Allow clean exits from the scheduler with scx_bpf_exit()

include/linux/sched/ext.h

@@ -56,7 +56,8 @@ enum scx_exit_kind {
 	SCX_EXIT_NONE,
 	SCX_EXIT_DONE,
 
-	SCX_EXIT_UNREG = 64,	/* BPF unregistration */
+	SCX_EXIT_UNREG = 64,	/* User-space initiated unregistration */
+	SCX_EXIT_UNREG_BPF,	/* BPF-initiated unregistration */
 	SCX_EXIT_SYSRQ,		/* requested by 'S' sysrq */
 
 	SCX_EXIT_ERROR = 1024,	/* runtime error, error msg contains details */
@@ -72,6 +73,9 @@ struct scx_exit_info {
 	/* %SCX_EXIT_* - broad category of the exit reason */
 	enum scx_exit_kind	kind;
 
+	/* exit code if gracefully exiting from BPF */
+	s64			exit_code;
+
 	/* textual representation of the above */
 	const char		*reason;
 

kernel/sched/ext.c

@@ -3304,7 +3304,9 @@ static const char *scx_exit_reason(enum scx_exit_kind kind)
 {
 	switch (kind) {
 	case SCX_EXIT_UNREG:
-		return "BPF scheduler unregistered";
+		return "Scheduler unregistered from user space";
+	case SCX_EXIT_UNREG_BPF:
+		return "Scheduler unregistered from BPF";
 	case SCX_EXIT_SYSRQ:
 		return "disabled by sysrq-S";
 	case SCX_EXIT_ERROR:
@@ -3606,8 +3608,9 @@ static void scx_ops_error_irq_workfn(struct irq_work *irq_work)
 
 static DEFINE_IRQ_WORK(scx_ops_error_irq_work, scx_ops_error_irq_workfn);
 
-__printf(2, 3) void scx_ops_error_kind(enum scx_exit_kind kind,
-				       const char *fmt, ...)
+__printf(3, 4) void scx_ops_exit_kind(enum scx_exit_kind kind,
+				      s64 exit_code,
+				      const char *fmt, ...)
 {
 	struct scx_exit_info *ei = scx_exit_info;
 	int none = SCX_EXIT_NONE;
@@ -3616,12 +3619,15 @@ __printf(2, 3) void scx_ops_error_kind(enum scx_exit_kind kind,
 	if (!atomic_try_cmpxchg(&scx_exit_kind, &none, kind))
 		return;
 
-	ei->bt_len = stack_trace_save(ei->bt, SCX_EXIT_BT_LEN, 1);
+	if (kind >= SCX_EXIT_ERROR)
+		ei->bt_len = stack_trace_save(ei->bt, SCX_EXIT_BT_LEN, 1);
 
 	va_start(args, fmt);
 	vscnprintf(ei->msg, SCX_EXIT_MSG_LEN, fmt, args);
 	va_end(args);
 
+	ei->exit_code = exit_code;
+
 	/*
 	 * Set ei->kind and ->reason for scx_dump_state(). They'll be set again
 	 * in scx_ops_disable_workfn().
@@ -5059,17 +5065,9 @@ struct scx_bpf_error_bstr_bufs {
 
 static DEFINE_PER_CPU(struct scx_bpf_error_bstr_bufs, scx_bpf_error_bstr_bufs);
 
-/**
- * scx_bpf_error_bstr - Indicate fatal error
- * @fmt: error message format string
- * @data: format string parameters packaged using ___bpf_fill() macro
- * @data__sz: @data len, must end in '__sz' for the verifier
- *
- * Indicate that the BPF scheduler encountered a fatal error and initiate ops
- * disabling.
- */
-__bpf_kfunc void scx_bpf_error_bstr(char *fmt, unsigned long long *data,
-				    u32 data__sz)
+static void bpf_exit_bstr_common(enum scx_exit_kind kind, s64 exit_code,
+				 char *fmt, unsigned long long *data,
+				 u32 data__sz)
 {
 	struct bpf_bprintf_data bprintf_data = { .get_bin_args = true };
 	struct scx_bpf_error_bstr_bufs *bufs;
@@ -5108,9 +5106,44 @@ __bpf_kfunc void scx_bpf_error_bstr(char *fmt, unsigned long long *data,
 		goto out_restore;
 	}
 
-	scx_ops_error_kind(SCX_EXIT_ERROR_BPF, "%s", bufs->msg);
+	scx_ops_exit_kind(kind, exit_code, "%s", bufs->msg);
 out_restore:
 	local_irq_restore(flags);
+
+}
+
+/**
+ * scx_bpf_exit_bstr - Gracefully exit the BPF scheduler.
+ * @exit_code: Exit value to pass to user space via struct scx_exit_info.
+ * @fmt: error message format string
+ * @data: format string parameters packaged using ___bpf_fill() macro
+ * @data__sz: @data len, must end in '__sz' for the verifier
+ *
+ * Indicate that the BPF scheduler wants to exit gracefully, and initiate ops
+ * disabling.
+ */
+__bpf_kfunc void scx_bpf_exit_bstr(s64 exit_code, char *fmt,
+				   unsigned long long *data, u32 data__sz)
+{
+	bpf_exit_bstr_common(SCX_EXIT_UNREG_BPF, exit_code, fmt, data,
+			     data__sz);
+}
+
+/**
+ * scx_bpf_error_bstr - Indicate fatal error
+ * @fmt: error message format string
+ * @data: format string parameters packaged using ___bpf_fill() macro
+ * @data__sz: @data len, must end in '__sz' for the verifier
+ *
+ * Indicate that the BPF scheduler encountered a fatal error and initiate ops
+ * disabling.
+ */
+__bpf_kfunc void scx_bpf_error_bstr(char *fmt, unsigned long long *data,
+				    u32 data__sz)
+{
+
+	bpf_exit_bstr_common(SCX_EXIT_ERROR_BPF, 0, fmt, data,
+			     data__sz);
 }
 
 /**
@@ -5200,6 +5233,7 @@ BTF_ID_FLAGS(func, scx_bpf_get_idle_cpumask, KF_ACQUIRE)
 BTF_ID_FLAGS(func, scx_bpf_get_idle_smtmask, KF_ACQUIRE)
 BTF_ID_FLAGS(func, scx_bpf_put_idle_cpumask, KF_RELEASE)
 BTF_ID_FLAGS(func, scx_bpf_error_bstr, KF_TRUSTED_ARGS)
+BTF_ID_FLAGS(func, scx_bpf_exit_bstr, KF_TRUSTED_ARGS)
 BTF_ID_FLAGS(func, scx_bpf_task_running, KF_RCU)
 BTF_ID_FLAGS(func, scx_bpf_task_cpu, KF_RCU)
 #ifdef CONFIG_CGROUP_SCHED

kernel/sched/ext.h

@@ -145,8 +145,12 @@ int scx_check_setscheduler(struct task_struct *p, int policy);
 bool scx_can_stop_tick(struct rq *rq);
 void init_sched_ext_class(void);
 
-__printf(2, 3) void scx_ops_error_kind(enum scx_exit_kind kind,
-				       const char *fmt, ...);
+__printf(3, 4) void scx_ops_exit_kind(enum scx_exit_kind kind,
+				      s64 exit_code,
+				      const char *fmt, ...);
+#define scx_ops_error_kind(__err, fmt, args...)					\
+	scx_ops_exit_kind(__err, 0, fmt, ##args)
+
 #define scx_ops_error(fmt, args...)						\
 	scx_ops_error_kind(SCX_EXIT_ERROR, fmt, ##args)
 

tools/sched_ext/include/scx/common.bpf.h

@@ -29,28 +29,53 @@ static inline void ___vmlinux_h_sanity_check___(void)
 }
 
 void scx_bpf_error_bstr(char *fmt, unsigned long long *data, u32 data_len) __ksym;
+void scx_bpf_exit_bstr(s64 exit_code, char *fmt,
+		       unsigned long long *data, u32 data__sz) __ksym;
 
 static inline __attribute__((format(printf, 1, 2)))
-void ___scx_bpf_error_format_checker(const char *fmt, ...) {}
+void ___scx_bpf_exit_format_checker(const char *fmt, ...) {}
+
+/*
+ * Helper macro for initializing the fmt and variadic argument inputs to both
+ * bstr exit kfuncs. Callers to this function should use ___fmt and ___param to
+ * refer to the initialized list of inputs to the bstr kfunc.
+ */
+#define scx_bpf_exit_preamble(fmt, args...)				\
+	static char ___fmt[] = fmt;					\
+	/*								\
+	 * Note that __param[] must have at least one			\
+	 * element to keep the verifier happy.				\
+	 */								\
+	unsigned long long ___param[___bpf_narg(args) ?: 1] = {};	\
+									\
+	_Pragma("GCC diagnostic push")					\
+	_Pragma("GCC diagnostic ignored \"-Wint-conversion\"")		\
+	___bpf_fill(___param, args);					\
+	_Pragma("GCC diagnostic pop")					\
+
+/*
+ * scx_bpf_exit() wraps the scx_bpf_exit_bstr() kfunc with variadic arguments
+ * instead of an array of u64. Using this macro will cause the scheduler to
+ * exit cleanly with the specified exit code being passed to user space.
+ */
+#define scx_bpf_exit(code, fmt, args...)					\
+({										\
+	scx_bpf_exit_preamble(fmt, args)					\
+	scx_bpf_exit_bstr(code, ___fmt, ___param, sizeof(___param));		\
+	___scx_bpf_exit_format_checker(fmt, ##args);				\
+})
 
 /*
  * scx_bpf_error() wraps the scx_bpf_error_bstr() kfunc with variadic arguments
- * instead of an array of u64. Note that __param[] must have at least one
- * element to keep the verifier happy.
+ * instead of an array of u64. Invoking this macro will cause the scheduler to
+ * exit in an erroneous state, with diagnostic information being passed to the
+ * user.
  */
 #define scx_bpf_error(fmt, args...)						\
 ({										\
-	static char ___fmt[] = fmt;						\
-	unsigned long long ___param[___bpf_narg(args) ?: 1] = {};		\
-										\
-	_Pragma("GCC diagnostic push")						\
-	_Pragma("GCC diagnostic ignored \"-Wint-conversion\"")			\
-	___bpf_fill(___param, args);						\
-	_Pragma("GCC diagnostic pop")						\
-										\
+	scx_bpf_exit_preamble(fmt, args)					\
 	scx_bpf_error_bstr(___fmt, ___param, sizeof(___param));			\
-										\
-	___scx_bpf_error_format_checker(fmt, ##args);				\
+	___scx_bpf_exit_format_checker(fmt, ##args);				\
 })
 
 s32 scx_bpf_create_dsq(u64 dsq_id, s32 node) __ksym;

tools/sched_ext/include/scx/user_exit_info.h

@@ -18,6 +18,7 @@ enum uei_sizes {
 
 struct user_exit_info {
 	int		kind;
+	s64		exit_code;
 	char		reason[UEI_REASON_LEN];
 	char		msg[UEI_MSG_LEN];
 };
@@ -39,6 +40,8 @@ struct user_exit_info {
 				  sizeof(__uei_name.msg), (__ei)->msg);		\
 	bpf_probe_read_kernel_str(__uei_name##_dump,				\
 				  __uei_name##_dump_len, (__ei)->dump);		\
+	if (bpf_core_field_exists((__ei)->exit_code))				\
+		__uei_name.exit_code = (__ei)->exit_code;			\
 	/* use __sync to force memory barrier */				\
 	__sync_val_compare_and_swap(&__uei_name.kind, __uei_name.kind,		\
 				    (__ei)->kind);				\

tools/testing/selftests/sched_ext/Makefile

@@ -165,6 +165,7 @@ auto-test-targets :=			\
 	enq_select_cpu_fails		\
 	ddsp_bogus_dsq_fail		\
 	ddsp_vtimelocal_fail		\
+	exit				\
 	init_enable_count		\
 	maximal				\
 	maybe_null			\

tools/testing/selftests/sched_ext/exit.bpf.c

@@ -0,0 +1,84 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2024 Meta Platforms, Inc. and affiliates.
+ * Copyright (c) 2024 David Vernet <[email protected]>
+ */
+
+#include <scx/common.bpf.h>
+
+char _license[] SEC("license") = "GPL";
+
+#include "exit_test.h"
+
+const volatile int exit_point;
+UEI_DEFINE(uei);
+
+#define EXIT_CLEANLY() scx_bpf_exit(exit_point, "%d", exit_point)
+
+s32 BPF_STRUCT_OPS(exit_select_cpu, struct task_struct *p,
+		   s32 prev_cpu, u64 wake_flags)
+{
+	bool found;
+
+	if (exit_point == EXIT_SELECT_CPU)
+		EXIT_CLEANLY();
+
+	return scx_bpf_select_cpu_dfl(p, prev_cpu, wake_flags, &found);
+}
+
+void BPF_STRUCT_OPS(exit_enqueue, struct task_struct *p, u64 enq_flags)
+{
+	if (exit_point == EXIT_ENQUEUE)
+		EXIT_CLEANLY();
+
+	scx_bpf_dispatch(p, SCX_DSQ_GLOBAL, SCX_SLICE_DFL, enq_flags);
+}
+
+void BPF_STRUCT_OPS(exit_dispatch, s32 cpu, struct task_struct *p)
+{
+	if (exit_point == EXIT_DISPATCH)
+		EXIT_CLEANLY();
+
+	scx_bpf_consume(SCX_DSQ_GLOBAL);
+}
+
+void BPF_STRUCT_OPS(exit_enable, struct task_struct *p)
+{
+	if (exit_point == EXIT_ENABLE)
+		EXIT_CLEANLY();
+}
+
+s32 BPF_STRUCT_OPS(exit_init_task, struct task_struct *p,
+		    struct scx_init_task_args *args)
+{
+	if (exit_point == EXIT_INIT_TASK)
+		EXIT_CLEANLY();
+
+	return 0;
+}
+
+void BPF_STRUCT_OPS(exit_exit, struct scx_exit_info *ei)
+{
+	UEI_RECORD(uei, ei);
+}
+
+s32 BPF_STRUCT_OPS_SLEEPABLE(exit_init)
+{
+	if (exit_point == EXIT_INIT)
+		EXIT_CLEANLY();
+
+	return 0;
+}
+
+SEC(".struct_ops.link")
+struct sched_ext_ops exit_ops = {
+	.select_cpu		= exit_select_cpu,
+	.enqueue		= exit_enqueue,
+	.dispatch		= exit_dispatch,
+	.init_task		= exit_init_task,
+	.enable			= exit_enable,
+	.exit			= exit_exit,
+	.init			= exit_init,
+	.name			= "exit",
+	.timeout_ms		= 1000U,
+};

tools/testing/selftests/sched_ext/exit.c

@@ -0,0 +1,55 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2024 Meta Platforms, Inc. and affiliates.
+ * Copyright (c) 2024 David Vernet <[email protected]>
+ */
+#include <bpf/bpf.h>
+#include <sched.h>
+#include <scx/common.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include "exit.bpf.skel.h"
+#include "scx_test.h"
+
+#include "exit_test.h"
+
+static enum scx_test_status run(void *ctx)
+{
+	enum exit_test_case tc;
+
+	for (tc = 0; tc < NUM_EXITS; tc++) {
+		struct exit *skel;
+		struct bpf_link *link;
+		char buf[16];
+
+		skel = exit__open();
+		skel->rodata->exit_point = tc;
+		exit__load(skel);
+		link = bpf_map__attach_struct_ops(skel->maps.exit_ops);
+		if (!link) {
+			SCX_ERR("Failed to attach scheduler");
+			exit__destroy(skel);
+			return SCX_TEST_FAIL;
+		}
+
+		/* Assumes uei.kind is written last */
+		while (skel->data->uei.kind == SCX_EXIT_NONE)
+			sched_yield();
+
+		SCX_EQ(skel->data->uei.kind, SCX_EXIT_UNREG_BPF);
+		SCX_EQ(skel->data->uei.exit_code, tc);
+		sprintf(buf, "%d", tc);
+		SCX_ASSERT(!strcmp(skel->data->uei.msg, buf));
+		bpf_link__destroy(link);
+		exit__destroy(skel);
+	}
+
+	return SCX_TEST_PASS;
+}
+
+struct scx_test exit_test = {
+	.name = "exit",
+	.description = "Verify we can cleanly exit a scheduler in multiple places",
+	.run = run,
+};
+REGISTER_SCX_TEST(&exit_test)