aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorHugh Dickins <hughd@google.com>2020-10-09 20:07:59 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2020-10-10 15:52:54 -0700
commit033b5d77551167f8c24ca862ce83d3e0745f9245 (patch)
treed12a8eced640eeba315bc6c92969f98766827551
parent6f2f486d57c4d562cdf4932320b66fbb878ab1c4 (diff)
downloadlinux-stericsson-033b5d77551167f8c24ca862ce83d3e0745f9245.tar.gz
mm/khugepaged: fix filemap page_to_pgoff(page) != offset
There have been elusive reports of filemap_fault() hitting its VM_BUG_ON_PAGE(page_to_pgoff(page) != offset, page) on kernels built with CONFIG_READ_ONLY_THP_FOR_FS=y. Suren has hit it on a kernel with CONFIG_READ_ONLY_THP_FOR_FS=y and CONFIG_NUMA is not set: and he has analyzed it down to how khugepaged without NUMA reuses the same huge page after collapse_file() failed (whereas NUMA targets its allocation to the respective node each time). And most of us were usually testing with CONFIG_NUMA=y kernels. collapse_file(old start) new_page = khugepaged_alloc_page(hpage) __SetPageLocked(new_page) new_page->index = start // hpage->index=old offset new_page->mapping = mapping xas_store(&xas, new_page) filemap_fault page = find_get_page(mapping, offset) // if offset falls inside hpage then // compound_head(page) == hpage lock_page_maybe_drop_mmap() __lock_page(page) // collapse fails xas_store(&xas, old page) new_page->mapping = NULL unlock_page(new_page) collapse_file(new start) new_page = khugepaged_alloc_page(hpage) __SetPageLocked(new_page) new_page->index = start // hpage->index=new offset new_page->mapping = mapping // mapping becomes valid again // since compound_head(page) == hpage // page_to_pgoff(page) got changed VM_BUG_ON_PAGE(page_to_pgoff(page) != offset) An initial patch replaced __SetPageLocked() by lock_page(), which did fix the race which Suren illustrates above. But testing showed that it's not good enough: if the racing task's __lock_page() gets delayed long after its find_get_page(), then it may follow collapse_file(new start)'s successful final unlock_page(), and crash on the same VM_BUG_ON_PAGE. It could be fixed by relaxing filemap_fault()'s VM_BUG_ON_PAGE to a check and retry (as is done for mapping), with similar relaxations in find_lock_entry() and pagecache_get_page(): but it's not obvious what else might get caught out; and khugepaged non-NUMA appears to be unique in exposing a page to page cache, then revoking, without going through a full cycle of freeing before reuse. Instead, non-NUMA khugepaged_prealloc_page() release the old page if anyone else has a reference to it (1% of cases when I tested). Although never reported on huge tmpfs, I believe its find_lock_entry() has been at similar risk; but huge tmpfs does not rely on khugepaged for its normal working nearly so much as READ_ONLY_THP_FOR_FS does. Reported-by: Denis Lisov <dennis.lissov@gmail.com> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=206569 Link: https://lore.kernel.org/linux-mm/?q=20200219144635.3b7417145de19b65f258c943%40linux-foundation.org Reported-by: Qian Cai <cai@lca.pw> Link: https://lore.kernel.org/linux-xfs/?q=20200616013309.GB815%40lca.pw Reported-and-analyzed-by: Suren Baghdasaryan <surenb@google.com> Fixes: 87c460a0bded ("mm/khugepaged: collapse_shmem() without freezing new_page") Signed-off-by: Hugh Dickins <hughd@google.com> Cc: stable@vger.kernel.org # v4.9+ Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-rw-r--r--mm/khugepaged.c12
1 files changed, 12 insertions, 0 deletions
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index cfa0dba5fd3b..3a494cfa56d2 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -914,6 +914,18 @@ static struct page *khugepaged_alloc_hugepage(bool *wait)
static bool khugepaged_prealloc_page(struct page **hpage, bool *wait)
{
+ /*
+ * If the hpage allocated earlier was briefly exposed in page cache
+ * before collapse_file() failed, it is possible that racing lookups
+ * have not yet completed, and would then be unpleasantly surprised by
+ * finding the hpage reused for the same mapping at a different offset.
+ * Just release the previous allocation if there is any danger of that.
+ */
+ if (*hpage && page_count(*hpage) > 1) {
+ put_page(*hpage);
+ *hpage = NULL;
+ }
+
if (!*hpage)
*hpage = khugepaged_alloc_hugepage(wait);