linux-kernel - Re: [PATCH] net: macb: ensure the device is available before accessing GEMGXL control registers

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:   Fri, 21 May 2021 09:51:46 -0700
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Zong Li <zong.li@...ive.com>, nicolas.ferre@...rochip.com,
        claudiu.beznea@...rochip.com, davem@...emloft.net, kuba@...nel.org,
        palmer@...belt.com, paul.walmsley@...ive.com,
        schwab@...ux-m68k.org, sboyd@...nel.org, aou@...s.berkeley.edu,
        mturquette@...libre.com, geert@...ux-m68k.org, yixun.lan@...il.com,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-riscv@...ts.infradead.org
Subject: Re: [PATCH] net: macb: ensure the device is available before
 accessing GEMGXL control registers



On 5/21/2021 5:48 AM, Zong Li wrote:
> If runtime power menagement is enabled, the gigabit ethernet PLL would
> be disabled after macb_probe(). During this period of time, the system
> would hang up if we try to access GEMGXL control registers.
> 
> We can't put runtime_pm_get/runtime_pm_put/ there due to the issue of
> sleep inside atomic section (7fa2955ff70ce453 ("sh_eth: Fix sleeping
> function called from invalid context"). Add the similar flag to ensure
> the device is available before accessing GEMGXL device.
> 
> Signed-off-by: Zong Li <zong.li@...ive.com>
> ---
>  drivers/net/ethernet/cadence/macb.h      | 2 ++
>  drivers/net/ethernet/cadence/macb_main.c | 7 +++++++
>  2 files changed, 9 insertions(+)
> 
> diff --git a/drivers/net/ethernet/cadence/macb.h b/drivers/net/ethernet/cadence/macb.h
> index d8d87213697c..acf5242ce715 100644
> --- a/drivers/net/ethernet/cadence/macb.h
> +++ b/drivers/net/ethernet/cadence/macb.h
> @@ -1309,6 +1309,8 @@ struct macb {
>  
>  	u32	rx_intr_mask;
>  
> +	unsigned int is_opened;
> +
>  	struct macb_pm_data pm_data;
>  	const struct macb_usrio_config *usrio;
>  };
> diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
> index 6bc7d41d519b..e079ed10ad91 100644
> --- a/drivers/net/ethernet/cadence/macb_main.c
> +++ b/drivers/net/ethernet/cadence/macb_main.c
> @@ -2781,6 +2781,8 @@ static int macb_open(struct net_device *dev)
>  	if (bp->ptp_info)
>  		bp->ptp_info->ptp_init(dev);
>  
> +	bp->is_opened = 1;
> +
>  	return 0;
>  
>  reset_hw:
> @@ -2818,6 +2820,8 @@ static int macb_close(struct net_device *dev)
>  	if (bp->ptp_info)
>  		bp->ptp_info->ptp_remove(dev);
>  
> +	bp->is_opened = 0;
> +
>  	pm_runtime_put(&bp->pdev->dev);
>  
>  	return 0;
> @@ -2867,6 +2871,9 @@ static struct net_device_stats *gem_get_stats(struct macb *bp)
>  	struct gem_stats *hwstat = &bp->hw_stats.gem;
>  	struct net_device_stats *nstat = &bp->dev->stats;
>  
> +	if (!bp->is_opened)
> +		return nstat;

The canonical way to do this check is to use netif_running(), and not
open code a boolean tracking whether a network device is opened or not.
-- 
Florian