Re: [PATCH net] RDMA/core: fix UAF in ib_get_eth_speed

[Denis Kirjanov <dkirjanov@suse.de>]

On 3/28/24 14:22, Denis Kirjanov wrote:
> call to ib_device_get_netdev from ib_get_eth_speed
> may lead to a race condition while accessing a netdevice
> instance since we don't hold the rtnl lock while checking
> the registration state:
> 	if (res && res->reg_state != NETREG_REGISTERED) {
> 
> Reported-by: syzbot+5fe14f2ff4ccbace9a26@syzkaller.appspotmail.com
> Fixes: d41861942fc55 ("IB/core: Add generic function to extract IB speed from netdev")

Please ignore, there is an issue on error path

> Signed-off-by: Denis Kirjanov <dkirjanov@suse.de>
> ---
>  drivers/infiniband/core/verbs.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
> index 94a7f3b0c71c..aa4f642e7de9 100644
> --- a/drivers/infiniband/core/verbs.c
> +++ b/drivers/infiniband/core/verbs.c
> @@ -1976,11 +1976,11 @@ int ib_get_eth_speed(struct ib_device *dev, u32 port_num, u16 *speed, u8 *width)
>  	if (rdma_port_get_link_layer(dev, port_num) != IB_LINK_LAYER_ETHERNET)
>  		return -EINVAL;
>  
> +	rtnl_lock();
>  	netdev = ib_device_get_netdev(dev, port_num);
>  	if (!netdev)
>  		return -ENODEV;
>  
> -	rtnl_lock();
>  	rc = __ethtool_get_link_ksettings(netdev, &lksettings);
>  	rtnl_unlock();
>