Public bug reported:
== Comment: #0 - Ping Tian Han <pt...@cn.ibm.com> - 2016-12-07 19:14:40 ==
+++ This bug was initially created as a clone of Bug #148750 +++
On roselp4, we got numastat as this:
% sudo numastat
node0 node1
numa_hit 0 0
numa_miss 0 292054
numa_foreign 0 292054
interleave_hit 0 4100
local_node 0 0
other_node 0 0
% uname -a
Linux roselp4 4.8.0-30-generic #32~16.04.1-Ubuntu SMP Fri Dec 2 03:43:52 UTC
2016 ppc64le ppc64le ppc64le GNU/Linux
%
The number of numa_miss and numa_foreign looks wrong, because the manual says
that "Each numa_miss has a numa_foreign on another node." and "Each
numa_foreign has a numa_miss on another node." But the numbers of node0 are all
zeros.
== Comment: #18 - JIA HE <heji...@cn.ibm.com> - 2017-02-20 00:24:26 ==
hi, Ping Tian Han
Please see the test result, is it what you expected?
before the patches
root@seedlp1:~# numastat
node0 node2 node3
numa_hit 0 0 0
numa_miss 0 352521 5466
numa_foreign 0 352521 5466
interleave_hit 0 5628 5402
local_node 0 0 0
other_node 0 0 0
root@seedlp1:~#
after the patches
xinhui locking
root@seedlp1:/boot# numastat
node0 node2 node3
numa_hit 0 238766 4837
numa_miss 0 0 0
numa_foreign 0 0 0
interleave_hit 0 4888 4777
local_node 0 0 0
other_node 0 238766 4837
== Comment: #19 - JIA HE <heji...@cn.ibm.com> - 2017-02-20 00:28:14 ==
New kernel is at
http://kernel.stglabs.ibm.com/~hejianet/bug149763/linux-image-4.8.0-37-generic_4.8.0-37.39_ppc64el.deb
md5: f3a57af05f0945f10b9d2841844f0fe4 linux-
image-4.8.0-37-generic_4.8.0-37.39bug149763_ppc64el.deb
== Comment: #20 - Ping Tian Han <pt...@cn.ibm.com> - 2017-02-20 00:57:30 ==
(In reply to comment #18)
> hi, Ping Tian Han
> Please see the test result, is it what you expected?
> before the patches
> root@seedlp1:~# numastat
> node0 node2 node3
> numa_hit 0 0 0
> numa_miss 0 352521 5466
> numa_foreign 0 352521 5466
> interleave_hit 0 5628 5402
> local_node 0 0 0
> other_node 0 0 0
> root@seedlp1:~#
>
> after the patches
> xinhui locking
> root@seedlp1:/boot# numastat
> node0 node2 node3
> numa_hit 0 238766 4837
> numa_miss 0 0 0
> numa_foreign 0 0 0
> interleave_hit 0 4888 4777
> local_node 0 0 0
> other_node 0 238766 4837
This looks good to me. Thanks.
Please cherry pick the patches below
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=2df26639e708a88dcc22171949da638a9998f3bc
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=41b6167e8f746b475668f1da78599fc4284f18db
** Affects: ubuntu
Importance: Undecided
Status: New
** Tags: architecture-ppc64le bugnameltc-149763 severity-high
targetmilestone-inin---
** Tags added: architecture-ppc64le bugnameltc-149763 severity-high
targetmilestone-inin---
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1672953
Title:
ISST-LTE:pVM:roselp4:ubuntu16.04.2: number of numa_miss and
numa_foreign wrong in numastat
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+bug/1672953/+subscriptions
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
