Re: 2.4.5-ac8 hardlocks when going to standby

[Remi Turk <remi@a2zis.com>]

On Wed, Jun 06, 2001 at 03:10:11PM +0200, Mikael Pettersson wrote:
> > On Tue, Jun 05, 2001 at 10:18:07PM +0100, Alan Cox wrote:
> > > Thanks. UP-APIC is a real candidate for this case.
> 
> Actually, I suspect apm.c is at fault here. Suspend works,
> which proves that the PM code in apic.c and nmi.c works.
> 
> But note how apm.c:send_event() ignores standby events and fails
> to propagate them to PM clients. Thus, Remi's box will have an
> activated local APIC and live NMI watchdog when the APM BIOS
> finally gets to do whatever it does at standby.
> It is fatal to have an active local APIC and NMI watchdog at suspend,
> and I can only assume that this is true for standby as well.
> 
> Please try changing apm.c:send_event() to propagate standbys to PM
> clients just like suspends. Does this fix the problem?

By applying the following patch (lookalike)?

arch/i386/kernel/apm.c:send_event():

	case APM_SYS_SUSPEND:
	case APM_CRITICAL_SUSPEND:
	case APM_USER_SUSPEND:
+	case APM_USER_STANDBY:
+	case APM_SYS_STANDBY:
		/* map all suspends to ACPI D3 */
		if (pm_send_all(PM_SUSPEND, (void
			*)3)) {

> 
> (Any why use standby in the first place? Any reason you don't
> want to / can't use suspend?)

Because IIRC I can only choose between power-off and standby
as functions of my powerbutton ;-)

> 
> /Mikael

-- 
Linux 2.4.5-ac9 #5 Wed Jun 6 18:30:24 CEST 2001