numa_node_of_cpu() returns warning when cpu_index > 79
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
numactl (Ubuntu) |
Fix Released
|
High
|
Unassigned | ||
Trusty |
Fix Released
|
High
|
Unassigned | ||
Utopic |
Fix Released
|
High
|
Unassigned | ||
Vivid |
Fix Released
|
High
|
Unassigned |
Bug Description
=======
SRU Justification
1. Impact: numactl fails on systems with > 79 cpus
2. Test case: see 'steps to reproduce' below in the description.
3. Regression potential: this is a straight cherrypick of an upstream commit and should be safe. It changes core code so if there were regressions we can't say they'd be contained in any way, but they should not affect other packages
=======
== Comment: #0 - TSAI-YANG JEA <email address hidden> - 2014-08-12 14:52:38 ==
---Problem Description---
In a ppc64le P8 node with SMT 8 mode, numa_node_of_cpu (80) returns following warning:
libnuma: Warning: /sys not mounted or invalid. Assuming one node: No such file or directory
---uname output---
Linux c656f2n01 3.13.0-30-generic #55-Ubuntu SMP Fri Jul 4 21:29:58 UTC 2014 ppc64le ppc64le ppc64le GNU/Linux
Machine Type = Tuletta P8 22L
---Debugger---
A debugger is not configured
---Steps to Reproduce---
Problem can be created with following test case on a P8 LE machine with SMT=8.
#include <dlfcn.h>
#include <stdio.h>
int main(int argc, char *argv[])
{
int dlopen_flags = RTLD_NOW|
int(
int node = 0;
int cpu;
if(argc != 2) {
return -1;
}
cpu = atoi(argv[1]);
printf("%d\n", cpu);
void *handle = dlopen(
if(handle) {
node_of_cpu = (__typeof_
node = node_of_cpu(cpu);
}
}
return 0;
}
To compile the test: gcc test.c -ldl -o test
To run: ./test 80
Other information:
$ numactl -H
available: 4 nodes (0-1,16-17)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 65115 MB
node 0 free: 2955 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 65473 MB
node 1 free: 3313 MB
node 16 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
node 16 size: 65457 MB
node 16 free: 9310 MB
node 17 cpus: 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 17 size: 65209 MB
node 17 free: 10999 MB
node distances:
node 0 1 16 17
0: 10 20 40 40
1: 20 10 40 40
16: 40 40 10 20
17: 40 40 20 10
Userspace tool common name: libnuma
The userspace tool has the following bit modes: 64-bit
== Comment: #4 - Kamalesh Babulal <email address hidden> - 2014-08-19 12:29:15 ==
numa_node_of_cpu
|
|_numa_
and tries to read /sys/devices/
That's the reason why your seeing the warning and wrong information about node. We should be skipping the nodes which are not set in the numa_nodes_ptr
tags: | added: architecture-ppc64le bugnameltc-114573 severity-medium targetmilestone-inin1404 |
Ubuntu Foundations Team Bug Bot (crichton) wrote : | #1 |
tags: | added: bot-comment |
affects: | ubuntu → numactl (Ubuntu) |
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #2 |
------- Comment From <email address hidden> 2014-08-21 14:33 EDT-------
I have changed the source package to numactl in the Launchpad, kindly let us know if you require more information.
tags: |
added: severity-high removed: severity-medium |
bugproxy (bugproxy) wrote : | #3 |
------- Comment From <email address hidden> 2014-11-11 19:13 EDT-------
In the note below you state "Using last trusty ISO, I installed numactl package". Our support requires us to be at 14.04.1, I went out to the Ubuntu numactl page: https:/
It looks like the latest trusty numactl package is numactl 2.0.9~rc5-1ubuntu2 in ppc64el which we have on our system.
Can you elaborate on where the fix is available, if it is not in 14.04.1 can a patch be provided?
Thanks. Mark
bugproxy (bugproxy) wrote : | #4 |
------- Comment From <email address hidden> 2014-11-18 15:49 EDT-------
We have found the issue - On a bare LE system, nodes are name 0, 1 , 16 , 17 and the library search for these nodes for the meminfo file however it search for nodes 0, 1, 2 for the cpumap file.
On a VM started by qemu, the nodes are named 0,1,2,3 and the problem doesn't occures.
In libnuma.c, function numa_node_of_cpu() doesn't use previously created list of valid nodes but use the entire array indexed with the maximum number of nodes available on the system.
Fix is to check node is valid before calling numa_node_
$ diff -urN libnuma.c libnuma.c.new
--- libnuma.c 2014-11-18 10:44:32.563683894 -0500
+++ libnuma.c.new 2014-11-18 10:44:04.431683894 -0500
@@ -1403,10 +1403,12 @@
bmp = numa_bitmask_
nnodes = numa_max_node();
for (node = 0; node <= nnodes; node++){
- numa_node_
- if (numa_bitmask_
- ret = node;
- goto end;
+ if (_getbit(
+ numa_node_
+ if (numa_bitmask_
+ ret = node;
+ goto end;
+ }
}
}
ret = -1;
bugproxy (bugproxy) wrote : new libnuma.so with fix | #5 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #6 |
------- Comment From <email address hidden> 2014-11-18 16:45 EDT-------
Sent patch to the community:
bugproxy (bugproxy) wrote : new libnuma.so with fix | #7 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #8 |
------- Comment From <email address hidden> 2014-11-18 19:13 EDT-------
Hi,
Are you sure the fix is in the new libnuma.so.1?
I copied it to our node and re-tested it. I didn't see any difference. The test still failed.
root@c656f2n01:
63898 51 libnuma.so.1 <---------- the fix libnuma.so.1
12964 51 libnuma.so.1.orig <----------- the original libnuma.so.1
bugproxy (bugproxy) wrote : new libnuma.so with fix | #9 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : | #10 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #11 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #12 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #13 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #14 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #15 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #16 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #17 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #18 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #19 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Changed in numactl (Ubuntu): | |
status: | New → Confirmed |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #20 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #21 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #22 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #23 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #24 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #25 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #26 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
tags: |
added: severity-critical removed: severity-high |
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #27 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #28 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #29 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #30 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #31 |
------- Comment From <email address hidden> 2014-12-08 18:16 EDT-------
Canonical, can we get this included in 15.04 also please?
tags: |
added: targetmilestone-inin14042 removed: targetmilestone-inin1404 |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #32 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #33 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #34 |
------- Comment From <email address hidden> 2014-12-09 16:00 EDT-------
Patch sent to <email address hidden> as commit 0c422464b672cb7
bugproxy (bugproxy) wrote : new libnuma.so with fix | #35 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #36 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #37 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #38 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Breno Leitão (breno-leitao) wrote : | #39 |
Canonical,
Do you agree on this patch? We sent it upstream but didn't received any comment so far.
Changed in numactl (Ubuntu): | |
assignee: | nobody → Taco Screen team (taco-screen-team) |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #40 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #41 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #42 |
------- Comment From <email address hidden> 2015-01-06 16:22 EDT-------
The patch sent to the mailing list was :
Subject: [PATCH] numactl-
Date: Tue, 09 Dec 2014 16:44:57 +0100
From: Thierry <email address hidden> <email address hidden>
To: <email address hidden>, <email address hidden>, <email address hidden>, <email address hidden>
Message-ID: <email address hidden>
References: <email address hidden>
Could you please take this patch in account as it is needed by couple of distros.
Thanks
-------- Forwarded Message --------
Subject: [PATCH] numactl-
Date: Tue, 18 Nov 2014 17:31:41 +0100
From: Thierry <email address hidden> <email address hidden>
To: <email address hidden>, Ian Wienand <email address hidden>
<email address hidden>
Subject: [PATCH] libnuma_
libnuma.so function numa_node_of_cpu() doesn't use previously
created list of valid nodes but use the entire array indexed with the
maximum number of nodes available on the system.
Fix is to check node is valid before calling numa_node_
which check that the cpu belongs to that node.
.
Signed-off-by: Thierry Fauck <email address hidden>
---
Index: numactl-
=======
--- numactl-
+++ numactl-
@@ -1403,10 +1403,12 @@ int numa_node_
bmp = numa_bitmask_
nnodes = numa_max_node();
for (node = 0; node <= nnodes; node++){
- numa_node_
- if (numa_bitmask_
- ret = node;
- goto end;
+ if (_getbit(
+ numa_node_
+ if (numa_bitmask_
+ ret = node;
+ goto end;
+ }
}
}
ret = -1;
--
Thierry Fauck @ linux.vnet.ibm
bugproxy (bugproxy) wrote : new libnuma.so with fix | #43 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #44 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Serge Hallyn (serge-hallyn) wrote : | #45 |
After looking at it for awhile, the patch looks correct to me.
bugproxy (bugproxy) wrote : new libnuma.so with fix | #46 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #47 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #48 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #49 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Breno Leitão (breno-leitao) wrote : | #50 |
Is it possible to have this patch accepted in Ubuntu?
bugproxy (bugproxy) wrote : new libnuma.so with fix | #51 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #52 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Steve Langasek (vorlon) wrote : | #53 |
Serge, as you've reviewed this patch, are you happy to upload this for 15.04?
Changed in numactl (Ubuntu): | |
assignee: | Taco Screen team (taco-screen-team) → Serge Hallyn (serge-hallyn) |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #54 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #55 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Serge Hallyn (serge-hallyn) wrote : | #56 |
@vorlon,
I'll look a bit more and then do that (or reject).
bugproxy (bugproxy) wrote : new libnuma.so with fix | #57 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #58 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Serge Hallyn (serge-hallyn) wrote : Re: [Bug 1358835] Re: numa_node_of_cpu() returns warning when cpu_index > 79 | #59 |
Thierry,
Could you comment on whether this upstream patch
https:/
would suffice for you?
Serge Hallyn (serge-hallyn) wrote : | #60 |
ping
Changed in numactl (Ubuntu): | |
importance: | Undecided → High |
status: | Confirmed → Incomplete |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #61 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #62 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #63 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #64 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #65 |
------- Comment From <email address hidden> 2015-02-24 10:34 EDT-------
Agree that patch mentioned is handling the issue.
Thanks
bugproxy (bugproxy) wrote : new libnuma.so with fix | #66 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #67 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Thierry FAUCK (thierry-j) wrote : | #68 |
Upstream patch is good to me
Changed in numactl (Ubuntu): | |
status: | Incomplete → In Progress |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #69 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #70 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #71 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #72 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Launchpad Janitor (janitor) wrote : | #73 |
This bug was fixed in the package numactl - 2.0.10-1ubuntu1
---------------
numactl (2.0.10-1ubuntu1) vivid; urgency=medium
* cherrypick add-check-
upstream to fix warning when cpu_index > 79 (LP: #1358835)
-- Serge Hallyn <email address hidden> Tue, 24 Feb 2015 23:03:30 -0600
Changed in numactl (Ubuntu): | |
status: | In Progress → Fix Released |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #74 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #75 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #76 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #77 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Changed in numactl (Ubuntu Trusty): | |
importance: | Undecided → High |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #78 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #79 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
description: | updated |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #80 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #81 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : new libnuma.so with fix | #82 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #83 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Adam Conrad (adconrad) wrote : Please test proposed package | #84 |
Hello bugproxy, or anyone else affected,
Accepted numactl into trusty-proposed. The package will build now and be available at https:/
Please help us by testing this new package. See https:/
If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-
Further information regarding the verification process can be found at https:/
Changed in numactl (Ubuntu Trusty): | |
status: | New → Fix Committed |
tags: | added: verification-needed |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #85 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #86 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
Changed in numactl (Ubuntu Utopic): | |
status: | New → Fix Committed |
bugproxy (bugproxy) wrote : new libnuma.so with fix | #87 |
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #88 |
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
tags: |
added: verification-done removed: verification-needed |
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #89 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
tags: |
added: verification-done-trusty verification-done-vivid removed: verification-done |
Adam Conrad (adconrad) wrote : | #91 |
Thierry, can you verify this on utopic as well? That would be great.
bugproxy (bugproxy) wrote : | #92 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
tags: | added: verification-needed-utopic |
Launchpad Janitor (janitor) wrote : | #94 |
This bug was fixed in the package numactl - 2.0.9~rc5-1ubuntu3
---------------
numactl (2.0.9~
* cherrypick add-check-
upstream to fix warning when cpu_index > 79 (LP: #1358835)
-- Serge Hallyn <email address hidden> Thu, 26 Feb 2015 14:17:00 -0600
Changed in numactl (Ubuntu Trusty): | |
status: | Fix Committed → Fix Released |
Brian Murray (brian-murray) wrote : Update Released | #95 |
The verification of the Stable Release Update for numactl has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regressions.
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #96 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
tags: | removed: verification-done-trusty verification-done-vivid verification-needed-utopic |
bugproxy (bugproxy) wrote : | #99 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
bugproxy (bugproxy) wrote : | #100 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
bugproxy (bugproxy) wrote : | #107 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Serge Hallyn (serge-hallyn) wrote : | #108 |
Thierry,
so should the patch in your comment #107 be applied to the package currently in trusty-updates?
Serge Hallyn (serge-hallyn) wrote : | #109 |
I'm confused as previously (between comments #88 and #89) you had marked this as verified. Do you remember if that was with the same version?
bugproxy (bugproxy) wrote : | #110 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Serge Hallyn (serge-hallyn) wrote : | #111 |
I'm still confused - if possible please give a concise answer (with no other context) - is the patch which I am pasting below all that is needed? Should that patch be sent upstream?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly parse %s)", fn);
+ numa_warn(
+ "(cannot open or correctly parse %s)", fn);
+ }
bitmask.maskp = (unsigned long *)mask;
bitmask.size = buflen_needed * 8;
numa_bitmask_
@@ -1355,11 +1357,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
- strerror(errno));
- numa_warn(
- "(cannot open or correctly parse %s)", fn);
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
+ strerror(errno));
+ numa_warn(
+ "(cannot open or correctly parse %s)", fn);
+ }
numa_bitmask_
err = -1;
}
bugproxy (bugproxy) wrote : | #112 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #113 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #114 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #115 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
Launchpad Janitor (janitor) wrote : | #116 |
This bug was fixed in the package numactl - 2.0.9~rc5-1ubuntu3
---------------
numactl (2.0.9~
* cherrypick add-check-
upstream to fix warning when cpu_index > 79 (LP: #1358835)
-- Serge Hallyn <email address hidden> Thu, 26 Feb 2015 14:17:00 -0600
Changed in numactl (Ubuntu Utopic): | |
status: | Fix Committed → Fix Released |
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #117 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #118 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #119 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #120 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
Serge Hallyn (serge-hallyn) wrote : | #121 |
Was the package 2.0.10-1ubuntu3 (in vivid and wily) ever tested? It has an additional upstream patch to fix a potential segv.
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #122 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #123 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #124 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #125 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #126 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #127 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #128 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #129 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #130 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #131 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #132 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #133 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #134 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #135 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #136 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #137 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #138 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : new libnuma.so with fix | #139 |
- new libnuma.so with fix Edit (31.1 KiB, application/x-deb)
------- Comment on attachment From <email address hidden> 2014-11-18 15:55 EDT-------
Please have a test
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #140 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #141 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #142 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Thierry FAUCK (thierry-j) wrote : | #144 |
After moving to wily and using -proposed deb archive, I get:
ubuntu@fenix:~$ apt-cache policy libnuma1
libnuma1:
Installed: 2.0.10-1ubuntu3
Candidate: 2.0.10-1ubuntu3
Version table:
*** 2.0.10-1ubuntu3 0
500 http://
100 /var/lib/
ubuntu@fenix:~$ cd NUMACTL/
ubuntu@
80
cpu 80 belongs to node 16
L$ numactl -H
available: 4 nodes (0-1,16-17)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 32499 MB
node 0 free: 18926 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 32721 MB
node 1 free: 28857 MB
node 16 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
node 16 size: 32721 MB
node 16 free: 28814 MB
node 17 cpus: 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 17 size: 32577 MB
node 17 free: 31739 MB
node distances:
node 0 1 16 17
0: 10 20 40 40
1: 20 10 40 40
16: 40 40 10 20
17: 40 40 20 10
when under 15.04 I got:
$ ./test 80
80
libnuma: Warning: /sys not mounted or invalid. Assuming one node: No such file or directory
cpu 80 belongs to node 16
so it is validated on Wily
tags: |
added: targetmilestone-inin1510 removed: targetmilestone-inin14042 |
bugproxy (bugproxy) wrote : | #145 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
tags: |
added: targetmilestone-inin14042 removed: targetmilestone-inin1510 |
Changed in numactl (Ubuntu Trusty): | |
status: | Fix Released → Confirmed |
Changed in numactl (Ubuntu Utopic): | |
status: | Fix Released → Confirmed |
importance: | Undecided → High |
Changed in numactl (Ubuntu Vivid): | |
status: | New → Confirmed |
importance: | Undecided → High |
Chris J Arges (arges) wrote : Please test proposed package | #149 |
Hello bugproxy, or anyone else affected,
Accepted numactl into trusty-proposed. The package will build now and be available at https:/
Please help us by testing this new package. See https:/
If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-
Further information regarding the verification process can be found at https:/
Changed in numactl (Ubuntu Trusty): | |
status: | Confirmed → Fix Committed |
tags: | added: verification-needed |
Changed in numactl (Ubuntu Utopic): | |
status: | Confirmed → Fix Committed |
Chris J Arges (arges) wrote : | #150 |
Hello bugproxy, or anyone else affected,
Accepted numactl into utopic-proposed. The package will build now and be available at https:/
Please help us by testing this new package. See https:/
If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-
Further information regarding the verification process can be found at https:/
Changed in numactl (Ubuntu Vivid): | |
status: | Confirmed → Fix Committed |
Chris J Arges (arges) wrote : | #151 |
Hello bugproxy, or anyone else affected,
Accepted numactl into vivid-proposed. The package will build now and be available at https:/
Please help us by testing this new package. See https:/
If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-
Further information regarding the verification process can be found at https:/
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #152 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Breno Leitão (breno-leitao) wrote : | #155 |
Still waiting for the fix to be released in 14.04 LTS.
bugproxy (bugproxy) wrote : | #156 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
tags: |
added: verification-done removed: verification-needed |
Launchpad Janitor (janitor) wrote : | #158 |
This bug was fixed in the package numactl - 2.0.10-1ubuntu2.1
---------------
numactl (2.0.10-1ubuntu2.1) vivid-proposed; urgency=medium
* d/p/libnuma-
(LP: #1358835)
-- Serge Hallyn <email address hidden> Thu, 21 May 2015 15:15:21 -0700
Changed in numactl (Ubuntu Vivid): | |
status: | Fix Committed → Fix Released |
Launchpad Janitor (janitor) wrote : | #159 |
This bug was fixed in the package numactl - 2.0.9~rc5-
---------------
numactl (2.0.9~
* d/p/libnuma-
(LP: #1358835)
* d/patches/
due to uninitialised mask. (LP: #1441388)
* d/rules: include simple-patchsys.mk to get patches applied
-- Serge Hallyn <email address hidden> Thu, 21 May 2015 15:30:52 -0700
Changed in numactl (Ubuntu Utopic): | |
status: | Fix Committed → Fix Released |
Launchpad Janitor (janitor) wrote : | #160 |
This bug was fixed in the package numactl - 2.0.9~rc5-
---------------
numactl (2.0.9~
* d/p/libnuma-
(LP: #1358835)
* d/patches/
due to uninitialised mask. (LP: #1441388)
* d/rules: include simple-patchsys.mk to get patches applied
-- Serge Hallyn <email address hidden> Thu, 21 May 2015 15:43:11 -0700
Changed in numactl (Ubuntu Trusty): | |
status: | Fix Committed → Fix Released |
bugproxy (bugproxy) wrote : | #161 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Serge Hallyn (serge-hallyn) wrote : | #166 |
> Has this patch been accepted by maintainers? The Canonical status is fix
By "this patch" d you mean:
libnuma.so On ppc64el, cpu number are not contigous
?
> released. Is there a fix that should be pulled and tested?
As far as I can tell this patch has never been sent to the libnuma
mailing list (I had assumed it had been). It is not in the upstream git
tree. Were you planning on sending it?
The bug is 'Fix Released' because it is fixed in our packages.
If you have no intention of sending the patch to the mailing list please
let us know, and I'll mail it.
*Please* reply, if possible, with a regular email to
<email address hidden> .
bugproxy (bugproxy) wrote : | #167 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : libnuma.so.1 sum 55709 | #169 |
- libnuma.so.1 sum 55709 Edit (31.1 KiB, application/x-deb)
------- Comment (attachment only) From <email address hidden> 2014-11-19 14:45 EDT-------
bugproxy (bugproxy) wrote : Patch for libnuma.c | #170 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #171 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
bugproxy (bugproxy) wrote : Patch for libnuma.c | #174 |
- Patch for libnuma.c Edit (525 bytes, text/plain)
------- Comment (attachment only) From <email address hidden> 2014-12-08 16:02 EDT-------
Changed in numactl (Ubuntu): | |
assignee: | Serge Hallyn (serge-hallyn) → nobody |
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla | #175 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Thierry FAUCK (thierry-j) wrote : | #178 |
Path has been pushed upstream: https:/
bugproxy (bugproxy) wrote : | #179 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Serge Hallyn (serge-hallyn) wrote : | #180 |
The patch in our package has the upstream patch except for one apparent
error - it checks for
if (f)
instead of
if (numa_bitmask_
I think it will be easiest to open a new bug to fix that. I'll note the new bug# here.
bugproxy (bugproxy) wrote : | #181 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Serge Hallyn (serge-hallyn) wrote : | #182 |
Opened bug 1535494
bugproxy (bugproxy) wrote : | #183 |
------- Comment From <email address hidden> 2015-03-05 10:05 EDT-------
Verified:
gcc test.c -ldl -o test
ubuntu@vm19:~$ ./test 80
80
cpu 80 belongs to node 2
ubuntu@vm19:~$ numactl -H
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
node 0 size: 969 MB
node 0 free: 445 MB
node 1 cpus: 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
node 1 size: 1022 MB
node 1 free: 777 MB
node 2 cpus: 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
node 2 size: 1022 MB
node 2 free: 37 MB
node 3 cpus: 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
node 3 size: 1017 MB
node 3 free: 907 MB
node distances:
node 0 1 2 3
0: 10 40 40 40
1: 40 10 40 40
2: 40 40 10 40
3: 40 40 40 10
ubuntu@vm19:~$ cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_
DISTRIB_
DISTRIB_
------- Comment From <email address hidden> 2015-03-19 05:43 EDT-------
The test described in the bug fails in Ubuntu 14.04.2.
The following packages were tried as well: libnuma1 2.0.9~rc5-1ubuntu3 and numactl 2.0.9~rc5-1ubuntu3
In all cases the test failed. The efix for 14.04.1 provided by Thierry Fauck worked but it was not used in 14.04.02
------- Comment From <email address hidden> 2015-03-25 10:41 EDT-------
Any updates on this?
------- Comment From <email address hidden> 2015-04-08 07:54 EDT-------
Commit 573609fb2711b6c
From 573609fb2711b6c
From: Thierry FAUCK - IBM LTC <email address hidden>
Date: Tue, 7 Apr 2015 17:53:34 +0200
Subject: [PATCH] libnuma.so On ppc64el, cpu number are not contigous
Commit 32075635db57c3d
/sys not mounted when it find a non existent cpu number, and that's the
default for multinode on ppc64el bare metal installed system. This patch
removes the message when the /sys entry doesn't exist, but use the return
code.
Signed-off-by: Thierry FAUCK - IBM LTC <email address hidden>
modified: libnuma.c
---
libnuma.c | 22 +++++++
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/libnuma.c b/libnuma.c
index 3717d5b..98aa10f 100644
--- a/libnuma.c
+++ b/libnuma.c
@@ -1276,11 +1276,13 @@ numa_node_
sprintf(fn, "/sys/devices/
f = fopen(fn, "r");
if (!f || getdelim(&line, &len, '\n', f) < 1) {
- numa_warn(
- "/sys not mounted or invalid. Assuming one node: %s",
+ if (f) {
+ numa_warn(
+ "/sys not mounted or invalid. Assuming one node: %s",
strerror(errno));
- numa_warn(
- "(cannot open or correctly pars...
Thank you for taking the time to report this bug and helping to make Ubuntu better. It seems that your bug report is not filed about a specific source package though, rather it is just filed against Ubuntu in general. It is important that bug reports be filed about source packages so that people interested in the package can find the bugs about it. You can find some hints about determining what package your bug might be about at https:/ /wiki.ubuntu. com/Bugs/ FindRightPackag e. You might also ask for help in the #ubuntu-bugs irc channel on Freenode.
To change the source package that this bug is filed about visit https:/ /bugs.launchpad .net/ubuntu/ +bug/1358835/ +editstatus and add the package name in the text box next to the word Package.
[This is an automated message. I apologize if it reached you inappropriately; please just reply to this message indicating so.]