linux-stable.git/fs/f2fs/node.c, branch v4.19.2

f2fs: fix to recover cold bit of inode block during POR

2018-11-13T19:08:39+00:00

commit ef2a007134b4eaa39264c885999f296577bc87d2 upstream.

Testcase to reproduce this bug:
1. mkfs.f2fs /dev/sdd
2. mount -t f2fs /dev/sdd /mnt/f2fs
3. touch /mnt/f2fs/file
4. sync
5. chattr +A /mnt/f2fs/file
6. xfs_io -f /mnt/f2fs/file -c "fsync"
7. godown /mnt/f2fs
8. umount /mnt/f2fs
9. mount -t f2fs /dev/sdd /mnt/f2fs
10. chattr -A /mnt/f2fs/file
11. xfs_io -f /mnt/f2fs/file -c "fsync"
12. umount /mnt/f2fs
13. mount -t f2fs /dev/sdd /mnt/f2fs
14. lsattr /mnt/f2fs/file

-----------------N- /mnt/f2fs/file

But actually, we expect the corrct result is:

-------A---------N- /mnt/f2fs/file

The reason is in step 9) we missed to recover cold bit flag in inode
block, so later, in fsync, we will skip write inode block due to below
condition check, result in lossing data in another SPOR.

f2fs_fsync_node_pages()
	if (!IS_DNODE(page) || !is_cold_node(page))
		continue;

Note that, I guess that some non-dir inode has already lost cold bit
during POR, so in order to reenable recovery for those inode, let's
try to recover cold bit in f2fs_iget() to save more fsynced data.

Fixes: c56675750d7c ("f2fs: remove unneeded set_cold_node()")
Cc:  4.17+
Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim 
Signed-off-by: Greg Kroah-Hartman

f2fs: fix missing up_read

2018-11-13T19:08:39+00:00

commit 89d13c38501df730cbb2e02c4499da1b5187119d upstream.

This patch fixes missing up_read call.

Fixes: c9b60788fc76 ("f2fs: fix to do sanity check with block address in main area")
Cc:  # 4.19+
Reviewed-by: Chao Yu 
Signed-off-by: Jaegeuk Kim 
Signed-off-by: Greg Kroah-Hartman

f2fs: fix to recover inode's crtime during POR

2018-11-13T19:08:35+00:00

[ Upstream commit 5cd1f387a13b5188b4edb4c834310302a85a6ea2 ]

Testcase to reproduce this bug:
1. mkfs.f2fs -O extra_attr -O inode_crtime /dev/sdd
2. mount -t f2fs /dev/sdd /mnt/f2fs
3. touch /mnt/f2fs/file
4. xfs_io -f /mnt/f2fs/file -c "fsync"
5. godown /mnt/f2fs
6. umount /mnt/f2fs
7. mount -t f2fs /dev/sdd /mnt/f2fs
8. xfs_io -f /mnt/f2fs/file -c "statx -r"

stat.btime.tv_sec = 0
stat.btime.tv_nsec = 0

This patch fixes to recover inode creation time fields during
mount.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

f2fs: rework fault injection handling to avoid a warning

2018-08-14T16:49:15+00:00

When CONFIG_F2FS_FAULT_INJECTION is disabled, we get a warning about an
unused label:

fs/f2fs/segment.c: In function '__submit_discard_cmd':
fs/f2fs/segment.c:1059:1: error: label 'submit' defined but not used [-Werror=unused-label]

This could be fixed by adding another #ifdef around it, but the more
reliable way of doing this seems to be to remove the other #ifdefs
where that is easily possible.

By defining time_to_inject() as a trivial stub, most of the checks for
CONFIG_F2FS_FAULT_INJECTION can go away. This also leads to nicer
formatting of the code.

Signed-off-by: Arnd Bergmann 
Reviewed-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: refresh recent accessed nat entry in lru list

2018-08-13T17:48:17+00:00

Introduce nat_list_lock to protect nm_i->nat_entries list, and manage
it as a LRU list, refresh location for therein recent accessed entries
in the list.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: fix to avoid broken of dnode block list

2018-08-10T23:19:05+00:00

f2fs recovery flow is relying on dnode block link list, it means fsynced
file recovery depends on previous dnode's persistence in the list, so
during fsync() we should wait on all regular inode's dnode writebacked
before issuing flush.

By this way, we can avoid dnode block list being broken by out-of-order
IO submission due to IO scheduler or driver.

Sheng Yong helps to do the test with this patch:

Target:/data (f2fs, -)
64MB / 32768KB / 4KB / 8

1 / PERSIST / Index

Base:
	SEQ-RD(MB/s)	SEQ-WR(MB/s)	RND-RD(IOPS)	RND-WR(IOPS)	Insert(TPS)	Update(TPS)	Delete(TPS)
1	867.82		204.15		41440.03	41370.54	680.8		1025.94		1031.08
2	871.87		205.87		41370.3		40275.2		791.14		1065.84		1101.7
3	866.52		205.69		41795.67	40596.16	694.69		1037.16		1031.48
Avg	868.7366667	205.2366667	41535.33333	40747.3		722.21		1042.98		1054.753333

After:
	SEQ-RD(MB/s)	SEQ-WR(MB/s)	RND-RD(IOPS)	RND-WR(IOPS)	Insert(TPS)	Update(TPS)	Delete(TPS)
1	798.81		202.5		41143		40613.87	602.71		838.08		913.83
2	805.79		206.47		40297.2		41291.46	604.44		840.75		924.27
3	814.83		206.17		41209.57	40453.62	602.85		834.66		927.91
Avg	806.4766667	205.0466667	40883.25667	40786.31667	603.3333333	837.83		922.0033333

Patched/Original:
	0.928332713	0.999074239	0.984300676	1.000957528	0.835398753	0.803303994	0.874141189

It looks like atomic write will suffer performance regression.

I suspect that the criminal is that we forcing to wait all dnode being in
storage cache before we issue PREFLUSH+FUA.

BTW, will commit ("f2fs: don't need to wait for node writes for atomic write")
cause the problem: we will lose data of last transaction after SPO, even if
atomic write return no error:

- atomic_open();
- write() P1, P2, P3;
- atomic_commit();
 - writeback data: P1, P2, P3;
 - writeback node: N1, N2, N3;  <--- If N1, N2 is not writebacked, N3 with fsync_mark is
writebacked, In SPOR, we won't find N3 since node chain is broken, turns out that losing
last transaction.
 - preflush + fua;
- power-cut

If we don't wait dnode writeback for atomic_write:

	SEQ-RD(MB/s)	SEQ-WR(MB/s)	RND-RD(IOPS)	RND-WR(IOPS)	Insert(TPS)	Update(TPS)	Delete(TPS)
1	779.91		206.03		41621.5		40333.16	716.9		1038.21		1034.85
2	848.51		204.35		40082.44	39486.17	791.83		1119.96		1083.77
3	772.12		206.27		41335.25	41599.65	723.29		1055.07		971.92
Avg	800.18		205.55		41013.06333	40472.99333	744.0066667	1071.08		1030.18

Patched/Original:
	0.92108464	1.001526693	0.987425886	0.993268102	1.030180511	1.026942031	0.976702294

SQLite's performance recovers.

Jaegeuk:
"Practically, I don't see db corruption becase of this. We can excuse to lose
the last transaction."

Finally, we decide to keep original implementation of atomic write interface
sematics that we don't wait all dnode writeback before preflush+fua submission.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: avoid f2fs_bug_on() in cp_error case

2018-08-10T23:19:05+00:00

There is a subtle race condition to invoke f2fs_bug_on() in shutdown tests. I've
confirmed that the last checkpoint is preserved in consistent state, so it'd be
fine to just return error at this moment.

Signed-off-by: Jaegeuk Kim

f2fs: let checkpoint flush dnode page of regular

2018-08-01T18:52:36+00:00

Fsyncer will wait on all dnode pages of regular writeback before flushing,
if there are async dnode pages blocked by IO scheduler, it may decrease
fsync's performance.

In this patch, we choose to let f2fs_balance_fs_bg() to trigger checkpoint
to flush these dnode pages of regular, so async IO of dnode page can be
elimitnated, making fsyncer only need to wait for sync IO.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: clean up with get_current_nat_page

2018-08-01T18:52:36+00:00

Just cleanup, no logic change.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: fix to propagate error from __get_meta_page()

2018-08-01T18:52:36+00:00

If caller of __get_meta_page() can handle error, let's propagate error
from __get_meta_page().

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim