On Fri, 2007-02-09 at 12:49 -0500, Len Brown wrote:

On Friday 09 February 2007 12:14, James Morris wrote:

quoted

This is being disabled in the guest kernel only.  The host and guest 
kernels are expected to be the same build.

Okay, but better to use disable_acpi()
indeed, since this would be the first code not already inside CONFIG_ACPI
to invoke disable_acpi(), we could define the inline as empty and you could
then scratch the #ifdef too.

Thanks Len!

This applies on top of that series.

== 
Len Brown [off-list ref] said:

Okay, but better to use disable_acpi()
indeed, since this would be the first code not already inside CONFIG_ACPI
to invoke disable_acpi(), we could define the inline as empty and you could
then scratch the #ifdef too.

Signed-off-by: Rusty Russell <redacted>

diff -r 85363b87e20b arch/i386/lguest/lguest.c

--- a/arch/i386/lguest/lguest.c	Sat Feb 10 01:52:37 2007 +1100
+++ b/arch/i386/lguest/lguest.c	Sat Feb 10 10:28:36 2007 +1100

@@ -555,10 +555,7 @@ static __attribute_used__ __init void lg
 	mce_disabled = 1;
 #endif
 
-#ifdef CONFIG_ACPI
-	acpi_disabled = 1;
-	acpi_ht = 0;
-#endif
+	disable_acpi();
 	if (boot->initrd_size) {
 		/* We stash this at top of memory. */
 		INITRD_START = boot->max_pfn*PAGE_SIZE - boot->initrd_size;

diff -r 85363b87e20b include/asm-i386/acpi.h

--- a/include/asm-i386/acpi.h	Sat Feb 10 01:52:37 2007 +1100
+++ b/include/asm-i386/acpi.h	Sat Feb 10 10:43:43 2007 +1100

@@ -127,6 +127,7 @@ extern int acpi_irq_balance_set(char *st
 #define acpi_ioapic 0
 static inline void acpi_noirq_set(void) { }
 static inline void acpi_disable_pci(void) { }
+static inline void disable_acpi(void) { }
 
 #endif	/* !CONFIG_ACPI */