** Also affects: linux-azure-edge (Ubuntu)
Importance: Undecided
Status: New
** Changed in: linux-azure-edge (Ubuntu)
Assignee: (unassigned) => Marcelo Cerri (mhcerri)
** Changed in: linux-azure-edge (Ubuntu Xenial)
Assignee: (unassigned) => Marcelo Cerri (mhcerri)
**
** Also affects: linux-azure (Ubuntu Xenial)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading netvsc issue on linux-azure
** Changed in: linux-azure (Ubuntu)
Status: Confirmed => Fix Committed
** Changed in: linux-azure (Ubuntu Xenial)
Status: New => Fix Committed
** Changed in: linux-azure (Ubuntu)
Assignee: (unassigned) => Marcelo Cerri (mhcerri)
** Changed in: linux-azure (Ubuntu Xenial)
https://lists.ubuntu.com/archives/kernel-team/2017-December/088827.html
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading netvsc issue on linux-azure 4.13.0-1001.1
@mhcerri
Thanks for the explanation! I built the kernels directly from the git repo and
I didn't use the binary packages (I thought they should be the same). What's
your repro-rate? Yesterday as Chris tested linux-azure-edge 4.13.0-1004.4 + my
pull request, the repro rate was only ~1%, meaning
Ok. I've tested it again using gen1 VMs and I got the same results. Both
linux-azure and linux-azure-edge show the problem, but for both cases I
need to unload, load and unload the driver again to trigger the issue.
I also tested the test kernel with the pull request and it shows the
same
@decui
The repos are right. I'm asking because some of the kernel under
http://kernel.ubuntu.com/~mhcerri/azure/ are testing kernels.
linux-azure and linux-azure-edge version numbers are not equivalent.
Thus, linux-azure 4.13.0-1001.1 is not equivalent to linux-azure-edge
4.13.0-1001.1. In fact,
@decui
Are you building the kernels directly from the git repo or are you using
binary packages from somewhere?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading
@mhcerri
FYI: I'm not using any kernel from http://kernel.ubuntu.com/~mhcerri/azure/.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading netvsc issue on linux-azure
@mhcerri
I'm using
https://git.launchpad.net/~canonical-kernel/ubuntu/+source/linux-azure/, e.g.
for 4.13.0-1004-azure-edge, I mean:
https://git.launchpad.net/~canonical-kernel/ubuntu/+source/linux-azure/commit/?h=azure-edge-next=21d8a99f88af972684618521cf19adafe24dc566.
Please let me know in
@mhcerri
I only tested Gen1 VM (4.13.0-1004-azure-edge) on my local Hyper-V host (WS
2016), and couldn't repro the issue, i.e. reloading hv_netvsc and changing MTU
worked fine.
As I understand, I think the bug was originally reported against
4.13.0-1001.1 (I did not test this version). That's
I couldn't properly verify the problem on azure. If I execute:
sudo sh -c 'modprobe -r hv_netvsc; modprobe hv_netvsc'
on an azure instance I lost SSH connection, but I don't see the hung
task messages on the boot diagnostics windows. The problem is probably
happening and maybe I'm not getting
Right now only Generation 1 VMs are used in Azure.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading netvsc issue on linux-azure 4.13.0-1001.1
To manage
Also, which specific builds of linux-azure and linux-azure-edge are you
using? Are you using any versions from
http://kernel.ubuntu.com/~mhcerri/azure/?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
Dexuan,
Are you using gen1 or gen2 VMs?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading netvsc issue on linux-azure 4.13.0-1001.1
To manage notifications about
This is indeed very odd. Examining both code bases, we don't see any
significant difference. The source code is the same, as well as the
configs and the module inclusion lists. We're checking if we can
reproduce bug 1736283.
@jpoulson probably knows the answer, but were both linux-azure and
Hrm, interesting. Thanks for the note, I'll circle around with the team
here to see what the discrepancy might be.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
[hyper-v] reloading
@leann Actually I'm not sure if the pull request in bug 1736283 can fix
this bug. It looks 4.13.0-1004-azure-edge has already fixed this bug
somehow.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
After having had a chat with Josh Poulson @Microsoft, it is believed
that the pull request in bu https://bugs.launchpad.net/ubuntu/+source
/linux-azure/+bug/1736283 should likely resolve the issue seen here. We
will post a test kernel in bug 1736283.
--
You received this bug notification
BTW, I tested "modprobe -r hv_netvsc; modprobe hv_netvsc" and "ip link
set dev eth0 mtu 3000; ip link set dev eth0 mtu 4500" with 4.13.0-1004
-azure-edge on WS 2016 (Version 1607, OS build 14393:1943).
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is
I can't reproduce the issue with 4.13.0-1004-azure-edge
(https://git.launchpad.net/~canonical-kernel/ubuntu/+source/linux-
azure/commit/?h=azure-edge-
next=21d8a99f88af972684618521cf19adafe24dc566)
It looks this bug has been fixed by some patch between linux-azure
4.13.0-1001.1 and
Status changed to 'Confirmed' because the bug affects multiple users.
** Changed in: linux-azure (Ubuntu)
Status: New => Confirmed
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1735546
Title:
Adding this issue here, as in the past we saw linked issues between
netvsc reload and mtu change.
The same hung task failure occurs when changing the MTU on the
interface.
Repro steps:
ip link set dev eth0 mtu 3000
ip link set dev eth0 mtu 4500
At this point the command will hang and then run
23 matches
Mail list logo