← Back to team overview

kernel-packages team mailing list archive

[Bug 1513980] Re: Ubuntu 14.04.03 did not detect Link down on Houston CU network adapter ports when cable pulled

 

git describe --contains 18824894dbec3eb2202fc92d52a0c8bd27c8a63f
v4.2-rc1~130^2~396

** Also affects: linux (Ubuntu Xenial)
   Importance: Medium
     Assignee: Canonical Kernel Team (canonical-kernel-team)
       Status: Triaged

** Also affects: linux (Ubuntu Wily)
   Importance: Undecided
       Status: New

** Changed in: linux (Ubuntu Wily)
       Status: New => Fix Released

** Changed in: linux (Ubuntu Xenial)
       Status: Triaged => Fix Released

-- 
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1513980

Title:
  Ubuntu 14.04.03 did not detect Link down on Houston CU network adapter
  ports when cable pulled

Status in linux package in Ubuntu:
  Fix Released
Status in linux source package in Wily:
  Fix Released
Status in linux source package in Xenial:
  Fix Released

Bug description:
  Problem Description:
  ====================

  I installed Houston CU (feature code 2CC1) PCIe2 10GbE SFP+ CU 4-port
  Converged Network Adapter in slot P1-C2 of tul117fp1 system. The
  system is configured as PowerNV and running Ubuntu 14.04.03

  - I configured 2 copper ports of the adapter and running network
  stress on it for several hours. I then pulled cable and waited for
  sometime.

  - The OS did not detect the link down. Even though the TX/RX traffic
  stopped on the ports.

  - Kernel log and dmesg did not have any entry to indicate link down
  after cable pulled.

  0007:00:00.0 PCI bridge: IBM Device 03dc
  0007:01:00.0 Ethernet controller: Emulex Corporation OneConnect NIC (Lancer) (rev 10)
  0007:01:00.1 Ethernet controller: Emulex Corporation OneConnect NIC (Lancer) (rev 10)
  0007:01:00.2 Ethernet controller: Emulex Corporation OneConnect NIC (Lancer) (rev 10)
  0007:01:00.3 Ethernet controller: Emulex Corporation OneConnect NIC (Lancer) (rev 10)
  0007:01:00.4 Fibre Channel: Emulex Corporation OneConnect FCoE Initiator (Lancer) (rev 10)
  0007:01:00.5 Fibre Channel: Emulex Corporation OneConnect FCoE Initiator (Lancer) (rev 10)

  - ifconfig -a cmd still show link up

  eth43     Link encap:Ethernet  HWaddr 00:90:fa:6e:e0:f2  
            inet addr:103.1.1.15  Bcast:103.1.1.255  Mask:255.255.255.0
            UP BROADCAST MULTICAST  MTU:1500  Metric:1
            RX packets:7543074880 errors:1 dropped:9 overruns:0 frame:1
            TX packets:5926170683 errors:0 dropped:0 overruns:0 carrier:0
            collisions:0 txqueuelen:1000 
            RX bytes:8997647070924 (8.9 TB)  TX bytes:7177651764733 (7.1 TB)

  eth44     Link encap:Ethernet  HWaddr 00:90:fa:6e:e0:f3  
            inet addr:103.1.2.15  Bcast:103.1.2.255  Mask:255.255.255.0
            UP BROADCAST MULTICAST  MTU:1500  Metric:1
            RX packets:5926170613 errors:1 dropped:0 overruns:0 frame:1
            TX packets:7543089239 errors:0 dropped:0 overruns:0 carrier:0
            collisions:0 txqueuelen:1000 
            RX bytes:7488707854629 (7.4 TB)  TX bytes:8619256135370 (8.6 TB)

  Paul, looking at the source code for the driver in Ubuntu 14.04.3 the
  driver doesn't log any messages after an update to the link status.

  The following commit upstream adds this functionality:

  https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=18824894dbec3eb2202fc92d52a0c8bd27c8a63f

  I'll create a test kernel with this patch applied and will attach it
  here so you can test this commit

  - I installed Ubuntu patched kernel

  - I setup and re-ran the test on Houston adapter. I verified that with
  the patch link Up/Down events now get logged properly when cable
  pulls.

  I verified the patch and it fixed the problem

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1513980/+subscriptions