Pentium M tests
Roberto Bagnara
bagnara at cs.unipr.it
Thu Mar 3 11:38:36 CET 2005
Torbjorn Granlund wrote:
> The GMP project needs help with porting to Pentium M
> (aka Centrino, aka Dothan).
>
> If you have some sort of Pentium M, please help.
>
> 1. Run config.guess from gmp 4.1.4 and send us the output.
$ /usr/local/distrib/gmp-4.1.4/config.guess
pentium3-pc-linux-gnu
> 2. Compile this C code, run it and send us the output.
$ gcc -W -Wall cpuid.c
[roberto at quark tmp]$ a.out
GenuineIntel
Highest value understood by CPUID: 2
Family 6
Model 13
Stepping 6
FPU VME DE PSE TSC MSR MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT CLFSH DS ACPI MMX FXSR SSE SSE2 SS TM PBE
CPUID arg = 1:
eax 000006d6
ebx 00000816
ecx 00000180
edx afe9fbbf
CPUID arg = 2:
eax 02b3b001
ebx 000000f0
ecx 00000000
edx 2c04307d
Intel Pentium III
" Intel(R) Pentium(R) M processor 1.70GHz"
> 3. Build and test the gmp 4.1.4 sources with this script. There
> are some instructions in the beginning of the script.
$ export SRCDIR=/usr/local/distrib/gmp-4.1.4
$ sh compare-asm-varants.sh
******************************************************************************
Set SRCDIR to point to the gmp source directory (defaults to ./)
Set CC to your compiler, including any ABI or ISA options (defaults to gcc)
Set MPNDIR to point to the proper mpn subdirectory (e.g., x86, sparc64, alpha)
(defaults to x86)
Ignore errors for some tests; they are unavoidable
******************************************************************************
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 24.0133
2 21.6746
3 15.6323
4 22.4086
5 8.9166
6 8.1562
8 7.2126
11 6.6069
15 6.2140
20 5.8131
27 5.6838
37 5.4426
51 5.2966
71 5.1676
99 5.0904
138 5.0562
193 5.0039
270 4.8955
377 4.9005
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 39.8225
2 22.7252
3 17.3602
4 11.7425
5 11.8441
6 10.5765
8 8.7989
11 8.6396
15 7.6284
20 6.6474
27 6.7267
37 6.2955
51 6.1322
71 5.7512
99 5.5601
138 5.5522
193 5.5106
270 5.3698
377 5.4849
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/sse2/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 20.4447
2 12.9808
3 10.4234
4 9.1237
5 8.2985
6 7.7369
8 6.9860
11 6.8145
15 6.1091
20 5.7581
27 5.4362
37 5.5018
51 5.0942
71 5.1737
99 5.0597
138 4.9382
193 4.8575
270 4.8389
377 4.8006
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 23.9978
2 15.1347
3 11.9868
4 10.5193
5 9.6379
6 8.8045
8 7.7324
11 6.9296
15 6.3272
20 6.2185
27 5.8143
37 5.5038
51 5.2549
71 5.2103
99 5.0528
138 4.9977
193 4.9512
270 4.9300
377 4.8820
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/mmx/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 13.0000
2 24.4844
3 20.4472
4 20.5466
5 17.1950
6 18.3278
8 17.1943
11 15.3362
15 14.9264
20 15.1286
27 14.4299
37 14.2625
51 14.1336
71 14.0691
99 14.0164
138 13.9928
193 13.9495
270 13.8815
377 13.7745
operation=mul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/mul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_mul_1.3
1 15.6527
2 22.9749
3 16.6773
4 15.8623
5 13.9145
6 14.3221
8 13.7995
11 12.2681
15 12.2113
20 12.2455
27 11.7712
37 11.5748
51 11.5077
71 11.3722
99 11.3636
138 11.4317
193 11.3451
270 11.3056
377 11.2378
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 40.8545
2 25.5663
3 18.7622
4 13.8159
5 13.5141
6 12.1060
8 10.4807
11 9.9441
15 9.2018
20 8.1746
27 8.0960
37 7.6970
51 7.4237
71 7.3125
99 7.2165
138 7.0350
193 7.0299
270 6.8648
377 6.8654
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 22.7104
2 16.2737
3 13.9847
4 12.8451
5 11.6339
6 10.7743
8 9.5401
11 8.6066
15 7.9826
20 7.5993
27 7.1724
37 6.9235
51 6.6961
71 6.6144
99 6.4765
138 6.4270
193 6.3615
270 6.3185
377 6.2955
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 27.1787
2 20.0365
3 17.8829
4 16.8666
5 16.2903
6 11.0576
8 9.9520
11 8.0000
15 7.6023
20 7.0315
27 6.6111
37 6.2757
51 6.0913
71 6.0141
99 5.9351
138 5.7821
193 5.7238
270 5.6690
377 5.6737
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/sse2/addmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 21.4734
2 14.7003
3 12.1566
4 11.0111
5 10.1495
6 9.5910
8 8.9512
11 8.4320
15 8.0299
20 7.7868
27 7.5912
37 7.3993
51 7.3016
71 7.5929
99 7.3771
138 7.3007
193 7.1813
270 7.1549
377 7.1160
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 14.7521
2 17.0357
3 13.9667
4 12.6458
5 11.7337
6 11.0980
8 9.6268
11 8.6178
15 7.9387
20 7.6848
27 7.1717
37 7.0045
51 6.7457
71 6.6262
99 6.5330
138 6.4275
193 6.3385
270 6.3333
377 6.2780
operation=addmul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_addmul_1.3
1 43.8880
2 30.1686
3 28.6066
4 25.5485
5 25.5286
6 24.0090
8 23.2177
11 22.9958
15 22.4303
20 21.8340
27 21.7485
37 21.5193
51 21.3209
71 21.2233
99 21.1434
138 21.0290
193 20.9793
270 21.0037
377 20.8408
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 40.8545
2 25.5163
3 18.7309
4 13.7889
5 13.5162
6 12.1079
8 10.4807
11 9.9487
15 9.1931
20 8.1746
27 8.0960
37 7.6970
51 7.4199
71 7.3125
99 7.2165
138 7.0350
193 7.0486
270 6.8648
377 6.8654
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 22.6889
2 16.2685
3 13.9795
4 12.8451
5 11.6339
6 10.7743
8 9.5401
11 8.6066
15 7.9826
20 7.5993
27 7.1724
37 6.9235
51 6.6961
71 6.6144
99 6.4765
138 6.4270
193 6.3615
270 6.3185
377 6.3130
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 27.2382
2 23.5085
3 17.8829
4 16.9228
5 16.2556
6 11.0576
8 9.9764
11 7.9884
15 7.5940
20 7.0329
27 6.6221
37 6.2757
51 6.1042
71 6.0141
99 5.9176
138 5.7754
193 5.7238
270 5.6727
377 5.6960
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/sse2/submul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 26.6268
2 16.6483
3 13.7256
4 12.2100
5 11.4125
6 10.8505
8 10.1726
11 9.5801
15 9.1896
20 8.8908
27 8.6924
37 8.5521
51 8.3816
71 8.6322
99 8.5276
138 8.3522
193 8.2650
270 8.1867
377 8.1406
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 14.7514
2 17.0418
3 13.9667
4 12.6458
5 11.7337
6 11.0986
8 9.6268
11 8.6155
15 7.9386
20 7.6848
27 7.1717
37 7.0045
51 6.7457
71 6.6080
99 6.5330
138 6.4275
193 6.3385
270 6.3333
377 6.2780
operation=submul_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/aorsmul_1.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_submul_1.3
1 43.8880
2 30.1647
3 28.6016
4 25.5485
5 25.5286
6 24.0090
8 23.2177
11 22.9958
15 22.4303
20 21.8340
27 21.7485
37 21.5193
51 21.3209
71 21.2233
99 21.1434
138 21.0290
193 20.9793
270 20.9704
377 20.8408
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 32.6696
2 20.9313
3 17.7340
4 16.1115
5 15.1363
6 14.4873
8 7.6208
11 6.0440
15 4.7953
20 4.9519
27 4.2739
37 3.8685
51 3.5955
71 3.4053
99 3.2433
138 3.1957
193 3.0510
270 2.9981
377 2.9952
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 20.2431
2 19.1698
3 12.5636
4 11.7254
5 10.4171
6 9.0263
8 8.7009
11 7.9060
15 7.3624
20 6.8994
27 6.6893
37 6.4258
51 6.2993
71 6.1763
99 6.0829
138 5.9969
193 5.9658
270 5.9741
377 5.9469
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/sse2/add_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 22.6952
2 14.9958
3 11.5437
4 10.0454
5 9.1237
6 8.3379
8 7.4054
11 6.8033
15 6.3621
20 6.0316
27 5.7963
37 5.6481
51 5.4625
71 5.5166
99 5.3963
138 5.3140
193 5.2617
270 5.2069
377 5.1720
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 32.6696
2 20.9313
3 17.6973
4 16.0804
5 15.1363
6 14.4900
8 7.6208
11 6.0440
15 4.7962
20 4.9519
27 4.2746
37 3.8685
51 3.5955
71 3.4053
99 3.2433
138 3.1957
193 3.0510
270 2.9981
377 2.9952
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 34.7333
2 25.5069
3 19.4127
4 17.8996
5 16.7623
6 15.1785
8 14.5536
11 7.7174
15 8.3135
20 6.2324
27 5.1111
37 5.1839
51 4.3627
71 4.5539
99 3.9466
138 3.7498
193 3.6597
270 3.7471
377 3.5853
operation=add_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_add_n.3
1 41.8702
2 21.9560
3 15.3208
4 12.5068
5 11.0448
6 9.0367
8 6.2562
11 7.1575
15 5.6553
20 5.2636
27 4.8825
37 4.4998
51 4.2624
71 4.0093
99 3.9275
138 3.8575
193 3.7701
270 3.6919
377 3.6767
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 32.6696
2 20.9313
3 17.7340
4 16.1092
5 15.1338
6 14.4819
8 7.6208
11 6.0419
15 4.7953
20 4.9519
27 4.2746
37 3.8685
51 3.5955
71 3.4053
99 3.2433
138 3.1957
193 3.0510
270 2.9981
377 2.9952
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 20.1729
2 19.1730
3 12.5636
4 11.7253
5 10.5187
6 9.1960
8 8.7023
11 7.9977
15 7.4305
20 6.9020
27 6.7286
37 6.4264
51 6.3189
71 6.1904
99 6.0930
138 5.9974
193 5.9663
270 5.9714
377 5.9475
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/sse2/sub_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 22.6653
2 15.0247
3 12.0109
4 10.0502
5 8.9377
6 8.4639
8 7.4054
11 6.7054
15 6.3119
20 6.0698
27 5.7633
37 5.5899
51 5.4485
71 5.3482
99 5.2780
138 5.3140
193 5.2567
270 5.2069
377 5.1746
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 32.6696
2 20.9313
3 17.7319
4 16.1092
5 15.1363
6 14.4873
8 7.6208
11 6.0419
15 4.7953
20 4.9524
27 4.2739
37 3.8697
51 3.5955
71 3.4053
99 3.2433
138 3.1957
193 3.0510
270 2.9981
377 2.9952
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 34.7357
2 25.5093
3 19.4127
4 17.8934
5 16.7593
6 15.1785
8 14.5569
11 7.7174
15 8.3135
20 6.2324
27 5.1111
37 5.1839
51 4.3627
71 4.5539
99 3.9466
138 3.7498
193 3.6597
270 3.7471
377 3.5853
operation=sub_n file=/usr/local/distrib/gmp-4.1.4/mpn/x86/aors_n.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_sub_n.3
1 41.8702
2 21.9560
3 15.3208
4 12.5068
5 11.0448
6 9.0378
8 6.2563
11 7.1575
15 5.6553
20 5.2636
27 4.8838
37 4.4970
51 4.2624
71 4.0072
99 3.9275
138 3.8575
193 3.7748
270 3.6919
377 3.6767
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 16.7009
2 11.0373
3 8.1723
4 6.4615
5 5.6241
6 5.2414
8 4.2915
11 3.8112
15 3.4727
20 3.2173
27 3.0698
37 2.9296
51 2.8256
71 2.7462
99 2.6901
138 2.7359
193 2.6862
270 2.6433
377 2.6160
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/mmx/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 14.8540
2 13.2814
3 9.6254
4 7.8298
5 7.3273
6 6.3879
8 5.2371
11 4.3636
15 3.6847
20 3.1658
27 2.8415
37 2.5423
51 2.3453
71 2.1872
99 2.0713
138 1.9894
193 1.9296
270 1.9413
377 1.8974
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/mmx/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 13.6248
2 11.4359
3 8.8066
4 7.0868
5 6.2329
6 5.6353
8 4.7635
11 4.5081
15 3.8891
20 3.6833
27 3.5215
37 3.3980
51 3.3018
71 3.0897
99 2.9978
138 2.8742
193 2.8326
270 2.7677
377 2.7492
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/lshift.asm
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/lshift.asm: 128: warning, using emms in place of femms, use for testing only
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/lshift.asm: 269: warning, using emms in place of femms, use for testing only
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/lshift.asm: 282: warning, using emms in place of femms, use for testing only
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 13.6195
2 14.5889
3 14.9867
4 11.3548
5 12.5291
6 9.3886
8 8.8287
11 3.7372
15 3.1834
20 2.7512
27 2.5784
37 2.3223
51 2.1995
71 2.0763
99 1.9933
138 1.9291
193 1.8874
270 1.8990
377 1.8687
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/mmx/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 14.8540
2 13.2772
3 9.6254
4 7.8307
5 7.3284
6 6.3864
8 5.2371
11 4.3636
15 3.6774
20 3.1658
27 2.8420
37 2.5433
51 2.3453
71 2.1879
99 2.0721
138 1.9894
193 1.9296
270 1.9413
377 1.8974
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/mmx/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 14.3823
2 13.2828
3 9.2822
4 7.4289
5 6.5351
6 5.9566
8 5.2333
11 5.0274
15 4.0691
20 3.3417
27 2.8641
37 2.4877
51 2.2045
71 1.9874
99 1.8560
138 1.7424
193 1.6763
270 1.6153
377 1.5664
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 38.8085
2 20.9370
3 14.6427
4 12.1244
5 10.4161
6 9.0205
8 7.9138
11 5.9513
15 4.9799
20 4.1415
27 3.5360
37 3.2879
51 2.8829
71 2.8028
99 2.5635
138 2.4245
193 2.3252
270 2.3292
377 2.2595
operation=lshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/mmx/lshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_lshift.3
1 14.8413
2 13.2814
3 9.6254
4 7.8298
5 7.3273
6 6.3864
8 5.2371
11 4.3636
15 3.6847
20 3.1658
27 2.8415
37 2.5423
51 2.3453
71 2.1872
99 2.0713
138 1.9894
193 1.9296
270 1.9413
377 1.8974
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 18.6968
2 12.0109
3 8.5563
4 6.6538
5 5.8230
6 5.1587
8 4.3911
11 3.8388
15 3.4800
20 3.2173
27 3.0698
37 2.9296
51 2.8265
71 2.7462
99 2.6901
138 2.7359
193 2.6862
270 2.6433
377 2.6163
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/mmx/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 15.5210
2 13.1630
3 9.9796
4 8.2086
5 7.0935
6 6.2205
8 5.2328
11 4.4558
15 3.5411
20 3.1141
27 2.7660
37 2.4872
51 2.3055
71 2.1590
99 2.0507
138 1.9734
193 1.9176
270 1.9303
377 1.8926
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/mmx/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 13.6248
2 11.4215
3 8.8200
4 7.1660
5 6.1892
6 5.4849
8 4.7225
11 4.1776
15 3.8884
20 3.6830
27 3.5212
37 3.3986
51 3.3615
71 3.2022
99 3.0788
138 2.8952
193 2.8694
270 2.7820
377 2.7565
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/rshift.asm
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/rshift.asm: 133: warning, using emms in place of femms, use for testing only
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/rshift.asm: 268: warning, using emms in place of femms, use for testing only
/usr/local/distrib/gmp-4.1.4/mpn/x86/k6/k62mmx/rshift.asm: 281: warning, using emms in place of femms, use for testing only
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 13.6281
2 12.2455
3 9.1574
4 7.4892
5 6.5534
6 5.6766
8 4.6660
11 3.9345
15 3.2726
20 2.7763
27 2.6055
37 2.3575
51 2.2153
71 2.0879
99 2.0004
138 1.9362
193 1.8976
270 1.8961
377 1.8770
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium4/mmx/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 15.5210
2 13.1586
3 9.9809
4 8.2042
5 7.0935
6 6.2205
8 5.2328
11 4.4558
15 3.5411
20 3.1141
27 2.7660
37 2.4879
51 2.3055
71 2.1590
99 2.0507
138 1.9734
193 1.9176
270 1.9303
377 1.8926
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/mmx/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 14.3791
2 13.3264
3 9.4378
4 7.5808
5 6.5367
6 5.9582
8 5.2333
11 4.8435
15 3.8894
20 3.2750
27 2.8082
37 2.4544
51 2.2041
71 1.9859
99 1.8250
138 1.7171
193 1.6500
270 1.5910
377 1.5609
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 35.7410
2 18.8931
3 14.5159
4 11.8680
5 10.0083
6 8.8494
8 7.6576
11 5.6518
15 4.9965
20 3.9895
27 3.4174
37 3.1594
51 2.8082
71 2.7585
99 2.5142
138 2.4014
193 2.3258
270 2.3130
377 2.2442
operation=rshift file=/usr/local/distrib/gmp-4.1.4/mpn/x86/pentium/mmx/rshift.asm
overhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_rshift.3
1 15.5227
2 13.1602
3 9.9809
4 8.2042
5 7.0935
6 6.2205
8 5.2328
11 4.4558
15 3.5411
20 3.1141
27 2.7660
37 2.4879
51 2.3055
71 2.1590
99 2.0507
138 1.9734
193 1.9176
270 1.9303
377 1.8926
operation=divrem_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/p6/mmx/divrem_1.asmoverhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_divrem_1.3
1 55.0457
2 47.1435
3 45.0556
4 44.1411
5 40.0259
6 38.0426
8 34.3719
11 31.5483
15 30.0587
20 28.3900
27 27.6765
37 26.7072
51 26.0937
71 25.7653
99 25.4061
138 25.2633
193 24.9102
270 24.8296
377 24.6220
operation=divrem_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/divrem_1.asm
.libs/libgmp.a(divrem_1.o)(.text+0x70): In function `__gmpn_divrem_1c':
/tmp/mpn/tmp-divrem_1.s:147: multiple definition of `__gmpn_divrem_1c'
/tmp/cc2gUsak.o(.text+0x0): first defined here
/usr/bin/ld: Warning: size of symbol `__gmpn_divrem_1c' changed from 38 in /tmp/cc2gUsak.o to 49 in .libs/libgmp.a(divrem_1.o)
.libs/libgmp.a(divrem_1.o)(.text+0xa1): In function `__gmpn_divrem_1':
/tmp/mpn/tmp-divrem_1.s:178: multiple definition of `__gmpn_divrem_1'
/tmp/cc2gUsak.o(.text+0x26): first defined here
/usr/bin/ld: Warning: size of symbol `__gmpn_divrem_1' changed from 104 in /tmp/cc2gUsak.o to 631 in .libs/libgmp.a(divrem_1.o)
collect2: ld returned 1 exit status
compare-asm-varants.sh: line 62: ./a.out: No such file or directory
operation=divrem_1 file=/usr/local/distrib/gmp-4.1.4/mpn/x86/k7/mmx/divrem_1.asmoverhead 6.13 cycles, precision 10000 units of 5.88e-10 secs, CPU freq 1701.02 MHz
mpn_divrem_1.3
1 55.5025
2 47.3750
3 51.7042
4 45.3856
5 41.6000
6 39.2319
8 36.0304
11 33.6182
15 31.7913
20 30.5250
27 29.8360
37 28.8575
51 28.3922
71 27.6878
99 27.5000
138 27.0217
193 26.8722
270 27.1574
377 27.0822
--
Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy
http://www.cs.unipr.it/~bagnara/
mailto:bagnara at cs.unipr.it
More information about the gmp-devel
mailing list