Revision 2bbd00aef0671bfe3c2ca5ba67097246257de125 authored by Johannes Weiner on 26 February 2021, 01:16:47 UTC, committed by Linus Torvalds on 26 February 2021, 17:41:00 UTC
On NOHZ, the periodic vmstat flushers on each CPU can go to sleep and won't wake up until stat changes are detected in the per-cpu deltas of the zone vmstat counters. In commit 75ef71840539 ("mm, vmstat: add infrastructure for per-node vmstats") per-node counters were introduced, and subsequently most stats were moved from the zone to the node level. However, the node counters weren't added to the NOHZ wakeup detection. In theory this can cause per-cpu errors to remain in the user-reported stats indefinitely. In practice this only affects a handful of sub counters (file_mapped, dirty and writeback e.g.) because other page state changes at the node level likely involve a change at the zone level as well (alloc and free, lru ops). Also, nobody has complained. Fix it up for completeness: wake up vmstat refreshing on node changes. Also remove the BUILD_BUG_ONs that assert counter size; we haven't relied on it since we added sizeof() to the range calculation in commit 13c9aaf7fa01 ("mm/vmstat.c: fix NUMA statistics updates"). Link: https://lkml.kernel.org/r/20210202184342.118513-1-hannes@cmpxchg.org Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent a052d4d
percpu-km.c
// SPDX-License-Identifier: GPL-2.0-only
/*
* mm/percpu-km.c - kernel memory based chunk allocation
*
* Copyright (C) 2010 SUSE Linux Products GmbH
* Copyright (C) 2010 Tejun Heo <tj@kernel.org>
*
* Chunks are allocated as a contiguous kernel memory using gfp
* allocation. This is to be used on nommu architectures.
*
* To use percpu-km,
*
* - define CONFIG_NEED_PER_CPU_KM from the arch Kconfig.
*
* - CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK must not be defined. It's
* not compatible with PER_CPU_KM. EMBED_FIRST_CHUNK should work
* fine.
*
* - NUMA is not supported. When setting up the first chunk,
* @cpu_distance_fn should be NULL or report all CPUs to be nearer
* than or at LOCAL_DISTANCE.
*
* - It's best if the chunk size is power of two multiple of
* PAGE_SIZE. Because each chunk is allocated as a contiguous
* kernel memory block using alloc_pages(), memory will be wasted if
* chunk size is not aligned. percpu-km code will whine about it.
*/
#if defined(CONFIG_SMP) && defined(CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK)
#error "contiguous percpu allocation is incompatible with paged first chunk"
#endif
#include <linux/log2.h>
static int pcpu_populate_chunk(struct pcpu_chunk *chunk,
int page_start, int page_end, gfp_t gfp)
{
return 0;
}
static void pcpu_depopulate_chunk(struct pcpu_chunk *chunk,
int page_start, int page_end)
{
/* nada */
}
static struct pcpu_chunk *pcpu_create_chunk(enum pcpu_chunk_type type,
gfp_t gfp)
{
const int nr_pages = pcpu_group_sizes[0] >> PAGE_SHIFT;
struct pcpu_chunk *chunk;
struct page *pages;
unsigned long flags;
int i;
chunk = pcpu_alloc_chunk(type, gfp);
if (!chunk)
return NULL;
pages = alloc_pages(gfp, order_base_2(nr_pages));
if (!pages) {
pcpu_free_chunk(chunk);
return NULL;
}
for (i = 0; i < nr_pages; i++)
pcpu_set_page_chunk(nth_page(pages, i), chunk);
chunk->data = pages;
chunk->base_addr = page_address(pages);
spin_lock_irqsave(&pcpu_lock, flags);
pcpu_chunk_populated(chunk, 0, nr_pages);
spin_unlock_irqrestore(&pcpu_lock, flags);
pcpu_stats_chunk_alloc();
trace_percpu_create_chunk(chunk->base_addr);
return chunk;
}
static void pcpu_destroy_chunk(struct pcpu_chunk *chunk)
{
const int nr_pages = pcpu_group_sizes[0] >> PAGE_SHIFT;
if (!chunk)
return;
pcpu_stats_chunk_dealloc();
trace_percpu_destroy_chunk(chunk->base_addr);
if (chunk->data)
__free_pages(chunk->data, order_base_2(nr_pages));
pcpu_free_chunk(chunk);
}
static struct page *pcpu_addr_to_page(void *addr)
{
return virt_to_page(addr);
}
static int __init pcpu_verify_alloc_info(const struct pcpu_alloc_info *ai)
{
size_t nr_pages, alloc_pages;
/* all units must be in a single group */
if (ai->nr_groups != 1) {
pr_crit("can't handle more than one group\n");
return -EINVAL;
}
nr_pages = (ai->groups[0].nr_units * ai->unit_size) >> PAGE_SHIFT;
alloc_pages = roundup_pow_of_two(nr_pages);
if (alloc_pages > nr_pages)
pr_warn("wasting %zu pages per chunk\n",
alloc_pages - nr_pages);
return 0;
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...