On Wed, Jul 4, 2012 at 11:29 AM, Michael Sioutis <papito.dit@gmail.com> wrote:
Thank you all very much!
Of course I don't want to present trace-based JITs as the best choice, but leave some hints on the "possible" advantages that it "possibly" offers, just to put an additional comment on my graphs.
So it has come down to this: Trace-based JIT compilation allows more specialized optimizations on code and is potentially more well suited for modern CPUs \cite{PyPy,Dynamo} ... here we also see the possible advantages of tracing JIT since ... scalable ... robust.
So I think I'm not lying to anyone who reads it :)
Michael, Personally, I would not cite Dynamo to assert anything about the benefits of tracing JITs. Some of Dynamo's success was due to the quirkiness of the HP PA-RISC architecture, which also certainly would not be a good reference for modern CPUs. Also, I believe that Firefox, Chrome and Safari JavaScript JITs now are all method-based. The Meta-tracing JITs is another design feature of PyPy and provides some early optimization for dynamic languages. But as far as tracing JITs in general, the "advantage" may be more about the JIT design decisions that a tracing JIT encourages versus the default design of a method JIT. One really needs to consider a tracing JIT versus a method JIT with extensive profiling, specialization, etc. In other words, it's not a dichotomy, but a distribution. - David