The idea behind the cache is to save get_pageblock_migratetype()
lookups during bulk freeing. A microbenchmark suggests this isn't
helping, though. The pcp migratetype can get stale, which means that
bulk freeing has an extra branch to check if the pageblock was
isolated while on the pcp.
While the variance overlaps, the cache write and the branch seem to
make this a net negative. The following test allocates and frees
batches of 10,000 pages (~3x the pcp high marks to trigger flushing):
Before:
8,668.48 msec task-clock # 99.735 CPUs utilized ( +- 2.90% )
19 context-switches # 4.341 /sec ( +- 3.24% )
0 cpu-migrations # 0.000 /sec
17,440 page-faults # 3.984 K/sec ( +- 2.90% )
41,758,692,473 cycles # 9.541 GHz ( +- 2.90% )
126,201,294,231 instructions # 5.98 insn per cycle ( +- 2.90% )
25,348,098,335 branches # 5.791 G/sec ( +- 2.90% )
33,436,921 branch-misses # 0.26% of all branches ( +- 2.90% )
0.0869148 +- 0.0000302 seconds time elapsed ( +- 0.03% )
After:
8,444.81 msec task-clock # 99.726 CPUs utilized ( +- 2.90% )
22 context-switches # 5.160 /sec ( +- 3.23% )
0 cpu-migrations # 0.000 /sec
17,443 page-faults # 4.091 K/sec ( +- 2.90% )
40,616,738,355 cycles # 9.527 GHz ( +- 2.90% )
126,383,351,792 instructions # 6.16 insn per cycle ( +- 2.90% )
25,224,985,153 branches # 5.917 G/sec ( +- 2.90% )
32,236,793 branch-misses # 0.25% of all branches ( +- 2.90% )
0.0846799 +- 0.0000412 seconds time elapsed ( +- 0.05% )
A side effect is that this also ensures that pages whose pageblock
gets stolen while on the pcplist end up on the right freelist and we
don't perform potentially type-incompatible buddy merges (or skip
merges when we shouldn't), which is likely beneficial to long-term
fragmentation management, although the effects would be harder to
measure. Settle for simpler and faster code as justification here.
v2:
- remove erroneous leftover VM_BUG_ON in pcp bulk freeing (Mike)
Acked-by: Zi Yan <ziy@xxxxxxxxxx>
Reviewed-by: Vlastimil Babka <vbabka@xxxxxxx>
Acked-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Tested-by: "Huang, Ying" <ying.huang@xxxxxxxxx>
Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
---
mm/page_alloc.c | 66 +++++++++++--------------------------------------
1 file changed, 14 insertions(+), 52 deletions(-)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 16241906a368..9665a2db3b53 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -206,24 +206,6 @@ EXPORT_SYMBOL(node_states);
gfp_t gfp_allowed_mask __read_mostly = GFP_BOOT_MASK;
-/*
- * A cached value of the page's pageblock's migratetype, used when the page is
- * put on a pcplist. Used to avoid the pageblock migratetype lookup when
- * freeing from pcplists in most cases, at the cost of possibly becoming stale.
- * Also the migratetype set in the page does not necessarily match the pcplist
- * index, e.g. page might have MIGRATE_CMA set but be on a pcplist with any
- * other index - this ensures that it will be put on the correct CMA freelist.
- */
-static inline int get_pcppage_migratetype(struct page *page)
-{
- return page->index;
-}
-
-static inline void set_pcppage_migratetype(struct page *page, int migratetype)
-{
- page->index = migratetype;
-}
-
#ifdef CONFIG_HUGETLB_PAGE_SIZE_VARIABLE
unsigned int pageblock_order __read_mostly;
#endif
@@ -1191,7 +1173,6 @@ static void free_pcppages_bulk(struct zone *zone, int count,
{
unsigned long flags;
unsigned int order;
- bool isolated_pageblocks;
struct page *page;
/*
@@ -1204,7 +1185,6 @@ static void free_pcppages_bulk(struct zone *zone, int count,
pindex = pindex - 1;
spin_lock_irqsave(&zone->lock, flags);
- isolated_pageblocks = has_isolate_pageblock(zone);
while (count > 0) {
struct list_head *list;
@@ -1220,23 +1200,19 @@ static void free_pcppages_bulk(struct zone *zone, int count,
order = pindex_to_order(pindex);
nr_pages = 1 << order;
do {
+ unsigned long pfn;
int mt;
page = list_last_entry(list, struct page, pcp_list);
- mt = get_pcppage_migratetype(page);
+ pfn = page_to_pfn(page);
+ mt = get_pfnblock_migratetype(page, pfn);
/* must delete to avoid corrupting pcp list */
list_del(&page->pcp_list);
count -= nr_pages;
pcp->count -= nr_pages;
- /* MIGRATE_ISOLATE page should not go to pcplists */
- VM_BUG_ON_PAGE(is_migrate_isolate(mt), page);
- /* Pageblock could have been isolated meanwhile */
- if (unlikely(isolated_pageblocks))
- mt = get_pageblock_migratetype(page);
-
- __free_one_page(page, page_to_pfn(page), zone, order, mt, FPI_NONE);
+ __free_one_page(page, pfn, zone, order, mt, FPI_NONE);
trace_mm_page_pcpu_drain(page, order, mt);
} while (count > 0 && !list_empty(list));
}
@@ -1575,7 +1551,6 @@ struct page *__rmqueue_smallest(struct zone *zone, unsigned int order,
continue;
del_page_from_free_list(page, zone, current_order);
expand(zone, page, order, current_order, migratetype);
- set_pcppage_migratetype(page, migratetype);
trace_mm_page_alloc_zone_locked(page, order, migratetype,
pcp_allowed_order(order) &&
migratetype < MIGRATE_PCPTYPES);
@@ -2182,7 +2157,7 @@ static int rmqueue_bulk(struct zone *zone, unsigned int order,
* pages are ordered properly.
*/
list_add_tail(&page->pcp_list, list);
- if (is_migrate_cma(get_pcppage_migratetype(page)))
+ if (is_migrate_cma(get_pageblock_migratetype(page)))
__mod_zone_page_state(zone, NR_FREE_CMA_PAGES,
-(1 << order));
}
@@ -2375,19 +2350,6 @@ void drain_all_pages(struct zone *zone)
__drain_all_pages(zone, false);
}
-static bool free_unref_page_prepare(struct page *page, unsigned long pfn,
- unsigned int order)
-{
- int migratetype;
-
- if (!free_pages_prepare(page, order))
- return false;
-
- migratetype = get_pfnblock_migratetype(page, pfn);
- set_pcppage_migratetype(page, migratetype);
- return true;
-}
-
static int nr_pcp_free(struct per_cpu_pages *pcp, int batch, int high, bool free_high)
{
int min_nr_free, max_nr_free;
@@ -2520,7 +2482,7 @@ void free_unref_page(struct page *page, unsigned int order)
unsigned long pfn = page_to_pfn(page);
int migratetype, pcpmigratetype;
- if (!free_unref_page_prepare(page, pfn, order))
+ if (!free_pages_prepare(page, order))
return;
/*
@@ -2530,7 +2492,7 @@ void free_unref_page(struct page *page, unsigned int order)
* get those areas back if necessary. Otherwise, we may have to free
* excessively into the page allocator
*/
- migratetype = pcpmigratetype = get_pcppage_migratetype(page);
+ migratetype = pcpmigratetype = get_pfnblock_migratetype(page, pfn);
if (unlikely(migratetype >= MIGRATE_PCPTYPES)) {
if (unlikely(is_migrate_isolate(migratetype))) {
free_one_page(page_zone(page), page, pfn, order, migratetype, FPI_NONE);
@@ -2569,14 +2531,14 @@ void free_unref_folios(struct folio_batch *folios)
if (order > 0 && folio_test_large_rmappable(folio))
folio_undo_large_rmappable(folio);
- if (!free_unref_page_prepare(&folio->page, pfn, order))
+ if (!free_pages_prepare(&folio->page, order))
continue;
/*
* Free isolated folios and orders not handled on the PCP
* directly to the allocator, see comment in free_unref_page.
*/
- migratetype = get_pcppage_migratetype(&folio->page);
+ migratetype = get_pfnblock_migratetype(&folio->page, pfn);
if (!pcp_allowed_order(order) ||
is_migrate_isolate(migratetype)) {
free_one_page(folio_zone(folio), &folio->page, pfn,
@@ -2593,10 +2555,11 @@ void free_unref_folios(struct folio_batch *folios)
for (i = 0; i < folios->nr; i++) {
struct folio *folio = folios->folios[i];
struct zone *zone = folio_zone(folio);
+ unsigned long pfn = folio_pfn(folio);
unsigned int order = (unsigned long)folio->private;
folio->private = NULL;
- migratetype = get_pcppage_migratetype(&folio->page);
+ migratetype = get_pfnblock_migratetype(&folio->page, pfn);
/* Different zone requires a different pcp lock */
if (zone != locked_zone) {
@@ -2613,9 +2576,8 @@ void free_unref_folios(struct folio_batch *folios)
pcp = pcp_spin_trylock(zone->per_cpu_pageset);
if (unlikely(!pcp)) {
pcp_trylock_finish(UP_flags);
- free_one_page(zone, &folio->page,
- folio_pfn(folio), order,
- migratetype, FPI_NONE);
+ free_one_page(zone, &folio->page, pfn,
+ order, migratetype, FPI_NONE);
locked_zone = NULL;
continue;
}
@@ -2784,7 +2746,7 @@ struct page *rmqueue_buddy(struct zone *preferred_zone, struct zone *zone,
}
}
__mod_zone_freepage_state(zone, -(1 << order),
- get_pcppage_migratetype(page));
+ get_pageblock_migratetype(page));
spin_unlock_irqrestore(&zone->lock, flags);
} while (check_new_pages(page, order));
--
2.44.0
