Subject : netperf ~50% regression with 2.6.33-rc1, bisect to 1b9508f Submitter : Lin Ming <ming.m.lin@intel.com> Date : 2010-01-25 10:03 References : http://marc.info/?l=linux-kernel&m=126441481427331&w=4 Handled-By : Mike Galbraith <efault@gmx.de> Handled-By : Peter Zijlstra <peterz@infradead.org> This entry is being used for tracking a regression from 2.6.. Please don't close it until the problem is fixed in the mainline.
Bisection turned up: commit 1b9508f6831e10d53256825de8904caa22d1ca2c Author: Mike Galbraith <efault@gmx.de> Date: Wed Nov 4 17:53:50 2009 +0100 sched: Rate-limit newidle Rate limit newidle to migration_cost. It's a win for all stages of sysbench oltp tests. Signed-off-by: Mike Galbraith <efault@gmx.de> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu> First-Bad-Commit : 1b9508f6831e10d53256825de8904caa22d1ca2c
On Monday 01 February 2010, Mike Galbraith wrote: > On Mon, 2010-02-01 at 01:22 +0100, Rafael J. Wysocki wrote: > > This message has been generated automatically as a part of a report > > of recent regressions. > > > > The following bug entry is on the current list of known regressions > > from 2.6.32. Please verify if it still should be listed and let me know > > (either way). > > Yes, it should remain open. We're currently waiting for some data from > Lin Ming. The regression itself isn't making much sense.. a kernel with > NEWIDLE disabled should show the same performance, but does not.
Sounds a lot like the bug I reported here: https://bugzilla.redhat.com/show_bug.cgi?id=559943 Intel e1000e NIC in a Sun X6270 blade.
Patch : http://patchwork.kernel.org/patch/78544/ Handled-By : Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Fixed by commit d306ebc28649b89877a22158fe0076f06cc46f60 .