[PATCH][BUG] fix bug in ACPI based CPU hotplug

[PATCH][BUG] fix bug in ACPI based CPU hotplug

[Kenji Kaneshige]

Hi,

I found a bug in ACPI based CPU hotplug code. Because of this bug,
ACPI based CPU hotplug will always fail if NR_CPUS is equal to or
more than 255.

Thanks,
Kenji Kaneshige


This patch fixes a serious bug in ACPI based CPU hotplug code. Because
of this bug, ACPI based CPU hotplug will always fail if NR_CPUS is
equal to or more than 255.

Signed-off-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: Motoyuki Itoh <motoyuki@soft.fujitsu.com>

---
 drivers/acpi/processor_core.c |   11 +++++------
 1 files changed, 5 insertions(+), 6 deletions(-)

Index: linux-2.6.16-rc5-mm2/drivers/acpi/processor_core.c
===================================================================
--- linux-2.6.16-rc5-mm2.orig/drivers/acpi/processor_core.c	2006-03-07 13:42:20.000000000 +0900
+++ linux-2.6.16-rc5-mm2/drivers/acpi/processor_core.c	2006-03-08 14:03:58.000000000 +0900
@@ -382,7 +382,7 @@
 
 /* Use the acpiid in MADT to map cpus in case of SMP */
 #ifndef CONFIG_SMP
-#define convert_acpiid_to_cpu(acpi_id) (0xff)
+#define convert_acpiid_to_cpu(acpi_id) (-1)
 #else
 
 #ifdef CONFIG_IA64
@@ -395,7 +395,7 @@
 #define ARCH_BAD_APICID		(0xff)
 #endif
 
-static u8 convert_acpiid_to_cpu(u8 acpi_id)
+static int convert_acpiid_to_cpu(u8 acpi_id)
 {
 	u16 apic_id;
 	int i;
@@ -421,7 +421,7 @@
 	acpi_status status = 0;
 	union acpi_object object = { 0 };
 	struct acpi_buffer buffer = { sizeof(union acpi_object), &object };
-	u8 cpu_index;
+	int cpu_index;
 	static int cpu0_initialized;
 
 	ACPI_FUNCTION_TRACE("acpi_processor_get_info");
@@ -466,8 +466,7 @@
 	cpu_index = convert_acpiid_to_cpu(pr->acpi_id);
 
 	/* Handle UP system running SMP kernel, with no LAPIC in MADT */
-	if (!cpu0_initialized && (cpu_index == 0xff) &&
-	    (num_online_cpus() == 1)) {
+	if (!cpu0_initialized && (cpu_index < 0) && (num_online_cpus() == 1)) {
 		cpu_index = 0;
 	}
 
@@ -480,7 +479,7 @@
 	 *  less than the max # of CPUs. They should be ignored _iff
 	 *  they are physically not present.
 	 */
-	if (cpu_index >= NR_CPUS) {
+	if (cpu_index < 0) {
 		if (ACPI_FAILURE
 		    (acpi_processor_hotadd_init(pr->handle, &pr->id))) {
 			ACPI_ERROR((AE_INFO,

Re: [PATCH][BUG] fix bug in ACPI based CPU hotplug

[Andrew Morton]

Kenji Kaneshige <kaneshige.kenji@soft.fujitsu.com> wrote:
>
> This patch fixes a serious bug in ACPI based CPU hotplug code. Because
>  of this bug, ACPI based CPU hotplug will always fail if NR_CPUS is
>  equal to or more than 255.

Looks rather similar to
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc5/2.6.16-rc5-mm3/broken-out/acpi-signedness-fix-2.patch
;)

I think they're functionally equivalent - the only difference is the ==-1
versus <0 comparisons.

Re: [PATCH][BUG] fix bug in ACPI based CPU hotplug

[Kenji Kaneshige]

Andrew Morton wrote:
> Kenji Kaneshige <kaneshige.kenji@soft.fujitsu.com> wrote:
> 
>>This patch fixes a serious bug in ACPI based CPU hotplug code. Because
>> of this bug, ACPI based CPU hotplug will always fail if NR_CPUS is
>> equal to or more than 255.
> 
> 
> Looks rather similar to
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc5/2.6.16-rc5-mm3/broken-out/acpi-signedness-fix-2.patch
> ;)
> 
> I think they're functionally equivalent - the only difference is the ==-1
> versus <0 comparisons.
> 
> 

Oh, I didn't noticed it.
Sorry for the noise.

Thanks,
Kenji Kaneshige