Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Sasha Levin <sashal@kernel.org>
To:     stable@vger.kernel.org, linux-kernel@vger.kernel.org
Cc:     alex chen <alex.chen@huawei.com>, Joel Becker <jlbec@evilplan.org>,
        Mark Fasheh <mfasheh@suse.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Sasha Levin <sashal@kernel.org>
Subject: [PATCH AUTOSEL 3.18 02/98] ocfs2: fix journal commit deadlock in ocfs2_convert_inline_data_to_extents
Date:   Thu, 25 Oct 2018 10:17:17 -0400
Message-Id: <20181025141853.214051-2-sashal@kernel.org>
In-Reply-To: <20181025141853.214051-1-sashal@kernel.org>
References: <20181025141853.214051-1-sashal@kernel.org>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

From: alex chen <alex.chen@huawei.com>

[ Upstream commit 15eba0fe3eeaeb1b80489c1ebb9d47d6d7003f57 ]

Similar to ocfs2_write_end_nolock() which is metioned at commit
136f49b91710 ("ocfs2: fix journal commit deadlock"), we should unlock
pages before ocfs2_commit_trans() in ocfs2_convert_inline_data_to_extents.

Otherwise, it will cause a deadlock with journal commit threads.

Signed-off-by: Alex Chen <alex.chen@huawei.com>
Reviewed-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/ocfs2/alloc.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index d9db90446c9d..3564b896a92e 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -6871,7 +6871,7 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		mlog_errno(ret);
-		goto out_unlock;
+		goto out;
 	}
 
 	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh,
@@ -6929,7 +6929,7 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 		if (ret) {
 			mlog_errno(ret);
 			need_free = 1;
-			goto out_commit;
+			goto out_unlock;
 		}
 
 		page_end = PAGE_CACHE_SIZE;
@@ -6962,12 +6962,16 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 		if (ret) {
 			mlog_errno(ret);
 			need_free = 1;
-			goto out_commit;
+			goto out_unlock;
 		}
 
 		inode->i_blocks = ocfs2_inode_sector_count(inode);
 	}
 
+out_unlock:
+	if (pages)
+		ocfs2_unlock_and_free_pages(pages, num_pages);
+
 out_commit:
 	if (ret < 0 && did_quota)
 		dquot_free_space_nodirty(inode,
@@ -6987,15 +6991,11 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 
 	ocfs2_commit_trans(osb, handle);
 
-out_unlock:
+out:
 	if (data_ac)
 		ocfs2_free_alloc_context(data_ac);
-
-out:
-	if (pages) {
-		ocfs2_unlock_and_free_pages(pages, num_pages);
+	if (pages)
 		kfree(pages);
-	}
 
 	return ret;
 }
-- 
2.17.1