Dan Sugalski <[EMAIL PROTECTED]> wrote:Optimized for speed, at least.
I'm finding that large subs seem to give imcc headaches. I'm not sure if it's O(n^2) or O(2^n) headaches, but definitely issues.
Live analysis and register allocation are the main problems. The former is O(n_lines * n_vars). The latter goes horribly slow in the case of spilling. Please run your program with the "-v" switch, you should see some statistics including spilled variables count.
Oh, yeah, lots of spilling. In a hacked up version of the PIR I see 50 spills in the main routine. Not a good thing, I think, and it takes forever even with Parrot built with optimizations. (Which does make quite a difference, though not enough)
--
Dan
--------------------------------------"it's like this"------------------- Dan Sugalski even samurai [EMAIL PROTECTED] have teddy bears and even teddy bears get drunk