mm/migrate: fix CPUHP state to update node demotion order
Subsystems (3)
PATCH 1/1 mm/migrate: fix CPUHP state to update node demotion order
The node demotion order needs to be updated during CPU hotplug. Because whether a NUMA node has CPU may influence the demotion order. The update function should be called during CPU online/offline after the node_states[N_CPU] has been updated. That is done in CPUHP_AP_ONLINE_DYN during CPU online and in CPUHP_MM_VMSTAT_DEAD during CPU offline. But in commit 884a6e5d1f93 ("mm/migrate: update node demotion order on hotplug events"), the function to update node demotion order is called in CPUHP_AP_ONLINE_DYN during CPU online/offline. This doesn't satisfy the order requirement. So in this patch, we added CPUHP_AP_MM_DEMOTION_ONLINE and CPUHP_MM_DEMOTION_DEAD to be called after CPUHP_AP_ONLINE_DYN and CPUHP_MM_VMSTAT_DEAD during CPU online and offline, and register the update function on them.
Changes:
v2:
- Revise state name to follow the naming convention per Thomas' comments.
- Use cpuhp_setup_state() to initialize the initial order per Mika's comments.
--- include/linux/cpuhotplug.h | 4 ++++ mm/migrate.c | 8 +++++--- 2 files changed, 9 insertions(+), 3 deletions(-)
include/linux/cpuhotplug.h +4 -0
@@ -72,6 +72,8 @@ enum cpuhp_state {CPUHP_SLUB_DEAD ,CPUHP_DEBUG_OBJ_DEAD ,CPUHP_MM_WRITEBACK_DEAD , +/* Must be after CPUHP_MM_VMSTAT_DEAD */ +CPUHP_MM_DEMOTION_DEAD ,CPUHP_MM_VMSTAT_DEAD ,CPUHP_SOFTIRQ_DEAD , CPUHP_NET_MVNETA_DEAD,
@@ -240,6 +242,8 @@ enum cpuhp_state {CPUHP_AP_BASE_CACHEINFO_ONLINE ,CPUHP_AP_ONLINE_DYN ,CPUHP_AP_ONLINE_DYN_END =CPUHP_AP_ONLINE_DYN +30 , +/* Must be after CPUHP_AP_ONLINE_DYN for node_states[N_CPU] update */ +CPUHP_AP_MM_DEMOTION_ONLINE ,CPUHP_AP_X86_HPET_ONLINE ,CPUHP_AP_X86_KVM_CLK_ONLINE , CPUHP_AP_DTPM_CPU_ONLINE,
mm/migrate.c +5 -3
@@ -3284,9 +3284,8 @@ static int __init migrate_on_reclaim_init(void) {int ret; - ret =cpuhp_setup_state (CPUHP_AP_ONLINE_DYN ,"migrate on reclaim" , - migration_online_cpu, - migration_offline_cpu); + ret =cpuhp_setup_state_nocalls (CPUHP_MM_DEMOTION_DEAD ,"mm/demotion:offline" , +NULL , migration_offline_cpu);/* * In the unlikely case that this fails, the automatic * migration targets may become suboptimal for nodes
@@ -3294,6 +3293,9 @@ static int __init migrate_on_reclaim_init(void)* rare case, do not bother trying to do anything special. */ WARN_ON (ret <0 ); + ret =cpuhp_setup_state (CPUHP_AP_MM_DEMOTION_ONLINE ,"mm/demotion:online" , + migration_online_cpu,NULL ); +WARN_ON (ret <0 );hotplug_memory_notifier (migrate_on_reclaim_callback,100 );return 0 ;
Signature · 1 line
2.30.2
Andrew Morton
maintainerQuoted 9 lines · On Mon, 27 Sep 2021 16:11:00 +0800 Huang Ying <[email protected]> wrote:
The node demotion order needs to be updated during CPU hotplug. Because whether a NUMA node has CPU may influence the demotion order. The update function should be called during CPU online/offline after the node_states[N_CPU] has been updated. That is done in
CPUHP_AP_ONLINE_DYNduring CPU online and inCPUHP_MM_VMSTAT_DEADduring CPU offline. But in commit884a6e5d1f93("mm/migrate: update node demotion order on hotplug events"), the function to update node demotion order is called inCPUHP_AP_ONLINE_DYNduring CPU online/offline. This doesn't satisfy the order requirement.
What are the user-visible runtime effects of this error?
Quoted 21 lines
So in
this patch, we addedCPUHP_AP_MM_DEMOTION_ONLINEandCPUHP_MM_DEMOTION_DEADto be called afterCPUHP_AP_ONLINE_DYNandCPUHP_MM_VMSTAT_DEADduring CPU online and offline, and register the update function on them.Fixes 884a6e5d1f93 ("mm/migrate: update node demotion order on hotplug events")Signed-off-by "Huang, Ying" <[email protected]>Cc Thomas Gleixner <[email protected]>Cc Dave Hansen <[email protected]>Cc Yang Shi <[email protected]>Cc Zi Yan <[email protected]>Cc Michal Hocko <[email protected]>Cc Wei Xu <[email protected]>Cc Oscar Salvador <[email protected]>Cc David Rientjes <[email protected]>Cc Dan Williams <[email protected]>Cc David Hildenbrand <[email protected]>Cc Greg Thelen <[email protected]>Cc Keith Busch <[email protected]>
Quoted 12 lines · Andrew Morton
Quoted 9 lines · On Mon, 27 Sep 2021 16:11:00 +0800 Huang Ying <[email protected]> wrote:
The node demotion order needs to be updated during CPU hotplug. Because whether a NUMA node has CPU may influence the demotion order. The update function should be called during CPU online/offline after the node_states[N_CPU] has been updated. That is done in
CPUHP_AP_ONLINE_DYNduring CPU online and inCPUHP_MM_VMSTAT_DEADduring CPU offline. But in commit884a6e5d1f93("mm/migrate: update node demotion order on hotplug events"), the function to update node demotion order is called inCPUHP_AP_ONLINE_DYNduring CPU online/offline. This doesn't satisfy the order requirement.
Quoted 1 line
What are the user-visible runtime effects of this error?
For example, there are 4 CPUs (P0, P1, P2, P3) in 2 sockets (P0, P1 in S0 and P2, P3 in S1), the demotion order is
- S0 -> NUMA_NO_NODE
- S1 -> NUMA_NO_NODE
After P2 and P3 is offlined, because S1 has no CPU now, the demotion order should have been changed to
- S0 -> S1
- S1 -> NO_NODE
but it isn't changed, because the order updating callback for CPU hotplug doesn't see the new nodemask. Now, if P1 is offlined, the demotion order is changed to the expected order as above.
I will update the patch description after adding the above description.
Best Regards,
Huang, Ying
Quoted 21 lines
Quoted 21 lines
So in
this patch, we addedCPUHP_AP_MM_DEMOTION_ONLINEandCPUHP_MM_DEMOTION_DEADto be called afterCPUHP_AP_ONLINE_DYNandCPUHP_MM_VMSTAT_DEADduring CPU online and offline, and register the update function on them.Fixes 884a6e5d1f93 ("mm/migrate: update node demotion order on hotplug events")Signed-off-by "Huang, Ying" <[email protected]>Cc Thomas Gleixner <[email protected]>Cc Dave Hansen <[email protected]>Cc Yang Shi <[email protected]>Cc Zi Yan <[email protected]>Cc Michal Hocko <[email protected]>Cc Wei Xu <[email protected]>Cc Oscar Salvador <[email protected]>Cc David Rientjes <[email protected]>Cc Dan Williams <[email protected]>Cc David Hildenbrand <[email protected]>Cc Greg Thelen <[email protected]>Cc Keith Busch <[email protected]>