Re: Instruction scheduling: loop unrolling

jacob@jacob.remcomp.fr (Jacob Navia)
15 Jun 1997 22:40:32 -0400

          From comp.compilers

Related articles
Re: Instruction scheduling with gcc on alpha Robert.Harley@inria.fr (1997-06-13)
Re: Instruction scheduling: loop unrolling jacob@jacob.remcomp.fr (1997-06-15)
| List of all articles for this month |

From: jacob@jacob.remcomp.fr (Jacob Navia)
Newsgroups: comp.compilers
Date: 15 Jun 1997 22:40:32 -0400
Organization: Compilers Central
References: 97-06-052
Keywords: optimize

In lcc-win32 you could just do (when you stick to 32 bit numbers)
pmul(vector,constant,n); where vector is a vector of 32 bits integers,
constant is a 32 bit number, and n is the number of elements of
'vector'.


I generate for that construct the new 'mmx' instruction for doing two
32 bits adds/multiplies in parallel using the mmx registers of the
pentium processor.


My compiler is the only one to offer a high level interface (in C) for
those new mmx instructions. It is freely available from:
http://www.remcomp.com/lcc-win32. The sources of the compiler are
included.
--
Jacob Navia Logiciels/Informatique
41 rue Maurice Ravel Tel (1) 48.23.51.44
93430 Villetaneuse Fax (1) 48.23.95.39
France
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.