From: Hamza Mahfooz <someguy@effective-light.com>
To: linux-kernel@vger.kernel.org
Cc: Thomas Renninger <trenn@suse.com>, Shuah Khan <shuah@kernel.org>,
linux-pm@vger.kernel.org,
Hamza Mahfooz <someguy@effective-light.com>,
Janakarajan Natarajan <Janakarajan.Natarajan@amd.com>
Subject: [PATCH 1/2] cpupower: implement the multi-cpu monitoring of programs
Date: Wed, 12 May 2021 07:26:57 -0400 [thread overview]
Message-ID: <20210512112658.89965-1-someguy@effective-light.com> (raw)
If we look inside cpupower/ToDo, the current 6th point makes mention of
a method to implement multi-cpu monitoring without introducing noise to
the tested program itself.
Suggested-by: Janakarajan Natarajan <Janakarajan.Natarajan@amd.com>
Signed-off-by: Hamza Mahfooz <someguy@effective-light.com>
---
.../utils/idle_monitor/cpupower-monitor.c | 28 +++++++++++++------
1 file changed, 20 insertions(+), 8 deletions(-)
diff --git a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
index 7c77045fef52..5fc9b38be4e5 100644
--- a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
+++ b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
@@ -278,7 +278,7 @@ void list_monitors(void)
}
}
-int fork_it(char **argv)
+int fork_it(int cpu, char **argv)
{
int status;
unsigned int num;
@@ -315,9 +315,9 @@ int fork_it(char **argv)
timediff = timespec_diff_us(start, end);
if (WIFEXITED(status))
- printf(_("%s took %.5f seconds and exited with status %d\n"),
- argv[0], timediff / (1000.0 * 1000),
- WEXITSTATUS(status));
+ printf(_("cpu %d: %s took %.5f seconds and exited with status %d\n"),
+ cpu, argv[0],
+ timediff / (1000.0 * 1000), WEXITSTATUS(status));
return 0;
}
@@ -388,7 +388,8 @@ int cmd_monitor(int argc, char **argv)
{
unsigned int num;
struct cpuidle_monitor *test_mon;
- int cpu;
+ int cpu, status;
+ pid_t child_pid;
cmdline(argc, argv);
cpu_count = get_cpu_topology(&cpu_top);
@@ -440,10 +441,21 @@ int cmd_monitor(int argc, char **argv)
/*
* if any params left, it must be a command to fork
*/
- if (argc - optind)
- fork_it(argv + optind);
- else
+ if (argc - optind) {
+ for (cpu = 0; cpu < cpu_count; cpu++) {
+ child_pid = fork();
+ if (!child_pid) {
+ bind_cpu(cpu);
+ fork_it(cpu, argv + optind);
+ exit(EXIT_SUCCESS);
+ } else if (waitpid(child_pid, &status, 0) == -1) {
+ perror("waitpid");
+ exit(EXIT_FAILURE);
+ }
+ }
+ } else {
do_interval_measure(interval);
+ }
/* ToDo: Topology parsing needs fixing first to do
this more generically */
--
2.31.1
next reply other threads:[~2021-05-12 11:27 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-12 11:26 Hamza Mahfooz [this message]
2021-05-12 11:26 ` [PATCH 2/2] cpupower: removed a completed task from the list Hamza Mahfooz
2021-05-12 14:45 ` Shuah Khan
2021-05-12 20:30 ` Hamza Mahfooz
2021-05-12 14:44 ` [PATCH 1/2] cpupower: implement the multi-cpu monitoring of programs Shuah Khan
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=20210512112658.89965-1-someguy@effective-light.com \
--to=someguy@effective-light.com \
--cc=Janakarajan.Natarajan@amd.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=shuah@kernel.org \
--cc=trenn@suse.com \
/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).