<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Hi Bengt,<br>
<br>
(surprise!)<br>
<br>
This is a useful change. Both webrevs look good. I have a slight
preference to the first one. I didn't like the use of
_allocation_failure on the second one here:<br>
<meta charset="utf-8">
<pre style="color: rgb(0, 0, 0); font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(238, 238, 238);"> 984 result = do_collection_pause(word_size, gc_count_before, &succeeded,
<span class="changed" style="color: blue;"> 985 GCCause::_allocation_failure);</span>
</pre>
given that the allocation didn't really fail - the eden is just
full. The other use of _allocation_failure (in the
VM_G1CollectorForAllocation closure) really implies that an
allocation did fail and G1 is trying to take evasive action.<br>
<br>
Tony<br class="Apple-interchange-newline">
<br>
<br>
<br>
<br>
<div class="moz-cite-prefix">On 8/21/13 7:47 AM, Bengt Rutisson
wrote:<br>
</div>
<blockquote cite="mid:5214D2E9.7080502@oracle.com" type="cite">
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
Hi All,<br>
<br>
Could I have a couple for reviews for this small change?<br>
<br>
<a moz-do-not-send="true" class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Ebrutisso/8022872/webrev.00/">http://cr.openjdk.java.net/~brutisso/8022872/webrev.00/</a><br>
<br>
This change makes sure that we report "humongous allocation" as
the GC cause when we trigger a young collection because we are
attempting a humongous allocation.<br>
<br>
Before the change a log entry for such a young GC looked like
this:<br>
<br>
[GC pause (G1 Evacuation Pause) (young) 1364M->1364M(2048M),
0.0004200 secs]<br>
<br>
After the change it looks like this:<br>
<br>
[GC pause (G1 Humongous Allocation) (young)
1364M->1364M(2047M), 0.0541270 secs]<br>
<br>
While doing this change I realized that the GC cause
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
_g1_inc_collection_pause was only used in one more place. The slow
path for allocations. There is already a cause called
_allocation_failure, so I suggest that we use that one instead.
This will eliminate the need for _g1_inc_collection_pause.<br>
<br>
In that case the webrev will look like this:<br>
<br>
<a moz-do-not-send="true" class="moz-txt-link-freetext"
href="http://cr.openjdk.java.net/%7Ebrutisso/8022872/webrev_incCause/">http://cr.openjdk.java.net/~brutisso/8022872/webrev_incCause/</a><br>
<br>
I prefer this cleanup, but I'm fine with just doing the first one
if we think _g1_inc_collection_pause cause to keep for some
reason.<br>
<br>
Thanks,<br>
Bengt<br>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Tony Printezis | Staff Software Engineer | Twitter
@TonyPrintezis
<a class="moz-txt-link-abbreviated" href="mailto:tprintezis@twitter.com">tprintezis@twitter.com</a>
</pre>
</body>
</html>