I don't think one should discount the possibility of bugs in the JIT
compiler uncovered by Lucene having a much broader impact in other
production code that hasn't yet been run for long periods of time on
Java 7. Just because something runs on Java 7 for an hour or two isn't
sufficient -- we've seen JIT bugs surface [1] that require days of
execution to get enough passes through critical code to JIT it to the
point of breakage.
Bugs in the JIT compiler should be taken extremely seriously. Crashes,
poor performance, etc, are not nearly as big of a deal as quietly and
mysteriously incorrect results.
Oracle had 5 days to change the default value of the optimization flag
in question -- and I believe they should have done just that. If I
wanted a JVM where I had to flip a bunch of knobs just to get it to run
correctly I'd use IBM's JVM.
--
Jess Holle
[1] This was in Tomcat and due to a JIT bug involving String.indexOf()
optimization on some Intel hardware that occurs from Java 6 Update 18
through 23.
On 8/3/2011 10:17 AM, Robert Casto wrote:
And why should Oracle have to worry about what the Lucene guys are
doing anyway. Now that they know about the bug they will test for it.
But they can't be expected to test every piece of software out there
that runs on Java. The Lucene guys should have tested against Java7
long ago and seen this issue and reported it or worked around it. I'm
just glad that a new version is out and I can't wait for Java8 and
hopefully get some long awaited for features, better date library, and
much more. With a great Java8 release, Java as a language will have a
long life ahead.
On Wed, Aug 3, 2011 at 11:05 AM, Kirk <[email protected]
<mailto:[email protected]>> wrote:
I don't think this is a denial of the bug, it's just that out of
the blocks there is a huge amount of FUD about a problem that the
Lucene guys could have looked at quite a while ago.
Regards,
Kirk
On Aug 3, 2011, at 12:03 AM, Mark Derricutt wrote:
If it was just "a broken lucene" then maybe, but this revealed
bugs in Hotspot itself.
I'd be curious to know how/why these bugs never revealed
themselves earlier in whatever tests hotspot currently has, and
whether or not any test cases have since been added to cover this
now.
Mark
On 3/08/2011, at 9:11 AM, Kirk wrote:
Bottom line, everyone knew it was coming.... the builds were
freely avaliable. I've been using 7 on my Mac for months so....
I don't have much sympathy in this case.
--
You received this message because you are subscribed to the
Google Groups "The Java Posse" group.
To post to this group, send email to [email protected]
<mailto:[email protected]>.
To unsubscribe from this group, send email to
[email protected]
<mailto:[email protected]>.
For more options, visit this group at
http://groups.google.com/group/javaposse?hl=en.
--
You received this message because you are subscribed to the Google
Groups "The Java Posse" group.
To post to this group, send email to [email protected]
<mailto:[email protected]>.
To unsubscribe from this group, send email to
[email protected]
<mailto:javaposse%[email protected]>.
For more options, visit this group at
http://groups.google.com/group/javaposse?hl=en.
--
Robert Casto
www.robertcasto.com <http://www.robertcasto.com>
www.sellerstoolbox.com <http://www.sellerstoolbox.com>
--
You received this message because you are subscribed to the Google
Groups "The Java Posse" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/javaposse?hl=en.
--
You received this message because you are subscribed to the Google Groups "The Java
Posse" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/javaposse?hl=en.
