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>

4 files changed, 146 insertions, 0 deletions

diff --git a/scripts/ipe/Makefile b/scripts/ipe/Makefile
new file mode 100644
index 000000000..e87553fbb
--- /dev/null
+++ b/scripts/ipe/Makefile
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+subdir-y := polgen
diff --git a/scripts/ipe/polgen/.gitignore b/scripts/ipe/polgen/.gitignore
new file mode 100644
index 000000000..b6f05cf3d
--- /dev/null
+++ b/scripts/ipe/polgen/.gitignore
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+polgen
diff --git a/scripts/ipe/polgen/Makefile b/scripts/ipe/polgen/Makefile
new file mode 100644
index 000000000..c20456a2f
--- /dev/null
+++ b/scripts/ipe/polgen/Makefile
@@ -0,0 +1,5 @@
+# SPDX-License-Identifier: GPL-2.0
+hostprogs-always-y	:= polgen
+HOST_EXTRACFLAGS += \
+	-I$(srctree)/include \
+	-I$(srctree)/include/uapi \
diff --git a/scripts/ipe/polgen/polgen.c b/scripts/ipe/polgen/polgen.c
new file mode 100644
index 000000000..01134cf89
--- /dev/null
+++ b/scripts/ipe/polgen/polgen.c
@@ -0,0 +1,137 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020-2024 Microsoft Corporation. All rights reserved.
+ */
+
+#include <stdlib.h>
+#include <stddef.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <errno.h>
+
+static void usage(const char *const name)
+{
+	printf("Usage: %s OutputFile (PolicyFile)\n", name);
+	exit(EINVAL);
+}
+
+static int policy_to_buffer(const char *pathname, char **buffer, size_t *size)
+{
+	size_t fsize;
+	size_t read;
+	char *lbuf;
+	int rc = 0;
+	FILE *fd;
+
+	fd = fopen(pathname, "r");
+	if (!fd) {
+		rc = errno;
+		goto out;
+	}
+
+	fseek(fd, 0, SEEK_END);
+	fsize = ftell(fd);
+	rewind(fd);
+
+	lbuf = malloc(fsize);
+	if (!lbuf) {
+		rc = ENOMEM;
+		goto out_close;
+	}
+
+	read = fread((void *)lbuf, sizeof(*lbuf), fsize, fd);
+	if (read != fsize) {
+		rc = -1;
+		goto out_free;
+	}
+
+	*buffer = lbuf;
+	*size = fsize;
+	fclose(fd);
+
+	return rc;
+
+out_free:
+	free(lbuf);
+out_close:
+	fclose(fd);
+out:
+	return rc;
+}
+
+static int write_boot_policy(const char *pathname, const char *buf, size_t size)
+{
+	FILE *fd;
+	size_t i;
+
+	fd = fopen(pathname, "w");
+	if (!fd)
+		return errno;
+
+	fprintf(fd, "/* This file is automatically generated.");
+	fprintf(fd, " Do not edit. */\n");
+	fprintf(fd, "#include <linux/stddef.h>\n");
+	fprintf(fd, "\nextern const char *const ipe_boot_policy;\n\n");
+	fprintf(fd, "const char *const ipe_boot_policy =\n");
+
+	if (!buf || size == 0) {
+		fprintf(fd, "\tNULL;\n");
+		fclose(fd);
+		return 0;
+	}
+
+	fprintf(fd, "\t\"");
+
+	for (i = 0; i < size; ++i) {
+		switch (buf[i]) {
+		case '"':
+			fprintf(fd, "\\\"");
+			break;
+		case '\'':
+			fprintf(fd, "'");
+			break;
+		case '\n':
+			fprintf(fd, "\\n\"\n\t\"");
+			break;
+		case '\\':
+			fprintf(fd, "\\\\");
+			break;
+		case '\t':
+			fprintf(fd, "\\t");
+			break;
+		case '\?':
+			fprintf(fd, "\\?");
+			break;
+		default:
+			fprintf(fd, "%c", buf[i]);
+		}
+	}
+	fprintf(fd, "\";\n");
+	fclose(fd);
+
+	return 0;
+}
+
+int main(int argc, const char *const argv[])
+{
+	char *policy = NULL;
+	size_t len = 0;
+	int rc = 0;
+
+	if (argc < 2)
+		usage(argv[0]);
+
+	if (argc > 2) {
+		rc = policy_to_buffer(argv[2], &policy, &len);
+		if (rc != 0)
+			goto cleanup;
+	}
+
+	rc = write_boot_policy(argv[1], policy, len);
+cleanup:
+	if (policy)
+		free(policy);
+	if (rc != 0)
+		perror("An error occurred during policy conversion: ");
+	return rc;
+}