All of lore.kernel.org
 help / color / mirror / Atom feed
* [GIT PULL] scheduler updates for v4.15
@ 2017-11-13  8:00 Ingo Molnar
  2017-11-13  9:23 ` Mike Galbraith
  0 siblings, 1 reply; 3+ messages in thread
From: Ingo Molnar @ 2017-11-13  8:00 UTC (permalink / raw)
  To: Linus Torvalds
  Cc: linux-kernel, Peter Zijlstra, Mike Galbraith, Thomas Gleixner,
	Andrew Morton

Linus,

Please pull the latest sched-core-for-linus git tree from:

   git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git sched-core-for-linus

   # HEAD: 765cc3a4b224e22bf524fabe40284a524f37cdd0 sched/core: Optimize sched_feat() for !CONFIG_SCHED_DEBUG builds

The main updates in this cycle were:

 - Group balancing enhancements and cleanups (Brendan Jackman)

 - Move CPU isolation related functionality into its separate 
   kernel/sched/isolation.c file, with related 'housekeeping_*()' namespace and 
   nomenclature et al. (Frederic Weisbecker)

 - Improve the interactive/cpu-intense fairness calculation (Josef Bacik)

 - Improve the PELT code and related cleanups (Peter Zijlstra)

 - Improve the logic of pick_next_task_fair() (Uladzislau Rezki)

 - Improve the RT IPI based balancing logic (Steven Rostedt)

 - Various micro-optimizations:

   - better !CONFIG_SCHED_DEBUG optimizations (Patrick Bellasi)
   - better idle loop (Cheng Jian)

 - ... plus misc fixes, cleanups and updates.

 Thanks,

	Ingo

------------------>
Brendan Jackman (7):
      sched/fair: Sync task util before slow-path wakeup
      sched/fair: Force balancing on NOHZ balance if local group has capacity
      sched/fair: Move select_task_rq_fair() slow-path into its own function
      sched/fair: Remove unnecessary comparison with -1
      sched/fair: Fix find_idlest_group() when local group is not allowed
      sched/fair: Fix usage of find_idlest_group() when no groups are allowed
      sched/fair: Fix usage of find_idlest_group() when the local group is idlest

Cheng Jian (1):
      sched/idle: Micro-optimize the idle loop

Dou Liyang (1):
      x86/tsc: Append the 'tsc=' description for the 'tsc=unstable' boot parameter

Frederic Weisbecker (12):
      sched/isolation: Move housekeeping related code to its own file
      sched/isolation, watchdog: Use housekeeping_cpumask() instead of ad-hoc version
      sched/isolation: Provide a dynamic off-case to housekeeping_any_cpu()
      sched/isolation: Make the housekeeping cpumask private
      sched/isolation: Use its own static key
      sched/isolation: Rename is_housekeeping_cpu() to housekeeping_cpu()
      sched/isolation: Split out new CONFIG_CPU_ISOLATION=y config from CONFIG_NO_HZ_FULL
      sched/isolation: Introduce housekeeping flags
      sched/isolation: Handle the nohz_full= parameter
      sched/isolation: Move isolcpus= handling to the housekeeping code
      sched/isolation: Add basic isolcpus flags
      sched/isolation: Document isolcpus= boot parameter flags, mark it deprecated

Josef Bacik (1):
      sched/fair: Calculate runnable_weight slightly differently

Luca Abeni (1):
      sched/deadline: Fix switching to -deadline

Matthias Kaehlcke (1):
      sched/sysctl: Fix attributes of some extern declarations

Patrick Bellasi (1):
      sched/core: Optimize sched_feat() for !CONFIG_SCHED_DEBUG builds

Peter Zijlstra (20):
      sched/fair: Clean up calc_cfs_shares()
      sched/fair: Add comment to calc_cfs_shares()
      sched/fair: Cure calc_cfs_shares() vs. reweight_entity()
      sched/fair: Remove se->load.weight from se->avg.load_sum
      sched/fair: Change update_load_avg() arguments
      sched/fair: Move enqueue migrate handling
      sched/fair: Rename {en,de}queue_entity_load_avg()
      sched/fair: Introduce {en,de}queue_load_avg()
      sched/fair: More accurate reweight_entity()
      sched/fair: Rewrite cfs_rq->removed_*avg
      sched/fair: Rewrite PELT migration propagation
      sched/fair: Propagate an effective runnable_load_avg
      sched/fair: Implement synchonous PELT detach on load-balance migrate
      sched/fair: Align PELT windows between cfs_rq and its se
      sched/fair: Implement more accurate async detach
      sched/fair: Update calc_group_*() comments
      sched/idle: Move quiet_vmstate() into the NOHZ code
      sched/debug: Rename task-state printing helpers
      sched/deadline: Rename __dl_clear() to __dl_sub()
      sched/topology: Restore SD_PREFER_SIBLING on MC domains

Rakib Mullick (1):
      sched/isolcpus: Fix "isolcpus=" boot parameter handling when !CONFIG_CPUMASK_OFFSTACK

Sebastian Andrzej Siewior (2):
      sched/rt: Add a helper to test for a RT task
      block/ioprio: Use a helper to check for RT prio

Steven Rostedt (Red Hat) (1):
      sched/rt: Simplify the IPI based RT balancing logic

Suravee Suthikulpanit (1):
      sched/topology: Introduce NUMA identity node sched domain

Uladzislau Rezki (1):
      sched/fair: Search a task from the tail of the queue

Vincent Guittot (1):
      sched/fair: Use reweight_entity() for set_user_nice()

luca abeni (2):
      sched/headers: Remove duplicate prototype of __dl_clear_params()
      sched/deadline: Use C bitfields for the state flags


 Documentation/admin-guide/kernel-parameters.txt |   40 +-
 drivers/base/cpu.c                              |   11 +-
 drivers/net/ethernet/tile/tilegx.c              |    6 +-
 fs/proc/array.c                                 |    2 +-
 include/linux/cpumask.h                         |   16 +
 include/linux/ioprio.h                          |    3 +-
 include/linux/sched.h                           |   19 +-
 include/linux/sched/isolation.h                 |   51 ++
 include/linux/sched/rt.h                        |   11 +
 include/linux/sched/sysctl.h                    |    6 +-
 include/linux/tick.h                            |   39 +-
 include/trace/events/sched.h                    |    2 +-
 init/Kconfig                                    |    7 +
 init/main.c                                     |    2 +
 kernel/cgroup/cpuset.c                          |   15 +-
 kernel/rcu/tree_plugin.h                        |    3 +-
 kernel/rcu/update.c                             |    3 +-
 kernel/sched/Makefile                           |    1 +
 kernel/sched/core.c                             |   56 +-
 kernel/sched/deadline.c                         |   21 +-
 kernel/sched/debug.c                            |   18 +-
 kernel/sched/fair.c                             | 1049 +++++++++++++++--------
 kernel/sched/idle.c                             |    4 +-
 kernel/sched/isolation.c                        |  155 ++++
 kernel/sched/rt.c                               |  316 +++----
 kernel/sched/sched.h                            |   73 +-
 kernel/sched/topology.c                         |   49 +-
 kernel/time/tick-sched.c                        |   33 +-
 kernel/trace/trace_output.c                     |   12 +-
 kernel/trace/trace_sched_wakeup.c               |    8 +-
 kernel/watchdog.c                               |   13 +-
 31 files changed, 1270 insertions(+), 774 deletions(-)
 create mode 100644 include/linux/sched/isolation.h
 create mode 100644 kernel/sched/isolation.c

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [GIT PULL] scheduler updates for v4.15
  2017-11-13  8:00 [GIT PULL] scheduler updates for v4.15 Ingo Molnar
@ 2017-11-13  9:23 ` Mike Galbraith
  2017-11-13 15:13   ` Ingo Molnar
  0 siblings, 1 reply; 3+ messages in thread
From: Mike Galbraith @ 2017-11-13  9:23 UTC (permalink / raw)
  To: Ingo Molnar, Linus Torvalds
  Cc: linux-kernel, Peter Zijlstra, Thomas Gleixner, Andrew Morton,
	Steven Rostedt

On Mon, 2017-11-13 at 09:00 +0100, Ingo Molnar wrote:
> 
>  - Improve the RT IPI based balancing logic (Steven Rostedt)

Well how 'bout that, serendipity strikes again:

I just got a report that a proggy Gregory Haskins wrote back in 2007
(preempt-test) was claiming RT was broken, but no, it's every kernel
containing b6366f048e0c.  What the proggy does is to start a gaggle of
RT tasks of descending priority, giving each an identical quantum of
work to do, expecting higher priority tasks to get it done before
lower.  That currently does not always happen, we occasionally do not
get a task pushed for long enough that a task that should have been
preempted by a push shortly after hitting the CPU, manages to remain
long enough to finish well before it should have been able to.

This patch fixed it up before I could even get my troubleshooting
wheels properly smoking (they spin a bit before getting traction).
 Given that, not to mention b6366f048e0c being capable of log jamming
even fairly modest boxen, perhaps this patch wants a stable tag?

	-Mike

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [GIT PULL] scheduler updates for v4.15
  2017-11-13  9:23 ` Mike Galbraith
@ 2017-11-13 15:13   ` Ingo Molnar
  0 siblings, 0 replies; 3+ messages in thread
From: Ingo Molnar @ 2017-11-13 15:13 UTC (permalink / raw)
  To: Mike Galbraith, Greg Kroah-Hartman
  Cc: Linus Torvalds, linux-kernel, Peter Zijlstra, Thomas Gleixner,
	Andrew Morton, Steven Rostedt


* Mike Galbraith <efault@gmx.de> wrote:

> On Mon, 2017-11-13 at 09:00 +0100, Ingo Molnar wrote:
> > 
> >  - Improve the RT IPI based balancing logic (Steven Rostedt)
> 
> Well how 'bout that, serendipity strikes again:
> 
> I just got a report that a proggy Gregory Haskins wrote back in 2007
> (preempt-test) was claiming RT was broken, but no, it's every kernel
> containing b6366f048e0c.  What the proggy does is to start a gaggle of
> RT tasks of descending priority, giving each an identical quantum of
> work to do, expecting higher priority tasks to get it done before
> lower.  That currently does not always happen, we occasionally do not
> get a task pushed for long enough that a task that should have been
> preempted by a push shortly after hitting the CPU, manages to remain
> long enough to finish well before it should have been able to.
> 
> This patch fixed it up before I could even get my troubleshooting
> wheels properly smoking (they spin a bit before getting traction).
>  Given that, not to mention b6366f048e0c being capable of log jamming
> even fairly modest boxen, perhaps this patch wants a stable tag?

Ok, agreed - I've forwarded 4bdced5c9a29 to Greg for -stable inclusion.

Thanks,

	Ingo

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2017-11-13 15:13 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-11-13  8:00 [GIT PULL] scheduler updates for v4.15 Ingo Molnar
2017-11-13  9:23 ` Mike Galbraith
2017-11-13 15:13   ` Ingo Molnar

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.