linux-kernel - Re: [PATCH v1] net: stmmac: enable MAC after MTL configuring

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]

Message-ID: <20241021130554.00005cf5@gmail.com>
Date: Mon, 21 Oct 2024 13:05:54 +0800
From: Furong Xu <0x1207@...il.com>
To: 2694439648@...com
Cc: alexandre.torgue@...s.st.com, joabreu@...opsys.com, davem@...emloft.net,
 edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
 mcoquelin.stm32@...il.com, netdev@...r.kernel.org,
 linux-stm32@...md-mailman.stormreply.com,
 linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
 hailong.fan@...ngine.com
Subject: Re: [PATCH v1] net: stmmac: enable MAC after MTL configuring

On Mon, 21 Oct 2024 09:03:05 +0800, 2694439648@...com wrote:

> From: "hailong.fan" <hailong.fan@...ngine.com>
> 
> DMA maybe block while ETH is opening,
> Adjust the enable sequence, put the MAC enable last
> 
> For example, ETH is directly connected to the switch,
> which never power down and sends broadcast packets at regular intervals.
> During the process of opening ETH, data may flow into the MTL FIFO,
> once MAC RX is enabled. and then, MTL will be set, such as FIFO size.
> Once enable DMA, There is a certain probability that DMA will read
> incorrect data from MTL FIFO, causing DMA to hang up.
> By read DMA_Debug_Status, you can be observed that the RPS remains at
> a certain value forever. The correct process should be to configure
> MAC/MTL/DMA before enabling DMA/MAC
> 
> Signed-off-by: hailong.fan <hailong.fan@...ngine.com>
> 

A Fixes: tag should be added.

>  static void dwxgmac2_dma_stop_rx(struct stmmac_priv *priv, void __iomem *ioaddr,
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index e21404822..c19ca62a4 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -3437,9 +3437,6 @@ static int stmmac_hw_setup(struct net_device *dev, bool ptp_register)
>  		priv->hw->rx_csum = 0;
>  	}
>  
> -	/* Enable the MAC Rx/Tx */
> -	stmmac_mac_set(priv, priv->ioaddr, true);
> -
>  	/* Set the HW DMA mode and the COE */
>  	stmmac_dma_operation_mode(priv);
>  
> @@ -3523,6 +3520,9 @@ static int stmmac_hw_setup(struct net_device *dev, bool ptp_register)
>  	/* Start the ball rolling... */
>  	stmmac_start_all_dma(priv);
>  
> +	/* Enable the MAC Rx/Tx */
> +	stmmac_mac_set(priv, priv->ioaddr, true);
> +

This sequence fix should be applied to stmmac_xdp_open() too.

>  	stmmac_set_hw_vlan_mode(priv, priv->hw);
>  
>  	return 0;

It is better to split this patch into individual patches, since you are
trying to fix an issue related to several previous commits:
dwmac4, dwxgmac2, stmmac_hw_setup() and stmmac_xdp_open()