[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1453580148-61861-2-git-send-email-jaegeuk@kernel.org>
Date: Sat, 23 Jan 2016 12:15:48 -0800
From: Jaegeuk Kim <jaegeuk@...nel.org>
To: linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
linux-f2fs-devel@...ts.sourceforge.net
Cc: Jaegeuk Kim <jaegeuk@...nel.org>
Subject: [PATCH 2/2] f2fs: fix to overcome inline_data floods
The scenario is:
1. create lots of node blocks
2. sync
3. write lots of inline_data
-> got panic due to no free space
In that case, we should flush node blocks when writing inline_data in #3,
and trigger gc as well.
Signed-off-by: Jaegeuk Kim <jaegeuk@...nel.org>
---
fs/f2fs/gc.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 0204433..8686231 100644
--- a/fs/f2fs/gc.c
+++ b/fs/f2fs/gc.c
@@ -838,8 +838,15 @@ gc_more:
if (gc_type == BG_GC && has_not_enough_free_secs(sbi, sec_freed)) {
gc_type = FG_GC;
+ /*
+ * If there is no victim and no prefree segment but still not
+ * enough free sections, we should flush dent/node blocks and do
+ * garbage collections.
+ */
if (__get_victim(sbi, &segno, gc_type) || prefree_segments(sbi))
write_checkpoint(sbi, &cpc);
+ else if (has_not_enough_free_secs(sbi, 0))
+ write_checkpoint(sbi, &cpc);
}
if (segno == NULL_SEGNO && !__get_victim(sbi, &segno, gc_type))
--
2.6.3
Powered by blists - more mailing lists