From: Tejun Heo <tj@kernel.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: sched-ext@lists.linux.dev, David Vernet <void@manifault.com>,
Andrea Righi <arighi@nvidia.com>,
Changwoo Min <changwoo@igalia.com>,
Emil Tsalapatis <emil@etsalapatis.com>,
linux-kernel@vger.kernel.org
Subject: [GIT PULL] sched_ext changes for v7.1
Date: Mon, 13 Apr 2026 08:27:32 -1000 [thread overview]
Message-ID: <d9b4c8a00c64192691e22954d09574d1@kernel.org> (raw)
Hello,
This depends on tip/sched-core and cgroup/for-7.1 and should be pulled
after both.
The following changes since commit 7e0ffb72de8aa3b25989c2d980e81b829c577010:
sched_ext: Fix stale direct dispatch state in ddsp_dsq_id (2026-04-03 07:14:49 -1000)
are available in the Git repository at:
https://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext.git tags/sched_ext-for-7.1
for you to fetch changes up to 7e311bafb9ad3a4711c08c00b09fb7839ada37f0:
tools/sched_ext: Add explicit cast from void* in RESIZE_ARRAY() (2026-04-13 06:14:11 -1000)
----------------------------------------------------------------
sched_ext: Changes for v7.1
- Cgroup sub-scheduler groundwork. Multiple BPF schedulers can be
attached to cgroups and the dispatch path is made hierarchical. This
involves substantial restructuring of the core dispatch, bypass,
watchdog, and dump paths to be per-scheduler, along with new
infrastructure for scheduler ownership enforcement, lifecycle
management, and cgroup subtree iteration. The enqueue path is not yet
updated and will follow in a later cycle.
- scx_bpf_dsq_reenq() generalized to support any DSQ including remote
local DSQs and user DSQs. Built on top of this, SCX_ENQ_IMMED
guarantees that tasks dispatched to local DSQs either run immediately
or get reenqueued back through ops.enqueue(), giving schedulers tighter
control over queueing latency. Also useful for opportunistic CPU
sharing across sub-schedulers.
- ops.dequeue() was only invoked when the core knew a task was in BPF
data structures, missing scheduling property change events and skipping
callbacks for non-local DSQ dispatches from ops.select_cpu(). Fixed to
guarantee exactly one ops.dequeue() call when a task leaves BPF
scheduler custody.
- Kfunc access validation moved from runtime to BPF verifier time,
removing runtime mask enforcement.
- Idle SMT sibling prioritization in the idle CPU selection path.
- Documentation, selftest, and tooling updates. Misc bug fixes and
cleanups.
- Merges from tip/sched-core, cgroup/for-7.1, and for-7.0-fixes to
resolve dependencies and conflicts for the above changes.
----------------------------------------------------------------
Andrea Righi (11):
sched_ext: Properly mark SCX-internal migrations via sticky_cpu
sched_ext: Add rq parameter to dispatch_enqueue()
sched_ext: Fix ops.dequeue() semantics
selftests/sched_ext: Add test to validate ops.dequeue() semantics
sched_ext: Pass full dequeue flags to ops.quiescent()
selftests/sched_ext: Update scx_bpf_dsq_move_to_local() in kselftests
sched_ext: idle: Prioritize idle SMT sibling
sched_ext: Guard cpu_smt_mask() with CONFIG_SCHED_SMT
tools/sched_ext: Add compat handling for sub-scheduler ops
sched_ext: Documentation: Clarify ops.dispatch() role in task lifecycle
sched_ext: Documentation: Add ops.dequeue() to task lifecycle
Cheng-Yang Chou (19):
sched_ext: Fix scx_bpf_reenqueue_local() silently reenqueuing nothing
sched_ext: Fix incomplete help text usage strings
sched_ext: Fix uninitialized ret in scx_alloc_and_add_sched()
sched_ext/selftests: Fix incorrect include guard comments
sched_ext: Update demo schedulers and selftests to use scx_bpf_task_set_dsq_vtime()
sched_ext: Update selftests to drop ops.cpu_acquire/release()
sched_ext: Fix slab-out-of-bounds in scx_alloc_and_add_sched()
selftests/sched_ext: Show failed test names in summary
sched_ext: Fix build errors and unused label warning in non-cgroup configs
tools/sched_ext: Add scx_bpf_sub_dispatch() compat wrapper
sched_ext: Fix invalid kobj cast in scx_uevent()
selftests/sched_ext: Skip rt_stall on older kernels and list skipped tests
tools/sched_ext: Regenerate autogen enum headers
sched_ext: Fix missing return after scx_error() in scx_dsq_move()
sched_ext: Fix missing SCX_EV_SUB_BYPASS_DISPATCH aggregation in scx_read_events()
sched_ext: Document why built-in DSQs are unsupported sources in scx_bpf_dsq_move_to_local()
tools/sched_ext: Fix off-by-one in scx_sdt payload zeroing
selftests/sched_ext: Improve runner error reporting for invalid arguments
sched_ext: Remove runtime kfunc mask enforcement
David Carlier (2):
sched_ext: Optimize sched_ext_entity layout for cache locality
selftests/sched_ext: Add missing error check for exit__load()
Ke Zhao (1):
tools/sched_ext: Update stale scx_ops_error() comment in fcg_cgroup_move()
Kuba Piecuch (3):
sched_ext: Documentation: improve accuracy of task lifecycle pseudo-code
sched_ext: Make string params of __ENUM_set() const
tools/sched_ext: Add explicit cast from void* in RESIZE_ARRAY()
Philipp Hahn (1):
sched: Prefer IS_ERR_OR_NULL over manual NULL check
Samuele Mariotti (1):
sched_ext: Fix missing warning in scx_set_task_state() default case
Tejun Heo (92):
Merge branch 'for-7.0-fixes' into for-7.1
Merge branch 'for-7.1' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup into for-7.1
sched_ext: Implement cgroup subtree iteration for scx_task_iter
sched_ext: Add @kargs to scx_fork()
sched/core: Swap the order between sched_post_fork() and cgroup_post_fork()
sched_ext: Update p->scx.disallow warning in scx_init_task()
sched_ext: Reorganize enable/disable path for multi-scheduler support
sched_ext: Introduce cgroup sub-sched support
sched_ext: Introduce scx_task_sched[_rcu]()
sched_ext: Introduce scx_prog_sched()
sched_ext: Enforce scheduling authority in dispatch and select_cpu operations
sched_ext: Enforce scheduler ownership when updating slice and dsq_vtime
sched_ext: scx_dsq_move() should validate the task belongs to the right scheduler
sched_ext: Refactor task init/exit helpers
sched_ext: Make scx_prio_less() handle multiple schedulers
sched_ext: Move default slice to per-scheduler field
sched_ext: Move aborting flag to per-scheduler field
sched_ext: Move bypass_dsq into scx_sched_pcpu
sched_ext: Move bypass state into scx_sched
sched_ext: Prepare bypass mode for hierarchical operation
sched_ext: Factor out scx_dispatch_sched()
sched_ext: When calling ops.dispatch() @prev must be on the same scx_sched
sched_ext: Separate bypass dispatch enabling from bypass depth tracking
sched_ext: Implement hierarchical bypass mode
sched_ext: Dispatch from all scx_sched instances
sched_ext: Move scx_dsp_ctx and scx_dsp_max_batch into scx_sched
sched_ext: Make watchdog sub-sched aware
sched_ext: Convert scx_dump_state() spinlock to raw spinlock
sched_ext: Support dumping multiple schedulers and add scheduler identification
sched_ext: Implement cgroup sub-sched enabling and disabling
sched_ext: Add scx_sched back pointer to scx_sched_pcpu
sched_ext: Make scx_bpf_reenqueue_local() sub-sched aware
sched_ext: Factor out scx_link_sched() and scx_unlink_sched()
sched_ext: Add rhashtable lookup for sub-schedulers
sched_ext: Add basic building blocks for nested sub-scheduler dispatching
Merge branch 'for-7.0-fixes' into for-7.1
sched_ext: Relocate scx_bpf_task_cgroup() and its BTF_ID to the end of kfunc section
sched_ext: Wrap global DSQs in per-node structure
sched_ext: Factor out pnode allocation and deallocation into helpers
sched_ext: Change find_global_dsq() to take CPU number instead of task
sched_ext: Relocate run_deferred() and its callees
sched_ext: Convert deferred_reenq_locals from llist to regular list
sched_ext: Wrap deferred_reenq_local_node into a struct
sched_ext: Introduce scx_bpf_dsq_reenq() for remote local DSQ reenqueue
sched_ext: Add reenq_flags plumbing to scx_bpf_dsq_reenq()
sched_ext: Add per-CPU data to DSQs
sched_ext: Factor out nldsq_cursor_next_task() and nldsq_cursor_lost_task()
sched_ext: Implement scx_bpf_dsq_reenq() for user DSQs
sched_ext: Optimize schedule_dsq_reenq() with lockless fast path
sched_ext: Simplify task state handling
sched_ext: Add SCX_TASK_REENQ_REASON flags
Revert "sched_ext: Use READ_ONCE() for the read side of dsq->nr update"
tools/sched_ext/include: Remove dead sdt_task_defs.h guard from common.h
tools/sched_ext/include: Sync bpf_arena_common.bpf.h with scx repo
tools/sched_ext/include: Add missing helpers to common.bpf.h
tools/sched_ext/include: Add __COMPAT_HAS_scx_bpf_select_cpu_and macro
tools/sched_ext/include: Add libbpf version guard for assoc_struct_ops
tools/sched_ext/include: Regenerate enum_defs.autogen.h
Merge branch 'for-7.0-fixes' into for-7.1
Merge branch 'sched/core' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into for-7.1
sched_ext: Replace system_unbound_wq with system_dfl_wq in scx_kobj_release()
sched_ext: Fix sub_detach op check to test the parent's ops
sched_ext: Add scx_dump_lock and dump_disabled
sched_ext: Always bounce scx_disable() through irq_work
sched_ext: Fix scx_sched_lock / rq lock ordering
sched_ext: Reject sub-sched attachment to a disabled parent
sched_ext: Split task_should_reenq() into local and user variants
sched_ext: Add scx_vet_enq_flags() and plumb dsq_id into preamble
sched_ext: Implement SCX_ENQ_IMMED
sched_ext: Plumb enq_flags through the consume path
sched_ext: Add enq_flags to scx_bpf_dsq_move_to_local()
sched_ext: Add SCX_OPS_ALWAYS_ENQ_IMMED ops flag
sched_ext: Use schedule_deferred_locked() in schedule_dsq_reenq()
sched_ext: Fix cgroup double-put on sub-sched abort path
sched_ext: Use kobject_put() for kobject_init_and_add() failure in scx_alloc_and_add_sched()
sched_ext: Use irq_work_queue_on() in schedule_deferred()
tools/sched_ext: Remove redundant SCX_ENQ_IMMED compat definition
Revert "docs: Raise minimum pahole version to 1.26 for KF_IMPLICIT_ARGS kfuncs"
Revert "selftests/sched_ext: Add tests for SCX_ENQ_IMMED and scx_bpf_dsq_reenq()"
Merge branch 'for-7.0-fixes' into for-7.1
Merge branch 'for-7.0-fixes' into for-7.1
sched_ext: Drop TRACING access to select_cpu kfuncs
sched_ext: Add select_cpu kfuncs to scx_kfunc_ids_unlocked
sched_ext: Track @p's rq lock across set_cpus_allowed_scx -> ops.set_cpumask
sched_ext: Fix ops.cgroup_move() invocation kf_mask and rq tracking
sched_ext: Decouple kfunc unlocked-context check from kf_mask
sched_ext: Drop redundant rq-locked check from scx_bpf_task_cgroup()
sched_ext: Add verifier-time kfunc context filter
sched_ext: Rename scx_kf_allowed_on_arg_tasks() to scx_kf_arg_task_ok()
sched_ext: Warn on task-based SCX op recursion
sched_ext: Drop spurious warning on kick during scheduler disable
tools/sched_ext: Kick home CPU for stranded tasks in scx_qmap
Zhao Mengmeng (3):
sched_ext: remove SCX_OPS_HAS_CGROUP_WEIGHT
tools/sched_ext: scx_pair: fix pair_ctx indexing for CPU pairs
scx_central: Defer timer start to central dispatch to fix init error
Zqiang (1):
sched_ext: Choose the right sch->ops.name to output in the print_scx_info()
fangqiurong (2):
sched_ext: Documentation: Fix scx_bpf_move_to_local kfunc name
selftests/sched_ext: Fix wrong DSQ ID in peek_dsq error message
zhidao su (5):
sched_ext: Fix typos in comments
sched_ext: Documentation: Document events sysfs file and module parameters
selftests/sched_ext: Return non-zero exit code on test failure
selftests/sched_ext: Add tests for SCX_ENQ_IMMED and scx_bpf_dsq_reenq()
docs: Raise minimum pahole version to 1.26 for KF_IMPLICIT_ARGS kfuncs
Documentation/scheduler/sched-ext.rst | 205 +-
include/linux/cgroup-defs.h | 4 +
include/linux/sched/ext.h | 109 +-
init/Kconfig | 4 +
kernel/fork.c | 6 +-
kernel/sched/core.c | 2 +-
kernel/sched/ext.c | 4199 +++++++++++++++-----
kernel/sched/ext.h | 4 +-
kernel/sched/ext_idle.c | 199 +-
kernel/sched/ext_idle.h | 2 +
kernel/sched/ext_internal.h | 344 +-
kernel/sched/sched.h | 12 +-
tools/sched_ext/include/scx/bpf_arena_common.bpf.h | 8 +-
tools/sched_ext/include/scx/common.bpf.h | 277 ++
tools/sched_ext/include/scx/common.h | 5 +-
tools/sched_ext/include/scx/compat.bpf.h | 57 +-
tools/sched_ext/include/scx/compat.h | 52 +-
tools/sched_ext/include/scx/enum_defs.autogen.h | 61 +-
tools/sched_ext/include/scx/enums.autogen.bpf.h | 11 +
tools/sched_ext/include/scx/enums.autogen.h | 4 +
tools/sched_ext/include/scx/enums.h | 2 +-
tools/sched_ext/scx_central.bpf.c | 66 +-
tools/sched_ext/scx_central.c | 26 +-
tools/sched_ext/scx_cpu0.bpf.c | 2 +-
tools/sched_ext/scx_flatcg.bpf.c | 24 +-
tools/sched_ext/scx_pair.c | 16 +-
tools/sched_ext/scx_qmap.bpf.c | 214 +-
tools/sched_ext/scx_qmap.c | 29 +-
tools/sched_ext/scx_sdt.bpf.c | 5 +-
tools/sched_ext/scx_sdt.c | 2 +-
tools/sched_ext/scx_simple.bpf.c | 8 +-
tools/sched_ext/scx_userland.c | 2 +-
tools/testing/selftests/sched_ext/Makefile | 1 +
tools/testing/selftests/sched_ext/dequeue.bpf.c | 389 ++
tools/testing/selftests/sched_ext/dequeue.c | 274 ++
tools/testing/selftests/sched_ext/exit.bpf.c | 2 +-
tools/testing/selftests/sched_ext/exit.c | 2 +-
tools/testing/selftests/sched_ext/exit_test.h | 2 +-
tools/testing/selftests/sched_ext/maximal.bpf.c | 17 +-
tools/testing/selftests/sched_ext/maximal.c | 3 +
tools/testing/selftests/sched_ext/numa.bpf.c | 2 +-
tools/testing/selftests/sched_ext/peek_dsq.bpf.c | 10 +-
tools/testing/selftests/sched_ext/reload_loop.c | 3 +
tools/testing/selftests/sched_ext/rt_stall.c | 5 +
tools/testing/selftests/sched_ext/runner.c | 40 +-
.../selftests/sched_ext/select_cpu_vtime.bpf.c | 8 +-
tools/testing/selftests/sched_ext/util.h | 2 +-
47 files changed, 5356 insertions(+), 1365 deletions(-)
create mode 100644 tools/testing/selftests/sched_ext/dequeue.bpf.c
create mode 100644 tools/testing/selftests/sched_ext/dequeue.c
--
tejun
next reply other threads:[~2026-04-13 18:27 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-04-13 18:27 Tejun Heo [this message]
2026-04-15 18:11 ` [GIT PULL] sched_ext changes for v7.1 pr-tracker-bot
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=d9b4c8a00c64192691e22954d09574d1@kernel.org \
--to=tj@kernel.org \
--cc=arighi@nvidia.com \
--cc=changwoo@igalia.com \
--cc=emil@etsalapatis.com \
--cc=linux-kernel@vger.kernel.org \
--cc=sched-ext@lists.linux.dev \
--cc=torvalds@linux-foundation.org \
--cc=void@manifault.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox