[PATCH] ipmi: fix refcount leak in i_ipmi_request()

[PATCH] ipmi: fix refcount leak in i_ipmi_request()

[Wentao Liang]

When a caller provides a `supplied_recv` message to i_ipmi_request(),
the function increments the user's `nr_msgs` reference count. If an
error occurs later, the out_err cleanup path only frees the recv_msg
if the function allocated it itself (i.e., !supplied_recv). In the
supplied_recv case the cleanup is skipped, leaving the reference count
elevated. The caller ipmi_request_supply_msgs() does not release the
supplied_recv on error, so the reference is permanently leaked.

Fix this by explicitly reverting the reference count operations when a
supplied recv_msg with a valid user pointer is present in the error
path: decrement nr_msgs and drop the user's kref.

Cc: stable@vger.kernel.org
Fixes: b52da4054ee0 ("ipmi: Rework user message limit handling")
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
---
 drivers/char/ipmi/ipmi_msghandler.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c
index 869ac87a4b6a..5b9d914cc7a9 100644
--- a/drivers/char/ipmi/ipmi_msghandler.c
+++ b/drivers/char/ipmi/ipmi_msghandler.c
@@ -2347,6 +2347,10 @@ static int i_ipmi_request(struct ipmi_user     *user,
 		if (smi_msg == NULL) {
 			if (!supplied_recv)
 				ipmi_free_recv_msg(recv_msg);
+			else if (recv_msg->user) {
+				atomic_dec(&recv_msg->user->nr_msgs);
+				kref_put(&recv_msg->user->refcount, free_ipmi_user);
+			}
 			return -ENOMEM;
 		}
 	}
@@ -2420,6 +2424,10 @@ static int i_ipmi_request(struct ipmi_user     *user,
 			ipmi_free_smi_msg(smi_msg);
 		if (!supplied_recv)
 			ipmi_free_recv_msg(recv_msg);
+		else if (recv_msg->user) {
+			atomic_dec(&recv_msg->user->nr_msgs);
+			kref_put(&recv_msg->user->refcount, free_ipmi_user);
+		}
 	}
 	return rv;
 }
-- 
2.34.1

Re: [PATCH] ipmi: fix refcount leak in i_ipmi_request()

[Corey Minyard]

On Wed, Jun 03, 2026 at 12:06:34PM +0000, Wentao Liang wrote:
> When a caller provides a `supplied_recv` message to i_ipmi_request(),
> the function increments the user's `nr_msgs` reference count. If an
> error occurs later, the out_err cleanup path only frees the recv_msg
> if the function allocated it itself (i.e., !supplied_recv). In the
> supplied_recv case the cleanup is skipped, leaving the reference count
> elevated. The caller ipmi_request_supply_msgs() does not release the
> supplied_recv on error, so the reference is permanently leaked.
> 
> Fix this by explicitly reverting the reference count operations when a
> supplied recv_msg with a valid user pointer is present in the error
> path: decrement nr_msgs and drop the user's kref.

This looks correct, it's in my next queue.

Thanks,

-corey

> 
> Cc: stable@vger.kernel.org
> Fixes: b52da4054ee0 ("ipmi: Rework user message limit handling")
> Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
> ---
>  drivers/char/ipmi/ipmi_msghandler.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c
> index 869ac87a4b6a..5b9d914cc7a9 100644
> --- a/drivers/char/ipmi/ipmi_msghandler.c
> +++ b/drivers/char/ipmi/ipmi_msghandler.c
> @@ -2347,6 +2347,10 @@ static int i_ipmi_request(struct ipmi_user     *user,
>  		if (smi_msg == NULL) {
>  			if (!supplied_recv)
>  				ipmi_free_recv_msg(recv_msg);
> +			else if (recv_msg->user) {
> +				atomic_dec(&recv_msg->user->nr_msgs);
> +				kref_put(&recv_msg->user->refcount, free_ipmi_user);
> +			}
>  			return -ENOMEM;
>  		}
>  	}
> @@ -2420,6 +2424,10 @@ static int i_ipmi_request(struct ipmi_user     *user,
>  			ipmi_free_smi_msg(smi_msg);
>  		if (!supplied_recv)
>  			ipmi_free_recv_msg(recv_msg);
> +		else if (recv_msg->user) {
> +			atomic_dec(&recv_msg->user->nr_msgs);
> +			kref_put(&recv_msg->user->refcount, free_ipmi_user);
> +		}
>  	}
>  	return rv;
>  }
> -- 
> 2.34.1
>