diff options
| author | Stuart Monteith <stuart.monteith@linaro.org> | 2018-01-10 14:43:19 +0000 |
|---|---|---|
| committer | Stuart Monteith <stuart.monteith@linaro.org> | 2018-01-15 10:42:59 +0000 |
| commit | 9d964531d5614bf23e93b85756a7cf2c35fcd219 (patch) | |
| tree | 6cfb744449ef874874ff249ba5077b3a3a174d02 | |
| parent | 6d8e195520595e261fd07d18be946fcbbe6670fa (diff) | |
Change jdk9 builds to jdk9u
The existing jdk9 files will still be available.
Change-Id: Iefd39449a5660d88c770ce00c29df07852d9c8b5
| -rw-r--r-- | panel.htm | 1 | ||||
| -rw-r--r-- | test9u.htm | 35 |
2 files changed, 36 insertions, 0 deletions
@@ -35,6 +35,7 @@ a:active { <a href="source.htm" target=contents><div class="panel-button">Source Code</div></a> <a href="test8u.htm" target=contents><div class="panel-button">JDK8u Test results</div></a> <a href="test9.htm" target=contents><div class="panel-button">JDK9 Test results</div></a> +<a href="test9u.htm" target=contents><div class="panel-button">JDK9u Test results</div></a> <a href="test10.htm" target=contents><div class="panel-button">JDK10 Test results</div></a> <a href="testX.htm" target=contents><div class="panel-button">JDK Test results</div></a> <a href="install.htm" target=contents><div class="panel-button">Install OpenJDK</div></a> diff --git a/test9u.htm b/test9u.htm new file mode 100644 index 0000000..6439683 --- /dev/null +++ b/test9u.htm @@ -0,0 +1,35 @@ +<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> +<HTML> +<HEAD> +<META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=iso-8859-1"> +<title>Engineering Releases for AARCH64</title> +</HEAD> +<BODY bgcolor=#ffffff> +<h2>Nightly Test Results</h2> +The following are the results of the nightly testing we perform on OpenJDK9u. The primary purpose of our overnight testing is to test for regressions in any code that is checked into the JDK9u code base. Whenever a change is checked into the AARCH64 OpenJDK9u source the nightly test rebuild both client and server versions of OpenJDK9u and runs the tests below. The complete build and test cycle takes approximately 20 hours so results are usually published at about 20:00 UTC the following day. +<ul> +<li><a href="jdk9u/openjdk-jtreg-nightly-tests">JTReg Test Results</a> +<li><a href="jdk9u/jcstress-nightly-runs">JCStress Test Results</a> +<li><a href="jdk9u/SPECjbb2015-results">SPECjbb2015 Results</a> +<li><a href="jdk9u/hadoop-terasort-benchmark-results">Hadoop Results<a> +</ul> +<p>The following is a brief description of each of the above tests +<dl> +<dt>JTreg +<dd>JTReg is the standard Java Test Regression test suite and contains three components, hotspot, langtools and jdk which test the corresponding components in the JDK. In total there are over 8500 individual unit tests in the JTReg test suite. We test both the client and server compilers against the JTReg test suite. +</dl> +<dl> +<dt>JCStress +<dd>JCStress is a test suite designed to stress test the atomic and synchronisation primitives of the JDK. The test suite contains over 11500 individual tests and runs for over 12 hours as part of the overnight testing. This test suite has been extremely useful and has shown up a number of rare and difficult to find problems. +</dl> +<dl> +<dt>SPECjbb2015 +<dd>SPECjbb2015 is a commercial benchmark from the SPEC Corporation. It models a transaction management system and measures the response times under increasing workloads. It is widely used so we chose this as one of our standard benchmarks to test for performance regression. +</dl> +<dl> +<dt>Hadoop +<dd>We choose Hadoop as an industry standard enterprise application to test for any performance regressions. Terasort is an industry standard benchmark for Hadoop. +</dl> +<p>Note: For the SPECjbb2015 and Hadoop benchmarks we publish relative performance figures. The baseline against which they were originally run is no longer meaningful, so changes to the performance figure are all that matter. +</body> +</html> |