Skip to contentSkip to conversation

tools/sched_ext: Receive updates from SCX repo

PATCH Applied Andrea Righi 2 participants 2 messages Aug 4, 2025 – Aug 11, 2025 Scheduler - sched_ext

lore
Share & tools
lore

Andrea Righi

author reviewer
#0 lore raw

PATCH 1/1 tools/sched_ext: Receive updates from SCX repo

Receive tools/sched_ext updates form https://github.com/sched-ext/scx to sync userspace bits:

- basic BPF arena allocator abstractions,

- additional process flags definitions,

- fixed is_migration_disabled() helper,

- separate out user_exit_info BPF and user space code.

This also fixes the following warning when building the selftests:

tools/sched_ext/include/scx/common.bpf.h:550:9: warning: 'likely' macro redefined [-Wmacro-redefined]
550 | #define likely(x) __builtin_expect(!!(x), 1)
| ^

---
 .../include/scx/bpf_arena_common.bpf.h        | 175 ++++++++++++++++++
 .../sched_ext/include/scx/bpf_arena_common.h  |  33 ++++
 tools/sched_ext/include/scx/common.bpf.h      | 102 +++++++++-
 tools/sched_ext/include/scx/common.h          |   5 +-
 tools/sched_ext/include/scx/compat.bpf.h      |   5 +
 .../include/scx/user_exit_info.bpf.h          |  40 ++++
 tools/sched_ext/include/scx/user_exit_info.h  |  49 +----
 .../include/scx/user_exit_info_common.h       |  30 +++
 tools/sched_ext/scx_central.bpf.c             |   2 +-
 tools/sched_ext/scx_central.c                 |   1 +
 tools/sched_ext/scx_flatcg.bpf.c              |   2 +-
 tools/sched_ext/scx_flatcg.c                  |   2 +
 tools/sched_ext/scx_simple.c                  |   2 +
 13 files changed, 388 insertions(+), 60 deletions(-)
 create mode 100644 tools/sched_ext/include/scx/bpf_arena_common.bpf.h
 create mode 100644 tools/sched_ext/include/scx/bpf_arena_common.h
 create mode 100644 tools/sched_ext/include/scx/user_exit_info.bpf.h
 create mode 100644 tools/sched_ext/include/scx/user_exit_info_common.h

ChangeLog:
- Import only the bare minimum changes required by the scheduler examples and the selftests
- Link to the previous patch:
https://lore.kernel.org/all/[email protected]/

tools/sched_ext/include/scx/bpf_arena_common.bpf.h new file +175 -0
new file mode 100644
index 0000000000000..4366fb3c91ce8
@@ -0,0 +1,175 @@
+/* SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) */
+/* Copyright (c) 2024 Meta Platforms, Inc. and affiliates. */
+#pragma once
+
+#ifndef PAGE_SIZE
+#define PAGE_SIZE __PAGE_SIZE
+/*
+ * for older kernels try sizeof(struct genradix_node)
+ * or flexible:
+ * static inline long __bpf_page_size(void) {
+ *   return bpf_core_enum_value(enum page_size_enum___l, __PAGE_SIZE___l) ?: sizeof(struct genradix_node);
+ * }
+ * but generated code is not great.
+ */
+#endif
+
+#if defined(__BPF_FEATURE_ADDR_SPACE_CAST) && !defined(BPF_ARENA_FORCE_ASM)
+#define __arena __attribute__((address_space(1)))
+#define __arena_global __attribute__((address_space(1)))
+#define cast_kern(ptr) /* nop for bpf prog. emitted by LLVM */
+#define cast_user(ptr) /* nop for bpf prog. emitted by LLVM */
+#else
+
+/* emit instruction:
+ * rX = rX .off = BPF_ADDR_SPACE_CAST .imm32 = (dst_as << 16) | src_as
+ *
+ * This is a workaround for LLVM compiler versions without
+ * __BPF_FEATURE_ADDR_SPACE_CAST that do not automatically cast between arena
+ * pointers and native kernel/userspace ones. In this case we explicitly do so
+ * with cast_kern() and cast_user(). E.g., in the Linux kernel tree,
+ * tools/testing/selftests/bpf includes tests that use these macros to implement
+ * linked lists and hashtables backed by arena memory. In sched_ext, we use
+ * cast_kern() and cast_user() for compatibility with older LLVM toolchains.
+ */
+#ifndef bpf_addr_space_cast
+#define bpf_addr_space_cast(var, dst_as, src_as)\
+	asm volatile(".byte 0xBF;		\
+		     .ifc %[reg], r0;		\
+		     .byte 0x00;		\
+		     .endif;			\
+		     .ifc %[reg], r1;		\
+		     .byte 0x11;		\
+		     .endif;			\
+		     .ifc %[reg], r2;		\
+		     .byte 0x22;		\
+		     .endif;			\
+		     .ifc %[reg], r3;		\
+		     .byte 0x33;		\
+		     .endif;			\
+		     .ifc %[reg], r4;		\
+		     .byte 0x44;		\
+		     .endif;			\
+		     .ifc %[reg], r5;		\
+		     .byte 0x55;		\
+		     .endif;			\
+		     .ifc %[reg], r6;		\
+		     .byte 0x66;		\
+		     .endif;			\
+		     .ifc %[reg], r7;		\
+		     .byte 0x77;		\
+		     .endif;			\
+		     .ifc %[reg], r8;		\
+		     .byte 0x88;		\
+		     .endif;			\
+		     .ifc %[reg], r9;		\
+		     .byte 0x99;		\
+		     .endif;			\
+		     .short %[off];		\
+		     .long %[as]"		\
+		     : [reg]"+r"(var)		\
+		     : [off]"i"(BPF_ADDR_SPACE_CAST) \
+		     , [as]"i"((dst_as << 16) | src_as));
+#endif
+
+#define __arena
+#define __arena_global SEC(".addr_space.1")
+#define cast_kern(ptr) bpf_addr_space_cast(ptr, 0, 1)
+#define cast_user(ptr) bpf_addr_space_cast(ptr, 1, 0)
+#endif
+
+void __arena* bpf_arena_alloc_pages(void *map, void __arena *addr, __u32 page_cnt,
+				    int node_id, __u64 flags) __ksym __weak;
+void bpf_arena_free_pages(void *map, void __arena *ptr, __u32 page_cnt) __ksym __weak;
+
+/*
+ * Note that cond_break can only be portably used in the body of a breakable
+ * construct, whereas can_loop can be used anywhere.
+ */
+#ifdef TEST
+#define can_loop true
+#define __cond_break(expr) expr
+#else
+#ifdef __BPF_FEATURE_MAY_GOTO
+#define can_loop					\
+	({ __label__ l_break, l_continue;		\
+	bool ret = true;				\
+	asm volatile goto("may_goto %l[l_break]"	\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: ret = false;				\
+	l_continue:;					\
+	ret;						\
+	})
+
+#define __cond_break(expr)				\
+	({ __label__ l_break, l_continue;		\
+	asm volatile goto("may_goto %l[l_break]"	\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: expr;					\
+	l_continue:;					\
+	})
+#else
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
+#define can_loop					\
+	({ __label__ l_break, l_continue;		\
+	bool ret = true;				\
+	asm volatile goto("1:.byte 0xe5;		\
+		      .byte 0;				\
+		      .long ((%l[l_break] - 1b - 8) / 8) & 0xffff;	\
+		      .short 0"				\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: ret = false;				\
+	l_continue:;					\
+	ret;						\
+	})
+
+#define __cond_break(expr)				\
+	({ __label__ l_break, l_continue;		\
+	asm volatile goto("1:.byte 0xe5;		\
+		      .byte 0;				\
+		      .long ((%l[l_break] - 1b - 8) / 8) & 0xffff;	\
+		      .short 0"				\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: expr;					\
+	l_continue:;					\
+	})
+#else
+#define can_loop					\
+	({ __label__ l_break, l_continue;		\
+	bool ret = true;				\
+	asm volatile goto("1:.byte 0xe5;		\
+		      .byte 0;				\
+		      .long (((%l[l_break] - 1b - 8) / 8) & 0xffff) << 16;	\
+		      .short 0"				\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: ret = false;				\
+	l_continue:;					\
+	ret;						\
+	})
+
+#define __cond_break(expr)				\
+	({ __label__ l_break, l_continue;		\
+	asm volatile goto("1:.byte 0xe5;		\
+		      .byte 0;				\
+		      .long (((%l[l_break] - 1b - 8) / 8) & 0xffff) << 16;	\
+		      .short 0"				\
+		      :::: l_break);			\
+	goto l_continue;				\
+	l_break: expr;					\
+	l_continue:;					\
+	})
+#endif /* __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ */
+#endif /* __BPF_FEATURE_MAY_GOTO */
+#endif /* TEST */
+
+#define cond_break __cond_break(break)
+#define cond_break_label(label) __cond_break(goto label)
+
+
+void bpf_preempt_disable(void) __weak __ksym;
+void bpf_preempt_enable(void) __weak __ksym;
tools/sched_ext/include/scx/bpf_arena_common.h new file +33 -0
new file mode 100644
index 0000000000000..10141db0b59d0
@@ -0,0 +1,33 @@
+/* SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) */
+/* Copyright (c) 2024 Meta Platforms, Inc. and affiliates. */
+#pragma once
+
+#ifndef arena_container_of
+#define arena_container_of(ptr, type, member)			\
+	({							\
+		void __arena *__mptr = (void __arena *)(ptr);	\
+		((type *)(__mptr - offsetof(type, member)));	\
+	})
+#endif
+
+/* Provide the definition of PAGE_SIZE. */
+#include <sys/user.h>
+
+#define __arena
+#define __arg_arena
+#define cast_kern(ptr) /* nop for user space */
+#define cast_user(ptr) /* nop for user space */
+char __attribute__((weak)) arena[1];
+
+#ifndef offsetof
+#define offsetof(type, member)  ((unsigned long)&((type *)0)->member)
+#endif
+
+static inline void __arena* bpf_arena_alloc_pages(void *map, void *addr, __u32 page_cnt,
+						  int node_id, __u64 flags)
+{
+	return NULL;
+}
+static inline void bpf_arena_free_pages(void *map, void __arena *ptr, __u32 page_cnt)
+{
+}
tools/sched_ext/include/scx/common.bpf.h +93 -9

194 lines not on this page · raw message

tools/sched_ext/include/scx/common.h +3 -2
index 1dc76bd842966..b3c6372bcf810 100644
@@ -75,8 +75,9 @@ typedef int64_t s64;
 #include "enums.h"
 
 /* not available when building kernel tools/sched_ext */
-#if __has_include(<lib/sdt_task.h>)
-#include <lib/sdt_task.h>
+#if __has_include(<lib/sdt_task_defs.h>)
+#include "bpf_arena_common.h"
+#include <lib/sdt_task_defs.h>
 #endif
 
 #endif	/* __SCHED_EXT_COMMON_H */
tools/sched_ext/include/scx/compat.bpf.h +5 -0
index 9252e1a00556f..36e0cd2fd4eda 100644
@@ -38,6 +38,7 @@ void scx_bpf_dispatch_from_dsq_set_slice___compat(struct bpf_iter_scx_dsq *it__i
 void scx_bpf_dispatch_from_dsq_set_vtime___compat(struct bpf_iter_scx_dsq *it__iter, u64 vtime) __ksym __weak;
 bool scx_bpf_dispatch_from_dsq___compat(struct bpf_iter_scx_dsq *it__iter, struct task_struct *p, u64 dsq_id, u64 enq_flags) __ksym __weak;
 bool scx_bpf_dispatch_vtime_from_dsq___compat(struct bpf_iter_scx_dsq *it__iter, struct task_struct *p, u64 dsq_id, u64 enq_flags) __ksym __weak;
+int bpf_cpumask_populate(struct cpumask *dst, void *src, size_t src__sz) __ksym __weak;
 
 #define scx_bpf_dsq_insert(p, dsq_id, slice, enq_flags)				\
 	(bpf_ksym_exists(scx_bpf_dsq_insert) ?					\
@@ -82,6 +83,10 @@ bool scx_bpf_dispatch_vtime_from_dsq___compat(struct bpf_iter_scx_dsq *it__iter,
 	  scx_bpf_dispatch_vtime_from_dsq___compat((it__iter), (p), (dsq_id), (enq_flags)) : \
 	  false))
 
+#define __COMPAT_bpf_cpumask_populate(cpumask, src, size__sz)		\
+	(bpf_ksym_exists(bpf_cpumask_populate) ?			\
+	 (bpf_cpumask_populate(cpumask, src, size__sz)) : -EOPNOTSUPP)
+
 #define scx_bpf_dispatch(p, dsq_id, slice, enq_flags)				\
 	_Static_assert(false, "scx_bpf_dispatch() renamed to scx_bpf_dsq_insert()")
 
tools/sched_ext/include/scx/user_exit_info.bpf.h new file +40 -0
new file mode 100644
index 0000000000000..e7ac6611a9901
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Define struct user_exit_info which is shared between BPF and userspace parts
+ * to communicate exit status and other information.
+ *
+ * Copyright (c) 2022 Meta Platforms, Inc. and affiliates.
+ * Copyright (c) 2022 Tejun Heo <[email protected]>
+ * Copyright (c) 2022 David Vernet <[email protected]>
+ */
+
+#ifndef __USER_EXIT_INFO_BPF_H
+#define __USER_EXIT_INFO_BPF_H
+
+#ifndef LSP
+#include "vmlinux.h"
+#endif
+#include <bpf/bpf_core_read.h>
+
+#include "user_exit_info_common.h"
+
+#define UEI_DEFINE(__name)							\
+	char RESIZABLE_ARRAY(data, __name##_dump);				\
+	const volatile u32 __name##_dump_len;					\
+	struct user_exit_info __name SEC(".data")
+
+#define UEI_RECORD(__uei_name, __ei) ({						\
+	bpf_probe_read_kernel_str(__uei_name.reason,				\
+				  sizeof(__uei_name.reason), (__ei)->reason);	\
+	bpf_probe_read_kernel_str(__uei_name.msg,				\
+				  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);				\
+})
+
+#endif /* __USER_EXIT_INFO_BPF_H */
tools/sched_ext/include/scx/user_exit_info.h +2 -47

62 lines not on this page · raw message

tools/sched_ext/include/scx/user_exit_info_common.h new file +30 -0
new file mode 100644
index 0000000000000..2d0981aedd898
@@ -0,0 +1,30 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Define struct user_exit_info which is shared between BPF and userspace parts
+ * to communicate exit status and other information.
+ *
+ * Copyright (c) 2022 Meta Platforms, Inc. and affiliates.
+ * Copyright (c) 2022 Tejun Heo <[email protected]>
+ * Copyright (c) 2022 David Vernet <[email protected]>
+ */
+#ifndef __USER_EXIT_INFO_COMMON_H
+#define __USER_EXIT_INFO_COMMON_H
+
+#ifdef LSP
+#include "../vmlinux.h"
+#endif
+
+enum uei_sizes {
+	UEI_REASON_LEN		= 128,
+	UEI_MSG_LEN		= 1024,
+	UEI_DUMP_DFL_LEN	= 32768,
+};
+
+struct user_exit_info {
+	int		kind;
+	s64		exit_code;
+	char		reason[UEI_REASON_LEN];
+	char		msg[UEI_MSG_LEN];
+};
+
+#endif /* __USER_EXIT_INFO_COMMON_H */
tools/sched_ext/scx_central.bpf.c +1 -1
index 50bc1737c167a..55df8b7988657 100644
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 /*
- * A central FIFO sched_ext scheduler which demonstrates the followings:
+ * A central FIFO sched_ext scheduler which demonstrates the following:
  *
  * a. Making all scheduling decisions from one CPU:
  *
tools/sched_ext/scx_central.c +1 -0
index 6ba6e610eeaa0..55931a4cd71c7 100644
@@ -61,6 +61,7 @@ int main(int argc, char **argv)
 	skel->rodata->nr_cpu_ids = libbpf_num_possible_cpus();
 	skel->rodata->slice_ns = __COMPAT_ENUM_OR_ZERO("scx_public_consts", "SCX_SLICE_DFL");
 
+	assert(skel->rodata->nr_cpu_ids > 0);
 	assert(skel->rodata->nr_cpu_ids <= INT32_MAX);
 
 	while ((opt = getopt(argc, argv, "s:c:pvh")) != -1) {
tools/sched_ext/scx_flatcg.bpf.c +1 -1
index fdc7170639e60..2c720e3ecad59 100644
@@ -950,5 +950,5 @@ SCX_OPS_DEFINE(flatcg_ops,
 	       .cgroup_move		= (void *)fcg_cgroup_move,
 	       .init			= (void *)fcg_init,
 	       .exit			= (void *)fcg_exit,
-	       .flags			= SCX_OPS_ENQ_EXITING,
+	       .flags			= SCX_OPS_HAS_CGROUP_WEIGHT | SCX_OPS_ENQ_EXITING,
 	       .name			= "flatcg");
tools/sched_ext/scx_flatcg.c +2 -0
index 6dd423eeb4ff9..cd85eb4011793 100644
@@ -6,6 +6,7 @@
  */
 #include <stdio.h>
 #include <signal.h>
+#include <assert.h>
 #include <unistd.h>
 #include <libgen.h>
 #include <limits.h>
@@ -137,6 +138,7 @@ int main(int argc, char **argv)
 	skel = SCX_OPS_OPEN(flatcg_ops, scx_flatcg);
 
 	skel->rodata->nr_cpus = libbpf_num_possible_cpus();
+	assert(skel->rodata->nr_cpus > 0);
 	skel->rodata->cgrp_slice_ns = __COMPAT_ENUM_OR_ZERO("scx_public_consts", "SCX_SLICE_DFL");
 
 	while ((opt = getopt(argc, argv, "s:i:dfvh")) != -1) {
tools/sched_ext/scx_simple.c +2 -0
index 76d83199545cb..06d4b13bf76bc 100644
@@ -7,6 +7,7 @@
 #include <stdio.h>
 #include <unistd.h>
 #include <signal.h>
+#include <assert.h>
 #include <libgen.h>
 #include <bpf/bpf.h>
 #include <scx/common.h>
@@ -41,6 +42,7 @@ static void sigint_handler(int simple)
 static void read_stats(struct scx_simple *skel, __u64 *stats)
 {
 	int nr_cpus = libbpf_num_possible_cpus();
+	assert(nr_cpus > 0);
 	__u64 cnts[2][nr_cpus];
 	__u32 idx;
 
Signature · 1 line
2.50.1
reviewer Applied
#1 lore raw
Quoted 11 lines · Andrea Righi

Receive tools/sched_ext updates form https://github.com/sched-ext/scx to sync userspace bits:

- basic BPF arena allocator abstractions,

- additional process flags definitions,

- fixed is_migration_disabled() helper,

- separate out user_exit_info BPF and user space code.

Quoted 8 lines

This also fixes the following warning when building the selftests:

tools/sched_ext/include/scx/common.bpf.h:550:9: warning: 'likely' macro redefined [-Wmacro-redefined]
550 | #define likely(x) __builtin_expect(!!(x), 1)
| ^

Applied to sched_ext/for-6.18.

Thanks.

Signature · 1 line
tejun