Skip to content

sched/mm: LRU drain flush on nohz_full

Frederic Weisbecker linux-mm lkml 4 versions – In review

Get series with b4
b4 am [email protected]

Versions newest first · linked by author and title

  1. v4 Latest6 patches 9 repliesIn review
  2. v36 patches 7 repliesSuperseded
  3. v26 patches 12 repliesSuperseded
Earlier versions · 1
VersionPostedPatchesRepliesStatus
v1 RFC6 patches15 repliesSuperseded

Changes in v4 from the cover letter

* Apply review from Oleg and K Prateek Nayak (handle io_uring kthreads and use guard)

* Confine this into a new CONFIG_NO_HZ_FULL_WORK because it is still experimental.

Patches in v4 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/6task_work: Provide means to check if a work is queuedinclude/linux/task_work.h kernel/task_work.c+19 -20 replies
2/6sched/fair: Use task_work_queued() on numa_workkernel/sched/fair.c+1 -40 replies
3/6sched: Use task_work_queued() on cid_workkernel/sched/core.c+1 -41 reply
4/6tick/nohz: Move nohz_full related fields out of hot task struct's placesinclude/linux/sched.h+8 -61 reply
5/6sched/isolation: Introduce isolated task workinclude/linux/sched.h include/linux/sched/isolation.h kernel/sched/core.c +3 more+58 -03 replies
6/6mm: Drain LRUs upon resume to userspace on nohz_full CPUsinclude/linux/pagevec.h include/linux/swap.h kernel/sched/isolation.c +1 more+35 -174 replies

Reviews trailers per patch

4 of 6 patches have a Reviewed-by or Acked-by.

PatchOleg NesterovValentin Schneider
1/6R-b R-b
2/6R-b R-b
3/6R-b R-b
4/6R-b
5/6
6/6

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.

Read the v4 thread

Cover letter [PATCH 0/6 v4] sched/mm: LRU drain flush on nohz_full

When LRUs are pending, the drain can be triggered remotely, whether the remote CPU is running in userspace in nohz_full mode or not. This kind of noise is expected to be caused by preparatory work before a task runs isolated in userspace. This patchset is a proposal to flush that before the task starts its critical work in userspace.

Changes since v3:

* Apply review from Oleg and K Prateek Nayak (handle io_uring kthreads and use guard)

* Confine this into a new CONFIG_NO_HZ_FULL_WORK because it is still experimental.

git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks.git
task/work-v4

HEAD: 87896fa0dc36b421533c9dc85dd32b61eaff887b

Thanks,
Frederic
---

Frederic Weisbecker (6):
task_work: Provide means to check if a work is queued sched/fair: Use task_work_queued() on numa_work
sched: Use task_work_queued() on cid_work
tick/nohz: Move nohz_full related fields out of hot task struct's places sched/isolation: Introduce isolated task work
mm: Drain LRUs upon resume to userspace on nohz_full CPUs

 include/linux/pagevec.h         | 18 ++----------------
 include/linux/sched.h           | 18 ++++++++++++------
 include/linux/sched/isolation.h | 17 +++++++++++++++++
 include/linux/swap.h            |  1 +
 include/linux/task_work.h       | 12 ++++++++++++
 kernel/sched/core.c             |  6 ++----
 kernel/sched/fair.c             |  5 +----
 kernel/sched/isolation.c        | 26 ++++++++++++++++++++++++++
 kernel/sched/sched.h            |  1 +
 kernel/task_work.c              |  9 +++++++--
 kernel/time/Kconfig             | 12 ++++++++++++
 mm/swap.c                       | 30 +++++++++++++++++++++++++++++-
 12 files changed, 122 insertions(+), 33 deletions(-)