netdev - Re: [PATCH net-next] netdevsim: 'support' multi-buf XDP

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] [day] [month] [year] [list]

Message-ID: <Z9AVs2laMDqYPp6S@localhost.localdomain>
Date: Tue, 11 Mar 2025 11:51:31 +0100
From: Michal Kubiak <michal.kubiak@...el.com>
To: Jakub Kicinski <kuba@...nel.org>
CC: <davem@...emloft.net>, <netdev@...r.kernel.org>, <edumazet@...gle.com>,
	<pabeni@...hat.com>, <andrew+netdev@...n.ch>, <horms@...nel.org>,
	<sdf@...ichev.me>
Subject: Re: [PATCH net-next] netdevsim: 'support' multi-buf XDP

On Tue, Mar 11, 2025 at 10:28:20AM +0100, Jakub Kicinski wrote:
> Don't error out on large MTU if XDP is multi-buf.
> The ping test now tests ping with XDP and high MTU.
> netdevsim doesn't actually run the prog (yet?) so
> it doesn't matter if the prog was multi-buf..
> 
> Signed-off-by: Jakub Kicinski <kuba@...nel.org>
> ---
>  drivers/net/netdevsim/bpf.c    | 3 ++-
>  drivers/net/netdevsim/netdev.c | 3 ++-
>  2 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/netdevsim/bpf.c b/drivers/net/netdevsim/bpf.c
> index 608953d4f98d..49537d3c4120 100644
> --- a/drivers/net/netdevsim/bpf.c
> +++ b/drivers/net/netdevsim/bpf.c
> @@ -296,7 +296,8 @@ static int nsim_setup_prog_checks(struct netdevsim *ns, struct netdev_bpf *bpf)
>  		NSIM_EA(bpf->extack, "attempt to load offloaded prog to drv");
>  		return -EINVAL;
>  	}
> -	if (ns->netdev->mtu > NSIM_XDP_MAX_MTU) {
> +	if (bpf->prog && !bpf->prog->aux->xdp_has_frags &&
> +	    ns->netdev->mtu > NSIM_XDP_MAX_MTU) {
>  		NSIM_EA(bpf->extack, "MTU too large w/ XDP enabled");

Would it make sense to extend this error message to indicate that single-buf
XDP is being used? For example: "MTU too large w/ single-buf XDP enabled"?
(Please consider this as a suggestion only.)

>  		return -EINVAL;
>  	}
> diff --git a/drivers/net/netdevsim/netdev.c b/drivers/net/netdevsim/netdev.c
> index d71fd2907cc8..a5e5e064927d 100644
> --- a/drivers/net/netdevsim/netdev.c
> +++ b/drivers/net/netdevsim/netdev.c
> @@ -116,7 +116,8 @@ static int nsim_change_mtu(struct net_device *dev, int new_mtu)
>  {
>  	struct netdevsim *ns = netdev_priv(dev);
>  
> -	if (ns->xdp.prog && new_mtu > NSIM_XDP_MAX_MTU)
> +	if (ns->xdp.prog && !ns->xdp.prog->aux->xdp_has_frags &&
> +	    new_mtu > NSIM_XDP_MAX_MTU)
>  		return -EBUSY;
>  
>  	WRITE_ONCE(dev->mtu, new_mtu);
> -- 
> 2.48.1
> 
> 

Thanks,
Reviewed-by: Michal Kubiak <michal.kubiak@...el.com>