From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 019A61C03 for ; Mon, 20 Mar 2023 15:10:27 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 711AFC433D2; Mon, 20 Mar 2023 15:10:27 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1679325027; bh=gwi4qgfkSnuoIiJJcxMq7HAed2DLyUpGGxuCINhs1L0=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=FZ98eXx8qp44l4XResvO9y+DZMCsh8f/hWxKh14znkEi5hDrtG3m+AIGr7Da5T+n/ zhjIpHIHczBRWq5KYXWhtXz5BNN/TfaUUK7rHmp00LZNGMomkCLUAqfbb9OAXjEXeJ LohWf++MrlWQGUFUGpImx18nzpDKxIYSVZ7lv+TY= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, "Darrick J. Wong" , Christoph Hellwig , Dave Chinner , Dave Chinner , Amir Goldstein Subject: [PATCH 5.10 85/99] xfs: dont leak btree cursor when insrec fails after a split Date: Mon, 20 Mar 2023 15:55:03 +0100 Message-Id: <20230320145446.977379316@linuxfoundation.org> X-Mailer: git-send-email 2.40.0 In-Reply-To: <20230320145443.333824603@linuxfoundation.org> References: <20230320145443.333824603@linuxfoundation.org> User-Agent: quilt/0.67 Precedence: bulk X-Mailing-List: patches@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: "Darrick J. Wong" commit a54f78def73d847cb060b18c4e4a3d1d26c9ca6d upstream. The recent patch to improve btree cycle checking caused a regression when I rebased the in-memory btree branch atop the 5.19 for-next branch, because in-memory short-pointer btrees do not have AG numbers. This produced the following complaint from kmemleak: unreferenced object 0xffff88803d47dde8 (size 264): comm "xfs_io", pid 4889, jiffies 4294906764 (age 24.072s) hex dump (first 32 bytes): 90 4d 0b 0f 80 88 ff ff 00 a0 bd 05 80 88 ff ff .M.............. e0 44 3a a0 ff ff ff ff 00 df 08 06 80 88 ff ff .D:............. backtrace: [] xfbtree_dup_cursor+0x49/0xc0 [xfs] [] xfs_btree_dup_cursor+0x3b/0x200 [xfs] [] __xfs_btree_split+0x6ad/0x820 [xfs] [] xfs_btree_split+0x60/0x110 [xfs] [] xfs_btree_make_block_unfull+0x19a/0x1f0 [xfs] [] xfs_btree_insrec+0x3aa/0x810 [xfs] [] xfs_btree_insert+0xb3/0x240 [xfs] [] xfs_rmap_insert+0x99/0x200 [xfs] [] xfs_rmap_map_shared+0x192/0x5f0 [xfs] [] xfs_rmap_map_raw+0x6b/0x90 [xfs] [] xrep_rmap_stash+0xd5/0x1d0 [xfs] [] xrep_rmap_visit_bmbt+0xa0/0xf0 [xfs] [] xrep_rmap_scan_iext+0x56/0xa0 [xfs] [] xrep_rmap_scan_ifork+0xd8/0x160 [xfs] [] xrep_rmap_scan_inode+0x35/0x80 [xfs] [] xrep_rmap_find_rmaps+0x10e/0x270 [xfs] I noticed that xfs_btree_insrec has a bunch of debug code that return out of the function immediately, without freeing the "new" btree cursor that can be returned when _make_block_unfull calls xfs_btree_split. Fix the error return in this function to free the btree cursor. Signed-off-by: Darrick J. Wong Reviewed-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Dave Chinner Signed-off-by: Amir Goldstein Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_btree.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) --- a/fs/xfs/libxfs/xfs_btree.c +++ b/fs/xfs/libxfs/xfs_btree.c @@ -3190,7 +3190,7 @@ xfs_btree_insrec( struct xfs_btree_block *block; /* btree block */ struct xfs_buf *bp; /* buffer for block */ union xfs_btree_ptr nptr; /* new block ptr */ - struct xfs_btree_cur *ncur; /* new btree cursor */ + struct xfs_btree_cur *ncur = NULL; /* new btree cursor */ union xfs_btree_key nkey; /* new block key */ union xfs_btree_key *lkey; int optr; /* old key/record index */ @@ -3270,7 +3270,7 @@ xfs_btree_insrec( #ifdef DEBUG error = xfs_btree_check_block(cur, block, level, bp); if (error) - return error; + goto error0; #endif /* @@ -3290,7 +3290,7 @@ xfs_btree_insrec( for (i = numrecs - ptr; i >= 0; i--) { error = xfs_btree_debug_check_ptr(cur, pp, i, level); if (error) - return error; + goto error0; } xfs_btree_shift_keys(cur, kp, 1, numrecs - ptr + 1); @@ -3375,6 +3375,8 @@ xfs_btree_insrec( return 0; error0: + if (ncur) + xfs_btree_del_cursor(ncur, error); return error; }