This is the initial pull request of sched_ext. The v7 patchset
(https://lkml.kernel.org/r/20240618212056.2833381-1-tj@kernel.org) is
applied on top of tip/sched/core + bpf/master as of Jun 18th.
tip/sched/core 793a62823d1c ("sched/core: Drop spinlocks on contention iff kernel is preempti
ble")
bpf/master f6afdaf72a ("Merge branch 'bpf-support-resilient-split-btf'")
Since then, the following pulls were made:
- v6.11-rc1 is pulled to keep up with the mainline.
- tip/sched/core was pulled several times:
- 7b9f6c864a, 0df340ceae, 5ac998574f, 0b1777f0fa: To resolve
conflicts. See each commit for details on conflicts and their
resolutions.
- d7b01aef9d: To receive fd03c5b858 ("sched: Rework pick_next_task()")
and related commits. @prev in added to sched_class->put_prev_task() and
put_prev_task() is reordered after ->pick_task(), which makes
sched_class->switch_class() unnecessary. The follow-up commits update
sched_ext accordingly and drop sched_class->switch_class().
- bpf/master was pulled to receive baebe9aaba ("bpf: allow passing struct
bpf_iter_<type> as kfunc arguments") and related changes in preparation
for the DSQ iterator patchset
To obtain the net sched_ext changes, diff against:
git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext.git for-6.12-base
which is the merge of:
tip/sched/core bc9057da1a ("sched/cpufreq: Use NSEC_PER_MSEC for deadline task")
bpf/master 2ad6d23f46 ("selftests/bpf: Do not update vmlinux.h unnecessarily")
Since the v7 patchset, the following changes were made:
- cpuperf support which was a part of the v6 patchset was posted separately
and then applied after reviews.
- cgroup support which was a part of the v6 patchset was posted seprately,
iterated and then applied.
- Improve integration with sched core.
- Double locking usage in migration paths dropped. Depend on
TASK_ON_RQ_MIGRATING synchronization instead.
- The BPF scheduler couldn't directly dispatch to the local DSQ of another
CPU using a SCX_DSQ_LOCAL_ON verdict. This caused difficulties around
handling non-wakeup enqueues. Updated so that SCX_DSQ_LOCAL_ON can be used
in the enqueue path too.
- DSQ iterator which was a part of the v6 patchset was posted separately.
The iterator itself was applied after a couple revisions. The associated
selective consumption kfunc can use further improvements and is still
being worked on.
- scx_bpf_dispatch[_vtime]_from_dsq() added to increase flexibility. A task
can now be transferred between two DSQs from almost any context. This
involved significant refactoring of migration code.
- Various fixes and improvements.
As the branch is based on top of tip/sched/core + bpf/master, please merge
after both are applied.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCZuOSuA4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGVZyAQDBU3WPkYKB8gl6a6YQ+/PzBXorOK7mioS9A2iJ
vBR3FgEAg1vtcss1S+2juWmVq7ItiFNWCqtXzUr/bVmL9CqqDwA=
=bOOC
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext support from Tejun Heo:
"This implements a new scheduler class called ‘ext_sched_class’, or
sched_ext, which allows scheduling policies to be implemented as BPF
programs.
The goals of this are:
- Ease of experimentation and exploration: Enabling rapid iteration
of new scheduling policies.
- Customization: Building application-specific schedulers which
implement policies that are not applicable to general-purpose
schedulers.
- Rapid scheduler deployments: Non-disruptive swap outs of scheduling
policies in production environments"
See individual commits for more documentation, but also the cover letter
for the latest series:
Link: https://lore.kernel.org/all/20240618212056.2833381-1-tj@kernel.org/
* tag 'sched_ext-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext: (110 commits)
sched: Move update_other_load_avgs() to kernel/sched/pelt.c
sched_ext: Don't trigger ops.quiescent/runnable() on migrations
sched_ext: Synchronize bypass state changes with rq lock
scx_qmap: Implement highpri boosting
sched_ext: Implement scx_bpf_dispatch[_vtime]_from_dsq()
sched_ext: Compact struct bpf_iter_scx_dsq_kern
sched_ext: Replace consume_local_task() with move_local_task_to_local_dsq()
sched_ext: Move consume_local_task() upward
sched_ext: Move sanity check and dsq_mod_nr() into task_unlink_from_dsq()
sched_ext: Reorder args for consume_local/remote_task()
sched_ext: Restructure dispatch_to_local_dsq()
sched_ext: Fix processs_ddsp_deferred_locals() by unifying DTL_INVALID handling
sched_ext: Make find_dsq_for_dispatch() handle SCX_DSQ_LOCAL_ON
sched_ext: Refactor consume_remote_task()
sched_ext: Rename scx_kfunc_set_sleepable to unlocked and relocate
sched_ext: Add missing static to scx_dump_data
sched_ext: Add missing static to scx_has_op[]
sched_ext: Temporarily work around pick_task_scx() being called without balance_scx()
sched_ext: Add a cgroup scheduler which uses flattened hierarchy
sched_ext: Add cgroup support
...
88264981f2