Return to README CVS log

Up to [local] / OpenXM_contrib / gmp / mpn / x86

Annotation of OpenXM_contrib/gmp/mpn/x86/README, Revision 1.1.1.1

1.1       maekawa     1:
                      2:                       X86 MPN SUBROUTINES
                      3:
                      4:
                      5: This directory contains mpn functions for various 80x86 chips.
                      6:
                      7:
                      8: CODE ORGANIZATION
                      9:
                     10:        x86              i386, i486, generic
                     11:        x86/pentium      Intel Pentium (P5, P54)
                     12:        x86/pentium/mmx  Intel Pentium with MMX (P55)
                     13:        x86/p6           Intel Pentium Pro
                     14:        x86/p6/mmx       Intel Pentium II, III
                     15:        x86/p6/p3mmx     Intel Pentium III
                     16:        x86/k6           AMD K6, K6-2, K6-3
                     17:        x86/k6/mmx
                     18:        x86/k6/k62mmx    AMD K6-2
                     19:        x86/k7           AMD Athlon
                     20:        x86/k7/mmx
                     21:
                     22:
                     23: The x86 directory is also the main support for P6 at the moment, and
                     24: is something of a blended style, meant to be reasonable on all x86s.
                     25:
                     26:
                     27:
                     28: STATUS
                     29:
                     30: The code is well-optimized for AMD and Intel chips, but not so well
                     31: optimized for Cyrix chips.
                     32:
                     33:
                     34:
                     35: RELEVANT OPTIMIZATION ISSUES
                     36:
                     37: For implementations with slow double shift instructions (SHLD and
                     38: SHRD), it might be better to mimic their operation with SHL+SHR+OR.
                     39: (M2 is likely to benefit from that, but not Pentium due to its slow
                     40: plain SHL and SHR.)