sched_ext: Maintain an online cid mask in the scheduler arenagrafted

Schedulers on the default cid mapping treat [0, nr_online_cids) as the
online set and restart on hotplug. Schedulers that install their own mapping
with scx_bpf_cid_override() have no way to learn which cids are online: the
count no longer identifies members and the CPU-form cpumask is unusable from
cid programs. This is an obvious hole in the cid API.

Add scx_bpf_online_cmask(), a kernel-maintained cmask in the scheduler's
arena, allocated alongside the per-CPU scratch masks and populated after the
cid mapping is finalized and before ops.init(), for child schedulers too.
The pointer stays valid through ops.exit() with no reference to take. It is
the arena offset as a void pointer, the same form struct_ops arena arguments
arrive in. The verifier types the void return as a scalar for the program's
arena cast.

The mask follows the SCX hotplug notifications: seeded from cpu_active_mask
and updated before ops.cid_online/offline() runs, so it lags cpu_online_mask
only inside a hotplug transition. Updates walk the scheduler list under the
lock that also serializes unlinking. Reads are live, not atomic snapshots.
Root initialization excludes hotplug.

v2: Reworded the getter kerneldoc (Andrea Righi).

Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>

1 files changed, 51 insertions, 0 deletions

diff --git a/include/drm/intel/pick.h b/include/drm/intel/pick.h
new file mode 100644
index 000000000..d976fab8f
--- /dev/null
+++ b/include/drm/intel/pick.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright © 2026 Intel Corporation */
+
+#ifndef _PICK_H_
+#define _PICK_H_
+
+/*
+ * Given the first two numbers __a and __b of arbitrarily many evenly spaced
+ * numbers, pick the 0-based __index'th value.
+ *
+ * Always prefer this over _PICK() if the numbers are evenly spaced.
+ */
+#define _PICK_EVEN(__index, __a, __b) ((__a) + (__index) * ((__b) - (__a)))
+
+/*
+ * Like _PICK_EVEN(), but supports 2 ranges of evenly spaced address offsets.
+ * @__c_index corresponds to the index in which the second range starts to be
+ * used. Using math interval notation, the first range is used for indexes [ 0,
+ * @__c_index), while the second range is used for [ @__c_index, ... ). Example:
+ *
+ * #define _FOO_A			0xf000
+ * #define _FOO_B			0xf004
+ * #define _FOO_C			0xf008
+ * #define _SUPER_FOO_A			0xa000
+ * #define _SUPER_FOO_B			0xa100
+ * #define FOO(x)			_MMIO(_PICK_EVEN_2RANGES(x, 3,		\
+ *					      _FOO_A, _FOO_B,			\
+ *					      _SUPER_FOO_A, _SUPER_FOO_B))
+ *
+ * This expands to:
+ *	0: 0xf000,
+ *	1: 0xf004,
+ *	2: 0xf008,
+ *	3: 0xa000,
+ *	4: 0xa100,
+ *	5: 0xa200,
+ *	...
+ */
+#define _PICK_EVEN_2RANGES(__index, __c_index, __a, __b, __c, __d)		\
+	(BUILD_BUG_ON_ZERO(!__is_constexpr(__c_index)) +			\
+	 ((__index) < (__c_index) ? _PICK_EVEN(__index, __a, __b) :		\
+				   _PICK_EVEN((__index) - (__c_index), __c, __d)))
+
+/*
+ * Given the arbitrary numbers in varargs, pick the 0-based __index'th number.
+ *
+ * Always prefer _PICK_EVEN() over this if the numbers are evenly spaced.
+ */
+#define _PICK(__index, ...) (((const u32 []){ __VA_ARGS__ })[__index])
+
+#endif