RE: Regression in reading /proc/stat in the newer kernels withlarge SMP and NUMA configurations

[Oberman, Laurence (HAS GSE)]

Hi Eric,

Just FYI, the same behavior is seen on multiple hardware platforms so it is definitely kernel changes. The earlier kernel on the DL785 platform was returning reads within 60 microseconds.
I have measured it on multiple platforms.

In this case we have a DL980 with HT enabled so you will see 128 CPUS. If I shut off HT, it does not make a difference.

[root@linuxbeast boot]# dmesg | grep nr_cpu_ids
NR_CPUS:4096 nr_cpumask_bits:128 nr_cpu_ids:128 nr_node_ids:8

-----Original Message-----
From: Eric Dumazet [mailto:eric.dumazet@xxxxxxxxx] 
Sent: Friday, October 14, 2011 12:59 PM
To: Oberman, Laurence (HAS GSE)
Cc: Seger, Mark; linux-kernel@xxxxxxxxxxxxxxx; Cabaniols, Sebastien
Subject: RE: Regression in reading /proc/stat in the newer kernels with large SMP and NUMA configurations

Le vendredi 14 octobre 2011 Ã 17:52 +0100, Oberman, Laurence (HAS GSE) a
Ãcrit :
> Standard RHEL kernel build.
> What the customer would be using.
> 
> [root@linuxbeast boot]# cat config-2.6.32-131.6.1.el6.x86_64 | grep NR
> CONFIG_NR_CPUS=4096

Then something must be wrong.... what are your nr_cpu_ids /
nr_cpumask_bits values ?


dmesg | grep nr_cpu_ids



N‹§²æìr¸›yúèšØb²X¬¶ÇvØ^–)Þ{.nÇ+‰·¥Š{±‘êçzX§¶›¡Ü}©ž²ÆzÚ&j:+v‰¨¾«‘êçzZ+€Ê+zf£¢·hšˆ§~†­†Ûiÿûàz¹®w¥¢¸?™¨è­Ú&¢)ßf”ù^jÇy§m…á@A«a¶Úÿ0¶ìh®å’i