<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
I have a new fix for this bug. My previous fix broke solaris-x86 (I
<br>
had not defined an early_initialize() for x86). This fix is
slightly<br>
smaller and has the virtue of moving the required initialization<br>
closer to where it is used.<br>
<br>
<a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~jmasa/8133023/webrev.03/index.html">http://cr.openjdk.java.net/~jmasa/8133023/webrev.03/index.html</a><br>
<br>
Testing: JPRT build on all platforms, checked by hand that the
correct number<br>
of GC worker threads are created on later Niagara platforms.<br>
<br>
Thanks.<br>
<br>
Jon<br>
<br>
<div class="moz-cite-prefix">On 11/12/2015 1:31 PM, Jon Masamitsu
wrote:<br>
</div>
<blockquote cite="mid:56450520.6060706@oracle.com" type="cite">
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<font face="Times New Roman, Times, serif">GC calls
VM_Version::calc_parallel_worker_threads() to determine<br>
the number of GC threads to create. On Sparc it checks for
newer<br>
Niagara hardware to decide the proper scaling of the GC threads
with<br>
the hardware threads. </font><font face="Times New Roman,
Times, serif"><font face="Times New Roman, Times, serif">calc_parallel_worker_threads()
was being called<br>
before enough information was gathered to determine the Sparc
hardware.<br>
<br>
Moved the gathering of information needed to earlier in the
JVM<br>
initialization.<br>
<br>
</font><a moz-do-not-send="true" class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Ejmasa/8133023/webrev.00/">http://cr.openjdk.java.net/~jmasa/8133023/webrev.00/</a><br>
<br>
<a moz-do-not-send="true" class="moz-txt-link-freetext"
href="https://bugs.openjdk.java.net/browse/JDK-8133023">https://bugs.openjdk.java.net/browse/JDK-8133023</a><br>
<br>
Thanks.<br>
<br>
Jon<br>
</font> </blockquote>
<br>
</body>
</html>