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, 54 insertions, 0 deletions

diff --git a/Documentation/driver-api/tee.rst b/Documentation/driver-api/tee.rst
new file mode 100644
index 000000000..4d58ac071
--- /dev/null
+++ b/Documentation/driver-api/tee.rst
@@ -0,0 +1,54 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+===============================================
+TEE (Trusted Execution Environment) driver API
+===============================================
+
+Kernel provides a TEE bus infrastructure where a Trusted Application is
+represented as a device identified via Universally Unique Identifier (UUID) and
+client drivers register a table of supported device UUIDs.
+
+TEE bus infrastructure registers following APIs:
+
+match():
+  iterates over the client driver UUID table to find a corresponding
+  match for device UUID. If a match is found, then this particular device is
+  probed via corresponding probe API registered by the client driver. This
+  process happens whenever a device or a client driver is registered with TEE
+  bus.
+
+uevent():
+  notifies user-space (udev) whenever a new device is registered on
+  TEE bus for auto-loading of modularized client drivers.
+
+TEE bus device enumeration is specific to underlying TEE implementation, so it
+is left open for TEE drivers to provide corresponding implementation.
+
+Then TEE client driver can talk to a matched Trusted Application using APIs
+listed in include/linux/tee_drv.h.
+
+TEE client driver example
+-------------------------
+
+Suppose a TEE client driver needs to communicate with a Trusted Application
+having UUID: ``ac6a4085-0e82-4c33-bf98-8eb8e118b6c2``, so driver registration
+snippet would look like::
+
+	static const struct tee_client_device_id client_id_table[] = {
+		{UUID_INIT(0xac6a4085, 0x0e82, 0x4c33,
+			   0xbf, 0x98, 0x8e, 0xb8, 0xe1, 0x18, 0xb6, 0xc2)},
+		{}
+	};
+
+	MODULE_DEVICE_TABLE(tee, client_id_table);
+
+	static struct tee_client_driver client_driver = {
+		.probe		= client_probe,
+		.remove		= client_remove,
+		.id_table	= client_id_table,
+		.driver		= {
+			.name		= DRIVER_NAME,
+		},
+	};
+
+	module_tee_client_driver(client_driver);