Improvements to powerpc32 asm code
Mark Rodenkirch
mrodenkirch@wi.rr.com
Mon, 2 Jun 2003 20:19:43 -0500
On Monday, June 2, 2003, at 07:55 PM, Kevin Ryde wrote:
> Mark Rodenkirch <mrodenkirch@wi.rr.com> writes:
>>
>> mpn_add_n mpn_add_n_new
>> 1 21.1270 #10.0658
>> 2 13.0883 #7.0432
>> 3 10.0633 #5.7040
>
> I take it this is -C. -CD is probably best for looking at cycles/limb
> loop speeds, since it cancels out fixed overheads. A plain -c can be
> used to look at total times on small sizes.
Yes, that was -C. Here are the -CD results if you are interested:
1 (21.1270) (#10.0622)
2 5.0496 #4.0293
3 4.0124 #3.0227
4 #4.0166 8.0599
5 #4.0338 5.0142
6 4.0311 #4.0187
7 #4.0071 4.0165
8 4.0611 #0.0051
9 #3.9867 5.0392
10 #4.0239 4.0240
11 4.0514 #4.0152
12 4.0084 #0.0511
13 #4.0405 4.9908
14 #4.0185 4.0358
15 4.0095 #4.0084
16 4.0664 #0.0258
17 #4.0111 5.0238
18 #4.0169 4.0433
19 4.0598 #4.0142
20 3.9686 #0.0000
190 4.5698 #4.1600
191 3.8153 #3.7427
192 3.8598 #0.0938
193 #3.9814 4.8766
194 4.6123 #3.7899
195 #3.5529 4.3497
196 4.1587 #-0.0851
197 #4.2556 5.4088
198 3.7425 #3.3844
199 #3.8848 4.2380
200 4.4803 #0.0000
1990 7.2464 #5.8824
1991 7.2464 #6.5911
1992 7.2464 #-0.7087
1993 #0.0000 5.8824
1994 #0.0000 0.0000
1995 7.9923 #5.8824
1996 6.5004 #0.4902
1997 #1.7305 5.9524
1998 #5.5159 5.9524
1999 #0.0000 0.0000
2000 7.2464 #0.0000
I don't fully comprehend these values compared to -C, but I hope you
find them useful.
--Mark