[Gmp-commit] /var/hg/gmp: Retune (mainly for DIV_1_VS_MUL_1_PERCENT).

mercurial at gmplib.org mercurial at gmplib.org
Mon Jan 2 12:02:54 UTC 2017


details:   /var/hg/gmp/rev/2d17a5171152
changeset: 17197:2d17a5171152
user:      Torbjorn Granlund <tg at gmplib.org>
date:      Mon Jan 02 13:02:52 2017 +0100
description:
Retune (mainly for DIV_1_VS_MUL_1_PERCENT).

diffstat:

 mpn/ia64/gmp-mparam.h                |    2 +
 mpn/powerpc64/mode64/p7/gmp-mparam.h |    2 +
 mpn/s390_32/esame/gmp-mparam.h       |  268 +++++++++++++++-------------------
 mpn/s390_64/gmp-mparam.h             |  250 ++++++++++++++++---------------
 mpn/x86/bd2/gmp-mparam.h             |    2 +
 mpn/x86_64/bd2/gmp-mparam.h          |    2 +
 mpn/x86_64/coreisbr/gmp-mparam.h     |    2 +
 7 files changed, 257 insertions(+), 271 deletions(-)

diffs (truncated from 637 to 300 lines):

diff -r faa6086f89b6 -r 2d17a5171152 mpn/ia64/gmp-mparam.h
--- a/mpn/ia64/gmp-mparam.h	Sun Jan 01 16:10:15 2017 +0100
+++ b/mpn/ia64/gmp-mparam.h	Mon Jan 02 13:02:52 2017 +0100
@@ -51,6 +51,8 @@
 #define DIVEXACT_1_THRESHOLD                 0  /* always (native) */
 #define BMOD_1_TO_MOD_1_THRESHOLD        MP_SIZE_T_MAX  /* never */
 
+#define DIV_1_VS_MUL_1_PERCENT             315
+
 #define MUL_TOOM22_THRESHOLD                39
 #define MUL_TOOM33_THRESHOLD               129
 #define MUL_TOOM44_THRESHOLD               214
diff -r faa6086f89b6 -r 2d17a5171152 mpn/powerpc64/mode64/p7/gmp-mparam.h
--- a/mpn/powerpc64/mode64/p7/gmp-mparam.h	Sun Jan 01 16:10:15 2017 +0100
+++ b/mpn/powerpc64/mode64/p7/gmp-mparam.h	Mon Jan 02 13:02:52 2017 +0100
@@ -51,6 +51,8 @@
 #define DIVEXACT_1_THRESHOLD                 0  /* always (native) */
 #define BMOD_1_TO_MOD_1_THRESHOLD           35
 
+#define DIV_1_VS_MUL_1_PERCENT             384
+
 #define MUL_TOOM22_THRESHOLD                22
 #define MUL_TOOM33_THRESHOLD                72
 #define MUL_TOOM44_THRESHOLD               202
diff -r faa6086f89b6 -r 2d17a5171152 mpn/s390_32/esame/gmp-mparam.h
--- a/mpn/s390_32/esame/gmp-mparam.h	Sun Jan 01 16:10:15 2017 +0100
+++ b/mpn/s390_32/esame/gmp-mparam.h	Mon Jan 02 13:02:52 2017 +0100
@@ -31,177 +31,147 @@
 #define GMP_LIMB_BITS 32
 #define GMP_LIMB_BYTES 4
 
-/* 4400 MHz IBM z10 running in 32-bit mode */
-/* FFT tuning limit = 15000000 */
-/* Generated by tuneup.c, 2014-03-12, gcc 4.7 */
+/* 4400 MHz IBM z196 running in 32-bit mode */
+/* FFT tuning limit = 0.5M */
+/* Generated by tuneup.c, 2017-01-02, gcc 4.9 */
 
-#define DIVREM_1_NORM_THRESHOLD              3
+#define DIVREM_1_NORM_THRESHOLD          MP_SIZE_T_MAX  /* never */
 #define DIVREM_1_UNNORM_THRESHOLD        MP_SIZE_T_MAX  /* never */
-#define MOD_1_1P_METHOD                      1
-#define MOD_1_NORM_THRESHOLD                 4
-#define MOD_1_UNNORM_THRESHOLD               8
-#define MOD_1N_TO_MOD_1_1_THRESHOLD         16
-#define MOD_1U_TO_MOD_1_1_THRESHOLD          8
+#define MOD_1_1P_METHOD                      2
+#define MOD_1_NORM_THRESHOLD             MP_SIZE_T_MAX  /* never */
+#define MOD_1_UNNORM_THRESHOLD           MP_SIZE_T_MAX  /* never */
+#define MOD_1N_TO_MOD_1_1_THRESHOLD         45
+#define MOD_1U_TO_MOD_1_1_THRESHOLD         18
 #define MOD_1_1_TO_MOD_1_2_THRESHOLD         0  /* never mpn_mod_1_1p */
-#define MOD_1_2_TO_MOD_1_4_THRESHOLD        38
-#define PREINV_MOD_1_TO_MOD_1_THRESHOLD     30
-#define USE_PREINV_DIVREM_1                  1
+#define MOD_1_2_TO_MOD_1_4_THRESHOLD     MP_SIZE_T_MAX
+#define PREINV_MOD_1_TO_MOD_1_THRESHOLD      3
+#define USE_PREINV_DIVREM_1                  0
 #define DIV_QR_1N_PI1_METHOD                 1
-#define DIV_QR_1_NORM_THRESHOLD              3
+#define DIV_QR_1_NORM_THRESHOLD          MP_SIZE_T_MAX  /* never */
 #define DIV_QR_1_UNNORM_THRESHOLD        MP_SIZE_T_MAX  /* never */
 #define DIV_QR_2_PI2_THRESHOLD           MP_SIZE_T_MAX  /* never */
-#define DIVEXACT_1_THRESHOLD                 0  /* always */
-#define BMOD_1_TO_MOD_1_THRESHOLD           56
+#define DIVEXACT_1_THRESHOLD                 6
+#define BMOD_1_TO_MOD_1_THRESHOLD            0  /* always */
 
-#define MUL_TOOM22_THRESHOLD                 8
-#define MUL_TOOM33_THRESHOLD                59
-#define MUL_TOOM44_THRESHOLD                88
-#define MUL_TOOM6H_THRESHOLD               125
-#define MUL_TOOM8H_THRESHOLD               169
+#define DIV_1_VS_MUL_1_PERCENT             320
 
-#define MUL_TOOM32_TO_TOOM43_THRESHOLD      57
-#define MUL_TOOM32_TO_TOOM53_THRESHOLD      55
-#define MUL_TOOM42_TO_TOOM53_THRESHOLD      57
-#define MUL_TOOM42_TO_TOOM63_THRESHOLD      56
-#define MUL_TOOM43_TO_TOOM54_THRESHOLD      82
+#define MUL_TOOM22_THRESHOLD                12
+#define MUL_TOOM33_THRESHOLD                81
+#define MUL_TOOM44_THRESHOLD               130
+#define MUL_TOOM6H_THRESHOLD               173
+#define MUL_TOOM8H_THRESHOLD               260
+
+#define MUL_TOOM32_TO_TOOM43_THRESHOLD      81
+#define MUL_TOOM32_TO_TOOM53_THRESHOLD      91
+#define MUL_TOOM42_TO_TOOM53_THRESHOLD      83
+#define MUL_TOOM42_TO_TOOM63_THRESHOLD      86
+#define MUL_TOOM43_TO_TOOM54_THRESHOLD     112
 
 #define SQR_BASECASE_THRESHOLD               0  /* always (native) */
-#define SQR_TOOM2_THRESHOLD                 14
-#define SQR_TOOM3_THRESHOLD                 90
-#define SQR_TOOM4_THRESHOLD                144
-#define SQR_TOOM6_THRESHOLD                196
-#define SQR_TOOM8_THRESHOLD                309
+#define SQR_TOOM2_THRESHOLD                 18
+#define SQR_TOOM3_THRESHOLD                 69
+#define SQR_TOOM4_THRESHOLD                178
+#define SQR_TOOM6_THRESHOLD                254
+#define SQR_TOOM8_THRESHOLD                406
 
-#define MULMID_TOOM42_THRESHOLD             24
+#define MULMID_TOOM42_THRESHOLD             30
 
 #define MULMOD_BNM1_THRESHOLD               12
-#define SQRMOD_BNM1_THRESHOLD               10
+#define SQRMOD_BNM1_THRESHOLD                7
 
-#define MUL_FFT_MODF_THRESHOLD             252  /* k = 5 */
+#define MUL_FFT_MODF_THRESHOLD             276  /* k = 5 */
 #define MUL_FFT_TABLE3                                      \
-  { {    252, 5}, {      9, 6}, {      5, 5}, {     13, 6}, \
-    {      7, 5}, {     15, 6}, {     13, 7}, {      7, 6}, \
-    {     15, 7}, {      9, 6}, {     19, 7}, {     11, 6}, \
-    {     23, 7}, {     13, 8}, {      7, 7}, {     15, 6}, \
-    {     31, 7}, {     19, 8}, {     11, 7}, {     23, 9}, \
-    {      7, 8}, {     15, 7}, {     31, 8}, {     19, 7}, \
-    {     39, 8}, {     27, 9}, {     15, 8}, {     39, 9}, \
-    {     23,10}, {     15, 9}, {     31, 8}, {     63, 9}, \
-    {     39, 8}, {     79, 9}, {     47,10}, {     31, 9}, \
-    {     79,10}, {     47,11}, {     31,10}, {     63, 9}, \
-    {    127, 8}, {    255,10}, {     79, 8}, {    319, 9}, \
-    {    175, 8}, {    351, 7}, {    703, 6}, {   1407,10}, \
-    {     95, 9}, {    191, 8}, {    383, 9}, {    207, 8}, \
-    {    415, 7}, {    831, 9}, {    223,11}, {     63,10}, \
-    {    127, 9}, {    255,10}, {    143, 9}, {    287, 8}, \
-    {    575, 9}, {    319,10}, {    175, 9}, {    351, 8}, \
-    {    703, 7}, {   1407,11}, {     95,10}, {    191, 9}, \
-    {    383,10}, {    207, 9}, {    415, 8}, {    831,12}, \
-    {     63,11}, {    127,10}, {    255, 9}, {    575,11}, \
-    {    159,10}, {    351, 9}, {    703, 8}, {   1407,11}, \
-    {    191,10}, {    415, 9}, {    831,11}, {    223,10}, \
-    {    447, 9}, {    895, 8}, {   1791,10}, {    479, 9}, \
-    {    959,12}, {    127,11}, {    287,10}, {    575,11}, \
-    {    351,10}, {    703, 9}, {   1407,12}, {    191,11}, \
-    {    415,10}, {    831,11}, {    447,10}, {    895, 9}, \
-    {   1791,11}, {    479,10}, {    959,13}, {    127,12}, \
-    {    255,11}, {    575,12}, {    319,11}, {    703,10}, \
-    {   1407,12}, {    383,11}, {    831,12}, {    447,11}, \
-    {    895,10}, {   1791,11}, {    959,10}, {   1919, 9}, \
-    {   3839,13}, {    255,12}, {    575,11}, {   1215,10}, \
-    {   2431,12}, {    703,11}, {   1407,13}, {    383,12}, \
-    {    895,11}, {   1791,12}, {    959,11}, {   1919,10}, \
-    {   3839,14}, {    255,13}, {    511,12}, {   1215,11}, \
-    {   2431,13}, {    639,12}, {   1471,13}, {    895,12}, \
-    {   1919,11}, {   3839,10}, {   7679,14}, {    511,13}, \
-    {   1023,12}, {   2047,13}, {   1151,12}, {   2431,13}, \
-    {   1407,14}, {    767,13}, {   8192,14}, {  16384,15}, \
+  { {    276, 5}, {     15, 6}, {      8, 5}, {     17, 6}, \
+    {      9, 5}, {     19, 6}, {     13, 7}, {      7, 6}, \
+    {     17, 7}, {      9, 6}, {     19, 7}, {     11, 6}, \
+    {     23, 7}, {     13, 8}, {      7, 7}, {     19, 8}, \
+    {     11, 7}, {     25, 8}, {     15, 7}, {     31, 8}, \
+    {     19, 7}, {     39, 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, 7}, {    511, 9}, {    143,10}, {     79, 9}, \
+    {    159, 8}, {    319, 9}, {    175, 8}, {    351,10}, \
+    {     95, 9}, {    191, 8}, {    383,11}, {     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}, \
+    {    383,12}, {     63,11}, {    127,10}, {    255, 9}, \
+    {    511,10}, {    271, 9}, {    543, 8}, {   1087,10}, \
+    {    287, 9}, {    575,11}, {    159,10}, {    351, 9}, \
+    {    703, 8}, {   1407,11}, {    191,10}, {    415, 9}, \
+    {    831,11}, {    223,10}, {    479, 9}, {    959, 8}, \
+    {   1919,12}, {   4096,13}, {   8192,14}, {  16384,15}, \
     {  32768,16} }
-#define MUL_FFT_TABLE3_SIZE 149
-#define MUL_FFT_THRESHOLD                 2240
+#define MUL_FFT_TABLE3_SIZE 89
+#define MUL_FFT_THRESHOLD                 2688
 
-#define SQR_FFT_MODF_THRESHOLD             244  /* k = 5 */
+#define SQR_FFT_MODF_THRESHOLD             240  /* k = 5 */
 #define SQR_FFT_TABLE3                                      \
-  { {    244, 5}, {     11, 6}, {      6, 5}, {     13, 6}, \
-    {      7, 5}, {     15, 6}, {      8, 5}, {     17, 6}, \
-    {     11, 5}, {     23, 6}, {     13, 7}, {      7, 6}, \
-    {     16, 7}, {      9, 6}, {     19, 7}, {     11, 6}, \
-    {     23, 7}, {     13, 8}, {      7, 7}, {     19, 8}, \
-    {     11, 7}, {     23, 9}, {      7, 8}, {     15, 7}, \
-    {     31, 8}, {     19, 7}, {     39, 8}, {     23, 9}, \
-    {     15, 8}, {     39, 9}, {     23,10}, {     15, 9}, \
-    {     31, 8}, {     63, 9}, {     39, 8}, {     79, 9}, \
-    {     47,10}, {     31, 9}, {     63, 8}, {    127, 9}, \
-    {     71, 8}, {    143, 7}, {    287,10}, {     47,11}, \
-    {     31,10}, {     63, 9}, {    127, 8}, {    255, 9}, \
-    {    143, 8}, {    287,10}, {     79, 9}, {    159, 8}, \
-    {    319, 9}, {    175, 8}, {    351, 7}, {    703, 6}, \
-    {   1407,10}, {     95, 9}, {    191, 8}, {    383,11}, \
-    {     63,10}, {    127, 9}, {    255,10}, {    143, 9}, \
+  { {    240, 5}, {     17, 6}, {     17, 7}, {      9, 6}, \
+    {     19, 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}, \
+    {    143,10}, {     47,11}, {     31,10}, {     63, 9}, \
+    {    127, 8}, {    255, 7}, {    511, 9}, {    143,10}, \
+    {     79, 9}, {    159, 8}, {    319, 9}, {    175, 8}, \
+    {    351, 7}, {    703,10}, {     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, 8}, {    703, 7}, {   1407,11}, \
-    {     95,10}, {    191, 9}, {    383,10}, {    207,12}, \
-    {     63,11}, {    127,10}, {    255, 9}, {    511,10}, \
-    {    287, 9}, {    575,11}, {    159,10}, {    351, 9}, \
-    {    703, 8}, {   1407,11}, {    191,10}, {    415,11}, \
-    {    223,10}, {    447, 9}, {    895, 8}, {   1791,10}, \
-    {    479, 9}, {    959,12}, {    127,11}, {    255,10}, \
-    {    511,11}, {    287,10}, {    575,11}, {    319,10}, \
-    {    639,11}, {    351,10}, {    703, 9}, {   1407,12}, \
-    {    191,11}, {    415,10}, {    831,11}, {    447,10}, \
-    {    895, 9}, {   1791,11}, {    479,13}, {    127,12}, \
-    {    255,11}, {    575,12}, {    319,11}, {    703,10}, \
-    {   1407,12}, {    383,11}, {    831,12}, {    447,11}, \
-    {    895,10}, {   1791,11}, {    959,10}, {   1919,13}, \
-    {    255,12}, {    511,11}, {   1023,12}, {    575,11}, \
-    {   1215,10}, {   2431,12}, {    703,11}, {   1407,13}, \
-    {    383,12}, {    895,11}, {   1791,12}, {    959,11}, \
-    {   1919,14}, {    255,13}, {    511,12}, {   1215,11}, \
-    {   2431,13}, {    639,12}, {   1471,11}, {   2943,10}, \
-    {   5887,13}, {    895,12}, {   1919,11}, {   3839,10}, \
-    {   7679,14}, {    511,13}, {   1023,12}, {   2047,13}, \
-    {   1151,12}, {   2431,13}, {   1407,12}, {   2943,11}, \
-    {   5887,14}, {    767,13}, {   8192,14}, {  16384,15}, \
-    {  32768,16} }
-#define SQR_FFT_TABLE3_SIZE 161
-#define SQR_FFT_THRESHOLD                 1728
+    {     95,10}, {    191, 9}, {    383,10}, {    207, 9}, \
+    {    415,12}, {     63,11}, {    127,10}, {    255, 9}, \
+    {    511,10}, {    287, 9}, {    575,11}, {    159,10}, \
+    {    351, 9}, {    703, 8}, {   1407,11}, {    191,10}, \
+    {    415, 9}, {    831,11}, {    223,10}, {    479,12}, \
+    {   4096,13}, {   8192,14}, {  16384,15}, {  32768,16} }
+#define SQR_FFT_TABLE3_SIZE 84
+#define SQR_FFT_THRESHOLD                 1856
 
 #define MULLO_BASECASE_THRESHOLD             0  /* always */
-#define MULLO_DC_THRESHOLD                  33
-#define MULLO_MUL_N_THRESHOLD             4392
-
-#define DC_DIV_QR_THRESHOLD                 43
-#define DC_DIVAPPR_Q_THRESHOLD             150
-#define DC_BDIV_QR_THRESHOLD                38
-#define DC_BDIV_Q_THRESHOLD                107
-
-#define INV_MULMOD_BNM1_THRESHOLD           14
-#define INV_NEWTON_THRESHOLD               165
-#define INV_APPR_THRESHOLD                 149
-
-#define BINV_NEWTON_THRESHOLD              147
-#define REDC_1_TO_REDC_N_THRESHOLD          43
+#define MULLO_DC_THRESHOLD                  27
+#define MULLO_MUL_N_THRESHOLD             5240
+#define SQRLO_BASECASE_THRESHOLD             0  /* always */
+#define SQRLO_DC_THRESHOLD                  65
+#define SQRLO_SQR_THRESHOLD               3470
 
-#define MU_DIV_QR_THRESHOLD                777
-#define MU_DIVAPPR_Q_THRESHOLD             942
-#define MUPI_DIV_QR_THRESHOLD               69
-#define MU_BDIV_QR_THRESHOLD               654
-#define MU_BDIV_Q_THRESHOLD                777
-
-#define POWM_SEC_TABLE  3,32,126,692,1486
+#define DC_DIV_QR_THRESHOLD                 32
+#define DC_DIVAPPR_Q_THRESHOLD             135
+#define DC_BDIV_QR_THRESHOLD                32
+#define DC_BDIV_Q_THRESHOLD                 80
 
-#define MATRIX22_STRASSEN_THRESHOLD         17
-#define HGCD_THRESHOLD                     103
-#define HGCD_APPR_THRESHOLD                144
-#define HGCD_REDUCE_THRESHOLD             1437
-#define GCD_DC_THRESHOLD                   275
-#define GCDEXT_DC_THRESHOLD                206
+#define INV_MULMOD_BNM1_THRESHOLD           42
+#define INV_NEWTON_THRESHOLD               177
+#define INV_APPR_THRESHOLD                 139
+
+#define BINV_NEWTON_THRESHOLD              179
+#define REDC_1_TO_REDC_N_THRESHOLD          39
+
+#define MU_DIV_QR_THRESHOLD                872
+#define MU_DIVAPPR_Q_THRESHOLD             998
+#define MUPI_DIV_QR_THRESHOLD               66
+#define MU_BDIV_QR_THRESHOLD               748
+#define MU_BDIV_Q_THRESHOLD                906
+
+#define POWM_SEC_TABLE  9,34,257,946,2913


More information about the gmp-commit mailing list