Skip to contentSkip to conversation

powerpc/pseries: unregister VPA when hot unplugging a CPU

PATCH Applied Laurent Dufour 3 participants 3 messages Nov 14, 2022 – Dec 8, 2022 PowerPC

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 powerpc/pseries: unregister VPA when hot unplugging a CPU

The VPA should unregister when offlining a CPU. Otherwise there could be a short window where 2 CPUs could share the same VPA.

This happens because the hypervisor is still keeping the VPA attached to the vCPU even if it became offline.

Here is a potential situation:
1. remove proc A,
2. add proc B. If proc B gets proc A's place in cpu_present_map, then it registers proc A's VPAs.
3. If proc B is then re-added to the LP, its threads are sharing VPAs with proc A briefly as they come online.

As the hypervisor may check for the VPA's yield_count field oddity, it may detects an unexpected value and kill the LPAR.

---
 arch/powerpc/platforms/pseries/hotplug-cpu.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/pseries/hotplug-cpu.c +1 -0
index e0a7ac5db15d..090ae5a1e0f5 100644
@@ -70,6 +70,7 @@ static void pseries_cpu_offline_self(void)
 		xics_teardown_cpu();
 
 	unregister_slb_shadow(hwcpu);
+	unregister_vpa(hwcpu);
 	rtas_stop_self();
 
 	/* Should never get here... */
Signature · 1 line
2.38.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 27 lines · Laurent Dufour

The VPA should unregister when offlining a CPU. Otherwise there could be a short window where 2 CPUs could share the same VPA.

This happens because the hypervisor is still keeping the VPA attached to the vCPU even if it became offline.

Here is a potential situation:
1. remove proc A,
2. add proc B. If proc B gets proc A's place in cpu_present_map, then it registers proc A's VPAs.
3. If proc B is then re-added to the LP, its threads are sharing VPAs with proc A briefly as they come online.

As the hypervisor may check for the VPA's yield_count field oddity, it may detects an unexpected value and kill the LPAR.

---
 arch/powerpc/platforms/pseries/hotplug-cpu.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/pseries/hotplug-cpu.c +0 -0
index e0a7ac5db15d..090ae5a1e0f5 100644

arch/powerpc/platforms/pseries/hotplug-cpu.c static void pseries_cpu_offline_self(void) · line 76

Quoted 7 lines
arch/powerpc/platforms/pseries/hotplug-cpu.c +1 -0
@@ -70,6 +70,7 @@ static void pseries_cpu_offline_self(void)
 		xics_teardown_cpu();
 
 	unregister_slb_shadow(hwcpu);
+	unregister_vpa(hwcpu);
 	rtas_stop_self();
 
 	/* Should never get here... */

I was wondering whether we could leave an active dispatch trace log buffer registered, which could interfere with releasing the VPA, but I verified that DTL has the appropriate cpuhp callback for that (dtl_worker_offline()).

Alternatively we could change the code to dynamically register and unregister VPAs only on processor add and remove, as opposed to CPU online/offline. But I can't see any significant advantage to that.

#2 lore raw
In reply to Laurent Dufour
Quoted 6 lines · Laurent Dufour

The VPA should unregister when offlining a CPU. Otherwise there could be a short window where 2 CPUs could share the same VPA.

This happens because the hypervisor is still keeping the VPA attached to the vCPU even if it became offline.

Quoted 8 lines

Here is a potential situation:
1. remove proc A,
2. add proc B. If proc B gets proc A's place in cpu_present_map, then it registers proc A's VPAs.
3. If proc B is then re-added to the LP, its threads are sharing VPAs with proc A briefly as they come online.

[...]

Applied to powerpc/next.

[1/1] powerpc/pseries: unregister VPA when hot unplugging a CPU

      https://git.kernel.org/powerpc/c/f6aa37c51ec0d053ee34c235bfe0e666618a3baf

cheers