Add a comment explaining the design intent behind rejecting built-in DSQs
(%SCX_DSQ_GLOBAL and %SCX_DSQ_LOCAL*) as sources. Local DSQs support
reenqueueing but the BPF scheduler cannot directly iterate or move tasks
from them. %SCX_DSQ_GLOBAL is similar but also doesn't support
reenqueueing because it maps to multiple per-node DSQs, making the scope
difficult to define.
Also annotate @dsq_id to make clear it must be a user-created DSQ.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_central currently assumes that ops.init() runs on the selected
central CPU and aborts otherwise. This is no longer true, as ops.init()
is invoked from the scx_enable_helper thread, which can run on any
CPU.
As a result, sched_setaffinity() from userspace doesn't work, causing
scx_central to fail when loading with:
[ 1985.319942] sched_ext: central: scx_central.bpf.c:314: init from non-central CPU
[ 1985.320317] scx_exit+0xa3/0xd0
[ 1985.320535] scx_bpf_error_bstr+0xbd/0x220
[ 1985.320840] bpf_prog_3a445a8163fa8149_central_init+0x103/0x1ba
[ 1985.321073] bpf__sched_ext_ops_init+0x40/0xa8
[ 1985.321286] scx_root_enable_workfn+0x507/0x1650
[ 1985.321461] kthread_worker_fn+0x260/0x940
[ 1985.321745] kthread+0x303/0x3e0
[ 1985.321901] ret_from_fork+0x589/0x7d0
[ 1985.322065] ret_from_fork_asm+0x1a/0x30
DEBUG DUMP
===================================================================
central: root
scx_enable_help[134] triggered exit kind 1025:
scx_bpf_error (scx_central.bpf.c:314: init from non-central CPU)
Fix this by:
- Defer bpf_timer_start() to the first dispatch on the central CPU.
- Initialize the BPF timer in central_init() and kick the central CPU
to guarantee entering the dispatch path on the central CPU immediately.
- Remove the unnecessary sched_setaffinity() call in userspace.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Zhao Mengmeng <zhaomengmeng@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
This reverts commit c50dcf5331.
The tests are superficial, likely AI-generated slop, and flaky. They
don't add actual value and just churn the selftests.
Signed-off-by: Tejun Heo <tj@kernel.org>
This reverts commit 9edd04c418.
The doc-only change is insufficient. The actual fix requires bumping the
minimum pahole version to 1.27 with proper Kconfig and Makefile.btf
changes, and should be routed through the BPF tree.
Link: https://lore.kernel.org/all/e0ca748d-3204-4160-b37d-0f76cbac8c6c@linux.dev
Signed-off-by: Tejun Heo <tj@kernel.org>
compat.bpf.h defined a fallback SCX_ENQ_IMMED macro using
__COMPAT_ENUM_OR_ZERO(). After 6bf36c68b0 ("tools/sched_ext:
Regenerate autogen enum headers") added SCX_ENQ_IMMED to the autogen
headers, including both triggers -Wmacro-redefined warnings.
The autogen definition through const volatile __weak already resolves to
0 on older kernels, providing the same backward compatibility. Remove
the now-redundant compat fallback.
Fixes: 6bf36c68b0 ("tools/sched_ext: Regenerate autogen enum headers")
Link: https://lore.kernel.org/r/20260326100313.338388-1-zhaomzhao@126.com
Reported-by: Zhao Mengmeng <zhaomengmeng@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_pair sizes pair_ctx to nr_cpu_ids / 2, so valid pair_ctx keys are
dense pair indexes in the range [0, nr_cpu_ids / 2).
However, the userspace setup code stores pair_id as the first CPU number
in each pair. On an 8-CPU system with "-S 1", that produces pair IDs
0, 2, 4 and 6 for pairs [0,1], [2,3], [4,5] and [6,7]. CPUs in the
latter half then look up pair_ctx with out-of-range keys and the BPF
scheduler aborts with:
EXIT: scx_bpf_error (scx_pair.bpf.c:328: failed to lookup pairc and
in_pair_mask for cpu[5])
Assign pair_id using a dense pair counter instead so that each CPU pair
maps to a valid pair_ctx entry. Besides, reject odd CPU configuration, as
scx_pair requires all CPUs to be paired.
Fixes: f0262b102c ("tools/sched_ext: add scx_pair scheduler")
Signed-off-by: Zhao Mengmeng <zhaomengmeng@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
025b1bd419 introduced SCX_EV_SUB_BYPASS_DISPATCH to track scheduling
of bypassed descendant tasks, and correctly increments it per-CPU and
displays it in sysfs and dump output. However, scx_read_events() which
aggregates per-CPU counters into a summary was not updated to include
this event, causing it to always read as zero in sysfs, in debug dumps,
and via the scx_bpf_events() kfunc.
Add the missing scx_agg_event() call for SCX_EV_SUB_BYPASS_DISPATCH.
Fixes: 025b1bd419 ("sched_ext: Implement hierarchical bypass mode")
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When scx_bpf_dsq_move[_vtime]() is called on a task that belongs to a
different scheduler, scx_error() is invoked to flag the violation.
scx_error() schedules an asynchronous scheduler teardown via irq_work
and returns immediately, so execution falls through and the DSQ move
proceeds on a cross-scheduler task regardless, potentially corrupting
DSQ state.
Add the missing return false so the function exits right after
reporting the error, consistent with the other early-exit checks in
the same function (e.g. scx_vet_enq_flags() failure at the top).
Fixes: bb4d9fd551 ("sched_ext: scx_dsq_move() should validate the task belongs to the right scheduler")
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
ops.dispatch() is invoked when a CPU becomes available. This can occur
when a task voluntarily yields the CPU, exhausts its time slice, or is
preempted for other reasons.
If the task is still runnable, refilling its time slice in
ops.dispatch() (either by the BPF scheduler or the sched_ext core)
allows it to continue running without triggering ops.stopping().
However, this behavior is not clearly reflected in the current task
lifecycle diagram.
Update the diagram to better represent this interaction.
Fixes: 9465f44d2d ("sched_ext: Documentation: Clarify time slice handling in task lifecycle")
Cc: stable@vger.kernel.org # v6.17+
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The print_scx_info() always output scx_root structure's->ops.name,
but for built with CONFIG_EXT_SUB_SCHED=y kernels, the tasks may be
attach an sub scx_sched structure. this commit therefore use the
scx_task_sched_rcu() to correctly get scx_sched structure to output
ops.name, and drop state check.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
Signed-off-by: Tejun Heo <tj@kernel.org>
rt_stall tests the ext DL server which was introduced in commit
cd959a3562 ("sched_ext: Add a DL server for sched_ext tasks").
On older kernels that lack this feature, the test calls ksft_exit_fail()
internally which terminates the entire runner process, preventing
subsequent tests from running.
Add a guard in setup() that checks for the ext_server field in struct rq
via __COMPAT_struct_has_field() and returns SCX_TEST_SKIP if not present.
Also print the names of skipped tests in the results summary, mirroring
the existing behavior for failed tests.
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Since Linux 7.0, kfuncs annotated with KF_IMPLICIT_ARGS require pahole
v1.26 or later. Without it, such kfuncs will have incorrect BTF
prototypes in vmlinux, causing BPF programs to fail to load with a
"func_proto incompatible with vmlinux" error. Many sched_ext kfuncs
are affected (e.g. scx_bpf_create_dsq, scx_bpf_kick_cpu).
The root cause: scripts/Makefile.btf passes --btf_features=decl_tag_kfuncs
to pahole only when pahole >= 1.26. Without that flag, pahole emits no
DECL_TAG BTF entries for __bpf_kfunc-annotated functions. As a result,
resolve_btfids/main.c::collect_kfuncs() finds no bpf_kfunc DECL_TAGs,
short-circuits at line 1002, and btf2btf() never creates the _impl
variants or strips the implicit 'aux' argument from the visible proto.
The vmlinux BTF retains the 3-param prototype while BPF programs declare
the 2-param version, triggering the mismatch.
Raise the minimum version in the requirements table from 1.22 to 1.26
and add a note explaining the failure mode, so users understand why
their BPF programs fail on distributions shipping pahole v1.25 (e.g.
Ubuntu 24.04 LTS).
Suggested-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: zhidao su <suzhidao@xiaomi.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When scx_alloc_and_add_sched() creates the sub-scheduler kset, it sets
sch->kobj as the parent. Because sch->kobj.kset points to scx_kset,
registering this sub-kset triggers a KOBJ_ADD uevent. The uevent walk
finds scx_kset and calls scx_uevent() with the sub-kset's kobject.
scx_uevent() unconditionally uses container_of() to cast the incoming
kobject to struct scx_sched, producing a wild pointer when the kobject
belongs to the kset itself rather than a scheduler instance. Accessing
sch->ops.name through this pointer causes a KASAN slab-out-of-bounds
read:
BUG: KASAN: slab-out-of-bounds in string+0x3b6/0x4c0
Read of size 1 at addr ffff888004d04348 by task scx_enable_help/748
Call Trace:
string+0x3b6/0x4c0
vsnprintf+0x3ec/0x1550
add_uevent_var+0x160/0x3a0
scx_uevent+0x22/0x30
kobject_uevent_env+0x5dc/0x1730
kset_register+0x192/0x280
scx_alloc_and_add_sched+0x130d/0x1c60
...
Fix this by checking the kobject's ktype against scx_ktype before
performing the cast, and returning 0 for non-matching kobjects.
Tested with vng and scx_qmap without triggering any KASAN errors.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Add a transparent compatibility wrapper for the scx_bpf_sub_dispatch()
kfunc in compat.bpf.h. This allows BPF schedulers using the sub-sched
dispatch feature to build and run on older kernels that lack the kfunc.
To avoid requiring code changes in individual schedulers, the
transparent wrapper pattern is used instead of a __COMPAT prefix. The
kfunc is declared with a ___compat suffix, while the static inline
wrapper retains the original scx_bpf_sub_dispatch() name.
When the kfunc is unavailable, the wrapper safely falls back to
returning false. This is acceptable because the dispatch path cannot
do anything useful without underlying sub-sched support anyway.
Tested scx_qmap on v6.14 successfully.
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Add three selftests covering features introduced in v7.1:
- dsq_reenq: Verify scx_bpf_dsq_reenq() on user DSQs triggers
ops.enqueue() with SCX_ENQ_REENQ and SCX_TASK_REENQ_KFUNC in
p->scx.flags.
- enq_immed: Verify SCX_OPS_ALWAYS_ENQ_IMMED slow path where tasks
dispatched to a busy CPU's local DSQ are re-enqueued through
ops.enqueue() with SCX_TASK_REENQ_IMMED.
- consume_immed: Verify SCX_ENQ_IMMED via the consume path using
scx_bpf_dsq_move_to_local___v2() with explicit SCX_ENQ_IMMED.
All three tests skip gracefully on kernels that predate the required
features by checking availability via __COMPAT_has_ksym() /
__COMPAT_read_enum() before loading.
Signed-off-by: zhidao su <suzhidao@xiaomi.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
schedule_deferred() uses irq_work_queue() which always queues on the
calling CPU. The deferred work can run from any CPU correctly, and the
_locked() path already processes remote rqs from the calling CPU. However,
when falling through to the irq_work path, queuing on the target CPU is
preferable as the work can run sooner via IPI delivery rather than waiting
for the calling CPU to re-enable IRQs.
Currently, only reenqueue operations use this path - either BPF-initiated
reenqueue targeting a remote rq, or IMMED reenqueue when the target CPU is
busy running userspace (not in balance or wakeup, so the _locked() fast
paths aren't available). Use irq_work_queue_on() to target the owning CPU.
This improves IMMED reenqueue latency when tasks are dispatched to
remote local DSQs. Testing on a 24-CPU AMD Ryzen 3900X with scx_qmap
-I -F 50 (ALWAYS_ENQ_IMMED, every 50th enqueue forced to prev_cpu's
local DSQ) under heavy mixed load (2x CPU oversubscription, yield and
context-switch pressure, SCHED_FIFO bursts, periodic fork storms, mixed
nice levels, C-states disabled), measuring local DSQ residence time
(insert to remove) over 5 x 120s runs (~1.2M tasks per set):
>128us outliers: 71 -> 39 (-45%)
>256us outliers: 59 -> 36 (-39%)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Extend SCX_OPS_OPEN() with compatibility handling for ops.sub_attach()
and ops.sub_detach(), allowing scx C schedulers with sub-scheduler
support to run on kernels both with and without its support.
Cc: Cheng-Yang Chou <yphbchou0911@gmail.com>
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Reviewed-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When building with SCHED_CLASS_EXT=y but CGROUPS=n, clang reports errors
for undeclared cgroup_put() and cgroup_get() calls, and a warning for the
unused err_stop_helper label.
EXT_SUB_SCHED is def_bool y depending only on SCHED_CLASS_EXT, but it
fundamentally requires cgroups (cgroup_path, cgroup_get, cgroup_put,
cgroup_id, etc.). Add the missing CGROUPS dependency to EXT_SUB_SCHED in
init/Kconfig.
Guard cgroup_put() and cgroup_get() in the common paths with:
#if defined(CONFIG_EXT_GROUP_SCHED) || defined(CONFIG_EXT_SUB_SCHED)
Guard the err_stop_helper label with #ifdef CONFIG_EXT_SUB_SCHED since
all gotos targeting it are inside that same ifdef block.
Tested with both CGROUPS enabled and disabled.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202603210903.IrKhPd6k-lkp@intel.com/
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Acked-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The function scx_ops_error() was dropped, but the
comment here is left pointing to the old name.
Update to be consistent with current API.
Signed-off-by: Ke Zhao <ke.zhao.kernel@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
runner.c always returned 0 regardless of test results. The kselftest
framework (tools/testing/selftests/kselftest/runner.sh) invokes the runner
binary and treats a non-zero exit code as a test failure; with the old
code, failed sched_ext tests were silently hidden from the parent harness
even though individual "not ok" TAP lines were emitted.
Return 1 when at least one test failed, 0 when all tests passed or were
skipped.
Signed-off-by: zhidao su <suzhidao@xiaomi.com>
Acked-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Two categories of sched_ext diagnostics are currently undocumented:
1. Per-scheduler events sysfs file
Each active BPF scheduler exposes a set of diagnostic counters at
/sys/kernel/sched_ext/<name>/events. These counters are defined
(with detailed comments) in kernel/sched/ext_internal.h but have
no corresponding documentation in sched-ext.rst. BPF scheduler
developers must read kernel source to understand what each counter
means.
Add a description of the events file, an example of its output, and
a brief explanation of every counter.
2. Module parameters
kernel/sched/ext.c registers two parameters under the sched_ext.
prefix (slice_bypass_us, bypass_lb_intv_us) via module_param_cb()
with MODULE_PARM_DESC() strings, but sched-ext.rst makes no mention
of them. Users who need to tune bypass-mode behavior have no
in-tree documentation to consult.
Add a "Module Parameters" section documenting both knobs: their
default values, valid ranges (taken from the set_*() validators in
ext.c), and the note from the source that they are primarily for
debugging.
No functional changes.
Signed-off-by: zhidao su <suzhidao@xiaomi.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
In the default built-in idle CPU selection policy, when @prev_cpu is
busy and no fully idle core is available, try to place the task on its
SMT sibling if that sibling is idle, before searching any other idle CPU
in the same LLC.
Migration to the sibling is cheap and keeps the task on the same core,
preserving L1 cache and reducing wakeup latency.
On large SMT systems this appears to consistently boost throughput by
roughly 2-3% on CPU-bound workloads (running a number of tasks equal to
the number of SMT cores).
Cc: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When tests fail, the runner only printed the failure count, making
it hard to tell which tests failed without scrolling through output.
Track failed test names in an array and print them after the summary
so failures are immediately visible at the end of the run.
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
ancestors[] is a flexible array member that needs level + 1 slots to
hold all ancestors including self (indices 0..level), but kzalloc_flex()
only allocates `level` slots:
sch = kzalloc_flex(*sch, ancestors, level);
...
sch->ancestors[level] = sch; /* one past the end */
For the root scheduler (level = 0), zero slots are allocated and
ancestors[0] is written immediately past the end of the object.
KASAN reports:
BUG: KASAN: slab-out-of-bounds in scx_alloc_and_add_sched+0x1c17/0x1d10
Write of size 8 at addr ffff888066b56538 by task scx_enable_help/667
The buggy address is located 0 bytes to the right of
allocated 1336-byte region [ffff888066b56000, ffff888066b56538)
Fix by passing level + 1 to kzalloc_flex().
Tested with vng + scx_lavd, KASAN no longer triggers.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
kobject_init_and_add() failure requires kobject_put() for proper cleanup, but
the error paths were using kfree(sch) possibly leaking the kobject name. The
kset_create_and_add() failure was already using kobject_put() correctly.
Switch the kobject_init_and_add() error paths to use kobject_put(). As the
release path puts the cgroup ref, make scx_alloc_and_add_sched() always
consume @cgrp via a new err_put_cgrp label at the bottom of the error chain
and update scx_sub_enable_workfn() accordingly.
Fixes: 17108735b4 ("sched_ext: Use dynamic allocation for scx_sched")
Reported-by: David Carlier <devnexen@gmail.com>
Link: https://lore.kernel.org/r/20260314134457.46216-1-devnexen@gmail.com
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
The abort path in scx_sub_enable_workfn() fell through to out_put_cgrp,
double-putting the cgroup ref already owned by sch->cgrp. It also skipped
kthread_flush_work() needed to flush the disable path.
Relocate the abort block above err_unlock_and_disable so it falls through to
err_disable.
Fixes: 337ec00b1d ("sched_ext: Implement cgroup sub-sched enabling and disabling")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
ops.cpu_acquire/release() are deprecated by commit a3f5d48222
("sched_ext: Allow scx_bpf_reenqueue_local() to be called from
anywhere") in favor of handling CPU preemption via the sched_switch
tracepoint.
In the maximal selftest, replace the cpu_acquire/release stubs with a
minimal sched_switch TP program. Attach all non-struct_ops programs
(including the new TP) via maximal__attach() after disabling auto-attach
for the maximal_ops struct_ops map, which is managed manually in run().
Apply the same fix to reload_loop, which also uses the maximal skeleton.
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Direct writes to p->scx.dsq_vtime are deprecated in favor of
scx_bpf_task_set_dsq_vtime(). Update scx_simple, scx_flatcg, and
select_cpu_vtime selftest to use the new kfunc with
scale_by_task_weight_inverse().
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Fix two mismatched closing comments in header include guards:
- util.h: closing comment says __SCX_TEST_H__ but the guard is
__SCX_TEST_UTIL_H__
- exit_test.h: closing comment has a spurious '#' character before
the guard name
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Under CONFIG_EXT_SUB_SCHED, the kzalloc() and kstrdup() failure
paths jump to err_stop_helper without first setting ret. The
function then returns ERR_PTR(ret) with ret uninitialized, which
can produce ERR_PTR(0) (NULL), causing the caller's IS_ERR() check
to pass and leading to a NULL pointer dereference.
Set ret = -ENOMEM before each goto to fix the error path.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
After commit 860683763e ("sched_ext: Add enq_flags to
scx_bpf_dsq_move_to_local()") some of the kselftests are failing to
build:
exit.bpf.c:44:34: error: too few arguments provided to function-like macro invocation
44 | scx_bpf_dsq_move_to_local(DSQ_ID);
Update the kselftests adding the new argument to
scx_bpf_dsq_move_to_local().
Fixes: 860683763e ("sched_ext: Add enq_flags to scx_bpf_dsq_move_to_local()")
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
schedule_dsq_reenq() always uses schedule_deferred() which falls back to
irq_work. However, callers like schedule_reenq_local() already hold the
target rq lock, and scx_bpf_dsq_reenq() may hold it via the ops callback.
Add a locked_rq parameter so schedule_dsq_reenq() can use
schedule_deferred_locked() when the target rq is already held. The locked
variant can use cheaper paths (balance callbacks, wakeup hooks) instead of
always bouncing through irq_work.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
SCX_ENQ_IMMED makes enqueue to local DSQs succeed only if the task can start
running immediately. Otherwise, the task is re-enqueued through ops.enqueue().
This provides tighter control but requires specifying the flag on every
insertion.
Add SCX_OPS_ALWAYS_ENQ_IMMED ops flag. When set, SCX_ENQ_IMMED is
automatically applied to all local DSQ enqueues including through
scx_bpf_dsq_move_to_local().
scx_qmap is updated with -I option to test the feature and -F option for
IMMED stress testing which forces every Nth enqueue to a busy local DSQ.
v2: - Cover scx_bpf_dsq_move_to_local() path (now has enq_flags via ___v2).
- scx_qmap: Remove sched_switch and cpu_release handlers (superseded by
kernel-side wakeup_preempt_scx()). Add -F for IMMED stress testing.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_bpf_dsq_move_to_local() moves a task from a non-local DSQ to the
current CPU's local DSQ. This is an indirect way of dispatching to a local
DSQ and should support enq_flags like direct dispatches do - e.g.
SCX_ENQ_HEAD for head-of-queue insertion and SCX_ENQ_IMMED for immediate
execution guarantees.
Add scx_bpf_dsq_move_to_local___v2() with an enq_flags parameter. The
original becomes a v1 compat wrapper passing 0. The compat macro is updated
to a three-level chain: v2 (7.1+) -> v1 (current) -> scx_bpf_consume
(pre-rename). All in-tree BPF schedulers are updated to pass 0.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add enq_flags parameter to consume_dispatch_q() and consume_remote_task(),
passing it through to move_{local,remote}_task_to_local_dsq(). All callers
pass 0.
No functional change. This prepares for SCX_ENQ_IMMED support on the consume
path.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add SCX_ENQ_IMMED enqueue flag for local DSQ insertions. Once a task is
dispatched with IMMED, it either gets on the CPU immediately and stays on it,
or gets reenqueued back to the BPF scheduler. It will never linger on a local
DSQ behind other tasks or on a CPU taken by a higher-priority class.
rq_is_open() uses rq->next_class to determine whether the rq is available,
and wakeup_preempt_scx() triggers reenqueue when a higher-priority class task
arrives. These capture all higher class preemptions. Combined with reenqueue
points in the dispatch path, all cases where an IMMED task would not execute
immediately are covered.
SCX_TASK_IMMED persists in p->scx.flags until the next fresh enqueue, so the
guarantee survives SAVE/RESTORE cycles. If preempted while running,
put_prev_task_scx() reenqueues through ops.enqueue() with
SCX_TASK_REENQ_PREEMPTED instead of silently placing the task back on the
local DSQ.
This enables tighter scheduling latency control by preventing tasks from
piling up on local DSQs. It also enables opportunistic CPU sharing across
sub-schedulers - without this, a sub-scheduler can stuff the local DSQ of a
shared CPU, making it difficult for others to use.
v2: - Rewrite is_curr_done() as rq_is_open() using rq->next_class and
implement wakeup_preempt_scx() to achieve complete coverage of all
cases where IMMED tasks could get stranded.
- Track IMMED persistently in p->scx.flags and reenqueue
preempted-while-running tasks through ops.enqueue().
- Bound deferred reenq cycles (SCX_REENQ_LOCAL_MAX_REPEAT).
- Misc renames, documentation.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add scx_vet_enq_flags() stub and call it from scx_dsq_insert_preamble() and
scx_dsq_move(). Pass dsq_id into preamble so the vetting function can
validate flag and DSQ combinations.
No functional change. This prepares for SCX_ENQ_IMMED which will populate the
vetting function.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Split task_should_reenq() into local_task_should_reenq() and
user_task_should_reenq(). The local variant takes reenq_flags by pointer.
No functional change. This prepares for SCX_ENQ_IMMED which will add
IMMED-specific logic to the local variant.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
exit__load(skel) was called without checking its return value.
Every other test in the suite wraps the load call with
SCX_FAIL_IF(). Add the missing check to be consistent with the
rest of the test suite.
Fixes: a5db7817af ("sched_ext: Add selftests")
Signed-off-by: David Carlier <devnexen@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Several demo schedulers and the selftest runner had usage strings
that omitted options which are actually supported:
- scx_central: add missing [-v]
- scx_pair: add missing [-v]
- scx_qmap: add missing [-S] and [-H]
- scx_userland: add missing [-v]
- scx_sdt: remove [-f] which no longer exists
- runner.c: add missing [-s], [-l], [-q]; drop [-h] which none of the
other sched_ext tools list in their usage lines
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_claim_exit() propagates exits to descendants under scx_sched_lock.
A sub-sched being attached concurrently could be missed if it links
after the propagation. Check the parent's exit_kind in scx_link_sched()
under scx_sched_lock to interlock against scx_claim_exit() - either the
parent sees the child in its iteration or the child sees the parent's
non-NONE exit_kind and fails attachment.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
There are two sites that nest rq lock inside scx_sched_lock:
- scx_bypass() takes scx_sched_lock then rq lock per CPU to propagate
per-cpu bypass flags and re-enqueue tasks.
- sysrq_handle_sched_ext_dump() takes scx_sched_lock to iterate all
scheds, scx_dump_state() then takes rq lock per CPU for dump.
And scx_claim_exit() takes scx_sched_lock to propagate exits to
descendants. It can be reached from scx_tick(), BPF kfuncs, and many
other paths with rq lock already held, creating the reverse ordering:
rq lock -> scx_sched_lock vs. scx_sched_lock -> rq lock
Fix by flipping scx_bypass() to take rq lock first, and dropping
scx_sched_lock from sysrq_handle_sched_ext_dump() as scx_sched_all is
already RCU-traversable and scx_dump_lock now prevents dumping a dead
sched. This makes the consistent ordering rq lock -> scx_sched_lock.
Reported-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Link: https://lore.kernel.org/r/20260309163025.2240221-1-yphbchou0911@gmail.com
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_disable() directly called kthread_queue_work() which can acquire
worker->lock, pi_lock and rq->__lock. This made scx_disable() unsafe to
call while holding locks that conflict with this chain - in particular,
scx_claim_exit() calls scx_disable() for each descendant while holding
scx_sched_lock, which nests inside rq->__lock in scx_bypass().
The error path (scx_vexit()) was already bouncing through irq_work to
avoid this issue. Generalize the pattern to all scx_disable() calls by
always going through irq_work. irq_work_queue() is lockless and safe to
call from any context, and the actual kthread_queue_work() call happens
in the irq_work handler outside any locks.
Rename error_irq_work to disable_irq_work to reflect the broader usage.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Add a dedicated scx_dump_lock and per-sched dump_disabled flag so that
debug dumping can be safely disabled during sched teardown without
relying on scx_sched_lock. This is a prep for the next patch which
decouples the sysrq dump path from scx_sched_lock to resolve a lock
ordering issue.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
sub_detach is the parent's op called to notify the parent that a child
is detaching. Test parent->ops.sub_detach instead of sch->ops.sub_detach.
Fixes: ebeca1f930 ("sched_ext: Introduce cgroup sub-sched support")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Prefer using IS_ERR_OR_NULL() over using IS_ERR() and a manual NULL
check.
Change generated with coccinelle.
Signed-off-by: Philipp Hahn <phahn-oss@avm.de>
Signed-off-by: Tejun Heo <tj@kernel.org>
c2a57380df ("sched: Replace use of system_unbound_wq with system_dfl_wq")
converted system_unbound_wq usages in ext.c but missed the queue_rcu_work()
call in scx_kobj_release() which was added later by the dynamic scx_sched
allocation conversion. Apply the same conversion.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Marco Crivellari <marco.crivellari@suse.com>
