexey.y.starikovskiy@linux.intel.com>
Date:   Tue Feb 13 02:35:50 2007 -0500

    ACPI: Disable wake GPEs only once.
    
    fixes Suspend/Resume regressions due to recent ACPICA update.
    
    Signed-off-by: Alexey Starikovskiy <alexey.y.starikovskiy@linux.intel.com>
    Signed-off-by: Len Brown <len.brown@intel.com>

diff --git a/drivers/acpi/events/evgpe.c b/drivers/acpi/events/evgpe.c
index dfac3ec..635ba44 100644
--- a/drivers/acpi/events/evgpe.c
+++ b/drivers/acpi/events/evgpe.c
@@ -636,17 +636,6 @@ acpi_ev_gpe_dispatch(struct acpi_gpe_event_info *gpe_event_info, u32 gpe_number)
 		}
 	}
 
-	if (!acpi_gbl_system_awake_and_running) {
-		/*
-		 * We just woke up because of a wake GPE. Disable any further GPEs
-		 * until we are fully up and running (Only wake GPEs should be enabled
-		 * at this time, but we just brute-force disable them all.)
-		 * 1) We must disable this particular wake GPE so it won't fire again
-		 * 2) We want to disable all wake GPEs, since we are now awake
-		 */
-		(void)acpi_hw_disable_all_gpes();
-	}
-
 	/*
 	 * Dispatch the GPE to either an installed handler, or the control method
 	 * associated with this GPE (_Lxx or _Exx). If a handler exists, we invoke