The quilt patch titled
Subject: mm/rmap: minimize folio->_nr_pages_mapped updates when batching PTE (un)mapping
has been removed from the -mm tree. Its filename was
mm-rmap-minimize-folio-_nr_pages_mapped-updates-when-batching-pte-unmapping.patch
This patch was dropped because it was merged into the mm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: David Hildenbrand <david@xxxxxxxxxx>
Subject: mm/rmap: minimize folio->_nr_pages_mapped updates when batching PTE (un)mapping
Date: Wed, 7 Aug 2024 13:55:15 +0200
It is not immediately obvious, but we can move the folio->_nr_pages_mapped
update out of the loop and reduce the number of atomic ops without
affecting the stats.
The important point to realize is that only removing the last PMD mapping
will result in _nr_pages_mapped going below ENTIRELY_MAPPED, not the
individual atomic_inc_return_relaxed() calls. Concurrent races with
removal of PMD mappings should be handled as expected, just like when we
would have such races right now on a single mapcount update.
In a simple munmap() microbenchmark [1] on 1 GiB of memory backed by the
same PTE-mapped folio size (only mapped by a single process such that they
will get completely unmapped), this change results in a speedup (positive
is good) per folio size on a x86-64 Intel machine of roughly (a bit of
noise expected):
* 16 KiB: +10%
* 32 KiB: +15%
* 64 KiB: +17%
* 128 KiB: +21%
* 256 KiB: +22%
* 512 KiB: +22%
* 1024 KiB: +23%
* 2048 KiB: +27%
[1] https://gitlab.com/davidhildenbrand/scratchspace/-/blob/main/pte-mapped-folio-benchmarks.c
Link: https://lkml.kernel.org/r/20240807115515.1640951-1-david@xxxxxxxxxx
Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/rmap.c | 27 +++++++++++++--------------
1 file changed, 13 insertions(+), 14 deletions(-)
--- a/mm/rmap.c~mm-rmap-minimize-folio-_nr_pages_mapped-updates-when-batching-pte-unmapping
+++ a/mm/rmap.c
@@ -1158,7 +1158,7 @@ static __always_inline unsigned int __fo
{
atomic_t *mapped = &folio->_nr_pages_mapped;
const int orig_nr_pages = nr_pages;
- int first, nr = 0;
+ int first = 0, nr = 0;
__folio_rmap_sanity_checks(folio, page, nr_pages, level);
@@ -1170,13 +1170,13 @@ static __always_inline unsigned int __fo
}
do {
- first = atomic_inc_and_test(&page->_mapcount);
- if (first) {
- first = atomic_inc_return_relaxed(mapped);
- if (first < ENTIRELY_MAPPED)
- nr++;
- }
+ first += atomic_inc_and_test(&page->_mapcount);
} while (page++, --nr_pages > 0);
+
+ if (first &&
+ atomic_add_return_relaxed(first, mapped) < ENTIRELY_MAPPED)
+ nr = first;
+
atomic_add(orig_nr_pages, &folio->_large_mapcount);
break;
case RMAP_LEVEL_PMD:
@@ -1527,7 +1527,7 @@ static __always_inline void __folio_remo
enum rmap_level level)
{
atomic_t *mapped = &folio->_nr_pages_mapped;
- int last, nr = 0, nr_pmdmapped = 0;
+ int last = 0, nr = 0, nr_pmdmapped = 0;
bool partially_mapped = false;
__folio_rmap_sanity_checks(folio, page, nr_pages, level);
@@ -1541,14 +1541,13 @@ static __always_inline void __folio_remo
atomic_sub(nr_pages, &folio->_large_mapcount);
do {
- last = atomic_add_negative(-1, &page->_mapcount);
- if (last) {
- last = atomic_dec_return_relaxed(mapped);
- if (last < ENTIRELY_MAPPED)
- nr++;
- }
+ last += atomic_add_negative(-1, &page->_mapcount);
} while (page++, --nr_pages > 0);
+ if (last &&
+ atomic_sub_return_relaxed(last, mapped) < ENTIRELY_MAPPED)
+ nr = last;
+
partially_mapped = nr && atomic_read(mapped);
break;
case RMAP_LEVEL_PMD:
_
Patches currently in -mm which might be from david@xxxxxxxxxx are
mm-rmap-use-folio-_mapcount-for-small-folios.patch
mm-always-inline-_compound_head-with-config_hugetlb_page_optimize_vmemmap=y.patch
selftests-mm-fix-charge_reserved_hugetlbsh-test.patch
