Skip to contentSkip to conversation

acpi: Fix CPU hot removal problem

PATCH Superseded canquan.shen 2 participants 3 messages Sep 8, 2011 – Sep 13, 2011 ACPI

v2 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 acpi: Fix CPU hot removal problem

We run Linux as a guest in Xen environment. When used xen tools (xm vcpu-set <n>) to hot add and remove vcpu to and from the guest,
we encountered the failure on vcpu removal. We found the reason is
that it didn't go to really remove cpu in the cpu removal code path.

This patch adds acpi_bus_trim in acpi_process_hotplug_notify to fix this issue. With this patch, it works fine for us.

---
 drivers/acpi/processor_driver.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)
drivers/acpi/processor_driver.c +5 -0
index a1c564f..114592f 100644
@@ -678,7 +678,11 @@ static void acpi_processor_hotplug_notify(acpi_handle handle,
             return;
         }
 
+        if (acpi_bus_trim(device, 1)) {
+            printk(KERN_ERR PREFIX
+                    "Removing device failed\n");
+            return;
+        }
         break;
     default:
         ACPI_DEBUG_PRINT((ACPI_DB_INFO,
Signature · 1 line
1.7.6
#1 lore raw
Quoted 29 lines · canquan.shen

We run Linux as a guest in Xen environment. When used xen tools (xm vcpu-set <n>) to hot add and remove vcpu to and from the guest, we encountered the failure on vcpu removal. We found the reason is that it didn't go to really remove cpu in the cpu removal code path.

This patch adds acpi_bus_trim in acpi_process_hotplug_notify to fix this issue. With this patch, it works fine for us.

---
drivers/acpi/processor_driver.c | 5 +++++

1 files changed, 5 insertions(+), 0 deletions(-)
drivers/acpi/processor_driver.c +5 -0
index a1c564f..114592f 100644
@@ -678,7 +678,11 @@ static void
 acpi_processor_hotplug_notify(acpi_handle handle,
 return;
 }
 
+ if (acpi_bus_trim(device, 1)) {
+ printk(KERN_ERR PREFIX
+ "Removing device failed\n");
+ return;
+ }
 break;
Quoted 2 lines

default:
ACPI_DEBUG_PRINT((ACPI_DB_INFO,

Hi Len,
I just see your email in linux-kernel maillist and known that your
kernel.org email is kaput, so I resend this email to your intel email address.

Thanks,
- Canquan Shen

#2 lore raw
Quoted 25 lines · canquan.shen
Quoted 24 lines · canquan.shen

We run Linux as a guest in Xen environment. When used xen tools (xm vcpu-set <n>) to hot add and remove vcpu to and from the guest, we encountered the failure on vcpu removal. We found the reason is that it didn't go to really remove cpu in the cpu removal code path.

This patch adds acpi_bus_trim in acpi_process_hotplug_notify to fix this issue. With this patch, it works fine for us.

---
drivers/acpi/processor_driver.c | 5 +++++

1 files changed, 5 insertions(+), 0 deletions(-)
drivers/acpi/processor_driver.c +0 -0
index a1c564f..114592f 100644
@@ -678,7 +678,11 @@ static void
 acpi_processor_hotplug_notify(acpi_handle handle,
 return;
 }
 
Quoted 8 lines
Quoted 8 lines
drivers/acpi/processor_driver.c +5 -0
+ if (acpi_bus_trim(device, 1)) {
+ printk(KERN_ERR PREFIX
+ "Removing device failed\n");
+ return;
+ }
 break;

default:
ACPI_DEBUG_PRINT((ACPI_DB_INFO,

I think this patch is OK. One nit is that kernel printks should never be constant strings. There's almost always more information that's useful in debugging. This error should at least tell us which driver is involved, and preferably which specific device (though we have to be careful in case the device struct is freed by acpi_bus_trim()).

Bjorn