[PATCH] ocfs2: fix deadlock in ocfs2_get_system_file_inode

[PATCH] ocfs2: fix deadlock in ocfs2_get_system_file_inode

[Kevin Paul Reddy Janagari]

commit: 7bf1823e010e8db2fb649c790bd1b449a75f52d8 upstream

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_fill_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

Link: https://lkml.kernel.org/r/20240924093257.7181-1-pvmohammedanees2003@gmail.com
Signed-off-by: Kevin Paul Reddy Janagari <kevinpaul468@gmail.com>
---
 fs/ocfs2/extent_map.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 70a768b623cf..f7672472fa82 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -973,7 +973,13 @@ 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;
+			mlog(ML_ERROR,
+				 "Inode #%llu ip_alloc_sem is temporarily unavailable\n",
+				 (unsigned long long)OCFS2_I(inode)->ip_blkno);
+			break;
+		}
 		rc = ocfs2_extent_map_get_blocks(inode, v_block + done,
 						 &p_block, &p_count, NULL);
 		up_read(&OCFS2_I(inode)->ip_alloc_sem);
-- 
2.39.5

Re: [PATCH] ocfs2: fix deadlock in ocfs2_get_system_file_inode

[Al Viro]

On Thu, Apr 24, 2025 at 09:59:11PM +0530, Kevin Paul Reddy Janagari wrote:
> This issue can be resolved by making the down_read -> down_read_try
> in the ocfs2_read_virt_blocks.

ITYM "can be papered over by..."; *IF* fill_super() is the only place where
these are taken in such order, this is strictly worse than the current
situation - you are taking a false positive from lockdep (fill_super is
not going to have any other threads accessing the same fs instance) and
adding random failures and syslog spew where none existed.

NAK in that form; it _may_ serve as a stopgap if fill_super is not
the only place where we do it in such order, but even then we need
to deal with the problem properly.

[PATCH] ocfs2: Fix deadlock in ocfs2_get_system_file_inode

[Mohammed Anees]

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

CPU1 -
ocfs2_get_system_file_inode - 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>
---
 fs/ocfs2/extent_map.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 70a768b62..f83d0a3b6 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -12,6 +12,7 @@
 #include <linux/slab.h>
 #include <linux/types.h>
 #include <linux/fiemap.h>
+#include <linux/delay.h>
 
 #include <cluster/masklog.h>
 
@@ -961,6 +962,8 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
 	int rc = 0;
 	u64 p_block, p_count;
 	int i, count, done = 0;
+	int retries, max_retries = 5;
+	int retry_delay_ms = 30;
 
 	trace_ocfs2_read_virt_blocks(
 	     inode, (unsigned long long)v_block, nr, bhs, flags,
@@ -973,7 +976,18 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
 	}
 
 	while (done < nr) {
-		down_read(&OCFS2_I(inode)->ip_alloc_sem);
+		retries = 0;
+		while (retries < max_retries) {
+			if (down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem))
+				break; // Lock acquired
+			msleep(retry_delay_ms);
+			retries++;
+		}
+		if (retries == max_retries) {
+			rc = -EAGAIN;
+			mlog(ML_ERROR, "Cannot acquire lock\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

Re: [PATCH] ocfs2: Fix deadlock in ocfs2_get_system_file_inode

[Joseph Qi]

On 9/22/24 1:50 AM, 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_write_begin - lock(&ocfs2_file_ip_alloc_sem_key);
> .


From the report link, it's ocfs2_mknod(), but not
ocfs2_write_begin().

> .
> .
> ocfs2_get_system_file_inode - lock(&osb->system_file_mutex);
> 
> CPU1 -
> ocfs2_get_system_file_inode - lock(&osb->system_file_mutex);

From the report link, it is in the flow of ocfs2_fill_super().
I'm not sure how it actually happens since user has to mount ocfs2
before doing any operations, e.g. create a file.
Anyway, since many flows will call ocfs2_get_system_file_inode(),
so it will theoretically happen.

> .
> .
> .
> 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>
> ---
>  fs/ocfs2/extent_map.c | 16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
> index 70a768b62..f83d0a3b6 100644
> --- a/fs/ocfs2/extent_map.c
> +++ b/fs/ocfs2/extent_map.c
> @@ -12,6 +12,7 @@
>  #include <linux/slab.h>
>  #include <linux/types.h>
>  #include <linux/fiemap.h>
> +#include <linux/delay.h>
>  
>  #include <cluster/masklog.h>
>  
> @@ -961,6 +962,8 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
>  	int rc = 0;
>  	u64 p_block, p_count;
>  	int i, count, done = 0;
> +	int retries, max_retries = 5;
> +	int retry_delay_ms = 30;
>  
>  	trace_ocfs2_read_virt_blocks(
>  	     inode, (unsigned long long)v_block, nr, bhs, flags,
> @@ -973,7 +976,18 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
>  	}
>  
>  	while (done < nr) {
> -		down_read(&OCFS2_I(inode)->ip_alloc_sem);
> +		retries = 0;
> +		while (retries < max_retries) {
> +			if (down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem))
> +				break; // Lock acquired
> +			msleep(retry_delay_ms);
> +			retries++;
> +		}

I'd like just use down_read_trylock() and fail directly if can't.

Thanks,
Joseph

> +		if (retries == max_retries) {
> +			rc = -EAGAIN;
> +			mlog(ML_ERROR, "Cannot acquire lock\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);