From patchwork Mon May 16 19:36:42 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Greg KH X-Patchwork-Id: 573261 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 77349C433FE for ; Mon, 16 May 2022 20:10:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S240654AbiEPUKJ (ORCPT ); Mon, 16 May 2022 16:10:09 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:57530 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1351019AbiEPUB4 (ORCPT ); Mon, 16 May 2022 16:01:56 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7B62A47549; Mon, 16 May 2022 12:56:29 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 183AF60FEB; Mon, 16 May 2022 19:56:29 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 2DA39C385AA; Mon, 16 May 2022 19:56:28 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1652730988; bh=35U1+jwEv92WmgOxTVCnczY3ljAOaMeAQs9IhU0Y48A=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=HMbIZmRVs4449I2rde98fT/bAtlS8j5psx9aGD3mQflN/PRaI22bh8itJtFk5M12z gwGRpi8YRJ6xrO9oHAnPZw2tPxNwooxvfboXeboHpe8vHP9OXz1FymwRnSUQ11JBMs BF6hd4FnnRu2Oh0m/TmEh3dMBw40qn2ahEPzAPtE= From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, "Darrick J. Wong" , Brian Foster , "Matthew Wilcox (Oracle)" , Sasha Levin Subject: [PATCH 5.17 068/114] block: Do not call folio_next() on an unreferenced folio Date: Mon, 16 May 2022 21:36:42 +0200 Message-Id: <20220516193627.444142895@linuxfoundation.org> X-Mailer: git-send-email 2.36.1 In-Reply-To: <20220516193625.489108457@linuxfoundation.org> References: <20220516193625.489108457@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: stable@vger.kernel.org From: Matthew Wilcox (Oracle) [ Upstream commit 170f37d6aa6ad4582eefd7459015de79e244536e ] It is unsafe to call folio_next() on a folio unless you hold a reference on it that prevents it from being split or freed. After returning from the iterator, iomap calls folio_end_writeback() which may drop the last reference to the page, or allow the page to be split. If that happens, the iterator will not advance far enough through the bio_vec, leading to assertion failures like the BUG() in folio_end_writeback() that checks we're not trying to end writeback on a page not currently under writeback. Other assertion failures were also seen, but they're all explained by this one bug. Fix the bug by remembering where the next folio starts before returning from the iterator. There are other ways of fixing this bug, but this seems the simplest. Reported-by: Darrick J. Wong Tested-by: Darrick J. Wong Reported-by: Brian Foster Tested-by: Brian Foster Signed-off-by: Matthew Wilcox (Oracle) Signed-off-by: Sasha Levin --- include/linux/bio.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/bio.h b/include/linux/bio.h index 117d7f248ac9..2ca54c084d5a 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -272,6 +272,7 @@ struct folio_iter { size_t offset; size_t length; /* private: for use by the iterator */ + struct folio *_next; size_t _seg_count; int _i; }; @@ -286,6 +287,7 @@ static inline void bio_first_folio(struct folio_iter *fi, struct bio *bio, PAGE_SIZE * (bvec->bv_page - &fi->folio->page); fi->_seg_count = bvec->bv_len; fi->length = min(folio_size(fi->folio) - fi->offset, fi->_seg_count); + fi->_next = folio_next(fi->folio); fi->_i = i; } @@ -293,9 +295,10 @@ static inline void bio_next_folio(struct folio_iter *fi, struct bio *bio) { fi->_seg_count -= fi->length; if (fi->_seg_count) { - fi->folio = folio_next(fi->folio); + fi->folio = fi->_next; fi->offset = 0; fi->length = min(folio_size(fi->folio), fi->_seg_count); + fi->_next = folio_next(fi->folio); } else if (fi->_i + 1 < bio->bi_vcnt) { bio_first_folio(fi, bio, fi->_i + 1); } else {