From: Tejun Heo <tj@kernel.org>
To: sched-ext@lists.linux.dev
Cc: David Vernet <void@manifault.com>,
Andrea Righi <arighi@nvidia.com>,
Changwoo Min <changwoo@igalia.com>,
Emil Tsalapatis <emil@etsalapatis.com>,
David Dai <david.dai@linux.dev>,
linux-kernel@vger.kernel.org, Tejun Heo <tj@kernel.org>
Subject: [PATCH 2/2] sched_ext: scx_qmap: Reset a child's cpuperf targets once its PERF revoke takes effect
Date: Wed, 7 Oct 2026 14:03:52 -1000 [thread overview]
Message-ID: <20261008000352.1689057-3-tj@kernel.org> (raw)
In-Reply-To: <20261008000352.1689057-1-tj@kernel.org>
A child that ran a cid at a low cpuperf target leaves the target there when
PERF is revoked, since the kernel resets targets only at root enable. Reset
it to SCX_CPUPERF_ONE from ops.sub_child_ecaps_updated() once the revoke is
in effect, a detach included. Count the deliveries in the hier stats.
Signed-off-by: Tejun Heo <tj@kernel.org>
---
tools/sched_ext/scx_qmap.bpf.c | 15 +++++++++++++++
tools/sched_ext/scx_qmap.c | 7 +++++--
tools/sched_ext/scx_qmap.h | 1 +
3 files changed, 21 insertions(+), 2 deletions(-)
diff --git a/tools/sched_ext/scx_qmap.bpf.c b/tools/sched_ext/scx_qmap.bpf.c
index 7f69394556b6..853db13d060e 100644
--- a/tools/sched_ext/scx_qmap.bpf.c
+++ b/tools/sched_ext/scx_qmap.bpf.c
@@ -2208,6 +2208,20 @@ void BPF_STRUCT_OPS(qmap_sub_ecaps_updated, s32 cid, u64 before, u64 after)
execute_partition();
}
+void BPF_STRUCT_OPS(qmap_sub_child_ecaps_updated, u64 cgroup_id, s32 cid, u64 before,
+ u64 after)
+{
+ __sync_fetch_and_add(&qa.nr_child_ecaps, 1);
+
+ /*
+ * A cid's cpuperf target stays where the last write put it. Reset it
+ * once a child's PERF revoke is in effect so that the child's last
+ * target doesn't outlive the cap.
+ */
+ if ((before & ~after) & SCX_CAP_PERF)
+ scx_bpf_cidperf_set(cid, SCX_CPUPERF_ONE);
+}
+
SCX_OPS_CID_DEFINE(qmap_ops,
.flags = SCX_OPS_ENQ_EXITING | SCX_OPS_TID_TO_TASK,
.select_cid = (void *)qmap_select_cid,
@@ -2232,6 +2246,7 @@ SCX_OPS_CID_DEFINE(qmap_ops,
.sub_caps_updated = (void *)qmap_sub_caps_updated,
.sub_ecaps_updated = (void *)qmap_sub_ecaps_updated,
.sub_cid_sched_updated = (void *)qmap_sub_cid_sched_updated,
+ .sub_child_ecaps_updated = (void *)qmap_sub_child_ecaps_updated,
.init_cids = (void *)qmap_init_cids,
.init = (void *)qmap_init,
.exit = (void *)qmap_exit,
diff --git a/tools/sched_ext/scx_qmap.c b/tools/sched_ext/scx_qmap.c
index e83d5e2147dd..17cf584a2a3d 100644
--- a/tools/sched_ext/scx_qmap.c
+++ b/tools/sched_ext/scx_qmap.c
@@ -113,6 +113,7 @@ struct hier_prev {
u64 nr_enq_blocked;
u64 nr_inject_attempts;
u64 nr_rescue_dsp;
+ u64 nr_child_ecaps;
};
/* current wall-clock time as "HH:MM:SS" for the startup and interval headers */
@@ -208,18 +209,20 @@ static void print_hier(struct qmap_arena *qa, struct hier_prev *prev, u64 own_cg
}
format_cid_ranges(qa, CID_SHARED, ranges, sizeof(ranges));
- printf("hier : nsub=%llu excl=%u shared=%s rr=%s reenq cap/immed +%llu/+%llu blocked=+%llu inj=+%llu rescue=+%llu\n",
+ printf("hier : nsub=%llu excl=%u shared=%s rr=%s reenq cap/immed +%llu/+%llu blocked=+%llu inj=+%llu rescue=+%llu child_ecaps=+%llu\n",
(unsigned long long)qa->nr_sub_scheds, qa->part.nr_excl, ranges, rr,
(unsigned long long)(qa->nr_reenq_cap - prev->nr_reenq_cap),
(unsigned long long)(qa->nr_reenq_immed - prev->nr_reenq_immed),
(unsigned long long)(qa->nr_enq_blocked - prev->nr_enq_blocked),
(unsigned long long)(qa->nr_inject_attempts - prev->nr_inject_attempts),
- (unsigned long long)(qa->nr_rescue_dsp - prev->nr_rescue_dsp));
+ (unsigned long long)(qa->nr_rescue_dsp - prev->nr_rescue_dsp),
+ (unsigned long long)(qa->nr_child_ecaps - prev->nr_child_ecaps));
prev->nr_reenq_cap = qa->nr_reenq_cap;
prev->nr_reenq_immed = qa->nr_reenq_immed;
prev->nr_enq_blocked = qa->nr_enq_blocked;
prev->nr_inject_attempts = qa->nr_inject_attempts;
prev->nr_rescue_dsp = qa->nr_rescue_dsp;
+ prev->nr_child_ecaps = qa->nr_child_ecaps;
/*
* alloc is the cid-time the partition handed each participant, and used
diff --git a/tools/sched_ext/scx_qmap.h b/tools/sched_ext/scx_qmap.h
index 949459d06a18..9e6eefb39735 100644
--- a/tools/sched_ext/scx_qmap.h
+++ b/tools/sched_ext/scx_qmap.h
@@ -195,6 +195,7 @@ struct qmap_arena {
u64 nr_enq_blocked; /* SCX_ENQ_BLOCKED dispatches */
u64 nr_inject_attempts; /* fault-injection: dispatches to an unheld cid */
u64 nr_rescue_dsp; /* SCX_ENQ_RESCUE dispatch attempts */
+ u64 nr_child_ecaps; /* ops.sub_child_ecaps_updated() deliveries */
u32 inject_mode; /* fault-injection mode (QMAP_INJ_*) */
};
--
2.55.0
next prev parent reply other threads:[~2026-10-08 0:03 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-08 0:03 [PATCHSET sched_ext/for-7.4] sched_ext: Add ops.sub_child_ecaps_updated() Tejun Heo
2026-10-08 0:03 ` [PATCH 1/2] sched_ext: Add ops.sub_child_ecaps_updated() to report a child's effective cap changes Tejun Heo
2026-10-08 0:03 ` Tejun Heo [this message]
2026-10-08 9:32 [PATCHSET v2 sched_ext/for-7.4] sched_ext: Add ops.sub_child_ecaps_updated() Tejun Heo
2026-10-08 9:32 ` [PATCH 2/2] sched_ext: scx_qmap: Reset a child's cpuperf targets once its PERF revoke takes effect Tejun Heo
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=20261008000352.1689057-3-tj@kernel.org \
--to=tj@kernel.org \
--cc=arighi@nvidia.com \
--cc=changwoo@igalia.com \
--cc=david.dai@linux.dev \
--cc=emil@etsalapatis.com \
--cc=linux-kernel@vger.kernel.org \
--cc=sched-ext@lists.linux.dev \
--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
all inboxes | Powered by JetHome®