<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">3rd Webrev for review please:<br>
<br>
This version does away with the common abstract base case
ConcurrentServiceThread, and pushes the code down to the concrete
classes. This may get cleaned up in a later fix to
ConcurrentGCThread.<br>
<br>
<br>
RFE: <a moz-do-not-send="true" class="issue-link"
data-issue-key="JDK-8138920"
href="https://bugs.openjdk.java.net/browse/JDK-8138920"
id="key-val" rel="4848143">JDK-8138920</a> Refactor the sampling
thread from ConcurrentG1RefineThread<br>
<br>
Webrev: <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~drwhite/8138920/webrev.03/">http://cr.openjdk.java.net/~drwhite/8138920/webrev.03/</a><br>
<br>
Passed jprt (finally!).<br>
<br>
Thanks for looking!<br>
<br>
- Derek<br>
<a moz-do-not-send="true" class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Edrwhite/8138920/webrev.01/"></a><br>
On 10/9/15 12:01 PM, Bengt Rutisson wrote:<br>
</div>
<blockquote cite="mid:5617E4C6.3040501@oracle.com" type="cite">
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
<br>
Hi Derek,<br>
<br>
Comments inlined.<br>
<br>
<div class="moz-cite-prefix">On 2015-10-09 00:29, Derek White
wrote:<br>
</div>
<blockquote cite="mid:5616EE55.4010902@oracle.com" type="cite">
<meta content="text/html; charset=utf-8"
http-equiv="Content-Type">
<div class="moz-cite-prefix">Another call for review:<br>
<br>
2nd webrev:<br>
<a moz-do-not-send="true" class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Edrwhite/8138920/webrev.02/">http://cr.openjdk.java.net/~drwhite/8138920/webrev.02/</a><br>
<br>
See changes and comments below:<br>
<br>
On 10/8/15 2:47 AM, Bengt Rutisson wrote:<br>
</div>
<blockquote cite="mid:5616118F.2060208@oracle.com" type="cite">
<meta content="text/html; charset=utf-8"
http-equiv="Content-Type">
<br>
Hi Derek,<br>
<br>
<div class="moz-cite-prefix">On 2015-10-07 17:19, Derek White
wrote:<br>
</div>
<blockquote cite="mid:5615381F.9010801@oracle.com" type="cite">
<meta content="text/html; charset=utf-8"
http-equiv="Content-Type">
<div class="moz-cite-prefix">Hi Bengt,<br>
<br>
On 10/7/15 4:02 AM, Bengt Rutisson wrote:<br>
</div>
<blockquote cite="mid:5614D1AA.5000105@oracle.com"
type="cite">
<meta content="text/html; charset=utf-8"
http-equiv="Content-Type">
Hi Derek,<br>
<br>
Thanks for fixing this!<br>
<br>
<div class="moz-cite-prefix">On 2015-10-06 19:51, Derek
White wrote:<br>
</div>
<blockquote cite="mid:56140A14.2030703@oracle.com"
type="cite">
<meta http-equiv="content-type" content="text/html;
charset=utf-8">
Refactor and cleanup the G1 concurrent thread classes:<br>
- Pull out a sampling thread class (now
ConcurrentG1SampleThread) from ConcurrentG1RefineThread.<br>
- Factor out an abstract base class
ConcurrentG1ServiceThread that is used by:<br>
- ConcurrentG1RefineThread<br>
- ConcurrentG1SampleThread<br>
- ConcurrentMarkThread<br>
- Made the handling of the "primary" refinement thread
more explicit.<br>
- Updated obsolete and confusing comments<br>
<br>
This is tech debt that also will allow disabling
concurrent refinement (if desired) and also fixes a P4
bug.<br>
Patch started by Thomas and improved and/or mangled
myself.<br>
<br>
RFE:
<meta http-equiv="content-type" content="text/html;
charset=utf-8">
<a moz-do-not-send="true" class="issue-link"
data-issue-key="JDK-8138920"
href="https://bugs.openjdk.java.net/browse/JDK-8138920"
id="key-val" rel="4848143">JDK-8138920</a> Refactor
the sampling thread from ConcurrentG1RefineThread<br>
<br>
Webrev: <a moz-do-not-send="true"
class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Edrwhite/8138920/webrev.01/">http://cr.openjdk.java.net/~drwhite/8138920/webrev.01/</a><br>
</blockquote>
<br>
Overall this looks really good to me. <br>
<br>
Some comments:<br>
<br>
No one seems to depend on
ConcurrentG1ServiceThread::vtime_accum(). All uses have
the specific subclass available. So, I don't think the
pure virtual declaration in ConcurrentG1ServiceThread is
needed. I'd just remove that and also make the
corresponding methods in the subclasses non-virtual.<br>
</blockquote>
<br>
OK. At some point we need to systematic rewrite of timing,
but that can wait.<br>
</blockquote>
<br>
Quite agree. <br>
<br>
<blockquote cite="mid:5615381F.9010801@oracle.com" type="cite">
<blockquote cite="mid:5614D1AA.5000105@oracle.com"
type="cite"> That more or less only leaves the run() and
stop() methods in ConcurrentG1ServiceThread. It is kind of
nice for the subclasses to get help with this, but I
wonder if it is not possible to improve the
ConcurrentGCThread class to help with this instead.
Basically I guess I am a little unsure if the extra class
ConcurrentG1ServiceThread is really needed.<br>
<br>
</blockquote>
I'll look at ConcurrentGCThread to see how well it could
cover these cases.<br>
</blockquote>
<br>
Thanks. I think it is worth a try. If it doesn't turn out well
we can keep the intermediate class, but I think it is worth
exploring.<br>
</blockquote>
<br>
I looked at pushing ConcurrentServiceThread up into
ConcurrentGCThread, but things got a little complicated.
ConcurrentG1RefineThread, ConcurrentMarkThread, and
ConcurrentSampleThread have a very "regularized" implementation
of the "termination protocol". G1StringDedupThread is slightly
off from this, and ConcurrentMarkSweepThread more so. Pushing
the shared code up into ConcurrentGCThread but not using it in
G1StringDedupThread and ConcurrentMarkSweepThread seems
confusing.<br>
<br>
There's a tension between providing a framework that handles all
shared factorizable code, but can become a straight jacket for
future code, and everyone doing everything separately and
differently. This webrev is somewhere in the middle. Some of the
changes between webrev.01 and .02 are to make the duplicated
code more similar, even though it's not shared.<br>
</blockquote>
<br>
I see. Thanks for investigating it!<br>
<br>
I think I agree with Per, though. The value of
ConcurrentServiceThread in its current form is IMHO not really
worth the extra complexity of having it. I would prefer to just
duplicate the code in ConcurrentG1RefineThread,
ConcurrentMarkThread, and ConcurrentSampleThread for now and then
have a separate change to try to clean this part up.<br>
<br>
<blockquote cite="mid:5616EE55.4010902@oracle.com" type="cite"> <br>
<blockquote cite="mid:5616118F.2060208@oracle.com" type="cite">
<blockquote cite="mid:5615381F.9010801@oracle.com" type="cite">
<blockquote cite="mid:5614D1AA.5000105@oracle.com"
type="cite"> Naming. The naming in G1 is a bit
inconsistent. Some files and classes are prefixed with G1
and some are not. But if they are called something with G1
it is normally a prefix. So, I would prefer the new
classes to be called G1Concurrent* instead of
ConcurrentG1*.<br>
</blockquote>
<br>
So we have:<br>
- ConcurrentG1RefineThread<br>
- ConcurrentMarkThread<br>
<br>
And I added:<br>
- ConcurrentG1SampleThread<br>
- ConcurrentG1ServiceThread<br>
<br>
And maybe I'm removing ConcurrentG1ServiceThread. In that
case I'd be inclined to rename:<br>
ConcurrentG1SampleThread => ConcurrentSampleThread </blockquote>
<br>
Sounds good. The ConcurrentG1Refine* classes are really the
only oddly named G1 classes, so I think it is better not to
let that naming spread.<br>
</blockquote>
This version includes the class renaming. </blockquote>
<br>
Thanks for fixing this.<br>
<br>
If we do decide to keep ConcurrentServiceThread around I think it
would be better to move ConcurrentSampleThread into its own file.
It is a separate enough entity to warrant its own file I think.<br>
<br>
<blockquote cite="mid:5616EE55.4010902@oracle.com" type="cite">
<blockquote cite="mid:5616118F.2060208@oracle.com" type="cite">
<br>
<blockquote cite="mid:5615381F.9010801@oracle.com" type="cite">
<blockquote cite="mid:5614D1AA.5000105@oracle.com"
type="cite"> You write "and also fixes a P4 bug". Which
bug is that?<br>
</blockquote>
<meta http-equiv="content-type" content="text/html;
charset=utf-8">
<a moz-do-not-send="true" class="issue-link"
data-issue-key="JDK-8136856"
href="https://bugs.openjdk.java.net/browse/JDK-8136856"
id="key-val" rel="4845182">JDK-8136856</a> G1 makes two
concurrent refinement threads with
-XX:G1ConcRefinementThreads=1<br>
(because the sampling thread "is-a" concurrent refinement
thread.<br>
</blockquote>
<br>
Ah. I see. Makes sense. Thanks.<br>
<br>
But it is still not possible to turn refinement off by setting
-XX:G1ConcRefinementThreads=0 since that is considered the
default, right?<br>
</blockquote>
<br>
I'm not sure about this. If I recall correctly, Thomas implied
that it was hard to disable concurrent refinement without
disabling the sampling thread too.<br>
</blockquote>
<br>
I'm thinking about this code in Arguments::set_g1_gc_flags():<br>
<br>
if (G1ConcRefinementThreads == 0) {<br>
FLAG_SET_DEFAULT(G1ConcRefinementThreads, ParallelGCThreads);<br>
}<br>
<br>
Could we change that, so that you could turn off refinement by
setting -XX:G1ConcRefinementThreads=0 ? It should probably be
handled as a separate fix though.<br>
<br>
Thanks,<br>
Bengt<br>
<br>
<blockquote cite="mid:5616EE55.4010902@oracle.com" type="cite"> <br>
<blockquote cite="mid:5616118F.2060208@oracle.com" type="cite">
<blockquote cite="mid:5615381F.9010801@oracle.com" type="cite">
I cant' recall how to mark a bug as blocking another bug.<br>
</blockquote>
<br>
You add a link (More->Link) to the other bug and choose
"block" or "is blocked by".<br>
<br>
Thanks,<br>
Bengt<br>
</blockquote>
Thanks for the tip!<br>
<br>
- Derek<br>
<br>
</blockquote>
<br>
</blockquote>
<br>
</body>
</html>