Re: [PATCH] f2fs: Fix deadlock in f2fs_gc() context during atomic files handling

[Jaegeuk Kim]

On 11/13, Sahitya Tummala wrote:
> The FS got stuck in the below stack when the storage is almost
> full/dirty condition (when FG_GC is being done).
> 
> schedule_timeout
> io_schedule_timeout
> congestion_wait
> f2fs_drop_inmem_pages_all
> f2fs_gc
> f2fs_balance_fs
> __write_node_page
> f2fs_fsync_node_pages
> f2fs_do_sync_file
> f2fs_ioctl
> 
> The root cause for this issue is there is a potential infinite loop
> in f2fs_drop_inmem_pages_all() for the case where gc_failure is true
> and when there an inode whose i_gc_failures[GC_FAILURE_ATOMIC] is
> not set. Fix this by keeping track of the total atomic files
> currently opened and using that to exit from this condition.
> 
> Fix-suggested-by: Chao Yu <yuchao0@xxxxxxxxxx>
> Signed-off-by: Chao Yu <yuchao0@xxxxxxxxxx>
> Signed-off-by: Sahitya Tummala <stummala@xxxxxxxxxxxxxx>
> ---
> v2:
> - change fix as per Chao's suggestion
> - decrement sbi->atomic_files protected under sbi->inode_lock[ATOMIC_FILE] and
>   only when atomic flag is cleared for the first time, otherwise, the count
>   goes to an invalid/high value as f2fs_drop_inmem_pages() can be called from
>   two contexts at the same time.
> 
>  fs/f2fs/f2fs.h    |  1 +
>  fs/f2fs/file.c    |  1 +
>  fs/f2fs/segment.c | 21 +++++++++++++++------
>  3 files changed, 17 insertions(+), 6 deletions(-)
> 
> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> index c681f51..e04a665 100644
> --- a/fs/f2fs/f2fs.h
> +++ b/fs/f2fs/f2fs.h
> @@ -1297,6 +1297,7 @@ struct f2fs_sb_info {
>  	unsigned int gc_mode;			/* current GC state */
>  	unsigned int next_victim_seg[2];	/* next segment in victim section */
>  	/* for skip statistic */
> +	unsigned int atomic_files;              /* # of opened atomic file */
>  	unsigned long long skipped_atomic_files[2];	/* FG_GC and BG_GC */
>  	unsigned long long skipped_gc_rwsem;		/* FG_GC only */
>  
> diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
> index f6c038e..22c4949 100644
> --- a/fs/f2fs/file.c
> +++ b/fs/f2fs/file.c
> @@ -1919,6 +1919,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
>  	spin_lock(&sbi->inode_lock[ATOMIC_FILE]);
>  	if (list_empty(&fi->inmem_ilist))
>  		list_add_tail(&fi->inmem_ilist, &sbi->inode_list[ATOMIC_FILE]);
> +	sbi->atomic_files++;
>  	spin_unlock(&sbi->inode_lock[ATOMIC_FILE]);
>  
>  	/* add inode in inmem_list first and set atomic_file */
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index da830fc..0b7a33b 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -288,6 +288,8 @@ void f2fs_drop_inmem_pages_all(struct f2fs_sb_info *sbi, bool gc_failure)
>  	struct list_head *head = &sbi->inode_list[ATOMIC_FILE];
>  	struct inode *inode;
>  	struct f2fs_inode_info *fi;
> +	unsigned int count = sbi->atomic_files;
> +	unsigned int looped = 0;
>  next:
>  	spin_lock(&sbi->inode_lock[ATOMIC_FILE]);
>  	if (list_empty(head)) {
> @@ -296,22 +298,26 @@ void f2fs_drop_inmem_pages_all(struct f2fs_sb_info *sbi, bool gc_failure)
>  	}
>  	fi = list_first_entry(head, struct f2fs_inode_info, inmem_ilist);
>  	inode = igrab(&fi->vfs_inode);
> +	if (inode)
> +		list_move_tail(&fi->inmem_ilist, head);
>  	spin_unlock(&sbi->inode_lock[ATOMIC_FILE]);
>  
>  	if (inode) {
>  		if (gc_failure) {
> -			if (fi->i_gc_failures[GC_FAILURE_ATOMIC])
> -				goto drop;
> -			goto skip;
> +			if (!fi->i_gc_failures[GC_FAILURE_ATOMIC])
> +				goto skip;
>  		}
> -drop:
>  		set_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST);
>  		f2fs_drop_inmem_pages(inode);
> +skip:
>  		iput(inode);
>  	}
> -skip:
>  	congestion_wait(BLK_RW_ASYNC, HZ/50);
>  	cond_resched();
> +	if (gc_failure) {
> +		if (++looped >= count)

There is a race condition when handling sbi->atomic_files?

> +			return;
> +	}
>  	goto next;
>  }
>  
> @@ -327,13 +333,16 @@ void f2fs_drop_inmem_pages(struct inode *inode)
>  		mutex_unlock(&fi->inmem_lock);
>  	}
>  
> -	clear_inode_flag(inode, FI_ATOMIC_FILE);
>  	fi->i_gc_failures[GC_FAILURE_ATOMIC] = 0;
>  	stat_dec_atomic_write(inode);
>  
>  	spin_lock(&sbi->inode_lock[ATOMIC_FILE]);
>  	if (!list_empty(&fi->inmem_ilist))
>  		list_del_init(&fi->inmem_ilist);
> +	if (f2fs_is_atomic_file(inode)) {
> +		clear_inode_flag(inode, FI_ATOMIC_FILE);
> +		sbi->atomic_files--;
> +	}
>  	spin_unlock(&sbi->inode_lock[ATOMIC_FILE]);
>  }
>  
> -- 
> Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc.
> Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.