Benchmark and test for htodw algos.

[KeepingRealBusy]

Alex,

Here are my AMD timings:

Code Select Expand


AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ (SSE3)



47      cycles for Fast version
22      cycles for Fast version under AMD
51      cycles for Small 1
53      cycles for Small 2
59      cycles for Small 3
50      cycles for Small 3.1
39      cycles for Small 4
15      cycles for MMX 1
27      cycles for MMX 2
24      cycles for SSE1

Other's Versions:
47      cycles for Axhex2dw improved by Hutch (1)
39      cycles for Axhex2dw improved by Hutch (2)

15      cycles for Lingo's SSE version
13      cycles for Lingo's BIG integer version
23      cycles for Jochen's WORD-Indexed version
33      cycles for Dave's version (with minor changes)


35      cycles for Fast version
22      cycles for Fast version under AMD
40      cycles for Small 1
36      cycles for Small 2
39      cycles for Small 3
67      cycles for Small 3.1
83      cycles for Small 4
20      cycles for MMX 1
27      cycles for MMX 2
24      cycles for SSE1

Other's Versions:
35      cycles for Axhex2dw improved by Hutch (1)
39      cycles for Axhex2dw improved by Hutch (2)

15      cycles for Lingo's SSE version
13      cycles for Lingo's BIG integer version
23      cycles for Jochen's WORD-Indexed version
33      cycles for Dave's version (with minor changes)


58      cycles for Fast version
39      cycles for Fast version under AMD
50      cycles for Small 1
40      cycles for Small 2
52      cycles for Small 3
50      cycles for Small 3.1
39      cycles for Small 4
16      cycles for MMX 1
27      cycles for MMX 2
24      cycles for SSE1

Other's Versions:
44      cycles for Axhex2dw improved by Hutch (1)
39      cycles for Axhex2dw improved by Hutch (2)

15      cycles for Lingo's SSE version
13      cycles for Lingo's BIG integer version
35      cycles for Jochen's WORD-Indexed version
33      cycles for Dave's version (with minor changes)

==========
Codesizes:
Axhex2dw_Unrolled:      396
Axhex2dw_Unrolled_AMD:  396
Axhex2dw1 - 1:  70
Axhex2dw2 - 2:  69
Axhex2dw3 - 3:  57
Axhex2dw3_1 - 3.1:      56
Axhex2dw3 - 4:  61
Axhex2dw_MMX:   128
Axhex2dw_MMX2:  160
Axhex2dw_SSE:   160
Alex_Short_Hutch:       59
Axhex2dw_Hutch2:        54
Hex2dwLingoSSE: 160
lingo_htodw:    1950
ax_jj_htodw:    174
krbhtodw:       547
--- ok ---