dumping ground for random patches and texts
 help / color / mirror / Atom feed
From: Eric Wong <e@80x24.org>
To: spew@80x24.org
Subject: [PATCH] gc-steps: probes for tracking each GC step
Date: Sat,  2 Jun 2018 13:58:20 +0000	[thread overview]
Message-ID: <20180602135820.6686-1-e@80x24.org> (raw)

This should be useful for tracking latency due to GC
---
 tool/gc-steps.stp | 55 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)
 create mode 100644 tool/gc-steps.stp

diff --git a/tool/gc-steps.stp b/tool/gc-steps.stp
new file mode 100644
index 0000000000..db61cd78ed
--- /dev/null
+++ b/tool/gc-steps.stp
@@ -0,0 +1,55 @@
+/*
+ * Logs each GC mark and sweep entry and exit with timestamps
+ *
+ * You will need to change "ruby" to point to an absolute path
+ * of the ruby binary.  Then run "stap /path/to/gc-steps.stp" and run ruby
+ */
+global marks;
+global sweeps;
+global mark_starts;
+global sweep_starts;
+
+function dump_stats() {
+  if (@count(marks)) {
+    printf("  marks % 8u % 8u % 9u % 9u % 9u\n",
+            @min(marks), @max(marks), @avg(marks),
+            @variance(marks), @count(marks));
+    println(@hist_log(marks));
+  }
+  if (@count(sweeps)) {
+    printf(" sweeps % 8u % 8u % 9u % 9u % 9u\n",
+            @min(sweeps), @max(sweeps), @avg(sweeps),
+            @variance(sweeps), @count(sweeps));
+    println(@hist_log(sweeps))
+  }
+}
+
+probe timer.s(5) {
+  dump_stats();
+}
+
+probe process("ruby").mark("gc__sweep__continue__enter")
+{
+  sweep_starts[pid()] = gettimeofday_us();
+}
+
+probe process("ruby").mark("gc__sweep__continue__exit")
+{
+  diff = gettimeofday_us() - sweep_starts[pid()];
+  sweeps <<< diff;
+}
+
+probe process("ruby").mark("gc__marks__continue__enter")
+{
+  mark_starts[pid()] = gettimeofday_us();
+}
+
+probe process("ruby").mark("gc__marks__continue__exit")
+{
+  diff = gettimeofday_us() - mark_starts[pid()];
+  marks <<< diff;
+}
+
+probe end {
+  dump_stats();
+}
-- 
EW


                 reply	other threads:[~2018-06-02 13:58 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20180602135820.6686-1-e@80x24.org \
    --to=e@80x24.org \
    --cc=spew@80x24.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 read-only IMAP folder(s) and NNTP newsgroup(s).