[PATCH net v2 0/2] ax25: fix reference counting issue of ax25_dev

[PATCH net v2 0/2] ax25: fix reference counting issue of ax25_dev

[Duoming Zhou]

The first patch changes kfree in ax25_dev_free to ax25_dev_put,
because the ax25_dev is managed by reference counting.

The second patch fixes potential reference counting leak issue
in ax25_addr_ax25dev.

You can see the former discussion in the following link:
https://lore.kernel.org/netdev/20240501060218.32898-1-duoming@zju.edu.cn/

Duoming Zhou (2):
  ax25: change kfree in ax25_dev_free to ax25_dev_free
  ax25: fix potential reference counting leak in ax25_addr_ax25dev

 net/ax25/ax25_dev.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

-- 
2.17.1

[PATCH net v2 1/2] ax25: change kfree in ax25_dev_free to ax25_dev_free

[Duoming Zhou]

The ax25_dev is managed by reference counting, so it should not be
deallocated directly by kfree() in ax25_dev_free(), replace it with
ax25_dev_put() instead.

Fixes: d01ffb9eee4a ("ax25: add refcount in ax25_dev to avoid UAF bugs")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
---
 net/ax25/ax25_dev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ax25/ax25_dev.c b/net/ax25/ax25_dev.c
index 282ec581c07..07723095c60 100644
--- a/net/ax25/ax25_dev.c
+++ b/net/ax25/ax25_dev.c
@@ -208,7 +208,7 @@ void __exit ax25_dev_free(void)
 		s        = ax25_dev;
 		netdev_put(ax25_dev->dev, &ax25_dev->dev_tracker);
 		ax25_dev = ax25_dev->next;
-		kfree(s);
+		ax25_dev_put(s);
 	}
 	ax25_dev_list = NULL;
 	spin_unlock_bh(&ax25_dev_lock);
-- 
2.17.1

Re: [PATCH net v2 1/2] ax25: change kfree in ax25_dev_free to ax25_dev_free

[Markus Elfring]

> The ax25_dev is managed by reference counting, so it should not be
> deallocated directly by kfree() in ax25_dev_free(), replace it with
> ax25_dev_put() instead.

You repeated a wording mistake in the summary phrase from a previous cover letter.
Please avoid confusion about desired code replacements.
How do you think about to append parentheses to involved function names?


Would you find the following change description a bit nicer?

   The object “ax25_dev” is managed by reference counting.
   Thus it should not be directly released by a kfree() call in ax25_dev_free().
   Replace it with a ax25_dev_put() call instead.


Would you like to extend patch version descriptions (or changelogs) accordingly?

Regards,
Markus

Re: [PATCH net v2 1/2] ax25: change kfree in ax25_dev_free to ax25_dev_free

[Dan Carpenter]

On Fri, May 03, 2024 at 07:36:54AM +0200, Markus Elfring wrote:
> > The ax25_dev is managed by reference counting, so it should not be
> > deallocated directly by kfree() in ax25_dev_free(), replace it with
> > ax25_dev_put() instead.
> 
> You repeated a wording mistake in the summary phrase from a previous cover letter.

Yeah.  That's true.  The subject should be changed to:

Subject: [PATCH] ax25: change kfree() in ax25_dev_free() to ax25_dev_put()

Another option would be:

Subject: [PATCH] ax25: use ax25_dev_put() in ax25_dev_free()

Otherwise the commit message is okay as-is.

regards,
dan carpenter

[PATCH net v2 2/2] ax25: fix potential reference counting leak in ax25_addr_ax25dev

[Duoming Zhou]

The reference counting of ax25_dev potentially increase more
than once in ax25_addr_ax25dev(), which will cause memory leak.

In order to fix the above issue, only increase the reference
counting of ax25_dev once, when the res is not null.

Fixes: d01ffb9eee4a ("ax25: add refcount in ax25_dev to avoid UAF bugs")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
---
 net/ax25/ax25_dev.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/ax25/ax25_dev.c b/net/ax25/ax25_dev.c
index 07723095c60..7c2ea7309b0 100644
--- a/net/ax25/ax25_dev.c
+++ b/net/ax25/ax25_dev.c
@@ -37,8 +37,9 @@ ax25_dev *ax25_addr_ax25dev(ax25_address *addr)
 	for (ax25_dev = ax25_dev_list; ax25_dev != NULL; ax25_dev = ax25_dev->next)
 		if (ax25cmp(addr, (const ax25_address *)ax25_dev->dev->dev_addr) == 0) {
 			res = ax25_dev;
-			ax25_dev_hold(ax25_dev);
 		}
+	if (res)
+		ax25_dev_hold(res);
 	spin_unlock_bh(&ax25_dev_lock);
 
 	return res;
-- 
2.17.1

Re: [PATCH net v2 2/2] ax25: fix potential reference counting leak in ax25_addr_ax25dev

[Markus Elfring]

How do you think about to append parentheses to the function name
in the summary phrase?


> The reference counting of ax25_dev potentially increase more
> than once in ax25_addr_ax25dev(), which will cause memory leak.
>
> In order to fix the above issue, only increase the reference
> counting of ax25_dev once, when the res is not null.

Would you find the following change description a bit nicer?

   The reference counter of the object “ax25_dev” can be increased multiple times
   in ax25_addr_ax25dev(). This will cause a memory leak so far.

   Thus move a needed function call behind a for loop
   and increase the reference counter only when the local variable “res”
   is not a null pointer.


…
> +++ b/net/ax25/ax25_dev.c
> @@ -37,8 +37,9 @@ ax25_dev *ax25_addr_ax25dev(ax25_address *addr)
…

Would you like to omit curly brackets in the affected function implementation?

Regards,
Markus

Re: [PATCH net v2 2/2] ax25: fix potential reference counting leak in ax25_addr_ax25dev

[Dan Carpenter]

Yeah, it's true that we should delete the curly braces around the if
block.  Otherwise checkpatch.pl -f will complain.

The commit message is fine as-is.  Please stop nit-picking.

regards,
dan carpenter

Re: [v2 2/2] ax25: fix potential reference counting leak in ax25_addr_ax25dev

[Markus Elfring]

> The commit message is fine as-is.  Please stop nit-picking.

I dare to point recurring improvement possibilities out.
I became curious if the change acceptance will ever grow accordingly.

Regards,
Markus

Re: [PATCH net v2 0/2] ax25: fix reference counting issue of ax25_dev

[Dan Carpenter]

On Fri, May 03, 2024 at 07:36:14AM +0800, Duoming Zhou wrote:
> The first patch changes kfree in ax25_dev_free to ax25_dev_put,
> because the ax25_dev is managed by reference counting.
> 
> The second patch fixes potential reference counting leak issue
> in ax25_addr_ax25dev.
> 
> You can see the former discussion in the following link:
> https://lore.kernel.org/netdev/20240501060218.32898-1-duoming@zju.edu.cn/
> 

Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>

regards,
dan carpenter