Re: [PATCH v3 00/10] Optimize cgroup context switch

[Ian Rogers]

Apologies, I missed the in-reply-to
<20190724223746.153620-1-irogers@xxxxxxxxxx>.

Ian

On Wed, Nov 13, 2019 at 4:30 PM Ian Rogers <irogers@xxxxxxxxxx> wrote:
>
> Avoid iterating over all per-CPU events during cgroup changing context
> switches by organizing events by cgroup.
>
> To make an efficient set of iterators, introduce a min max heap
> utility with test.
>
> These patches include a caching algorithm to improve the search for
> the first event in a group by Kan Liang <kan.liang@xxxxxxxxxxxxxxx> as
> well as rebasing hit "optimize event_filter_match during sched_in"
> from https://lkml.org/lkml/2019/8/7/771.
>
> The v2 patch set was modified by Peter Zijlstra in his perf/cgroup
> branch:
> https://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git
>
> These patches follow Peter's reorganization and his fixes to the
> perf_cpu_context min_heap storage code.
>
> Ian Rogers (8):
>   lib: introduce generic min max heap
>   perf: Use min_max_heap in visit_groups_merge
>   perf: Add per perf_cpu_context min_heap storage
>   perf/cgroup: Grow per perf_cpu_context heap storage
>   perf/cgroup: Order events in RB tree by cgroup id
>   perf: simplify and rename visit_groups_merge
>   perf: cache perf_event_groups_first for cgroups
>   perf: optimize event_filter_match during sched_in
>
> Kan Liang (1):
>   perf/cgroup: Do not switch system-wide events in cgroup switch
>
> Peter Zijlstra (1):
>   perf/cgroup: Reorder perf_cgroup_connect()
>
>  include/linux/min_max_heap.h | 134 +++++++++
>  include/linux/perf_event.h   |  14 +
>  kernel/events/core.c         | 512 ++++++++++++++++++++++++++++-------
>  lib/Kconfig.debug            |  10 +
>  lib/Makefile                 |   1 +
>  lib/test_min_max_heap.c      | 194 +++++++++++++
>  6 files changed, 769 insertions(+), 96 deletions(-)
>  create mode 100644 include/linux/min_max_heap.h
>  create mode 100644 lib/test_min_max_heap.c
>
> --
> 2.24.0.432.g9d3f5f5b63-goog
>