[f2fs-dev] [PATCH AUTOSEL 5.1 016/141] f2fs: fix to avoid deadloop if data_flush is on

[f2fs-dev] [PATCH AUTOSEL 5.1 016/141] f2fs: fix to avoid deadloop if data_flush is on

[Sasha Levin]

From: Chao Yu <yuchao0@huawei.com>

[ Upstream commit 040d2bb318d1aea4f28cc22504b44e446666c86e ]

As Hagbard Celine reported:

[  615.697824] INFO: task kworker/u16:5:344 blocked for more than 120 seconds.
[  615.697825]       Not tainted 5.0.15-gentoo-f2fslog #4
[  615.697826] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
disables this message.
[  615.697827] kworker/u16:5   D    0   344      2 0x80000000
[  615.697831] Workqueue: writeback wb_workfn (flush-259:0)
[  615.697832] Call Trace:
[  615.697836]  ? __schedule+0x2c5/0x8b0
[  615.697839]  schedule+0x32/0x80
[  615.697841]  schedule_preempt_disabled+0x14/0x20
[  615.697842]  __mutex_lock.isra.8+0x2ba/0x4d0
[  615.697845]  ? log_store+0xf5/0x260
[  615.697848]  f2fs_write_data_pages+0x133/0x320
[  615.697851]  ? trace_hardirqs_on+0x2c/0xe0
[  615.697854]  do_writepages+0x41/0xd0
[  615.697857]  __filemap_fdatawrite_range+0x81/0xb0
[  615.697859]  f2fs_sync_dirty_inodes+0x1dd/0x200
[  615.697861]  f2fs_balance_fs_bg+0x2a7/0x2c0
[  615.697863]  ? up_read+0x5/0x20
[  615.697865]  ? f2fs_do_write_data_page+0x2cb/0x940
[  615.697867]  f2fs_balance_fs+0xe5/0x2c0
[  615.697869]  __write_data_page+0x1c8/0x6e0
[  615.697873]  f2fs_write_cache_pages+0x1e0/0x450
[  615.697878]  f2fs_write_data_pages+0x14b/0x320
[  615.697880]  ? trace_hardirqs_on+0x2c/0xe0
[  615.697883]  do_writepages+0x41/0xd0
[  615.697885]  __filemap_fdatawrite_range+0x81/0xb0
[  615.697887]  f2fs_sync_dirty_inodes+0x1dd/0x200
[  615.697889]  f2fs_balance_fs_bg+0x2a7/0x2c0
[  615.697891]  f2fs_write_node_pages+0x51/0x220
[  615.697894]  do_writepages+0x41/0xd0
[  615.697897]  __writeback_single_inode+0x3d/0x3d0
[  615.697899]  writeback_sb_inodes+0x1e8/0x410
[  615.697902]  __writeback_inodes_wb+0x5d/0xb0
[  615.697904]  wb_writeback+0x28f/0x340
[  615.697906]  ? cpumask_next+0x16/0x20
[  615.697908]  wb_workfn+0x33e/0x420
[  615.697911]  process_one_work+0x1a1/0x3d0
[  615.697913]  worker_thread+0x30/0x380
[  615.697915]  ? process_one_work+0x3d0/0x3d0
[  615.697916]  kthread+0x116/0x130
[  615.697918]  ? kthread_create_worker_on_cpu+0x70/0x70
[  615.697921]  ret_from_fork+0x3a/0x50

There is still deadloop in below condition:

d A
- do_writepages
 - f2fs_write_node_pages
  - f2fs_balance_fs_bg
   - f2fs_sync_dirty_inodes
    - f2fs_write_cache_pages
     - mutex_lock(&sbi->writepages)	-- lock once
     - __write_data_page
      - f2fs_balance_fs_bg
       - f2fs_sync_dirty_inodes
        - f2fs_write_data_pages
         - mutex_lock(&sbi->writepages)	-- lock again

Thread A			Thread B
- do_writepages
 - f2fs_write_node_pages
  - f2fs_balance_fs_bg
   - f2fs_sync_dirty_inodes
    - .cp_task = current
				- f2fs_sync_dirty_inodes
				 - .cp_task = current
				 - filemap_fdatawrite
				 - .cp_task = NULL
    - filemap_fdatawrite
     - f2fs_write_cache_pages
      - enter f2fs_balance_fs_bg since .cp_task is NULL
    - .cp_task = NULL

Change as below to avoid this:
- add condition to avoid holding .writepages mutex lock in path
of data flush
- introduce mutex lock sbi.flush_lock to exclude concurrent data
flush in background.

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/data.c    | 3 +++
 fs/f2fs/f2fs.h    | 1 +
 fs/f2fs/segment.c | 4 ++++
 fs/f2fs/super.c   | 1 +
 4 files changed, 9 insertions(+)

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 9d3c11e09a03..59262327b5ec 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2231,6 +2231,9 @@ static inline bool __should_serialize_io(struct inode *inode,
 		return false;
 	if (IS_NOQUOTA(inode))
 		return false;
+	/* to avoid deadlock in path of data flush */
+	if (F2FS_I(inode)->cp_task)
+		return false;
 	if (wbc->sync_mode != WB_SYNC_ALL)
 		return true;
 	if (get_dirty_pages(inode) >= SM_I(F2FS_I_SB(inode))->min_seq_blocks)
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index e2cf567dcbd7..2ba2478d0663 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -1200,6 +1200,7 @@ struct f2fs_sb_info {
 	/* for inode management */
 	struct list_head inode_list[NR_INODE_TYPE];	/* dirty inode list */
 	spinlock_t inode_lock[NR_INODE_TYPE];	/* for dirty inode list lock */
+	struct mutex flush_lock;		/* for flush exclusion */
 
 	/* for extent tree cache */
 	struct radix_tree_root extent_tree_root;/* cache extent cache entries */
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 3d6efbfe180f..e9778f06ac0b 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -546,9 +546,13 @@ void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi)
 		if (test_opt(sbi, DATA_FLUSH)) {
 			struct blk_plug plug;
 
+			mutex_lock(&sbi->flush_lock);
+
 			blk_start_plug(&plug);
 			f2fs_sync_dirty_inodes(sbi, FILE_INODE);
 			blk_finish_plug(&plug);
+
+			mutex_unlock(&sbi->flush_lock);
 		}
 		f2fs_sync_fs(sbi->sb, true);
 		stat_inc_bg_cp_count(sbi->stat_info);
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index f2aaa2cc6b3e..6bc1a15b52ad 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -3261,6 +3261,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		INIT_LIST_HEAD(&sbi->inode_list[i]);
 		spin_lock_init(&sbi->inode_lock[i]);
 	}
+	mutex_init(&sbi->flush_lock);
 
 	f2fs_init_extent_cache_info(sbi);
 
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

[f2fs-dev] [PATCH AUTOSEL 5.1 028/141] f2fs: Fix accounting for unusable blocks

[Sasha Levin]

From: Daniel Rosenberg <drosen@google.com>

[ Upstream commit a4c3ecaaadac5693f555cfef1c9eecf4c39df818 ]

Fixes possible underflows when dealing with unusable blocks.

Signed-off-by: Daniel Rosenberg <drosen@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/f2fs.h | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 2ba2478d0663..f1157d5c62bb 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -1761,8 +1761,12 @@ static inline int inc_valid_block_count(struct f2fs_sb_info *sbi,
 
 	if (!__allow_reserved_blocks(sbi, inode, true))
 		avail_user_block_count -= F2FS_OPTION(sbi).root_reserved_blocks;
-	if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
-		avail_user_block_count -= sbi->unusable_block_count;
+	if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED))) {
+		if (avail_user_block_count > sbi->unusable_block_count)
+			avail_user_block_count -= sbi->unusable_block_count;
+		else
+			avail_user_block_count = 0;
+	}
 	if (unlikely(sbi->total_valid_block_count > avail_user_block_count)) {
 		diff = sbi->total_valid_block_count - avail_user_block_count;
 		if (diff > *count)
@@ -1958,7 +1962,7 @@ static inline int inc_valid_node_count(struct f2fs_sb_info *sbi,
 					struct inode *inode, bool is_inode)
 {
 	block_t	valid_block_count;
-	unsigned int valid_node_count;
+	unsigned int valid_node_count, user_block_count;
 	int err;
 
 	if (is_inode) {
@@ -1985,10 +1989,11 @@ static inline int inc_valid_node_count(struct f2fs_sb_info *sbi,
 
 	if (!__allow_reserved_blocks(sbi, inode, false))
 		valid_block_count += F2FS_OPTION(sbi).root_reserved_blocks;
+	user_block_count = sbi->user_block_count;
 	if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
-		valid_block_count += sbi->unusable_block_count;
+		user_block_count -= sbi->unusable_block_count;
 
-	if (unlikely(valid_block_count > sbi->user_block_count)) {
+	if (unlikely(valid_block_count > user_block_count)) {
 		spin_unlock(&sbi->stat_lock);
 		goto enospc;
 	}
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

[f2fs-dev] [PATCH AUTOSEL 5.1 029/141] f2fs: Lower threshold for disable_cp_again

[Sasha Levin]

From: Daniel Rosenberg <drosen@google.com>

[ Upstream commit ae4ad7ea09d32ff1b6fb908ff12f8c1bd5241b29 ]

The existing threshold for allowable holes at checkpoint=disable time is
too high. The OVP space contains reserved segments, which are always in
the form of free segments. These must be subtracted from the OVP value.

The current threshold is meant to be the maximum value of holes of a
single type we can have and still guarantee that we can fill the disk
without failing to find space for a block of a given type.

If the disk is full, ignoring current reserved, which only helps us,
the amount of unused blocks is equal to the OVP area. Of that, there
are reserved segments, which must be free segments, and the rest of the
ovp area, which can come from either free segments or holes. The maximum
possible amount of holes is OVP-reserved.

Now, consider the disk when mounting with checkpoint=disable.
We must be able to fill all available free space with either data or
node blocks. When we start with checkpoint=disable, holes are locked to
their current type. Say we have H of one type of hole, and H+X of the
other. We can fill H of that space with arbitrary typed blocks via SSR.
For the remaining H+X blocks, we may not have any of a given block type
left at all. For instance, if we were to fill the disk entirely with
blocks of the type with fewer holes, the H+X blocks of the opposite type
would not be used. If H+X > OVP-reserved, there would be more holes than
could possibly exist, and we would have failed to find a suitable block
earlier on, leading to a crash in update_sit_entry.

If H+X <= OVP-reserved, then the holes end up effectively masked by the OVP
region in this case.

Signed-off-by: Daniel Rosenberg <drosen@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/segment.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index e9778f06ac0b..60373930b1b7 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -876,7 +876,9 @@ void f2fs_dirty_to_prefree(struct f2fs_sb_info *sbi)
 int f2fs_disable_cp_again(struct f2fs_sb_info *sbi)
 {
 	struct dirty_seglist_info *dirty_i = DIRTY_I(sbi);
-	block_t ovp = overprovision_segments(sbi) << sbi->log_blocks_per_seg;
+	int ovp_hole_segs =
+		(overprovision_segments(sbi) - reserved_segments(sbi));
+	block_t ovp_holes = ovp_hole_segs << sbi->log_blocks_per_seg;
 	block_t holes[2] = {0, 0};	/* DATA and NODE */
 	struct seg_entry *se;
 	unsigned int segno;
@@ -891,10 +893,10 @@ int f2fs_disable_cp_again(struct f2fs_sb_info *sbi)
 	}
 	mutex_unlock(&dirty_i->seglist_lock);
 
-	if (holes[DATA] > ovp || holes[NODE] > ovp)
+	if (holes[DATA] > ovp_holes || holes[NODE] > ovp_holes)
 		return -EAGAIN;
 	if (is_sbi_flag_set(sbi, SBI_CP_DISABLED_QUICK) &&
-		dirty_segments(sbi) > overprovision_segments(sbi))
+		dirty_segments(sbi) > ovp_hole_segs)
 		return -EAGAIN;
 	return 0;
 }
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

[f2fs-dev] [PATCH AUTOSEL 5.1 080/141] f2fs: fix is_idle() check for discard type

[Sasha Levin]

From: Sahitya Tummala <stummala@codeaurora.org>

[ Upstream commit 56659ce838456c6f2315ce8a4bd686ac4b23e9d1 ]

The discard thread should issue upto dpolicy->max_requests at once
and wait for all those discard requests at once it reaches
dpolicy->max_requests. It should then sleep for dpolicy->min_interval
timeout before issuing the next batch of discard requests. But in the
current code of is_idle(), it checks for dcc_info->queued_discard and
aborts issuing the discard batch of max_requests. This
dcc_info->queued_discard will be true always once one discard command
is issued.

It is thus resulting into this type of discard request pattern -

- Issue discard request#1
- is_idle() returns false, discard thread waits for request#1 and then
  sleeps for min_interval 50ms.
- Issue discard request#2
- is_idle() returns false, discard thread waits for request#2 and then
  sleeps for min_interval 50ms.
- and so on for all other discard requests, assuming f2fs is idle w.r.t
  other conditions.

With this fix, the pattern will look like this -

- Issue discard request#1
- Issue discard request#2
  and so on upto max_requests of 8
- Issue discard request#8
- wait for min_interval 50ms.

Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/f2fs.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index f1157d5c62bb..1024550ad11a 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -2194,7 +2194,7 @@ static inline bool is_idle(struct f2fs_sb_info *sbi, int type)
 		get_pages(sbi, F2FS_DIO_WRITE))
 		return false;
 
-	if (SM_I(sbi) && SM_I(sbi)->dcc_info &&
+	if (type != DISCARD_TIME && SM_I(sbi) && SM_I(sbi)->dcc_info &&
 			atomic_read(&SM_I(sbi)->dcc_info->queued_discard))
 		return false;
 
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

[f2fs-dev] [PATCH AUTOSEL 5.1 115/141] f2fs: fix to avoid long latency during umount

[Sasha Levin]

From: Heng Xiao <heng.xiao@unisoc.com>

[ Upstream commit 6e0cd4a9dd4df1a0afcb454f1e654b5c80685913 ]

In umount, we give an constand time to handle pending discard, previously,
in __issue_discard_cmd() we missed to check timeout condition in loop,
result in delaying long time, fix it.

Signed-off-by: Heng Xiao <heng.xiao@unisoc.com>
[Chao Yu: add commit message]
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/segment.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 60373930b1b7..c7ed3022c4e7 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -1483,6 +1483,10 @@ static int __issue_discard_cmd(struct f2fs_sb_info *sbi,
 		list_for_each_entry_safe(dc, tmp, pend_list, list) {
 			f2fs_bug_on(sbi, dc->state != D_PREP);
 
+			if (dpolicy->timeout != 0 &&
+				f2fs_time_over(sbi, dpolicy->timeout))
+				break;
+
 			if (dpolicy->io_aware && i < dpolicy->io_aware_gran &&
 						!is_idle(sbi, DISCARD_TIME)) {
 				io_interrupted = true;
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

[f2fs-dev] [PATCH AUTOSEL 5.1 116/141] f2fs: avoid out-of-range memory access

[Sasha Levin]

From: Ocean Chen <oceanchen@google.com>

[ Upstream commit 56f3ce675103e3fb9e631cfb4131fc768bc23e9a ]

blkoff_off might over 512 due to fs corrupt or security
vulnerability. That should be checked before being using.

Use ENTRIES_IN_SUM to protect invalid value in cur_data_blkoff.

Signed-off-by: Ocean Chen <oceanchen@google.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/f2fs/segment.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index c7ed3022c4e7..798e33720c74 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -3404,6 +3404,11 @@ static int read_compacted_summaries(struct f2fs_sb_info *sbi)
 		seg_i = CURSEG_I(sbi, i);
 		segno = le32_to_cpu(ckpt->cur_data_segno[i]);
 		blk_off = le16_to_cpu(ckpt->cur_data_blkoff[i]);
+		if (blk_off > ENTRIES_IN_SUM) {
+			f2fs_bug_on(sbi, 1);
+			f2fs_put_page(page, 1);
+			return -EFAULT;
+		}
 		seg_i->next_segno = segno;
 		reset_curseg(sbi, i, 0);
 		seg_i->alloc_type = ckpt->alloc_type[i];
-- 
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel