Messages in this thread Patch in this message |  | | From | Omar Sandoval <> | Subject | [PATCH 2/3] btrfs: handle race on ENOMEM in alloc_extent_buffer | Date | Tue, 17 Feb 2015 02:51:08 -0800 |
| |
Consider the following interleaving of overlapping calls to alloc_extent_buffer:
Call 1:
- Successfully allocates a few pages with find_or_create_page - find_or_create_page fails, goto free_eb - Unlocks the allocated pages
Call 2: - Calls find_or_create_page and gets a page in call 1's extent_buffer - Finds that the page is already associated with an extent_buffer - Grabs a reference to the half-written extent_buffer and calls mark_extent_buffer_accessed on it
mark_extent_buffer_accessed will then try to call mark_page_accessed on a null page and panic.
The fix is to clear page->private of the half-written extent_buffer's pages all at once while holding mapping->private_lock.
Signed-off-by: Omar Sandoval <osandov@osandov.com> --- fs/btrfs/extent_io.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index c73df6a..6024db9 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -4850,6 +4850,7 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info, mark_extent_buffer_accessed(exists, p); goto free_eb; } + exists = NULL; /* * Do this so attach doesn't complain and we need to @@ -4913,13 +4914,24 @@ again: return eb; free_eb: + spin_lock(&mapping->private_lock); for (i = 0; i < num_pages; i++) { - if (eb->pages[i]) - unlock_page(eb->pages[i]); - } + struct page *page = eb->pages[i]; + if (page) { + unlock_page(page); + ClearPagePrivate(page); + set_page_private(page, 0); + /* One for the page private */ + page_cache_release(page); + /* One for when we alloced the page */ + page_cache_release(page); + } + } + spin_unlock(&mapping->private_lock); WARN_ON(!atomic_dec_and_test(&eb->refs)); - btrfs_release_extent_buffer(eb); + __free_extent_buffer(eb); + return exists; } -- 2.3.0
|  |