sched/numa: Complete scanning of partial VMAs regardless of PID activity

NUMA Balancing skips VMAs when the current task has not trapped a NUMA
fault within the VMA. If the VMA is skipped then mm->numa_scan_offset
advances and a task that is trapping faults within the VMA may never
fully update PTEs within the VMA.

Force tasks to update PTEs for partially scanned PTEs. The VMA will
be tagged for NUMA hints by some task but this removes some of the
benefit of tracking PID activity within a VMA. A follow-on patch
will mitigate this problem.

The test cases and machines evaluated did not trigger the corner case so
the performance results are neutral with only small changes within the
noise from normal test-to-test variance. However, the next patch makes
the corner case easier to trigger.

Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Raghavendra K T <raghavendra.kt@amd.com>
Link: https://lore.kernel.org/r/20231010083143.19593-6-mgorman@techsingularity.net

diff --git a/include/linux/sched/numa_balancing.h b/include/linux/sched/numa_balancing.h
index c127a1509e2faad8ca856e9866791500dacc2e4b..7dcc0bdfddbbf1dfea8bf2c1f35e7745c107230f 100644 (file)
--- a/include/linux/sched/numa_balancing.h
+++ b/include/linux/sched/numa_balancing.h
@@ -21,6 +21,7 @@ enum numa_vmaskip_reason {
        NUMAB_SKIP_INACCESSIBLE,
        NUMAB_SKIP_SCAN_DELAY,
        NUMAB_SKIP_PID_INACTIVE,
+       NUMAB_SKIP_IGNORE_PID,
 };
 
 #ifdef CONFIG_NUMA_BALANCING

diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h
index d82a04d6a1bc723345ea203a486867090434af2c..bfc07c10541a109d44fd6aa93526f50c7f486a60 100644 (file)
--- a/include/trace/events/sched.h
+++ b/include/trace/events/sched.h
@@ -670,7 +670,8 @@ DEFINE_EVENT(sched_numa_pair_template, sched_swap_numa,
        EM( NUMAB_SKIP_SHARED_RO,               "shared_ro" )   \
        EM( NUMAB_SKIP_INACCESSIBLE,            "inaccessible" )        \
        EM( NUMAB_SKIP_SCAN_DELAY,              "scan_delay" )  \
-       EMe(NUMAB_SKIP_PID_INACTIVE,            "pid_inactive" )
+       EM( NUMAB_SKIP_PID_INACTIVE,            "pid_inactive" )        \
+       EMe(NUMAB_SKIP_IGNORE_PID,              "ignore_pid_inactive" )
 
 /* Redefine for export. */
 #undef EM

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index ce36969625bdb5dfbf6aeaabee8fb44662283836..ab79013f6e9120818ec7d0116a168ff56fa5e85b 100644 (file)
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -3113,7 +3113,7 @@ static void reset_ptenuma_scan(struct task_struct *p)
        p->mm->numa_scan_offset = 0;
 }
 
-static bool vma_is_accessed(struct vm_area_struct *vma)
+static bool vma_is_accessed(struct mm_struct *mm, struct vm_area_struct *vma)
 {
        unsigned long pids;
        /*
@@ -3126,7 +3126,19 @@ static bool vma_is_accessed(struct vm_area_struct *vma)
                return true;
 
        pids = vma->numab_state->pids_active[0] | vma->numab_state->pids_active[1];
-       return test_bit(hash_32(current->pid, ilog2(BITS_PER_LONG)), &pids);
+       if (test_bit(hash_32(current->pid, ilog2(BITS_PER_LONG)), &pids))
+               return true;
+
+       /*
+        * Complete a scan that has already started regardless of PID access, or
+        * some VMAs may never be scanned in multi-threaded applications:
+        */
+       if (mm->numa_scan_offset > vma->vm_start) {
+               trace_sched_skip_vma_numa(mm, vma, NUMAB_SKIP_IGNORE_PID);
+               return true;
+       }
+
+       return false;
 }
 
 #define VMA_PID_RESET_PERIOD (4 * sysctl_numa_balancing_scan_delay)
@@ -3270,7 +3282,7 @@ static void task_numa_work(struct callback_head *work)
                }
 
                /* Do not scan the VMA if task has not accessed */
-               if (!vma_is_accessed(vma)) {
+               if (!vma_is_accessed(mm, vma)) {
                        trace_sched_skip_vma_numa(mm, vma, NUMAB_SKIP_PID_INACTIVE);
                        continue;
                }