<!DOCTYPE html>
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>Hi Nico,</p>
    <p>I would definitely suspect GC. You've already eliminated
      full/incremental GC. Does Cuis keep statistics on when scavenges
      are done? It's a sub-millisecond difference between the peaks.
      ~700 microseconds seems like maybe a bit long for a scavenge, but
      not unbelievable.</p>
    <p>To eliminate memory allocation or GC effects, you could try
      running a full GC between each repetition of the test and see
      whether there is still a bimodal distribution.</p>
    <p>Regards,</p>
    <p>-Martin </p>
    <div class="moz-cite-prefix">On 5/12/26 12:17 PM, Nicolás Papagna
      Maldonado via Cuis-dev wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CADGn7BOcjrKLSNU+eNrW=+x6waSLx7Qc3Z4wNjkrm2=Gvrg8Mw@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div dir="ltr">
          <div>Hi folks,</div>
          <div><br>
          </div>
          <div>I'm in the process of doing performance benchmarks of my
            Code Coverage tool for my thesis, and I've encountered
            something interesting while measuring the test execution
            times of the Aconcagua package.</div>
          <div><br>
          </div>
          <div><b>Observation:</b><br>
            When running the Aconcagua test suite 150 times in a clean
            image, the execution times are not normally distributed.
            Instead, they fall into two distinct "modes":<br>
            <br>
            <img src="cid:part1.jniQixs9.90R0lbBR@hand2mouse.com"
              alt="image.png" width="535" height="423" class=""><br>
            A Base Mode (~6.28 ms): Representing roughly 72% of the
            iterations.<br>
            A Spike Mode (~7.0 ms): Representing roughly 28% of the
            iterations.<br>
            <br>
            As the next plot shows, the spikes are not random; they
            seem to be kind-of deterministic (roughly every 4th
            iteration).</div>
          <div>I think these perturbations are visible because I'm
            taking very small measurements (~6.28 ms).</div>
          <div><img src="cid:part2.Yfn7bBwb.HJnOFyjo@hand2mouse.com"
              alt="image.png" width="535" height="422" class=""><br>
            <br>
            <b>Question:</b><br>
            I am trying to isolate whether this "heartbeat" is
            originating from within the Cuis image or the VM/OS layer
            (I'm thinking of VM timer/events, maybe UI timers?)</div>
          <div>Does anyone know what could be causing this?</div>
          <div>Are there any known events that run at this particular
            frequency?</div>
          <div><br>
          </div>
          <div>I repeated the experiment a couple of times and the
            perturbations are consistent (in terms of ms and frequency).</div>
          <div><br>
          </div>
          <div><b>Methodology:</b></div>
          <div>
            <ul>
              <li>Measurements were taken in batches of 50 test runs.</li>
              <li>Each batch was run in a fresh image, waiting 2 minutes
                for the VM to stabilize Cuis code.</li>
              <li>10 warmup runs were done and discarded before
                measuring test run times.</li>
              <li>Measurements were taken using Time primHighResClock
                and converted to ms using Time
                highResTimerTicksPerMillisecond.</li>
              <li>VM stats were collected before and after measurements.
                A full GC was run before taking measurements.</li>
              <li>No GC activity was observed while measuring the tests
                run (VM stats report no full nor incremental GCs).</li>
              <li>The test suite was created and run from a workspace</li>
            </ul>
          </div>
          <div><b>Cuis environment:</b><br>
            <ul>
              <li>Cuis Smalltalk commit: <a
href="https://github.com/Cuis-Smalltalk/Cuis-Smalltalk-Dev/commit/b221722348cc46ebb9f5d9f3a057a3f361aa3d55"
                  moz-do-not-send="true" class="moz-txt-link-freetext">https://github.com/Cuis-Smalltalk/Cuis-Smalltalk-Dev/commit/b221722348cc46ebb9f5d9f3a057a3f361aa3d55</a></li>
              <li>Code Coverage commit: <a
href="https://github.com/npapagna/cuis-code-coverage/commit/db19de02250a3fd50dfc4e0b4d0893b656e31e9b"
                  moz-do-not-send="true" class="moz-txt-link-freetext">https://github.com/npapagna/cuis-code-coverage/commit/db19de02250a3fd50dfc4e0b4d0893b656e31e9b</a></li>
              <li>Aconcagua commit: <a
href="https://github.com/Cuis-Smalltalk/Measures/commit/05aa02aada64f7b64898a5e1452ae8fb76f79427"
                  moz-do-not-send="true" class="moz-txt-link-freetext">https://github.com/Cuis-Smalltalk/Measures/commit/05aa02aada64f7b64898a5e1452ae8fb76f79427</a></li>
            </ul>
            <b>Machine environment:</b><br>
            <ul>
              <li>Model Name: MacBook Pro</li>
              <li>Model Identifier: MacBookPro18,3</li>
              <li>Model Number: MKGR3LL/A</li>
              <li>Chip: Apple M1 Pro</li>
              <li>Total Number of Cores: 8 (6 Performance and 2
                Efficiency)</li>
              <li>Memory: 16 GB</li>
              <li>System Firmware Version: 13822.81.10</li>
              <li>OS Loader Version: 13822.81.102</li>
              <li>APPLE SSD AP0512R (NVMe, avg read 5671 MB/s, avg
                write 5670 MB/s)</li>
              <li>macOs 26.3 (25D125)</li>
              <li>A dedicated fresh profile was created for the
                benchmarks (indexing, time machine, screen lock/saver,
                display shutdown, bluetooth and wifi disabled).</li>
            </ul>
          </div>
        </div>
        <div dir="ltr">
          <div>Thanks in advance,</div>
          <div>Nico PM</div>
          <div><br>
          </div>
        </div>
      </div>
      <br>
      <fieldset class="moz-mime-attachment-header"></fieldset>
    </blockquote>
  </body>
</html>