[PATCH v2 00/11] Fixes and improvement (including hard affinity!) for Credit2

From: Dario Faggioli <dario.faggioli@citrix.com>
To: xen-devel@lists.xenproject.org
Cc: Juergen Gross <jgross@suse.com>,
	Justin Weaver <jtweaver@hawaii.edu>,
	George Dunlap <george.dunlap@eu.citrix.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	Tianyang Chen <tiche@seas.upenn.edu>,
	Robert VanVossen <robert.vanvossen@dornerworks.com>,
	Uma Sharma <uma.sharma523@gmail.com>,
	Josh Whitehead <josh.whitehead@dornerworks.com>,
	Meng Xu <mengxu@cis.upenn.edu>, Jan Beulich <JBeulich@suse.com>
Subject: [PATCH v2 00/11] Fixes and improvement (including hard affinity!) for Credit2
Date: Wed, 06 Apr 2016 19:22:30 +0200	[thread overview]
Message-ID: <20160406170023.25877.15622.stgit@Solace.fritz.box> (raw)

Hi,

Here's v2 of this series:
 http://lists.xenproject.org/archives/html/xen-devel/2016-03/msg02544.html

It took a bit, but the changes in that tricky logic of switching scheduler for
pCPUs needed thorough re-testing. :-/

In any case, the series is smaller, and a few patches are reviewed and acked
already:
 - it's smaller because I followed George's suggestion and folded a few patches
   into one;
 - the patches that already have acks from the proper maintainers are (they're
   also marked with a 'k' in the git provided summary, at the very bottom):

     xen: sched: implement .init_pdata in Credit, Credit2 and RTDS
     xen: sched: move pCPU initialization in an helper
     xen: sched: improve credit2 bootparams' scope, placement and signedness
     xen: sched: on Credit2, don't reprogram the timer if idle
     xen: sched: fix per-socket runqueue creation in credit2
     xen: sched: per-core runqueues as default in credit2
     xen: sched: implement vcpu hard affinity in Credit2

So, it's really "only" these 4 that are in demand for people's (mostly,
George's) attention:

     xen: sched: make implementing .alloc_pdata optional
     xen: sched: close potential races when switching scheduler to CPUs
     xen: sched: allow for choosing credit2 runqueues configuration at boot
     xen: sched: privde some scratch space for not putting cpumasks on stack

Among which "xen: sched: close potential races when switching scheduler to
CPUs" is the trickiest, and the one that has underwent the most changes
(following review comments). Others, should be pieces of cake. :-P

There's a git branch for the series here:

 git://xenbits.xen.org/people/dariof/xen.git  rel/sched/credit2/fix-runq-and-haff-v2
 http://xenbits.xen.org/gitweb/?p=people/dariof/xen.git;a=shortlog;h=refs/heads/rel/sched/credit2/fix-runq-and-haff-v2

Thanks in advance and Regards,
Dario
---
Dario Faggioli (9):
      xen: sched: make implementing .alloc_pdata optional
    k xen: sched: implement .init_pdata in Credit, Credit2 and RTDS
    k xen: sched: move pCPU initialization in an helper
      xen: sched: close potential races when switching scheduler to CPUs
    k xen: sched: on Credit2, don't reprogram the timer if idle
    k xen: sched: fix per-socket runqueue creation in credit2
      xen: sched: allow for choosing credit2 runqueues configuration at boot
    k xen: sched: per-core runqueues as default in credit2
      xen: sched: privde some scratch space for not putting cpumasks on stack

Justin Weaver (1):
    k xen: sched: implement vcpu hard affinity in Credit2

Uma Sharma (1):
    k xen: sched: improve credit2 bootparams' scope, placement and signedness

 docs/misc/xen-command-line.markdown |   19 ++
 xen/common/sched_arinc653.c         |   31 --
 xen/common/sched_credit.c           |  103 ++++++--
 xen/common/sched_credit2.c          |  436 ++++++++++++++++++++++++-----------
 xen/common/sched_rt.c               |  117 ++++-----
 xen/common/schedule.c               |   76 +++++-
 xen/include/xen/sched-if.h          |    7 +
 7 files changed, 496 insertions(+), 293 deletions(-)
--
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel