From: Andreas Hollmann <hollmann@in.tum.de>
To: linux-numa <linux-numa@vger.kernel.org>
Subject: System uses numa Interleave mode as default policy?
Date: Sat, 4 Jan 2014 23:32:30 +0100 [thread overview]
Message-ID: <CAGz0_-3V4TSBn0juiLVfu4yFta8aNsqJYN7jBbi_jP_h2fAxpw@mail.gmail.com> (raw)
Hi,
it seems that my system (4-socket system) is
interleaving pages across all numa nodes.
The default should be local allocation?
I tried this out using the stream benchmark and pinned
the single thread to a single cpu with taskset.
numa_balancing is disabled.
Can I configure the default memory policy or is there
something wrong?
Configuration:
$ cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-linux
root=UUID=294db3db-d278-4b20-9199-f5ac6926efaf rw console=tty0
console=ttyS0,115200 numa_balancing=disable nomodeset quiet
$ uname -a
Linux inwest 3.12.6-1-ARCH #1 SMP PREEMPT Fri Dec 20 19:39:00 CET 2013
x86_64 GNU/Linux
Here is how i produced the output:
wget www.cs.virginia.edu/stream/FTP/Code/stream.c
gcc -mcmodel=medium -O -DSTREAM_ARRAY_SIZE=100000000 stream.c -o stream.100M
taskset -c 0 ./stream.100M 1> /dev/null & watch -n 1 numastat -cp $!
Per-node process memory usage (in MBs) for PID 2055 (stream.100M)
Node 0 Node 1 Node 2 Node 3 Total
------ ------ ------ ------ -----
Huge 0 0 0 0 0
Heap 576 576 569 568 2289
Stack 0 0 0 0 0
Private 0 0 0 0 0
------- ------ ------ ------ ------ -----
Total 576 576 569 568 2289
numactl -l taskset -c 0 ./stream.100M 1> /dev/null & watch -n 1 numastat -cp $!
Per-node process memory usage (in MBs) for PID 2333 (stream.100M)
Node 0 Node 1 Node 2 Node 3 Total
------ ------ ------ ------ -----
Huge 0 0 0 0 0
Heap 576 573 568 572 2289
Stack 0 0 0 0 0
Private 0 0 0 0 0
------- ------ ------ ------ ------ -----
Total 576 573 568 572 2289
numactl -m 0 taskset -c 0 ./stream.100M 1> /dev/null & watch -n 1
numastat -cp $!
Per-node process memory usage (in MBs) for PID 2394 (stream.100M)
Node 0 Node 1 Node 2 Node 3 Total
------ ------ ------ ------ -----
Huge 0 0 0 0 0
Heap 2289 0 0 0 2289
Stack 0 0 0 0 0
Private 0 0 0 0 0
------- ------ ------ ------ ------ -----
Total 2289 0 0 0 2289
---
$ numactl -s 0 taskset -c 0 ./stream.100M
policy: default
preferred node: current
physcpubind: 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 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
cpubind: 0 1 2 3
nodebind: 0 1 2 3
membind: 0 1 2 3
Thanks,
Andreas
reply other threads:[~2014-01-04 22:32 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAGz0_-3V4TSBn0juiLVfu4yFta8aNsqJYN7jBbi_jP_h2fAxpw@mail.gmail.com \
--to=hollmann@in.tum.de \
--cc=linux-numa@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).