The patch titled hugetlb: fix quota management for private mappings has been removed from the -mm tree. Its filename was hugetlb-fix-quota-management-for-private-mappings.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ Subject: hugetlb: fix quota management for private mappings From: Adam Litke <agl@xxxxxxxxxx> The hugetlbfs quota management system was never taught to handle MAP_PRIVATE mappings when that support was added. Currently, quota is debited at page instantiation and credited at file truncation. This approach works correctly for shared pages but is incomplete for private pages. In addition to hugetlb_no_page(), private pages can be instantiated by hugetlb_cow(); but this function does not respect quotas. Private huge pages are treated very much like normal, anonymous pages. They are not "backed" by the hugetlbfs file and are not stored in the mapping's radix tree. This means that private pages are invisible to truncate_hugepages() so that function will not credit the quota. This patch (based on a prototype provided by Ken Chen) moves quota crediting for all pages into free_huge_page(). page->private is used to store a pointer to the mapping to which this page belongs. This is used to credit quota on the appropriate hugetlbfs instance. Signed-off-by: Adam Litke <agl@xxxxxxxxxx> Cc: Ken Chen <kenchen@xxxxxxxxxx> Cc: Ken Chen <kenchen@xxxxxxxxxx> Cc: Andy Whitcroft <apw@xxxxxxxxxxxx> Cc: Dave Hansen <haveblue@xxxxxxxxxx> Cc: David Gibson <hermes@xxxxxxxxxxxxxxxxxxxxx> Cc: William Lee Irwin III <wli@xxxxxxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- fs/hugetlbfs/inode.c | 1 - mm/hugetlb.c | 13 ++++++++++--- 2 files changed, 10 insertions(+), 4 deletions(-) diff -puN fs/hugetlbfs/inode.c~hugetlb-fix-quota-management-for-private-mappings fs/hugetlbfs/inode.c --- a/fs/hugetlbfs/inode.c~hugetlb-fix-quota-management-for-private-mappings +++ a/fs/hugetlbfs/inode.c @@ -364,7 +364,6 @@ static void truncate_hugepages(struct in ++next; truncate_huge_page(page); unlock_page(page); - hugetlb_put_quota(mapping); freed++; } huge_pagevec_release(&pvec); diff -puN mm/hugetlb.c~hugetlb-fix-quota-management-for-private-mappings mm/hugetlb.c --- a/mm/hugetlb.c~hugetlb-fix-quota-management-for-private-mappings +++ a/mm/hugetlb.c @@ -116,7 +116,9 @@ static void update_and_free_page(struct static void free_huge_page(struct page *page) { int nid = page_to_nid(page); + struct address_space *mapping; + mapping = (struct address_space *) page_private(page); BUG_ON(page_count(page)); INIT_LIST_HEAD(&page->lru); @@ -129,6 +131,9 @@ static void free_huge_page(struct page * enqueue_huge_page(page); } spin_unlock(&hugetlb_lock); + if (mapping) + hugetlb_put_quota(mapping); + set_page_private(page, 0); } /* @@ -388,8 +393,10 @@ static struct page *alloc_huge_page(stru page = alloc_huge_page_shared(vma, addr); else page = alloc_huge_page_private(vma, addr); - if (page) + if (page) { set_page_refcounted(page); + set_page_private(page, (unsigned long) vma->vm_file->f_mapping); + } return page; } @@ -730,6 +737,8 @@ static int hugetlb_cow(struct mm_struct set_huge_ptep_writable(vma, address, ptep); return 0; } + if (hugetlb_get_quota(vma->vm_file->f_mapping)) + return VM_FAULT_SIGBUS; page_cache_get(old_page); new_page = alloc_huge_page(vma, address); @@ -796,7 +805,6 @@ retry: err = add_to_page_cache(page, mapping, idx, GFP_KERNEL); if (err) { put_page(page); - hugetlb_put_quota(mapping); if (err == -EEXIST) goto retry; goto out; @@ -830,7 +838,6 @@ out: backout: spin_unlock(&mm->page_table_lock); - hugetlb_put_quota(mapping); unlock_page(page); put_page(page); goto out; _ Patches currently in -mm which might be from agl@xxxxxxxxxx are origin.patch hugetlb-allow-sticky-directory-mount-option.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