From: Steven Rostedt <rostedt@goodmis.org>
To: linux-kernel@vger.kernel.org
Cc: Ingo Molnar <mingo@kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
Kalesh Singh <kaleshsingh@google.com>
Subject: [for-next][PATCH 12/15] tracing/histogram: Optimize division by a power of 2
Date: Wed, 27 Oct 2021 12:09:52 -0400 [thread overview]
Message-ID: <20211027161001.968134870@goodmis.org> (raw)
In-Reply-To: 20211027160940.084904334@goodmis.org
From: Kalesh Singh <kaleshsingh@google.com>
The division is a slow operation. If the divisor is a power of 2, use a
shift instead.
Results were obtained using Android's version of perf (simpleperf[1]) as
described below:
1. hist_field_div() is modified to call 2 test functions:
test_hist_field_div_[not]_optimized(); passing them the
same args. Use noinline and volatile to ensure these are
not optimized out by the compiler.
2. Create a hist event trigger that uses division:
events/kmem/rss_stat$ echo 'hist:keys=common_pid:x=size/<divisor>'
>> trigger
events/kmem/rss_stat$ echo 'hist:keys=common_pid:vals=$x'
>> trigger
3. Run Android's lmkd_test[2] to generate rss_stat events, and
record CPU samples with Android's simpleperf:
simpleperf record -a --exclude-perf --post-unwind=yes -m 16384 -g
-f 2000 -o perf.data
== Results ==
Divisor is a power of 2 (divisor == 32):
test_hist_field_div_not_optimized | 8,717,091 cpu-cycles
test_hist_field_div_optimized | 1,643,137 cpu-cycles
If the divisor is a power of 2, the optimized version is ~5.3x faster.
Divisor is not a power of 2 (divisor == 33):
test_hist_field_div_not_optimized | 4,444,324 cpu-cycles
test_hist_field_div_optimized | 5,497,958 cpu-cycles
If the divisor is not a power of 2, as expected, the optimized version is
slightly slower (~24% slower).
[1] https://android.googlesource.com/platform/system/extras/+/master/simpleperf/doc/README.md
[2] https://cs.android.com/android/platform/superproject/+/master:system/memory/lmkd/tests/lmkd_test.cpp
Link: https://lkml.kernel.org/r/20211025200852.3002369-7-kaleshsingh@google.com
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Suggested-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
kernel/trace/trace_events_hist.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
index 71b453576d85..452daad7cfb3 100644
--- a/kernel/trace/trace_events_hist.c
+++ b/kernel/trace/trace_events_hist.c
@@ -304,6 +304,10 @@ static u64 hist_field_div(struct hist_field *hist_field,
if (!val2)
return -1;
+ /* Use shift if the divisor is a power of 2 */
+ if (!(val2 & (val2 - 1)))
+ return val1 >> __ffs64(val2);
+
return div64_u64(val1, val2);
}
--
2.33.0
next prev parent reply other threads:[~2021-10-27 16:10 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-27 16:09 [for-next][PATCH 00/15] tracing: More updates for 5.16 Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 01/15] lib/bootconfig: Fix the xbc_get_info kerneldoc Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 02/15] samples/kretprobes: Fix return value if register_kretprobe() failed Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 03/15] docs, kprobes: Remove invalid URL and add new reference Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 04/15] test_kprobes: Move it from kernel/ to lib/ Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 05/15] MAINTAINERS: Update KPROBES and TRACING entries Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 06/15] selftests/ftrace: Stop tracing while reading the trace file by default Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 07/15] tracing: Add support for creating hist trigger variables from literal Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 08/15] tracing: Add division and multiplication support for hist triggers Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 09/15] tracing: Fix operator precedence for hist triggers expression Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 10/15] tracing/histogram: Simplify handling of .sym-offset in expressions Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 11/15] tracing/histogram: Covert expr to const if both operands are constants Steven Rostedt
2021-10-27 16:09 ` Steven Rostedt [this message]
2021-10-27 16:09 ` [for-next][PATCH 13/15] tracing/histogram: Document expression arithmetic and constants Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 14/15] ftrace: disable preemption when recursion locked Steven Rostedt
2021-10-27 16:17 ` Steven Rostedt
2021-10-27 16:09 ` [for-next][PATCH 15/15] ftrace: do CPU checking after preemption disabled Steven Rostedt
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=20211027161001.968134870@goodmis.org \
--to=rostedt@goodmis.org \
--cc=akpm@linux-foundation.org \
--cc=kaleshsingh@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.