netdev - Re: [PATCH v2 1/2] ag71xx: Handle allocation errors in ag71xx_rings

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

Date:   Mon, 2 Mar 2020 20:22:03 +0100
From:   Oleksij Rempel <linux@...pel-privat.de>
To:     Hauke Mehrtens <hauke@...ke-m.de>, davem@...emloft.net
Cc:     netdev@...r.kernel.org, chris.snook@...il.com, jcliburn@...il.com
Subject: Re: [PATCH v2 1/2] ag71xx: Handle allocation errors in
 ag71xx_rings_init()

Hi Hauke,

Am 02.03.20 um 01:02 schrieb Hauke Mehrtens:
> Free the allocated resources in ag71xx_rings_init() in case
> ag71xx_ring_rx_init() returns an error.
>
> This is only a potential problem, I did not ran into this one.

Thank you for spotting it!

> Signed-off-by: Hauke Mehrtens <hauke@...ke-m.de>
> Fixes: d51b6ce441d3 ("net: ethernet: add ag71xx driver")
> ---
>
> v2:
>  * rebased onm top of "net: ag71xx: port to phylink"
>
>  drivers/net/ethernet/atheros/ag71xx.c | 22 ++++++++++++++++++----
>  1 file changed, 18 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c
> index 02b7705393ca..38683224b70b 100644
> --- a/drivers/net/ethernet/atheros/ag71xx.c
> +++ b/drivers/net/ethernet/atheros/ag71xx.c
> @@ -1169,6 +1169,7 @@ static int ag71xx_rings_init(struct ag71xx *ag)
>  	struct ag71xx_ring *tx = &ag->tx_ring;
>  	struct ag71xx_ring *rx = &ag->rx_ring;
>  	int ring_size, tx_size;
> +	int ret;
>
>  	ring_size = BIT(tx->order) + BIT(rx->order);
>  	tx_size = BIT(tx->order);
> @@ -1181,9 +1182,8 @@ static int ag71xx_rings_init(struct ag71xx *ag)
>  					   ring_size * AG71XX_DESC_SIZE,
>  					   &tx->descs_dma, GFP_KERNEL);
>  	if (!tx->descs_cpu) {
> -		kfree(tx->buf);
> -		tx->buf = NULL;
> -		return -ENOMEM;
> +		ret = -ENOMEM;
> +		goto err_free_buf;
>  	}
>
>  	rx->buf = &tx->buf[tx_size];
> @@ -1191,7 +1191,21 @@ static int ag71xx_rings_init(struct ag71xx *ag)
>  	rx->descs_dma = tx->descs_dma + tx_size * AG71XX_DESC_SIZE;
>
>  	ag71xx_ring_tx_init(ag);
> -	return ag71xx_ring_rx_init(ag);
> +	ret = ag71xx_ring_rx_init(ag);
> +	if (ret)
> +		goto err_free_dma;
> +
> +	return 0;
> +
> +err_free_dma:
> +	dma_free_coherent(&ag->pdev->dev, ring_size * AG71XX_DESC_SIZE,
> +			  tx->descs_cpu, tx->descs_dma);
> +	rx->buf = NULL;
> +err_free_buf:
> +	kfree(tx->buf);
> +	tx->buf = NULL;
> +
> +	return ret;
>  }

Hm... i think it will be better to reuse ag71xx_rings_free() at this place.

>  static void ag71xx_rings_free(struct ag71xx *ag)
>


--
Regards,
Oleksij