RFR(XS): [TEST] add test to verify EnableTrace works

David Holmes david.holmes at oracle.com
Wed Sep 12 01:31:24 UTC 2018


FYI I also hacked JPRT to run this test, so I can confirm it passes as 
expected on Oracle JDK on Solaris, Windows and Linux.

Cheers,
David

On 11/09/2018 11:50 AM, David Holmes wrote:
> Looks fine now.
> 
> FWIW didn't need a new webrev just to remove othervm.
> 
> Thanks,
> David
> 
> On 11/09/2018 4:57 AM, Hohensee, Paul wrote:
>> Copied, http://cr.openjdk.java.net/~phh/8209863/webrev.03/
>>
>> On 9/10/18, 10:51 AM, "hotspot-runtime-dev on behalf of Hohensee, 
>> Paul" <hotspot-runtime-dev-bounces at openjdk.java.net on behalf of 
>> hohensee at amazon.com> wrote:
>>
>>      Because I copied it to the wrong directory! :(
>>      Paul
>>      On 9/10/18, 9:55 AM, "hotspot-runtime-dev on behalf of Daniel D. 
>> Daugherty" <hotspot-runtime-dev-bounces at openjdk.java.net on behalf of 
>> daniel.daugherty at oracle.com> wrote:
>>          Why is this webrev using '8206075' when previous webrev used 
>> '8209863'?
>>          Dan
>>          On 9/10/18 12:47 PM, Liu, Xin wrote:
>>          > You are right. I updated it. othervm doesn't help much. I 
>> remove it.
>>          >   Here is CR.
>>          > http://cr.openjdk.java.net/~phh/8206075/webrev.03/
>>          >
>>          > thanks,
>>          > --lx
>>          >
>>          >
>>          > On 9/7/18, 4:08 PM, "David Holmes" 
>> <david.holmes at oracle.com> wrote:
>>          >
>>          >      On 8/09/2018 2:40 AM, Liu, Xin wrote:
>>          >      > Hi, David,
>>          >      > Thanks you for the commend.
>>          >      >
>>          >      > I added 'othervm' in last minute. The test grabs 
>> Classload or Compilation events when JVM initializes. Without 
>> `othervm`, the test might not capture them because the JVM has already 
>> did that.
>>          >      > 'othervm' can make the test more consistent.
>>          >
>>          >      No it won't because you are capturing events from JVMs 
>> that you
>>          >      explicitly launch with ProcessTools. The VM launched 
>> by "othervm" does
>>          >      not have EnableTracing applied to it and nobody looks 
>> at its output anyway.
>>          >
>>          >      David
>>          >
>>          >      > Thanks you for reviewing it.
>>          >      >
>>          >      > Thanks,
>>          >      > --lx
>>          >      >
>>          >      >
>>          >      > On 9/6/18, 5:17 PM, "David Holmes" 
>> <david.holmes at oracle.com> wrote:
>>          >      >
>>          >      >      Hi,
>>          >      >
>>          >      >      On 7/09/2018 6:51 AM, Liu, Xin wrote:
>>          >      >      > Hi, David,
>>          >      >      > Could you review this new revision?
>>          >      >      > Webrev: 
>> http://cr.openjdk.java.net/~phh/8209863/webrev.02/
>>          >      >      > bug: 
>> https://bugs.openjdk.java.net/browse/JDK-8209863
>>          >      >
>>          >      >      This looks fine to me.
>>          >      >
>>          >      >      One nit:
>>          >      >
>>          >      >      29  * @run main/othervm
>>          >      >
>>          >      >      You don't need othervm for this test as it just 
>> launches seperate VMs
>>          >      >      anyway. No need for updated webrev.
>>          >      >
>>          >      >      Thanks,
>>          >      >      David
>>          >      >
>>          >      >      > I should verify new tests using both 
>> Openjdk/Oraclejdk. Thank you for heading up.
>>          >      >      >
>>          >      >      > Thanks,
>>          >      >      > --lx
>>          >      >      >
>>          >      >      > On 9/4/18, 11:05 PM, "Liu Xin" 
>> <navy.xliu at gmail.com> wrote:
>>          >      >      >
>>          >      >      >      Hi, David,
>>          >      >      >
>>          >      >      >      Thanks to investigate the closed 
>> implementation.
>>          >      >      >
>>          >      >      >      Indeed, OracleJDK shows “JAVA(TM)” 
>> instead of OpenJDK.
>>          >      >      >
>>          >      >      >      $java -XX:+EnableTracing -version
>>          >      >      >      java version "1.8.0_171"
>>          >      >      >      Java(TM) SE Runtime Environment (build 
>> 1.8.0_171-b11)
>>          >      >      >      Java HotSpot(TM) 64-Bit Server VM (build 
>> 25.171-b11, mixed mode)
>>          >      >      >
>>          >      >      >      I will add a guard for the Test.  
>> Furthermore, I think I need to check if it exits with 0 for OracleJDK.
>>          >      >      >      The right behavior of OracleJDK8 is to 
>> accept this flag yet suppress all events.
>>          >      >      >
>>          >      >      >      Thanks,
>>          >      >      >      —lx
>>          >      >      >
>>          >      >      >
>>          >      >      >      > On Sep 4, 2018, at 9:51 PM, David 
>> Holmes <david.holmes at oracle.com> wrote:
>>          >      >      >      >
>>          >      >      >      > Hi Liu Xin,
>>          >      >      >      >
>>          >      >      >      > On 4/09/2018 7:45 PM, David Holmes wrote:
>>          >      >      >      >> On 4/09/2018 10:12 AM, Liu, Xin wrote:
>>          >      >      >      >>> Hi, David,
>>          >      >      >      >>> Thank you for reviewing it.
>>          >      >      >      >>>
>>          >      >      >      >>> If you are using OracleJDK, it won't 
>> have any output. OracleJDK doesn't suffer from crash. I guess It 
>> suppress events unconditionally.
>>          >      >      >      >>> For JDK8u, "-XX:+EnableTracing 
>> -version" will dump some events.  It won't crash if it has this patch 
>> http://hg.openjdk.java.net/jdk8u/jdk8u-dev/hotspot/rev/80ee2541504e
>>          >      >      >      >>> The attachment is the output.
>>          >      >      >      >> If the Oracle JDK doesn't produce any 
>> output then this test will fail on the Oracle JDK so can't be accepted 
>> in current form.
>>          >      >      >      >> There's something going on with 
>> EnableTracing that I'm not understanding. I'll try to take a closer 
>> look tomorrow ( was on a day off today).
>>          >      >      >      >
>>          >      >      >      > Okay ... the Oracle JDK replaces the 
>> OpenJDK tracing facility with JFR which completely ignores the 
>> EnableTracing flag.
>>          >      >      >      >
>>          >      >      >      > If the test is simply verifying that 
>> -XX:+EnableTracing doesn't crash then it can reduce to just:
>>          >      >      >      >
>>          >      >      >      > /*
>>          >      >      >      > * @test
>>          >      >      >      > * @summary Ensure -XX:+EnableTracing 
>> doesn't crash
>>          >      >      >      > * @run main/othervm -XX:+EnableTracing 
>> TestEnableTracing
>>          >      >      >      > */
>>          >      >      >      > public class TestEnableTracing {
>>          >      >      >      >    public static void main(String[] 
>> args) {
>>          >      >      >      >    }
>>          >      >      >      > }
>>          >      >      >      >
>>          >      >      >      > If you want to actually verify some of 
>> the trace output then you will need to guard the shouldMatch/Contain 
>> statements with something like:
>>          >      >      >      >
>>          >      >      >      > if 
>> (output.getOutput().contains("OpenJDK")) {
>>          >      >      >      > ...
>>          >      >      >      > }
>>          >      >      >      >
>>          >      >      >      > I don't think there is a direct way to 
>> ensure the test is only executed on an OpenJDK binary.
>>          >      >      >      >
>>          >      >      >      > Thanks,
>>          >      >      >      > David
>>          >      >      >      >
>>          >      >      >      >> Thanks,
>>          >      >      >      >> David
>>          >      >      >      >>>
>>          >      >      >      >>> Output of jtreg
>>          >      >      >      >>> ###
>>          >      >      >      >>> # Running suite single for openJDK
>>          >      >      >      >>> # Results are available in 
>> build/brazil-open-single-tests
>>          >      >      >      >>> ###
>>          >      >      >      >>> Directory "build/JTwork" not found: 
>> creating
>>          >      >      >      >>> Passed: 
>> runtime/EnableTracing/TestEnableTracing.java
>>          >      >      >      >>> Test results: passed: 1
>>          >      >      >      >>> Report written to 
>> /local/home/xxinliu/ws-openjdk8/build/OpenJDK8CompatibilityTests/OpenJDK8CompatibilityTests-1.0/RHEL5_64/DEV.STD.PTHREAD/build/brazil-open-single-tests/html/report.html 
>>
>>          >      >      >      >>> Results written to 
>> /local/home/xxinliu/ws-openjdk8/build/OpenJDK8CompatibilityTests/OpenJDK8CompatibilityTests-1.0/RHEL5_64/DEV.STD.PTHREAD/build/JTwork 
>>
>>          >      >      >      >>>
>>          >      >      >      >>> Thanks,
>>          >      >      >      >>> --lx
>>          >      >      >      >>>
>>          >      >      >      >>>
>>          >      >      >      >>> On 9/2/18, 10:01 PM, "David Holmes" 
>> <david.holmes at oracle.com> wrote:
>>          >      >      >      >>>
>>          >      >      >      >>>      Hi,
>>          >      >      >      >>>      On 25/08/2018 2:35 AM, Liu, Xin 
>> wrote:
>>          >      >      >      >>>      > Hi, hotspot-Runtime Developers,
>>          >      >      >      >>>      >
>>          >      >      >      >>>      > Could you review this small 
>> tests?
>>          >      >      >      >>>      > It adds a new test to verify 
>> +EnableTracing works for jdk8u.
>>          >      >      >      >>>      I don't see any output produced 
>> with just "-XX:+EnableTracing -version". ??
>>          >      >      >      >>>        >
>>          >      >      >      >>>     
>> /java/re/jdk/8u172/promoted/latest/binaries/linux-amd64/fastdebug/bin/java 
>>
>>          >      >      >      >>>      -XX:+EnableTracing -version
>>          >      >      >      >>>      java version "1.8.0_172-fastdebug"
>>          >      >      >      >>>      Java(TM) SE Runtime Environment 
>> (build 1.8.0_172-fastdebug-b35)
>>          >      >      >      >>>      Java HotSpot(TM) 64-Bit Server 
>> VM (build 25.172-b35-fastdebug, mixed mode)
>>          >      >      >      >>>      Thanks,
>>          >      >      >      >>>      David
>>          >      >      >      >>>      > BUG:  
>> https://bugs.openjdk.java.net/browse/JDK-8209863
>>          >      >      >      >>>      > WEBREV: 
>> http://cr.openjdk.java.net/~phh/8209863/webrev.00/
>>          >      >      >      >>>      >
>>          >      >      >      >>>      > Thanks,
>>          >      >      >      >>>      > -lx
>>          >      >      >      >>>      >
>>          >      >      >      >>>
>>          >      >      >
>>          >      >      >
>>          >      >      >
>>          >      >
>>          >      >
>>          >
>>          >
>>


More information about the hotspot-runtime-dev mailing list