netdev - Re: [PATCH] net: mvpp2: Check for null pcs in mvpp2_acpi

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, 11 Feb 2022 18:18:22 -0600
From:   Jeremy Linton <jeremy.linton@....com>
To:     "Russell King (Oracle)" <linux@...linux.org.uk>
Cc:     netdev@...r.kernel.org, mw@...ihalf.com, davem@...emloft.net,
        kuba@...nel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] net: mvpp2: Check for null pcs in mvpp2_acpi_start()

Hi,

On 2/11/22 17:49, Russell King (Oracle) wrote:
> On Fri, Feb 11, 2022 at 05:42:35PM -0600, Jeremy Linton wrote:
>> Booting a MACCHIATObin with 5.17 the system OOPs with
>> a null pointer deref when the network is started. This
>> is caused by the pcs->ops structure being null on this
>> particular platform/firmware.
> 
> pcs->ops should never be NULL. I'm surprised this fix results in any
> kind of working networking.
> 
> Instead, the initialilsation of port->pcs_*.ops needs to be moved out
> of the if (!mvpp2_use_acpi_compat_mode(..)) block. Please try this:

That appears to fix it as well, shall I re-post this with your fix, or 
will you?

> 
> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> index b45cc7bfcdb5..0fb65940c0a5 100644
> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> @@ -7036,6 +7036,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
>   	dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
>   	dev->dev.of_node = port_node;
>   
> +	port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> +	port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> +
>   	if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
>   		port->phylink_config.dev = &dev->dev;
>   		port->phylink_config.type = PHYLINK_NETDEV;
> @@ -7106,9 +7109,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
>   				  port->phylink_config.supported_interfaces);
>   		}
>   
> -		port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> -		port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> -
>   		phylink = phylink_create(&port->phylink_config, port_fwnode,
>   					 phy_mode, &mvpp2_phylink_ops);
>   		if (IS_ERR(phylink)) {
> 
> Thanks.
> 

I did a bit more testing and as a side note, it seems ethtool tosses 
these errors too.


netlink error: failed to retrieve link settings
netlink error: Unknown error 524
netlink error: failed to retrieve link settings
netlink error: Unknown error 524
Settings for enamrvl110i0:
         Link detected: no