util-linux.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Qais Yousef <qais.yousef@arm.com>
To: util-linux@vger.kernel.org
Cc: Karel Zak <kzak@redhat.com>,
	Bernhard Voelker <mail@bernhard-voelker.de>,
	"Peter Zijlstra (Intel)" <peterz@infradead.org>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Patrick Bellasi <patrick.bellasi@matbug.net>,
	Vincent Donnefort <vincent.donnefort@arm.com>,
	Qais Yousef <qais.yousef@arm.com>
Subject: [PATCH v2 2/5] Add uclampset schedutil
Date: Sat, 30 Jan 2021 20:50:36 +0000	[thread overview]
Message-ID: <20210130205039.581764-3-qais.yousef@arm.com> (raw)
In-Reply-To: <20210130205039.581764-1-qais.yousef@arm.com>

Utilization clamping is a new kernel feature that got merged in 5.3. It
allows controlling the performance of a process by manipulating the
utilization such that the task appears bigger or smaller than what it
really is.

There's a system-wide control to to restrict what maximum values the
process are allowed to use.

Man page added in a later patch attempts to explain the usage in more
detail.

Signed-off-by: Qais Yousef <qais.yousef@arm.com>
---
 .gitignore              |   1 +
 include/pathnames.h     |   4 +
 schedutils/sched_attr.h |  30 ++++
 schedutils/uclampset.c  | 337 ++++++++++++++++++++++++++++++++++++++++
 4 files changed, 372 insertions(+)
 create mode 100644 schedutils/uclampset.c

diff --git a/.gitignore b/.gitignore
index a6f379146..c01d2f644 100644
--- a/.gitignore
+++ b/.gitignore
@@ -180,3 +180,4 @@ ylwrap
 /wipefs
 /write
 /zramctl
+/uclampset
diff --git a/include/pathnames.h b/include/pathnames.h
index 8f6233749..69a0b5524 100644
--- a/include/pathnames.h
+++ b/include/pathnames.h
@@ -207,4 +207,8 @@
 #define _PATH_DEV_RFKILL	"/dev/rfkill"
 #define _PATH_SYS_RFKILL	"/sys/class/rfkill"
 
+#define _PATH_PROC_KERNEL(file)	"/proc/sys/kernel" #file
+#define _PATH_PROC_UCLAMP_MIN	_PATH_PROC_KERNEL(/sched_util_clamp_min)
+#define _PATH_PROC_UCLAMP_MAX	_PATH_PROC_KERNEL(/sched_util_clamp_max)
+
 #endif /* PATHNAMES_H */
diff --git a/schedutils/sched_attr.h b/schedutils/sched_attr.h
index b39d37b5d..39d8532a3 100644
--- a/schedutils/sched_attr.h
+++ b/schedutils/sched_attr.h
@@ -13,6 +13,8 @@
  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  *
  * Copyright (C) 2004 Robert Love
+ * Copyright (C) 2020-2021 Qais Yousef
+ * Copyright (C) 2020-2021 Arm Ltd
  */
 #ifndef UTIL_LINUX_SCHED_ATTR_H
 #define UTIL_LINUX_SCHED_ATTR_H
@@ -42,6 +44,30 @@
 # define SCHED_FLAG_RESET_ON_FORK 0x01
 #endif
 
+#if defined(__linux__) && !defined(SCHED_FLAG_RECLAIM)
+# define SCHED_FLAG_RECLAIM 0x02
+#endif
+
+#if defined(__linux__) && !defined(SCHED_FLAG_DL_OVERRUN)
+# define SCHED_FLAG_DL_OVERRUN 0x04
+#endif
+
+#if defined(__linux__) && !defined(SCHED_FLAG_KEEP_POLICY)
+# define SCHED_FLAG_KEEP_POLICY 0x08
+#endif
+
+#if defined(__linux__) && !defined(SCHED_FLAG_KEEP_PARAMS)
+# define SCHED_FLAG_KEEP_PARAMS 0x10
+#endif
+
+#if defined(__linux__) && !defined(SCHED_FLAG_UTIL_CLAMP_MIN)
+# define SCHED_FLAG_UTIL_CLAMP_MIN 0x20
+#endif
+
+#if defined(__linux__) && !defined(SCHED_FLAG_UTIL_CLAMP_MAX)
+# define SCHED_FLAG_UTIL_CLAMP_MAX 0x40
+#endif
+
 #if defined (__linux__)
 # include <sys/syscall.h>
 #endif
@@ -73,6 +99,10 @@ struct sched_attr {
 	uint64_t sched_runtime;
 	uint64_t sched_deadline;
 	uint64_t sched_period;
+
+	/* UTILIZATION CLAMPING */
+	uint32_t sched_util_min;
+	uint32_t sched_util_max;
 };
 
 static int sched_setattr(pid_t pid, const struct sched_attr *attr, unsigned int flags)
diff --git a/schedutils/uclampset.c b/schedutils/uclampset.c
new file mode 100644
index 000000000..b9a02aae1
--- /dev/null
+++ b/schedutils/uclampset.c
@@ -0,0 +1,337 @@
+/*
+ * uclampset.c - change utilization clamping attributes of a task or the system
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Copyright (C) 2020-2021 Qais Yousef
+ * Copyright (C) 2020-2021 Arm Ltd
+ */
+
+#include <errno.h>
+#include <getopt.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "closestream.h"
+#include "path.h"
+#include "pathnames.h"
+#include "procutils.h"
+#include "sched_attr.h"
+#include "strutils.h"
+
+#define NOT_SET		-2U
+
+struct uclampset {
+	unsigned int util_min;
+	unsigned int util_max;
+
+	pid_t pid;
+	unsigned int	all_tasks:1,		/* all threads of the PID */
+			system:1,
+			util_min_set:1,		/* indicates -m option was passed */
+			util_max_set:1,		/* indicates -M option was passed */
+			reset_on_fork:1,
+			verbose:1;
+	char *cmd;
+};
+
+static void __attribute__((__noreturn__)) usage(void)
+{
+	FILE *out = stdout;
+
+	fprintf(out,
+		_(" %1$s [options]\n"
+		  " %1$s [options] --pid <pid> | --system | <command> <arg>...\n"),
+		program_invocation_short_name);
+
+	fputs(USAGE_SEPARATOR, out);
+	fputs(_("Show or change the utilization clamping attributes of a process or the system.\n"), out);
+	fputs(_("Utilization range: [0:1024]\n"), out);
+	fputs(_("Use special -1 value to reset to system's default\n"), out);
+
+	fputs(USAGE_OPTIONS, out);
+	fputs(_(" -m                   util_min value to set\n"), out);
+	fputs(_(" -M                   util_max value to set\n"), out);
+	fputs(_(" -a, --all-tasks      operate on all the tasks (threads) for a given pid\n"), out);
+	fputs(_(" -p, --pid            operate on existing given pid\n"), out);
+	fputs(_(" -s, --system         operate on system\n"), out);
+	fputs(_(" -R, --reset-on-fork  set reset-on-fork flag\n"), out);
+	fputs(_(" -v, --verbose        display status information\n"), out);
+
+	fputs(USAGE_SEPARATOR, out);
+	printf(USAGE_HELP_OPTIONS(22));
+
+	printf(USAGE_MAN_TAIL("uclampset(1)"));
+	exit(EXIT_SUCCESS);
+}
+
+static void show_uclamp_pid_info(pid_t pid, char *cmd)
+{
+	struct sched_attr sa;
+	char *comm;
+
+	/* don't display "pid 0" as that is confusing */
+	if (!pid)
+		pid = getpid();
+
+	if (sched_getattr(pid, &sa, sizeof(sa), 0) != 0)
+		err(EXIT_FAILURE, _("failed to get pid %d's uclamp values"), pid);
+
+	if (cmd)
+		comm = cmd;
+	else
+		comm = proc_get_command_name(pid);
+
+	printf(_("%s (%d) util_clamp: min: %d max: %d\n"),
+	       comm ? : "uknown", pid, sa.sched_util_min, sa.sched_util_max);
+
+	if (!cmd)
+		free(comm);
+}
+
+static unsigned int read_uclamp_sysfs(char *filename)
+{
+	unsigned int val;
+
+	if (ul_path_read_u32(NULL, &val, filename) != 0)
+		err(EXIT_FAILURE, _("cannot read %s"), filename);
+
+	return val;
+}
+
+static void write_uclamp_sysfs(char *filename, unsigned int val)
+{
+	if (ul_path_write_u64(NULL, val, filename) != 0)
+		err(EXIT_FAILURE, _("cannot write %s"), filename);
+}
+
+static void show_uclamp_system_info(void)
+{
+	unsigned int min, max;
+
+	min = read_uclamp_sysfs(_PATH_PROC_UCLAMP_MIN);
+	max = read_uclamp_sysfs(_PATH_PROC_UCLAMP_MAX);
+
+	printf(_("System util_clamp: min: %u max: %u\n"), min, max);
+}
+
+static void show_uclamp_info(struct uclampset *ctl)
+{
+	if (ctl->system) {
+		show_uclamp_system_info();
+	} else if (ctl->all_tasks) {
+		pid_t tid;
+		struct proc_tasks *ts = proc_open_tasks(ctl->pid);
+
+		if (!ts)
+			err(EXIT_FAILURE, _("cannot obtain the list of tasks"));
+
+		while (!proc_next_tid(ts, &tid))
+			show_uclamp_pid_info(tid, NULL);
+
+		proc_close_tasks(ts);
+	} else {
+		show_uclamp_pid_info(ctl->pid, ctl->cmd);
+	}
+}
+
+static int set_uclamp_one(struct uclampset *ctl, pid_t pid)
+{
+	struct sched_attr sa;
+
+	if (sched_getattr(pid, &sa, sizeof(sa), 0) != 0)
+		err(EXIT_FAILURE, _("failed to get pid %d's uclamp values"), pid);
+
+	if (ctl->util_min_set)
+		sa.sched_util_min = ctl->util_min;
+	if (ctl->util_max_set)
+		sa.sched_util_max = ctl->util_max;
+
+	sa.sched_flags = SCHED_FLAG_KEEP_POLICY |
+			 SCHED_FLAG_KEEP_PARAMS |
+			 SCHED_FLAG_UTIL_CLAMP_MIN |
+			 SCHED_FLAG_UTIL_CLAMP_MAX;
+
+	if (ctl->reset_on_fork)
+		sa.sched_flags |= SCHED_FLAG_RESET_ON_FORK;
+
+	return sched_setattr(pid, &sa, 0);
+}
+
+static void set_uclamp_pid(struct uclampset *ctl)
+{
+	if (ctl->all_tasks) {
+		pid_t tid;
+		struct proc_tasks *ts = proc_open_tasks(ctl->pid);
+
+		if (!ts)
+			err(EXIT_FAILURE, _("cannot obtain the list of tasks"));
+
+		while (!proc_next_tid(ts, &tid))
+			if (set_uclamp_one(ctl, tid) == -1)
+				err(EXIT_FAILURE, _("failed to set tid %d's uclamp values"), tid);
+
+		proc_close_tasks(ts);
+
+	} else if (set_uclamp_one(ctl, ctl->pid) == -1) {
+		err(EXIT_FAILURE, _("failed to set pid %d's uclamp values"), ctl->pid);
+	}
+}
+
+static void set_uclamp_system(struct uclampset *ctl)
+{
+	if (!ctl->util_min_set)
+		ctl->util_min = read_uclamp_sysfs(_PATH_PROC_UCLAMP_MIN);
+
+	if (!ctl->util_max_set)
+		ctl->util_max = read_uclamp_sysfs(_PATH_PROC_UCLAMP_MAX);
+
+	if (ctl->util_min > ctl->util_max) {
+		errno = EINVAL;
+		err(EXIT_FAILURE, _("util_min must be <= util_max"));
+	}
+
+	write_uclamp_sysfs(_PATH_PROC_UCLAMP_MIN, ctl->util_min);
+	write_uclamp_sysfs(_PATH_PROC_UCLAMP_MAX, ctl->util_max);
+}
+
+static void validate_util(int val)
+{
+	if (val > 1024 || val < -1) {
+		errno = EINVAL;
+		err(EXIT_FAILURE, _("%d out of range"), val);
+	}
+}
+
+int main(int argc, char **argv)
+{
+	struct uclampset _ctl = {
+		.pid = -1,
+		.util_min = NOT_SET,
+		.util_max = NOT_SET,
+		.cmd = NULL
+	};
+	struct uclampset *ctl = &_ctl;
+	int c;
+
+	static const struct option longopts[] = {
+		{ "all-tasks",		no_argument, NULL, 'a' },
+		{ "pid",		no_argument, NULL, 'p' },
+		{ "system",		no_argument, NULL, 's' },
+		{ "reset-on-fork",	no_argument, NULL, 'R' },
+		{ "help",		no_argument, NULL, 'h' },
+		{ "verbose",		no_argument, NULL, 'v' },
+		{ "version",		no_argument, NULL, 'V' },
+		{ NULL,			no_argument, NULL, 0 }
+	};
+
+	setlocale(LC_ALL, "");
+	bindtextdomain(PACKAGE, LOCALEDIR);
+	textdomain(PACKAGE);
+	close_stdout_atexit();
+
+	while((c = getopt_long(argc, argv, "+asRphmMvV", longopts, NULL)) != -1)
+	{
+		switch (c) {
+		case 'a':
+			ctl->all_tasks = 1;
+			break;
+		case 'p':
+			errno = 0;
+			ctl->pid = strtos32_or_err(argv[optind], _("invalid PID argument"));
+			optind++;
+			break;
+		case 's':
+			ctl->system = 1;
+			break;
+		case 'R':
+			ctl->reset_on_fork = 1;
+			break;
+		case 'v':
+			ctl->verbose = 1;
+			break;
+		case 'm':
+			ctl->util_min = strtos32_or_err(argv[optind], _("invalid util_min argument"));
+			ctl->util_min_set = 1;
+			validate_util(ctl->util_min);
+			optind++;
+			break;
+		case 'M':
+			ctl->util_max = strtos32_or_err(argv[optind], _("invalid util_max argument"));
+			ctl->util_max_set = 1;
+			validate_util(ctl->util_max);
+			optind++;
+			break;
+		case 'V':
+			print_version(EXIT_SUCCESS);
+			/* fallthrough */
+		case 'h':
+			usage();
+		default:
+			errtryhelp(EXIT_FAILURE);
+		}
+	}
+
+	if (argc == 1) {
+		usage();
+		exit(EXIT_FAILURE);
+	}
+
+	/* all_tasks implies --pid */
+	if (ctl->all_tasks && ctl->pid == -1) {
+		errno = EINVAL;
+		err(EXIT_FAILURE, _("missing -p option"));
+	}
+
+	if (!ctl->util_min_set && !ctl->util_max_set) {
+		/* -p or -s must be passed */
+		if (!ctl->system && ctl->pid == -1) {
+			usage();
+			exit(EXIT_FAILURE);
+		}
+
+		show_uclamp_info(ctl);
+		return EXIT_SUCCESS;
+	}
+
+	/* ensure there's a command to execute if no -s or -p */
+	if (!ctl->system && ctl->pid == -1) {
+		if (argc <= optind) {
+			errno = EINVAL;
+			err(EXIT_FAILURE, _("no cmd to execute"));
+		}
+
+		argv += optind;
+		ctl->cmd = argv[0];
+	}
+
+	if (ctl->pid == -1)
+		ctl->pid = 0;
+
+	if (ctl->system)
+		set_uclamp_system(ctl);
+	else
+		set_uclamp_pid(ctl);
+
+	if (ctl->verbose)
+		show_uclamp_info(ctl);
+
+	if (ctl->cmd) {
+		execvp(ctl->cmd, argv);
+		errexec(ctl->cmd);
+	}
+
+	return EXIT_SUCCESS;
+}
-- 
2.25.1


  parent reply	other threads:[~2021-01-30 20:51 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-01-30 20:50 [PATCH v2 0/5] Add a new uclampset utility Qais Yousef
2021-01-30 20:50 ` [PATCH v2 1/5] Move sched_attr struct and syscall definitions into header file Qais Yousef
2021-01-30 20:50 ` Qais Yousef [this message]
2021-02-01 16:12   ` [PATCH v2 2/5] Add uclampset schedutil Karel Zak
2021-02-01 17:32     ` Qais Yousef
2021-02-02  8:01       ` Karel Zak
2021-02-02 11:02         ` Qais Yousef
2021-02-02 15:40           ` Karel Zak
2021-02-02 16:29             ` Qais Yousef
2021-01-30 20:50 ` [PATCH v2 3/5] uclampset: Add man page Qais Yousef
2021-01-30 20:50 ` [PATCH v2 4/5] uclampset: Plump into the build system Qais Yousef
2021-01-30 20:50 ` [PATCH v2 5/5] uclampset: Plumb in bash-completion Qais Yousef

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210130205039.581764-3-qais.yousef@arm.com \
    --to=qais.yousef@arm.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=kzak@redhat.com \
    --cc=mail@bernhard-voelker.de \
    --cc=patrick.bellasi@matbug.net \
    --cc=peterz@infradead.org \
    --cc=util-linux@vger.kernel.org \
    --cc=vincent.donnefort@arm.com \
    --cc=vincent.guittot@linaro.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).