Skip to contentSkip to conversation

sched/numa: Skip VMA scanning on memory pinned to one NUMA node via cpuset.mem

PATCH Superseded Libo Chen 2 participants 9 messages Apr 24, 2025

Subsystems (4)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Libo Chen

author
#0 lore raw

v4 cover sched/numa: Skip VMA scanning on memory pinned to one NUMA node via cpuset.mem

v1->v2:
1. add perf improvment numbers in commit log. Yet to find perf diff on will-it-scale, so not included here. Plan to run more workloads.
2. add tracepoint.
3. To peterz's comment, this will make it impossible to attract tasks to those memory just like other VMA skippings. This is the current implementation, I think we can improve that in the future, but at the moment it's probabaly better to keep it consistent.

v2->v3:
1. add enable_cpuset() based on Mel's suggestion but again I think it's redundant
2. print out nodemask with %*p.. format in the tracepoint

v3->v4:
1. fix an unsafe dereference of a pointer to content not on ring buffer, namely mem_allowed_ptr in the tracepoint.

Libo Chen (2):
sched/numa: Skip VMA scanning on memory pinned to one NUMA node via
cpuset.mems
sched/numa: Add tracepoint that tracks the skipping of numa balancing
due to cpuset memory pinning

 include/trace/events/sched.h | 31 +++++++++++++++++++++++++++++++
 kernel/sched/fair.c          |  9 +++++++++
 2 files changed, 40 insertions(+)
Signature · 1 line
2.43.5
author
#1 lore raw

PATCH v4 1/2 sched/numa: Skip VMA scanning on memory pinned to one NUMA node via cpuset.mems

When the memory of the current task is pinned to one NUMA node by cgroup, there is no point in continuing the rest of VMA scanning and hinting page faults as they will just be overhead. With this change, there will be no more unnecessary PTE updates or page faults in this scenario.

We have seen up to a 6x improvement on a typical java workload running on VMs with memory and CPU pinned to one NUMA node via cpuset in a two-socket AARCH64 system. With the same pinning, on a 18-cores-per-socket Intel platform, we have seen 20% improvment in a microbench that creates a 30-vCPU selftest KVM guest with 4GB memory, where each vCPU reads 4KB pages in a fixed number of loops.

---
 kernel/sched/fair.c | 7 +++++++
 1 file changed, 7 insertions(+)
kernel/sched/fair.c +7 -0
index e43993a4e580..c9903b1b3948 100644
@@ -3329,6 +3329,13 @@ static void task_numa_work(struct callback_head *work)
 	if (p->flags & PF_EXITING)
 		return;
 
+	/*
+	 * Memory is pinned to only one NUMA node via cpuset.mems, naturally
+	 * no page can be migrated.
+	 */
+	if (cpusets_enabled() && nodes_weight(cpuset_current_mems_allowed) == 1)
+		return;
+
 	if (!mm->numa_next_scan) {
 		mm->numa_next_scan = now +
 			msecs_to_jiffies(sysctl_numa_balancing_scan_delay);
Signature · 1 line
2.43.5
author
#2 lore raw

PATCH v4 2/2 sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

In reply to Libo Chen

Unlike sched_skip_vma_numa tracepoint which tracks skipped VMAs, this tracks the task subjected to cpuset.mems pinning and prints out its allowed memory node mask.

---
 include/trace/events/sched.h | 31 +++++++++++++++++++++++++++++++
 kernel/sched/fair.c          |  4 +++-
 2 files changed, 34 insertions(+), 1 deletion(-)
include/trace/events/sched.h +31 -0
index 8994e97d86c1..91f9dc177dad 100644
@@ -745,6 +745,37 @@ TRACE_EVENT(sched_skip_vma_numa,
 		  __entry->vm_end,
 		  __print_symbolic(__entry->reason, NUMAB_SKIP_REASON))
 );
+
+TRACE_EVENT(sched_skip_cpuset_numa,
+
+	TP_PROTO(struct task_struct *tsk, nodemask_t *mem_allowed_ptr),
+
+	TP_ARGS(tsk, mem_allowed_ptr),
+
+	TP_STRUCT__entry(
+		__array( char,		comm,		TASK_COMM_LEN		)
+		__field( pid_t,		pid					)
+		__field( pid_t,		tgid					)
+		__field( pid_t,		ngid					)
+		__array( unsigned long, mem_allowed, BITS_TO_LONGS(MAX_NUMNODES))
+	),
+
+	TP_fast_assign(
+		memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN);
+		__entry->pid		 = task_pid_nr(tsk);
+		__entry->tgid		 = task_tgid_nr(tsk);
+		__entry->ngid		 = task_numa_group_id(tsk);
+		memcpy(__entry->mem_allowed, mem_allowed_ptr->bits,
+		       sizeof(__entry->mem_allowed));
+	),
+
+	TP_printk("comm=%s pid=%d tgid=%d ngid=%d mem_nodes_allowed=%*pbl",
+		  __entry->comm,
+		  __entry->pid,
+		  __entry->tgid,
+		  __entry->ngid,
+		  MAX_NUMNODES, __entry->mem_allowed)
+);
 #endif /* CONFIG_NUMA_BALANCING */
 
 /*
kernel/sched/fair.c +3 -1
index c9903b1b3948..cc892961ce15 100644
@@ -3333,8 +3333,10 @@ static void task_numa_work(struct callback_head *work)
 	 * Memory is pinned to only one NUMA node via cpuset.mems, naturally
 	 * no page can be migrated.
 	 */
-	if (cpusets_enabled() && nodes_weight(cpuset_current_mems_allowed) == 1)
+	if (cpusets_enabled() && nodes_weight(cpuset_current_mems_allowed) == 1) {
+		trace_sched_skip_cpuset_numa(current, &cpuset_current_mems_allowed);
 		return;
+	}
 
 	if (!mm->numa_next_scan) {
 		mm->numa_next_scan = now +
Signature · 1 line
2.43.5
maintainer
#3 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 17 lines · On Wed, 23 Apr 2025 17:01:46 -0700 Libo Chen <[email protected]> wrote:

+++ b/include/trace/events/sched.h

include/trace/events/sched.h +12 -0
@@ -745,6 +745,37 @@ TRACE_EVENT(sched_skip_vma_numa,
 		  __entry->vm_end,
 		  __print_symbolic(__entry->reason, NUMAB_SKIP_REASON))
 );
+
+TRACE_EVENT(sched_skip_cpuset_numa,
+
+	TP_PROTO(struct task_struct *tsk, nodemask_t *mem_allowed_ptr),
+
+	TP_ARGS(tsk, mem_allowed_ptr),
+
+	TP_STRUCT__entry(
+		__array( char,		comm,		TASK_COMM_LEN		)
+		__field( pid_t,		pid					)
+		__field( pid_t,		tgid					)
+		__field( pid_t,		ngid					)

include/trace/events/sched.h TRACE_EVENT(sched_skip_vma_numa, · line 769

Quoted 10 lines
include/trace/events/sched.h +10 -0
+		__array( unsigned long, mem_allowed, BITS_TO_LONGS(MAX_NUMNODES))
+	),
+
+	TP_fast_assign(
+		memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN);
+		__entry->pid		 = task_pid_nr(tsk);
+		__entry->tgid		 = task_tgid_nr(tsk);
+		__entry->ngid		 = task_numa_group_id(tsk);
+		memcpy(__entry->mem_allowed, mem_allowed_ptr->bits,
+		       sizeof(__entry->mem_allowed));

Is mem_allowed->bits guaranteed to be the size of BITS_TO_LONGS(MAX_NUM_NODES) in size? If not, then memcpy will read beyond that size.

-- Steve

Quoted 9 lines
include/trace/events/sched.h +9 -0
+	),
+
+	TP_printk("comm=%s pid=%d tgid=%d ngid=%d mem_nodes_allowed=%*pbl",
+		  __entry->comm,
+		  __entry->pid,
+		  __entry->tgid,
+		  __entry->ngid,
+		  MAX_NUMNODES, __entry->mem_allowed)
+);
author
#4 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 25 lines · Steven Rostedt
Quoted 22 lines · On Wed, 23 Apr 2025 17:01:46 -0700 Libo Chen <[email protected]> wrote:

+++ b/include/trace/events/sched.h

include/trace/events/sched.h +17 -0
@@ -745,6 +745,37 @@ TRACE_EVENT(sched_skip_vma_numa,
 		  __entry->vm_end,
 		  __print_symbolic(__entry->reason, NUMAB_SKIP_REASON))
 );
+
+TRACE_EVENT(sched_skip_cpuset_numa,
+
+	TP_PROTO(struct task_struct *tsk, nodemask_t *mem_allowed_ptr),
+
+	TP_ARGS(tsk, mem_allowed_ptr),
+
+	TP_STRUCT__entry(
+		__array( char,		comm,		TASK_COMM_LEN		)
+		__field( pid_t,		pid					)
+		__field( pid_t,		tgid					)
+		__field( pid_t,		ngid					)
+		__array( unsigned long, mem_allowed, BITS_TO_LONGS(MAX_NUMNODES))
+	),
+
+	TP_fast_assign(
+		memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN);
Quoted 9 lines

include/trace/events/sched.h TRACE_EVENT(sched_skip_vma_numa, · line 769

Quoted 5 lines
include/trace/events/sched.h +5 -0
+		__entry->pid		 = task_pid_nr(tsk);
+		__entry->tgid		 = task_tgid_nr(tsk);
+		__entry->ngid		 = task_numa_group_id(tsk);
+		memcpy(__entry->mem_allowed, mem_allowed_ptr->bits,
+		       sizeof(__entry->mem_allowed));

Is mem_allowed->bits guaranteed to be the size of BITS_TO_LONGS(MAX_NUM_NODES) in size? If not, then memcpy will read beyond that size.

Yes, evidence can be found in the definitions of nodemask_t and DECLARE_BITMAP:

// include/linux/nodemask_types.h
typedef struct { DECLARE_BITMAP(bits, MAX_NUMNODES); } nodemask_t;

// include/linux/types.h
#define DECLARE_BITMAP(name,bits) \
unsigned long name[BITS_TO_LONGS(bits)]

Thanks,
Libo

Quoted 13 lines

-- Steve

Quoted 9 lines
include/trace/events/sched.h +9 -0
+	),
+
+	TP_printk("comm=%s pid=%d tgid=%d ngid=%d mem_nodes_allowed=%*pbl",
+		  __entry->comm,
+		  __entry->pid,
+		  __entry->tgid,
+		  __entry->ngid,
+		  MAX_NUMNODES, __entry->mem_allowed)
+);
maintainer
#5 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 12 lines · On Wed, 23 Apr 2025 17:36:30 -0700 Libo Chen <[email protected]> wrote:
Quoted 11 lines

include/trace/events/sched.h TRACE_EVENT(sched_skip_vma_numa, · line 769

Quoted 7 lines
include/trace/events/sched.h +7 -0
+	TP_fast_assign(
+		memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN);
+		__entry->pid		 = task_pid_nr(tsk);
+		__entry->tgid		 = task_tgid_nr(tsk);
+		__entry->ngid		 = task_numa_group_id(tsk);
+		memcpy(__entry->mem_allowed, mem_allowed_ptr->bits,
+		       sizeof(__entry->mem_allowed));  

Is mem_allowed->bits guaranteed to be the size of BITS_TO_LONGS(MAX_NUM_NODES) in size? If not, then memcpy will read beyond that size.

Quoted 9 lines

Yes, evidence can be found in the definitions of nodemask_t and DECLARE_BITMAP:

// include/linux/nodemask_types.h
typedef struct { DECLARE_BITMAP(bits, MAX_NUMNODES); } nodemask_t;

// include/linux/types.h
#define DECLARE_BITMAP(name,bits) \
unsigned long name[BITS_TO_LONGS(bits)]

Hmm, I wonder then if we should add in TP_fast_assign():

	BUILD_BUG_ON(sizeof(nodemask_t) != BITS_TO_LONGS(MAX_NUM_NODES) * sizeof(long));

-- Steve

author
#6 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 20 lines · Steven Rostedt
Quoted 17 lines · On Wed, 23 Apr 2025 17:36:30 -0700 Libo Chen <[email protected]> wrote:
Quoted 11 lines

include/trace/events/sched.h TRACE_EVENT(sched_skip_vma_numa, · line 769

Quoted 7 lines
include/trace/events/sched.h +7 -0
+	TP_fast_assign(
+		memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN);
+		__entry->pid		 = task_pid_nr(tsk);
+		__entry->tgid		 = task_tgid_nr(tsk);
+		__entry->ngid		 = task_numa_group_id(tsk);
+		memcpy(__entry->mem_allowed, mem_allowed_ptr->bits,
+		       sizeof(__entry->mem_allowed));  

Is mem_allowed->bits guaranteed to be the size of BITS_TO_LONGS(MAX_NUM_NODES) in size? If not, then memcpy will read beyond that size.

Yes, evidence can be found in the definitions of nodemask_t and DECLARE_BITMAP:

// include/linux/nodemask_types.h
typedef struct { DECLARE_BITMAP(bits, MAX_NUMNODES); } nodemask_t;

Quoted 9 lines
Quoted 4 lines

// include/linux/types.h
#define DECLARE_BITMAP(name,bits) \
unsigned long name[BITS_TO_LONGS(bits)]

Hmm, I wonder then if we should add in TP_fast_assign():

	BUILD_BUG_ON(sizeof(nodemask_t) != BITS_TO_LONGS(MAX_NUM_NODES) * sizeof(long));

to guard against potential changes in nodemask_t definition?

Quoted 1 line

-- Steve

maintainer
#7 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 6 lines · On Wed, 23 Apr 2025 18:12:55 -0700 Libo Chen <[email protected]> wrote:
Quoted 4 lines

Hmm, I wonder then if we should add in TP_fast_assign():

	BUILD_BUG_ON(sizeof(nodemask_t) != BITS_TO_LONGS(MAX_NUM_NODES) * sizeof(long));

to guard against potential changes in nodemask_t definition?

Correct.

Whenever there's an implicit dependency like this, where if something were to change it can cause a bug in the kernel, it's always better to have a build time check to catch it before it becomes an issue.

-- Steve

author
#8 lore raw

Re: [PATCH v4 2/2] sched/numa: Add tracepoint that tracks the skipping of numa balancing due to cpuset memory pinning

Quoted 8 lines · Steven Rostedt
Quoted 5 lines · On Wed, 23 Apr 2025 18:12:55 -0700 Libo Chen <[email protected]> wrote:
Quoted 4 lines

Hmm, I wonder then if we should add in TP_fast_assign():

	BUILD_BUG_ON(sizeof(nodemask_t) != BITS_TO_LONGS(MAX_NUM_NODES) * sizeof(long));
Quoted 8 lines
Quoted 1 line

to guard against potential changes in nodemask_t definition?

Correct.

Whenever there's an implicit dependency like this, where if something were to change it can cause a bug in the kernel, it's always better to have a build time check to catch it before it becomes an issue.

Okay that's reasonable. I will add it~