The patch titled
Subject: mm/page_alloc: use only one PCP list for THP-sized allocations
has been added to the -mm mm-unstable branch. Its filename is
mm-page_alloc-use-only-one-pcp-list-for-thp-sized-allocations.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-page_alloc-use-only-one-pcp-list-for-thp-sized-allocations.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: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Subject: mm/page_alloc: use only one PCP list for THP-sized allocations
Date: Fri, 24 Jun 2022 13:54:18 +0100
The per_cpu_pages is cache-aligned on a standard x86-64 distribution
configuration but a later patch will add a new field which would push the
structure into the next cache line. Use only one list to store THP-sized
pages on the per-cpu list. This assumes that the vast majority of
THP-sized allocations are GFP_MOVABLE but even if it was another type, it
would not contribute to serious fragmentation that potentially causes a
later THP allocation failure. Align per_cpu_pages on the cacheline
boundary to ensure there is no false cache sharing.
After this patch, the structure sizing is;
struct per_cpu_pages {
int count; /* 0 4 */
int high; /* 4 4 */
int batch; /* 8 4 */
short int free_factor; /* 12 2 */
short int expire; /* 14 2 */
struct list_head lists[13]; /* 16 208 */
/* size: 256, cachelines: 4, members: 6 */
/* padding: 32 */
} __attribute__((__aligned__(64)));
Link: https://lkml.kernel.org/r/20220624125423.6126-3-mgorman@xxxxxxxxxxxxxxxxxxx
Signed-off-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Tested-by: Minchan Kim <minchan@xxxxxxxxxx>
Acked-by: Minchan Kim <minchan@xxxxxxxxxx>
Acked-by: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Marcelo Tosatti <mtosatti@xxxxxxxxxx>
Cc: Marek Szyprowski <m.szyprowski@xxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Nicolas Saenz Julienne <nsaenzju@xxxxxxxxxx>
Cc: Yu Zhao <yuzhao@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/mmzone.h | 11 +++++++----
mm/page_alloc.c | 4 ++--
2 files changed, 9 insertions(+), 6 deletions(-)
--- a/include/linux/mmzone.h~mm-page_alloc-use-only-one-pcp-list-for-thp-sized-allocations
+++ a/include/linux/mmzone.h
@@ -355,15 +355,18 @@ enum zone_watermarks {
};
/*
- * One per migratetype for each PAGE_ALLOC_COSTLY_ORDER plus one additional
- * for pageblock size for THP if configured.
+ * One per migratetype for each PAGE_ALLOC_COSTLY_ORDER. One additional list
+ * for THP which will usually be GFP_MOVABLE. Even if it is another type,
+ * it should not contribute to serious fragmentation causing THP allocation
+ * failures.
*/
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
#define NR_PCP_THP 1
#else
#define NR_PCP_THP 0
#endif
-#define NR_PCP_LISTS (MIGRATE_PCPTYPES * (PAGE_ALLOC_COSTLY_ORDER + 1 + NR_PCP_THP))
+#define NR_LOWORDER_PCP_LISTS (MIGRATE_PCPTYPES * (PAGE_ALLOC_COSTLY_ORDER + 1))
+#define NR_PCP_LISTS (NR_LOWORDER_PCP_LISTS + NR_PCP_THP)
/*
* Shift to encode migratetype and order in the same integer, with order
@@ -389,7 +392,7 @@ struct per_cpu_pages {
/* Lists of pages, one per migrate type stored on the pcp-lists */
struct list_head lists[NR_PCP_LISTS];
-};
+} ____cacheline_aligned_in_smp;
struct per_cpu_zonestat {
#ifdef CONFIG_SMP
--- a/mm/page_alloc.c~mm-page_alloc-use-only-one-pcp-list-for-thp-sized-allocations
+++ a/mm/page_alloc.c
@@ -653,7 +653,7 @@ static inline unsigned int order_to_pind
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
if (order > PAGE_ALLOC_COSTLY_ORDER) {
VM_BUG_ON(order != pageblock_order);
- base = PAGE_ALLOC_COSTLY_ORDER + 1;
+ return NR_LOWORDER_PCP_LISTS;
}
#else
VM_BUG_ON(order > PAGE_ALLOC_COSTLY_ORDER);
@@ -667,7 +667,7 @@ static inline int pindex_to_order(unsign
int order = pindex / MIGRATE_PCPTYPES;
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
- if (order > PAGE_ALLOC_COSTLY_ORDER)
+ if (pindex == NR_LOWORDER_PCP_LISTS)
order = pageblock_order;
#else
VM_BUG_ON(order > PAGE_ALLOC_COSTLY_ORDER);
_
Patches currently in -mm which might be from mgorman@xxxxxxxxxxxxxxxxxxx are
mm-page_alloc-add-page-buddy_list-and-page-pcp_list.patch
mm-page_alloc-use-only-one-pcp-list-for-thp-sized-allocations.patch
mm-page_alloc-split-out-buddy-removal-code-from-rmqueue-into-separate-helper.patch
mm-page_alloc-remove-mistaken-page-==-null-check-in-rmqueue.patch
mm-page_alloc-protect-pcp-lists-with-a-spinlock.patch
mm-page_alloc-replace-local_lock-with-normal-spinlock.patch
