Luke Gorrie

Seems like everybody is happy with JAM as the suggested hotel so I've booked in there too. I plan to arrive Friday afternoon and hang out with whoever is around...

(JFYI I didn't merge this on my sweep through open PRs just now because based on the discussion it seems like changes are pending.)

@fsfod The duplicate work should only be for instructions after the last snapshot though, right? I am surprised that the side trace took exit 2 but is repeating work that...

@fsfod I tested removing the `global = 0` line from the parent. Then there is no `TBAR` barrier but the child is still repeating all of the same instructions to...

This is a welcome perspective. I'm feeling confident that porting the VM to C is the right approach. Also because @javierguerragiraldez perspective (https://github.com/raptorjit/raptorjit/issues/172#issuecomment-390194890) that the current interpreter has multiple problems...

OK. In either case it seems like the JIT is the place to focus our optimization efforts. The interpreter has to be simple so that we can evolve and maintain...

@fsfod pointed out on Twitter than `ins_AD` in the assembler code is not actually loading `RA` and `RD` out of the instruction and so apparently these registers still hold their...

Question is how I will translate code like this to C now. I suppose that if `FUNCC` wants to access the operands of its caller instruction it could locate them...

Could also be that I should consider aligning my C VM more closely with the asm VM e.g. allowing the predecessor instruction to perform the decoding of the following instruction....

Now I see the code where I overlooked this detail about the `D` register in calls. Here is the standard `ins_next` macro that instructions use to decode the following instruction...