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, 131 insertions, 0 deletions

diff --git a/lib/dump_stack.c b/lib/dump_stack.c
new file mode 100644
index 000000000..f0c78b5b5
--- /dev/null
+++ b/lib/dump_stack.c
@@ -0,0 +1,131 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Provide a default dump_stack() function for architectures
+ * which don't implement their own.
+ */
+
+#include <linux/kernel.h>
+#include <linux/buildid.h>
+#include <linux/export.h>
+#include <linux/sched.h>
+#include <linux/sched/debug.h>
+#include <linux/smp.h>
+#include <linux/atomic.h>
+#include <linux/kexec.h>
+#include <linux/utsname.h>
+#include <linux/stop_machine.h>
+
+static char dump_stack_arch_desc_str[128];
+
+/**
+ * dump_stack_set_arch_desc - set arch-specific str to show with task dumps
+ * @fmt: printf-style format string
+ * @...: arguments for the format string
+ *
+ * The configured string will be printed right after utsname during task
+ * dumps.  Usually used to add arch-specific system identifiers.  If an
+ * arch wants to make use of such an ID string, it should initialize this
+ * as soon as possible during boot.
+ */
+void __init dump_stack_set_arch_desc(const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	vsnprintf(dump_stack_arch_desc_str, sizeof(dump_stack_arch_desc_str),
+		  fmt, args);
+	va_end(args);
+}
+
+#if IS_ENABLED(CONFIG_STACKTRACE_BUILD_ID)
+#define BUILD_ID_FMT " %20phN"
+#define BUILD_ID_VAL vmlinux_build_id
+#else
+#define BUILD_ID_FMT "%s"
+#define BUILD_ID_VAL ""
+#endif
+
+/**
+ * dump_stack_print_info - print generic debug info for dump_stack()
+ * @log_lvl: log level
+ *
+ * Arch-specific dump_stack() implementations can use this function to
+ * print out the same debug information as the generic dump_stack().
+ */
+void dump_stack_print_info(const char *log_lvl)
+{
+	printk("%sCPU: %d UID: %u PID: %d Comm: %.20s %s%s %s %.*s %s " BUILD_ID_FMT "\n",
+	       log_lvl, raw_smp_processor_id(),
+	       __kuid_val(current_real_cred()->euid),
+	       current->pid, current->comm,
+	       kexec_crash_loaded() ? "Kdump: loaded " : "",
+	       print_tainted(),
+	       init_utsname()->release,
+	       (int)strcspn(init_utsname()->version, " "),
+	       init_utsname()->version, preempt_model_str(), BUILD_ID_VAL);
+
+	if (get_taint())
+		printk("%s%s\n", log_lvl, print_tainted_verbose());
+
+	if (dump_stack_arch_desc_str[0] != '\0')
+		printk("%sHardware name: %s\n",
+		       log_lvl, dump_stack_arch_desc_str);
+
+	print_worker_info(log_lvl, current);
+	print_stop_info(log_lvl, current);
+	print_scx_info(log_lvl, current);
+}
+
+/**
+ * show_regs_print_info - print generic debug info for show_regs()
+ * @log_lvl: log level
+ *
+ * show_regs() implementations can use this function to print out generic
+ * debug information.
+ */
+void show_regs_print_info(const char *log_lvl)
+{
+	dump_stack_print_info(log_lvl);
+}
+
+static void __dump_stack(const char *log_lvl)
+{
+	dump_stack_print_info(log_lvl);
+	show_stack(NULL, NULL, log_lvl);
+}
+
+/**
+ * dump_stack_lvl - dump the current task information and its stack trace
+ * @log_lvl: log level
+ *
+ * Architectures can override this implementation by implementing its own.
+ */
+asmlinkage __visible void dump_stack_lvl(const char *log_lvl)
+{
+	bool in_panic = panic_on_this_cpu();
+	unsigned long flags;
+
+	/*
+	 * Permit this cpu to perform nested stack dumps while serialising
+	 * against other CPUs, unless this CPU is in panic.
+	 *
+	 * When in panic, non-panic CPUs are not permitted to store new
+	 * printk messages so there is no need to synchronize the output.
+	 * This avoids potential deadlock in panic() if another CPU is
+	 * holding and unable to release the printk_cpu_sync.
+	 */
+	if (!in_panic)
+		printk_cpu_sync_get_irqsave(flags);
+
+	__dump_stack(log_lvl);
+
+	if (!in_panic)
+		printk_cpu_sync_put_irqrestore(flags);
+}
+EXPORT_SYMBOL(dump_stack_lvl);
+
+asmlinkage __visible void dump_stack(void)
+{
+	dump_stack_lvl(KERN_DEFAULT);
+}
+EXPORT_SYMBOL(dump_stack);