Skip to contentSkip to conversation

cgroup:cpuset:separate legacy cgroup v1 code and put under config option

PATCH Applied Chen Ridong 3 participants 15 messages Aug 30, 2024

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

v4 cover cgroup:cpuset:separate legacy cgroup v1 code and put under config option

Cgroups v2 have been around for a while and many users have fully adopted them, so they never use cgroups v1 features and functionality. Yet they have to "pay" for the cgroup v1 support anyway:
1) the kernel binary contains an unused cgroup v1 code,
2) some code paths have additional checks which are not needed,
3) some common structures like task_struct and mem_cgroup contain unused cgroup v1-specific members.

Cgroup memory controller has already separated legacy code to memory-v1.c. So it is time to do the same thing for cpuset controller.

This patchset aims to do:
1) moving cgroup v1-specific cpuset code to the new cpuset-v1.c file,
2) putting definitions shared by cpuset.c and cpuset-v1.c into the cpuset-internal.h header,
3) introducing the CONFIG_CPUSETS_V1 config option, turned off by default,
4) making cpuset-v1.c to compile only if CONFIG_CPUSETS_V1 is set.

---
V4:
- Rename legacy_files to cpuset1_files.
- Revert rebuild_sched_domain and fmeter_init rename.

V3:
- Delete blank line at the end of file.
- Rename some generic functions name with cpuset_/cpuset1_ prefix.

V2:
- Update to base on the latest cgroup/for-6.12.
- Add CONFIG_CPUSETS_V1 for cpuset_memory_pressure_bump.

Chen Ridong (12):
cgroup/cpuset: introduce cpuset-v1.c cgroup/cpuset: move common code to cpuset-internal.h cgroup/cpuset: move memory_pressure to cpuset-v1.c cgroup/cpuset: move relax_domain_level to cpuset-v1.c cgroup/cpuset: move memory_spread to cpuset-v1.c cgroup/cpuset: add callback_lock helper cgroup/cpuset: move legacy hotplug update to cpuset-v1.c cgroup/cpuset: move validate_change_legacy to cpuset-v1.c cgroup/cpuset: move v1 interfaces to cpuset-v1.c cgroup/cpuset: rename functions shared between v1 and v2 cgroup/cpuset: guard cpuset-v1 code under CONFIG_CPUSETS_V1 cgroup/cpuset: add sefltest for cpuset v1

 MAINTAINERS                                   |   3 +
 include/linux/cpuset.h                        |   4 +
 init/Kconfig                                  |  13 +
 kernel/cgroup/Makefile                        |   1 +
 kernel/cgroup/cpuset-internal.h               | 304 ++++++
 kernel/cgroup/cpuset-v1.c                     | 562 +++++++++++
 kernel/cgroup/cpuset.c                        | 892 +-----------------
 .../selftests/cgroup/test_cpuset_v1_base.sh   |  77 ++
 8 files changed, 1005 insertions(+), 851 deletions(-)
 create mode 100644 kernel/cgroup/cpuset-internal.h
 create mode 100644 kernel/cgroup/cpuset-v1.c
 create mode 100755 tools/testing/selftests/cgroup/test_cpuset_v1_base.sh
Signature · 1 line
2.34.1
#1 lore raw

PATCH v4 1/12 cgroup/cpuset: introduce cpuset-v1.c

This patch introduces the cgroup/cpuset-v1.c source file which will be used for all legacy (cgroup v1) cpuset cgroup code. It also introduces cgroup/cpuset-internal.h to keep declarations shared between cgroup/cpuset.c and cpuset/cpuset-v1.c.

As of now, let's compile it if CONFIG_CPUSET is set. Later on it can be switched to use a separate config option, so that the legacy code won't be compiled if not required.

---
 MAINTAINERS                     | 2 ++
 kernel/cgroup/Makefile          | 2 +-
 kernel/cgroup/cpuset-internal.h | 6 ++++++
 kernel/cgroup/cpuset-v1.c       | 3 +++
 4 files changed, 12 insertions(+), 1 deletion(-)
 create mode 100644 kernel/cgroup/cpuset-internal.h
 create mode 100644 kernel/cgroup/cpuset-v1.c
MAINTAINERS +2 -0
index 82e3924816d2..3b5ec1cafd95 100644
@@ -5698,6 +5698,8 @@ S:	Maintained
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup.git
 F:	Documentation/admin-guide/cgroup-v1/cpusets.rst
 F:	include/linux/cpuset.h
+F:	kernel/cgroup/cpuset-internal.h
+F:	kernel/cgroup/cpuset-v1.c
 F:	kernel/cgroup/cpuset.c
 F:	tools/testing/selftests/cgroup/test_cpuset.c
 F:	tools/testing/selftests/cgroup/test_cpuset_prs.sh
kernel/cgroup/Makefile +1 -1
index 12f8457ad1f9..005ac4c675cb 100644
@@ -4,6 +4,6 @@ obj-y := cgroup.o rstat.o namespace.o cgroup-v1.o freezer.o
 obj-$(CONFIG_CGROUP_FREEZER) += legacy_freezer.o
 obj-$(CONFIG_CGROUP_PIDS) += pids.o
 obj-$(CONFIG_CGROUP_RDMA) += rdma.o
-obj-$(CONFIG_CPUSETS) += cpuset.o
+obj-$(CONFIG_CPUSETS) += cpuset.o cpuset-v1.o
 obj-$(CONFIG_CGROUP_MISC) += misc.o
 obj-$(CONFIG_CGROUP_DEBUG) += debug.o
kernel/cgroup/cpuset-internal.h new file +6 -0
new file mode 100644
index 000000000000..034de3cbf3ad
@@ -0,0 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+#ifndef __CPUSET_INTERNAL_H
+#define __CPUSET_INTERNAL_H
+
+#endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c new file +3 -0
new file mode 100644
index 000000000000..bdec4b196986
@@ -0,0 +1,3 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include "cpuset-internal.h"
Signature · 1 line
2.34.1
#2 lore raw

PATCH v4 2/12 cgroup/cpuset: move common code to cpuset-internal.h

In reply to Chen Ridong

Move some declarations that will be used for cpuset v1 and v2, including 'cpuset struct', 'cpuset_flagbits_t', cpuset_filetype_t,etc. No logical change.

---
 kernel/cgroup/cpuset-internal.h | 235 +++++++++++++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 236 +-------------------------------
 2 files changed, 236 insertions(+), 235 deletions(-)
kernel/cgroup/cpuset-internal.h +235 -0
index 034de3cbf3ad..ffea3eefebdf 100644
@@ -3,4 +3,239 @@
 #ifndef __CPUSET_INTERNAL_H
 #define __CPUSET_INTERNAL_H
 
+#include <linux/union_find.h>
+#include <linux/cpumask.h>
+#include <linux/spinlock.h>
+#include <linux/cpuset.h>
+#include <linux/cgroup.h>
+
+/* See "Frequency meter" comments, below. */
+
+struct fmeter {
+	int cnt;		/* unprocessed events count */
+	int val;		/* most recent output value */
+	time64_t time;		/* clock (secs) when val computed */
+	spinlock_t lock;	/* guards read or write of above */
+};
+
+/*
+ * Invalid partition error code
+ */
+enum prs_errcode {
+	PERR_NONE = 0,
+	PERR_INVCPUS,
+	PERR_INVPARENT,
+	PERR_NOTPART,
+	PERR_NOTEXCL,
+	PERR_NOCPUS,
+	PERR_HOTPLUG,
+	PERR_CPUSEMPTY,
+	PERR_HKEEPING,
+	PERR_ACCESS,
+};
+
+/* bits in struct cpuset flags field */
+typedef enum {
+	CS_ONLINE,
+	CS_CPU_EXCLUSIVE,
+	CS_MEM_EXCLUSIVE,
+	CS_MEM_HARDWALL,
+	CS_MEMORY_MIGRATE,
+	CS_SCHED_LOAD_BALANCE,
+	CS_SPREAD_PAGE,
+	CS_SPREAD_SLAB,
+} cpuset_flagbits_t;
+
+/* The various types of files and directories in a cpuset file system */
+
+typedef enum {
+	FILE_MEMORY_MIGRATE,
+	FILE_CPULIST,
+	FILE_MEMLIST,
+	FILE_EFFECTIVE_CPULIST,
+	FILE_EFFECTIVE_MEMLIST,
+	FILE_SUBPARTS_CPULIST,
+	FILE_EXCLUSIVE_CPULIST,
+	FILE_EFFECTIVE_XCPULIST,
+	FILE_ISOLATED_CPULIST,
+	FILE_CPU_EXCLUSIVE,
+	FILE_MEM_EXCLUSIVE,
+	FILE_MEM_HARDWALL,
+	FILE_SCHED_LOAD_BALANCE,
+	FILE_PARTITION_ROOT,
+	FILE_SCHED_RELAX_DOMAIN_LEVEL,
+	FILE_MEMORY_PRESSURE_ENABLED,
+	FILE_MEMORY_PRESSURE,
+	FILE_SPREAD_PAGE,
+	FILE_SPREAD_SLAB,
+} cpuset_filetype_t;
+
+struct cpuset {
+	struct cgroup_subsys_state css;
+
+	unsigned long flags;		/* "unsigned long" so bitops work */
+
+	/*
+	 * On default hierarchy:
+	 *
+	 * The user-configured masks can only be changed by writing to
+	 * cpuset.cpus and cpuset.mems, and won't be limited by the
+	 * parent masks.
+	 *
+	 * The effective masks is the real masks that apply to the tasks
+	 * in the cpuset. They may be changed if the configured masks are
+	 * changed or hotplug happens.
+	 *
+	 * effective_mask == configured_mask & parent's effective_mask,
+	 * and if it ends up empty, it will inherit the parent's mask.
+	 *
+	 *
+	 * On legacy hierarchy:
+	 *
+	 * The user-configured masks are always the same with effective masks.
+	 */
+
+	/* user-configured CPUs and Memory Nodes allow to tasks */
+	cpumask_var_t cpus_allowed;
+	nodemask_t mems_allowed;
+
+	/* effective CPUs and Memory Nodes allow to tasks */
+	cpumask_var_t effective_cpus;
+	nodemask_t effective_mems;
+
+	/*
+	 * Exclusive CPUs dedicated to current cgroup (default hierarchy only)
+	 *
+	 * The effective_cpus of a valid partition root comes solely from its
+	 * effective_xcpus and some of the effective_xcpus may be distributed
+	 * to sub-partitions below & hence excluded from its effective_cpus.
+	 * For a valid partition root, its effective_cpus have no relationship
+	 * with cpus_allowed unless its exclusive_cpus isn't set.
+	 *
+	 * This value will only be set if either exclusive_cpus is set or
+	 * when this cpuset becomes a local partition root.
+	 */
+	cpumask_var_t effective_xcpus;
+
+	/*
+	 * Exclusive CPUs as requested by the user (default hierarchy only)
+	 *
+	 * Its value is independent of cpus_allowed and designates the set of
+	 * CPUs that can be granted to the current cpuset or its children when
+	 * it becomes a valid partition root. The effective set of exclusive
+	 * CPUs granted (effective_xcpus) depends on whether those exclusive
+	 * CPUs are passed down by its ancestors and not yet taken up by
+	 * another sibling partition root along the way.
+	 *
+	 * If its value isn't set, it defaults to cpus_allowed.
+	 */
+	cpumask_var_t exclusive_cpus;
+
+	/*
+	 * This is old Memory Nodes tasks took on.
+	 *
+	 * - top_cpuset.old_mems_allowed is initialized to mems_allowed.
+	 * - A new cpuset's old_mems_allowed is initialized when some
+	 *   task is moved into it.
+	 * - old_mems_allowed is used in cpuset_migrate_mm() when we change
+	 *   cpuset.mems_allowed and have tasks' nodemask updated, and
+	 *   then old_mems_allowed is updated to mems_allowed.
+	 */
+	nodemask_t old_mems_allowed;
+
+	struct fmeter fmeter;		/* memory_pressure filter */
+
+	/*
+	 * Tasks are being attached to this cpuset.  Used to prevent
+	 * zeroing cpus/mems_allowed between ->can_attach() and ->attach().
+	 */
+	int attach_in_progress;
+
+	/* for custom sched domain */
+	int relax_domain_level;
+
+	/* number of valid local child partitions */
+	int nr_subparts;
+
+	/* partition root state */
+	int partition_root_state;
+
+	/*
+	 * number of SCHED_DEADLINE tasks attached to this cpuset, so that we
+	 * know when to rebuild associated root domain bandwidth information.
+	 */
+	int nr_deadline_tasks;
+	int nr_migrate_dl_tasks;
+	u64 sum_migrate_dl_bw;
+
+	/* Invalid partition error code, not lock protected */
+	enum prs_errcode prs_err;
+
+	/* Handle for cpuset.cpus.partition */
+	struct cgroup_file partition_file;
+
+	/* Remote partition silbling list anchored at remote_children */
+	struct list_head remote_sibling;
+
+	/* Used to merge intersecting subsets for generate_sched_domains */
+	struct uf_node node;
+};
+
+static inline struct cpuset *css_cs(struct cgroup_subsys_state *css)
+{
+	return css ? container_of(css, struct cpuset, css) : NULL;
+}
+
+/* Retrieve the cpuset for a task */
+static inline struct cpuset *task_cs(struct task_struct *task)
+{
+	return css_cs(task_css(task, cpuset_cgrp_id));
+}
+
+static inline struct cpuset *parent_cs(struct cpuset *cs)
+{
+	return css_cs(cs->css.parent);
+}
+
+/* convenient tests for these bits */
+static inline bool is_cpuset_online(struct cpuset *cs)
+{
+	return test_bit(CS_ONLINE, &cs->flags) && !css_is_dying(&cs->css);
+}
+
+static inline int is_cpu_exclusive(const struct cpuset *cs)
+{
+	return test_bit(CS_CPU_EXCLUSIVE, &cs->flags);
+}
+
+static inline int is_mem_exclusive(const struct cpuset *cs)
+{
+	return test_bit(CS_MEM_EXCLUSIVE, &cs->flags);
+}
+
+static inline int is_mem_hardwall(const struct cpuset *cs)
+{
+	return test_bit(CS_MEM_HARDWALL, &cs->flags);
+}
+
+static inline int is_sched_load_balance(const struct cpuset *cs)
+{
+	return test_bit(CS_SCHED_LOAD_BALANCE, &cs->flags);
+}
+
+static inline int is_memory_migrate(const struct cpuset *cs)
+{
+	return test_bit(CS_MEMORY_MIGRATE, &cs->flags);
+}
+
+static inline int is_spread_page(const struct cpuset *cs)
+{
+	return test_bit(CS_SPREAD_PAGE, &cs->flags);
+}
+
+static inline int is_spread_slab(const struct cpuset *cs)
+{
+	return test_bit(CS_SPREAD_SLAB, &cs->flags);
+}
+
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset.c +1 -235

284 lines not on this page · raw message

Signature · 1 line
2.34.1
#3 lore raw

PATCH v4 3/12 cgroup/cpuset: move memory_pressure to cpuset-v1.c

In reply to Chen Ridong

Collection of memory_pressure can be enabled by writing 1 to the cpuset file 'memory_pressure_enabled', which is only for cpuset-v1. Therefore, move the corresponding code to cpuset-v1.c.

Currently, the 'fmeter_init' and 'fmeter_getrate' functions are called at cpuset.c, so expose them to cpuset.c.

---
 kernel/cgroup/cpuset-internal.h |   7 ++
 kernel/cgroup/cpuset-v1.c       | 134 ++++++++++++++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 134 --------------------------------
 3 files changed, 141 insertions(+), 134 deletions(-)
kernel/cgroup/cpuset-internal.h +7 -0
index ffea3eefebdf..7911c86bf012 100644
@@ -238,4 +238,11 @@ static inline int is_spread_slab(const struct cpuset *cs)
 	return test_bit(CS_SPREAD_SLAB, &cs->flags);
 }
 
+/*
+ * cpuset-v1.c
+ */
+
+void fmeter_init(struct fmeter *fmp);
+int fmeter_getrate(struct fmeter *fmp);
+
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +134 -0

137 lines not on this page · raw message

kernel/cgroup/cpuset.c +0 -134

146 lines not on this page · raw message

Signature · 1 line
2.34.1
#4 lore raw

PATCH v4 4/12 cgroup/cpuset: move relax_domain_level to cpuset-v1.c

In reply to Chen Ridong

Setting domain level is not supported at cpuset v2, so move corresponding code into cpuset-v1.c.

The 'cpuset_write_s64' and 'cpuset_read_s64' are only used for setting domain level, move them to cpuset-v1.c. Currently, expose to cpuset.c. After cpuset legacy interface files are move to cpuset-v1.c, they can be static. The 'rebuild_sched_domains_locked' is exposed to cpuset-v1.c.

The change from original code is that using 'cpuset_lock' and 'cpuset_unlock' functions to lock or unlock cpuset_mutex.

---
 kernel/cgroup/cpuset-internal.h |  6 +++-
 kernel/cgroup/cpuset-v1.c       | 59 +++++++++++++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 62 ++-------------------------------
 3 files changed, 66 insertions(+), 61 deletions(-)
kernel/cgroup/cpuset-internal.h +5 -1
index 7911c86bf012..1058a45f05ec 100644
@@ -238,11 +238,15 @@ static inline int is_spread_slab(const struct cpuset *cs)
 	return test_bit(CS_SPREAD_SLAB, &cs->flags);
 }
 
+void rebuild_sched_domains_locked(void);
+
 /*
  * cpuset-v1.c
  */
-
 void fmeter_init(struct fmeter *fmp);
 int fmeter_getrate(struct fmeter *fmp);
+int cpuset_write_s64(struct cgroup_subsys_state *css, struct cftype *cft,
+			    s64 val);
+s64 cpuset_read_s64(struct cgroup_subsys_state *css, struct cftype *cft);
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +59 -0

62 lines not on this page · raw message

kernel/cgroup/cpuset.c +2 -60

92 lines not on this page · raw message

Signature · 1 line
2.34.1
#5 lore raw

PATCH v4 5/12 cgroup/cpuset: move memory_spread to cpuset-v1.c

In reply to Chen Ridong

'memory_spread' is only set in cpuset v1. move corresponding code into cpuset-v1.c.

Currently, 'cpuset_update_task_spread_flags' and 'update_tasks_flags' are exposed to cpuset.c.

---
 kernel/cgroup/cpuset-internal.h |  3 +++
 kernel/cgroup/cpuset-v1.c       | 42 +++++++++++++++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 42 ---------------------------------
 3 files changed, 45 insertions(+), 42 deletions(-)
kernel/cgroup/cpuset-internal.h +3 -0
index 1058a45f05ec..02c4b0c74fa9 100644
@@ -248,5 +248,8 @@ int fmeter_getrate(struct fmeter *fmp);
 int cpuset_write_s64(struct cgroup_subsys_state *css, struct cftype *cft,
 			    s64 val);
 s64 cpuset_read_s64(struct cgroup_subsys_state *css, struct cftype *cft);
+void cpuset_update_task_spread_flags(struct cpuset *cs,
+					struct task_struct *tsk);
+void update_tasks_flags(struct cpuset *cs);
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +42 -0

45 lines not on this page · raw message

kernel/cgroup/cpuset.c +0 -42

54 lines not on this page · raw message

Signature · 1 line
2.34.1
#6 lore raw

PATCH v4 6/12 cgroup/cpuset: add callback_lock helper

In reply to Chen Ridong

To modify cpuset, both cpuset_mutex and callback_lock are needed. Add helpers for cpuset-v1 to get callback_lock.

---
 kernel/cgroup/cpuset-internal.h |  2 ++
 kernel/cgroup/cpuset.c          | 10 ++++++++++
 2 files changed, 12 insertions(+)
kernel/cgroup/cpuset-internal.h +2 -0
index 02c4b0c74fa9..9a60dd6681e4 100644
@@ -239,6 +239,8 @@ static inline int is_spread_slab(const struct cpuset *cs)
 }
 
 void rebuild_sched_domains_locked(void);
+void callback_lock_irq(void);
+void callback_unlock_irq(void);
 
 /*
  * cpuset-v1.c
kernel/cgroup/cpuset.c +10 -0
index 0a3347e4dddc..2b2dc963299b 100644
@@ -269,6 +269,16 @@ void cpuset_unlock(void)
 
 static DEFINE_SPINLOCK(callback_lock);
 
+void callback_lock_irq(void)
+{
+	spin_lock_irq(&callback_lock);
+}
+
+void callback_unlock_irq(void)
+{
+	spin_unlock_irq(&callback_lock);
+}
+
 static struct workqueue_struct *cpuset_migrate_mm_wq;
 
 static DECLARE_WAIT_QUEUE_HEAD(cpuset_attach_wq);
Signature · 1 line
2.34.1
#7 lore raw

PATCH v4 7/12 cgroup/cpuset: move legacy hotplug update to cpuset-v1.c

In reply to Chen Ridong

There are some differents about hotplug update between cpuset v1 and cpuset v2. Move the legacy code to cpuset-v1.c.

'update_tasks_cpumask' and 'update_tasks_nodemask' are both used in cpuset v1 and cpuset v2, declare them in cpuset-internal.h.

The change from original code is that use callback_lock helpers to get callback_lock lock/unlock.

---
 kernel/cgroup/cpuset-internal.h |  5 ++
 kernel/cgroup/cpuset-v1.c       | 91 +++++++++++++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 96 +--------------------------------
 3 files changed, 98 insertions(+), 94 deletions(-)
kernel/cgroup/cpuset-internal.h +5 -0
index 9a60dd6681e4..7cd30ad809d5 100644
@@ -241,6 +241,8 @@ static inline int is_spread_slab(const struct cpuset *cs)
 void rebuild_sched_domains_locked(void);
 void callback_lock_irq(void);
 void callback_unlock_irq(void);
+void update_tasks_cpumask(struct cpuset *cs, struct cpumask *new_cpus);
+void update_tasks_nodemask(struct cpuset *cs);
 
 /*
  * cpuset-v1.c
@@ -253,5 +255,8 @@ s64 cpuset_read_s64(struct cgroup_subsys_state *css, struct cftype *cft);
 void cpuset_update_task_spread_flags(struct cpuset *cs,
 					struct task_struct *tsk);
 void update_tasks_flags(struct cpuset *cs);
+void hotplug_update_tasks_legacy(struct cpuset *cs,
+			    struct cpumask *new_cpus, nodemask_t *new_mems,
+			    bool cpus_updated, bool mems_updated);
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +91 -0

100 lines not on this page · raw message

kernel/cgroup/cpuset.c +2 -94

120 lines not on this page · raw message

Signature · 1 line
2.34.1
#8 lore raw

PATCH v4 8/12 cgroup/cpuset: move validate_change_legacy to cpuset-v1.c

In reply to Chen Ridong

The validate_change_legacy functions is used for v1, move it to cpuset-v1.c. And two micro 'cpuset_for_each_child' and 'cpuset_for_each_descendant_pre' are common for v1 and v2, move them to cpuset-internal.h.

---
 kernel/cgroup/cpuset-internal.h | 29 +++++++++++++
 kernel/cgroup/cpuset-v1.c       | 45 ++++++++++++++++++++
 kernel/cgroup/cpuset.c          | 73 ---------------------------------
 3 files changed, 74 insertions(+), 73 deletions(-)
kernel/cgroup/cpuset-internal.h +29 -0
index 7cd30ad809d5..07551ff0812e 100644
@@ -238,6 +238,34 @@ static inline int is_spread_slab(const struct cpuset *cs)
 	return test_bit(CS_SPREAD_SLAB, &cs->flags);
 }
 
+/**
+ * cpuset_for_each_child - traverse online children of a cpuset
+ * @child_cs: loop cursor pointing to the current child
+ * @pos_css: used for iteration
+ * @parent_cs: target cpuset to walk children of
+ *
+ * Walk @child_cs through the online children of @parent_cs.  Must be used
+ * with RCU read locked.
+ */
+#define cpuset_for_each_child(child_cs, pos_css, parent_cs)		\
+	css_for_each_child((pos_css), &(parent_cs)->css)		\
+		if (is_cpuset_online(((child_cs) = css_cs((pos_css)))))
+
+/**
+ * cpuset_for_each_descendant_pre - pre-order walk of a cpuset's descendants
+ * @des_cs: loop cursor pointing to the current descendant
+ * @pos_css: used for iteration
+ * @root_cs: target cpuset to walk ancestor of
+ *
+ * Walk @des_cs through the online descendants of @root_cs.  Must be used
+ * with RCU read locked.  The caller may modify @pos_css by calling
+ * css_rightmost_descendant() to skip subtree.  @root_cs is included in the
+ * iteration and the first node to be visited.
+ */
+#define cpuset_for_each_descendant_pre(des_cs, pos_css, root_cs)	\
+	css_for_each_descendant_pre((pos_css), &(root_cs)->css)		\
+		if (is_cpuset_online(((des_cs) = css_cs((pos_css)))))
+
 void rebuild_sched_domains_locked(void);
 void callback_lock_irq(void);
 void callback_unlock_irq(void);
@@ -258,5 +286,6 @@ void update_tasks_flags(struct cpuset *cs);
 void hotplug_update_tasks_legacy(struct cpuset *cs,
 			    struct cpumask *new_cpus, nodemask_t *new_mems,
 			    bool cpus_updated, bool mems_updated);
+int validate_change_legacy(struct cpuset *cur, struct cpuset *trial);
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +45 -0

48 lines not on this page · raw message

kernel/cgroup/cpuset.c +0 -73

91 lines not on this page · raw message

Signature · 1 line
2.34.1
#9 lore raw

PATCH v4 9/12 cgroup/cpuset: move v1 interfaces to cpuset-v1.c

In reply to Chen Ridong

Move legacy cpuset controller interfaces files and corresponding code into cpuset-v1.c. 'update_flag', 'cpuset_write_resmask' and 'cpuset_common_seq_show' are also used for v1, so declare them in cpuset-internal.h.

'cpuset_write_s64', 'cpuset_read_s64' and 'fmeter_getrate' are only used cpuset-v1.c now, make it static.

---
 kernel/cgroup/cpuset-internal.h |   9 +-
 kernel/cgroup/cpuset-v1.c       | 194 ++++++++++++++++++++++++++++++-
 kernel/cgroup/cpuset.c          | 195 +-------------------------------
 3 files changed, 199 insertions(+), 199 deletions(-)
kernel/cgroup/cpuset-internal.h +5 -4
index 07551ff0812e..a6c71c86e58d 100644
@@ -271,15 +271,16 @@ void callback_lock_irq(void);
 void callback_unlock_irq(void);
 void update_tasks_cpumask(struct cpuset *cs, struct cpumask *new_cpus);
 void update_tasks_nodemask(struct cpuset *cs);
+int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, int turning_on);
+ssize_t cpuset_write_resmask(struct kernfs_open_file *of,
+				    char *buf, size_t nbytes, loff_t off);
+int cpuset_common_seq_show(struct seq_file *sf, void *v);
 
 /*
  * cpuset-v1.c
  */
+extern struct cftype legacy_files[];
 void fmeter_init(struct fmeter *fmp);
-int fmeter_getrate(struct fmeter *fmp);
-int cpuset_write_s64(struct cgroup_subsys_state *css, struct cftype *cft,
-			    s64 val);
-s64 cpuset_read_s64(struct cgroup_subsys_state *css, struct cftype *cft);
 void cpuset_update_task_spread_flags(struct cpuset *cs,
 					struct task_struct *tsk);
 void update_tasks_flags(struct cpuset *cs);
kernel/cgroup/cpuset-v1.c +191 -3

215 lines not on this page · raw message

kernel/cgroup/cpuset.c +3 -192

234 lines not on this page · raw message

Signature · 1 line
2.34.1
#10 lore raw

PATCH v4 10/12 cgroup/cpuset: rename functions shared between v1 and v2

In reply to Chen Ridong

Some functions name declared in cpuset-internel.h are generic. To avoid confilicting with other variables for the same name, rename these functions with cpuset_/cpuset1_ prefix to make them unique to cpuset.

---
 kernel/cgroup/cpuset-internal.h | 20 ++++++-------
 kernel/cgroup/cpuset-v1.c       | 40 ++++++++++++-------------
 kernel/cgroup/cpuset.c          | 52 ++++++++++++++++-----------------
 3 files changed, 56 insertions(+), 56 deletions(-)
kernel/cgroup/cpuset-internal.h +10 -10
index a6c71c86e58d..f36419d688bd 100644
@@ -267,11 +267,11 @@ static inline int is_spread_slab(const struct cpuset *cs)
 		if (is_cpuset_online(((des_cs) = css_cs((pos_css)))))
 
 void rebuild_sched_domains_locked(void);
-void callback_lock_irq(void);
-void callback_unlock_irq(void);
-void update_tasks_cpumask(struct cpuset *cs, struct cpumask *new_cpus);
-void update_tasks_nodemask(struct cpuset *cs);
-int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, int turning_on);
+void cpuset_callback_lock_irq(void);
+void cpuset_callback_unlock_irq(void);
+void cpuset_update_tasks_cpumask(struct cpuset *cs, struct cpumask *new_cpus);
+void cpuset_update_tasks_nodemask(struct cpuset *cs);
+int cpuset_update_flag(cpuset_flagbits_t bit, struct cpuset *cs, int turning_on);
 ssize_t cpuset_write_resmask(struct kernfs_open_file *of,
 				    char *buf, size_t nbytes, loff_t off);
 int cpuset_common_seq_show(struct seq_file *sf, void *v);
@@ -279,14 +279,14 @@ int cpuset_common_seq_show(struct seq_file *sf, void *v);
 /*
  * cpuset-v1.c
  */
-extern struct cftype legacy_files[];
+extern struct cftype cpuset1_files[];
 void fmeter_init(struct fmeter *fmp);
-void cpuset_update_task_spread_flags(struct cpuset *cs,
+void cpuset1_update_task_spread_flags(struct cpuset *cs,
 					struct task_struct *tsk);
-void update_tasks_flags(struct cpuset *cs);
-void hotplug_update_tasks_legacy(struct cpuset *cs,
+void cpuset1_update_tasks_flags(struct cpuset *cs);
+void cpuset1_hotplug_update_tasks(struct cpuset *cs,
 			    struct cpumask *new_cpus, nodemask_t *new_mems,
 			    bool cpus_updated, bool mems_updated);
-int validate_change_legacy(struct cpuset *cur, struct cpuset *trial);
+int cpuset1_validate_change(struct cpuset *cur, struct cpuset *trial);
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset-v1.c +20 -20
index 0ccc440c468a..25c1d7b77e2f 100644
@@ -209,7 +209,7 @@ static s64 cpuset_read_s64(struct cgroup_subsys_state *css, struct cftype *cft)
  * Call with callback_lock or cpuset_mutex held. The check can be skipped
  * if on default hierarchy.
  */
-void cpuset_update_task_spread_flags(struct cpuset *cs,
+void cpuset1_update_task_spread_flags(struct cpuset *cs,
 					struct task_struct *tsk)
 {
 	if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys))
@@ -227,21 +227,21 @@ void cpuset_update_task_spread_flags(struct cpuset *cs,
 }
 
 /**
- * update_tasks_flags - update the spread flags of tasks in the cpuset.
+ * cpuset1_update_tasks_flags - update the spread flags of tasks in the cpuset.
  * @cs: the cpuset in which each task's spread flags needs to be changed
  *
  * Iterate through each task of @cs updating its spread flags.  As this
  * function is called with cpuset_mutex held, cpuset membership stays
  * stable.
  */
-void update_tasks_flags(struct cpuset *cs)
+void cpuset1_update_tasks_flags(struct cpuset *cs)
 {
 	struct css_task_iter it;
 	struct task_struct *task;
 
 	css_task_iter_start(&cs->css, 0, &it);
 	while ((task = css_task_iter_next(&it)))
-		cpuset_update_task_spread_flags(cs, task);
+		cpuset1_update_task_spread_flags(cs, task);
 	css_task_iter_end(&it);
 }
 
@@ -282,27 +282,27 @@ static void cpuset_migrate_tasks_workfn(struct work_struct *work)
 	kfree(s);
 }
 
-void hotplug_update_tasks_legacy(struct cpuset *cs,
+void cpuset1_hotplug_update_tasks(struct cpuset *cs,
 			    struct cpumask *new_cpus, nodemask_t *new_mems,
 			    bool cpus_updated, bool mems_updated)
 {
 	bool is_empty;
 
-	callback_lock_irq();
+	cpuset_callback_lock_irq();
 	cpumask_copy(cs->cpus_allowed, new_cpus);
 	cpumask_copy(cs->effective_cpus, new_cpus);
 	cs->mems_allowed = *new_mems;
 	cs->effective_mems = *new_mems;
-	callback_unlock_irq();
+	cpuset_callback_unlock_irq();
 
 	/*
-	 * Don't call update_tasks_cpumask() if the cpuset becomes empty,
+	 * Don't call cpuset_update_tasks_cpumask() if the cpuset becomes empty,
 	 * as the tasks will be migrated to an ancestor.
 	 */
 	if (cpus_updated && !cpumask_empty(cs->cpus_allowed))
-		update_tasks_cpumask(cs, new_cpus);
+		cpuset_update_tasks_cpumask(cs, new_cpus);
 	if (mems_updated && !nodes_empty(cs->mems_allowed))
-		update_tasks_nodemask(cs);
+		cpuset_update_tasks_nodemask(cs);
 
 	is_empty = cpumask_empty(cs->cpus_allowed) ||
 		   nodes_empty(cs->mems_allowed);
@@ -345,10 +345,10 @@ static int is_cpuset_subset(const struct cpuset *p, const struct cpuset *q)
 }
 
 /*
- * validate_change_legacy() - Validate conditions specific to legacy (v1)
+ * cpuset1_validate_change() - Validate conditions specific to legacy (v1)
  *                            behavior.
  */
-int validate_change_legacy(struct cpuset *cur, struct cpuset *trial)
+int cpuset1_validate_change(struct cpuset *cur, struct cpuset *trial)
 {
 	struct cgroup_subsys_state *css;
 	struct cpuset *c, *par;
@@ -421,28 +421,28 @@ static int cpuset_write_u64(struct cgroup_subsys_state *css, struct cftype *cft,
 
 	switch (type) {
 	case FILE_CPU_EXCLUSIVE:
-		retval = update_flag(CS_CPU_EXCLUSIVE, cs, val);
+		retval = cpuset_update_flag(CS_CPU_EXCLUSIVE, cs, val);
 		break;
 	case FILE_MEM_EXCLUSIVE:
-		retval = update_flag(CS_MEM_EXCLUSIVE, cs, val);
+		retval = cpuset_update_flag(CS_MEM_EXCLUSIVE, cs, val);
 		break;
 	case FILE_MEM_HARDWALL:
-		retval = update_flag(CS_MEM_HARDWALL, cs, val);
+		retval = cpuset_update_flag(CS_MEM_HARDWALL, cs, val);
 		break;
 	case FILE_SCHED_LOAD_BALANCE:
-		retval = update_flag(CS_SCHED_LOAD_BALANCE, cs, val);
+		retval = cpuset_update_flag(CS_SCHED_LOAD_BALANCE, cs, val);
 		break;
 	case FILE_MEMORY_MIGRATE:
-		retval = update_flag(CS_MEMORY_MIGRATE, cs, val);
+		retval = cpuset_update_flag(CS_MEMORY_MIGRATE, cs, val);
 		break;
 	case FILE_MEMORY_PRESSURE_ENABLED:
 		cpuset_memory_pressure_enabled = !!val;
 		break;
 	case FILE_SPREAD_PAGE:
-		retval = update_flag(CS_SPREAD_PAGE, cs, val);
+		retval = cpuset_update_flag(CS_SPREAD_PAGE, cs, val);
 		break;
 	case FILE_SPREAD_SLAB:
-		retval = update_flag(CS_SPREAD_SLAB, cs, val);
+		retval = cpuset_update_flag(CS_SPREAD_SLAB, cs, val);
 		break;
 	default:
 		retval = -EINVAL;
@@ -458,7 +458,7 @@ static int cpuset_write_u64(struct cgroup_subsys_state *css, struct cftype *cft,
  * for the common functions, 'private' gives the type of file
  */
 
-struct cftype legacy_files[] = {
+struct cftype cpuset1_files[] = {
 	{
 		.name = "cpus",
 		.seq_show = cpuset_common_seq_show,
kernel/cgroup/cpuset.c +26 -26

198 lines not on this page · raw message

Signature · 1 line
2.34.1
#11 lore raw

PATCH v4 11/12 cgroup/cpuset: guard cpuset-v1 code under CONFIG_CPUSETS_V1

In reply to Chen Ridong

This patch introduces CONFIG_CPUSETS_V1 and guard cpuset-v1 code under CONFIG_CPUSETS_V1. The default value of CONFIG_CPUSETS_V1 is N, so that user who adopted v2 don't have 'pay' for cpuset v1.

---
 include/linux/cpuset.h          |  4 ++++
 init/Kconfig                    | 13 +++++++++++++
 kernel/cgroup/Makefile          |  3 ++-
 kernel/cgroup/cpuset-internal.h | 12 ++++++++++++
 kernel/cgroup/cpuset.c          |  2 ++
 5 files changed, 33 insertions(+), 1 deletion(-)
include/linux/cpuset.h +4 -0
index 2a6981eeebf8..835e7b793f6a 100644
@@ -99,6 +99,7 @@ static inline bool cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask)
 extern int cpuset_mems_allowed_intersects(const struct task_struct *tsk1,
 					  const struct task_struct *tsk2);
 
+#ifdef CONFIG_CPUSETS_V1
 #define cpuset_memory_pressure_bump() 				\
 	do {							\
 		if (cpuset_memory_pressure_enabled)		\
@@ -106,6 +107,9 @@ extern int cpuset_mems_allowed_intersects(const struct task_struct *tsk1,
 	} while (0)
 extern int cpuset_memory_pressure_enabled;
 extern void __cpuset_memory_pressure_bump(void);
+#else
+static inline void cpuset_memory_pressure_bump(void) { }
+#endif
 
 extern void cpuset_task_status_allowed(struct seq_file *m,
 					struct task_struct *task);
init/Kconfig +13 -0
index a465ea9525bd..8bf091354bea 100644
@@ -1143,6 +1143,19 @@ config CPUSETS
 
 	  Say N if unsure.
 
+config CPUSETS_V1
+	bool "Legacy cgroup v1 cpusets controller"
+	depends on CPUSETS
+	default n
+	help
+	  Legacy cgroup v1 cpusets controller which has been deprecated by
+	  cgroup v2 implementation. The v1 is there for legacy applications
+	  which haven't migrated to the new cgroup v2 interface yet. If you
+	  do not have any such application then you are completely fine leaving
+	  this option disabled.
+
+	  Say N if unsure.
+
 config PROC_PID_CPUSET
 	bool "Include legacy /proc/<pid>/cpuset file"
 	depends on CPUSETS
kernel/cgroup/Makefile +2 -1
index 005ac4c675cb..a5c9359d516f 100644
@@ -4,6 +4,7 @@ obj-y := cgroup.o rstat.o namespace.o cgroup-v1.o freezer.o
 obj-$(CONFIG_CGROUP_FREEZER) += legacy_freezer.o
 obj-$(CONFIG_CGROUP_PIDS) += pids.o
 obj-$(CONFIG_CGROUP_RDMA) += rdma.o
-obj-$(CONFIG_CPUSETS) += cpuset.o cpuset-v1.o
+obj-$(CONFIG_CPUSETS) += cpuset.o
+obj-$(CONFIG_CPUSETS_V1) += cpuset-v1.o
 obj-$(CONFIG_CGROUP_MISC) += misc.o
 obj-$(CONFIG_CGROUP_DEBUG) += debug.o
kernel/cgroup/cpuset-internal.h +12 -0
index f36419d688bd..8c113d46ddd3 100644
@@ -279,6 +279,7 @@ int cpuset_common_seq_show(struct seq_file *sf, void *v);
 /*
  * cpuset-v1.c
  */
+#ifdef CONFIG_CPUSETS_V1
 extern struct cftype cpuset1_files[];
 void fmeter_init(struct fmeter *fmp);
 void cpuset1_update_task_spread_flags(struct cpuset *cs,
@@ -288,5 +289,16 @@ void cpuset1_hotplug_update_tasks(struct cpuset *cs,
 			    struct cpumask *new_cpus, nodemask_t *new_mems,
 			    bool cpus_updated, bool mems_updated);
 int cpuset1_validate_change(struct cpuset *cur, struct cpuset *trial);
+#else
+static inline void fmeter_init(struct fmeter *fmp) {}
+static inline void cpuset1_update_task_spread_flags(struct cpuset *cs,
+					struct task_struct *tsk) {}
+static inline void cpuset1_update_tasks_flags(struct cpuset *cs) {}
+static inline void cpuset1_hotplug_update_tasks(struct cpuset *cs,
+			    struct cpumask *new_cpus, nodemask_t *new_mems,
+			    bool cpus_updated, bool mems_updated) {}
+static inline int cpuset1_validate_change(struct cpuset *cur,
+				struct cpuset *trial) { return 0; }
+#endif /* CONFIG_CPUSETS_V1 */
 
 #endif /* __CPUSET_INTERNAL_H */
kernel/cgroup/cpuset.c +2 -0
index b63c577cacd7..9e6faaece5fa 100644
@@ -3617,7 +3617,9 @@ struct cgroup_subsys cpuset_cgrp_subsys = {
 	.can_fork	= cpuset_can_fork,
 	.cancel_fork	= cpuset_cancel_fork,
 	.fork		= cpuset_fork,
+#ifdef CONFIG_CPUSETS_V1
 	.legacy_cftypes	= cpuset1_files,
+#endif
 	.dfl_cftypes	= dfl_files,
 	.early_init	= true,
 	.threaded	= true,
Signature · 1 line
2.34.1
#12 lore raw

PATCH v4 12/12 cgroup/cpuset: add sefltest for cpuset v1

In reply to Chen Ridong

There is only hotplug test for cpuset v1, just add base read/write test for cpuset v1.

---
 MAINTAINERS                                   |  1 +
 .../selftests/cgroup/test_cpuset_v1_base.sh   | 77 +++++++++++++++++++
 2 files changed, 78 insertions(+)
 create mode 100755 tools/testing/selftests/cgroup/test_cpuset_v1_base.sh
MAINTAINERS +1 -0
index 3b5ec1cafd95..b59f54e1e30d 100644
@@ -5703,6 +5703,7 @@ F:	kernel/cgroup/cpuset-v1.c
 F:	kernel/cgroup/cpuset.c
 F:	tools/testing/selftests/cgroup/test_cpuset.c
 F:	tools/testing/selftests/cgroup/test_cpuset_prs.sh
+F:	tools/testing/selftests/cgroup/test_cpuset_v1_base.sh
 
 CONTROL GROUP - MEMORY RESOURCE CONTROLLER (MEMCG)
 M:	Johannes Weiner <[email protected]>
tools/testing/selftests/cgroup/test_cpuset_v1_base.sh new file +77 -0

77 lines not on this page · raw message

Signature · 1 line
2.34.1
maintainer Acked-by on the v4 series
#13 lore raw
In reply to Chen Ridong
Quoted 54 lines · Chen Ridong

Cgroups v2 have been around for a while and many users have fully adopted them, so they never use cgroups v1 features and functionality. Yet they have to "pay" for the cgroup v1 support anyway:
1) the kernel binary contains an unused cgroup v1 code,
2) some code paths have additional checks which are not needed,
3) some common structures like task_struct and mem_cgroup contain unused cgroup v1-specific members.

Cgroup memory controller has already separated legacy code to memory-v1.c. So it is time to do the same thing for cpuset controller.

This patchset aims to do:
1) moving cgroup v1-specific cpuset code to the new cpuset-v1.c file,
2) putting definitions shared by cpuset.c and cpuset-v1.c into the cpuset-internal.h header,
3) introducing the CONFIG_CPUSETS_V1 config option, turned off by default,
4) making cpuset-v1.c to compile only if CONFIG_CPUSETS_V1 is set.

---
V4:
- Rename legacy_files to cpuset1_files.
- Revert rebuild_sched_domain and fmeter_init rename.

V3:
- Delete blank line at the end of file.
- Rename some generic functions name with cpuset_/cpuset1_ prefix.

V2:
- Update to base on the latest cgroup/for-6.12.
- Add CONFIG_CPUSETS_V1 for cpuset_memory_pressure_bump.

Chen Ridong (12):
cgroup/cpuset: introduce cpuset-v1.c cgroup/cpuset: move common code to cpuset-internal.h cgroup/cpuset: move memory_pressure to cpuset-v1.c cgroup/cpuset: move relax_domain_level to cpuset-v1.c cgroup/cpuset: move memory_spread to cpuset-v1.c cgroup/cpuset: add callback_lock helper cgroup/cpuset: move legacy hotplug update to cpuset-v1.c cgroup/cpuset: move validate_change_legacy to cpuset-v1.c cgroup/cpuset: move v1 interfaces to cpuset-v1.c cgroup/cpuset: rename functions shared between v1 and v2 cgroup/cpuset: guard cpuset-v1 code under CONFIG_CPUSETS_V1 cgroup/cpuset: add sefltest for cpuset v1

  MAINTAINERS                                   |   3 +
  include/linux/cpuset.h                        |   4 +
  init/Kconfig                                  |  13 +
  kernel/cgroup/Makefile                        |   1 +
  kernel/cgroup/cpuset-internal.h               | 304 ++++++
  kernel/cgroup/cpuset-v1.c                     | 562 +++++++++++
  kernel/cgroup/cpuset.c                        | 892 +-----------------
  .../selftests/cgroup/test_cpuset_v1_base.sh   |  77 ++
  8 files changed, 1005 insertions(+), 851 deletions(-)
Quoted 4 lines

create mode 100644 kernel/cgroup/cpuset-internal.h
create mode 100644 kernel/cgroup/cpuset-v1.c
create mode 100755 tools/testing/selftests/cgroup/test_cpuset_v1_base.sh

For the series,

maintainer Applied
#14 lore raw
In reply to Chen Ridong
Quoted 11 lines · Chen Ridong

Cgroups v2 have been around for a while and many users have fully adopted them, so they never use cgroups v1 features and functionality. Yet they have to "pay" for the cgroup v1 support anyway:
1) the kernel binary contains an unused cgroup v1 code,
2) some code paths have additional checks which are not needed,
3) some common structures like task_struct and mem_cgroup contain unused cgroup v1-specific members.

Cgroup memory controller has already separated legacy code to memory-v1.c. So it is time to do the same thing for cpuset controller.

Quoted 6 lines

This patchset aims to do:
1) moving cgroup v1-specific cpuset code to the new cpuset-v1.c file,
2) putting definitions shared by cpuset.c and cpuset-v1.c into the cpuset-internal.h header,
3) introducing the CONFIG_CPUSETS_V1 config option, turned off by default,
4) making cpuset-v1.c to compile only if CONFIG_CPUSETS_V1 is set.

Applied the series to cgroup/for-6.12.

Thanks.

Signature · 1 line
tejun