diff options
author | Torbjorn Granlund <tege@gmplib.org> | 2010-01-23 23:37:29 +0100 |
---|---|---|
committer | Torbjorn Granlund <tege@gmplib.org> | 2010-01-23 23:37:29 +0100 |
commit | 57a693b6f8862e6203566a8f72e7f576d2ec69e2 (patch) | |
tree | f8baea10b08a15cae44e4f0edebcd95a0c6c6782 | |
parent | 89039de0c1a819eb025f87916bbedae8819b0f3d (diff) | |
download | gmp-57a693b6f8862e6203566a8f72e7f576d2ec69e2.tar.gz |
Add FFT_TABLE3 tables for more machines.
-rw-r--r-- | mpn/alpha/ev6/gmp-mparam.h | 96 | ||||
-rw-r--r-- | mpn/pa32/hppa2_0/gmp-mparam.h | 71 | ||||
-rw-r--r-- | mpn/powerpc32/gmp-mparam.h | 91 | ||||
-rw-r--r-- | mpn/powerpc64/mode64/p5/gmp-mparam.h | 114 | ||||
-rw-r--r-- | mpn/x86/k6/gmp-mparam.h | 71 | ||||
-rw-r--r-- | mpn/x86/p6/sse2/gmp-mparam.h | 90 | ||||
-rw-r--r-- | mpn/x86/pentium/mmx/gmp-mparam.h | 95 | ||||
-rw-r--r-- | mpn/x86_64/gmp-mparam.h | 8 | ||||
-rw-r--r-- | mpn/x86_64/pentium4/gmp-mparam.h | 224 |
9 files changed, 727 insertions, 133 deletions
diff --git a/mpn/alpha/ev6/gmp-mparam.h b/mpn/alpha/ev6/gmp-mparam.h index 259f4cccc..e047f523d 100644 --- a/mpn/alpha/ev6/gmp-mparam.h +++ b/mpn/alpha/ev6/gmp-mparam.h @@ -61,13 +61,91 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 20 #define SQRMOD_BNM1_THRESHOLD 23 -#define MUL_FFT_TABLE { 368, 864, 1984, 3840, 9216, 20480, 81920, 327680, 0 } -#define MUL_FFT_MODF_THRESHOLD 464 -#define MUL_FFT_THRESHOLD 7808 - -#define SQR_FFT_TABLE { 432, 864, 1856, 4864, 9216, 20480, 81920, 327680, 0 } -#define SQR_FFT_MODF_THRESHOLD 408 -#define SQR_FFT_THRESHOLD 4736 +#define MUL_FFT_MODF_THRESHOLD 480 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 480, 5}, { 18, 6}, { 10, 5}, { 21, 6}, \ + { 11, 5}, { 23, 6}, { 12, 5}, { 25, 6}, \ + { 19, 7}, { 10, 6}, { 25, 7}, { 13, 6}, \ + { 27, 7}, { 14, 6}, { 29, 7}, { 25, 8}, \ + { 13, 7}, { 29, 8}, { 15, 7}, { 32, 8}, \ + { 17, 7}, { 35, 8}, { 19, 7}, { 39, 8}, \ + { 29, 9}, { 15, 8}, { 37, 9}, { 19, 8}, \ + { 41, 9}, { 23, 8}, { 51, 9}, { 27, 8}, \ + { 55, 9}, { 31, 8}, { 63, 9}, { 35, 8}, \ + { 71, 9}, { 39,10}, { 23, 9}, { 55,10}, \ + { 31, 9}, { 67,10}, { 39, 9}, { 83,10}, \ + { 47, 9}, { 99,10}, { 55,11}, { 31,10}, \ + { 79,11}, { 47,10}, { 103,12}, { 31,11}, \ + { 63,10}, { 135,11}, { 79,10}, { 167,11}, \ + { 95,10}, { 191,11}, { 111,12}, { 63,11}, \ + { 127,10}, { 255,11}, { 143,10}, { 287, 9}, \ + { 575,11}, { 159,10}, { 319,12}, { 95,11}, \ + { 191,10}, { 383,11}, { 207,13}, { 63,12}, \ + { 127,11}, { 255,10}, { 511,11}, { 271,10}, \ + { 543,11}, { 287,10}, { 575,12}, { 159,11}, \ + { 319,10}, { 639,11}, { 351,10}, { 703,12}, \ + { 191,11}, { 383,10}, { 767,11}, { 415,10}, \ + { 831,11}, { 447,13}, { 127,12}, { 255,11}, \ + { 543,12}, { 287,11}, { 575,10}, { 1151,12}, \ + { 319,11}, { 639,12}, { 351,11}, { 703,13}, \ + { 191,12}, { 383,11}, { 767,12}, { 415,11}, \ + { 831,12}, { 447,11}, { 895,14}, { 127,13}, \ + { 255,12}, { 543,11}, { 1087,12}, { 575,11}, \ + { 1151,12}, { 607,13}, { 319,12}, { 671,11}, \ + { 1343,12}, { 703,13}, { 383,12}, { 831,13}, \ + { 447,12}, { 927,14}, { 255,13}, { 511,12}, \ + { 1087,13}, { 575,12}, { 1151,13}, { 639,12}, \ + { 1279,13}, { 703,12}, { 1407,14}, { 383,13}, \ + { 767,15}, { 255,14}, { 511,13}, { 1215,14}, \ + { 639,13}, { 1407,14}, { 767,13}, { 1663,14}, \ + { 895,13}, { 1791,15}, { 32768,16}, { 65536,17}, \ + { 131072,18}, { 262144,19}, { 524288,20}, {1048576,21}, \ + {2097152,22}, {4194304,23}, {8388608,24} } +#define MUL_FFT_TABLE3_SIZE 151 +#define MUL_FFT_THRESHOLD 7552 + +#define SQR_FFT_MODF_THRESHOLD 476 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 476, 5}, { 19, 6}, { 10, 5}, { 23, 6}, \ + { 12, 5}, { 25, 6}, { 27, 7}, { 14, 6}, \ + { 29, 7}, { 28, 8}, { 15, 7}, { 31, 8}, \ + { 29, 9}, { 15, 8}, { 35, 9}, { 19, 8}, \ + { 41, 9}, { 23, 8}, { 49, 9}, { 27,10}, \ + { 15, 9}, { 35, 8}, { 71, 9}, { 39,10}, \ + { 23, 9}, { 51,11}, { 15,10}, { 31, 9}, \ + { 67,10}, { 39, 9}, { 79,10}, { 47, 9}, \ + { 95,10}, { 55,11}, { 31,10}, { 79,11}, \ + { 47,10}, { 103,12}, { 31,11}, { 63,10}, \ + { 135,11}, { 79,10}, { 159, 9}, { 319,11}, \ + { 95,10}, { 191, 9}, { 383,11}, { 111,12}, \ + { 63,11}, { 127,10}, { 255, 9}, { 511,10}, \ + { 271,11}, { 143,10}, { 287, 9}, { 575,10}, \ + { 303,11}, { 159,10}, { 319,12}, { 95,11}, \ + { 191,10}, { 383, 9}, { 767,13}, { 63,12}, \ + { 127,11}, { 255,10}, { 511,11}, { 271,10}, \ + { 543,11}, { 287,10}, { 575,11}, { 303,12}, \ + { 159,11}, { 319,10}, { 639,11}, { 335,10}, \ + { 671,11}, { 351,10}, { 703,11}, { 367,10}, \ + { 735,12}, { 191,11}, { 383,10}, { 767,11}, \ + { 415,10}, { 831,11}, { 447,10}, { 895,13}, \ + { 127,12}, { 255,11}, { 543,12}, { 287,11}, \ + { 575,10}, { 1151,11}, { 607,12}, { 319,11}, \ + { 671,12}, { 351,11}, { 735,13}, { 191,12}, \ + { 383,11}, { 767,12}, { 415,11}, { 831,12}, \ + { 447,11}, { 895,12}, { 479,14}, { 127,13}, \ + { 255,12}, { 575,11}, { 1151,12}, { 607,13}, \ + { 319,12}, { 735,13}, { 383,12}, { 831,13}, \ + { 447,12}, { 959,14}, { 255,13}, { 511,12}, \ + { 1023,13}, { 575,12}, { 1215,13}, { 639,12}, \ + { 1279,13}, { 703,12}, { 1407,14}, { 383,13}, \ + { 767,12}, { 1535,13}, { 831,12}, { 1663,13}, \ + { 895,12}, { 1791,15}, { 255,14}, { 511,13}, \ + { 1215,14}, { 639,13}, { 1407,14}, { 767,13}, \ + { 1663,14}, { 895,13}, { 1791,15}, { 32768,16}, \ + { 65536,17}, { 131072,18}, { 262144,19}, { 524288,20}, \ + {1048576,21}, {2097152,22}, {4194304,23}, {8388608,24} } +#define SQR_FFT_TABLE3_SIZE 152 +#define SQR_FFT_THRESHOLD 5248 #define MULLO_BASECASE_THRESHOLD 0 /* always */ #define MULLO_DC_THRESHOLD 130 @@ -101,7 +179,3 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define GET_STR_PRECOMPUTE_THRESHOLD 19 #define SET_STR_DC_THRESHOLD 3754 #define SET_STR_PRECOMPUTE_THRESHOLD 8097 - -#define MUL_FFT_TABLE2 {{1,4}, {273,5}, {609,6}, {1729,7}, {4097,8}, {8961,9}, {11777,8}, {12545,9}, {13825,10}, {15361,9}, {19969,10}, {23553,9}, {26113,10}, {31745,9}, {35841,10}, {39937,9}, {42497,10}, {48129,9}, {50689,10}, {56321,11}, {63489,10}, {80897,11}, {96257,10}, {105473,12}, {126977,11}, {129025,10}, {143361,11}, {161793,10}, {171009,11}, {227329,12}, {258049,11}, {292865,10}, {310273,11}, {326657,12}, {389121,11}, {391169,10}, {394241,11}, {397313,10}, {408577,11}, {423937,13}, {516097,12}, {520193,11}, {620545,12}, {651265,11}, {751617,12}, {782337,11}, {915457,13}, {1040385,12}, {1044481,11}, {1112065,12}, {1175553,11}, {1243137,12}, {1306625,11}, {1374209,12}, {1437697,11}, {1439745,13}, {1564673,12}, {1568769,11}, {1636353,12}, {1961985,14}, {MP_SIZE_T_MAX, 0}} - -#define SQR_FFT_TABLE2 {{1,4}, {305,5}, {801,6}, {1729,7}, {3713,8}, {3841,7}, {4097,8}, {10497,9}, {11777,8}, {12545,9}, {20481,10}, {23553,9}, {26113,11}, {30721,10}, {31745,9}, {34305,10}, {48641,9}, {50177,10}, {56321,11}, {63489,10}, {80897,11}, {96257,10}, {105473,12}, {126977,11}, {129025,10}, {140289,11}, {227329,12}, {258049,11}, {260097,10}, {261121,11}, {266241,10}, {275457,11}, {292865,10}, {310785,11}, {325633,10}, {326657,12}, {389121,11}, {391169,10}, {408577,11}, {409601,13}, {516097,12}, {520193,11}, {621569,12}, {651265,11}, {751617,12}, {913409,13}, {1040385,12}, {1044481,11}, {1112065,12}, {1175553,11}, {1243137,12}, {1306625,11}, {1374209,12}, {1384449,11}, {1388545,12}, {1425409,11}, {1429505,12}, {1437697,11}, {1505281,13}, {1564673,12}, {1568769,11}, {1636353,12}, {1961985,14}, {MP_SIZE_T_MAX, 0}} diff --git a/mpn/pa32/hppa2_0/gmp-mparam.h b/mpn/pa32/hppa2_0/gmp-mparam.h index f0bda14a2..44543c748 100644 --- a/mpn/pa32/hppa2_0/gmp-mparam.h +++ b/mpn/pa32/hppa2_0/gmp-mparam.h @@ -23,8 +23,6 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ /* 552 MHz PA8600 (gcc61.fsffrance.org) */ -/* Generated by tuneup.c, 2009-11-29, gcc 4.2 */ - #define DIVREM_1_NORM_THRESHOLD 3 #define DIVREM_1_UNNORM_THRESHOLD 4 #define MOD_1_NORM_THRESHOLD 3 @@ -60,17 +58,74 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 13 #define SQRMOD_BNM1_THRESHOLD 15 -#define MUL_FFT_TABLE { 304, 672, 896, 2560, 10240, 24576, 98304, 0 } -#define MUL_FFT_MODF_THRESHOLD 296 +#define MUL_FFT_MODF_THRESHOLD 244 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 244, 5}, { 8, 4}, { 17, 5}, { 13, 6}, \ + { 7, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 13, 7}, { 7, 6}, { 17, 7}, { 9, 6}, \ + { 19, 7}, { 11, 6}, { 24, 7}, { 13, 8}, \ + { 7, 7}, { 19, 8}, { 11, 7}, { 25, 8}, \ + { 15, 7}, { 33, 8}, { 23, 9}, { 15, 8}, \ + { 39, 9}, { 23,10}, { 15, 9}, { 31, 8}, \ + { 67, 9}, { 39, 8}, { 79, 9}, { 47,10}, \ + { 31, 9}, { 71, 8}, { 143, 9}, { 79,10}, \ + { 47,11}, { 31,10}, { 63, 9}, { 127, 8}, \ + { 255, 9}, { 135, 8}, { 271, 9}, { 143,10}, \ + { 79, 9}, { 159, 8}, { 319, 9}, { 175, 8}, \ + { 351,10}, { 95, 9}, { 191, 8}, { 383, 9}, \ + { 207,11}, { 63,10}, { 127, 9}, { 255, 8}, \ + { 511, 9}, { 271,10}, { 143, 9}, { 287, 8}, \ + { 575,10}, { 159, 9}, { 319,10}, { 175, 9}, \ + { 351,11}, { 95,10}, { 191, 9}, { 383,10}, \ + { 207, 9}, { 415,12}, { 63,11}, { 127,10}, \ + { 255, 9}, { 511,10}, { 271, 9}, { 543, 8}, \ + { 1087,10}, { 287, 9}, { 575,10}, { 303,11}, \ + { 159,10}, { 351, 9}, { 703, 8}, { 1407,11}, \ + { 191,10}, { 415, 9}, { 831,11}, { 223, 9}, \ + { 895,10}, { 479,12}, { 127,11}, { 255,10}, \ + { 543, 9}, { 1087,11}, { 287,10}, { 607, 9}, \ + { 1215,11}, { 351,10}, { 703, 9}, { 1407,12}, \ + { 191,11}, { 415,10}, { 831,11}, { 479,13}, \ + { 8192,14}, { 16384,15}, { 32768,16} } +#define MUL_FFT_TABLE3_SIZE 107 #define MUL_FFT_THRESHOLD 2112 -#define SQR_FFT_TABLE { 304, 672, 1152, 2560, 6144, 24576, 98304, 0 } -#define SQR_FFT_MODF_THRESHOLD 232 -#define SQR_FFT_THRESHOLD 1728 +#define SQR_FFT_MODF_THRESHOLD 240 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 240, 5}, { 8, 4}, { 17, 5}, { 19, 6}, \ + { 17, 7}, { 9, 6}, { 20, 7}, { 11, 6}, \ + { 23, 7}, { 13, 8}, { 7, 7}, { 19, 8}, \ + { 11, 7}, { 25, 8}, { 15, 7}, { 33, 8}, \ + { 19, 7}, { 39, 8}, { 23, 9}, { 15, 8}, \ + { 39, 9}, { 23,10}, { 15, 9}, { 31, 8}, \ + { 63, 9}, { 47,10}, { 31, 9}, { 63, 8}, \ + { 127, 9}, { 71, 8}, { 143, 9}, { 79,10}, \ + { 47,11}, { 31,10}, { 63, 9}, { 127, 8}, \ + { 255, 7}, { 511, 9}, { 135, 8}, { 271, 9}, \ + { 143,10}, { 79, 9}, { 159, 8}, { 319, 9}, \ + { 175, 8}, { 351, 7}, { 703,10}, { 95, 9}, \ + { 191, 8}, { 383, 9}, { 207,11}, { 63,10}, \ + { 127, 9}, { 255, 8}, { 511, 9}, { 271,10}, \ + { 143, 9}, { 287, 8}, { 575,10}, { 159, 9}, \ + { 319,10}, { 175, 9}, { 351, 8}, { 703,11}, \ + { 95,10}, { 191, 9}, { 383,10}, { 207, 9}, \ + { 415,12}, { 63,11}, { 127,10}, { 255, 9}, \ + { 511,10}, { 271, 9}, { 543, 8}, { 1087,10}, \ + { 287, 9}, { 575,11}, { 159,10}, { 319, 9}, \ + { 639,10}, { 351, 9}, { 703, 8}, { 1407,11}, \ + { 191,10}, { 415, 9}, { 831,11}, { 223, 8}, \ + { 1791,10}, { 479, 9}, { 959,12}, { 127,11}, \ + { 255,10}, { 543,11}, { 287,10}, { 607,11}, \ + { 319,10}, { 639,11}, { 351,10}, { 703, 9}, \ + { 1407,12}, { 191,11}, { 415,10}, { 831,11}, \ + { 479,10}, { 959,13}, { 8192,14}, { 16384,15}, \ + { 32768,16} } +#define SQR_FFT_TABLE3_SIZE 109 +#define SQR_FFT_THRESHOLD 1600 #define MULLO_BASECASE_THRESHOLD 0 /* always */ #define MULLO_DC_THRESHOLD 90 -#define MULLO_MUL_N_THRESHOLD 3176 +#define MULLO_MUL_N_THRESHOLD 4167 #define DC_DIV_QR_THRESHOLD 100 #define DC_DIVAPPR_Q_THRESHOLD 342 diff --git a/mpn/powerpc32/gmp-mparam.h b/mpn/powerpc32/gmp-mparam.h index f4239c77d..c1f25532b 100644 --- a/mpn/powerpc32/gmp-mparam.h +++ b/mpn/powerpc32/gmp-mparam.h @@ -66,12 +66,95 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 9 #define SQRMOD_BNM1_THRESHOLD 14 -#define MUL_FFT_TABLE { 304, 672, 896, 2560, 6144, 24576, 98304, 0 } -#define MUL_FFT_MODF_THRESHOLD 320 +#define MUL_FFT_MODF_THRESHOLD 284 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 284, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 13, 7}, { 7, 6}, { 17, 7}, { 9, 6}, \ + { 20, 7}, { 11, 6}, { 23, 7}, { 13, 8}, \ + { 7, 7}, { 19, 8}, { 11, 7}, { 25, 9}, \ + { 7, 8}, { 15, 7}, { 33, 8}, { 19, 7}, \ + { 39, 8}, { 23, 7}, { 47, 9}, { 15, 8}, \ + { 39, 9}, { 23, 8}, { 47,10}, { 15, 9}, \ + { 31, 8}, { 67, 9}, { 39, 8}, { 79, 9}, \ + { 47, 8}, { 95,10}, { 31, 9}, { 71, 8}, \ + { 143, 9}, { 79,10}, { 47, 9}, { 95,11}, \ + { 31,10}, { 63, 9}, { 127, 8}, { 255, 9}, \ + { 135, 8}, { 271, 9}, { 143,10}, { 79, 9}, \ + { 159, 8}, { 319, 9}, { 175,10}, { 95, 9}, \ + { 191, 8}, { 383, 9}, { 207, 8}, { 415,11}, \ + { 63,10}, { 127, 9}, { 255, 8}, { 511, 9}, \ + { 271,10}, { 143, 9}, { 287, 8}, { 575,10}, \ + { 159, 9}, { 319,10}, { 175,11}, { 95,10}, \ + { 191, 9}, { 383,10}, { 207, 9}, { 415, 8}, \ + { 831,12}, { 63,11}, { 127,10}, { 255, 9}, \ + { 511,10}, { 271, 9}, { 543, 8}, { 1087,10}, \ + { 287, 9}, { 575,11}, { 159,10}, { 319, 9}, \ + { 639,10}, { 351, 9}, { 703,11}, { 191,10}, \ + { 415, 9}, { 831,11}, { 223,10}, { 447, 9}, \ + { 895,10}, { 479, 9}, { 959,12}, { 127,11}, \ + { 255,10}, { 543, 9}, { 1087,11}, { 287,10}, \ + { 607,11}, { 319,10}, { 639,11}, { 351,10}, \ + { 703, 9}, { 1407,12}, { 191,11}, { 383,10}, \ + { 767,11}, { 415,10}, { 831,11}, { 447,10}, \ + { 895,11}, { 479,10}, { 959,13}, { 127,12}, \ + { 255,11}, { 543,10}, { 1087,11}, { 607,12}, \ + { 319,11}, { 639,10}, { 1279,11}, { 703,10}, \ + { 1407,12}, { 383,11}, { 831,12}, { 447,11}, \ + { 959,10}, { 1919,13}, { 255,12}, { 511,11}, \ + { 1087,12}, { 575,11}, { 1215,10}, { 2431,12}, \ + { 639,11}, { 1279,12}, { 703,11}, { 1407,13}, \ + { 383,12}, { 959,11}, { 1919,14}, { 255,13}, \ + { 511,12}, { 1215,11}, { 2431,13}, { 639,12}, \ + { 1471,13}, { 767,12}, { 1599,13}, { 895,12}, \ + { 1919,14}, { 511,13}, { 1023,12}, { 2111,13}, \ + { 1151,12}, { 2431,13}, { 1407,14}, { 767,13}, \ + { 1535,12}, { 3071,13}, { 1919,12}, { 3839,15}, \ + { 32768,16} } +#define MUL_FFT_TABLE3_SIZE 165 #define MUL_FFT_THRESHOLD 3712 -#define SQR_FFT_TABLE { 272, 672, 1152, 2560, 10240, 40960, 98304, 0 } -#define SQR_FFT_MODF_THRESHOLD 288 +#define SQR_FFT_MODF_THRESHOLD 248 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 248, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 17, 7}, { 9, 6}, { 20, 7}, { 11, 6}, \ + { 23, 7}, { 13, 8}, { 7, 7}, { 19, 8}, \ + { 11, 7}, { 25, 9}, { 7, 8}, { 15, 7}, \ + { 33, 8}, { 19, 7}, { 39, 8}, { 23, 7}, \ + { 47, 8}, { 27, 9}, { 15, 8}, { 39, 9}, \ + { 23, 8}, { 47,10}, { 15, 9}, { 31, 8}, \ + { 63, 9}, { 39, 8}, { 79, 9}, { 47,10}, \ + { 31, 9}, { 63, 8}, { 127, 9}, { 71, 8}, \ + { 143, 9}, { 79,10}, { 47, 9}, { 95,11}, \ + { 31,10}, { 63, 9}, { 127, 8}, { 255, 7}, \ + { 511, 9}, { 143,10}, { 79, 9}, { 159, 8}, \ + { 319, 9}, { 175, 8}, { 351,10}, { 95, 9}, \ + { 191, 8}, { 383, 9}, { 207, 8}, { 415,11}, \ + { 63,10}, { 127, 9}, { 255, 8}, { 511,10}, \ + { 143, 9}, { 287, 8}, { 575,10}, { 159, 9}, \ + { 319,10}, { 175, 9}, { 351,11}, { 95,10}, \ + { 191, 9}, { 383,10}, { 207, 9}, { 415, 8}, \ + { 831,12}, { 63,11}, { 127,10}, { 255, 9}, \ + { 511,10}, { 271, 9}, { 543,10}, { 287, 9}, \ + { 575,11}, { 159,10}, { 319, 9}, { 639,10}, \ + { 351, 9}, { 703, 8}, { 1407, 9}, { 735,11}, \ + { 191,10}, { 415, 9}, { 831,11}, { 223,10}, \ + { 447, 9}, { 895,10}, { 479,12}, { 127,11}, \ + { 255,10}, { 543,11}, { 287,10}, { 607,11}, \ + { 319,10}, { 639,11}, { 351,10}, { 703, 9}, \ + { 1407,12}, { 191,11}, { 383,13}, { 127,12}, \ + { 255,11}, { 543,10}, { 1087,11}, { 575,10}, \ + { 1151,12}, { 319,11}, { 703,10}, { 1407,12}, \ + { 383,11}, { 831,12}, { 447,11}, { 959,13}, \ + { 255,12}, { 511,11}, { 1087,12}, { 575,11}, \ + { 1215,12}, { 639,11}, { 1279,12}, { 703,11}, \ + { 1407,13}, { 383,12}, { 959,14}, { 255,13}, \ + { 511,12}, { 1215,11}, { 2431,13}, { 639,12}, \ + { 1471,13}, { 767,12}, { 1599,13}, { 895,12}, \ + { 1919,14}, { 511,13}, { 1023,12}, { 2111,13}, \ + { 1151,12}, { 2431,13}, { 1407,12}, { 2815,14}, \ + { 767,13}, { 1535,12}, { 3199,13}, { 1919,15}, \ + { 32768,16} } +#define SQR_FFT_TABLE3_SIZE 153 #define SQR_FFT_THRESHOLD 2688 #define MULLO_BASECASE_THRESHOLD 0 /* always */ diff --git a/mpn/powerpc64/mode64/p5/gmp-mparam.h b/mpn/powerpc64/mode64/p5/gmp-mparam.h index 0062f3287..1ed896814 100644 --- a/mpn/powerpc64/mode64/p5/gmp-mparam.h +++ b/mpn/powerpc64/mode64/p5/gmp-mparam.h @@ -55,17 +55,119 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 13 #define SQRMOD_BNM1_THRESHOLD 9 -#define MUL_FFT_TABLE { 336, 736, 1728, 2816, 7168, 20480, 81920, 196608, 0 } -#define MUL_FFT_MODF_THRESHOLD 352 +#define MUL_FFT_MODF_THRESHOLD 348 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 348, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 10, 5}, { 21, 6}, { 21, 7}, { 11, 6}, \ + { 23, 7}, { 12, 6}, { 25, 7}, { 21, 8}, \ + { 11, 7}, { 25, 8}, { 13, 7}, { 27, 8}, \ + { 15, 7}, { 31, 8}, { 21, 9}, { 11, 8}, \ + { 27, 9}, { 15, 8}, { 33, 9}, { 19, 8}, \ + { 39, 9}, { 23, 8}, { 47, 9}, { 27,10}, \ + { 15, 9}, { 39,10}, { 23, 9}, { 51,11}, \ + { 15,10}, { 31, 9}, { 67,10}, { 39, 9}, \ + { 79,10}, { 47, 9}, { 95,10}, { 55,11}, \ + { 31,10}, { 79,11}, { 47,10}, { 95,12}, \ + { 31,11}, { 63,10}, { 127, 9}, { 255,10}, \ + { 135,11}, { 79,10}, { 159, 9}, { 319,11}, \ + { 95,10}, { 191, 9}, { 383,12}, { 63,11}, \ + { 127,10}, { 255, 9}, { 511,10}, { 271,11}, \ + { 143,10}, { 287, 9}, { 575,10}, { 319,12}, \ + { 95,11}, { 191,10}, { 383,13}, { 63,12}, \ + { 127,11}, { 255,10}, { 511,11}, { 271,10}, \ + { 543,11}, { 287,10}, { 575, 9}, { 1151,11}, \ + { 319,10}, { 639,11}, { 351,10}, { 703,12}, \ + { 191,11}, { 383,10}, { 767,11}, { 415,12}, \ + { 223,11}, { 447,13}, { 127,12}, { 255,11}, \ + { 511,10}, { 1023,11}, { 543,10}, { 1087,12}, \ + { 287,11}, { 575,10}, { 1151,12}, { 319,11}, \ + { 639,12}, { 351,11}, { 703,13}, { 191,12}, \ + { 383,11}, { 767,12}, { 415,11}, { 831,12}, \ + { 447,11}, { 895,14}, { 127,13}, { 255,12}, \ + { 511,11}, { 1023,12}, { 543,11}, { 1087,10}, \ + { 2175,12}, { 575,11}, { 1151,12}, { 607,13}, \ + { 319,12}, { 639,11}, { 1279,12}, { 671,11}, \ + { 1343,12}, { 703,11}, { 1407,13}, { 383,12}, \ + { 767,11}, { 1535,12}, { 831,13}, { 447,12}, \ + { 959,11}, { 1919,14}, { 255,13}, { 511,12}, \ + { 1087,11}, { 2175,13}, { 575,12}, { 1215,11}, \ + { 2431,10}, { 4863,13}, { 639,12}, { 1343,13}, \ + { 703,12}, { 1407,14}, { 383,13}, { 767,12}, \ + { 1535,13}, { 831,12}, { 1663,13}, { 959,12}, \ + { 1919,11}, { 3839,15}, { 255,14}, { 511,13}, \ + { 1087,12}, { 2175,13}, { 1215,12}, { 2431,11}, \ + { 4863,14}, { 639,13}, { 1343,12}, { 2687,13}, \ + { 1407,12}, { 2815,13}, { 1471,12}, { 2943,14}, \ + { 767,13}, { 1599,12}, { 3199,13}, { 1663,14}, \ + { 895,13}, { 1919,12}, { 3839,15}, { 511,14}, \ + { 1023,13}, { 2175,14}, { 1151,13}, { 2431,12}, \ + { 4863,14}, { 1279,13}, { 2687,14}, { 1407,13}, \ + { 2943,15}, { 767,14}, { 1535,13}, { 3199,14}, \ + { 1663,13}, { 3327,14}, { 1919,13}, { 3839,16}, \ + { 511,15}, { 1023,14}, { 2431,13}, { 4863,15}, \ + { 1279,14}, { 2943,12}, { 11775,15}, { 1535,14}, \ + { 3327,15}, { 1791,14}, { 16384,15}, { 32768,16}, \ + { 65536,17}, { 131072,18}, { 262144,19}, { 524288,20}, \ + {1048576,21}, {2097152,22}, {4194304,23}, {8388608,24} } +#define MUL_FFT_TABLE3_SIZE 208 #define MUL_FFT_THRESHOLD 3712 -#define SQR_FFT_TABLE { 304, 608, 1472, 2816, 7168, 20480, 49152, 196608, 0 } -#define SQR_FFT_MODF_THRESHOLD 320 +#define SQR_FFT_MODF_THRESHOLD 272 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 272, 5}, { 15, 6}, { 8, 5}, { 17, 6}, \ + { 19, 7}, { 17, 8}, { 9, 7}, { 21, 8}, \ + { 11, 7}, { 24, 8}, { 13, 7}, { 27, 8}, \ + { 15, 7}, { 31, 8}, { 21, 9}, { 11, 8}, \ + { 27, 9}, { 15, 8}, { 33, 9}, { 19, 8}, \ + { 39, 9}, { 23, 8}, { 47, 9}, { 27,10}, \ + { 15, 9}, { 39,10}, { 23, 9}, { 47,11}, \ + { 15,10}, { 31, 9}, { 63,10}, { 47,11}, \ + { 31,10}, { 71, 9}, { 143,10}, { 79,11}, \ + { 47,10}, { 95,12}, { 31,11}, { 63,10}, \ + { 127, 9}, { 255, 8}, { 511, 9}, { 271,10}, \ + { 143,11}, { 79,10}, { 159, 9}, { 319,10}, \ + { 175, 9}, { 351,11}, { 95,10}, { 191, 9}, \ + { 383,10}, { 207, 9}, { 415,12}, { 63,11}, \ + { 127,10}, { 255, 9}, { 511,10}, { 271,11}, \ + { 143,10}, { 287, 9}, { 575,11}, { 159,10}, \ + { 319,11}, { 175,10}, { 351,12}, { 95,11}, \ + { 191,10}, { 383,11}, { 207,10}, { 415,13}, \ + { 63,12}, { 127,11}, { 255,10}, { 511,11}, \ + { 271,10}, { 543,11}, { 287,10}, { 575,12}, \ + { 159,11}, { 319,10}, { 639,11}, { 351,10}, \ + { 703,12}, { 191,11}, { 383,10}, { 767,11}, \ + { 415,12}, { 223,11}, { 447,10}, { 895,11}, \ + { 479,10}, { 959,12}, { 255,11}, { 511,10}, \ + { 1023,11}, { 543,12}, { 287,11}, { 575,12}, \ + { 319,11}, { 639,12}, { 351,11}, { 703,13}, \ + { 191,12}, { 383,11}, { 767,12}, { 415,11}, \ + { 831,12}, { 447,11}, { 895,12}, { 479,11}, \ + { 959,13}, { 255,12}, { 511,11}, { 1023,12}, \ + { 543,11}, { 1087,12}, { 575,13}, { 319,12}, \ + { 639,11}, { 1279,12}, { 703,11}, { 1407,13}, \ + { 383,12}, { 831,13}, { 447,12}, { 959,14}, \ + { 255,13}, { 511,12}, { 1087,13}, { 575,12}, \ + { 1215,13}, { 639,12}, { 1279,13}, { 703,12}, \ + { 1407,14}, { 383,13}, { 831,12}, { 1663,13}, \ + { 959,12}, { 1919,15}, { 255,14}, { 511,13}, \ + { 1023,12}, { 2047,13}, { 1087,12}, { 2175,13}, \ + { 1215,14}, { 639,13}, { 1407,12}, { 2815,14}, \ + { 767,13}, { 1663,14}, { 895,13}, { 1919,15}, \ + { 511,14}, { 1023,13}, { 2175,14}, { 1151,13}, \ + { 2431,12}, { 4863,14}, { 1407,13}, { 2815,15}, \ + { 767,14}, { 1663,13}, { 3327,14}, { 1919,13}, \ + { 3839,16}, { 511,15}, { 1023,14}, { 2431,13}, \ + { 4863,15}, { 1279,14}, { 2943,13}, { 5887,12}, \ + { 11775,15}, { 1535,14}, { 3327,15}, { 1791,14}, \ + { 16384,15}, { 32768,16}, { 65536,17}, { 131072,18}, \ + { 262144,19}, { 524288,20}, {1048576,21}, {2097152,22}, \ + {4194304,23}, {8388608,24} } +#define SQR_FFT_TABLE3_SIZE 190 #define SQR_FFT_THRESHOLD 2752 #define MULLO_BASECASE_THRESHOLD 2 #define MULLO_DC_THRESHOLD 31 -#define MULLO_MUL_N_THRESHOLD 6633 +#define MULLO_MUL_N_THRESHOLD 7246 #define DC_DIV_QR_THRESHOLD 29 #define DC_DIVAPPR_Q_THRESHOLD 102 @@ -93,5 +195,5 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define GET_STR_DC_THRESHOLD 13 #define GET_STR_PRECOMPUTE_THRESHOLD 23 -#define SET_STR_DC_THRESHOLD 532 +#define SET_STR_DC_THRESHOLD 632 #define SET_STR_PRECOMPUTE_THRESHOLD 1790 diff --git a/mpn/x86/k6/gmp-mparam.h b/mpn/x86/k6/gmp-mparam.h index 6a103b7e4..168ea065e 100644 --- a/mpn/x86/k6/gmp-mparam.h +++ b/mpn/x86/k6/gmp-mparam.h @@ -56,13 +56,70 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 13 #define SQRMOD_BNM1_THRESHOLD 17 -#define MUL_FFT_TABLE { 368, 672, 1152, 3584, 10240, 24576, 163840, 0 } -#define MUL_FFT_MODF_THRESHOLD 624 -#define MUL_FFT_THRESHOLD 7808 - -#define SQR_FFT_TABLE { 400, 672, 1408, 4608, 10240, 24576, 98304, 393216, 0 } -#define SQR_FFT_MODF_THRESHOLD 528 -#define SQR_FFT_THRESHOLD 9600 +#define MUL_FFT_MODF_THRESHOLD 476 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 476, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 11, 5}, { 23, 6}, { 17, 7}, { 9, 6}, \ + { 19, 7}, { 11, 6}, { 23, 7}, { 13, 6}, \ + { 27, 7}, { 15, 6}, { 31, 7}, { 17, 6}, \ + { 35, 7}, { 21, 8}, { 11, 7}, { 27, 8}, \ + { 15, 7}, { 35, 8}, { 19, 7}, { 39, 8}, \ + { 23, 7}, { 47, 8}, { 27, 9}, { 15, 8}, \ + { 31, 7}, { 63, 8}, { 39, 9}, { 23, 8}, \ + { 51,10}, { 15, 9}, { 31, 8}, { 67, 9}, \ + { 47,10}, { 31, 9}, { 79,10}, { 47, 9}, \ + { 95,11}, { 31,10}, { 63, 9}, { 135,10}, \ + { 79, 9}, { 167,10}, { 95, 9}, { 191,10}, \ + { 111,11}, { 63,10}, { 127, 9}, { 255,10}, \ + { 143, 9}, { 287,10}, { 159,11}, { 95,10}, \ + { 191, 9}, { 383,12}, { 63,11}, { 127,10}, \ + { 255, 9}, { 511,10}, { 271, 9}, { 543,10}, \ + { 287,11}, { 159,10}, { 351,11}, { 191,10}, \ + { 415, 9}, { 831,11}, { 223,12}, { 127,11}, \ + { 255,10}, { 543,11}, { 287,10}, { 575,11}, \ + { 351,10}, { 703,12}, { 191,11}, { 415,10}, \ + { 831,13}, { 127,12}, { 255,11}, { 543,10}, \ + { 1087,11}, { 575,12}, { 319,11}, { 703,12}, \ + { 383,11}, { 831,12}, { 447,11}, { 895,13}, \ + { 255,12}, { 511,11}, { 1087,12}, { 575,11}, \ + { 1151,12}, { 703,13}, { 383,12}, { 959,14}, \ + { 255,13}, { 511,12}, { 1215,13}, { 8192,14}, \ + { 16384,15}, { 32768,16} } +#define MUL_FFT_TABLE3_SIZE 106 +#define MUL_FFT_THRESHOLD 7424 + +#define SQR_FFT_MODF_THRESHOLD 432 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 432, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 11, 5}, { 23, 6}, { 21, 7}, { 11, 6}, \ + { 24, 7}, { 13, 6}, { 27, 7}, { 15, 6}, \ + { 31, 7}, { 21, 8}, { 11, 7}, { 29, 8}, \ + { 15, 7}, { 35, 8}, { 19, 7}, { 39, 8}, \ + { 23, 7}, { 49, 8}, { 27, 9}, { 15, 8}, \ + { 39, 9}, { 23, 7}, { 93, 8}, { 47, 7}, \ + { 95, 8}, { 51,10}, { 15, 9}, { 31, 8}, \ + { 67, 9}, { 39, 8}, { 79, 9}, { 47, 8}, \ + { 95, 9}, { 55,10}, { 31, 9}, { 71, 8}, \ + { 143, 9}, { 79,10}, { 47, 9}, { 95,11}, \ + { 31,10}, { 63, 9}, { 135,10}, { 79, 9}, \ + { 167,10}, { 95, 9}, { 191,11}, { 63,10}, \ + { 127, 9}, { 255,10}, { 143, 9}, { 287, 8}, \ + { 575,10}, { 159, 9}, { 319,11}, { 95,10}, \ + { 191,12}, { 63,11}, { 127,10}, { 255, 9}, \ + { 511,10}, { 271, 9}, { 543,10}, { 287,11}, \ + { 159,10}, { 319, 9}, { 639,10}, { 351, 9}, \ + { 703,11}, { 191,10}, { 415,11}, { 223,12}, \ + { 127,11}, { 255,10}, { 543,11}, { 287,10}, \ + { 607,11}, { 319,10}, { 639,11}, { 351,10}, \ + { 703,12}, { 191,11}, { 415,10}, { 831,13}, \ + { 127,12}, { 255,11}, { 543,10}, { 1087,11}, \ + { 607,12}, { 319,11}, { 703,12}, { 383,11}, \ + { 831,12}, { 447,13}, { 255,12}, { 511,11}, \ + { 1087,12}, { 575,11}, { 1215,12}, { 703,13}, \ + { 383,12}, { 895,14}, { 255,13}, { 511,12}, \ + { 1215,13}, { 8192,14}, { 16384,15}, { 32768,16} } +#define SQR_FFT_TABLE3_SIZE 112 +#define SQR_FFT_THRESHOLD 7040 #define MULLO_BASECASE_THRESHOLD 3 #define MULLO_DC_THRESHOLD 60 diff --git a/mpn/x86/p6/sse2/gmp-mparam.h b/mpn/x86/p6/sse2/gmp-mparam.h index 7fa81b97a..2e7eb838a 100644 --- a/mpn/x86/p6/sse2/gmp-mparam.h +++ b/mpn/x86/p6/sse2/gmp-mparam.h @@ -62,17 +62,89 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 13 #define SQRMOD_BNM1_THRESHOLD 18 -#define MUL_FFT_TABLE { 400, 928, 1664, 3584, 10240, 24576, 163840, 0 } -#define MUL_FFT_MODF_THRESHOLD 720 -#define MUL_FFT_THRESHOLD 7552 - -#define SQR_FFT_TABLE { 400, 928, 1664, 3584, 10240, 40960, 163840, 0 } -#define SQR_FFT_MODF_THRESHOLD 592 +#define MUL_FFT_MODF_THRESHOLD 565 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 565, 5}, { 25, 6}, { 13, 5}, { 27, 6}, \ + { 25, 7}, { 13, 6}, { 28, 7}, { 15, 6}, \ + { 31, 7}, { 17, 6}, { 35, 7}, { 27, 8}, \ + { 15, 7}, { 35, 8}, { 19, 7}, { 41, 8}, \ + { 23, 7}, { 47, 8}, { 27, 9}, { 15, 8}, \ + { 31, 7}, { 63, 8}, { 39, 9}, { 23, 5}, \ + { 383, 4}, { 991, 5}, { 511, 6}, { 267, 7}, \ + { 157, 8}, { 91, 9}, { 47, 8}, { 111, 9}, \ + { 63, 8}, { 127, 9}, { 79,10}, { 47, 9}, \ + { 95,11}, { 31,10}, { 63, 9}, { 135,10}, \ + { 79, 9}, { 159,10}, { 95,11}, { 63,10}, \ + { 143, 9}, { 287,10}, { 159,11}, { 95,10}, \ + { 191,12}, { 63,11}, { 127,10}, { 255, 9}, \ + { 511,10}, { 271, 9}, { 543,10}, { 287,11}, \ + { 159,10}, { 335, 9}, { 671,11}, { 191,10}, \ + { 383, 9}, { 767,10}, { 399, 9}, { 799,10}, \ + { 415,11}, { 223,12}, { 127,11}, { 255,10}, \ + { 543, 9}, { 1087,11}, { 287,10}, { 607,11}, \ + { 319,10}, { 671,12}, { 191,11}, { 383,10}, \ + { 799,11}, { 415,10}, { 831,13}, { 127,12}, \ + { 255,11}, { 543,10}, { 1087,11}, { 607,10}, \ + { 1215,12}, { 319,11}, { 671,10}, { 1343,11}, \ + { 735,10}, { 1471,12}, { 383,11}, { 799,10}, \ + { 1599,11}, { 863,12}, { 447,11}, { 959,13}, \ + { 255,12}, { 511,11}, { 1087,12}, { 575,11}, \ + { 1215,12}, { 639,11}, { 1343,12}, { 703,11}, \ + { 1471,13}, { 383,12}, { 831,11}, { 1727,12}, \ + { 959,14}, { 255,13}, { 511,12}, { 1215,13}, \ + { 639,12}, { 1471,11}, { 2943,13}, { 767,12}, \ + { 1727,13}, { 895,12}, { 1919,14}, { 511,13}, \ + { 1023,12}, { 2111,13}, { 1151,12}, { 2431,13}, \ + { 1407,12}, { 2815,14}, { 767,13}, { 1663,12}, \ + { 3455,13}, { 8192,14}, { 16384,15}, { 32768,16} } +#define MUL_FFT_TABLE3_SIZE 132 +#define MUL_FFT_THRESHOLD 6784 + +#define SQR_FFT_MODF_THRESHOLD 472 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 472, 5}, { 25, 6}, { 13, 5}, { 27, 6}, \ + { 25, 7}, { 13, 6}, { 27, 7}, { 15, 6}, \ + { 31, 7}, { 17, 6}, { 35, 7}, { 27, 8}, \ + { 15, 7}, { 35, 8}, { 19, 7}, { 41, 8}, \ + { 23, 7}, { 49, 8}, { 27, 9}, { 15, 8}, \ + { 39, 9}, { 23, 8}, { 51,10}, { 15, 9}, \ + { 31, 8}, { 63, 4}, { 1023, 8}, { 67, 9}, \ + { 39, 5}, { 639, 4}, { 1471, 6}, { 383, 7}, \ + { 209, 8}, { 119, 9}, { 63, 7}, { 255, 8}, \ + { 139, 9}, { 71, 8}, { 143, 9}, { 79,10}, \ + { 47, 9}, { 95,11}, { 31,10}, { 63, 9}, \ + { 135,10}, { 79, 9}, { 159, 8}, { 319, 9}, \ + { 167,10}, { 95,11}, { 63,10}, { 143, 9}, \ + { 287,10}, { 159,11}, { 95,10}, { 191,12}, \ + { 63,11}, { 127,10}, { 255, 9}, { 543, 8}, \ + { 1087,10}, { 287, 9}, { 575,11}, { 159,10}, \ + { 319, 9}, { 639,10}, { 335, 9}, { 671,10}, \ + { 351, 9}, { 703,11}, { 191,10}, { 383, 9}, \ + { 767,10}, { 399, 9}, { 799,10}, { 415, 9}, \ + { 831,11}, { 223,12}, { 127,11}, { 255,10}, \ + { 543, 9}, { 1087,11}, { 287,10}, { 607, 9}, \ + { 1215,11}, { 319,10}, { 671, 9}, { 1343,11}, \ + { 351,10}, { 703,12}, { 191,11}, { 383,10}, \ + { 799,11}, { 415,10}, { 831,13}, { 127,12}, \ + { 255,11}, { 543,10}, { 1087,11}, { 607,12}, \ + { 319,11}, { 671,10}, { 1343,11}, { 735,12}, \ + { 383,11}, { 799,10}, { 1599,11}, { 863,12}, \ + { 447,11}, { 959,13}, { 255,12}, { 511,11}, \ + { 1087,12}, { 575,11}, { 1215,12}, { 639,11}, \ + { 1343,12}, { 703,11}, { 1471,13}, { 383,12}, \ + { 767,11}, { 1599,12}, { 831,11}, { 1727,12}, \ + { 959,14}, { 255,13}, { 511,12}, { 1215,13}, \ + { 639,12}, { 1471,13}, { 767,12}, { 1727,13}, \ + { 895,12}, { 1919,14}, { 511,13}, { 1023,12}, \ + { 2111,13}, { 1151,12}, { 2431,13}, { 1407,14}, \ + { 767,13}, { 1663,12}, { 3455,13}, { 8192,14}, \ + { 16384,15}, { 32768,16} } +#define SQR_FFT_TABLE3_SIZE 146 #define SQR_FFT_THRESHOLD 5760 #define MULLO_BASECASE_THRESHOLD 0 /* always */ #define MULLO_DC_THRESHOLD 34 -#define MULLO_MUL_N_THRESHOLD 13463 +#define MULLO_MUL_N_THRESHOLD 8907 #define DC_DIV_QR_THRESHOLD 19 #define DC_DIVAPPR_Q_THRESHOLD 56 @@ -102,7 +174,3 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define GET_STR_PRECOMPUTE_THRESHOLD 22 #define SET_STR_DC_THRESHOLD 587 #define SET_STR_PRECOMPUTE_THRESHOLD 983 - -#define MUL_FFT_TABLE2 {{1,4}, {337,5}, {801,6}, {1601,7}, {3457,8}, {3841,7}, {4481,8}, {4865,7}, {5249,8}, {6913,9}, {7681,8}, {9985,9}, {11777,8}, {13057,10}, {16897,8}, {22273,9}, {28417,10}, {30721,9}, {40961,10}, {48641,9}, {49153,11}, {63489,10}, {64513,9}, {72705,10}, {112641,11}, {129025,10}, {162817,11}, {194561,10}, {196609,12}, {258049,11}, {260097,10}, {278529,11}, {392705,9}, {395265,10}, {407553,11}, {452609,12}, {520193,11}, {522241,10}, {556545,11}, {587777,10}, {622081,11}, {653313,10}, {687105,12}, {782337,11}, {784385,10}, {818177,11}, {819201,13}, {1040385,12}, {1044481,11}, {1244161,12}, {1306625,11}, {1506305,12}, {1568769,11}, {1701889,10}, {1704961,12}, {1830913,11}, {1964033,13}, {2088961,12}, {2093057,11}, {2226177,12}, {2355201,11}, {2488321,12}, {2617345,11}, {2750465,12}, {2879489,11}, {3012609,13}, {3137537,12}, {3141633,11}, {3274753,12}, {3403777,11}, {3536897,12}, {3928065,14}, {MP_SIZE_T_MAX, 0}} - -#define SQR_FFT_TABLE2 {{1,4}, {337,5}, {801,6}, {1601,7}, {3457,8}, {3841,7}, {4481,8}, {4865,7}, {5249,8}, {5889,7}, {6273,8}, {6913,9}, {7681,8}, {9985,9}, {11777,8}, {13057,10}, {16641,8}, {17153,9}, {20993,7}, {25985,8}, {29697,9}, {40449,10}, {48129,9}, {49153,11}, {63489,10}, {64513,9}, {72705,10}, {97281,11}, {129025,10}, {162817,11}, {194561,10}, {195585,12}, {258049,11}, {260097,10}, {262145,9}, {264705,10}, {265729,9}, {267265,10}, {268289,9}, {269313,10}, {272897,9}, {278273,11}, {325633,10}, {359425,9}, {359937,11}, {391169,10}, {413697,11}, {456705,12}, {520193,11}, {522241,10}, {577537,11}, {581633,10}, {583681,11}, {585729,10}, {622081,11}, {653313,10}, {687617,11}, {718849,10}, {720897,12}, {782337,11}, {784385,10}, {818689,11}, {850945,13}, {1040385,12}, {1044481,11}, {1244161,12}, {1306625,11}, {1506305,12}, {1568769,11}, {1701889,10}, {1703937,12}, {1830913,11}, {1965057,13}, {2088961,12}, {2093057,11}, {2226177,12}, {2355201,11}, {2488321,12}, {2617345,11}, {2750465,12}, {2879489,11}, {3012609,13}, {3137537,12}, {3141633,11}, {3274753,12}, {3403777,11}, {3536897,12}, {3930113,14}, {MP_SIZE_T_MAX, 0}} diff --git a/mpn/x86/pentium/mmx/gmp-mparam.h b/mpn/x86/pentium/mmx/gmp-mparam.h index 2aa26ab65..958a32224 100644 --- a/mpn/x86/pentium/mmx/gmp-mparam.h +++ b/mpn/x86/pentium/mmx/gmp-mparam.h @@ -35,7 +35,7 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MOD_1U_TO_MOD_1_1_THRESHOLD 12 #define MOD_1_1_TO_MOD_1_2_THRESHOLD 0 #define MOD_1_2_TO_MOD_1_4_THRESHOLD 11 -#define PREINV_MOD_1_TO_MOD_1_THRESHOLD MP_SIZE_T_MAX /* never */ +#define PREINV_MOD_1_TO_MOD_1_THRESHOLD 63 #define USE_PREINV_DIVREM_1 0 #define DIVEXACT_1_THRESHOLD 0 /* always (native) */ #define BMOD_1_TO_MOD_1_THRESHOLD 51 @@ -61,29 +61,78 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MULMOD_BNM1_THRESHOLD 11 #define SQRMOD_BNM1_THRESHOLD 12 -#define MUL_FFT_TABLE { 304, 672, 1152, 3584, 10240, 40960, 163840, 0 } -#define MUL_FFT_MODF_THRESHOLD 368 -#define MUL_FFT_THRESHOLD 4736 - -#define SQR_FFT_TABLE { 272, 672, 1152, 4608, 10240, 40960, 98304, 393216, 0 } -#define SQR_FFT_MODF_THRESHOLD 288 +#define MUL_FFT_MODF_THRESHOLD 364 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 364, 5}, { 15, 6}, { 8, 5}, { 17, 6}, \ + { 9, 5}, { 19, 6}, { 17, 7}, { 9, 6}, \ + { 21, 7}, { 11, 6}, { 23, 7}, { 15, 6}, \ + { 31, 7}, { 21, 8}, { 11, 7}, { 27, 8}, \ + { 15, 7}, { 33, 8}, { 19, 7}, { 39, 8}, \ + { 23, 7}, { 47, 8}, { 27, 9}, { 15, 8}, \ + { 31, 7}, { 63, 8}, { 39, 9}, { 23, 8}, \ + { 47,10}, { 15, 9}, { 31, 8}, { 67, 9}, \ + { 39, 8}, { 79, 9}, { 47, 8}, { 95, 9}, \ + { 55,10}, { 31, 9}, { 79,10}, { 47, 9}, \ + { 95,11}, { 31,10}, { 63, 9}, { 135,10}, \ + { 79, 9}, { 159, 8}, { 319, 9}, { 167,10}, \ + { 95, 9}, { 191, 8}, { 383,11}, { 63,10}, \ + { 127, 9}, { 255,10}, { 143, 9}, { 287,10}, \ + { 159, 9}, { 319,11}, { 95,10}, { 191, 9}, \ + { 383,12}, { 63,11}, { 127,10}, { 271, 9}, \ + { 543,10}, { 287,11}, { 159,10}, { 351,11}, \ + { 191,10}, { 415,11}, { 223,12}, { 127,11}, \ + { 255,10}, { 511,11}, { 287,10}, { 575,11}, \ + { 351,12}, { 191,11}, { 415,13}, { 127,12}, \ + { 255,11}, { 575,12}, { 319,11}, { 703,12}, \ + { 383,11}, { 831,12}, { 447,13}, { 8192,14}, \ + { 16384,15}, { 32768,16} } +#define MUL_FFT_TABLE3_SIZE 90 +#define MUL_FFT_THRESHOLD 3520 + +#define SQR_FFT_MODF_THRESHOLD 340 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 340, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 17, 7}, { 9, 6}, { 21, 7}, { 11, 6}, \ + { 23, 7}, { 15, 6}, { 31, 7}, { 21, 8}, \ + { 11, 7}, { 29, 8}, { 15, 7}, { 33, 8}, \ + { 19, 7}, { 39, 8}, { 27, 7}, { 55, 9}, \ + { 15, 8}, { 31, 7}, { 65, 8}, { 43, 9}, \ + { 23, 8}, { 47,10}, { 15, 9}, { 31, 8}, \ + { 67, 9}, { 39, 8}, { 83, 9}, { 47, 8}, \ + { 95,10}, { 31, 9}, { 63, 8}, { 127, 9}, \ + { 79,10}, { 47, 9}, { 95,11}, { 31,10}, \ + { 63, 9}, { 127, 8}, { 255, 9}, { 135,10}, \ + { 79, 9}, { 159, 8}, { 319,10}, { 95, 9}, \ + { 191,11}, { 63,10}, { 127, 9}, { 255, 8}, \ + { 511, 9}, { 271,10}, { 143, 9}, { 287, 8}, \ + { 575, 9}, { 303,10}, { 159, 9}, { 319,11}, \ + { 95,10}, { 191, 9}, { 383,10}, { 207,12}, \ + { 63,11}, { 127,10}, { 271, 9}, { 543,10}, \ + { 287, 9}, { 575,10}, { 303,11}, { 159,10}, \ + { 351,11}, { 191,10}, { 415,11}, { 223,10}, \ + { 447,12}, { 127,11}, { 255,10}, { 543,11}, \ + { 287,10}, { 607,11}, { 351,12}, { 191,11}, \ + { 479,13}, { 127,12}, { 255,11}, { 575,12}, \ + { 319,11}, { 703,12}, { 383,11}, { 767,12}, \ + { 447,13}, { 8192,14}, { 16384,15}, { 32768,16} } +#define SQR_FFT_TABLE3_SIZE 96 #define SQR_FFT_THRESHOLD 5504 #define MULLO_BASECASE_THRESHOLD 0 /* always */ -#define MULLO_DC_THRESHOLD 45 -#define MULLO_MUL_N_THRESHOLD 9174 +#define MULLO_DC_THRESHOLD 48 +#define MULLO_MUL_N_THRESHOLD 6633 #define DC_DIV_QR_THRESHOLD 43 #define DC_DIVAPPR_Q_THRESHOLD 170 #define DC_BDIV_QR_THRESHOLD 43 -#define DC_BDIV_Q_THRESHOLD 113 +#define DC_BDIV_Q_THRESHOLD 110 -#define INV_MULMOD_BNM1_THRESHOLD 54 -#define INV_NEWTON_THRESHOLD 179 -#define INV_APPR_THRESHOLD 173 +#define INV_MULMOD_BNM1_THRESHOLD 30 +#define INV_NEWTON_THRESHOLD 177 +#define INV_APPR_THRESHOLD 171 -#define BINV_NEWTON_THRESHOLD 197 -#define REDC_1_TO_REDC_N_THRESHOLD 47 +#define BINV_NEWTON_THRESHOLD 194 +#define REDC_1_TO_REDC_N_THRESHOLD 50 #define MU_DIV_QR_THRESHOLD 1142 #define MU_DIVAPPR_Q_THRESHOLD 1142 @@ -91,13 +140,13 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MU_BDIV_QR_THRESHOLD 942 #define MU_BDIV_Q_THRESHOLD 1017 -#define MATRIX22_STRASSEN_THRESHOLD 12 -#define HGCD_THRESHOLD 88 -#define GCD_DC_THRESHOLD 293 -#define GCDEXT_DC_THRESHOLD 233 +#define MATRIX22_STRASSEN_THRESHOLD 13 +#define HGCD_THRESHOLD 92 +#define GCD_DC_THRESHOLD 283 +#define GCDEXT_DC_THRESHOLD 221 #define JACOBI_BASE_METHOD 2 -#define GET_STR_DC_THRESHOLD 14 -#define GET_STR_PRECOMPUTE_THRESHOLD 26 -#define SET_STR_DC_THRESHOLD 272 -#define SET_STR_PRECOMPUTE_THRESHOLD 920 +#define GET_STR_DC_THRESHOLD 18 +#define GET_STR_PRECOMPUTE_THRESHOLD 31 +#define SET_STR_DC_THRESHOLD 490 +#define SET_STR_PRECOMPUTE_THRESHOLD 994 diff --git a/mpn/x86_64/gmp-mparam.h b/mpn/x86_64/gmp-mparam.h index 8ebc943fc..b412f30f5 100644 --- a/mpn/x86_64/gmp-mparam.h +++ b/mpn/x86_64/gmp-mparam.h @@ -187,11 +187,11 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MATRIX22_STRASSEN_THRESHOLD 17 #define HGCD_THRESHOLD 139 -#define GCD_DC_THRESHOLD 555 -#define GCDEXT_DC_THRESHOLD 496 +#define GCD_DC_THRESHOLD 501 +#define GCDEXT_DC_THRESHOLD 474 #define JACOBI_BASE_METHOD 1 -#define GET_STR_DC_THRESHOLD 9 +#define GET_STR_DC_THRESHOLD 17 #define GET_STR_PRECOMPUTE_THRESHOLD 23 -#define SET_STR_DC_THRESHOLD 248 +#define SET_STR_DC_THRESHOLD 266 #define SET_STR_PRECOMPUTE_THRESHOLD 1648 diff --git a/mpn/x86_64/pentium4/gmp-mparam.h b/mpn/x86_64/pentium4/gmp-mparam.h index 9196e35e1..461633013 100644 --- a/mpn/x86_64/pentium4/gmp-mparam.h +++ b/mpn/x86_64/pentium4/gmp-mparam.h @@ -31,81 +31,187 @@ along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. */ #define MOD_1_NORM_THRESHOLD 0 /* always */ #define MOD_1_UNNORM_THRESHOLD 0 /* always */ #define MOD_1N_TO_MOD_1_1_THRESHOLD 6 -#define MOD_1U_TO_MOD_1_1_THRESHOLD 4 -#define MOD_1_1_TO_MOD_1_2_THRESHOLD 8 -#define MOD_1_2_TO_MOD_1_4_THRESHOLD 20 -#define PREINV_MOD_1_TO_MOD_1_THRESHOLD 10 +#define MOD_1U_TO_MOD_1_1_THRESHOLD 5 +#define MOD_1_1_TO_MOD_1_2_THRESHOLD 10 +#define MOD_1_2_TO_MOD_1_4_THRESHOLD 34 +#define PREINV_MOD_1_TO_MOD_1_THRESHOLD 16 #define USE_PREINV_DIVREM_1 1 /* native */ #define DIVEXACT_1_THRESHOLD 0 /* always (native) */ -#define BMOD_1_TO_MOD_1_THRESHOLD 21 +#define BMOD_1_TO_MOD_1_THRESHOLD 24 -#define MUL_TOOM22_THRESHOLD 12 +#define MUL_TOOM22_THRESHOLD 14 #define MUL_TOOM33_THRESHOLD 81 -#define MUL_TOOM44_THRESHOLD 130 -#define MUL_TOOM6H_THRESHOLD 197 -#define MUL_TOOM8H_THRESHOLD 482 +#define MUL_TOOM44_THRESHOLD 214 +#define MUL_TOOM6H_THRESHOLD 286 +#define MUL_TOOM8H_THRESHOLD 430 -#define MUL_TOOM32_TO_TOOM43_THRESHOLD 89 -#define MUL_TOOM32_TO_TOOM53_THRESHOLD 154 -#define MUL_TOOM42_TO_TOOM53_THRESHOLD 94 -#define MUL_TOOM42_TO_TOOM63_THRESHOLD 96 +#define MUL_TOOM32_TO_TOOM43_THRESHOLD 81 +#define MUL_TOOM32_TO_TOOM53_THRESHOLD 138 +#define MUL_TOOM42_TO_TOOM53_THRESHOLD 82 +#define MUL_TOOM42_TO_TOOM63_THRESHOLD 80 #define SQR_BASECASE_THRESHOLD 0 /* always (native) */ -#define SQR_TOOM2_THRESHOLD 22 -#define SQR_TOOM3_THRESHOLD 89 -#define SQR_TOOM4_THRESHOLD 244 -#define SQR_TOOM6_THRESHOLD 318 -#define SQR_TOOM8_THRESHOLD 502 +#define SQR_TOOM2_THRESHOLD 20 +#define SQR_TOOM3_THRESHOLD 81 +#define SQR_TOOM4_THRESHOLD 226 +#define SQR_TOOM6_THRESHOLD 303 +#define SQR_TOOM8_THRESHOLD 454 #define MULMOD_BNM1_THRESHOLD 9 -#define SQRMOD_BNM1_THRESHOLD 13 - -#define MUL_FFT_TABLE { 240, 544, 1600, 2816, 7168, 20480, 49152, 327680, 0 } -#define MUL_FFT_MODF_THRESHOLD 272 -#define MUL_FFT_THRESHOLD 4864 - -#define SQR_FFT_TABLE { 272, 544, 1600, 2816, 7168, 20480, 49152, 196608, 0 } -#define SQR_FFT_MODF_THRESHOLD 272 -#define SQR_FFT_THRESHOLD 4224 +#define SQRMOD_BNM1_THRESHOLD 11 + + +#define MUL_FFT_MODF_THRESHOLD 240 /* k = 5 */ +#define MUL_FFT_TABLE3 \ + { { 240, 5}, { 9, 4}, { 19, 5}, { 11, 6}, \ + { 6, 5}, { 13, 6}, { 7, 5}, { 15, 6}, \ + { 8, 5}, { 17, 6}, { 9, 5}, { 19, 6}, \ + { 13, 7}, { 7, 6}, { 15, 7}, { 8, 6}, \ + { 17, 7}, { 9, 6}, { 19, 7}, { 11, 6}, \ + { 23, 7}, { 13, 8}, { 7, 7}, { 17, 8}, \ + { 9, 7}, { 21, 8}, { 11, 7}, { 23, 8}, \ + { 13, 9}, { 7, 8}, { 15, 7}, { 31, 8}, \ + { 21, 9}, { 11, 8}, { 25,10}, { 7, 9}, \ + { 15, 8}, { 33, 9}, { 19, 8}, { 39, 9}, \ + { 23, 8}, { 47, 9}, { 27,10}, { 15, 9}, \ + { 39,10}, { 23, 9}, { 51,11}, { 15,10}, \ + { 31, 9}, { 63,10}, { 39, 9}, { 79,10}, \ + { 47,11}, { 31,10}, { 63, 9}, { 127, 8}, \ + { 255,10}, { 79, 9}, { 159,11}, { 47,10}, \ + { 95,12}, { 31,11}, { 63,10}, { 127, 9}, \ + { 287,11}, { 79,10}, { 159, 9}, { 319,10}, \ + { 175,11}, { 95,10}, { 191, 9}, { 383,10}, \ + { 207, 9}, { 415,11}, { 111,10}, { 223,12}, \ + { 63,11}, { 127,10}, { 255,11}, { 143,10}, \ + { 287,11}, { 159,10}, { 319,11}, { 175,12}, \ + { 95,11}, { 223,13}, { 63,12}, { 127,11}, \ + { 287,10}, { 575,12}, { 159,11}, { 319,10}, \ + { 639,11}, { 351,12}, { 191,11}, { 383,12}, \ + { 223,11}, { 447,13}, { 127,12}, { 255,11}, \ + { 511,12}, { 287,11}, { 575,12}, { 319,11}, \ + { 639,12}, { 351,13}, { 191,12}, { 415,11}, \ + { 831,12}, { 447,14}, { 127,13}, { 255,12}, \ + { 511,11}, { 1023,12}, { 543,11}, { 1087,10}, \ + { 2175,12}, { 575,13}, { 319,12}, { 639,11}, \ + { 1279,12}, { 703,11}, { 1407,13}, { 383,12}, \ + { 767,11}, { 1535,12}, { 831,11}, { 1663,13}, \ + { 447,14}, { 255,13}, { 511,12}, { 1023,11}, \ + { 2047,12}, { 1087,11}, { 2175,13}, { 575,12}, \ + { 1151,11}, { 2303,12}, { 1215,11}, { 2431,10}, \ + { 4863,13}, { 639,12}, { 1279,11}, { 2559,13}, \ + { 703,12}, { 1407,14}, { 383,13}, { 767,12}, \ + { 1535,13}, { 831,12}, { 1663,13}, { 895,15}, \ + { 255,14}, { 511,13}, { 1023,12}, { 2047,13}, \ + { 1087,12}, { 2175,13}, { 1215,12}, { 2431,11}, \ + { 4863,14}, { 639,13}, { 1407,12}, { 2815,13}, \ + { 1471,14}, { 767,13}, { 1663,14}, { 895,13}, \ + { 1791,12}, { 3583,13}, { 1919,12}, { 3839,15}, \ + { 511,14}, { 1023,13}, { 2175,14}, { 1151,13}, \ + { 2303,12}, { 4607,13}, { 2431,12}, { 4863,14}, \ + { 1279,13}, { 2687,14}, { 1407,13}, { 2815,15}, \ + { 767,14}, { 1791,13}, { 3583,14}, { 1919,13}, \ + { 3839,12}, { 7679,16}, { 511,15}, { 1023,14}, \ + { 2303,13}, { 4607,14}, { 2431,13}, { 4863,15}, \ + { 1279,14}, { 2943,13}, { 5887,15}, { 1535,14}, \ + { 3199,15}, { 1791,14}, { 3839,13}, { 7679,16}, \ + { 1023,15}, { 2047,14}, { 4351,15}, { 2303,14}, \ + { 4863,15}, { 2815,14}, { 5887,13}, { 11775,16}, \ + { 1535,15}, { 3071,14}, { 6655,15}, { 32768,16}, \ + { 65536,17}, { 131072,18}, { 262144,19}, { 524288,20}, \ + {1048576,21}, {2097152,22}, {4194304,23}, {8388608,24} } +#define MUL_FFT_TABLE3_SIZE 224 +#define MUL_FFT_THRESHOLD 2752 + +#define SQR_FFT_MODF_THRESHOLD 240 /* k = 5 */ +#define SQR_FFT_TABLE3 \ + { { 240, 5}, { 11, 6}, { 6, 5}, { 13, 6}, \ + { 15, 7}, { 8, 6}, { 19, 7}, { 10, 6}, \ + { 21, 7}, { 13, 8}, { 7, 7}, { 21, 8}, \ + { 11, 7}, { 25, 8}, { 13, 9}, { 7, 8}, \ + { 15, 7}, { 31, 8}, { 21, 9}, { 11, 8}, \ + { 25,10}, { 7, 9}, { 15, 8}, { 33, 9}, \ + { 19, 8}, { 39, 9}, { 23, 8}, { 47, 9}, \ + { 27,10}, { 15, 9}, { 39,10}, { 23, 9}, \ + { 51,11}, { 15,10}, { 31, 9}, { 63, 8}, \ + { 127,10}, { 39, 9}, { 79,10}, { 47,11}, \ + { 31,10}, { 63, 9}, { 127, 8}, { 255,10}, \ + { 71, 9}, { 143, 7}, { 575,10}, { 79,11}, \ + { 47,12}, { 31,11}, { 63,10}, { 127, 9}, \ + { 255,10}, { 143, 9}, { 287,11}, { 79, 9}, \ + { 319,10}, { 191, 9}, { 383,10}, { 207,12}, \ + { 63,11}, { 127,10}, { 255,11}, { 143,10}, \ + { 287, 9}, { 575,11}, { 159,10}, { 319,11}, \ + { 175,10}, { 351,12}, { 95,11}, { 191,10}, \ + { 383,11}, { 223,13}, { 63,12}, { 127,11}, \ + { 287,12}, { 159,11}, { 351,12}, { 191,11}, \ + { 383,10}, { 767,11}, { 415,12}, { 223,11}, \ + { 447,13}, { 127,12}, { 255,11}, { 511,12}, \ + { 287,11}, { 575,12}, { 319,11}, { 639,12}, \ + { 351,13}, { 191,12}, { 383,11}, { 767,12}, \ + { 415,11}, { 831,12}, { 447,14}, { 127,13}, \ + { 255,12}, { 511,11}, { 1023,12}, { 543,11}, \ + { 1087,10}, { 2175,12}, { 575,13}, { 319,12}, \ + { 639,11}, { 1279,12}, { 671,11}, { 1343,12}, \ + { 703,13}, { 383,12}, { 767,11}, { 1535,12}, \ + { 831,13}, { 447,14}, { 255,13}, { 511,12}, \ + { 1023,11}, { 2047,12}, { 1087,11}, { 2175,13}, \ + { 575,12}, { 1151,11}, { 2303,12}, { 1215,13}, \ + { 639,12}, { 1279,11}, { 2559,12}, { 1343,13}, \ + { 703,14}, { 383,13}, { 767,12}, { 1535,13}, \ + { 831,12}, { 1663,15}, { 255,14}, { 511,13}, \ + { 1023,12}, { 2047,13}, { 1087,12}, { 2175,13}, \ + { 1151,12}, { 2303,13}, { 1215,14}, { 639,13}, \ + { 1343,12}, { 2687,13}, { 1407,12}, { 2815,14}, \ + { 767,13}, { 1663,14}, { 895,13}, { 1791,12}, \ + { 3583,13}, { 1919,12}, { 3839,15}, { 511,14}, \ + { 1023,13}, { 2175,14}, { 1151,13}, { 2303,12}, \ + { 4607,13}, { 2431,12}, { 4863,14}, { 1279,13}, \ + { 2687,14}, { 1407,13}, { 2815,15}, { 767,14}, \ + { 1535,13}, { 3071,14}, { 1791,13}, { 3583,14}, \ + { 1919,13}, { 3839,12}, { 7679,16}, { 511,15}, \ + { 1023,14}, { 2175,13}, { 4351,14}, { 2303,13}, \ + { 4607,14}, { 2431,13}, { 4863,15}, { 1279,14}, \ + { 2815,13}, { 5631,14}, { 2943,13}, { 5887,12}, \ + { 11775,15}, { 1535,14}, { 3199,15}, { 1791,14}, \ + { 3583,13}, { 7167,14}, { 3839,13}, { 7679,16}, \ + { 1023,15}, { 2047,14}, { 4351,15}, { 2303,14}, \ + { 4863,15}, { 2815,14}, { 5887,13}, { 11775,16}, \ + { 1535,15}, { 3071,14}, { 6655,15}, { 32768,16}, \ + { 65536,17}, { 131072,18}, { 262144,19}, { 524288,20}, \ + {1048576,21}, {2097152,22}, {4194304,23}, {8388608,24} } +#define SQR_FFT_TABLE3_SIZE 220 +#define SQR_FFT_THRESHOLD 2752 #define MULLO_BASECASE_THRESHOLD 0 /* always */ -#define MULLO_DC_THRESHOLD 28 -#define MULLO_MUL_N_THRESHOLD 8648 +#define MULLO_DC_THRESHOLD 27 +#define MULLO_MUL_N_THRESHOLD 5240 -#define DC_DIV_QR_THRESHOLD 32 -#define DC_DIVAPPR_Q_THRESHOLD 68 +#define DC_DIV_QR_THRESHOLD 28 +#define DC_DIVAPPR_Q_THRESHOLD 60 #define DC_BDIV_QR_THRESHOLD 31 -#define DC_BDIV_Q_THRESHOLD 57 +#define DC_BDIV_Q_THRESHOLD 60 -#define INV_MULMOD_BNM1_THRESHOLD 116 +#define INV_MULMOD_BNM1_THRESHOLD 99 #define INV_NEWTON_THRESHOLD 266 -#define INV_APPR_THRESHOLD 114 +#define INV_APPR_THRESHOLD 110 -#define BINV_NEWTON_THRESHOLD 242 +#define BINV_NEWTON_THRESHOLD 262 #define REDC_1_TO_REDC_2_THRESHOLD 15 -#define REDC_2_TO_REDC_N_THRESHOLD 50 - -#define MU_DIV_QR_THRESHOLD 1589 -#define MU_DIVAPPR_Q_THRESHOLD 1528 -#define MUPI_DIV_QR_THRESHOLD 315 -#define MU_BDIV_QR_THRESHOLD 1499 -#define MU_BDIV_Q_THRESHOLD 1499 - -#define MATRIX22_STRASSEN_THRESHOLD 19 -#define HGCD_THRESHOLD 101 -#define GCD_DC_THRESHOLD 242 -#define GCDEXT_DC_THRESHOLD 222 +#define REDC_2_TO_REDC_N_THRESHOLD 47 + +#define MU_DIV_QR_THRESHOLD 979 +#define MU_DIVAPPR_Q_THRESHOLD 1078 +#define MUPI_DIV_QR_THRESHOLD 102 +#define MU_BDIV_QR_THRESHOLD 792 +#define MU_BDIV_Q_THRESHOLD 942 + +#define MATRIX22_STRASSEN_THRESHOLD 21 +#define HGCD_THRESHOLD 97 +#define GCD_DC_THRESHOLD 217 +#define GCDEXT_DC_THRESHOLD 237 #define JACOBI_BASE_METHOD 1 -#define GET_STR_DC_THRESHOLD 13 -#define GET_STR_PRECOMPUTE_THRESHOLD 25 -#define SET_STR_DC_THRESHOLD 532 -#define SET_STR_PRECOMPUTE_THRESHOLD 1561 - -/* These tables need updating. */ - -#define MUL_FFT_TABLE2 {{1,4}, {145,5}, {353,6}, {961,7}, {2689,8}, {6913,10}, {7169,9}, {7681,8}, {8449,9}, {13825,10}, {15361,9}, {19969,10}, {23553,9}, {26113,11}, {30721,10}, {31745,9}, {34561,10}, {48129,11}, {63489,10}, {81921,11}, {96257,10}, {97793,12}, {126977,11}, {133121,10}, {135169,11}, {137217,10}, {139777,9}, {141313,10}, {145409,11}, {161793,10}, {163841,11}, {165889,10}, {169985,11}, {172033,10}, {177153,11}, {206849,12}, {212993,11}, {217089,12}, {258049,11}, {358401,12}, {389121,11}, {391169,10}, {397313,11}, {410625,10}, {421889,11}, {450561,13}, {516097,12}, {520193,11}, {587777,12}, {651265,11}, {718849,12}, {790529,11}, {800769,12}, {815105,11}, {821249,12}, {833537,11}, {845825,12}, {915457,13}, {1040385,12}, {1437697,13}, {1564673,12}, {1830913,14}, {2088961,12}, {2355201,13}, {2613249,12}, {2879489,13}, {3137537,12}, {3405825,13}, {3661825,14}, {4186113,12}, {4454401,13}, {4714497,11}, {4979713,13}, {MP_SIZE_T_MAX, 0}} -#define MUL_FFT_TABLE2_SIZE 73 - -#define SQR_FFT_TABLE2 {{1,4}, {177,5}, {417,6}, {961,7}, {3073,8}, {6913,10}, {7169,9}, {7681,8}, {8449,9}, {13825,10}, {15361,9}, {19969,10}, {23553,9}, {26113,11}, {30721,10}, {48129,11}, {63489,10}, {65537,8}, {65793,10}, {80897,11}, {96257,12}, {126977,11}, {129025,10}, {136193,11}, {137217,10}, {146433,9}, {146945,11}, {161793,10}, {176129,11}, {195073,10}, {201729,11}, {202753,10}, {206849,11}, {219137,12}, {258049,11}, {260097,10}, {262145,11}, {270337,10}, {272385,11}, {279553,10}, {280577,11}, {286721,10}, {287745,11}, {359425,12}, {389121,11}, {423937,13}, {516097,12}, {520193,11}, {587777,12}, {651265,11}, {718849,12}, {782337,11}, {802817,12}, {806913,11}, {849921,12}, {915457,13}, {1040385,12}, {1437697,13}, {1564673,12}, {1830913,14}, {2080769,13}, {2088961,12}, {2355201,13}, {2613249,12}, {2752513,13}, {2785281,12}, {2801665,13}, {2818049,12}, {2838529,13}, {2863105,12}, {2867201,13}, {3137537,12}, {3403777,13}, {3661825,14}, {4177921,13}, {4186113,12}, {4452353,13}, {4714497,11}, {4979713,13}, {MP_SIZE_T_MAX, 0}} -#define SQR_FFT_TABLE2_SIZE 79 +#define GET_STR_DC_THRESHOLD 12 +#define GET_STR_PRECOMPUTE_THRESHOLD 23 +#define SET_STR_DC_THRESHOLD 572 +#define SET_STR_PRECOMPUTE_THRESHOLD 1588 |