From 296bb8c3314b9de714ad0056ed1b9445f450cb83 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim <jaegeuk@kernel.org>
Date: Mon, 26 Jul 2021 09:12:15 -0700
Subject: [PATCH] f2fs: do not submit NEW_ADDR to read node block
Git-commit: b7ec2061737f12c33e45beeb967d17f31abc1ada
Patch-mainline: v5.15-rc1
References: stable-5.14.6
[ Upstream commit b7ec2061737f12c33e45beeb967d17f31abc1ada ]
After the below patch, give cp is errored, we drop dirty node pages. This
can give NEW_ADDR to read node pages. Don't do WARN_ON() which gives
generic/475 failure.
Fixes: 28607bf3aa6f ("f2fs: drop dirty node pages when cp is in error status")
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Acked-by: Takashi Iwai <tiwai@suse.de>
---
fs/f2fs/node.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 0be9e2d7120e..1b0fe6e64b7d 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1321,7 +1321,8 @@ static int read_node_page(struct page *page, int op_flags)
if (err)
return err;
- if (unlikely(ni.blk_addr == NULL_ADDR) ||
+ /* NEW_ADDR can be seen, after cp_error drops some dirty node pages */
+ if (unlikely(ni.blk_addr == NULL_ADDR || ni.blk_addr == NEW_ADDR) ||
is_sbi_flag_set(sbi, SBI_IS_SHUTDOWN)) {
ClearPageUptodate(page);
return -ENOENT;
--
2.26.2