linux-kernel - Re: [PATCH] net/mlx5: Add vport return value checks

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20211231224615.oneibk4ks5wofgf7@sx1>
Date:   Fri, 31 Dec 2021 14:46:15 -0800
From:   Saeed Mahameed <saeed@...nel.org>
To:     Zizhuang Deng <sunsetdzz@...il.com>, mbloch@...dia.com
Cc:     saeedm@...dia.com, leon@...nel.org, davem@...emloft.net,
        kuba@...nel.org, netdev@...r.kernel.org,
        linux-rdma@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] net/mlx5: Add vport return value checks

On Thu, Dec 30, 2021 at 01:25:58PM +0800, Zizhuang Deng wrote:
>add missing vport return value checks for recent code, as in [1].
>
>Ref:
>[1] https://lkml.org/lkml/2020/11/1/315
>

Where did this patch come from ? real bug ? or just aligning the code to
be according the link below ? 
because all the use-cases below are supposed to be guaranteed to have a
valid vport object for uplink/pf/and ecpf vportrs, i am not against the
patch, I am just trying to understand if there is a hidden bug somewhere .. 



>Signed-off-by: Zizhuang Deng <sunsetdzz@...il.com>
>---
> .../mellanox/mlx5/core/eswitch_offloads.c     | 20 +++++++++++++++++++
> 1 file changed, 20 insertions(+)
>
>diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
>index f4eaa5893886..fda214021738 100644
>--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
>+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c

[...]

>@@ -1309,11 +1317,15 @@ static void esw_del_fdb_peer_miss_rules(struct mlx5_eswitch *esw)
>
> 	if (mlx5_ecpf_vport_exists(esw->dev)) {
> 		vport = mlx5_eswitch_get_vport(esw, MLX5_VPORT_ECPF);
>+		if (IS_ERR(vport))
>+			return;

memleak, we need to hit kvfree(flows) below, 
instead of returning you should make the del_flow conditional and continue
to next steps in the esw_del_fdb_peer_miss_rules routine, 
e.g:
if (vport)
	mlx5_del_flow_rules(flows[vport->index]);

> 		mlx5_del_flow_rules(flows[vport->index]);
> 	}
>
> 	if (mlx5_core_is_ecpf_esw_manager(esw->dev)) {
> 		vport = mlx5_eswitch_get_vport(esw, MLX5_VPORT_PF);
>+		if (IS_ERR(vport))
>+			return;

ditto 

> 		mlx5_del_flow_rules(flows[vport->index]);
> 	}
> 	kvfree(flows);
>@@ -2385,6 +2397,9 @@ static int esw_set_uplink_slave_ingress_root(struct mlx5_core_dev *master,
> 	if (master) {
> 		esw = master->priv.eswitch;
> 		vport = mlx5_eswitch_get_vport(esw, MLX5_VPORT_UPLINK);
>+		if (IS_ERR(vport))
>+			return PTR_ERR(vport);
>+
> 		MLX5_SET(set_flow_table_root_in, in, table_of_other_vport, 1);
> 		MLX5_SET(set_flow_table_root_in, in, table_vport_number,
> 			 MLX5_VPORT_UPLINK);
>@@ -2405,6 +2420,9 @@ static int esw_set_uplink_slave_ingress_root(struct mlx5_core_dev *master,
> 	} else {
> 		esw = slave->priv.eswitch;
> 		vport = mlx5_eswitch_get_vport(esw, MLX5_VPORT_UPLINK);
>+		if (IS_ERR(vport))
>+			return PTR_ERR(vport);
>+
> 		ns = mlx5_get_flow_vport_acl_namespace(slave,
> 						       MLX5_FLOW_NAMESPACE_ESW_INGRESS,
> 						       vport->index);
>@@ -2590,6 +2608,8 @@ static void esw_unset_master_egress_rule(struct mlx5_core_dev *dev)
>
> 	vport = mlx5_eswitch_get_vport(dev->priv.eswitch,
> 				       dev->priv.eswitch->manager_vport);
>+	if (IS_ERR(vport))
>+		return;
>
> 	esw_acl_egress_ofld_cleanup(vport);
> }
>-- 
>2.25.1
>