<div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">
One more piece of information about "Morph compileAll" in my image:<div><br></div><div><div><span style="font-size: 13.3333px">CommitHash: 118a02c69 Plugins: 202107232247 (Jul 23)</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">44.4% {572ms} Morph class(Behavior)>>sourceCodeAt:</span></div><div><span style="font-size: 13.3333px">10.6% {136ms} Compiler>>compile:in:notifying:ifFail:</span></div><div><span style="font-size: 13.3333px">9.1% {117ms} Morph class(Behavior)>>basicAddSelector:withMethod:</span></div><div><span style="font-size: 13.3333px">8.2% {106ms} MethodNode>>generate:</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">CommitHash: 534349675 Plugins: 202108200036 (Aug 20)</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">93.0% {48312ms} Morph class(Behavior)>>basicAddSelector:withMethod:</span></div><div><span style="font-size: 13.3333px"> 93.0% {48310ms} CompiledMethod>>methodClass:</span></div></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">Note that #methodClass: ends up in:</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">CompiledCode >> #literalAt:put:</span></div><div><span style="font-size: 13.3333px">CompiledCode >> #objectAt:put: (primitive 69)</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">Maybe the slow down in primitive 69 in this particular situation can give somebody a clue? Hmmm...</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">***</span></div><div><br></div><div>I captured the VM parameters but could not find anything suspicious:</div><div><br></div><div><div><span style="font-size: 13.3333px">report := [</span></div><div><span style="font-size: 13.3333px"><span style="white-space:pre"> </span>FileStream forceNewFileNamed: (Time highResClock // Time highResClockTicksPerMillisecond) asString, '.txt'</span></div><div><span style="font-size: 13.3333px"><span style="white-space:pre"> </span>do: [:stream | SystemReporter new reportVMParameters: stream]].</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">report value.</span></div><div><span style="font-size: 13.3333px">Morph compileAll.</span></div><div><span style="font-size: 13.3333px">report value.</span></div></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">See the attached logs to do a diff.</span></div><div><span style="font-size: 13.3333px"><br></span></div><div><span style="font-size: 13.3333px">Best,</span></div><div><span style="font-size: 13.3333px">Marcel</span></div><div class="mb_sig"></div>
<blockquote class="history_container" type="cite" style="border-left-style: solid;border-width: 1px;margin-top: 20px;margin-left: 0px;padding-left: 10px;min-width: 500px">
<p style="color: #AAAAAA; margin-top: 10px;">Am 20.08.2021 12:21:53 schrieb Marcel Taeumel <marcel.taeumel@hpi.de>:</p><div style="font-family:Arial,Helvetica,sans-serif"><div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">
Hi all --<div><br></div><div>I found the culprit. The regression started with the introduction of <span style="white-space: pre-wrap;font-size: 10pt"><b>FastCPrimitiveFlag</b> in VMMaker.oscog-eem.2995. That new feature needs more adjustments, I suppose. :-)</span></div><div><br></div><div>My previous numbers were kind of wrong. I compared cog.spur against stack.spur. The latter being, in comparison, blazingly fast in #compileAll. Something is wrong with current cog.spur, at least on Windows (64-bit).</div><div><br></div><div>CommitHash: d8576b0 Plugins: 2021<b>0820</b>0011 <br></div><div><br></div><div>stack.pur</div><div><div>Morph methodDict size. " 1493 "</div><div>[Morph compileAll] timeToRun " 1293 "</div></div><div><br></div><div>cog.spur</div><div><div>Morph methodDict size. " 1493 "</div><div>[Morph compileAll] timeToRun " 23018 "</div><div><br></div><div>Here are more tests with cog.spur:</div><div><br></div><div>CommitHash: 128d14994 Plugins: 2021<b>0505</b>0056</div><div>cog.spur -- [Morph compileAll] timeToRun " 1294 "</div><div><div><br></div><div>CommitHash: 9d4e82529 Plugins: 2021<b>0712</b>1800<br></div><div>cog.spur -- [Morph compileAll] timeToRun " 910 "</div><div><br></div><div>CommitHash: c0d1a5574 Plugins: 2021<b>0721</b>1710 -- <span style="font-size: 10pt">VMMaker.oscog-eem.2991</span><br></div><div>cog.spur -- [Morph compileAll] timeToRun " 901 "<br></div><div><br></div><div>CommitHash: 118a02c69 Plugins: 2021<b>0723</b>2247<br></div><div>cog.spur -- [Morph compileAll] timeToRun " 984 "</div><div><br></div><div><i>(BUILD BROKEN -- VMMaker.oscog-eem.2995)</i></div><div><i>(VM CRASHES between VMMaker.oscog-eem.2996 and VMMaker.oscog-eem.3009)</i><br></div><div><br></div><div>CommitHash: 03c4ef31b Plugins: 2021<b>0801</b>0511 -- VMMaker.oscog-eem.3009<br></div><div>cog.spur -- [Morph compileAll] timeToRun " 20787 "<br></div><div><br></div><div>***</div><div><br></div><div>Note that #compileAll does not write into the .changes file.<br></div><div><br></div><div>Note that the last "good" commit hash is 118a02c69f3e4c8c29e8ba867a78bd39f9a53cb4 (still using VMMaker.oscog-eem.2991). I added the tag "last-before-FastCPrimitiveFlag" to this commit.</div><div><br></div><div>Now let's review and fix this. ^__^</div><div><br></div><div>Best,</div><div>Marcel</div><div><br></div></div></div><div class="mb_sig"></div>
<blockquote class="history_container" type="cite" style="border-left-style: solid;border-width: 1px;margin-top: 20px;margin-left: 0px;padding-left: 10px;min-width: 500px">
<p style="color: #AAAAAA; margin-top: 10px;">Am 19.08.2021 15:55:34 schrieb Marcel Taeumel <marcel.taeumel@hpi.de>:</p><div style="font-family:Arial,Helvetica,sans-serif"><div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">
<div>[Morph compileAll] timeToRunWithoutGC</div><div><br></div><div>VM <span style="font-size: 10pt">202108160631 : </span><span style="font-size: 10pt">24607 milliseconds</span></div><div><span style="font-size: 10pt">VM </span><span style="font-size: 10pt">202108010511 : 1078 milliseconds</span></div><div><span style="font-size: 10pt"><br></span></div><div><span style="font-size: 10pt">Best,</span></div><div><span style="font-size: 10pt">Marcel</span></div><div class="mb_sig"></div><blockquote class="history_container" type="cite" style="border-left-style: solid;border-width: 1px;margin-top: 20px;margin-left: 0px;padding-left: 10px;min-width: 500px">
<p style="color: #AAAAAA; margin-top: 10px;">Am 19.08.2021 15:40:03 schrieb Marcel Taeumel <marcel.taeumel@hpi.de>:</p><div style="font-family:Arial,Helvetica,sans-serif"><div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">Hi all!<div class="mb_sig"></div><div><br></div><div>There is a serious performance regression in recent VM (<span style="font-size: 10pt">202108160631) </span><span style="font-size: 10pt">regarding code compilation. I could only narrow it down to CompiledMethod >> #methodClass:. Still investigating. The start-up is also very slow. Maybe GC-related?</span></div><div><span style="font-size: 10pt"><br></span></div><div>This regression was not there in 202108010511.</div><div><br></div><div>Best,</div><div>Marcel</div></div></div></blockquote>
</div></div></blockquote></div></div></blockquote></div>