Re: [PATCH] net/mlx5: DR, Prefer kcalloc over open coded arithmetic

From: Kees Cook <keescook@chromium.org>
To: Len Baker <len.baker@gmx.com>
Cc: Saeed Mahameed <saeedm@nvidia.com>,
	Leon Romanovsky <leon@kernel.org>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Yevgeny Kliteynik <kliteyn@nvidia.com>,
	Alex Vesker <valex@nvidia.com>, Erez Shitrit <erezsh@nvidia.com>,
	Jianbo Liu <jianbol@nvidia.com>,
	netdev@vger.kernel.org, linux-rdma@vger.kernel.org,
	linux-hardening@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] net/mlx5: DR, Prefer kcalloc over open coded arithmetic
Date: Mon, 20 Sep 2021 21:06:35 -0700	[thread overview]
Message-ID: <202109202105.9E901893@keescook> (raw)
In-Reply-To: <20210905074936.15723-1-len.baker@gmx.com>

On Sun, Sep 05, 2021 at 09:49:36AM +0200, Len Baker wrote:
> As noted in the "Deprecated Interfaces, Language Features, Attributes,
> and Conventions" documentation [1], size calculations (especially
> multiplication) should not be performed in memory allocator (or similar)
> function arguments due to the risk of them overflowing. This could lead
> to values wrapping around and a smaller allocation being made than the
> caller was expecting. Using those allocations could lead to linear
> overflows of heap memory and other misbehaviors.
> 
> So, refactor the code a bit to use the purpose specific kcalloc()
> function instead of the argument size * count in the kzalloc() function.
> 
> [1] https://www.kernel.org/doc/html/v5.14/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments
> 
> Signed-off-by: Len Baker <len.baker@gmx.com>
> ---
>  .../net/ethernet/mellanox/mlx5/core/steering/dr_action.c  | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c
> index 6475ba35cf6b..e8957dad3bb1 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c
> @@ -716,6 +716,7 @@ mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn,
>  	struct mlx5dr_action *action;
>  	bool reformat_req = false;
>  	u32 num_of_ref = 0;
> +	u32 ref_act_cnt;
>  	int ret;
>  	int i;
> 
> @@ -724,11 +725,14 @@ mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn,
>  		return NULL;
>  	}
> 
> -	hw_dests = kzalloc(sizeof(*hw_dests) * num_of_dests, GFP_KERNEL);
> +	hw_dests = kcalloc(num_of_dests, sizeof(*hw_dests), GFP_KERNEL);
>  	if (!hw_dests)
>  		return NULL;
> 
> -	ref_actions = kzalloc(sizeof(*ref_actions) * num_of_dests * 2, GFP_KERNEL);
> +	if (unlikely(check_mul_overflow(num_of_dests, 2u, &ref_act_cnt)))
> +		goto free_hw_dests;
> +
> +	ref_actions = kcalloc(ref_act_cnt, sizeof(*ref_actions), GFP_KERNEL);

In the future, consider array3_size(), but this is fine too. :)

-Kees

>  	if (!ref_actions)
>  		goto free_hw_dests;
> 
> --
> 2.25.1
> 

-- 
Kees Cook