[PATCH 0/2] sched_ext: minor doc and comment fixes

[PATCH 0/2] sched_ext: minor doc and comment fixes

[Tao Cui]

From: Tao Cui <cuitao@kylinos.cn>

Two small doc and comment fixes noticed while
reading the code. No functional changes.

Tao Cui (2):
  Documentation: sched_ext: fix events sysfs path and show_state example
  sched_ext: fix stale references in doc comments

 Documentation/scheduler/sched-ext.rst | 9 +++++----
 kernel/sched/ext/inlines.h            | 2 +-
 kernel/sched/ext/internal.h           | 5 ++---
 3 files changed, 8 insertions(+), 8 deletions(-)

-- 
2.43.0

[PATCH 1/2] Documentation: sched_ext: fix events sysfs path and show_state example

[Tao Cui]

From: Tao Cui <cuitao@kylinos.cn>

The events file is under the scheduler's sysfs kobject
(/sys/kernel/sched_ext/root/events for the root scheduler), not the
nonexistent "<scheduler-name>/events" path. Also add the missing "aborting"
line to the scx_show_state.py example.

Signed-off-by: Tao Cui <cuitao@kylinos.cn>
---
 Documentation/scheduler/sched-ext.rst | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/Documentation/scheduler/sched-ext.rst b/Documentation/scheduler/sched-ext.rst
index ad2fff3c0593..dd4853762426 100644
--- a/Documentation/scheduler/sched-ext.rst
+++ b/Documentation/scheduler/sched-ext.rst
@@ -93,13 +93,13 @@ scheduler has been loaded):
     # cat /sys/kernel/sched_ext/enable_seq
     1
 
-Each running scheduler also exposes a per-scheduler ``events`` file under
-``/sys/kernel/sched_ext/<scheduler-name>/events`` that tracks diagnostic
-counters. Each counter occupies one ``name value`` line:
+Each running scheduler exposes an ``events`` file under its sysfs kobject
+(``/sys/kernel/sched_ext/root/events`` for the root scheduler) that tracks
+diagnostic counters. Each counter occupies one ``name value`` line:
 
 .. code-block:: none
 
-    # cat /sys/kernel/sched_ext/simple/events
+    # cat /sys/kernel/sched_ext/root/events
     SCX_EV_SELECT_CPU_FALLBACK 0
     SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE 0
     SCX_EV_DISPATCH_KEEP_LAST 123
@@ -153,6 +153,7 @@ detailed information:
     switching_all : 1
     switched_all  : 1
     enable_state  : enabled (2)
+    aborting      : 0
     bypass_depth  : 0
     nr_rejected   : 0
     enable_seq    : 1
-- 
2.43.0

Re: [PATCH 1/2] Documentation: sched_ext: fix events sysfs path and show_state example

[Tejun Heo]

Hello, Tao.

On Tue, Aug 11, 2026 at 03:38:00PM +0800, Tao Cui wrote:
...
> @@ -153,6 +153,7 @@ detailed information:
>      switching_all : 1
>      switched_all  : 1
>      enable_state  : enabled (2)
> +    aborting      : 0
>      bypass_depth  : 0
>      nr_rejected   : 0
>      enable_seq    : 1

scx_root->aborting is a bool and drgn prints Python bools, so the actual
output here is "aborting      : False". Other than that, looks good.

Thanks.

-- 
tejun

[PATCH 2/2] sched_ext: fix stale references in doc comments

[Tao Cui]

From: Tao Cui <cuitao@kylinos.cn>

- scx_bpf_dispatch() was renamed to scx_bpf_dsq_insert()
- drop reference to undefined %SCX_DEQ_SAVE
- @name is in /sys/kernel/sched_ext/root/ops, not a sysctl

Signed-off-by: Tao Cui <cuitao@kylinos.cn>
---
 kernel/sched/ext/inlines.h  | 2 +-
 kernel/sched/ext/internal.h | 5 ++---
 2 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/kernel/sched/ext/inlines.h b/kernel/sched/ext/inlines.h
index d2d074cee1cb..812d67d18f54 100644
--- a/kernel/sched/ext/inlines.h
+++ b/kernel/sched/ext/inlines.h
@@ -14,7 +14,7 @@
 #include "cid.h"
 
 /*
- * One user of this function is scx_bpf_dispatch() which can be called
+ * One user of this function is scx_bpf_dsq_insert() which can be called
  * recursively as sub-sched dispatches nest. Always inline to reduce stack usage
  * from the call frame.
  */
diff --git a/kernel/sched/ext/internal.h b/kernel/sched/ext/internal.h
index b699e7c1103f..06aab2f75208 100644
--- a/kernel/sched/ext/internal.h
+++ b/kernel/sched/ext/internal.h
@@ -485,7 +485,6 @@ struct sched_ext_ops {
 	 * - sleeping (%SCX_DEQ_SLEEP)
 	 * - being moved to another CPU
 	 * - being temporarily taken off the queue for an attribute change
-	 *   (%SCX_DEQ_SAVE)
 	 *
 	 * This and ->dequeue() are related but not coupled. This operation
 	 * notifies @p's state transition and may not be preceded by ->dequeue()
@@ -966,8 +965,8 @@ struct sched_ext_ops {
 	 * @name: BPF scheduler's name
 	 *
 	 * Must be a non-zero valid BPF object name including only isalnum(),
-	 * '_' and '.' chars. Shows up in kernel.sched_ext_ops sysctl while the
-	 * BPF scheduler is enabled.
+	 * '_' and '.' chars. Exposed via
+	 * /sys/kernel/sched_ext/root/ops while the BPF scheduler is enabled.
 	 */
 	char name[SCX_OPS_NAME_LEN];
 
-- 
2.43.0

Re: [PATCH 2/2] sched_ext: fix stale references in doc comments

[Tejun Heo]

Hello, Tao.

On Tue, Aug 11, 2026 at 03:38:01PM +0800, Tao Cui wrote:
> - * One user of this function is scx_bpf_dispatch() which can be called
> + * One user of this function is scx_bpf_dsq_insert() which can be called
>   * recursively as sub-sched dispatches nest. Always inline to reduce stack usage
>   * from the call frame.

This one isn't a leftover of the scx_bpf_dispatch() rename.
scx_bpf_dsq_insert() doesn't call scx_dispatch_sched() and can't nest.
The comment is a typo for scx_bpf_sub_dispatch(), which a parent's
ops.dispatch() calls to trigger dispatching on a child scheduler, so the
calls nest as sub-scheds do. Please use scx_bpf_sub_dispatch() instead.

> 	 * - sleeping (%SCX_DEQ_SLEEP)
> 	 * - being moved to another CPU
> 	 * - being temporarily taken off the queue for an attribute change
> -	 *   (%SCX_DEQ_SAVE)

SCX_DEQ_SAVE was never defined, but these dequeues are flagged with
%SCX_DEQ_SCHED_CHANGE which ops.quiescent() does receive. Please replace
the reference instead of dropping it.

> +	 * '_' and '.' chars. Exposed via
> +	 * /sys/kernel/sched_ext/root/ops while the BPF scheduler is enabled.

@name applies to sub-schedulers too and they show it the same way under
their own directories. Maybe something like "Exposed via the ops file in
the scheduler's sysfs directory, /sys/kernel/sched_ext/root/ops for the
root scheduler, while the BPF scheduler is enabled."

The first hunk no longer applies to sched_ext/for-7.3. Please base v2 on
that branch.

Thanks.

-- 
tejun