Hacker News Reader

Comment by polskibus | Hacker News Reader

I wonder what is the level of .NET's CLR optimization in comparison to this new optimizer for C++. I hope they will reuse some of those ideas to make .NET better.

View on HN

The optimizer will be ported to .NET Native soon, it should help code quality quite a lot. The JIT is a completely different matter though: it's a different project, done by a different team - copy-paste of the source code is definitely not going to work. JIT compilers also have different priorities, unlikely they would port all parts of a large optimizer.

Hopefully, they will just recompile the binaries, and the CLR will just get faster...

That's not really the way it works. The CLR compiles .NET code to native machine code (either at runtime or ahead of time), which means you need to implement these optimizations at the code generation layer of the CLR. You can make the CLR implementation itself faster, but that's negligible compared to the run time of the generated code.

Are you saying the "code generation layer" of the CLR isn't itself written in c++?

All of this stuff trickles down without a doubt. I realize it isn't going to magically optimize the bytecode, however, the thing that runs the bytecode will be faster.

The CLR doesn't have any bytecode interpreter that could get faster with this change. It's a JIT compiler that compiles the bytecode to native code.

Interpreter or not, the JIT compiler will certainly get faster with the c++ optimizations. The libraries that are linked against will get faster. The underlying OS components will get faster. Since .NET assemblies are mostly just glue code between framework code, which is mostly c++, everything should get a speedup out of this.

Reply on news.ycombinator.com