←back to thread

Implementation of a Java Processor on a FPGA (2016)

(mavmatrix.uta.edu)
56 points mghackerlady | 2 comments | 20 Nov 25 06:40 UTC | HN request time: 0s | source
Show context
larsbrinkhoff ◴[20 Nov 25 07:35 UTC] No.45989967[source]▶
What happened to

1. Sun's JavaStation, 2. ARM's Jazelle, ??? 3. Profit!

replies(4): >>45990000 #>>45990288 #>>45990667 #>>45993727 #
dehrmann ◴[20 Nov 25 07:40 UTC] No.45990000[source]▶
It's more like JITs got good.
replies(1): >>45990334 #
ck45 ◴[20 Nov 25 08:33 UTC] No.45990334[source]▶
I never understood why AOT never took off for Java. The write once run anywhere quickly faded as an argument, the number of platforms that a software package needs to support is rather small.
replies(5): >>45990420 #>>45990438 #>>45990583 #>>45990837 #>>45994530 #
1. gf000 ◴[20 Nov 25 08:44 UTC] No.45990420{3}[source]▶
Well, one aspect is how dynamic the platform is.

It simply defaults to an open world where you could just load a class from any source at any time to subclass something, or straight up apply some transformation to classes as they load via instrumentation. And defaults matter, so AOT compilation is not completely trivial (though it's not too bad either with GraalVM's native image, given that the framework you use (if any) supports it).

Meanwhile most "AOT-first" languages assume a closed-world where everything "that could ever exist" is already known fully.

replies(1): >>45990450 #
2. pjmlp ◴[20 Nov 25 08:48 UTC] No.45990450[source]▶
Except when they support dynamic linking they pay the indirect call cost that JITs can remove.