Re: [slub shrink] 0f6934bf16: +191.9% vmstat.system.cs

2014-01-29 Thread Fengguang Wu
Hi Dave, I got more complete results for https://github.com/hansendc/linux.git slub-reshrink-for-Fengguang-20140113, It looks mostly good. There are regressions in - netperf.Throughput_Mbps - dbench.throughput-MB/sec - xfstests.generic.256.seconds However the others are all improveme

Re: [slub shrink] 0f6934bf16: +191.9% vmstat.system.cs

2014-01-17 Thread Fengguang Wu
Hi Dave, I retested the will-it-scale/read2 case with perf profile enabled, and here are the new comparison results. It shows that there are increased overheads in shmem_getpage_gfp(). If you'd like to collect more data, feel free to tell me. 9a0bb2966efbf30 0f6934bf1695682e7ced973f6 -

Re: [slub shrink] 0f6934bf16: +191.9% vmstat.system.cs

2014-01-16 Thread Fengguang Wu
On Thu, Jan 16, 2014 at 11:12:19AM -0800, Dave Hansen wrote: > On 01/15/2014 07:07 PM, kernel test robot wrote: > > 9a0bb2966efbf30 0f6934bf1695682e7ced973f6 > > --- - > >8235933 ~ 2% +80.6% 14872911 ~ 3% > > lkp-sbx04/micro/will-it-scale/read2

Re: [slub shrink] 0f6934bf16: +191.9% vmstat.system.cs

2014-01-16 Thread Dave Hansen
On 01/15/2014 07:07 PM, kernel test robot wrote: > 9a0bb2966efbf30 0f6934bf1695682e7ced973f6 > --- - >8235933 ~ 2% +80.6% 14872911 ~ 3% > lkp-sbx04/micro/will-it-scale/read2 >8235933 +80.6% 14872911 TOTAL interrupts.RES > >

[slub shrink] 0f6934bf16: +191.9% vmstat.system.cs

2014-01-15 Thread kernel test robot
Hi Dave, We noticed increased context switches in the will-it-scale/read2 test case. Test box is a 4S IVB-EX server. The other change is increased oltp.request_latency_max_ms in a NHM desktop. commit 0f6934bf1695682e7ced973f67d57ab9e124c325 Author: Dave Hansen AuthorDate: Mon Jan 13 07:40:46