Allocations profiler

[Aleksey Shipilev]

On 26.02.2015 22:56, Matt Warren wrote:
>> * I am not sure how they are able to pull off low-overhead profiling like
> *>* that.*
> 
> If the question is talking about the same post
> (http://blog.matttproud.com/2014/09/go-testing-patterns-benchmark-setup-and.html),
> it seems like the memory usage is something provided by the go runtime
> itself.

So, we do the lightweight GC churn esitmates with "-prof gc" now,
starting from JMH 1.8. It is not allocation-site accurate, but without a
significant overhead, and can be normalized per benchmark op:
 http://mail.openjdk.java.net/pipermail/jmh-dev/2015-April/001785.html

Thanks,
-Aleksey