hugetlb: Update -mm patches to fix pool resizing
Subsystems (2)
PATCH 1/1 hugetlb: Update -mm patches to fix pool resizing
Hi Andrew. Here is a port of my explicit resizing corner-case fix that will apply on top of the dynamic pool resizing patches now in -mm. Thanks.
Quoted 20 lines
From the original mainline patch notes... Changes in V2:
- Removed now unnecessary check as suggested by Ken ChenWhen shrinking the size of the hugetlb pool via the
nr_hugepagessysctl, we are careful to keep enough pages around to satisfy reservations. But the calculation is flawed for the following scenario:Action Pool Counters (Total, Free, Resv) ====== ============= Set pool to 1 page 1 1 0 Map 1 page MAP_PRIVATE 1 1 0 Touch the page to fault it in 1 0 0 Set pool to 3 pages 3 2 0 Map 2 pages MAP_SHARED 3 2 2 Set pool to 2 pages 2 1 2 <-- Mistake, should be 3 2 2 Touch the 2 shared pages 2 0 1 <-- Program crashes hereThe last touch above will terminate the process due to lack of huge pages.
Quoted 7 lines
This patch corrects the calculation so that it factors in pages being used for private mappings. Andrew, this is a standalone fix suitable for mainline. It is also now corrected in my latest dynamic pool resizing patchset which I will send out soon.
Signed-off-by Adam Litke <[email protected]>Acked-by Ken Chen <[email protected]>
---
hugetlb.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-)
mm/hugetlb.c +4 -5
@@ -297,14 +297,14 @@ static void try_to_free_low(unsigned long count)for (i =0 ; i <MAX_NUMNODES ; ++i) {struct page *page, *next;list_for_each_entry_safe (page, next, &hugepage_freelists[i], lru) { +if (count >= nr_huge_pages) +return ;if (PageHighMem (page))continue ;list_del (&page->lru);update_and_free_page (page); free_huge_pages--; free_huge_pages_node[page_to_nid (page)]--; -if (count >= nr_huge_pages) -return ; } } }
@@ -344,8 +344,6 @@ static unsigned long set_max_huge_pages(unsigned long count) goto out; } -if (count >= persistent_huge_pages) - goto out;/* * Decrease the pool size
@@ -354,7 +352,8 @@ static unsigned long set_max_huge_pages(unsigned long count)* pages into surplus state as needed so the pool will shrink * to the desired size as pages become free. */ - min_count =max (count, resv_huge_pages); + min_count = resv_huge_pages + nr_huge_pages - free_huge_pages; + min_count =max (count, min_count);try_to_free_low (min_count);while (min_count < persistent_huge_pages) {struct page *page =dequeue_huge_page (NULL ,0 );
Signature · 8 lines
Adam Litke - (agl at us.ibm.com) IBM Linux Technology Center -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"[email protected]"> [email protected] </a>