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, 118 insertions, 0 deletions

diff --git a/lib/extable.c b/lib/extable.c
new file mode 100644
index 000000000..9c9f40bd2
--- /dev/null
+++ b/lib/extable.c
@@ -0,0 +1,118 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Derived from arch/ppc/mm/extable.c and arch/i386/mm/extable.c.
+ *
+ * Copyright (C) 2004 Paul Mackerras, IBM Corp.
+ */
+
+#include <linux/bsearch.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/sort.h>
+#include <linux/uaccess.h>
+#include <linux/extable.h>
+
+#ifndef ARCH_HAS_RELATIVE_EXTABLE
+#define ex_to_insn(x)	((x)->insn)
+#else
+static inline unsigned long ex_to_insn(const struct exception_table_entry *x)
+{
+	return (unsigned long)&x->insn + x->insn;
+}
+#endif
+
+#ifndef ARCH_HAS_RELATIVE_EXTABLE
+#define swap_ex		NULL
+#else
+static void swap_ex(void *a, void *b, int size)
+{
+	struct exception_table_entry *x = a, *y = b, tmp;
+	int delta = b - a;
+
+	tmp = *x;
+	x->insn = y->insn + delta;
+	y->insn = tmp.insn - delta;
+
+#ifdef swap_ex_entry_fixup
+	swap_ex_entry_fixup(x, y, tmp, delta);
+#else
+	x->fixup = y->fixup + delta;
+	y->fixup = tmp.fixup - delta;
+#endif
+}
+#endif /* ARCH_HAS_RELATIVE_EXTABLE */
+
+/*
+ * The exception table needs to be sorted so that the binary
+ * search that we use to find entries in it works properly.
+ * This is used both for the kernel exception table and for
+ * the exception tables of modules that get loaded.
+ */
+static int cmp_ex_sort(const void *a, const void *b)
+{
+	const struct exception_table_entry *x = a, *y = b;
+
+	/* avoid overflow */
+	if (ex_to_insn(x) > ex_to_insn(y))
+		return 1;
+	if (ex_to_insn(x) < ex_to_insn(y))
+		return -1;
+	return 0;
+}
+
+void sort_extable(struct exception_table_entry *start,
+		  struct exception_table_entry *finish)
+{
+	sort(start, finish - start, sizeof(struct exception_table_entry),
+	     cmp_ex_sort, swap_ex);
+}
+
+#ifdef CONFIG_MODULES
+/*
+ * If the exception table is sorted, any referring to the module init
+ * will be at the beginning or the end.
+ */
+void trim_init_extable(struct module *m)
+{
+	/*trim the beginning*/
+	while (m->num_exentries &&
+	       within_module_init(ex_to_insn(&m->extable[0]), m)) {
+		m->extable++;
+		m->num_exentries--;
+	}
+	/*trim the end*/
+	while (m->num_exentries &&
+	       within_module_init(ex_to_insn(&m->extable[m->num_exentries - 1]),
+				  m))
+		m->num_exentries--;
+}
+#endif /* CONFIG_MODULES */
+
+static int cmp_ex_search(const void *key, const void *elt)
+{
+	const struct exception_table_entry *_elt = elt;
+	unsigned long _key = *(unsigned long *)key;
+
+	/* avoid overflow */
+	if (_key > ex_to_insn(_elt))
+		return 1;
+	if (_key < ex_to_insn(_elt))
+		return -1;
+	return 0;
+}
+
+/*
+ * Search one exception table for an entry corresponding to the
+ * given instruction address, and return the address of the entry,
+ * or NULL if none is found.
+ * We use a binary search, and thus we assume that the table is
+ * already sorted.
+ */
+const struct exception_table_entry *
+search_extable(const struct exception_table_entry *base,
+	       const size_t num,
+	       unsigned long value)
+{
+	return bsearch(&value, base, num,
+		       sizeof(struct exception_table_entry), cmp_ex_search);
+}