netdev - RE: [PATCH net v3] tun: add missing rx stats accounting in tun_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] [thread-next>] [day] [month] [year] [list]

Message-ID: <ed6fd9c514ae4a449580d11c9c6ba8e7@huawei.com>
Date: Thu, 18 Jan 2024 12:55:26 +0000
From: wangyunjian <wangyunjian@...wei.com>
To: Willem de Bruijn <willemdebruijn.kernel@...il.com>, "jasowang@...hat.com"
	<jasowang@...hat.com>, "kuba@...nel.org" <kuba@...nel.org>,
	"davem@...emloft.net" <davem@...emloft.net>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>, xudingke
	<xudingke@...wei.com>
Subject: RE: [PATCH net v3] tun: add missing rx stats accounting in
 tun_xdp_act

> -----Original Message-----
> From: Willem de Bruijn [mailto:willemdebruijn.kernel@...il.com]
> Sent: Wednesday, January 17, 2024 11:42 PM
> To: wangyunjian <wangyunjian@...wei.com>;
> willemdebruijn.kernel@...il.com; jasowang@...hat.com; kuba@...nel.org;
> davem@...emloft.net
> Cc: netdev@...r.kernel.org; linux-kernel@...r.kernel.org; xudingke
> <xudingke@...wei.com>; wangyunjian <wangyunjian@...wei.com>
> Subject: Re: [PATCH net v3] tun: add missing rx stats accounting in tun_xdp_act
> 
> Yunjian Wang wrote:
> > The TUN can be used as vhost-net backend, and it is necessary to count
> > the packets transmitted from TUN to vhost-net/virtio-net. However,
> > there are some places in the receive path that were not taken into
> > account when using XDP. The commit 8ae1aff0b331 ("tuntap: split out
> > XDP logic") only includes dropped counter for XDP_DROP, XDP_ABORTED,
> > and invalid XDP actions. It would be beneficial to also include new
> > accounting for successfully received bytes using
> > dev_sw_netstats_rx_add and introduce new dropped counter for XDP errors
> on XDP_TX and XDP_REDIRECT.
> 
> From the description it is clear that these are two separate changes wrapped
> into one patch. I should have flagged this previously.

Do I need to split these two modifications into 2 patches?
1. only fix dropped counter
2. add new accounting for successfully received bytes
Or:
Only fix dropped counter?

> 
> Ack on returning the error counter that was previously present and matches
> the Fixes tag.
> 
> For the second change, I had to check a few other XDP capable drivers to verify
> that it is indeed common to count such packets.
> 
> > Fixes: 8ae1aff0b331 ("tuntap: split out XDP logic")
> > Signed-off-by: Yunjian Wang <wangyunjian@...wei.com>
> > ---
> > v3: update commit log and code
> > v2: add Fixes tag
> > ---
> >  drivers/net/tun.c | 14 +++++++++-----
> >  1 file changed, 9 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/net/tun.c b/drivers/net/tun.c index
> > afa5497f7c35..0704a17e74e1 100644
> > --- a/drivers/net/tun.c
> > +++ b/drivers/net/tun.c
> > @@ -1625,18 +1625,15 @@ static struct sk_buff *__tun_build_skb(struct
> > tun_file *tfile,  static int tun_xdp_act(struct tun_struct *tun, struct
> bpf_prog *xdp_prog,
> >  		       struct xdp_buff *xdp, u32 act)  {
> > -	int err;
> > +	unsigned int datasize = xdp->data_end - xdp->data;
> > +	int err = 0;
> >
> >  	switch (act) {
> >  	case XDP_REDIRECT:
> >  		err = xdp_do_redirect(tun->dev, xdp, xdp_prog);
> > -		if (err)
> > -			return err;
> >  		break;
> >  	case XDP_TX:
> >  		err = tun_xdp_tx(tun->dev, xdp);
> > -		if (err < 0)
> > -			return err;
> >  		break;
> >  	case XDP_PASS:
> >  		break;
> > @@ -1651,6 +1648,13 @@ static int tun_xdp_act(struct tun_struct *tun,
> struct bpf_prog *xdp_prog,
> >  		break;
> >  	}
> >
> > +	if (err < 0) {
> > +		act = err;
> > +		dev_core_stats_rx_dropped_inc(tun->dev);
> > +	} else if (act == XDP_REDIRECT || act == XDP_TX) {
> > +		dev_sw_netstats_rx_add(tun->dev, datasize);
> > +	}
> > +
> 
> Let's avoid adding yet another branch and just do these operations in the case
> statements, like XDP_DROP.

Fix it like this?
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -1625,18 +1625,25 @@ static struct sk_buff *__tun_build_skb(struct tun_file *tfile,
 static int tun_xdp_act(struct tun_struct *tun, struct bpf_prog *xdp_prog,
                       struct xdp_buff *xdp, u32 act)
 {
+       unsigned int datasize = xdp->data_end - xdp->data;
        int err;
 
        switch (act) {
        case XDP_REDIRECT:
                err = xdp_do_redirect(tun->dev, xdp, xdp_prog);
-               if (err)
+               if (err) {
+                       dev_core_stats_rx_dropped_inc(tun->dev);
                        return err;
+               }
+               dev_sw_netstats_rx_add(tun->dev, datasize);
                break;
        case XDP_TX:
                err = tun_xdp_tx(tun->dev, xdp);
-               if (err < 0)
+               if (err < 0) {
+                       dev_core_stats_rx_dropped_inc(tun->dev);
                        return err;
+               }
+               dev_sw_netstats_rx_add(tun->dev, datasize);
                break;
        case XDP_PASS:

> 
> >  	return act;
> >  }
> >
> > --
> > 2.41.0
> >
> 
>