Re: kern/58124: lagg(4) fails to work with lacp protocol on mcx(4)

[s ymgch <s.ymgch228%gmail.com@localhost>]

The following reply was made to PR kern/58124; it has been noted by GNATS.

From: s ymgch <s.ymgch228%gmail.com@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost, netbsd-bugs%netbsd.org@localhost
Subject: Re: kern/58124: lagg(4) fails to work with lacp protocol on mcx(4)
Date: Mon, 8 Apr 2024 10:24:55 +0900

 Hi,
 
 Could you check "ifconfig -m mcx0" and "ifconfig -m mcx1"?
 LACP in lagg(4) needs "mediaopt full-duplex" to work, but mcx(4) may
 be missing IFM_FDX.
 
 On Mon, Apr 8, 2024 at 3:05=E2=80=AFAM <vezhlys%gmail.com@localhost> wrote:
 >
 > >Number:         58124
 > >Category:       kern
 > >Synopsis:       lagg(4) fails to work with lacp protocol on mcx(4)
 > >Confidential:   no
 > >Severity:       serious
 > >Priority:       medium
 > >Responsible:    kern-bug-people
 > >State:          open
 > >Class:          sw-bug
 > >Submitter-Id:   net
 > >Arrival-Date:   Sun Apr 07 18:05:00 +0000 2024
 > >Originator:     Andrius V
 > >Release:        NetBSD 10
 > >Organization:
 > >Environment:
 > >Description:
 > Hi,
 >
 > I am trying to configure lagg(4) interface using lacp (802.3ad) protocol =
 between
 > my NAS server (amd64 machine) and Mikrotik CRS309-1G-8S+ router (RouterOS=
 ).
 > My network adapter is QNAP QXG-10G2SF-CX4 (Mellanox ConnectX-4 Lx mcx(4) =
 SFP+ DAC).
 > Unfortunately I can't manage to make it work and communicate with my swit=
 ch or
 > receive IP address from my router's DHCP server.
 >
 > Switch should be configured correctly, I used it for Linux setup for year=
 s, also
 > deprecated agr(4) interface "sometimes" works with certain manual actions=
  applied.
 > agr(4) is unstable though and looses connection soon or after reboot requ=
 iring
 > manual interaction again, thus also unusable in general. It has "LACP Par=
 tner System ID"
 > in the interface, but it stays empty for NetBSD (should be mac address of=
  nas server).
 >
 > I am using dhcpcd DHCP client.
 >
 > Alternatively I tried loadbalance protocol, it works without issues in co=
 mparison.
 >
 > Besides these I have 2x igc(4) interfaces, but I am not using them for li=
 nk aggregation.
 >
 > ifconfig.lagg0
 > create
 > laggproto lacp laggport mcx0 laggport mcx1
 > up
 >
 > /etc/rc.conf
 > ...
 > net_interfaces=3D"igc1 lagg0"
 > dhcpcd=3DYES
 > dhcpcd_flags=3D"-qLM lagg0 igc1"
 > ...
 >
 > ifconfig lagg0
 > lagg0: flags=3D0x8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
 >         capabilities=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM=
 _Tx>
 >         capabilities=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CS=
 UM_Tx>
 >         capabilities=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         enabled=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM_Tx>
 >         enabled=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CSUM_Tx=
 >
 >         enabled=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         ec_capabilities=3D0x7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU>
 >         ec_enabled=3D0x2<VLAN_HWTAGGING>
 >         laggproto lacp
 >         laggport:
 >                 mcx0 pri=3D32768 flags=3D0
 >                 mcx1 pri=3D32768 flags=3D0
 >         address: xx:xx:xx:xx:xx:xx
 >         status: no carrier
 >
 > mcx0: flags=3D0x8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
 >         capabilities=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM=
 _Tx>
 >         capabilities=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CS=
 UM_Tx>
 >         capabilities=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         enabled=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM_Tx>
 >         enabled=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CSUM_Tx=
 >
 >         enabled=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         ec_capabilities=3D0x7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU>
 >         ec_enabled=3D0x2<VLAN_HWTAGGING>
 >         address: xx:xx:xx:xx:xx:xx
 >         media: Ethernet autoselect (10GBASE-CR1)
 >         status: active
 >
 > mcx1: flags=3D0x8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
 >         capabilities=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM=
 _Tx>
 >         capabilities=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CS=
 UM_Tx>
 >         capabilities=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         enabled=3D0x3ff00<IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx,TCP4CSUM_Tx>
 >         enabled=3D0x3ff00<UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx,TCP6CSUM_Tx=
 >
 >         enabled=3D0x3ff00<UDP6CSUM_Rx,UDP6CSUM_Tx>
 >         ec_capabilities=3D0x7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU>
 >         ec_enabled=3D0x2<VLAN_HWTAGGING>
 >         address: xx:xx:xx:xx:xx:xx
 >         media: Ethernet autoselect (10GBASE-CR1)
 >         status: active
 >         link xx:xx:xx:xx:xx:xx
 >
 > dmesg for mcx:
 > mcx0 at pci9 dev 0 function 0: Mellanox Technologies ConnectX-4 Lx (rev. =
 0x00)
 > mcx0: FW 14.22.1002
 > mcx0: Ethernet address xx:xx:xx:xx:xx:xx
 > mcx1 at pci9 dev 0 function 1: Mellanox Technologies ConnectX-4 Lx (rev. =
 0x00)
 > mcx1: FW 14.22.1002
 > mcx1: Ethernet address xx:xx:xx:xx:xx:xx
 >
 > >How-To-Repeat:
 > ifconfig lagg0 create
 > ifconfig lagg0 laggproto lacp laggport mcx0 laggport mcx1
 > ifconfig lagg0 up
 >
 > Start dhcpcd service or run dhcpcd lagg0 and observe that no IP is assign=
 ed, "status: no carrier" is displayed.
 > >Fix:
 >