public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
From: Ido Schimmel <idosch@idosch.org>
To: Ziyang Xuan <william.xuanziyang@huawei.com>
Cc: davem@davemloft.net, edumazet@google.com, kuba@kernel.org,
	pabeni@redhat.com, bigeasy@linutronix.de,
	wsa+renesas@sang-engineering.com, kaber@trash.net,
	netdev@vger.kernel.org
Subject: Re: [PATCH net] vlan: Fix to delete vid only when by_dev has hw filter capable in vlan_vids_del_by_dev()
Date: Tue, 1 Aug 2023 17:07:57 +0300	[thread overview]
Message-ID: <ZMkRvQFwMdjmcCh0@shredder> (raw)
In-Reply-To: <20230801095943.3650567-1-william.xuanziyang@huawei.com>

On Tue, Aug 01, 2023 at 05:59:43PM +0800, Ziyang Xuan wrote:
> diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
> index 0beb44f2fe1f..79cf4f033b66 100644
> --- a/net/8021q/vlan_core.c
> +++ b/net/8021q/vlan_core.c
> @@ -436,8 +436,11 @@ void vlan_vids_del_by_dev(struct net_device *dev,
>  	if (!vlan_info)
>  		return;
>  
> -	list_for_each_entry(vid_info, &vlan_info->vid_list, list)
> +	list_for_each_entry(vid_info, &vlan_info->vid_list, list) {
> +		if (!vlan_hw_filter_capable(by_dev, vid_info->proto))
> +			continue;
>  		vlan_vid_del(dev, vid_info->proto, vid_info->vid);

vlan_vids_add_by_dev() does not have this check which means that memory
will leak [1] if the device is enslaved after the bond already has a
VLAN upper.

I believe the correct solution is to explicitly set the STAG-related
features [3] in the bond driver in a similar fashion to how it's done
for the CTAG features. That way the bond driver will always propagate
the VLAN info to its slaves.

Please check the team driver as well. I think it's suffering from the
same problem. If so, please fix it in a separate patch.

[1]
unreferenced object 0xffff888103efbd00 (size 256):
  comm "ip", pid 351, jiffies 4294763177 (age 19.697s)
  hex dump (first 32 bytes):                                                          
    00 10 7a 11 81 88 ff ff 00 00 00 00 00 00 00 00  ..z.............
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:                                                                          
    [<ffffffff81a88c0a>] kmalloc_trace+0x2a/0xe0                     
    [<ffffffff840f349c>] vlan_vid_add+0x30c/0x790
    [<ffffffff840f4a68>] vlan_vids_add_by_dev+0x148/0x390
    [<ffffffff82eea5e8>] bond_enslave+0xaf8/0x5d50
    [<ffffffff837fbfd1>] do_set_master+0x1c1/0x220       
    [<ffffffff8380711c>] do_setlink+0xa0c/0x3fa0  
    [<ffffffff8381ee09>] __rtnl_newlink+0xc09/0x18c0
    [<ffffffff8381fb2c>] rtnl_newlink+0x6c/0xa0 
    [<ffffffff8381d4ee>] rtnetlink_rcv_msg+0x43e/0xe00
    [<ffffffff83a2c920>] netlink_rcv_skb+0x170/0x440
    [<ffffffff83a2a2cf>] netlink_unicast+0x53f/0x810  
    [<ffffffff83a2af0b>] netlink_sendmsg+0x96b/0xe90
    [<ffffffff83708e0f>] ____sys_sendmsg+0x30f/0xa70
    [<ffffffff837129fa>] ___sys_sendmsg+0x13a/0x1e0 
    [<ffffffff83712c6c>] __sys_sendmsg+0x11c/0x1f0  
    [<ffffffff842fe5c8>] do_syscall_64+0x38/0x80   
unreferenced object 0xffff888112ffab60 (size 32):
  comm "ip", pid 351, jiffies 4294763177 (age 19.697s)
  hex dump (first 32 bytes):
    a0 bd ef 03 81 88 ff ff a0 bd ef 03 81 88 ff ff  ................
    88 a8 0a 00 01 00 00 00 cc cc cc cc cc cc cc cc  ................
  backtrace:
    [<ffffffff81a88c0a>] kmalloc_trace+0x2a/0xe0
    [<ffffffff840f3599>] vlan_vid_add+0x409/0x790
    [<ffffffff840f4a68>] vlan_vids_add_by_dev+0x148/0x390
    [<ffffffff82eea5e8>] bond_enslave+0xaf8/0x5d50
    [<ffffffff837fbfd1>] do_set_master+0x1c1/0x220
    [<ffffffff8380711c>] do_setlink+0xa0c/0x3fa0
    [<ffffffff8381ee09>] __rtnl_newlink+0xc09/0x18c0
    [<ffffffff8381fb2c>] rtnl_newlink+0x6c/0xa0
    [<ffffffff8381d4ee>] rtnetlink_rcv_msg+0x43e/0xe00
    [<ffffffff83a2c920>] netlink_rcv_skb+0x170/0x440
    [<ffffffff83a2a2cf>] netlink_unicast+0x53f/0x810
    [<ffffffff83a2af0b>] netlink_sendmsg+0x96b/0xe90
    [<ffffffff83708e0f>] ____sys_sendmsg+0x30f/0xa70
    [<ffffffff837129fa>] ___sys_sendmsg+0x13a/0x1e0
    [<ffffffff83712c6c>] __sys_sendmsg+0x11c/0x1f0
    [<ffffffff842fe5c8>] do_syscall_64+0x38/0x80

[2]
#!/bin/bash

ip link add name dummy1 type dummy
ip link add bond1 type bond mode 802.3ad
ip link add link bond1 name bond1.10 type vlan id 10 protocol 802.1ad
ip link set dev dummy1 master bond1
ip link del dev dummy1

[3]
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 484c9e3e5e82..447b06ea4fc9 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -5901,7 +5901,9 @@ void bond_setup(struct net_device *bond_dev)
 
        bond_dev->hw_features = BOND_VLAN_FEATURES |
                                NETIF_F_HW_VLAN_CTAG_RX |
-                               NETIF_F_HW_VLAN_CTAG_FILTER;
+                               NETIF_F_HW_VLAN_CTAG_FILTER |
+                               NETIF_F_HW_VLAN_STAG_RX |
+                               NETIF_F_HW_VLAN_STAG_FILTER;
 
        bond_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL;
        bond_dev->features |= bond_dev->hw_features;

> +	}
>  }
>  EXPORT_SYMBOL(vlan_vids_del_by_dev);
>  
> -- 
> 2.25.1
> 
> 

  reply	other threads:[~2023-08-01 14:08 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-01  9:59 [PATCH net] vlan: Fix to delete vid only when by_dev has hw filter capable in vlan_vids_del_by_dev() Ziyang Xuan
2023-08-01 14:07 ` Ido Schimmel [this message]
2023-08-02 10:04   ` Ziyang Xuan (William)

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=ZMkRvQFwMdjmcCh0@shredder \
    --to=idosch@idosch.org \
    --cc=bigeasy@linutronix.de \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kaber@trash.net \
    --cc=kuba@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=william.xuanziyang@huawei.com \
    --cc=wsa+renesas@sang-engineering.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox