Re: [PATCH] Disable APIC on reboot.

[Mikael Pettersson]

davej@xxxxxxxxxx writes:
 > diff -urpN --exclude-from=/home/davej/.exclude bk-linus/arch/i386/kernel/reboot.c linux-2.5/arch/i386/kernel/reboot.c
 > --- bk-linus/arch/i386/kernel/reboot.c	2003-05-13 11:51:12.000000000 +0100
 > +++ linux-2.5/arch/i386/kernel/reboot.c	2003-07-16 02:54:29.000000000 +0100
 > @@ -8,6 +8,7 @@
 >  #include <linux/interrupt.h>
 >  #include <linux/mc146818rtc.h>
 >  #include <asm/uaccess.h>
 > +#include <asm/apic.h>
 >  #include "mach_reboot.h"
 >  
 >  /*
 > @@ -250,6 +251,19 @@ void machine_restart(char * __unused)
 >  	 */
 >  	smp_send_stop();
 >  	disable_IO_APIC();
 > +#else
 > +#ifdef CONFIG_X86_LOCAL_APIC
 > +	{
 > +	unsigned int l, h;
 > +
 > +	local_irq_disable();
 > +	disable_local_APIC();
 > +	rdmsr(MSR_IA32_APICBASE, l, h);
 > +	l &= ~MSR_IA32_APICBASE_ENABLE;
 > +	wrmsr(MSR_IA32_APICBASE, l, h);
 > +	local_irq_enable();
 > +}
 > +#endif

I agree we should probably disable the local APIC at reboot if we
enabled it previously, but this patch is broken. CONFIG_X86_LOCAL_APIC
doesn't imply that the CPU actually has one, and even if it does, the
access method may be different (e.g. P5 vs P6/K7/P4, and who knows how
the future C3 with local APIC will do it).

Only apic.c knows how the local APIC was initialised (if at all), so the
"disable it dammit" procedure needs to be in apic.c too.

Was the original bug report posted to LKML? I don't remember seeing it.

/Mikael
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/