<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=Windows-1252">
</head>
<body>
<div dir="ltr">
<div></div>
<div>
<div>
<div dir="ltr">
<div dir="ltr" style="">Thanks Ron. </div>
<div dir="ltr" style=""><br>
</div>
<div dir="ltr" style="">What about native code that doesn’t make an upcall but does a blocking operation directly (via OS syscalls or similar)? <span style="font-size: inherit;">Are such cases also rare? </span></div>
<div dir="ltr" style=""><span style="font-size: inherit;"><br>
</span></div>
<div dir="ltr" style=""><span style="font-size: inherit;">What would be the advice if someone is making blocking calls from their downcalls? It might not always be possible to rewrite it in Java. </span></div>
<div dir="ltr" style=""><span style="font-size: inherit;"><br>
</span></div>
<div dir="ltr" style=""><span style="font-size: inherit;">If I am reading this correctly [1], it seems like the golang implementation is able to unmount goroutines. </span></div>
<div dir="ltr" style=""><span style="font-size: inherit;"><br>
</span></div>
<div dir="ltr" style="">[1] <a rel="noreferrer noopener" href="https://stackoverflow.com/a/28354879">https://stackoverflow.com/a/28354879</a> </div>
</div>
</div>
<div id="ms-outlook-mobile-signature" dir="ltr"></div>
</div>
</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> Ron Pressler <ron.pressler@oracle.com><br>
<b>Sent:</b> Monday, June 26, 2023 5:47:32 PM<br>
<b>To:</b> Danish Nawab <dnawab@outlook.com><br>
<b>Cc:</b> loom-dev@openjdk.org <loom-dev@openjdk.org><br>
<b>Subject:</b> Re: Native interop with Virtual Threads</font>
<div> </div>
</div>
<div class="BodyFragment"><font size="2"><span style="font-size:11pt;">
<div class="PlainText">Virtual thread stacks move around in memory, as they’re stored in the Java heap. Frames of native code are hard to move around because they may contain pointers into the stack itself, which Java doesn’t know about. Because it isn’t much
of a problem in practice — “upcalls" from native code to Java are rare in general, and blocking operations in upcalls are rarer still — it isn’t a challenge worth solving at the moment.<br>
<br>
— Ron<br>
<br>
> On 26 Jun 2023, at 10:22, Danish Nawab <dnawab@outlook.com> wrote:<br>
> <br>
> Both the JEP [1] and the original Loom proposal [2] state that running native code will pin the virtual thread to its carrier.<br>
> <br>
> I am curious to learn more about the reasons and implementation challenges that cause this limitation. Does it have something to do with how the native code is managed by the JVM? Could you please point me in the right direction?
<br>
> <br>
> [1] <a href="https://openjdk.org/jeps/444">https://openjdk.org/jeps/444</a><br>
> [2] <a href="https://cr.openjdk.org/~rpressler/loom/Loom-Proposal.html">https://cr.openjdk.org/~rpressler/loom/Loom-Proposal.html</a><br>
<br>
<br>
</div>
</span></font></div>
</body>
</html>