syzbot has found a possible deadlock in ocfs2_get_system_file_inode [1].

The scenario is depicted here,

	CPU0					CPU1
lock(&ocfs2_file_ip_alloc_sem_key);
                               lock(&osb->system_file_mutex);
                               lock(&ocfs2_file_ip_alloc_sem_key);
lock(&osb->system_file_mutex);

The function calls which could lead to this are:

CPU0
ocfs2_mknod - lock(&ocfs2_file_ip_alloc_sem_key);
.
.
.
ocfs2_get_system_file_inode - lock(&osb->system_file_mutex);

CPU1 -
ocfs2_file_super - lock(&osb->system_file_mutex);
.
.
.
ocfs2_read_virt_blocks - lock(&ocfs2_file_ip_alloc_sem_key);

This issue can be resolved by making the down_read -> down_read_try
in the ocfs2_read_virt_blocks.

[1] https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd

Reported-and-tested-by: syzbot+e0055ea09f1f5e6fabdd@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd
Signed-off-by: Mohammed Anees <pvmohammedanees2003@gmail.com>
---
v2:
- Remove retries when doing down_read_trylock() and fail directly
---
 fs/ocfs2/extent_map.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletions(-)

diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 70a768b62..48919464a 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -973,7 +973,12 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
 	}
 
 	while (done < nr) {
-		down_read(&OCFS2_I(inode)->ip_alloc_sem);
+		if (!down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem))
+			rc = -EAGAIN;
+		if (rc) {
+			mlog(ML_ERROR, "Resource is temporarily unavailable\n");
+			break;
+		}
 		rc = ocfs2_extent_map_get_blocks(inode, v_block + done,
 						 &p_block, &p_count, NULL);
 		up_read(&OCFS2_I(inode)->ip_alloc_sem);
-- 
2.46.0

On 9/23/24 9:28 PM, Mohammed Anees wrote:
> syzbot has found a possible deadlock in ocfs2_get_system_file_inode [1].
> 
> The scenario is depicted here,
> 
> 	CPU0					CPU1
> lock(&ocfs2_file_ip_alloc_sem_key);
>                                lock(&osb->system_file_mutex);
>                                lock(&ocfs2_file_ip_alloc_sem_key);
> lock(&osb->system_file_mutex);
> 
> The function calls which could lead to this are:
> 
> CPU0
> ocfs2_mknod - lock(&ocfs2_file_ip_alloc_sem_key);
> .
> .
> .
> ocfs2_get_system_file_inode - lock(&osb->system_file_mutex);
> 
> CPU1 -
> ocfs2_file_super - lock(&osb->system_file_mutex);
> .
> .
> .
> ocfs2_read_virt_blocks - lock(&ocfs2_file_ip_alloc_sem_key);
> 
> This issue can be resolved by making the down_read -> down_read_try
> in the ocfs2_read_virt_blocks.
> 
> [1] https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd
> 
> Reported-and-tested-by: syzbot+e0055ea09f1f5e6fabdd@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd
> Signed-off-by: Mohammed Anees <pvmohammedanees2003@gmail.com>
> ---
> v2:
> - Remove retries when doing down_read_trylock() and fail directly
> ---
>  fs/ocfs2/extent_map.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
> index 70a768b62..48919464a 100644
> --- a/fs/ocfs2/extent_map.c
> +++ b/fs/ocfs2/extent_map.c
> @@ -973,7 +973,12 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
>  	}
>  
>  	while (done < nr) {
> -		down_read(&OCFS2_I(inode)->ip_alloc_sem);
> +		if (!down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem))
> +			rc = -EAGAIN;
> +		if (rc) {
> +			mlog(ML_ERROR, "Resource is temporarily unavailable\n");
> +			break;
> +		}

Or could be simplified to:

if (!down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem)) {
	rc = -EAGAIN;
	mlog(ML_ERROR, "Inode #%llu ip_alloc_sem is temporarily unavailable\n",
	     (unsigned long long)OCFS2_I(inode)->ip_blkno);
	break;
}

Thanks,
Joseph

>  		rc = ocfs2_extent_map_get_blocks(inode, v_block + done,
>  						 &p_block, &p_count, NULL);
>  		up_read(&OCFS2_I(inode)->ip_alloc_sem);