[prev in list] [next in list] [prev in thread] [next in thread]
List: openjdk-serviceability-dev
Subject: Re: RFR: 7153107 com/sun/jdi/InterruptHangTest.java fails in nightlies
From: Staffan Larsen <staffan.larsen () oracle ! com>
Date: 2016-03-22 18:22:41
Message-ID: 08A03ACD-1FD7-4A68-B4C5-7C9D399BFC23 () oracle ! com
[Download RAW message or body]
Thanks Serguei!
> On 22 mars 2016, at 19:10, serguei.spitsyn@oracle.com wrote:
>
> On 3/22/16 11:09, Staffan Larsen wrote:
> >
> > > On 22 mars 2016, at 18:15, \
> > > <mailto:serguei.spitsyn@oracle.com>serguei.spitsyn@oracle.com \
> > > <mailto:serguei.spitsyn@oracle.com> wrote:
> > > Hi Staffan,
> > >
> > > It looks good.
> > > One question though.
> > >
> > > I do not see that the timeout is increased in the \
> > > com/sun/jdi/InterruptHangTest.java. It remains 20 sec as was before.
> > >
> > > + float timeoutFactor = \
> > > Float.parseFloat(System.getProperty("test.timeout.factor", "1.0"));+ \
> > > long sleepSeconds = (long)(20 * timeoutFactor);
> > > ...
> > > - Thread.sleep(20000);
> > > + Thread.sleep(sleepSeconds * 1000);
> > > Or an increase is because of the "test.timeout.factor" property?
> > Yes, the increase will be because outside systems will set the timeout factor.
> Ok, thanks. Reviewed. Thanks, Serguei
> >
> > /Staffan
> > > Thanks, Serguei On 3/22/16 07:14, Staffan Larsen wrote:
> > > > Anyone up for a review of this change?
> > > >
> > > > > On 21 mars 2016, at 09:16, Staffan Larsen <staffan.larsen@oracle.com> \
> > > > > <mailto:staffan.larsen@oracle.com> wrote:
> > > > > Hi,
> > > > >
> > > > > For a while we have had various timeout related problem in the JDI tests. \
> > > > > The root cause of this seems to be the Indy String Concatenation work which \
> > > > > makes stepping over the first string concatenation horrendously slow in \
> > > > > debug version of the jdk. It is not ISC itself that is to blame, instead it \
> > > > > gets the blame for the full setup of the invokedynamic framework. See \
> > > > > JDK-8151887 for a longer discussion about this.
> > > > > One of the tests that fail because of this is \
> > > > > com/sun/jdi/InterruptHangTest.java. In this fix, I have simply increased \
> > > > > the timeout that is used internally in the test to take the jtreg \
> > > > > timeoutfactor into consideration. This seems to allow enough time for \
> > > > > stepping through the indy setup code.
> > > > > I have also added timestamped logging to the TestScaffold superclass to \
> > > > > make it easier to see what is taking a long time and how far a test has \
> > > > > proceeded.
> > > > > webrev: http://cr.openjdk.java.net/~sla/7153107/webrev.01/ \
> > > > > <http://cr.openjdk.java.net/%7Esla/7153107/webrev.01/>
> > > > > bug: https://bugs.openjdk.java.net/browse/JDK-7153107 \
> > > > > <https://bugs.openjdk.java.net/browse/JDK-7153107>
> > > > > Thanks,
> > > > > /Staffan
> > > > >
[Attachment #3 (unknown)]
<html><head><meta http-equiv="Content-Type" content="text/html \
charset=windows-1252"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: \
space; -webkit-line-break: after-white-space;" class="">Thanks Serguei!<div \
class=""><br class=""><div><blockquote type="cite" class=""><div class="">On 22 mars \
2016, at 19:10, <a href="mailto:serguei.spitsyn@oracle.com" \
class="">serguei.spitsyn@oracle.com</a> wrote:</div><br \
class="Apple-interchange-newline"><div class="">
<meta content="text/html; charset=windows-1252" http-equiv="Content-Type" \
class="">
<div bgcolor="#FFFFFF" text="#000000" class="">
<div class="moz-cite-prefix">On 3/22/16 11:09, Staffan Larsen wrote:<br class="">
</div>
<blockquote cite="mid:E5F4DB1A-FF25-4FC6-8B40-1DAFE87FBF64@oracle.com" \
type="cite" class=""> <meta http-equiv="Content-Type" content="text/html;
charset=windows-1252" class="">
<br class="">
<div class="">
<blockquote type="cite" class="">
<div class="">On 22 mars 2016, at 18:15, <a moz-do-not-send="true" \
href="mailto:serguei.spitsyn@oracle.com" class=""></a><a \
class="moz-txt-link-abbreviated" \
href="mailto:serguei.spitsyn@oracle.com">serguei.spitsyn@oracle.com</a> wrote:</div>
<br class="Apple-interchange-newline">
<div class="">
<meta content="text/html; charset=windows-1252" http-equiv="Content-Type" \
class=""> <div bgcolor="#FFFFFF" text="#000000" class="">
<div class="moz-cite-prefix">Hi Staffan,<br class="">
<br class="">
It looks good.<br class="">
One question though.<br class="">
<br class="">
I do not see that the timeout is increased in the
com/sun/jdi/InterruptHangTest.java.<br class="">
It remains 20 sec as was before.<br class="">
<br class="">
<meta http-equiv="content-type" content="text/html;
charset=windows-1252" class="">
<meta http-equiv="content-type" content="text/html;
charset=windows-1252" class="">
<pre class=""><span class="new"><meta http-equiv="content-type" \
content="text/html; charset=windows-1252" class=""><span class="new">+ \
float timeoutFactor = Float.parseFloat(System.getProperty("test.timeout.factor", \
"1.0"));</span>+ long sleepSeconds = (long)(20 * timeoutFactor);
...
</span></pre><span class="new"></span><meta http-equiv="content-type" \
content="text/html; charset=windows-1252" class=""><pre class=""><span \
class="removed">- Thread.sleep(20000);</span> <span \
class="new">+ Thread.sleep(sleepSeconds * \
1000);</span></pre>
Or an increase is because of the <span class="new">"test.timeout.factor" \
property?</span></div></div></div></blockquote><div class=""> </div><div \
class="">Yes, the increase will be because outside systems will set the timeout \
factor.</div></div></blockquote> Ok, thanks.
Reviewed.
Thanks,
Serguei
<blockquote cite="mid:E5F4DB1A-FF25-4FC6-8B40-1DAFE87FBF64@oracle.com" type="cite" \
class=""><div class=""><div class=""> </div><div class="">/Staffan</div><div \
class=""> </div>
<blockquote type="cite" class=""><div class=""><div bgcolor="#FFFFFF" text="#000000" \
class=""><div class="moz-cite-prefix"><span class="new">
Thanks,
Serguei
</span><meta http-equiv="content-type" content="text/html; charset=windows-1252" \
class="">
On 3/22/16 07:14, Staffan Larsen wrote:
</div><blockquote cite="mid:FD30E9F5-3F6E-4FD2-9893-6832E2ACE1C1@oracle.com" \
type="cite" class=""><pre class="" wrap="">Anyone up for a review of this change?
</pre><blockquote type="cite" class=""><pre class="" wrap="">On 21 mars 2016, at \
09:16, Staffan Larsen <a moz-do-not-send="true" class="moz-txt-link-rfc2396E" \
href="mailto:staffan.larsen@oracle.com"><staffan.larsen@oracle.com></a> wrote:
Hi,
For a while we have had various timeout related problem in the JDI tests. The root \
cause of this seems to be the Indy String Concatenation work which makes stepping \
over the first string concatenation horrendously slow in debug version of the jdk. It \
is not ISC itself that is to blame, instead it gets the blame for the full setup of \
the invokedynamic framework. See JDK-8151887 for a longer discussion about this.
One of the tests that fail because of this is com/sun/jdi/InterruptHangTest.java. In \
this fix, I have simply increased the timeout that is used internally in the test to \
take the jtreg timeoutfactor into consideration. This seems to allow enough time for \
stepping through the indy setup code.
I have also added timestamped logging to the TestScaffold superclass to make it \
easier to see what is taking a long time and how far a test has proceeded.
webrev: <a moz-do-not-send="true" class="moz-txt-link-freetext" \
href="http://cr.openjdk.java.net/%7Esla/7153107/webrev.01/">http://cr.openjdk.java.net/~sla/7153107/webrev.01/</a>
bug: <a moz-do-not-send="true" class="moz-txt-link-freetext" \
href="https://bugs.openjdk.java.net/browse/JDK-7153107">https://bugs.openjdk.java.net/browse/JDK-7153107</a>
Thanks,
/Staffan
</pre></blockquote>
</blockquote>
</div></div></blockquote></div>
</blockquote>
</div></div></blockquote></div><br class=""></div></body></html>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic