This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: gcc/resource.c mark_target_live_regs Question


>>>>> "Brad" == Brad Lucier <lucier@math.purdue.edu> writes:

    Brad> I'm all for compile-time performance, especially in places
    Brad> where it takes > 10 times longer than it should take (or
    Brad> than it took earlier).  But I'm most definitely *not* in
    Brad> favor of achieving compile-time performance by using
    Brad> quadratic (or worse) algorithms and then putting artificial
    Brad> limiters on the amount of input data considered.

Nobody disagrees with you, in the abstract.

The point is that we don't have a choice.  

Right now, we have quadratic algorithms, and no easy fix.  That means
that we need limiters to achieve something reasonable in the short
term.  

I'm sure that if you (or someone else) implements a more efficient
delay slot filler, it will be adopted quickly.

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]