6.6-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nhat Pham nphamcs@gmail.com
commit 3a75cb05d53f4a6823a32deb078de1366954a804 upstream.
In cachestat, we access the folio from the page cache's xarray to compute its page offset, and check for its dirty and writeback flags. However, we do not hold a reference to the folio before performing these actions, which means the folio can concurrently be released and reused as another folio/page/slab.
Get around this altogether by just using xarray's existing machinery for the folio page offsets and dirty/writeback states.
This changes behavior for tmpfs files to now always report zeroes in their dirty and writeback counters. This is okay as tmpfs doesn't follow conventional writeback cache behavior: its pages get "cleaned" during swapout, after which they're no longer resident etc.
Link: https://lkml.kernel.org/r/20240220153409.GA216065@cmpxchg.org Fixes: cf264e1329fb ("cachestat: implement cachestat syscall") Reported-by: Jann Horn jannh@google.com Suggested-by: Matthew Wilcox willy@infradead.org Signed-off-by: Nhat Pham nphamcs@gmail.com Signed-off-by: Johannes Weiner hannes@cmpxchg.org Tested-by: Jann Horn jannh@google.com Cc: stable@vger.kernel.org [6.4+] Signed-off-by: Andrew Morton akpm@linux-foundation.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- mm/filemap.c | 51 ++++++++++++++++++++++++++------------------------- 1 file changed, 26 insertions(+), 25 deletions(-)
--- a/mm/filemap.c +++ b/mm/filemap.c @@ -4159,28 +4159,40 @@ static void filemap_cachestat(struct add
rcu_read_lock(); xas_for_each(&xas, folio, last_index) { + int order; unsigned long nr_pages; pgoff_t folio_first_index, folio_last_index;
+ /* + * Don't deref the folio. It is not pinned, and might + * get freed (and reused) underneath us. + * + * We *could* pin it, but that would be expensive for + * what should be a fast and lightweight syscall. + * + * Instead, derive all information of interest from + * the rcu-protected xarray. + */ + if (xas_retry(&xas, folio)) continue;
+ order = xa_get_order(xas.xa, xas.xa_index); + nr_pages = 1 << order; + folio_first_index = round_down(xas.xa_index, 1 << order); + folio_last_index = folio_first_index + nr_pages - 1; + + /* Folios might straddle the range boundaries, only count covered pages */ + if (folio_first_index < first_index) + nr_pages -= first_index - folio_first_index; + + if (folio_last_index > last_index) + nr_pages -= folio_last_index - last_index; + if (xa_is_value(folio)) { /* page is evicted */ void *shadow = (void *)folio; bool workingset; /* not used */ - int order = xa_get_order(xas.xa, xas.xa_index); - - nr_pages = 1 << order; - folio_first_index = round_down(xas.xa_index, 1 << order); - folio_last_index = folio_first_index + nr_pages - 1; - - /* Folios might straddle the range boundaries, only count covered pages */ - if (folio_first_index < first_index) - nr_pages -= first_index - folio_first_index; - - if (folio_last_index > last_index) - nr_pages -= folio_last_index - last_index;
cs->nr_evicted += nr_pages;
@@ -4198,24 +4210,13 @@ static void filemap_cachestat(struct add goto resched; }
- nr_pages = folio_nr_pages(folio); - folio_first_index = folio_pgoff(folio); - folio_last_index = folio_first_index + nr_pages - 1; - - /* Folios might straddle the range boundaries, only count covered pages */ - if (folio_first_index < first_index) - nr_pages -= first_index - folio_first_index; - - if (folio_last_index > last_index) - nr_pages -= folio_last_index - last_index; - /* page is in cache */ cs->nr_cache += nr_pages;
- if (folio_test_dirty(folio)) + if (xas_get_mark(&xas, PAGECACHE_TAG_DIRTY)) cs->nr_dirty += nr_pages;
- if (folio_test_writeback(folio)) + if (xas_get_mark(&xas, PAGECACHE_TAG_WRITEBACK)) cs->nr_writeback += nr_pages;
resched: