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, 97 insertions, 0 deletions

diff --git a/arch/powerpc/boot/opal.c b/arch/powerpc/boot/opal.c
new file mode 100644
index 000000000..b69818ce5
--- /dev/null
+++ b/arch/powerpc/boot/opal.c
@@ -0,0 +1,97 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2016 IBM Corporation.
+ */
+
+#include "ops.h"
+#include "stdio.h"
+#include "io.h"
+#include <libfdt.h>
+#include "../include/asm/opal-api.h"
+
+/* Global OPAL struct used by opal-call.S */
+struct opal {
+	u64 base;
+	u64 entry;
+} opal;
+
+static u32 opal_con_id;
+
+/* see opal-wrappers.S */
+int64_t opal_console_write(int64_t term_number, u64 *length, const u8 *buffer);
+int64_t opal_console_read(int64_t term_number, uint64_t *length, u8 *buffer);
+int64_t opal_console_write_buffer_space(uint64_t term_number, uint64_t *length);
+int64_t opal_console_flush(uint64_t term_number);
+int64_t opal_poll_events(uint64_t *outstanding_event_mask);
+
+void opal_kentry(unsigned long fdt_addr, void *vmlinux_addr);
+
+static int opal_con_open(void)
+{
+	/*
+	 * When OPAL loads the boot kernel it stashes the OPAL base and entry
+	 * address in r8 and r9 so the kernel can use the OPAL console
+	 * before unflattening the devicetree. While executing the wrapper will
+	 * probably trash r8 and r9 so this kentry hook restores them before
+	 * entering the decompressed kernel.
+	 */
+	platform_ops.kentry = opal_kentry;
+	return 0;
+}
+
+static void opal_con_putc(unsigned char c)
+{
+	int64_t rc;
+	uint64_t olen, len;
+
+	do {
+		rc = opal_console_write_buffer_space(opal_con_id, &olen);
+		len = be64_to_cpu(olen);
+		if (rc)
+			return;
+		opal_poll_events(NULL);
+	} while (len < 1);
+
+
+	olen = cpu_to_be64(1);
+	opal_console_write(opal_con_id, &olen, &c);
+}
+
+static void opal_con_close(void)
+{
+	opal_console_flush(opal_con_id);
+}
+
+static void opal_init(void)
+{
+	void *opal_node;
+
+	opal_node = finddevice("/ibm,opal");
+	if (!opal_node)
+		return;
+	if (getprop(opal_node, "opal-base-address", &opal.base, sizeof(u64)) < 0)
+		return;
+	opal.base = be64_to_cpu(opal.base);
+	if (getprop(opal_node, "opal-entry-address", &opal.entry, sizeof(u64)) < 0)
+		return;
+	opal.entry = be64_to_cpu(opal.entry);
+}
+
+int opal_console_init(void *devp, struct serial_console_data *scdp)
+{
+	opal_init();
+
+	if (devp) {
+		int n = getprop(devp, "reg", &opal_con_id, sizeof(u32));
+		if (n != sizeof(u32))
+			return -1;
+		opal_con_id = be32_to_cpu(opal_con_id);
+	} else
+		opal_con_id = 0;
+
+	scdp->open = opal_con_open;
+	scdp->putc = opal_con_putc;
+	scdp->close = opal_con_close;
+
+	return 0;
+}