From f1a7941243c102a44e8847e3b94ff4ff3ec56f25 Mon Sep 17 00:00:00 2001 From: Shakeel Butt Date: Mon, 24 Oct 2022 05:28:41 +0000 Subject: mm: convert mm's rss stats into percpu_counter Currently mm_struct maintains rss_stats which are updated on page fault and the unmapping codepaths. For page fault codepath the updates are cached per thread with the batch of TASK_RSS_EVENTS_THRESH which is 64. The reason for caching is performance for multithreaded applications otherwise the rss_stats updates may become hotspot for such applications. However this optimization comes with the cost of error margin in the rss stats. The rss_stats for applications with large number of threads can be very skewed. At worst the error margin is (nr_threads * 64) and we have a lot of applications with 100s of threads, so the error margin can be very high. Internally we had to reduce TASK_RSS_EVENTS_THRESH to 32. Recently we started seeing the unbounded errors for rss_stats for specific applications which use TCP rx0cp. It seems like vm_insert_pages() codepath does not sync rss_stats at all. This patch converts the rss_stats into percpu_counter to convert the error margin from (nr_threads * 64) to approximately (nr_cpus ^ 2). However this conversion enable us to get the accurate stats for situations where accuracy is more important than the cpu cost. This patch does not make such tradeoffs - we can just use percpu_counter_add_local() for the updates and percpu_counter_sum() (or percpu_counter_sync() + percpu_counter_read) for the readers. At the moment the readers are either procfs interface, oom_killer and memory reclaim which I think are not performance critical and should be ok with slow read. However I think we can make that change in a separate patch. Link: https://lkml.kernel.org/r/20221024052841.3291983-1-shakeelb@google.com Signed-off-by: Shakeel Butt Cc: Marek Szyprowski Signed-off-by: Andrew Morton --- kernel/fork.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) (limited to 'kernel/fork.c') diff --git a/kernel/fork.c b/kernel/fork.c index 08969f5aa38d..0fef202434c3 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -753,7 +753,7 @@ static void check_mm(struct mm_struct *mm) "Please make sure 'struct resident_page_types[]' is updated as well"); for (i = 0; i < NR_MM_COUNTERS; i++) { - long x = atomic_long_read(&mm->rss_stat.count[i]); + long x = percpu_counter_sum(&mm->rss_stat[i]); if (unlikely(x)) pr_alert("BUG: Bad rss-counter state mm:%p type:%s val:%ld\n", @@ -779,6 +779,8 @@ static void check_mm(struct mm_struct *mm) */ void __mmdrop(struct mm_struct *mm) { + int i; + BUG_ON(mm == &init_mm); WARN_ON_ONCE(mm == current->mm); WARN_ON_ONCE(mm == current->active_mm); @@ -788,6 +790,9 @@ void __mmdrop(struct mm_struct *mm) check_mm(mm); put_user_ns(mm->user_ns); mm_pasid_drop(mm); + + for (i = 0; i < NR_MM_COUNTERS; i++) + percpu_counter_destroy(&mm->rss_stat[i]); free_mm(mm); } EXPORT_SYMBOL_GPL(__mmdrop); @@ -1107,6 +1112,8 @@ static void mm_init_uprobes_state(struct mm_struct *mm) static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p, struct user_namespace *user_ns) { + int i; + mt_init_flags(&mm->mm_mt, MM_MT_FLAGS); mt_set_external_lock(&mm->mm_mt, &mm->mmap_lock); atomic_set(&mm->mm_users, 1); @@ -1148,10 +1155,17 @@ static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p, if (init_new_context(p, mm)) goto fail_nocontext; + for (i = 0; i < NR_MM_COUNTERS; i++) + if (percpu_counter_init(&mm->rss_stat[i], 0, GFP_KERNEL_ACCOUNT)) + goto fail_pcpu; + mm->user_ns = get_user_ns(user_ns); lru_gen_init_mm(mm); return mm; +fail_pcpu: + while (i > 0) + percpu_counter_destroy(&mm->rss_stat[--i]); fail_nocontext: mm_free_pgd(mm); fail_nopgd: -- cgit v1.2.3