<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Hi Thomas,</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">>You are talking about rewriting the NMT backend tracking virtual memory regions lock-free,
right</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">Only VirtualMemorySummary, not VirtualMemoryTracker!</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">>Are you sure this is even worth optimizing? I doubt a typical jvm run involves more
than a few dozen/hundred mmap calls. That is not exactly hot. </span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">> Then, are you sure that swapping a lock synchronization with a bunch of atomic variable
updates brings that much performance gain? Atomic updates are still expensive. <br>
>Do we even have much contention? I would have thought that concurrently happening mmap calls are very rare.
<br>
>Reducing contention can be done a lot simpler, first by swapping ThreadCritical for a targeted mutex as we always wanted.
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">Unfortunately, Afshin did not list his branch in the e-mail, here's a link :</span><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><a href="https://github.com/afshin-zafari/jdk/commits/_nmt_light" id="LPlnk499540">https://github.com/afshin-zafari/jdk/commits/_nmt_light</a></span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Let's make sure that we're not talking past each other: What Afshin is proposing is replacing the summary mode tracking. We do have some preliminary performance results showing that these are effective (see second commit in branch).</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Basic idea is this:</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
The summary mode today still requires the VirtualMemoryTracker to be active, but this is only necessary because of MEMFLAGS not being passed when releasing memory, only when reserving/committing it. If you add MEMFLAGs as a parameter to release_memory etc,
then summary mode is independent of the VMT and the rest of Afshin's patch are natural progressions from the consequences of that.</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">>If that is not good enough, change the underlying data structure. A linked list of
regions is not that effective. There are data structures that are better suited for this, e.g. interval trees.
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">>But a very simple and possibly very effective change could be to change the structure
from a list to an array. That's a lot more cache friendly. If you optimise it right (e.g. similar to the "CachedNMTInformation" cache here:
<a href="https://github.com/tstuefe/jdk/blob/24a9916aedecbde76f5c6d79c001133b750f431d/src/hotspot/share/nmt/memMapPrinter.cpp#L84" id="OWA2086da9f-a512-1197-eada-6e76151c2875" class="OWAAutoLink">
https://github.com/tstuefe/jdk/blob/24a9916aedecbde76f5c6d79c001133b750f431d/src/hotspot/share/nmt/memMapPrinter.cpp#L84</a> ) this can be orders of magnitude faster.
</span></div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">Yes, I have a couple of branches for this work. You can find the most current one in
</span><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><a href="https://github.com/jdksjolen/jdk/tree/nmt-memory-view" id="LPlnk217573">https://github.com/jdksjolen/jdk/tree/nmt-memory-view</a></span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">I really like Afshin's patch because it will also help this work. Especially, if release_memory
etc pass along the MEMFLAGS then we can use this information to skip scanning a lot of regions.</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">So, I see this patch as differentiating further between summary and detailed mode and
also helping out writing a faster virtual memory tracker.<br>
<br>
>Are you sure this is even worth optimizing? I doubt a typical jvm run involves more than a few dozen/hundred mmap calls. That is not exactly hot.
<br>
Are you sure? A ZGC enabled JVM can have a lot of NMT calls, and Metaspace does a lot of small and contiguous allocations. We do see a performance degradation with just summary mode.</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">I hope that this brings some clarity to the issue.</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"><br>
</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">All the best,</span></div>
<div class="elementToProof"><span style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">Johan<br>
</span></div>
<div id="appendonsend"></div>
<hr style="display:inline-block;width:98%" tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" style="font-size:11pt" color="#000000"><b>From:</b> hotspot-runtime-dev <hotspot-runtime-dev-retn@openjdk.org> on behalf of Thomas Stuefe <tstuefe@redhat.com><br>
<b>Sent:</b> Wednesday, November 29, 2023 19:20<br>
<b>To:</b> Afshin Zafari <afshin.zafari@oracle.com><br>
<b>Cc:</b> hotspot-runtime-dev@openjdk.org <hotspot-runtime-dev@openjdk.org><br>
<b>Subject:</b> Re: New NMT Lightweight mode</font>
<div> </div>
</div>
<div>
<div dir="ltr">Hi Ashin,
<div><br>
</div>
<div>I did not see your mail, therefore I replied in JBS. </div>
<div><br>
</div>
<div>Cheers, Thomas</div>
</div>
<br>
<div class="x_gmail_quote">
<div dir="ltr" class="x_gmail_attr">On Wed, Nov 29, 2023 at 6:09 PM Afshin Zafari <<a href="mailto:afshin.zafari@oracle.com">afshin.zafari@oracle.com</a>> wrote:<br>
</div>
<blockquote class="x_gmail_quote" style="margin:0px 0px 0px 0.8ex; border-left-width:1px; border-left-style:solid; border-left-color:rgb(204,204,204); padding-left:1ex">
<div class="x_msg-1601002534520522468">
<div dir="ltr">
<div style="margin:0px; min-height:68px">
<div style="margin:0px 0px 0px 48px; height:24px">
<div style="text-align:left"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)">Dear all,</span></div>
<div style="text-align:left"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)">The JBS RFE
<a href="https://bugs.openjdk.org/browse/JDK-8320977" id="x_m_-1601002534520522468OWAf362d0bb-f39f-6df4-a5d7-3052ebccdac9" rel="noopener noreferrer" target="_blank" style="margin:0px">
https://bugs.openjdk.org/browse/JDK-8320977</a> is created for implementing a new Lightweight mode for NativeMemoryTracking (NMT) of JDK.</span></div>
</div>
</div>
<div style="margin:12px 16px 0px 52px; padding:0px 0px 2px">
<div style="direction:ltr; text-align:left; margin:0px"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)">Since the changes for this
RFE impact a large number of files and code in JDK, it would be great if I get comments while prototyping it.</span></div>
<div style="direction:ltr; text-align:left; margin:0px"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)"><br>
</span></div>
<div style="direction:ltr; text-align:left; margin:0px"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)">Best,</span></div>
<div style="direction:ltr; text-align:left; margin:0px"><span style="letter-spacing:normal; font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; font-weight:400; color:rgb(0,0,0)">Afshin</span></div>
</div>
<div style="font-family:Aptos,Aptos_EmbeddedFont,Aptos_MSFontService,Calibri,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0)">
<br>
</div>
</div>
</div>
</blockquote>
</div>
</div>
</body>
</html>