All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ido Schimmel <idosch@idosch.org>
To: Radu Rendec <rrendec@redhat.com>
Cc: Nikolay Aleksandrov <razor@blackwall.org>,
	Roopa Prabhu <roopa@nvidia.com>,
	bridge@lists.linux.dev, netdev@vger.kernel.org,
	Simon Horman <horms@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Jakub Kicinski <kuba@kernel.org>,
	Eric Dumazet <edumazet@google.com>,
	"David S. Miller" <davem@davemloft.net>
Subject: Re: [PATCH net-next] net/bridge: Add skb drop reasons to the most common drop points
Date: Sun, 15 Dec 2024 14:48:56 +0200	[thread overview]
Message-ID: <Z17QOJZ1DMzoc5Cj@shredder> (raw)
In-Reply-To: <c8ab80bb8e3735d301104f29d7f04275ad054214.camel@redhat.com>

On Fri, Dec 13, 2024 at 03:33:44PM -0500, Radu Rendec wrote:
> On Tue, 2024-12-10 at 11:18 +0200, Nikolay Aleksandrov wrote:
> > On 12/9/24 00:18, Radu Rendec wrote:
> > > The bridge input code may drop frames for various reasons and at various
> > > points in the ingress handling logic. Currently kfree_skb() is used
> > > everywhere, and therefore no drop reason is specified. Add drop reasons
> > > to the most common drop points.
> > > 
> > > The purpose of this patch is to address the most common drop points on
> > > the bridge ingress path. It does not exhaustively add drop reasons to
> > > the entire bridge code. The intention here is to incrementally add drop
> > > reasons to the rest of the bridge code in follow up patches.
> > > 
> > > Most of the skb drop points that are addressed in this patch can be
> > > easily tested by sending crafted packets. The diagram below shows a
> > > simple test configuration, and some examples using `packit`(*) are
> > > also included. The bridge is set up with STP disabled.
> > > (*) https://github.com/resurrecting-open-source-projects/packit
> > > 
> > > The following changes were *not* tested:
> > > * SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT in br_multicast_flood(). I could
> > >   not find an easy way to make a crafted packet get there.
> > > * SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD in br_handle_frame_finish()
> > >   when the port state is BR_STATE_DISABLED, because in that case the
> > >   frame is already dropped in the switch/case block at the end of
> > >   br_handle_frame().
> > > 
> > >     +---+---+
> > >     |  br0  |
> > >     +---+---+
> > >         |
> > >     +---+---+  veth pair  +-------+
> > >     | veth0 +-------------+ xeth0 |
> > >     +-------+             +-------+
> > > 
> > > SKB_DROP_REASON_MAC_INVALID_SOURCE - br_handle_frame()
> > > packit -t UDP -s 192.168.0.1 -d 192.168.0.2 -S 8000 -D 8000 \
> > >   -e 01:22:33:44:55:66 -E aa:bb:cc:dd:ee:ff -c 1 \
> > >   -p '0x de ad be ef' -i xeth0
> > > 
> > > SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL - br_handle_frame()
> > > packit -t UDP -s 192.168.0.1 -d 192.168.0.2 -S 8000 -D 8000 \
> > >   -e 02:22:33:44:55:66 -E 01:80:c2:00:00:01 -c 1 \
> > >   -p '0x de ad be ef' -i xeth0
> > > 
> > > SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD - br_handle_frame()
> > > bridge link set dev veth0 state 0 # disabled
> > > packit -t UDP -s 192.168.0.1 -d 192.168.0.2 -S 8000 -D 8000 \
> > >   -e 02:22:33:44:55:66 -E aa:bb:cc:dd:ee:ff -c 1 \
> > >   -p '0x de ad be ef' -i xeth0
> > > 
> > > SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD - br_handle_frame_finish()
> > > bridge link set dev veth0 state 2 # learning
> > > packit -t UDP -s 192.168.0.1 -d 192.168.0.2 -S 8000 -D 8000 \
> > >   -e 02:22:33:44:55:66 -E aa:bb:cc:dd:ee:ff -c 1 \
> > >   -p '0x de ad be ef' -i xeth0
> > > 
> > > SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT - br_flood()
> > > packit -t UDP -s 192.168.0.1 -d 192.168.0.2 -S 8000 -D 8000 \
> > >   -e 02:22:33:44:55:66 -E aa:bb:cc:dd:ee:ff -c 1 \
> > >   -p '0x de ad be ef' -i xeth0
> > > 
> > > Signed-off-by: Radu Rendec <rrendec@redhat.com>
> > > ---
> > >  include/net/dropreason-core.h | 18 ++++++++++++++++++
> > >  net/bridge/br_forward.c       |  4 ++--
> > >  net/bridge/br_input.c         | 24 +++++++++++++++---------
> > >  3 files changed, 35 insertions(+), 11 deletions(-)
> > > 
> > 
> > Hi,
> > Thanks for working on this, a few comments below.
> 
> Sure, thanks for reviewing! Please see my comments below.
> 
> > > diff --git a/include/net/dropreason-core.h b/include/net/dropreason-core.h
> > > index c29282fabae6..1f2ae5b387c1 100644
> > > --- a/include/net/dropreason-core.h
> > > +++ b/include/net/dropreason-core.h
> > > @@ -108,6 +108,9 @@
> > >  	FN(TUNNEL_TXINFO)		\
> > >  	FN(LOCAL_MAC)			\
> > >  	FN(ARP_PVLAN_DISABLE)		\
> > > +	FN(MAC_IEEE_MAC_CONTROL)	\
> > > +	FN(BRIDGE_INGRESS_PORT_NFWD)	\
> > > +	FN(BRIDGE_NO_EGRESS_PORT)	\
> > >  	FNe(MAX)
> > >  
> > >  /**
> > > @@ -502,6 +505,21 @@ enum skb_drop_reason {
> > >  	 * enabled.
> > >  	 */
> > >  	SKB_DROP_REASON_ARP_PVLAN_DISABLE,
> > > +	/**
> > > +	 * @SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL: the destination MAC address
> > > +	 * is an IEEE MAC Control address.
> > > +	 */
> > > +	SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL,
> > > +	/**
> > > +	 * @SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD: the STP state of the
> > > +	 * ingress bridge port does not allow frames to be forwarded.
> > > +	 */
> > > +	SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD,
> > 
> > Since this is used only when the port state causes the packet to drop, why not
> > rename it to something that suggests it was the state?
> 
> Yes, Ido had a similar suggestion [1], so it's clear that it must be
> renamed. I will go with SKB_DROP_REASON_BRIDGE_INGRESS_STP_STATE in the
> next version, unless you have a better idea.
> 
> > > +	/**
> > > +	 * SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT: no eligible egress port was
> > > +	 * found while attempting to flood the frame.
> > > +	 */
> > > +	SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT,
> > >  	/**
> > >  	 * @SKB_DROP_REASON_MAX: the maximum of core drop reasons, which
> > >  	 * shouldn't be used as a real 'reason' - only for tracing code gen
> > > diff --git a/net/bridge/br_forward.c b/net/bridge/br_forward.c
> > > index e19b583ff2c6..e33e2f4fc3d9 100644
> > > --- a/net/bridge/br_forward.c
> > > +++ b/net/bridge/br_forward.c
> > > @@ -249,7 +249,7 @@ void br_flood(struct net_bridge *br, struct sk_buff *skb,
> > >  
> > >  out:
> > >  	if (!local_rcv)
> > > -		kfree_skb(skb);
> > > +		kfree_skb_reason(skb, SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT);
> > 
> > This is not entirely correct, we can get here if we had an error forwarding
> > the packet to some port, but it may already have been forwarded to others.
> > The reason should distinguish between those two cases.
> 
> I will follow Ido's suggestion [2] and rename SKB_DROP_REASON_VXLAN_NO_REMOTE
> to SKB_DROP_REASON_NO_TX_TARGET, and then use that.
> 
> But it will only cover the case when there are no errors, so I still
> need a different reason for the error case. I looked, and I couldn't
> find an existing one that's close enough, so I think I should create a
> new one. How about SKB_DROP_REASON_TX_ERROR? I would not use "BRIDGE"
> in the name because I'm thinking it may be reused elsewhere, outside
> the bridge module.

AFAICT the only possible error is skb_clone() failure and this is
supposed to be covered by 'SKB_DROP_REASON_NOMEM'.

> 
> > >  }
> > >  
> > >  #ifdef CONFIG_BRIDGE_IGMP_SNOOPING
> > > @@ -349,6 +349,6 @@ void br_multicast_flood(struct net_bridge_mdb_entry *mdst,
> > >  
> > >  out:
> > >  	if (!local_rcv)
> > > -		kfree_skb(skb);
> > > +		kfree_skb_reason(skb, SKB_DROP_REASON_BRIDGE_NO_EGRESS_PORT);
> > 
> > Same comment as above (br_flood).
> > 
> > >  }
> > >  #endif
> > > diff --git a/net/bridge/br_input.c b/net/bridge/br_input.c
> > > index ceaa5a89b947..fc00e172e1e1 100644
> > > --- a/net/bridge/br_input.c
> > > +++ b/net/bridge/br_input.c
> > > @@ -96,8 +96,10 @@ int br_handle_frame_finish(struct net *net, struct sock *sk, struct sk_buff *skb
> > >  	if (br_mst_is_enabled(br)) {
> > >  		state = BR_STATE_FORWARDING;
> > >  	} else {
> > > -		if (p->state == BR_STATE_DISABLED)
> > > -			goto drop;
> > > +		if (p->state == BR_STATE_DISABLED) {
> > > +			kfree_skb_reason(skb, SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD);
> > > +			return 0;
> > > +		}
> > >  
> > >  		state = p->state;
> > >  	}
> > > @@ -155,8 +157,10 @@ int br_handle_frame_finish(struct net *net, struct sock *sk, struct sk_buff *skb
> > >  		}
> > >  	}
> > >  
> > > -	if (state == BR_STATE_LEARNING)
> > > -		goto drop;
> > > +	if (state == BR_STATE_LEARNING) {
> > > +		kfree_skb_reason(skb, SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD);
> > > +		return 0;
> > > +	}>  
> > >  	BR_INPUT_SKB_CB(skb)->brdev = br->dev;
> > >  	BR_INPUT_SKB_CB(skb)->src_port_isolated = !!(p->flags & BR_ISOLATED);
> > > @@ -331,8 +335,10 @@ static rx_handler_result_t br_handle_frame(struct sk_buff **pskb)
> > >  	if (unlikely(skb->pkt_type == PACKET_LOOPBACK))
> > >  		return RX_HANDLER_PASS;
> > >  
> > > -	if (!is_valid_ether_addr(eth_hdr(skb)->h_source))
> > > -		goto drop;
> > > +	if (!is_valid_ether_addr(eth_hdr(skb)->h_source)) {
> > > +		kfree_skb_reason(skb, SKB_DROP_REASON_MAC_INVALID_SOURCE);
> > > +		return RX_HANDLER_CONSUMED;
> > > +	}
> > >  
> > >  	skb = skb_share_check(skb, GFP_ATOMIC);
> > >  	if (!skb)
> > > @@ -374,7 +380,8 @@ static rx_handler_result_t br_handle_frame(struct sk_buff **pskb)
> > >  			return RX_HANDLER_PASS;
> > >  
> > >  		case 0x01:	/* IEEE MAC (Pause) */
> > > -			goto drop;
> > > +			kfree_skb_reason(skb, SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL);
> > > +			return RX_HANDLER_CONSUMED;
> > >  
> > >  		case 0x0E:	/* 802.1AB LLDP */
> > >  			fwd_mask |= p->br->group_fwd_mask;
> > > @@ -423,8 +430,7 @@ static rx_handler_result_t br_handle_frame(struct sk_buff **pskb)
> > >  
> > >  		return nf_hook_bridge_pre(skb, pskb);
> > >  	default:
> > > -drop:
> > > -		kfree_skb(skb);
> > > +		kfree_skb_reason(skb, SKB_DROP_REASON_BRIDGE_INGRESS_PORT_NFWD);
> > >  	}
> > >  	return RX_HANDLER_CONSUMED;
> > >  }
> 
> [1] https://lore.kernel.org/bridge/Z1sLyqZQCjbcCOde@shredder/
> [2] https://lore.kernel.org/bridge/Z1sUsSFfBC9GoiIA@shredder/
> 
> --
> Best regards,
> Radu
> 

  reply	other threads:[~2024-12-15 12:49 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-08 22:18 [PATCH net-next] net/bridge: Add skb drop reasons to the most common drop points Radu Rendec
2024-12-10  9:18 ` Nikolay Aleksandrov
2024-12-12 16:52   ` Ido Schimmel
2024-12-13 20:33   ` Radu Rendec
2024-12-15 12:48     ` Ido Schimmel [this message]
2024-12-12 16:14 ` Ido Schimmel
2024-12-13 20:44   ` Radu Rendec
2024-12-15 12:33     ` Ido Schimmel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Z17QOJZ1DMzoc5Cj@shredder \
    --to=idosch@idosch.org \
    --cc=bridge@lists.linux.dev \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=horms@kernel.org \
    --cc=kuba@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=razor@blackwall.org \
    --cc=roopa@nvidia.com \
    --cc=rrendec@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.