Perf regression since b72

[Marcus Lagergren]

At least for us, the InlineSmallCode workaround pretty much compensated entirely for the loss of tiered. This may not be the case for you, but I hope so. Benchmarking with tiered is very "noisy" with lots of standard deviation.

/M

On Apr 1, 2013, at 9:23 PM, Christian Thalinger <christian.thalinger at oracle.com> wrote:

> 
> On Mar 30, 2013, at 1:56 AM, Charles Oliver Nutter <headius at headius.com> wrote:
> 
>> I've been fiddling about with performance a bit again recently, and have noticed a perf degradation since b72. I mentioned this to the Nashorn guys and Marcus discovered that InlineSmallCode=2000 helped them get back to b72 performance. I can confirm this on JRuby as well, but in any case it seems that something has regressed.
>> 
>> Here's some numbers with JRuby. Numbers are for b72, hotspot-comp, and hotspot-comp with InlineSmallCode=2000. You can see that current hotspot-comp builds do not perform as well as b72 unless that flag is passed.
>> 
>> https://gist.github.com/headius/de7f99b52847c2436ee4
>> 
>> I have not yet started to explore the inlining or assembly output, but I wanted to confirm that others are seeing this degradation.
>> 
>> My build of hotspot-comp is current.
>> 
>> I do have some benchmarks that look fine without the additional flag (neural_net, for example), so I'm confused what's different in the degraded cases.
> 
> I'm not sure if Marcus talked to you already about this but there are two fixes that caused this regression:
> 
> 1.  http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8007439
> 
> which was fixed by:
> 
> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8010222
> 
> in HS25-b24.
> 
> 2.  We turned off tiered compilation:
> 
> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8005811
> 
> The InlineSmallCode=2000 is a way to work around the tiered slowdown.
> 
> -- Chris
> 
>> 
>> - Charlie
>> _______________________________________________
>> mlvm-dev mailing list
>> mlvm-dev at openjdk.java.net
>> http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev
> 
> _______________________________________________
> mlvm-dev mailing list
> mlvm-dev at openjdk.java.net
> http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/mlvm-dev/attachments/20130401/ae0dda28/attachment.html