From: Robin Dong Subject: [PATCH] ext4: s_dirtyclusters_counter should tranform to unit of cluster before assigning to "dirty_clusters" in ext4_has_free_clusters() Date: Wed, 1 Feb 2012 11:17:50 +0800 Message-ID: <1328066270-4461-1-git-send-email-hao.bigrat@gmail.com> Cc: Robin Dong , "Theodore Ts'o" To: linux-ext4@vger.kernel.org Return-path: Received: from mail-tul01m020-f174.google.com ([209.85.214.174]:47333 "EHLO mail-tul01m020-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754824Ab2BADSE (ORCPT ); Tue, 31 Jan 2012 22:18:04 -0500 Received: by obcva7 with SMTP id va7so730079obc.19 for ; Tue, 31 Jan 2012 19:18:02 -0800 (PST) Sender: linux-ext4-owner@vger.kernel.org List-ID: From: Robin Dong Creating 4-byte files until ENOSPC in a delay-allocation and bigalloc ext4 fs and then sync it, the dmseg will report like: [ 482.154538] EXT4-fs (sdb6): delayed block allocation failed for inode 1664 at logical offset 0 with max blocks 1 with error -28 [ 482.154540] EXT4-fs (sdb6): This should not happen!! Data will be lost The reason is ext4_has_free_clusters reporting wrong result. Actually, the unit of sbi->s_dirtyclusters_counter is block, so we should tranform it to cluster for argument "dirty_clusters", just like "free_clusters". Cc: "Theodore Ts'o" Reported-by: Eric Sandeen Signed-off-by: Robin Dong --- fs/ext4/balloc.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index f9e2cd8..8ff10c4 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -427,7 +427,7 @@ static int ext4_has_free_clusters(struct ext4_sb_info *sbi, if (free_clusters - (nclusters + root_clusters + dirty_clusters) < EXT4_FREECLUSTERS_WATERMARK) { free_clusters = EXT4_C2B(sbi, percpu_counter_sum_positive(fcc)); - dirty_clusters = percpu_counter_sum_positive(dcc); + dirty_clusters = EXT4_C2B(sbi, percpu_counter_sum_positive(dcc)); } /* Check whether we have space after accounting for current * dirty clusters & root reserved clusters. -- 1.7.4.1