[Bug 102334] [BXT]15% performance drop in SynMark v7.0 Multithread use-case

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=102334

Chris Wilson <chris at chris-wilson.co.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |NEEDINFO

--- Comment #6 from Chris Wilson <chris at chris-wilson.co.uk> ---
Gave it a cursory test and it appears to have fixed this particular regression,
so pushed to topic/core-for-CI for wider testing:

commit 367a3aa35e7bed10a74bed139f6728916f5d1508
Author: Peter Zijlstra <peterz at infradead.org>
Date:   Tue Aug 1 23:43:07 2017 +0200

    sched/fair: Fix wake_affine() for !NUMA_BALANCING

It should only affect !llc platforms, but the logic is slightly different than
the original prior to

commit 3fed382b46baac83703130fe4cd3d9147f427fb9
Author: Rik van Riel <riel at redhat.com>
Date:   Fri Jun 23 12:55:29 2017 -0400

    sched/numa: Implement NUMA node level wake_affine()

so testing definitely required.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20170824/377a7f4d/attachment.html>