| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <Z6ytgCGkdl07DewQ@x1>
Date: Wed, 12 Feb 2025 15:17:36 +0100
From: Arnaldo Carvalho de Melo <acme@...nel.org>
To: Ian Rogers <irogers@...gle.com>
Cc: Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...nel.org>,
Adrian Hunter <adrian.hunter@...el.com>,
Kan Liang <kan.liang@...ux.intel.com>, Hao Ge <gehao@...inos.cn>,
James Clark <james.clark@...aro.org>,
Howard Chu <howardchu95@...il.com>,
Dominique Martinet <asmadeus@...ewreck.org>,
Levi Yun <yeoreum.yun@....com>, Xu Yang <xu.yang_2@....com>,
Tengda Wu <wutengda@...weicloud.com>,
Yang Jihong <yangjihong1@...wei.com>,
linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1 01/10] perf bench evlist-open-close: Reduce scope of 2
variables
On Sat, Jan 11, 2025 at 11:01:34AM -0800, Ian Rogers wrote:
> Make 2 global variables local. Reduces ELF binary size by removing
> relocations. For a no flags build, the perf binary size is reduced by
> 4,144 bytes on x86-64.
I'm trying to reproduce your results:
$ gcc --version | head -1
gcc (GCC) 14.2.1 20240912 (Red Hat 14.2.1-3)
$
$ rm -rf /tmp/build/$(basename $PWD)/ ; mkdir -p /tmp/build/$(basename $PWD)/
$ make -k O=/tmp/build/$(basename $PWD)/ -C tools/perf install-bin
Without your patch:
$ ls -la ~/bin/perf
-rwxr-xr-x. 2 acme acme 11411680 Feb 12 10:57 /home/acme/bin/perf
$ size ~/bin/perf
text data bss dec hex filename
10071297 302496 34540 10408333 9ed18d /home/acme/bin/perf
$
Then, with your patch:
$ git log --oneline -1
abd904389b3f0807 (HEAD -> perf-tools-next) perf bench evlist-open-close: Reduce scope of 2 variables
$ perf -v
perf version 6.13.rc2.gabd904389b3f
$ size ~/bin/perf
text data bss dec hex filename
10072001 301568 34540 10408109 9ed0ad /home/acme/bin/perf
$
$ ls -la ~/bin/perf
-rwxr-xr-x. 2 acme acme 11411632 Feb 12 11:02 /home/acme/bin/perf
$
So a more modest 224 bytes reduction in the perf binary size.
In the distant past several of these moves from global to local were
made, for instance:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d20deb64e0490ee9442b5181bc08a62d2cadcb90
I tried but didn't find the before/after effects on binary size...
Anyways,
Reviewed-by: Arnaldo Carvalho de Melo <acme@...hat.com>
Tested-by: Arnaldo Carvalho de Melo <acme@...hat.com>
Thanks,
- Arnaldo
> Signed-off-by: Ian Rogers <irogers@...gle.com>
> ---
> tools/perf/bench/evlist-open-close.c | 42 +++++++++++++++-------------
> 1 file changed, 23 insertions(+), 19 deletions(-)
>
> diff --git a/tools/perf/bench/evlist-open-close.c b/tools/perf/bench/evlist-open-close.c
> index 5a27691469ed..79cedcf94a39 100644
> --- a/tools/perf/bench/evlist-open-close.c
> +++ b/tools/perf/bench/evlist-open-close.c
> @@ -46,25 +46,6 @@ static struct record_opts opts = {
> .ctl_fd_ack = -1,
> };
>
> -static const struct option options[] = {
> - OPT_STRING('e', "event", &event_string, "event", "event selector. use 'perf list' to list available events"),
> - OPT_INTEGER('n', "nr-events", &nr_events,
> - "number of dummy events to create (default 1). If used with -e, it clones those events n times (1 = no change)"),
> - OPT_INTEGER('i', "iterations", &iterations, "Number of iterations used to compute average (default=100)"),
> - OPT_BOOLEAN('a', "all-cpus", &opts.target.system_wide, "system-wide collection from all CPUs"),
> - OPT_STRING('C', "cpu", &opts.target.cpu_list, "cpu", "list of cpus where to open events"),
> - OPT_STRING('p', "pid", &opts.target.pid, "pid", "record events on existing process id"),
> - OPT_STRING('t', "tid", &opts.target.tid, "tid", "record events on existing thread id"),
> - OPT_STRING('u', "uid", &opts.target.uid_str, "user", "user to profile"),
> - OPT_BOOLEAN(0, "per-thread", &opts.target.per_thread, "use per-thread mmaps"),
> - OPT_END()
> -};
> -
> -static const char *const bench_usage[] = {
> - "perf bench internals evlist-open-close <options>",
> - NULL
> -};
> -
> static int evlist__count_evsel_fds(struct evlist *evlist)
> {
> struct evsel *evsel;
> @@ -225,6 +206,29 @@ static char *bench__repeat_event_string(const char *evstr, int n)
>
> int bench_evlist_open_close(int argc, const char **argv)
> {
> + const struct option options[] = {
> + OPT_STRING('e', "event", &event_string, "event",
> + "event selector. use 'perf list' to list available events"),
> + OPT_INTEGER('n', "nr-events", &nr_events,
> + "number of dummy events to create (default 1). If used with -e, it clones those events n times (1 = no change)"),
> + OPT_INTEGER('i', "iterations", &iterations,
> + "Number of iterations used to compute average (default=100)"),
> + OPT_BOOLEAN('a', "all-cpus", &opts.target.system_wide,
> + "system-wide collection from all CPUs"),
> + OPT_STRING('C', "cpu", &opts.target.cpu_list, "cpu",
> + "list of cpus where to open events"),
> + OPT_STRING('p', "pid", &opts.target.pid, "pid",
> + "record events on existing process id"),
> + OPT_STRING('t', "tid", &opts.target.tid, "tid",
> + "record events on existing thread id"),
> + OPT_STRING('u', "uid", &opts.target.uid_str, "user", "user to profile"),
> + OPT_BOOLEAN(0, "per-thread", &opts.target.per_thread, "use per-thread mmaps"),
> + OPT_END()
> + };
> + const char *const bench_usage[] = {
> + "perf bench internals evlist-open-close <options>",
> + NULL
> + };
> char *evstr, errbuf[BUFSIZ];
> int err;
>
> --
> 2.47.1.613.gc27f4b7a9f-goog
>
Powered by blists - more mailing lists