1. Patchew
  2. linux
  3. [PATCH v2 00/10] Add kernel cmdline option for rt_group_sched
  4. View patch

[PATCH v2 06/10] sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED

Michal Koutný posted 10 patches 9 months, 1 week ago
[PATCH v2 06/10] sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED
Posted by Michal Koutný 9 months, 1 week ago 
When RT_GROUPs are compiled but not exposed, their bandwidth cannot
be configured (and it is not initialized for non-root task_groups neither).
Therefore bypass any checks of task vs task_group bandwidth.

This will achieve behavior very similar to setups that have
!CONFIG_RT_GROUP_SCHED and attach cpu controller to cgroup v2 hierarchy.
(On a related note, this may allow having RT tasks with
CONFIG_RT_GROUP_SCHED and cgroup v2 hierarchy.)

Signed-off-by: Michal Koutný <mkoutny@suse.com>
---
 kernel/sched/core.c     | 6 +++++-
 kernel/sched/rt.c       | 2 +-
 kernel/sched/syscalls.c | 3 ++-
 3 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 5b67b4704a5ed..a418e7bc6a123 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -9166,11 +9166,15 @@ static int cpu_cgroup_can_attach(struct cgroup_taskset *tset)
 	struct task_struct *task;
 	struct cgroup_subsys_state *css;
 
+	if (!rt_group_sched_enabled())
+		goto scx_check;
+
 	cgroup_taskset_for_each(task, css, tset) {
 		if (!sched_rt_can_attach(css_tg(css), task))
 			return -EINVAL;
 	}
-#endif
+scx_check:
+#endif /* CONFIG_RT_GROUP_SCHED */
 	return scx_cgroup_can_attach(tset);
 }
 
diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index a427c3f560b71..f25fe2862a7df 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
@@ -2866,7 +2866,7 @@ static int sched_rt_global_constraints(void)
 int sched_rt_can_attach(struct task_group *tg, struct task_struct *tsk)
 {
 	/* Don't accept real-time tasks when there is no way for them to run */
-	if (rt_task(tsk) && tg->rt_bandwidth.rt_runtime == 0)
+	if (rt_group_sched_enabled() && rt_task(tsk) && tg->rt_bandwidth.rt_runtime == 0)
 		return 0;
 
 	return 1;
diff --git a/kernel/sched/syscalls.c b/kernel/sched/syscalls.c
index 8629a87628ebf..7b1689af9ff1e 100644
--- a/kernel/sched/syscalls.c
+++ b/kernel/sched/syscalls.c
@@ -634,7 +634,8 @@ int __sched_setscheduler(struct task_struct *p,
 		 * Do not allow real-time tasks into groups that have no runtime
 		 * assigned.
 		 */
-		if (rt_bandwidth_enabled() && rt_policy(policy) &&
+		if (rt_group_sched_enabled() &&
+				rt_bandwidth_enabled() && rt_policy(policy) &&
 				task_group(p)->rt_bandwidth.rt_runtime == 0 &&
 				!task_group_is_autogroup(task_group(p))) {
 			retval = -EPERM;
-- 
2.48.1
Re: [PATCH v2 06/10] sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED
Posted by Peter Zijlstra 8 months, 2 weeks ago 
On Mon, Mar 10, 2025 at 06:04:38PM +0100, Michal Koutný wrote:
> When RT_GROUPs are compiled but not exposed, their bandwidth cannot
> be configured (and it is not initialized for non-root task_groups neither).
> Therefore bypass any checks of task vs task_group bandwidth.
> 
> This will achieve behavior very similar to setups that have
> !CONFIG_RT_GROUP_SCHED and attach cpu controller to cgroup v2 hierarchy.
> (On a related note, this may allow having RT tasks with
> CONFIG_RT_GROUP_SCHED and cgroup v2 hierarchy.)

Can we make it so that cgroup-v2 is explicitly disallowed for now? As I
said earlier, we're looking at a new implemention with a incompatible
interface.
Re: [PATCH v2 06/10] sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED
Posted by Michal Koutný 8 months, 2 weeks ago 
On Wed, Apr 02, 2025 at 02:02:21PM +0200, Peter Zijlstra <peterz@infradead.org> wrote:
> Can we make it so that cgroup-v2 is explicitly disallowed for now? As I
> said earlier, we're looking at a new implemention with a incompatible
> interface.

I meant here that
- rt_group_sched=0 -> cgroup v2 works but there's no RT group scheduling
- rt_group_sched=1 -> cgroup v2 doesn't work (prohibit RT tasks in
                      non-root groups)

I.e. there is no new function for cgroup v2 besides that it is possible
to switch to RT group scheduling (with v1) without recompiling the
kernel.

Michal
[tip: sched/core] sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED
Posted by tip-bot2 for Michal Koutný 8 months, 1 week ago 
The following commit has been merged into the sched/core branch of tip:

Commit-ID:     277e0909754e9f3c82def97150d2f3ea700098f1
Gitweb:        https://git.kernel.org/tip/277e0909754e9f3c82def97150d2f3ea700098f1
Author:        Michal Koutný <mkoutny@suse.com>
AuthorDate:    Mon, 10 Mar 2025 18:04:38 +01:00
Committer:     Peter Zijlstra <peterz@infradead.org>
CommitterDate: Tue, 08 Apr 2025 20:55:54 +02:00

sched: Bypass bandwitdh checks with runtime disabled RT_GROUP_SCHED

When RT_GROUPs are compiled but not exposed, their bandwidth cannot
be configured (and it is not initialized for non-root task_groups neither).
Therefore bypass any checks of task vs task_group bandwidth.

This will achieve behavior very similar to setups that have
!CONFIG_RT_GROUP_SCHED and attach cpu controller to cgroup v2 hierarchy.
(On a related note, this may allow having RT tasks with
CONFIG_RT_GROUP_SCHED and cgroup v2 hierarchy.)

Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-7-mkoutny@suse.com
---
 kernel/sched/core.c     | 6 +++++-
 kernel/sched/rt.c       | 2 +-
 kernel/sched/syscalls.c | 3 ++-
 3 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 32fb4c1..6900ce5 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -9206,11 +9206,15 @@ static int cpu_cgroup_can_attach(struct cgroup_taskset *tset)
 	struct task_struct *task;
 	struct cgroup_subsys_state *css;
 
+	if (!rt_group_sched_enabled())
+		goto scx_check;
+
 	cgroup_taskset_for_each(task, css, tset) {
 		if (!sched_rt_can_attach(css_tg(css), task))
 			return -EINVAL;
 	}
-#endif
+scx_check:
+#endif /* CONFIG_RT_GROUP_SCHED */
 	return scx_cgroup_can_attach(tset);
 }
 
diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index efa22ba..5e82bfe 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
@@ -2864,7 +2864,7 @@ static int sched_rt_global_constraints(void)
 int sched_rt_can_attach(struct task_group *tg, struct task_struct *tsk)
 {
 	/* Don't accept real-time tasks when there is no way for them to run */
-	if (rt_task(tsk) && tg->rt_bandwidth.rt_runtime == 0)
+	if (rt_group_sched_enabled() && rt_task(tsk) && tg->rt_bandwidth.rt_runtime == 0)
 		return 0;
 
 	return 1;
diff --git a/kernel/sched/syscalls.c b/kernel/sched/syscalls.c
index 2bf5281..547c1f0 100644
--- a/kernel/sched/syscalls.c
+++ b/kernel/sched/syscalls.c
@@ -634,7 +634,8 @@ change:
 		 * Do not allow real-time tasks into groups that have no runtime
 		 * assigned.
 		 */
-		if (rt_bandwidth_enabled() && rt_policy(policy) &&
+		if (rt_group_sched_enabled() &&
+				rt_bandwidth_enabled() && rt_policy(policy) &&
 				task_group(p)->rt_bandwidth.rt_runtime == 0 &&
 				!task_group_is_autogroup(task_group(p))) {
 			retval = -EPERM;

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.