[PATCH V2] mm/vmstat: Add events for THP migration without split

From: Anshuman Khandual
Date: Thu Jun 04 2020 - 00:02:06 EST


Add the following new VM events which will help in validating THP migration
without split. Statistics reported through these new events will help in
performance debugging.

1. THP_MIGRATION_SUCCESS
2. THP_MIGRATION_FAILURE

THP_MIGRATION_FAILURE in particular represents an event when a THP could
not be migrated as a single entity following an allocation failure and
ended up getting split into constituent normal pages before being retried.
This event, along with PGMIGRATE_SUCCESS and PGMIGRATE_FAIL will help in
quantifying and analyzing THP migration events including both success and
failure cases.

Cc: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Cc: Zi Yan <ziy@xxxxxxxxxx>
Cc: John Hubbard <jhubbard@xxxxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: linux-mm@xxxxxxxxx
Cc: linux-kernel@xxxxxxxxxxxxxxx
[hughd: fixed oops on NULL newpage]
Signed-off-by: Anshuman Khandual <anshuman.khandual@xxxxxxx>
---
Changes in V2:

- Dropped PMD reference both from code and commit message per Matthew
- Added documentation and updated the commit message per Daniel

Changes in V1: (https://patchwork.kernel.org/patch/11564497/)

- Changed function name as thp_pmd_migration_success() per John
- Folded in a fix (https://patchwork.kernel.org/patch/11563009/) from Hugh

Changes in RFC V2: (https://patchwork.kernel.org/patch/11554861/)

- Decopupled and renamed VM events from their implementation per Zi and John
- Added THP_PMD_MIGRATION_FAILURE VM event upon allocation failure and split

Changes in RFC V1: (https://patchwork.kernel.org/patch/11542055/)

Documentation/vm/page_migration.rst | 15 +++++++++++++++
include/linux/vm_event_item.h | 4 ++++
mm/migrate.c | 23 +++++++++++++++++++++++
mm/vmstat.c | 4 ++++
4 files changed, 46 insertions(+)

diff --git a/Documentation/vm/page_migration.rst b/Documentation/vm/page_migration.rst
index 1d6cd7db4e43..67e9b067fed7 100644
--- a/Documentation/vm/page_migration.rst
+++ b/Documentation/vm/page_migration.rst
@@ -253,5 +253,20 @@ which are function pointers of struct address_space_operations.
PG_isolated is alias with PG_reclaim flag so driver shouldn't use the flag
for own purpose.

+Quantifying Migration
+=====================
+Following events can be used to quantify page migration.
+
+- PGMIGRATE_SUCCESS
+- PGMIGRATE_FAIL
+- THP_MIGRATION_SUCCESS
+- THP_MIGRATION_FAILURE
+
+THP_MIGRATION_FAILURE in particular represents an event when a THP could not be
+migrated as a single entity following an allocation failure and ended up getting
+split into constituent normal pages before being retried. This event, along with
+PGMIGRATE_SUCCESS and PGMIGRATE_FAIL will help in quantifying and analyzing THP
+migration events including both success and failure cases.
+
Christoph Lameter, May 8, 2006.
Minchan Kim, Mar 28, 2016.
diff --git a/include/linux/vm_event_item.h b/include/linux/vm_event_item.h
index ffef0f279747..6459265461df 100644
--- a/include/linux/vm_event_item.h
+++ b/include/linux/vm_event_item.h
@@ -91,6 +91,10 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT,
THP_ZERO_PAGE_ALLOC_FAILED,
THP_SWPOUT,
THP_SWPOUT_FALLBACK,
+#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
+ THP_MIGRATION_SUCCESS,
+ THP_MIGRATION_FAILURE,
+#endif
#endif
#ifdef CONFIG_MEMORY_BALLOON
BALLOON_INFLATE,
diff --git a/mm/migrate.c b/mm/migrate.c
index 7160c1556f79..0bb1dbb891bb 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1170,6 +1170,20 @@ static int __unmap_and_move(struct page *page, struct page *newpage,
#define ICE_noinline
#endif

+#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
+static inline void thp_migration_success(bool success)
+{
+ if (success)
+ count_vm_event(THP_MIGRATION_SUCCESS);
+ else
+ count_vm_event(THP_MIGRATION_FAILURE);
+}
+#else
+static inline void thp_migration_success(bool success)
+{
+}
+#endif
+
/*
* Obtain the lock on page, remove all ptes and migrate the page
* to the newly allocated page in newpage.
@@ -1232,6 +1246,14 @@ static ICE_noinline int unmap_and_move(new_page_t get_new_page,
* we want to retry.
*/
if (rc == MIGRATEPAGE_SUCCESS) {
+ /*
+ * When the page to be migrated has been freed from under
+ * us, that is considered a MIGRATEPAGE_SUCCESS, but no
+ * newpage has been allocated. It should not be counted
+ * as a successful THP migration.
+ */
+ if (newpage && PageTransHuge(newpage))
+ thp_migration_success(true);
put_page(page);
if (reason == MR_MEMORY_FAILURE) {
/*
@@ -1474,6 +1496,7 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
unlock_page(page);
if (!rc) {
list_safe_reset_next(page, page2, lru);
+ thp_migration_success(false);
goto retry;
}
}
diff --git a/mm/vmstat.c b/mm/vmstat.c
index 96d21a792b57..4ce1ab2e9704 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1274,6 +1274,10 @@ const char * const vmstat_text[] = {
"thp_zero_page_alloc_failed",
"thp_swpout",
"thp_swpout_fallback",
+#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
+ "thp_migration_success",
+ "thp_migration_failure",
+#endif
#endif
#ifdef CONFIG_MEMORY_BALLOON
"balloon_inflate",
--
2.20.1