From ashipile at redhat.com  Wed Jun  7 13:25:20 2017
From: ashipile at redhat.com (ashipile at redhat.com)
Date: Wed, 07 Jun 2017 13:25:20 +0000
Subject: hg: code-tools/jmh: 7901990: Make sure only a single profiler of
 given type is used
Message-ID: <201706071325.v57DPKwL006122@aojmv0008.oracle.com>

Changeset: a128fd4a5901
Author:    shade
Date:      2017-06-07 15:19 +0200
URL:       http://hg.openjdk.java.net/code-tools/jmh/rev/a128fd4a5901

7901990: Make sure only a single profiler of given type is used

+ jmh-core-it/src/test/java/org/openjdk/jmh/it/profilers/DuplicateProfilerTest.java
! jmh-core/src/main/java/org/openjdk/jmh/runner/Runner.java


From jzillmann at googlemail.com  Wed Jun 14 04:09:46 2017
From: jzillmann at googlemail.com (Johannes Zillmann)
Date: Wed, 14 Jun 2017 06:09:46 +0200
Subject: Visualizing JMH benchmark results
In-Reply-To: <CAEEKBgWQuUhFjWW2P8skrQsAKsWWSDWt1k3y5Wgp-na_B6eZbQ@mail.gmail.com>
References: <CAEEKBgWQuUhFjWW2P8skrQsAKsWWSDWt1k3y5Wgp-na_B6eZbQ@mail.gmail.com>
Message-ID: <1D11CF23-093D-4B0F-BAA1-09AD5BAC9299@googlemail.com>

Hi there,

just had the same questions a few month back and started to write some tools? 
The outcome is:
	A online visualizer: http://jmh.morethan.io
	A gradle report plugin: https://plugins.gradle.org/plugin/io.morethan.jmhreport

Both are still under development, but i find them already pretty useful (They speed up my digestion of the benchmark results, esp. if the benchmark suite was bigger).
Feedback welcome!

cheers
Johannes

PS: Lately somebody pointed out that there is https://nilskp.github.io/jmh-charts/ for several years now. So there was a special tool for that already to the time of the request ;)?.
(Doesn?t load in my Safari, but Chrome, etc.. is working fine!).

> On 28 Jun 2014, at 21:15, Behrooz Nobakht <nobeh5 at gmail.com> wrote:
> 
> Hi,
> 
> Since JMH generates CSV results, there are already various tools that can
> be used for visualization purposes. However, I was wondering what would the
> JMH community/authors recommend from their experience?
> 
> Specifically, Aleksey, what do you use for the visualizations in this post?
> http://shipilev.net/blog/2014/nanotrusting-nanotime/
> 
> Thanks,
> Behrooz
> 


From barteks2x at gmail.com  Wed Jun 14 07:21:20 2017
From: barteks2x at gmail.com (Bartosz Skrzypczak)
Date: Wed, 14 Jun 2017 09:21:20 +0200
Subject: Random SIGSEGV when running benchmark with perfasm - JMH, hsdis or
 JVM issue?
Message-ID: <CAKGpm2tHmopeyDd3v_FUej=6YSGtUsYSM5tE87Kd_C0JPQzebA@mail.gmail.com>

I already talked about it when I asked about "Random
java.util.NoSuchElementException exceptions when using perfasm"
 back in january 2017.  I know this is probably not the right place to
ask but I don't have any better idea. I think it may be the same
crash, but it was so long ago that I'm not sure. I am able to
reproduce it only with JMH with perfasm on my project so far, and is
uses hsdis, which means it may not necessarily be JVM issue. And I'm
profiling the same code part of code (a big part of code) as back
then. The problematic frame I'm getting is "V  [libjvm.so+0x8e095d]
Monitor::lock_without_safepoint_check()+0x3d".

I'm not really sure where I should even report it. I just happened to
get a core dump from it so this is the first time I may have enough
information to actually report it somewhere.

Could I get some information as to what would be the appropriate place
to report it?

From nitsanw at yahoo.com  Wed Jun 14 07:59:11 2017
From: nitsanw at yahoo.com (Nitsan Wakart)
Date: Wed, 14 Jun 2017 07:59:11 +0000 (UTC)
Subject: Random SIGSEGV when running benchmark with perfasm - JMH, hsdis
 or JVM issue?
In-Reply-To: <CAKGpm2tHmopeyDd3v_FUej=6YSGtUsYSM5tE87Kd_C0JPQzebA@mail.gmail.com>
References: <CAKGpm2tHmopeyDd3v_FUej=6YSGtUsYSM5tE87Kd_C0JPQzebA@mail.gmail.com>
Message-ID: <451163255.804622.1497427151684@mail.yahoo.com>

Hi,Please include basic information:- A link to the hs_err file(post it somewhere if you can)- A link to your original email- Versions: OS/JVM/hsdis- HWThanks,Nitsan 

    On Wednesday, June 14, 2017 9:22 AM, Bartosz Skrzypczak <barteks2x at gmail.com> wrote:
 

 I already talked about it when I asked about "Random
java.util.NoSuchElementException exceptions when using perfasm"
 back in january 2017.? I know this is probably not the right place to
ask but I don't have any better idea. I think it may be the same
crash, but it was so long ago that I'm not sure. I am able to
reproduce it only with JMH with perfasm on my project so far, and is
uses hsdis, which means it may not necessarily be JVM issue. And I'm
profiling the same code part of code (a big part of code) as back
then. The problematic frame I'm getting is "V? [libjvm.so+0x8e095d]
Monitor::lock_without_safepoint_check()+0x3d".

I'm not really sure where I should even report it. I just happened to
get a core dump from it so this is the first time I may have enough
information to actually report it somewhere.

Could I get some information as to what would be the appropriate place
to report it?


   

From leonardo.f.gomes at gmail.com  Wed Jun 14 11:46:59 2017
From: leonardo.f.gomes at gmail.com (Leonardo Gomes)
Date: Wed, 14 Jun 2017 13:46:59 +0200
Subject: Visualizing JMH benchmark results
In-Reply-To: <1D11CF23-093D-4B0F-BAA1-09AD5BAC9299@googlemail.com>
References: <CAEEKBgWQuUhFjWW2P8skrQsAKsWWSDWt1k3y5Wgp-na_B6eZbQ@mail.gmail.com>
 <1D11CF23-093D-4B0F-BAA1-09AD5BAC9299@googlemail.com>
Message-ID: <CAEAdJ9r+fqWMxXKqRoJRR5MPM76w0Za6iAQkhb1PBoP97EuFuw@mail.gmail.com>

Great work!

It would be nice if we could use it with
https://github.com/blackboard/jmh-jenkins
And/or if you could provide a REST API, so that one could call
http://jmh.morethan.io/ passing a JSON payload that corresponds to a JMH
result.

Cheers,
Leo.

On Wed, Jun 14, 2017 at 6:09 AM, Johannes Zillmann <jzillmann at googlemail.com
> wrote:

> Hi there,
>
> just had the same questions a few month back and started to write some
> tools?
> The outcome is:
>         A online visualizer: http://jmh.morethan.io
>         A gradle report plugin: https://plugins.gradle.org/
> plugin/io.morethan.jmhreport
>
> Both are still under development, but i find them already pretty useful
> (They speed up my digestion of the benchmark results, esp. if the benchmark
> suite was bigger).
> Feedback welcome!
>
> cheers
> Johannes
>
> PS: Lately somebody pointed out that there is
> https://nilskp.github.io/jmh-charts/ for several years now. So there was
> a special tool for that already to the time of the request ;)?.
> (Doesn?t load in my Safari, but Chrome, etc.. is working fine!).
>
> > On 28 Jun 2014, at 21:15, Behrooz Nobakht <nobeh5 at gmail.com> wrote:
> >
> > Hi,
> >
> > Since JMH generates CSV results, there are already various tools that can
> > be used for visualization purposes. However, I was wondering what would
> the
> > JMH community/authors recommend from their experience?
> >
> > Specifically, Aleksey, what do you use for the visualizations in this
> post?
> > http://shipilev.net/blog/2014/nanotrusting-nanotime/
> >
> > Thanks,
> > Behrooz
> >
>
>

From jzillmann at googlemail.com  Wed Jun 14 13:49:26 2017
From: jzillmann at googlemail.com (Johannes Zillmann)
Date: Wed, 14 Jun 2017 15:49:26 +0200
Subject: Visualizing JMH benchmark results
In-Reply-To: <CAEAdJ9r+fqWMxXKqRoJRR5MPM76w0Za6iAQkhb1PBoP97EuFuw@mail.gmail.com>
References: <CAEEKBgWQuUhFjWW2P8skrQsAKsWWSDWt1k3y5Wgp-na_B6eZbQ@mail.gmail.com>
 <1D11CF23-093D-4B0F-BAA1-09AD5BAC9299@googlemail.com>
 <CAEAdJ9r+fqWMxXKqRoJRR5MPM76w0Za6iAQkhb1PBoP97EuFuw@mail.gmail.com>
Message-ID: <05DFEFE7-3840-44CE-BA3C-125D18501D76@googlemail.com>

Thanks Leo,

> On 14 Jun 2017, at 13:46, Leonardo Gomes <leonardo.f.gomes at gmail.com> wrote:
> 
> Great work!
> 
> It would be nice if we could use it with https://github.com/blackboard/jmh-jenkins <https://github.com/blackboard/jmh-jenkins>
a Jenkins implementation is high on the list i want to have ;)
Are you the maintainer of that plugin ? Its relative easy to integrate the javascript visualizer into an offline app? thats what i?ve done for the gradle plugin as well...

> And/or if you could provide a REST API, so that one could call http://jmh.morethan.io/ <http://jmh.morethan.io/> passing a JSON payload that corresponds to a JMH result.
That would be handy, but the site is currently server-less (deployed as gihub pages app) and thus cannot take POST requests (it?s pure local javascript)?

best
Johannes


> 
> Cheers,
> Leo.
> 
> On Wed, Jun 14, 2017 at 6:09 AM, Johannes Zillmann <jzillmann at googlemail.com <mailto:jzillmann at googlemail.com>> wrote:
> Hi there,
> 
> just had the same questions a few month back and started to write some tools?
> The outcome is:
>         A online visualizer: http://jmh.morethan.io <http://jmh.morethan.io/>
>         A gradle report plugin: https://plugins.gradle.org/plugin/io.morethan.jmhreport <https://plugins.gradle.org/plugin/io.morethan.jmhreport>
> 
> Both are still under development, but i find them already pretty useful (They speed up my digestion of the benchmark results, esp. if the benchmark suite was bigger).
> Feedback welcome!
> 
> cheers
> Johannes
> 
> PS: Lately somebody pointed out that there is https://nilskp.github.io/jmh-charts/ <https://nilskp.github.io/jmh-charts/> for several years now. So there was a special tool for that already to the time of the request ;)?.
> (Doesn?t load in my Safari, but Chrome, etc.. is working fine!).
> 
> > On 28 Jun 2014, at 21:15, Behrooz Nobakht <nobeh5 at gmail.com <http://gmail.com/>> wrote:
> >
> > Hi,
> >
> > Since JMH generates CSV results, there are already various tools that can
> > be used for visualization purposes. However, I was wondering what would the
> > JMH community/authors recommend from their experience?
> >
> > Specifically, Aleksey, what do you use for the visualizations in this post?
> > http://shipilev.net/blog/2014/nanotrusting-nanotime/ <http://shipilev.net/blog/2014/nanotrusting-nanotime/>
> >
> > Thanks,
> > Behrooz
> >
> 
> 


From shade at redhat.com  Fri Jun 30 14:30:12 2017
From: shade at redhat.com (Aleksey Shipilev)
Date: Fri, 30 Jun 2017 16:30:12 +0200
Subject: Random SIGSEGV when running benchmark with perfasm - JMH, hsdis
 or JVM issue?
In-Reply-To: <451163255.804622.1497427151684@mail.yahoo.com>
References: <CAKGpm2tHmopeyDd3v_FUej=6YSGtUsYSM5tE87Kd_C0JPQzebA@mail.gmail.com>
 <451163255.804622.1497427151684@mail.yahoo.com>
Message-ID: <b8f5d3e3-697b-e807-1cb0-f32e8ececb0f@redhat.com>

What Nitsan said.

You can report it to bugs.java.com, but it would be handier if you could add the
basic information here. The stack trace and the hs_err might be revealing.

-Aleksey

On 06/14/2017 09:59 AM, Nitsan Wakart wrote:
> Hi,Please include basic information:- A link to the hs_err file(post it somewhere if you can)- A link to your original email- Versions: OS/JVM/hsdis- HWThanks,Nitsan 
> 
>     On Wednesday, June 14, 2017 9:22 AM, Bartosz Skrzypczak <barteks2x at gmail.com> wrote:
>  
> 
>  I already talked about it when I asked about "Random
> java.util.NoSuchElementException exceptions when using perfasm"
>  back in january 2017.  I know this is probably not the right place to
> ask but I don't have any better idea. I think it may be the same
> crash, but it was so long ago that I'm not sure. I am able to
> reproduce it only with JMH with perfasm on my project so far, and is
> uses hsdis, which means it may not necessarily be JVM issue. And I'm
> profiling the same code part of code (a big part of code) as back
> then. The problematic frame I'm getting is "V  [libjvm.so+0x8e095d]
> Monitor::lock_without_safepoint_check()+0x3d".
> 
> I'm not really sure where I should even report it. I just happened to
> get a core dump from it so this is the first time I may have enough
> information to actually report it somewhere.
> 
> Could I get some information as to what would be the appropriate place
> to report it?
> 
> 
>    
>