The patch titled
Subject: memory-hotplug: fix pages missed by race rather than failing
has been added to the -mm tree. Its filename is
memory-hotplug-fix-pages-missed-by-race-rather-than-failing.patch
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/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Minchan Kim <minchan@xxxxxxxxxx>
Subject: memory-hotplug: fix pages missed by race rather than failing
If race between allocation and isolation in memory-hotplug offline
happens, some pages could be in MIGRATE_MOVABLE of free_list although the
pageblock's migratetype of the page is MIGRATE_ISOLATE.
The race could be detected by get_freepage_migratetype in
__test_page_isolated_in_pageblock. If it is detected, now EBUSY gets
bubbled all the way up and the hotplug operations fails.
But better idea is instead of returning and failing memory-hotremove, move
the free page to the correct list at the time it is detected. It could
enhance memory-hotremove operation success ratio although the race is
really rare.
Suggested by Mel Gorman.
Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
Reviewed-by: Yasuaki Ishimatsu <isimatu.yasuaki@xxxxxxxxxxxxxx>
Acked-by: Mel Gorman <mgorman@xxxxxxx>
Cc: Xishi Qiu <qiuxishi@xxxxxxxxxx>
Cc: Wen Congyang <wency@xxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/page-isolation.h | 4 ++++
mm/page_alloc.c | 2 +-
mm/page_isolation.c | 15 +++++++++++++--
3 files changed, 18 insertions(+), 3 deletions(-)
diff -puN include/linux/page-isolation.h~memory-hotplug-fix-pages-missed-by-race-rather-than-failing include/linux/page-isolation.h
--- a/include/linux/page-isolation.h~memory-hotplug-fix-pages-missed-by-race-rather-than-failing
+++ a/include/linux/page-isolation.h
@@ -6,6 +6,10 @@ bool has_unmovable_pages(struct zone *zo
void set_pageblock_migratetype(struct page *page, int migratetype);
int move_freepages_block(struct zone *zone, struct page *page,
int migratetype);
+int move_freepages(struct zone *zone,
+ struct page *start_page, struct page *end_page,
+ int migratetype);
+
/*
* Changes migrate type in [start_pfn, end_pfn) to be MIGRATE_ISOLATE.
* If specified range includes migrate types other than MOVABLE or CMA,
diff -puN mm/page_alloc.c~memory-hotplug-fix-pages-missed-by-race-rather-than-failing mm/page_alloc.c
--- a/mm/page_alloc.c~memory-hotplug-fix-pages-missed-by-race-rather-than-failing
+++ a/mm/page_alloc.c
@@ -918,7 +918,7 @@ static int fallbacks[MIGRATE_TYPES][4] =
* Note that start_page and end_pages are not aligned on a pageblock
* boundary. If alignment is required, use move_freepages_block()
*/
-static int move_freepages(struct zone *zone,
+int move_freepages(struct zone *zone,
struct page *start_page, struct page *end_page,
int migratetype)
{
diff -puN mm/page_isolation.c~memory-hotplug-fix-pages-missed-by-race-rather-than-failing mm/page_isolation.c
--- a/mm/page_isolation.c~memory-hotplug-fix-pages-missed-by-race-rather-than-failing
+++ a/mm/page_isolation.c
@@ -194,8 +194,19 @@ __test_page_isolated_in_pageblock(unsign
}
page = pfn_to_page(pfn);
if (PageBuddy(page)) {
- if (get_freepage_migratetype(page) != MIGRATE_ISOLATE)
- break;
+ /*
+ * If race between isolatation and allocation happens,
+ * some free pages could be in MIGRATE_MOVABLE list
+ * although pageblock's migratation type of the page
+ * is MIGRATE_ISOLATE. Catch it and move the page into
+ * MIGRATE_ISOLATE list.
+ */
+ if (get_freepage_migratetype(page) != MIGRATE_ISOLATE) {
+ struct page *end_page = page +
+ (1 << page_order(page)) - 1;
+ move_freepages(page_zone(page), page, end_page,
+ MIGRATE_ISOLATE);
+ }
pfn += 1 << page_order(page);
}
else if (page_count(page) == 0 &&
_
Patches currently in -mm which might be from minchan@xxxxxxxxxx are
memory-hotplug-reset-pgdat-kswapd-to-null-if-creating-kernel-thread-fails.patch
linux-next.patch
mm-remove-__gfp_no_kswapd.patch
mm-fix-nonuniform-page-status-when-writing-new-file-with-small-buffer.patch
mm-fix-nonuniform-page-status-when-writing-new-file-with-small-buffer-fix.patch
mm-fix-nonuniform-page-status-when-writing-new-file-with-small-buffer-fix-fix.patch
mm-compaction-update-comment-in-try_to_compact_pages.patch
mm-vmscan-scale-number-of-pages-reclaimed-by-reclaim-compaction-based-on-failures.patch
mm-vmscan-scale-number-of-pages-reclaimed-by-reclaim-compaction-based-on-failures-fix.patch
mm-compaction-capture-a-suitable-high-order-page-immediately-when-it-is-made-available.patch
mm-change-enum-migrate_mode-with-bitwise-type.patch
mm-support-migrate_discard.patch
mm-support-migrate_discard-fix.patch
mm-page_alloc-use-get_freepage_migratetype-instead-of-page_private.patch
mm-remain-migratetype-in-freed-page.patch
memory-hotplug-bug-fix-race-between-isolation-and-allocation.patch
memory-hotplug-fix-pages-missed-by-race-rather-than-failing.patch
memory-hotplug-fix-pages-missed-by-race-rather-than-failng-fix.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
