The patch titled
Subject: sched/numa: implement access PID reset logic
has been added to the -mm mm-unstable branch. Its filename is
sched-numa-implement-access-pid-reset-logic.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/sched-numa-implement-access-pid-reset-logic.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Raghavendra K T <raghavendra.kt@xxxxxxx>
Subject: sched/numa: implement access PID reset logic
Date: Wed, 1 Mar 2023 17:49:02 +0530
This helps to ensure that only recently accessed PIDs scan the VMAs.
Current implementation: (idea supported by PeterZ)
1. Accessing PID information is maintained in two windows.
access_pids[1] being newest.
2. Reset old access PID info i.e. access_pid[0] every (4 *
sysctl_numa_balancing_scan_delay) interval after initial scan delay
period expires.
The above interval seemed to be experimentally optimum since it avoids
frequent reset of access info as well as helps clearing the old access
info regularly. The reset logic is implemented in scan path.
Link: https://lkml.kernel.org/r/f7a675f66d1442d048b4216b2baf94515012c405.1677672277.git.raghavendra.kt@xxxxxxx
Signed-off-by: Raghavendra K T <raghavendra.kt@xxxxxxx>
Suggested-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Cc: Bharata B Rao <bharata@xxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Cc: Disha Talreja <dishaa.talreja@xxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Mike Rapoport <rppt@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
--- a/include/linux/mm.h~sched-numa-implement-access-pid-reset-logic
+++ a/include/linux/mm.h
@@ -1672,8 +1672,8 @@ static inline void vma_set_access_pid_bi
unsigned int pid_bit;
pid_bit = current->pid % BITS_PER_LONG;
- if (vma->numab_state && !test_bit(pid_bit, &vma->numab_state->access_pids)) {
- __set_bit(pid_bit, &vma->numab_state->access_pids);
+ if (vma->numab_state && !test_bit(pid_bit, &vma->numab_state->access_pids[1])) {
+ __set_bit(pid_bit, &vma->numab_state->access_pids[1]);
}
}
#else /* !CONFIG_NUMA_BALANCING */
--- a/include/linux/mm_types.h~sched-numa-implement-access-pid-reset-logic
+++ a/include/linux/mm_types.h
@@ -477,7 +477,8 @@ struct vma_lock {
struct vma_numab_state {
unsigned long next_scan;
- unsigned long access_pids;
+ unsigned long next_pid_reset;
+ unsigned long access_pids[2];
};
/*
--- a/kernel/sched/fair.c~sched-numa-implement-access-pid-reset-logic
+++ a/kernel/sched/fair.c
@@ -2930,6 +2930,7 @@ static void reset_ptenuma_scan(struct ta
static bool vma_is_accessed(struct vm_area_struct *vma)
{
+ unsigned long pids;
/*
* Allow unconditional access first two times, so that all the (pages)
* of VMAs get prot_none fault introduced irrespective of accesses.
@@ -2939,10 +2940,12 @@ static bool vma_is_accessed(struct vm_ar
if (READ_ONCE(current->mm->numa_scan_seq) < 2)
return true;
- return test_bit(current->pid % BITS_PER_LONG,
- &vma->numab_state->access_pids);
+ pids = vma->numab_state->access_pids[0] | vma->numab_state->access_pids[1];
+ return test_bit(current->pid % BITS_PER_LONG, &pids);
}
+#define VMA_PID_RESET_PERIOD (4 * sysctl_numa_balancing_scan_delay)
+
/*
* The expensive part of numa migration is done from task_work context.
* Triggered from task_tick_numa().
@@ -3051,6 +3054,10 @@ static void task_numa_work(struct callba
vma->numab_state->next_scan = now +
msecs_to_jiffies(sysctl_numa_balancing_scan_delay);
+
+ /* Reset happens after 4 times scan delay of scan start */
+ vma->numab_state->next_pid_reset = vma->numab_state->next_scan +
+ msecs_to_jiffies(VMA_PID_RESET_PERIOD);
}
/*
@@ -3065,6 +3072,18 @@ static void task_numa_work(struct callba
if (!vma_is_accessed(vma))
continue;
+ /*
+ * RESET access PIDs regularly for old VMAs. Resetting after checking
+ * vma for recent access to avoid clearing PID info before access..
+ */
+ if (mm->numa_scan_seq &&
+ time_after(jiffies, vma->numab_state->next_pid_reset)) {
+ vma->numab_state->next_pid_reset = vma->numab_state->next_pid_reset +
+ msecs_to_jiffies(VMA_PID_RESET_PERIOD);
+ vma->numab_state->access_pids[0] = READ_ONCE(vma->numab_state->access_pids[1]);
+ vma->numab_state->access_pids[1] = 0;
+ }
+
do {
start = max(start, vma->vm_start);
end = ALIGN(start + (pages << PAGE_SHIFT), HPAGE_SIZE);
_
Patches currently in -mm which might be from raghavendra.kt@xxxxxxx are
sched-numa-enhance-vma-scanning-logic.patch
sched-numa-implement-access-pid-reset-logic.patch
sched-numa-use-hash_32-to-mix-up-pids-accessing-vma.patch
