diff options
author | Jeff Johnston <jjohnstn@redhat.com> | 2008-09-04 21:46:14 +0400 |
---|---|---|
committer | Jeff Johnston <jjohnstn@redhat.com> | 2008-09-04 21:46:14 +0400 |
commit | 06e514022dd5d284b85c26fe1b4ef8bd453bc204 (patch) | |
tree | 74a59eac0ebcbed39931201ebac83c6c518c0fae /newlib/libm/machine | |
parent | db04da927987ae655176a26f3e1472a54eb60a92 (diff) |
2008-09-04 Ken Werner <ken.werner@de.ibm.com>
* libm/machine/spu/headers/acoshf4.h: Rebase against current simdmath.
* libm/machine/spu/headers/asinhd2.h: Likewise.
* libm/machine/spu/headers/atanhd2.h: Likewise.
* libm/machine/spu/headers/atanhf4.h: Likewise.
* libm/machine/spu/headers/erff4.h: Likewise.
* libm/machine/spu/headers/expd2.h: Likewise.
* libm/machine/spu/headers/ldexpd2.h: Likewise.
* libm/machine/spu/headers/lgammaf4.h: Likewise.
* libm/machine/spu/headers/logbf4.h: Likewise.
* libm/machine/spu/headers/nextafterd2.h: Likewise.
* libm/machine/spu/headers/nextafterf4.h: Likewise.
* libm/machine/spu/headers/recipd2.h: Likewise.
* libm/machine/spu/headers/simdmath.h: Likewise.
* libm/machine/spu/headers/acoshd2.: Likewise.
Diffstat (limited to 'newlib/libm/machine')
-rw-r--r-- | newlib/libm/machine/spu/headers/acoshd2.h | 93 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/acoshf4.h | 73 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/asinhd2.h | 56 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/atanhd2.h | 79 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/atanhf4.h | 53 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/erff4.h | 355 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/expd2.h | 67 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/ldexpd2.h | 5 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/lgammaf4.h | 495 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/logbf4.h | 19 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/nextafterd2.h | 15 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/nextafterf4.h | 4 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/recipd2.h | 23 | ||||
-rw-r--r-- | newlib/libm/machine/spu/headers/simdmath.h | 9 |
14 files changed, 989 insertions, 357 deletions
diff --git a/newlib/libm/machine/spu/headers/acoshd2.h b/newlib/libm/machine/spu/headers/acoshd2.h index b9bfd0cb2..ead310e57 100644 --- a/newlib/libm/machine/spu/headers/acoshd2.h +++ b/newlib/libm/machine/spu/headers/acoshd2.h @@ -91,79 +91,78 @@ * Taylor Series Coefficients * for x around 1. */ -#define ACOSH_TAY01 1.0000000000000000000000000000000000000000000000000000000000000000000000E0 /* 1 / 1 */ -#define ACOSH_TAY02 -8.3333333333333333333333333333333333333333333333333333333333333333333333E-2 /* 1 / 12 */ -#define ACOSH_TAY03 1.8750000000000000000000000000000000000000000000000000000000000000000000E-2 /* 3 / 160 */ -#define ACOSH_TAY04 -5.5803571428571428571428571428571428571428571428571428571428571428571429E-3 /* 5 / 896 */ -#define ACOSH_TAY05 1.8988715277777777777777777777777777777777777777777777777777777777777778E-3 /* 35 / 18432 */ -#define ACOSH_TAY06 -6.9912997159090909090909090909090909090909090909090909090909090909090909E-4 /* 63 / 90112 */ -#define ACOSH_TAY07 2.7113694411057692307692307692307692307692307692307692307692307692307692E-4 /* 231 / 851968 */ -#define ACOSH_TAY08 -1.0910034179687500000000000000000000000000000000000000000000000000000000E-4 /* 143 / 1310720 */ -#define ACOSH_TAY09 4.5124222250545726102941176470588235294117647058823529411764705882352941E-5 /* 6435 / 142606336 */ -#define ACOSH_TAY10 -1.9065643611707185444078947368421052631578947368421052631578947368421053E-5 /* 12155 / 637534208 */ -#define ACOSH_TAY11 8.1936873140789213634672619047619047619047619047619047619047619047619048E-6 /* 46189 / 5637144576 */ -#define ACOSH_TAY12 -3.5705692742181860882302989130434782608695652173913043478260869565217391E-6 /* 88179 / 24696061952 */ -#define ACOSH_TAY13 1.5740259550511837005615234375000000000000000000000000000000000000000000E-6 /* 676039 / 429496729600 */ -#define ACOSH_TAY14 -7.0068819224144573564882631655092592592592592592592592592592592592592593E-7 /* 1300075 / 1855425871872 */ -#define ACOSH_TAY15 3.1453306166503321507881427633351293103448275862068965517241379310344828E-7 /* 5014575 / 15942918602752 */ -#if 0 -#define ACOSH_TAY16 -1.4221629293564136230176494967552923387096774193548387096774193548387097E-7 /* 9694845 / 68169720922112 */ -#define ACOSH_TAY17 6.4711106776113328206437555226412686434659090909090909090909090909090909E-8 /* 100180065 / 1548112371908608 */ -#define ACOSH_TAY18 -2.9609409781171182528071637664522443498883928571428571428571428571428571E-8 /* 116680311 / 3940649673949184 */ -#define ACOSH_TAY19 1.3615438056281793767600509061201198680980785472972972972972972972972973E-8 /* 2268783825 / 166633186212708352 */ -#endif +#define SDM_ACOSHD2_TAY01 1.000000000000000000000000000000000E0 /* 1 / 1 */ +#define SDM_ACOSHD2_TAY02 -8.333333333333333333333333333333333E-2 /* 1 / 12 */ +#define SDM_ACOSHD2_TAY03 1.875000000000000000000000000000000E-2 /* 3 / 160 */ +#define SDM_ACOSHD2_TAY04 -5.580357142857142857142857142857142E-3 /* 5 / 896 */ +#define SDM_ACOSHD2_TAY05 1.898871527777777777777777777777777E-3 /* 35 / 18432 */ +#define SDM_ACOSHD2_TAY06 -6.991299715909090909090909090909090E-4 /* 63 / 90112 */ +#define SDM_ACOSHD2_TAY07 2.711369441105769230769230769230769E-4 /* 231 / 851968 */ +#define SDM_ACOSHD2_TAY08 -1.091003417968750000000000000000000E-4 /* 143 / 1310720 */ +#define SDM_ACOSHD2_TAY09 4.512422225054572610294117647058823E-5 /* 6435 / 142606336 */ +#define SDM_ACOSHD2_TAY10 -1.906564361170718544407894736842105E-5 /* 12155 / 637534208 */ +#define SDM_ACOSHD2_TAY11 8.193687314078921363467261904761904E-6 /* 46189 / 5637144576 */ +#define SDM_ACOSHD2_TAY12 -3.570569274218186088230298913043478E-6 /* 88179 / 24696061952 */ +#define SDM_ACOSHD2_TAY13 1.574025955051183700561523437500000E-6 /* 676039 / 429496729600 */ +#define SDM_ACOSHD2_TAY14 -7.006881922414457356488263165509259E-7 /* 1300075 / 1855425871872 */ +#define SDM_ACOSHD2_TAY15 3.145330616650332150788142763335129E-7 /* 5014575 / 15942918602752 */ static __inline vector double _acoshd2(vector double x) { vec_uchar16 dup_even = ((vec_uchar16) { 0,1,2,3, 0,1,2,3, 8,9,10,11, 8,9,10,11 }); vec_double2 minus_oned = spu_splats(-1.0); vec_double2 twod = spu_splats(2.0); - vec_double2 xminus1; - vec_float4 xf; /* Where we switch from taylor to formula */ vec_float4 switch_approx = spu_splats(1.15f); - vec_uint4 use_form; vec_double2 result, fresult, mresult;; - xf = spu_roundtf(x); + vec_double2 xminus1 = spu_add(x, minus_oned); + vec_float4 xf = spu_roundtf(x); xf = spu_shuffle(xf, xf, dup_even); + vec_ullong2 use_form = (vec_ullong2)spu_cmpgt(xf, switch_approx); + + vec_double2 sqrtargformula = spu_madd(x, x, minus_oned); + vec_double2 sqrtargtaylor = spu_mul(xminus1, twod); + vec_double2 sqrtarg = spu_sel(sqrtargtaylor, sqrtargformula, use_form); + + vec_double2 sqrtresult = _sqrtd2(sqrtarg); + /* * Formula: * acosh = ln(x + sqrt(x^2 - 1)) */ - fresult = _sqrtd2(spu_madd(x, x, minus_oned)); - fresult = spu_add(x, fresult); + fresult = spu_add(x, sqrtresult); fresult = _logd2(fresult); /* * Taylor Series */ - xminus1 = spu_add(x, minus_oned); - - mresult = spu_madd(xminus1, spu_splats(ACOSH_TAY15), spu_splats(ACOSH_TAY14)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY13)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY12)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY11)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY10)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY09)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY08)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY07)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY06)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY05)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY04)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY03)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY02)); - mresult = spu_madd(xminus1, mresult, spu_splats(ACOSH_TAY01)); - - mresult = spu_mul(mresult, _sqrtd2(spu_mul(xminus1, twod))); + mresult = spu_splats(SDM_ACOSHD2_TAY15); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY14)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY13)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY12)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY11)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY10)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY09)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY08)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY07)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY06)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY05)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY04)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY03)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY02)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHD2_TAY01)); + + + mresult = spu_mul(mresult, sqrtresult); + /* * Select series or formula */ - use_form = spu_cmpgt(xf, switch_approx); - result = spu_sel(mresult, fresult, (vec_ullong2)use_form); + result = spu_sel(mresult, fresult, use_form); return result; } diff --git a/newlib/libm/machine/spu/headers/acoshf4.h b/newlib/libm/machine/spu/headers/acoshf4.h index 3efc35ffc..5cf94d8e5 100644 --- a/newlib/libm/machine/spu/headers/acoshf4.h +++ b/newlib/libm/machine/spu/headers/acoshf4.h @@ -89,26 +89,26 @@ * Taylor Series Coefficients * for x around 1. */ -#define ACOSH_TAY01 1.0000000000000000000000000000000000000000000000000000000000000000000000E0 /* 1 / 1 */ -#define ACOSH_TAY02 -8.3333333333333333333333333333333333333333333333333333333333333333333333E-2 /* 1 / 12 */ -#define ACOSH_TAY03 1.8750000000000000000000000000000000000000000000000000000000000000000000E-2 /* 3 / 160 */ -#define ACOSH_TAY04 -5.5803571428571428571428571428571428571428571428571428571428571428571429E-3 /* 5 / 896 */ -#define ACOSH_TAY05 1.8988715277777777777777777777777777777777777777777777777777777777777778E-3 /* 35 / 18432 */ -#define ACOSH_TAY06 -6.9912997159090909090909090909090909090909090909090909090909090909090909E-4 /* 63 / 90112 */ -#define ACOSH_TAY07 2.7113694411057692307692307692307692307692307692307692307692307692307692E-4 /* 231 / 851968 */ -#define ACOSH_TAY08 -1.0910034179687500000000000000000000000000000000000000000000000000000000E-4 /* 143 / 1310720 */ -#define ACOSH_TAY09 4.5124222250545726102941176470588235294117647058823529411764705882352941E-5 /* 6435 / 142606336 */ -#define ACOSH_TAY10 -1.9065643611707185444078947368421052631578947368421052631578947368421053E-5 /* 12155 / 637534208 */ -#define ACOSH_TAY11 8.1936873140789213634672619047619047619047619047619047619047619047619048E-6 /* 46189 / 5637144576 */ -#define ACOSH_TAY12 -3.5705692742181860882302989130434782608695652173913043478260869565217391E-6 /* 88179 / 24696061952 */ -#define ACOSH_TAY13 1.5740259550511837005615234375000000000000000000000000000000000000000000E-6 /* 676039 / 429496729600 */ -#define ACOSH_TAY14 -7.0068819224144573564882631655092592592592592592592592592592592592592593E-7 /* 1300075 / 1855425871872 */ -#define ACOSH_TAY15 3.1453306166503321507881427633351293103448275862068965517241379310344828E-7 /* 5014575 / 15942918602752 */ +#define SDM_ACOSHF4_TAY01 1.00000000000000000000000000000000000E0f /* 1 / 1 */ +#define SDM_ACOSHF4_TAY02 -8.33333333333333333333333333333333333E-2f /* 1 / 12 */ +#define SDM_ACOSHF4_TAY03 1.87500000000000000000000000000000000E-2f /* 3 / 160 */ +#define SDM_ACOSHF4_TAY04 -5.58035714285714285714285714285714286E-3f /* 5 / 896 */ +#define SDM_ACOSHF4_TAY05 1.89887152777777777777777777777777778E-3f /* 35 / 18432 */ +#define SDM_ACOSHF4_TAY06 -6.99129971590909090909090909090909091E-4f /* 63 / 90112 */ +#define SDM_ACOSHF4_TAY07 2.71136944110576923076923076923076923E-4f /* 231 / 851968 */ +#define SDM_ACOSHF4_TAY08 -1.09100341796875000000000000000000000E-4f /* 143 / 1310720 */ +#define SDM_ACOSHF4_TAY09 4.51242222505457261029411764705882353E-5f /* 6435 / 142606336 */ +#define SDM_ACOSHF4_TAY10 -1.90656436117071854440789473684210526E-5f /* 12155 / 637534208 */ +#define SDM_ACOSHF4_TAY11 8.19368731407892136346726190476190476E-6f /* 46189 / 5637144576 */ +#define SDM_ACOSHF4_TAY12 -3.57056927421818608823029891304347826E-6f /* 88179 / 24696061952 */ +#define SDM_ACOSHF4_TAY13 1.57402595505118370056152343750000000E-6f /* 676039 / 429496729600 */ +#define SDM_ACOSHF4_TAY14 -7.00688192241445735648826316550925926E-7f /* 1300075 / 1855425871872 */ +#define SDM_ACOSHF4_TAY15 3.14533061665033215078814276333512931E-7f /* 5014575 / 15942918602752 */ #if 0 -#define ACOSH_TAY16 -1.4221629293564136230176494967552923387096774193548387096774193548387097E-7 /* 9694845 / 68169720922112 */ -#define ACOSH_TAY17 6.4711106776113328206437555226412686434659090909090909090909090909090909E-8 /* 100180065 / 1548112371908608 */ -#define ACOSH_TAY18 -2.9609409781171182528071637664522443498883928571428571428571428571428571E-8 /* 116680311 / 3940649673949184 */ -#define ACOSH_TAY19 1.3615438056281793767600509061201198680980785472972972972972972972972973E-8 /* 2268783825 / 166633186212708352 */ +#define SDM_ACOSHF4_TAY16 -1.42216292935641362301764949675529234E-7f /* 9694845 / 68169720922112 */ +#define SDM_ACOSHF4_TAY17 6.47111067761133282064375552264126864E-8f /* 100180065 / 1548112371908608 */ +#define SDM_ACOSHF4_TAY18 -2.96094097811711825280716376645224435E-8f /* 116680311 / 3940649673949184 */ +#define SDM_ACOSHF4_TAY19 1.36154380562817937676005090612011987E-8f /* 2268783825 / 166633186212708352 */ #endif @@ -117,6 +117,7 @@ static __inline vector float _acoshf4(vector float x) { vec_float4 minus_onef = spu_splats(-1.0f); vec_float4 twof = spu_splats(2.0f); + vec_float4 largef = spu_splats(2.5e19f); vec_float4 xminus1; /* Where we switch from taylor to formula */ vec_float4 switch_approx = spu_splats(2.0f); @@ -129,30 +130,33 @@ static __inline vector float _acoshf4(vector float x) * acosh = ln(x + sqrt(x^2 - 1)) */ fresult = _sqrtf4(spu_madd(x, x, minus_onef)); - fresult = spu_add(x, fresult); + fresult = spu_add(x, spu_sel(fresult, x, spu_cmpgt(x, largef))); fresult = _logf4(fresult); + fresult = (vec_float4)spu_add((vec_uint4)fresult, spu_splats(2u)); /* * Taylor Series */ xminus1 = spu_add(x, minus_onef); - mresult = spu_madd(xminus1, spu_splats((float)ACOSH_TAY15), spu_splats((float)ACOSH_TAY14)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY13)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY12)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY11)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY10)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY09)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY08)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY07)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY06)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY05)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY04)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY03)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY02)); - mresult = spu_madd(xminus1, mresult, spu_splats((float)ACOSH_TAY01)); + mresult = spu_splats(SDM_ACOSHF4_TAY15); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY14)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY13)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY12)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY11)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY10)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY09)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY08)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY07)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY06)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY05)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY04)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY03)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY02)); + mresult = spu_madd(xminus1, mresult, spu_splats(SDM_ACOSHF4_TAY01)); mresult = spu_mul(mresult, _sqrtf4(spu_mul(xminus1, twof))); + mresult = (vec_float4)spu_add((vec_uint4)mresult, spu_splats(1u)); /* * Select series or formula @@ -160,6 +164,7 @@ static __inline vector float _acoshf4(vector float x) use_form = spu_cmpgt(x, switch_approx); result = spu_sel(mresult, fresult, use_form); + return result; } diff --git a/newlib/libm/machine/spu/headers/asinhd2.h b/newlib/libm/machine/spu/headers/asinhd2.h index 542d59858..f5fd6f6f7 100644 --- a/newlib/libm/machine/spu/headers/asinhd2.h +++ b/newlib/libm/machine/spu/headers/asinhd2.h @@ -93,43 +93,27 @@ * Maclaurin Series Coefficients * for x near 0. */ -#define ASINH_MAC01 1.0000000000000000000000000000000000000000000000000000000000000000000000E0 -#define ASINH_MAC03 -1.6666666666666666666666666666666666666666666666666666666666666666666667E-1 -#define ASINH_MAC05 7.5000000000000000000000000000000000000000000000000000000000000000000000E-2 -#define ASINH_MAC07 -4.4642857142857142857142857142857142857142857142857142857142857142857143E-2 -#define ASINH_MAC09 3.0381944444444444444444444444444444444444444444444444444444444444444444E-2 -#define ASINH_MAC11 -2.2372159090909090909090909090909090909090909090909090909090909090909091E-2 -#define ASINH_MAC13 1.7352764423076923076923076923076923076923076923076923076923076923076923E-2 -#define ASINH_MAC15 -1.3964843750000000000000000000000000000000000000000000000000000000000000E-2 -#define ASINH_MAC17 1.1551800896139705882352941176470588235294117647058823529411764705882353E-2 -#if 0 -#define ASINH_MAC19 -9.7616095291940789473684210526315789473684210526315789473684210526315789E-3 -#define ASINH_MAC21 8.3903358096168154761904761904761904761904761904761904761904761904761905E-3 -#define ASINH_MAC23 -7.3125258735988451086956521739130434782608695652173913043478260869565217E-3 -#define ASINH_MAC25 6.4472103118896484375000000000000000000000000000000000000000000000000000E-3 -#define ASINH_MAC27 -5.7400376708419234664351851851851851851851851851851851851851851851851852E-3 -#define ASINH_MAC29 5.1533096823199041958512931034482758620689655172413793103448275862068966E-3 -#define ASINH_MAC31 -4.6601434869150961599042338709677419354838709677419354838709677419354839E-3 -#define ASINH_MAC33 4.2409070936793630773370916193181818181818181818181818181818181818181818E-3 -#define ASINH_MAC35 -3.8809645588376692363194056919642857142857142857142857142857142857142857E-3 -#define ASINH_MAC37 3.5692053938259345454138678473395270270270270270270270270270270270270270E-3 -#define ASINH_MAC39 -3.2970595034734847453924325796274038461538461538461538461538461538461538E-3 -#define ASINH_MAC41 3.0578216492580306693548109473251714939024390243902439024390243902439024E-3 -#define ASINH_MAC43 -2.8461784011089421678767647854117460029069767441860465116279069767441860E-3 -#endif +#define SDM_ASINHD2_MAC01 1.000000000000000000000000000000000000000000E0 +#define SDM_ASINHD2_MAC03 -1.666666666666666666666666666666666666666667E-1 +#define SDM_ASINHD2_MAC05 7.500000000000000000000000000000000000000000E-2 +#define SDM_ASINHD2_MAC07 -4.464285714285714285714285714285714285714286E-2 +#define SDM_ASINHD2_MAC09 3.038194444444444444444444444444444444444444E-2 +#define SDM_ASINHD2_MAC11 -2.237215909090909090909090909090909090909091E-2 +#define SDM_ASINHD2_MAC13 1.735276442307692307692307692307692307692308E-2 +#define SDM_ASINHD2_MAC15 -1.396484375000000000000000000000000000000000E-2 +#define SDM_ASINHD2_MAC17 1.155180089613970588235294117647058823529412E-2 static __inline vector double _asinhd2(vector double x) { vec_double2 sign_mask = spu_splats(-0.0); vec_double2 oned = spu_splats(1.0); - vec_uchar16 dup_even = ((vec_uchar16) { 0,1,2,3, 0,1,2,3, 8,9,10,11, 8,9,10,11 }); + vec_uchar16 dup_even = ((vec_uchar16) { 0,1,2,3, 0,1,2,3, 8,9,10,11, 8,9,10,11 }); vec_uint4 infminus1 = spu_splats(0x7FEFFFFFU); vec_uint4 isinfnan; vec_double2 xabs, xsqu; vec_uint4 xabshigh; - /* Where we switch from maclaurin to formula */ - vec_float4 switch_approx = spu_splats(0.165f); + vec_float4 switch_approx = spu_splats(0.165f); /* Where we switch from maclaurin to formula */ vec_uint4 use_form; vec_float4 xf; vec_double2 result, fresult, mresult; @@ -153,14 +137,16 @@ static __inline vector double _asinhd2(vector double x) /* * Maclaurin Series approximation */ - mresult = spu_madd(xsqu, spu_splats(ASINH_MAC17), spu_splats(ASINH_MAC15)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC13)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC11)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC09)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC07)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC05)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC03)); - mresult = spu_madd(xsqu, mresult, spu_splats(ASINH_MAC01)); + + mresult = spu_splats(SDM_ASINHD2_MAC17); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC15)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC13)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC11)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC09)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC07)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC05)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC03)); + mresult = spu_madd(xsqu, mresult, spu_splats(SDM_ASINHD2_MAC01)); mresult = spu_mul(xabs, mresult); diff --git a/newlib/libm/machine/spu/headers/atanhd2.h b/newlib/libm/machine/spu/headers/atanhd2.h index b1bc33ef9..ab167d956 100644 --- a/newlib/libm/machine/spu/headers/atanhd2.h +++ b/newlib/libm/machine/spu/headers/atanhd2.h @@ -72,7 +72,7 @@ * Special Cases: * - atanh(1) = Infinity * - atanh(-1) = -Infinity - * - atanh(x) for |x| > 1 = NaN + * - atanh(x) for |x| > 1 = Undefined * */ @@ -80,37 +80,27 @@ * Maclaurin Series Coefficients * for x near 0. */ -#define ATANH_MAC01 1.0000000000000000000000000000000000000000000000000000000000000000000000E0 -#define ATANH_MAC03 3.3333333333333333333333333333333333333333333333333333333333333333333333E-1 -#define ATANH_MAC05 2.0000000000000000000000000000000000000000000000000000000000000000000000E-1 -#define ATANH_MAC07 1.4285714285714285714285714285714285714285714285714285714285714285714286E-1 -#define ATANH_MAC09 1.1111111111111111111111111111111111111111111111111111111111111111111111E-1 -#define ATANH_MAC11 9.0909090909090909090909090909090909090909090909090909090909090909090909E-2 -#define ATANH_MAC13 7.6923076923076923076923076923076923076923076923076923076923076923076923E-2 +#define SMD_DP_ATANH_MAC01 1.000000000000000000000000000000E0 +#define SMD_DP_ATANH_MAC03 3.333333333333333333333333333333E-1 +#define SMD_DP_ATANH_MAC05 2.000000000000000000000000000000E-1 +#define SMD_DP_ATANH_MAC07 1.428571428571428571428571428571E-1 +#define SMD_DP_ATANH_MAC09 1.111111111111111111111111111111E-1 +#define SMD_DP_ATANH_MAC11 9.090909090909090909090909090909E-2 +#define SMD_DP_ATANH_MAC13 7.692307692307692307692307692308E-2 +#define SMD_DP_ATANH_MAC15 6.666666666666666666666666666667E-2 +#define SMD_DP_ATANH_MAC17 5.882352941176470588235294117647E-2 #if 0 -#define ATANH_MAC15 6.6666666666666666666666666666666666666666666666666666666666666666666667E-2 -#define ATANH_MAC17 5.8823529411764705882352941176470588235294117647058823529411764705882353E-2 -#define ATANH_MAC19 5.2631578947368421052631578947368421052631578947368421052631578947368421E-2 -#define ATANH_MAC21 4.7619047619047619047619047619047619047619047619047619047619047619047619E-2 -#define ATANH_MAC23 4.3478260869565217391304347826086956521739130434782608695652173913043478E-2 -#define ATANH_MAC25 4.0000000000000000000000000000000000000000000000000000000000000000000000E-2 -#define ATANH_MAC27 3.7037037037037037037037037037037037037037037037037037037037037037037037E-2 -#define ATANH_MAC29 3.4482758620689655172413793103448275862068965517241379310344827586206897E-2 -#define ATANH_MAC31 3.2258064516129032258064516129032258064516129032258064516129032258064516E-2 -#define ATANH_MAC33 3.0303030303030303030303030303030303030303030303030303030303030303030303E-2 -#define ATANH_MAC35 2.8571428571428571428571428571428571428571428571428571428571428571428571E-2 -#define ATANH_MAC37 2.7027027027027027027027027027027027027027027027027027027027027027027027E-2 -#define ATANH_MAC39 2.5641025641025641025641025641025641025641025641025641025641025641025641E-2 -#define ATANH_MAC41 2.4390243902439024390243902439024390243902439024390243902439024390243902E-2 -#define ATANH_MAC43 2.3255813953488372093023255813953488372093023255813953488372093023255814E-2 -#define ATANH_MAC45 2.2222222222222222222222222222222222222222222222222222222222222222222222E-2 -#define ATANH_MAC47 2.1276595744680851063829787234042553191489361702127659574468085106382979E-2 -#define ATANH_MAC49 2.0408163265306122448979591836734693877551020408163265306122448979591837E-2 -#define ATANH_MAC51 1.9607843137254901960784313725490196078431372549019607843137254901960784E-2 -#define ATANH_MAC53 1.8867924528301886792452830188679245283018867924528301886792452830188679E-2 -#define ATANH_MAC55 1.8181818181818181818181818181818181818181818181818181818181818181818182E-2 -#define ATANH_MAC57 1.7543859649122807017543859649122807017543859649122807017543859649122807E-2 -#define ATANH_MAC59 1.6949152542372881355932203389830508474576271186440677966101694915254237E-2 +#define SMD_DP_ATANH_MAC19 5.263157894736842105263157894737E-2 +#define SMD_DP_ATANH_MAC21 4.761904761904761904761904761905E-2 +#define SMD_DP_ATANH_MAC23 4.347826086956521739130434782609E-2 +#define SMD_DP_ATANH_MAC25 4.000000000000000000000000000000E-2 +#define SMD_DP_ATANH_MAC27 3.703703703703703703703703703704E-2 +#define SMD_DP_ATANH_MAC29 3.448275862068965517241379310345E-2 +#define SMD_DP_ATANH_MAC31 3.225806451612903225806451612903E-2 +#define SMD_DP_ATANH_MAC33 3.030303030303030303030303030303E-2 +#define SMD_DP_ATANH_MAC35 2.857142857142857142857142857143E-2 +#define SMD_DP_ATANH_MAC37 2.702702702702702702702702702703E-2 +#define SMD_DP_ATANH_MAC39 2.564102564102564102564102564103E-2 #endif @@ -120,12 +110,9 @@ static __inline vector double _atanhd2(vector double x) vec_double2 sign_mask = spu_splats(-0.0); vec_double2 oned = spu_splats(1.0); vec_double2 onehalfd = spu_splats(0.5); - vec_uint4 infminus1 = spu_splats(0x7FEFFFFFU); - vec_uint4 isinfnan; - vec_uint4 xabshigh; vec_double2 xabs, xsqu; /* Where we switch from maclaurin to formula */ - vec_float4 switch_approx = spu_splats(0.08f); + vec_float4 switch_approx = spu_splats(0.125f); vec_uint4 use_form; vec_float4 xf; vec_double2 result, fresult, mresult;; @@ -147,12 +134,14 @@ static __inline vector double _atanhd2(vector double x) /* * Taylor Series */ - mresult = spu_madd(xsqu, spu_splats(ATANH_MAC13), spu_splats(ATANH_MAC11)); - mresult = spu_madd(xsqu, mresult, spu_splats(ATANH_MAC09)); - mresult = spu_madd(xsqu, mresult, spu_splats(ATANH_MAC07)); - mresult = spu_madd(xsqu, mresult, spu_splats(ATANH_MAC05)); - mresult = spu_madd(xsqu, mresult, spu_splats(ATANH_MAC03)); - mresult = spu_madd(xsqu, mresult, spu_splats(ATANH_MAC01)); + mresult = spu_madd(xsqu, spu_splats(SMD_DP_ATANH_MAC17), spu_splats(SMD_DP_ATANH_MAC15)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC13)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC11)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC09)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC07)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC05)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC03)); + mresult = spu_madd(xsqu, mresult, spu_splats(SMD_DP_ATANH_MAC01)); mresult = spu_mul(xabs, mresult); @@ -162,10 +151,10 @@ static __inline vector double _atanhd2(vector double x) use_form = spu_cmpgt(xf, switch_approx); result = spu_sel(mresult, fresult, (vec_ullong2)use_form); - /* Infinity and NaN */ - xabshigh = (vec_uint4)spu_shuffle(xabs, xabs, dup_even); - isinfnan = spu_cmpgt(xabshigh, infminus1); - result = spu_sel(result, x, (vec_ullong2)isinfnan); + /* + * Spec says results are undefined for |x| > 1, so + * no boundary tests needed here. + */ /* Restore sign - atanh is an anti-symmetric */ result = spu_sel(result, x, (vec_ullong2)sign_mask); diff --git a/newlib/libm/machine/spu/headers/atanhf4.h b/newlib/libm/machine/spu/headers/atanhf4.h index d29ef74d4..11ef786f2 100644 --- a/newlib/libm/machine/spu/headers/atanhf4.h +++ b/newlib/libm/machine/spu/headers/atanhf4.h @@ -52,6 +52,7 @@ #define _ATANHF4_H_ 1 #include <spu_intrinsics.h> +#include <math.h> #include "logf4.h" /* @@ -72,8 +73,7 @@ * Special Cases: * - atanh(1) = HUGE_VALF * - atanh(-1) = -HUGE_VALF - * - The result is undefined for x outside of the domain [-1,1], - * since single-precision NaN is not supported on the SPU. + * - The result is undefined for x outside of the domain [-1,1]. * */ @@ -81,34 +81,28 @@ * Maclaurin Series Coefficients * for x near 0. */ -#define ATANH_MAC01 1.0000000000000000000000000000000000000000000000000000000000000000000000E0 -#define ATANH_MAC03 3.3333333333333333333333333333333333333333333333333333333333333333333333E-1 -#define ATANH_MAC05 2.0000000000000000000000000000000000000000000000000000000000000000000000E-1 -#define ATANH_MAC07 1.4285714285714285714285714285714285714285714285714285714285714285714286E-1 +#define SDM_SP_ATANH_MAC01 1.000000000000000000000000000000E0 +#define SDM_SP_ATANH_MAC03 3.333333333333333333333333333333E-1 +#define SDM_SP_ATANH_MAC05 2.000000000000000000000000000000E-1 +#define SDM_SP_ATANH_MAC07 1.428571428571428571428571428571E-1 #if 0 -#define ATANH_MAC09 1.1111111111111111111111111111111111111111111111111111111111111111111111E-1 -#define ATANH_MAC11 9.0909090909090909090909090909090909090909090909090909090909090909090909E-2 -#define ATANH_MAC13 7.6923076923076923076923076923076923076923076923076923076923076923076923E-2 -#define ATANH_MAC15 6.6666666666666666666666666666666666666666666666666666666666666666666667E-2 -#define ATANH_MAC17 5.8823529411764705882352941176470588235294117647058823529411764705882353E-2 -#define ATANH_MAC19 5.2631578947368421052631578947368421052631578947368421052631578947368421E-2 -#define ATANH_MAC21 4.7619047619047619047619047619047619047619047619047619047619047619047619E-2 -#define ATANH_MAC23 4.3478260869565217391304347826086956521739130434782608695652173913043478E-2 -#define ATANH_MAC25 4.0000000000000000000000000000000000000000000000000000000000000000000000E-2 -#define ATANH_MAC27 3.7037037037037037037037037037037037037037037037037037037037037037037037E-2 -#define ATANH_MAC29 3.4482758620689655172413793103448275862068965517241379310344827586206897E-2 +#define SDM_SP_ATANH_MAC09 1.111111111111111111111111111111E-1 +#define SDM_SP_ATANH_MAC11 9.090909090909090909090909090909E-2 +#define SDM_SP_ATANH_MAC13 7.692307692307692307692307692308E-2 +#define SDM_SP_ATANH_MAC15 6.666666666666666666666666666667E-2 #endif static __inline vector float _atanhf4(vector float x) { + vec_uint4 one = spu_splats(1u); vec_float4 sign_mask = spu_splats(-0.0f); vec_float4 onef = spu_splats(1.0f); vec_float4 onehalff = spu_splats(0.5f); + vec_float4 huge = spu_splats(HUGE_VALF); vec_float4 result, fresult, mresult;; vec_float4 xabs, xsqu; /* Where we switch from maclaurin to formula */ - //vec_float4 switch_approx = spu_splats(0.4661f); vec_float4 switch_approx = spu_splats(0.165f); vec_uint4 use_form; @@ -126,18 +120,33 @@ static __inline vector float _atanhf4(vector float x) /* * Taylor Series */ - mresult = spu_madd(xsqu, spu_splats((float)ATANH_MAC07), spu_splats((float)ATANH_MAC05)); - mresult = spu_madd(xsqu, mresult, spu_splats((float)ATANH_MAC03)); - mresult = spu_madd(xsqu, mresult, spu_splats((float)ATANH_MAC01)); + mresult = spu_madd(xsqu, spu_splats((float)SDM_SP_ATANH_MAC07), spu_splats((float)SDM_SP_ATANH_MAC05)); + mresult = spu_madd(xsqu, mresult, spu_splats((float)SDM_SP_ATANH_MAC03)); + mresult = spu_madd(xsqu, mresult, spu_splats((float)SDM_SP_ATANH_MAC01)); mresult = spu_mul(xabs, mresult); - /* * Choose between series and formula */ use_form = spu_cmpgt(xabs, switch_approx); result = spu_sel(mresult, fresult, use_form); + /* + * Correct for accumulated truncation error. Currently reduces rms of + * absolute error by about 50% + */ + result = (vec_float4)spu_add((vec_uint4)result, spu_and(one, spu_cmpgt(xabs, spu_splats(0.0f)))); + result = (vec_float4)spu_add((vec_uint4)result, spu_and(one, spu_cmpgt(xabs, spu_splats(0.25f)))); + + /* + * Check Boundary Conditions + */ + result = spu_sel(result, huge, spu_cmpeq(xabs, onef)); + + /* + * Spec says |x| > 1, result is undefined, so no additional + * boundary checks needed. + */ /* Preserve sign - atanh is anti-symmetric */ result = spu_sel(result, x, (vec_uint4)sign_mask); diff --git a/newlib/libm/machine/spu/headers/erff4.h b/newlib/libm/machine/spu/headers/erff4.h index eafd9be12..e2eed2301 100644 --- a/newlib/libm/machine/spu/headers/erff4.h +++ b/newlib/libm/machine/spu/headers/erff4.h @@ -53,11 +53,6 @@ #include <spu_intrinsics.h> -#include "expf4.h" -#include "recipf4.h" -#include "divf4.h" -#include "erf_utils.h" - /* * FUNCTION * vector float _erff4(vector float x) @@ -75,54 +70,346 @@ static __inline vector float _erff4(vector float x) { - vector float onehalff = spu_splats(0.5f); - vector float zerof = spu_splats(0.0f); - vector float onef = spu_splats(1.0f); - vector float sign_mask = spu_splats(-0.0f); + vec_float4 sign_maskf = spu_splats(-0.0f); + vec_float4 zerof = spu_splats(0.0f); + vec_float4 onef = spu_splats(1.0f); + vec_float4 clamp = spu_splats(3.9199876f); + vec_float4 xabs = spu_andc(x, sign_maskf); + vec_float4 xsign = spu_and(x, sign_maskf); + vec_float4 result; - /* This is where we switch from Taylor Series to Continued Fraction approximation */ - vec_float4 approx_point = spu_splats(0.89f); - vec_float4 xabs, xsqu, xsign; - vec_uint4 isinf; - vec_float4 tresult, presult, result; + /* + * First thing we do is setup the description of each partition. + * This consists of: + * - Start x of partition + * - Offset (used for evaluating power series expanded around a point) + * - Truncation adjustment. + */ - xsign = spu_and(x, sign_mask); - /* Force Denorms to 0 */ - x = spu_add(x, zerof); + /*************************************************************** + * REGION 0: Approximation Near 0 from Above + * + */ +#define SDM_ERFF4_0_START 0.0f +#define SDM_ERFF4_0_OFF 0.0f +#define SDM_ERFF4_0_TRUNC 2u - xabs = spu_andc(x, sign_mask); - xsqu = spu_mul(x, x); +#define SDM_ERFF4_0_00 0.0f +#define SDM_ERFF4_0_01 1.12837916709551257389615890312154f +#define SDM_ERFF4_0_02 0.0f +#define SDM_ERFF4_0_03 -0.37612638903183752463205296770955f +#define SDM_ERFF4_0_04 0.0f +#define SDM_ERFF4_0_05 0.11283791670955125738961589031073f +#define SDM_ERFF4_0_06 0.0f +#define SDM_ERFF4_0_07 -0.02686617064513125175943235483588f +#define SDM_ERFF4_0_08 0.0f +#define SDM_ERFF4_0_09 0.00522397762544218784211184677371f +#define SDM_ERFF4_0_10 0.0f +//#define SDM_ERFF4_0_11 -0.00085483270234508528325466583569f - /* - * Taylor Series Expansion near Zero + + + /*************************************************************** + * REGION 1: Above 0 and Below 1 */ - TAYLOR_ERFF4(xabs, xsqu, tresult); +#define SDM_ERFF4_1_START 0.07f +#define SDM_ERFF4_1_OFF 0.0625f +#define SDM_ERFF4_1_TRUNC 1u - /* - * Continued Fraction Approximation of Erfc(). - * erf = 1 - erfc +#define SDM_ERFF4_1_00 0.0704319777223870780505900559232967439190042883f +#define SDM_ERFF4_1_01 1.1239800336253906104888456836298420746260842545f +#define SDM_ERFF4_1_02 -0.0702487521015869131555528552268651296641302713f +#define SDM_ERFF4_1_03 -0.3717329798708974154481338589088279778060226856f +#define SDM_ERFF4_1_04 0.0350329063214945152846051348331892508611482993f +#define SDM_ERFF4_1_05 0.1106440713032318617523250293018186620702780982f +#define SDM_ERFF4_1_06 -0.0116471931712158678624014740659716890227703402f +#define SDM_ERFF4_1_07 -0.0261358409084263503958678377968739965222786482f +#define SDM_ERFF4_1_08 0.0029041996223118476954500365511415181291113910f +#define SDM_ERFF4_1_09 0.0050416329596619035812041623972929782386498567f +#define SDM_ERFF4_1_10 -0.0005793225670734356072895029723913210064918149f +//#define SDM_ERFF4_1_11 -0.0008184112733188406359323913130525859730689332f + + + + /*************************************************************** + * REGION 2: + */ +#define SDM_ERFF4_2_START 0.13f +#define SDM_ERFF4_2_OFF 0.1875f +#define SDM_ERFF4_2_TRUNC 1u + +#define SDM_ERFF4_2_00 0.2091176770593758483008706390019410965937912290f +#define SDM_ERFF4_2_01 1.0893988034775673230502318110338693557898033315f +#define SDM_ERFF4_2_02 -0.2042622756520438730719184645688505042105881396f +#define SDM_ERFF4_2_03 -0.3376001500360169568827541289401834722369442864f +#define SDM_ERFF4_2_04 0.0997374392832245473983976877777590352590762400f +#define SDM_ERFF4_2_05 0.0937997370645632460099464120987231140266525679f +#define SDM_ERFF4_2_06 -0.0324591340420617488485277008302392706957527828f +#define SDM_ERFF4_2_07 -0.0205943885488331791711970665266474471714543313f +#define SDM_ERFF4_2_08 0.0079208906865255014554772269570592999495375181f +#define SDM_ERFF4_2_09 0.0036744273281123333893101007014150883409965011f +#define SDM_ERFF4_2_10 -0.0015459493690754127608506357908913858038162608f +//#define SDM_ERFF4_2_11 -0.0005485671070180836650399266219057172124875094f + + + + /*************************************************************** + * REGION 3: + */ +#define SDM_ERFF4_3_START 0.25f +#define SDM_ERFF4_3_OFF 0.5f +#define SDM_ERFF4_3_TRUNC 2u + +#define SDM_ERFF4_3_00 0.5204998778130465376827466538919645287364515699f +#define SDM_ERFF4_3_01 0.8787825789354447940937239548244578983625218956f +#define SDM_ERFF4_3_02 -0.4393912894677223970468619774122289491812609947f +#define SDM_ERFF4_3_03 -0.1464637631559074656822873258040763163937536583f +#define SDM_ERFF4_3_04 0.1830797039448843321028591572550953954921920811f +#define SDM_ERFF4_3_05 0.0073231881577953732841143662902038158196876832f +#define SDM_ERFF4_3_06 -0.0500417857449350507747815029830594081011991688f +#define SDM_ERFF4_3_07 0.0054052103069442040906558417856266259621504328f +#define SDM_ERFF4_3_08 0.0100475885141180567975497704160236877764167320f +#define SDM_ERFF4_3_09 -0.0021674118390300459951330548378744759122422210f +#define SDM_ERFF4_3_10 -0.0015694967741624277200510981457278746801387524f +//#define SDM_ERFF4_3_11 0.0004973489167651373192082360776274483020158863f + + + + /*************************************************************** + * REGION 4: + */ +#define SDM_ERFF4_4_START 0.77f +#define SDM_ERFF4_4_OFF 1.0f +#define SDM_ERFF4_4_TRUNC 1u + +#define SDM_ERFF4_4_00 0.8427007929497148693412206350826092590442f +#define SDM_ERFF4_4_01 0.4151074974205947033402682494413373653605f +#define SDM_ERFF4_4_02 -0.4151074974205947033402682494413373653605f +#define SDM_ERFF4_4_03 0.1383691658068649011134227498137791217898f +#define SDM_ERFF4_4_04 0.0691845829034324505567113749068895608946f +#define SDM_ERFF4_4_05 -0.0691845829034324505567113749068895608946f +#define SDM_ERFF4_4_06 0.0046123055268954967037807583271259707263f +#define SDM_ERFF4_4_07 0.0151547181597994891695653487891281895293f +#define SDM_ERFF4_4_08 -0.0047770307242846215860586425530947553951f +#define SDM_ERFF4_4_09 -0.0018851883701199847638468972527538689873f +#define SDM_ERFF4_4_10 0.0012262875805634852347353603488787303121f +//#define SDM_ERFF4_4_11 0.0000855239913717274641321540324726821411f + + + + /*************************************************************** + * REGION 5: + */ +#define SDM_ERFF4_5_START 1.36f +#define SDM_ERFF4_5_OFF 1.875f +#define SDM_ERFF4_5_TRUNC 1u + +#define SDM_ERFF4_5_00 0.99199005767011997029646305969122440092668f +#define SDM_ERFF4_5_01 0.03354582842421607459425032786195496507386f +#define SDM_ERFF4_5_02 -0.06289842829540513986421936474116555951979f +#define SDM_ERFF4_5_03 0.06744109256118439996552409663913862770819f +#define SDM_ERFF4_5_04 -0.04225988151097532834627238568547061029869f +#define SDM_ERFF4_5_05 0.01146258336487617627004706027236136941544f +#define SDM_ERFF4_5_06 0.00410518713321247739022655684589964019683f +#define SDM_ERFF4_5_07 -0.00492839390823910723763257456562751425198f +#define SDM_ERFF4_5_08 0.00143050168737012207687743571780226012058f +#define SDM_ERFF4_5_09 0.00036225644575338665306295794978774160986f +#define SDM_ERFF4_5_10 -0.00039015757824554169745459780322413823624f +//#define SDM_ERFF4_5_11 0.00007372993782406230817649249567932577159f + + + + /*************************************************************** + * REGION 6: + */ +#define SDM_ERFF4_6_START 2.0f +#define SDM_ERFF4_6_OFF 2.5f +#define SDM_ERFF4_6_TRUNC 1u + +#define SDM_ERFF4_6_00 0.999593047982555041060435784260025087279f +#define SDM_ERFF4_6_01 0.002178284230352709720386678564097264007f +#define SDM_ERFF4_6_02 -0.005445710575881774300966696410243160031f +#define SDM_ERFF4_6_03 0.008350089549685387261482267829039512051f +#define SDM_ERFF4_6_04 -0.008622375078479475976530602649551670054f +#define SDM_ERFF4_6_05 0.006117348213573859798085922300839816434f +#define SDM_ERFF4_6_06 -0.002798490157050356237996774544152735014f +#define SDM_ERFF4_6_07 0.000542410061327906884739143174194854432f +#define SDM_ERFF4_6_08 0.000260670173895134533751630061303802055f +#define SDM_ERFF4_6_09 -0.000250285386311056635227961206817778392f +#define SDM_ERFF4_6_10 0.000078801328907504400502579703621546608f +//#define SDM_ERFF4_6_11 5.137004620216358263402877651297096663210e-6f + + + + /*************************************************************** + * REGION 7: */ - CONTFRAC_ERFCF4(xabs, xsqu, presult); - presult = spu_sub(onef, presult); +#define SDM_ERFF4_7_START 2.75f +#define SDM_ERFF4_7_OFF 3.5f +#define SDM_ERFF4_7_TRUNC 1u + +#define SDM_ERFF4_7_00 0.999999256901627658587254476316243904363263f +#define SDM_ERFF4_7_01 5.399426777384782511586818937495781413007869e-6f +#define SDM_ERFF4_7_02 -0.000018897993720846738790553866281235234945f +#define SDM_ERFF4_7_03 0.000042295509756180796340763415010383621069f +#define SDM_ERFF4_7_04 -0.000067717810833034147332818020841092925222f +#define SDM_ERFF4_7_05 0.000082116282239393567363716204674415008991f +#define SDM_ERFF4_7_06 -0.000077744246390483389302250766562526063763f +#define SDM_ERFF4_7_07 0.000058192750619199206596604051163855823527f +#define SDM_ERFF4_7_08 -0.000034259175422410008064403380504975403351f +#define SDM_ERFF4_7_09 0.000015330768263696827211862952666453348031f +#define SDM_ERFF4_7_10 -4.641017709492666503521243665632827470977627e-6f +//#define SDM_ERFF4_7_11 4.447037356176705948450355327103423490366212e-7f + + + + + + /*************************************************************** + * Now we load the description of each partition. + */ + + /* Start point for each partition */ + vec_float4 r1start = spu_splats(SDM_ERFF4_1_START); + vec_float4 r2start = spu_splats(SDM_ERFF4_2_START); + vec_float4 r3start = spu_splats(SDM_ERFF4_3_START); + vec_float4 r4start = spu_splats(SDM_ERFF4_4_START); + vec_float4 r5start = spu_splats(SDM_ERFF4_5_START); + vec_float4 r6start = spu_splats(SDM_ERFF4_6_START); + vec_float4 r7start = spu_splats(SDM_ERFF4_7_START); + + /* X Offset for each partition */ + vec_float4 xoffseta = (vec_float4) {SDM_ERFF4_0_OFF, SDM_ERFF4_1_OFF, SDM_ERFF4_2_OFF, SDM_ERFF4_3_OFF}; + vec_float4 xoffsetb = (vec_float4) {SDM_ERFF4_4_OFF, SDM_ERFF4_5_OFF, SDM_ERFF4_6_OFF, SDM_ERFF4_7_OFF}; + + /* Truncation Correction for each partition */ + vec_uint4 tcorra = (vec_uint4) {SDM_ERFF4_0_TRUNC, SDM_ERFF4_1_TRUNC, SDM_ERFF4_2_TRUNC, SDM_ERFF4_3_TRUNC}; + vec_uint4 tcorrb = (vec_uint4) {SDM_ERFF4_4_TRUNC, SDM_ERFF4_5_TRUNC, SDM_ERFF4_6_TRUNC, SDM_ERFF4_7_TRUNC}; + + /* The coefficients for each partition */ + vec_float4 c00a = (vec_float4) {SDM_ERFF4_0_00, SDM_ERFF4_1_00, SDM_ERFF4_2_00, SDM_ERFF4_3_00}; + vec_float4 c01a = (vec_float4) {SDM_ERFF4_0_01, SDM_ERFF4_1_01, SDM_ERFF4_2_01, SDM_ERFF4_3_01}; + vec_float4 c02a = (vec_float4) {SDM_ERFF4_0_02, SDM_ERFF4_1_02, SDM_ERFF4_2_02, SDM_ERFF4_3_02}; + vec_float4 c03a = (vec_float4) {SDM_ERFF4_0_03, SDM_ERFF4_1_03, SDM_ERFF4_2_03, SDM_ERFF4_3_03}; + vec_float4 c04a = (vec_float4) {SDM_ERFF4_0_04, SDM_ERFF4_1_04, SDM_ERFF4_2_04, SDM_ERFF4_3_04}; + vec_float4 c05a = (vec_float4) {SDM_ERFF4_0_05, SDM_ERFF4_1_05, SDM_ERFF4_2_05, SDM_ERFF4_3_05}; + vec_float4 c06a = (vec_float4) {SDM_ERFF4_0_06, SDM_ERFF4_1_06, SDM_ERFF4_2_06, SDM_ERFF4_3_06}; + vec_float4 c07a = (vec_float4) {SDM_ERFF4_0_07, SDM_ERFF4_1_07, SDM_ERFF4_2_07, SDM_ERFF4_3_07}; + vec_float4 c08a = (vec_float4) {SDM_ERFF4_0_08, SDM_ERFF4_1_08, SDM_ERFF4_2_08, SDM_ERFF4_3_08}; + vec_float4 c09a = (vec_float4) {SDM_ERFF4_0_09, SDM_ERFF4_1_09, SDM_ERFF4_2_09, SDM_ERFF4_3_09}; + vec_float4 c10a = (vec_float4) {SDM_ERFF4_0_10, SDM_ERFF4_1_10, SDM_ERFF4_2_10, SDM_ERFF4_3_10}; + + vec_float4 c00b = (vec_float4) {SDM_ERFF4_4_00, SDM_ERFF4_5_00, SDM_ERFF4_6_00, SDM_ERFF4_7_00}; + vec_float4 c01b = (vec_float4) {SDM_ERFF4_4_01, SDM_ERFF4_5_01, SDM_ERFF4_6_01, SDM_ERFF4_7_01}; + vec_float4 c02b = (vec_float4) {SDM_ERFF4_4_02, SDM_ERFF4_5_02, SDM_ERFF4_6_02, SDM_ERFF4_7_02}; + vec_float4 c03b = (vec_float4) {SDM_ERFF4_4_03, SDM_ERFF4_5_03, SDM_ERFF4_6_03, SDM_ERFF4_7_03}; + vec_float4 c04b = (vec_float4) {SDM_ERFF4_4_04, SDM_ERFF4_5_04, SDM_ERFF4_6_04, SDM_ERFF4_7_04}; + vec_float4 c05b = (vec_float4) {SDM_ERFF4_4_05, SDM_ERFF4_5_05, SDM_ERFF4_6_05, SDM_ERFF4_7_05}; + vec_float4 c06b = (vec_float4) {SDM_ERFF4_4_06, SDM_ERFF4_5_06, SDM_ERFF4_6_06, SDM_ERFF4_7_06}; + vec_float4 c07b = (vec_float4) {SDM_ERFF4_4_07, SDM_ERFF4_5_07, SDM_ERFF4_6_07, SDM_ERFF4_7_07}; + vec_float4 c08b = (vec_float4) {SDM_ERFF4_4_08, SDM_ERFF4_5_08, SDM_ERFF4_6_08, SDM_ERFF4_7_08}; + vec_float4 c09b = (vec_float4) {SDM_ERFF4_4_09, SDM_ERFF4_5_09, SDM_ERFF4_6_09, SDM_ERFF4_7_09}; + vec_float4 c10b = (vec_float4) {SDM_ERFF4_4_10, SDM_ERFF4_5_10, SDM_ERFF4_6_10, SDM_ERFF4_7_10}; + + + vec_uchar16 shuffle0 = (vec_uchar16) spu_splats(0x00010203); + vec_uchar16 shuffle1 = (vec_uchar16) spu_splats(0x04050607); + vec_uchar16 shuffle2 = (vec_uchar16) spu_splats(0x08090A0B); + vec_uchar16 shuffle3 = (vec_uchar16) spu_splats(0x0C0D0E0F); + vec_uchar16 shuffle4 = (vec_uchar16) spu_splats(0x10111213); + vec_uchar16 shuffle5 = (vec_uchar16) spu_splats(0x14151617); + vec_uchar16 shuffle6 = (vec_uchar16) spu_splats(0x18191A1B); + vec_uchar16 shuffle7 = (vec_uchar16) spu_splats(0x1C1D1E1F); + /* - * Select the appropriate approximation. + * Determine the shuffle pattern based on which partition + * each element of x is in. */ - result = spu_sel(tresult, presult, spu_cmpgt(xabs, approx_point)); + + vec_uchar16 gt_r1start = (vec_uchar16)spu_cmpabsgt(x, r1start); + vec_uchar16 gt_r2start = (vec_uchar16)spu_cmpabsgt(x, r2start); + vec_uchar16 gt_r3start = (vec_uchar16)spu_cmpabsgt(x, r3start); + vec_uchar16 gt_r4start = (vec_uchar16)spu_cmpabsgt(x, r4start); + vec_uchar16 gt_r5start = (vec_uchar16)spu_cmpabsgt(x, r5start); + vec_uchar16 gt_r6start = (vec_uchar16)spu_cmpabsgt(x, r6start); + vec_uchar16 gt_r7start = (vec_uchar16)spu_cmpabsgt(x, r7start); + + vec_uchar16 shufflepattern; + shufflepattern = spu_sel(shuffle0, shuffle1, gt_r1start); + shufflepattern = spu_sel(shufflepattern, shuffle2, gt_r2start); + shufflepattern = spu_sel(shufflepattern, shuffle3, gt_r3start); + shufflepattern = spu_sel(shufflepattern, shuffle4, gt_r4start); + shufflepattern = spu_sel(shufflepattern, shuffle5, gt_r5start); + shufflepattern = spu_sel(shufflepattern, shuffle6, gt_r6start); + shufflepattern = spu_sel(shufflepattern, shuffle7, gt_r7start); + + + + /* Use the shuffle pattern to select the coefficients */ + + vec_float4 coeff_10 = spu_shuffle(c10a, c10b, shufflepattern); + vec_float4 coeff_09 = spu_shuffle(c09a, c09b, shufflepattern); + vec_float4 coeff_08 = spu_shuffle(c08a, c08b, shufflepattern); + vec_float4 coeff_07 = spu_shuffle(c07a, c07b, shufflepattern); + vec_float4 coeff_06 = spu_shuffle(c06a, c06b, shufflepattern); + vec_float4 coeff_05 = spu_shuffle(c05a, c05b, shufflepattern); + vec_float4 coeff_04 = spu_shuffle(c04a, c04b, shufflepattern); + vec_float4 coeff_03 = spu_shuffle(c03a, c03b, shufflepattern); + vec_float4 coeff_02 = spu_shuffle(c02a, c02b, shufflepattern); + vec_float4 coeff_01 = spu_shuffle(c01a, c01b, shufflepattern); + vec_float4 coeff_00 = spu_shuffle(c00a, c00b, shufflepattern); + + vec_float4 xoffset = spu_shuffle(xoffseta, xoffsetb, shufflepattern); + vec_uint4 tcorrection = spu_shuffle(tcorra, tcorrb, shufflepattern); + /* - * Special cases/errors. + * We've completed the coeff. setup. Now we actually do the + * approximation below. */ - /* x = +/- infinite, return +/-1 */ - isinf = spu_cmpeq((vec_uint4)xabs, 0x7F800000); - result = spu_sel(result, onef, isinf); + /* Adjust x value here (for approximations about a point) */ + vec_float4 xappr = spu_sub(xabs, xoffset); + + + /* Now we do the multiplies. + * Use Horner's method. + */ + result = coeff_10; + result = spu_madd(xappr, result, coeff_09); + result = spu_madd(xappr, result, coeff_08); + result = spu_madd(xappr, result, coeff_07); + result = spu_madd(xappr, result, coeff_06); + result = spu_madd(xappr, result, coeff_05); + result = spu_madd(xappr, result, coeff_04); + result = spu_madd(xappr, result, coeff_03); + result = spu_madd(xappr, result, coeff_02); + result = spu_madd(xappr, result, coeff_01); + result = spu_madd(xappr, result, coeff_00); + + + /* Adjust due to systematic truncation. Note that the correction + * value is always non-negative, so the result is cast as uint + * to do the adjustment. + */ + result = (vec_float4)spu_add((vec_uint4)result, tcorrection); + /* - * Preserve sign in result, since erf(-x) = -erf(x) + * Special Cases */ + + /* Erf(0) = 0 */ + result = spu_sel(result, zerof, spu_cmpeq(xabs, zerof)); + + /* Erf(infinity) = 1 */ + result = spu_sel(result, onef, spu_cmpgt(xabs, clamp)); + + + /* Preserve sign in result, since erf(-x) = -erf(x) */ result = spu_or(result, xsign); return result; diff --git a/newlib/libm/machine/spu/headers/expd2.h b/newlib/libm/machine/spu/headers/expd2.h index 19d77dd51..1a8581292 100644 --- a/newlib/libm/machine/spu/headers/expd2.h +++ b/newlib/libm/machine/spu/headers/expd2.h @@ -56,7 +56,6 @@ #include <spu_intrinsics.h> #include "floord2.h" -#include "ldexpd2.h" #define LOG2E 1.4426950408889634073599 // 1/log(2) @@ -101,15 +100,12 @@ static __inline vector double _expd2(vector double x) { - vec_uchar16 even2odd = ((vec_uchar16){0x80, 0x80, 0x80, 0x80, 0, 1, 2, 3, - 0x80, 0x80, 0x80, 0x80, 8, 9, 10, 11}); - // log(2) in extended machine representable precision vec_double2 ln2_hi = spu_splats(6.9314575195312500E-1); // 3FE62E4000000000 vec_double2 ln2_lo = spu_splats(1.4286068203094172E-6); // 3EB7F7D1CF79ABCA - // coefficients for the power series + // vec_double2 f01 = spu_splats(1.00000000000000000000E0); // 1/(1!) vec_double2 f02 = spu_splats(5.00000000000000000000E-1); // 1/(2!) vec_double2 f03 = spu_splats(1.66666666666666666667E-1); // 1/(3!) vec_double2 f04 = spu_splats(4.16666666666666666667E-2); // 1/(4!) @@ -126,19 +122,42 @@ static __inline vector double _expd2(vector double x) vec_double2 rx = _floord2(spu_madd(x,spu_splats(LOG2E),spu_splats(0.5))); // extract the exponent of reduction - vec_int4 nint = spu_convts(spu_roundtf(rx),0); - vec_llong2 n = spu_extend(spu_shuffle(nint, nint, even2odd)); + vec_int4 exp = spu_convts(spu_roundtf(rx),0); // reduce the input to within [ -ln(2)/2 ... ln(2)/2 ] vec_double2 r; r = spu_nmsub(rx,ln2_hi,x); r = spu_nmsub(rx,ln2_lo,r); - vec_double2 result; vec_double2 r2 = spu_mul(r,r); // Use Horner's method on the power series + /* result = ((((c12*x + c11)*x + c10)*x + c9)*x + c8)*x + c7)*x + c6)*x^6 + + ((((((c5*x + c4)*x + c3)*x + c2)*x + c1)*x + c0 + */ + +#ifdef __SPU_EDP__ + vec_double2 p1, p2, r4, r6; + + p1 = spu_madd(f12, r, f11); + p2 = spu_madd(f05, r, f04); + r4 = spu_mul(r2, r2); + p1 = spu_madd(p1, r, f10); + p2 = spu_madd(p2, r, f03); + p1 = spu_madd(p1, r, f09); + p2 = spu_madd(p2, r, f02); + p1 = spu_madd(p1, r, f08); + r6 = spu_mul(r2, r4); + p1 = spu_madd(p1, r, f07); + p2 = spu_madd(p2, r2, r); + p1 = spu_madd(p1, r, f06); + + result = spu_madd(r6, p1, p2); + result = spu_add(result, spu_splats(1.0)); + +#else + result = spu_madd(r,f12,f11); result = spu_madd(result,r,f10); result = spu_madd(result,r,f09); @@ -152,14 +171,38 @@ static __inline vector double _expd2(vector double x) result = spu_madd(result,r2,r); result = spu_add(result,spu_splats(1.0)); - // Scale the result - result = _ldexpd2(result, n); +#endif /* __SPU_EDP__ */ - return result; + // Scale the result - basically a call to ldexpd2() + vec_int4 e1, e2; + vec_int4 min = spu_splats(-2044); + vec_int4 max = spu_splats(2046); + vec_uint4 cmp_min, cmp_max; + vec_uint4 shift = (vec_uint4) { 20, 32, 20, 32 }; + vec_double2 f1, f2; -} + /* Clamp the specified exponent to the range -2044 to 2046. + */ + cmp_min = spu_cmpgt(exp, min); + cmp_max = spu_cmpgt(exp, max); + exp = spu_sel(min, exp, cmp_min); + exp = spu_sel(exp, max, cmp_max); + /* Generate the factors f1 = 2^e1 and f2 = 2^e2 + */ + e1 = spu_rlmaska(exp, -1); + e2 = spu_sub(exp, e1); + + f1 = (vec_double2)spu_sl(spu_add(e1, 1023), shift); + f2 = (vec_double2)spu_sl(spu_add(e2, 1023), shift); + + /* Compute the product x * 2^e1 * 2^e2 + */ + result = spu_mul(spu_mul(result, f1), f2); + + return result; +} #endif /* _EXPD2_H_ */ #endif /* __SPU__ */ diff --git a/newlib/libm/machine/spu/headers/ldexpd2.h b/newlib/libm/machine/spu/headers/ldexpd2.h index f00afceb2..758852b5d 100644 --- a/newlib/libm/machine/spu/headers/ldexpd2.h +++ b/newlib/libm/machine/spu/headers/ldexpd2.h @@ -70,10 +70,7 @@ static __inline vector double _ldexpd2(vector double x, vector signed long long { vec_uchar16 odd_to_even = ((vec_uchar16) { 4,5,6,7, 0x80,0x80,0x80,0x80, 12,13,14,15, 0x80,0x80,0x80,0x80 }); - vec_uchar16 dup_even = ((vec_uchar16) { 0,1,2,3, 0,1,2,3, - 8,9,10,11, 8,9,10,11}); vec_int4 exp; - vec_uint4 exphi; vec_int4 e1, e2; vec_int4 min = spu_splats(-2044); vec_int4 max = spu_splats(2046); @@ -84,8 +81,6 @@ static __inline vector double _ldexpd2(vector double x, vector signed long long exp = (vec_int4)spu_shuffle(llexp, llexp, odd_to_even); - exphi = (vec_uint4)spu_shuffle(llexp, llexp, dup_even); - /* Clamp the specified exponent to the range -2044 to 2046. */ diff --git a/newlib/libm/machine/spu/headers/lgammaf4.h b/newlib/libm/machine/spu/headers/lgammaf4.h index 956a0e1c4..555d7acba 100644 --- a/newlib/libm/machine/spu/headers/lgammaf4.h +++ b/newlib/libm/machine/spu/headers/lgammaf4.h @@ -53,11 +53,13 @@ #define _LGAMMAF4_H_ 1 #include <spu_intrinsics.h> -#include "lgammad2.h" -#include "recipf4.h" + #include "logf4.h" -#include "sinf4.h" +#include "divf4.h" +#include "recipf4.h" #include "truncf4.h" +#include "sinf4.h" + /* * FUNCTION @@ -68,129 +70,423 @@ * function for the corresponding elements of the input vector. * * C99 Special Cases: - * lgamma(0) returns +infinite + * lgamma(0) returns +infinity * lgamma(1) returns +0 * lgamma(2) returns +0 - * lgamma(negative integer) returns +infinite - * lgamma(+infinite) returns +infinite - * lgamma(-infinite) returns +infinite + * lgamma(negative integer) returns +infinity + * lgamma(+infinity) returns +infinity + * lgamma(-infinity) returns +infinity * * Other Cases: * lgamma(Nan) returns Nan - * lgamma(Denorm) treated as lgamma(0) and returns +infinite + * lgamma(Denorm) treated as lgamma(0) and returns +infinity * */ - static __inline vector float _lgammaf4(vector float x) { + vec_float4 result; + vec_float4 halflog2pi = spu_splats(9.189385332046727417803297364056E-1f); + vec_float4 logpi = spu_splats(1.1447298858494001741434273513530587116472948129153f); vec_float4 inff = (vec_float4)spu_splats(0x7F800000); vec_float4 zerof = spu_splats(0.0f); - vec_float4 pi = spu_splats((float)PI); + vec_float4 onef = spu_splats(1.0f); + vec_float4 twof = spu_splats(2.0f); vec_float4 sign_maskf = spu_splats(-0.0f); + vec_float4 pi = spu_splats(3.14159265358979323846264338328f); - vector unsigned int gt0; - - /* This is where we switch from near zero approx. */ - vec_float4 mac_switch = spu_splats(0.16f); - vec_float4 shift_switch = spu_splats(6.0f); - - vec_float4 inv_x, inv_xsqu; - vec_float4 xtrunc, xstirling; - vec_float4 sum, xabs; - vec_uint4 isnaninf, isshifted; - vec_float4 result, stresult, shresult, mresult, nresult; + /* + * Unfortunately, some of the approximation methods for lgamma require + * other basic math computations. Get those out of the way now. The + * compiler seems to good a good job of scheduling this code with + * the code that follows. + */ + vec_uint4 gt0 = spu_cmpgt(x, zerof); + vec_float4 xabs = spu_andc(x, sign_maskf); + vec_float4 ln_x = _logf4(xabs); + vec_float4 inv_x = _recipf4(xabs); + vec_float4 xtrunc = _truncf4(x); + vec_float4 inv_xsqu = spu_mul(inv_x, inv_x); + vec_uint4 isnaninf = spu_cmpgt((vec_uint4)xabs, 0x7F7FFFFF); + vec_uint4 ret_zero = spu_or(spu_cmpeq(x, onef), spu_cmpeq(x, twof)); - /* Force Denorms to 0 */ - x = spu_add(x, zerof); - xabs = spu_andc(x, sign_maskf); + /* + * First thing we do is setup the description of each partition. + * This consists of: + * - Start x of partition + * - Offset (used for evaluating power series expanded around a point) + * - Truncation adjustment. + * - Is approx method in region a rational approximation or just a polynomial + * - The coefficients used in the poly or rational approximation + */ - gt0 = spu_cmpgt(x, zerof); - xtrunc = _truncf4(x); - /* - * For 0 < x <= 0.16. - * Approximation Near Zero + /*************************************************************** + * REGION 0: Approximation Near 0 from Above * * Use Maclaurin Expansion of lgamma() * * lgamma(z) = -ln(z) - z * EulerMascheroni + Sum[(-1)^n * z^n * Zeta(n)/n] */ - mresult = spu_madd(xabs, spu_splats((float)ZETA_06_DIV_06), spu_splats((float)ZETA_05_DIV_05)); - mresult = spu_madd(xabs, mresult, spu_splats((float)ZETA_04_DIV_04)); - mresult = spu_madd(xabs, mresult, spu_splats((float)ZETA_03_DIV_03)); - mresult = spu_madd(xabs, mresult, spu_splats((float)ZETA_02_DIV_02)); - mresult = spu_mul(xabs, spu_mul(xabs, mresult)); - mresult = spu_sub(mresult, spu_add(_logf4(xabs), spu_mul(xabs, spu_splats((float)EULER_MASCHERONI)))); +#define SDM_LGF4_0_START 0.0f +#define SDM_LGF4_0_OFF 0.0f +#define SDM_LGF4_0_TRUNC 2u +#define SDM_LGF4_0_RATIONAL 0x0u - /* - * For 0.16 < x <= 6.0, we are going to push value - * out to an area where Stirling's approximation is - * accurate. Let's use a constant of 6. - * - * Use the recurrence relation: - * lgamma(x + 1) = ln(x) + lgamma(x) - * - * Note that we shift x here, before Stirling's calculation, - * then after Stirling's, we adjust the result. +#define SDM_LGF4_0_00 0.0f +#define SDM_LGF4_0_01 -0.5772156649015328606065121f +#define SDM_LGF4_0_02 0.8224670334241132182362076f +#define SDM_LGF4_0_03 -0.4006856343865314284665794f +#define SDM_LGF4_0_04 0.2705808084277845478790009f +#define SDM_LGF4_0_05 -0.2073855510286739852662731f +#define SDM_LGF4_0_06 1.6955717699740818995241965496515E-1f +#define SDM_LGF4_0_07 -1.4404989676884611811997107854997E-1f +#define SDM_LGF4_0_08 1.2550966952474304242233565481358E-1f +#define SDM_LGF4_0_09 -1.1133426586956469049087252991471E-1f +#define SDM_LGF4_0_10 1.0009945751278180853371459589003E-1f +#define SDM_LGF4_0_11 -9.0954017145829042232609298411497E-2f + + + + /*************************************************************** + * REGION 1: Above 0 and Below 1 + */ +#define SDM_LGF4_1_START 0.20f +#define SDM_LGF4_1_OFF 0.0f +#define SDM_LGF4_1_TRUNC 0u +#define SDM_LGF4_1_RATIONAL 0xFFFFFFFFu + +/* Numerator */ +#define SDM_LGF4_1_06 5.5247592697706124892083167601451981186889952720891079f +#define SDM_LGF4_1_07 188.42248906442882644741346270888237140890625699348872f +#define SDM_LGF4_1_08 730.89115027907050579364152184942040244662318995470771f +#define SDM_LGF4_1_09 -517.93391251349155395618464682404141737699116911423096f +#define SDM_LGF4_1_10 -866.81293419754982917624255525168901081630973644141406f +#define SDM_LGF4_1_11 459.90872804523394478152324135956113729930154636775805f + +/* Denominator */ +#define SDM_LGF4_1_00 1.0f +#define SDM_LGF4_1_01 62.356015559548850893358835861387218304619374633480009f +#define SDM_LGF4_1_02 553.64875642095755724931612658933597252336243693499682f +#define SDM_LGF4_1_03 997.28805670393557265195865662557219661414263910835386f +#define SDM_LGF4_1_04 257.10520661440946455560646958565998121417179154677712f +#define SDM_LGF4_1_05 -15.398409585547124178878369413880017200739911288666830f + + + + /*************************************************************** + * REGION 2: Above 0 and Below 1 + */ +#define SDM_LGF4_2_START 0.60f +#define SDM_LGF4_2_OFF 0.69f +#define SDM_LGF4_2_TRUNC 1u +#define SDM_LGF4_2_RATIONAL 0x0u + +/* This is a power series expanson of LogGamma around 0.69 */ +#define SDM_LGF4_2_00 0.27321026793030387025442491383648273204234f +#define SDM_LGF4_2_01 -1.24869016926209356266849815723905575347988f +#define SDM_LGF4_2_02 1.44985879780363867173410158693003578927407f +#define SDM_LGF4_2_03 -1.11686573274718166516744313082147691068190f +#define SDM_LGF4_2_04 1.14079150485439143731395820215710950729505f +#define SDM_LGF4_2_05 -1.29512166953091144888197173527810141620764f +#define SDM_LGF4_2_06 1.55206382120790061136858894716459302629069f +#define SDM_LGF4_2_07 -1.92227237154565289482911310272968704445560f +#define SDM_LGF4_2_08 2.43478939488445894670349784581009987461638f +#define SDM_LGF4_2_09 -3.13512449573283650741385084753752461908870f +#define SDM_LGF4_2_10 4.08851456399492725127969680590409811177590f +#define SDM_LGF4_2_11 5.38629680478093362448042704719642976375265f + + + + /*************************************************************** + * REGION 3: Around 1 + */ +#define SDM_LGF4_3_START 0.74f +#define SDM_LGF4_3_OFF 1.0f +#define SDM_LGF4_3_TRUNC 2u +#define SDM_LGF4_3_RATIONAL 0x0u + +#define SDM_LGF4_3_11 -0.90954017145829042232609298411497266951691494159836e-1f +#define SDM_LGF4_3_10 0.10009945751278180853371459589003190170060195315645f +#define SDM_LGF4_3_09 -0.11133426586956469049087252991471245116506731682165f +#define SDM_LGF4_3_08 0.12550966952474304242233565481358155815737009883123f +#define SDM_LGF4_3_07 -0.14404989676884611811997107854997096565712336579503f +#define SDM_LGF4_3_06 0.16955717699740818995241965496515342131696958167214f +#define SDM_LGF4_3_05 -0.20738555102867398526627309729140683361141618390038f +#define SDM_LGF4_3_04 0.27058080842778454787900092413529197569368773797968f +#define SDM_LGF4_3_03 -0.40068563438653142846657938717048333025499543078016f +#define SDM_LGF4_3_02 0.82246703342411321823620758332301259460947495060340f +#define SDM_LGF4_3_01 -0.57721566490153286060651209008240243104215933593992f +#define SDM_LGF4_3_00 0.0f + + + + /*************************************************************** + * REGION 4: Above 1 to Below 2 + */ + +#define SDM_LGF4_4_START 1.25f +#define SDM_LGF4_4_OFF 1.4616321449683623412626595423257213284681962040064f +#define SDM_LGF4_4_TRUNC 1u +#define SDM_LGF4_4_RATIONAL 0x0u + +#define SDM_LGF4_4_00 -0.12148629053584960809551455717769158215135617313000f +#define SDM_LGF4_4_01 0.0f +#define SDM_LGF4_4_02 0.48383612272381058521372238085482537020562860838860f +#define SDM_LGF4_4_03 -0.14758772299453070203095509395083641661852764909458f +#define SDM_LGF4_4_04 0.064624940238912752656100346425238557063086033931734f +#define SDM_LGF4_4_05 -0.032788541088481305500850258549331278505894787737970f +#define SDM_LGF4_4_06 0.017970675115210394292863824811126161810628596070981f +#define SDM_LGF4_4_07 -0.010314223036636387275160254800730296612070784399082f +#define SDM_LGF4_4_08 0.0061005360205178884031365656884883648099463048507839f +#define SDM_LGF4_4_09 -0.0036845696083163732546953776004972425913603137160767f +#define SDM_LGF4_4_10 0.00225976482322181046596248251178293952686321035f +#define SDM_LGF4_4_11 -0.00140225144590445083080002880374741201782467331f + + + + /*************************************************************** + * REGION 5: Around 2 + */ + +#define SDM_LGF4_5_START 1.50f +#define SDM_LGF4_5_OFF 2.0f +#define SDM_LGF4_5_TRUNC 1u +#define SDM_LGF4_5_RATIONAL 0x0u + +#define SDM_LGF4_5_00 0.0f +#define SDM_LGF4_5_01 0.42278433509846713939348790991759756895784066406008f +#define SDM_LGF4_5_02 0.32246703342411321823620758332301259460947495060340f +#define SDM_LGF4_5_03 -0.6735230105319809513324605383714999692166209744683e-1f +#define SDM_LGF4_5_04 0.2058080842778454787900092413529197569368773797968e-1f +#define SDM_LGF4_5_05 -0.738555102867398526627309729140683361141618390038e-2f +#define SDM_LGF4_5_06 0.289051033074152328575298829848675465030291500547e-2f +#define SDM_LGF4_5_07 -0.119275391170326097711393569282810851426622293789e-2f +#define SDM_LGF4_5_08 0.50966952474304242233565481358155815737009883123e-3f +#define SDM_LGF4_5_09 -0.22315475845357937976141880360134005395620571054e-3f +#define SDM_LGF4_5_10 0.9945751278180853371459589003190170060195315645e-4f +#define SDM_LGF4_5_11 -0.44926236738133141700207502406357860782403250745e-4f + + + + /*************************************************************** + * REGION 6: Above 2 to Below Stirlings + */ + +#define SDM_LGF4_6_START 2.48f +#define SDM_LGF4_6_OFF 0.0f +#define SDM_LGF4_6_TRUNC 2u +#define SDM_LGF4_6_RATIONAL 0xFFFFFFFFu + +/* Numerator */ +#define SDM_LGF4_6_06 2.8952045264375719070927153893062450394256201846894266f +#define SDM_LGF4_6_07 0.9017557380149600532583460408941390566399250566546766f +#define SDM_LGF4_6_08 -5.0120743649109868270726470406381462995568837028633266f +#define SDM_LGF4_6_09 0.5723176665030477945174549923532715487712277062412760f +#define SDM_LGF4_6_10 0.6107282478237180956153912232438073421489100296366786f +#define SDM_LGF4_6_11 0.0312308625200519550078820867041868696010490562277303f + +/* Denominator */ +#define SDM_LGF4_6_00 1.0f +#define SDM_LGF4_6_01 4.3592151369378598515798083402849838078885877442021500f +#define SDM_LGF4_6_02 2.6245676641191702420707093818412405820501009602499853f +#define SDM_LGF4_6_03 0.3438846837443412565179153619145215759074092780311669f +#define SDM_LGF4_6_04 0.0078092905528158343621764949220712317164193605131159f +#define SDM_LGF4_6_05 -0.000015217018272713076443927141674684568030697337620f + + + + /*************************************************************** + * REGION 7: Stirlings - Above 6.0 * */ - isshifted = spu_cmpgt(shift_switch, x); - xstirling = spu_sel(xabs, spu_add(xabs, spu_splats(6.0f)), isshifted); - inv_x = _recipf4(xstirling); - inv_xsqu = spu_mul(inv_x, inv_x); +#define SDM_LGF4_7_START 7.80f +#define SDM_LGF4_7_OFF 0.0f +#define SDM_LGF4_7_TRUNC 5u +#define SDM_LGF4_7_RATIONAL 0x0u + +#define SDM_LGF4_7_00 8.3333333333333333333333333333333333333333333333333333333333333333333333E-2f +#define SDM_LGF4_7_01 -2.7777777777777777777777777777777777777777777777777777777777777777777778E-3f +#define SDM_LGF4_7_02 7.9365079365079365079365079365079365079365079365079365079365079365079365E-4f +#define SDM_LGF4_7_03 -5.9523809523809523809523809523809523809523809523809523809523809523809524E-4f +#define SDM_LGF4_7_04 8.4175084175084175084175084175084175084175084175084175084175084175084175E-4f +#define SDM_LGF4_7_05 -1.9175269175269175269175269175269175269175269175269175269175269175269175E-3f +#define SDM_LGF4_7_06 6.4102564102564102564102564102564102564102564102564102564102564102564103E-3f +#define SDM_LGF4_7_07 0.0f +#define SDM_LGF4_7_08 0.0f +#define SDM_LGF4_7_09 0.0f +#define SDM_LGF4_7_10 0.0f +#define SDM_LGF4_7_11 0.0f + /* - * For 6.0 < x < infinite - * - * Use Stirling's Series. - * - * 1 1 1 1 1 - * lgamma(x) = --- ln (2*pi) + (z - ---) ln(x) - x + --- - ----- + ------ ... - * 2 2 12x 360x^3 1260x^5 - * - * + * Now we load the description of each partition. */ - sum = spu_madd(inv_xsqu, spu_splats((float)STIRLING_10), spu_splats((float)STIRLING_09)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_08)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_07)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_06)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_05)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_04)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_03)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_02)); - sum = spu_madd(sum, inv_xsqu, spu_splats((float)STIRLING_01)); - sum = spu_mul(sum, inv_x); - - stresult = spu_madd(spu_sub(xstirling, spu_splats(0.5f)), _logf4(xstirling), spu_splats((float)HALFLOG2PI)); - stresult = spu_sub(stresult, xstirling); - stresult = spu_add(stresult, sum); + + /* Start point for each partition */ + vec_float4 r1start = spu_splats(SDM_LGF4_1_START); + vec_float4 r2start = spu_splats(SDM_LGF4_2_START); + vec_float4 r3start = spu_splats(SDM_LGF4_3_START); + vec_float4 r4start = spu_splats(SDM_LGF4_4_START); + vec_float4 r5start = spu_splats(SDM_LGF4_5_START); + vec_float4 r6start = spu_splats(SDM_LGF4_6_START); + vec_float4 r7start = spu_splats(SDM_LGF4_7_START); + + /* X Offset for each partition */ + vec_float4 xoffseta = (vec_float4) {SDM_LGF4_0_OFF, SDM_LGF4_1_OFF, SDM_LGF4_2_OFF, SDM_LGF4_3_OFF}; + vec_float4 xoffsetb = (vec_float4) {SDM_LGF4_4_OFF, SDM_LGF4_5_OFF, SDM_LGF4_6_OFF, SDM_LGF4_7_OFF}; + + /* Truncation Correction for each partition */ + vec_uint4 tcorra = (vec_uint4) {SDM_LGF4_0_TRUNC, SDM_LGF4_1_TRUNC, SDM_LGF4_2_TRUNC, SDM_LGF4_3_TRUNC}; + vec_uint4 tcorrb = (vec_uint4) {SDM_LGF4_4_TRUNC, SDM_LGF4_5_TRUNC, SDM_LGF4_6_TRUNC, SDM_LGF4_7_TRUNC}; + + /* Is partition a Rational Approximation */ + vec_uint4 israta = (vec_uint4) {SDM_LGF4_0_RATIONAL, SDM_LGF4_1_RATIONAL, SDM_LGF4_2_RATIONAL, SDM_LGF4_3_RATIONAL}; + vec_uint4 isratb = (vec_uint4) {SDM_LGF4_4_RATIONAL, SDM_LGF4_5_RATIONAL, SDM_LGF4_6_RATIONAL, SDM_LGF4_7_RATIONAL}; + + /* The polynomial coefficients for all partitions */ + vec_float4 c00a = (vec_float4) {SDM_LGF4_0_00, SDM_LGF4_1_00, SDM_LGF4_2_00, SDM_LGF4_3_00}; + vec_float4 c01a = (vec_float4) {SDM_LGF4_0_01, SDM_LGF4_1_01, SDM_LGF4_2_01, SDM_LGF4_3_01}; + vec_float4 c02a = (vec_float4) {SDM_LGF4_0_02, SDM_LGF4_1_02, SDM_LGF4_2_02, SDM_LGF4_3_02}; + vec_float4 c03a = (vec_float4) {SDM_LGF4_0_03, SDM_LGF4_1_03, SDM_LGF4_2_03, SDM_LGF4_3_03}; + vec_float4 c04a = (vec_float4) {SDM_LGF4_0_04, SDM_LGF4_1_04, SDM_LGF4_2_04, SDM_LGF4_3_04}; + vec_float4 c05a = (vec_float4) {SDM_LGF4_0_05, SDM_LGF4_1_05, SDM_LGF4_2_05, SDM_LGF4_3_05}; + vec_float4 c06a = (vec_float4) {SDM_LGF4_0_06, SDM_LGF4_1_06, SDM_LGF4_2_06, SDM_LGF4_3_06}; + vec_float4 c07a = (vec_float4) {SDM_LGF4_0_07, SDM_LGF4_1_07, SDM_LGF4_2_07, SDM_LGF4_3_07}; + vec_float4 c08a = (vec_float4) {SDM_LGF4_0_08, SDM_LGF4_1_08, SDM_LGF4_2_08, SDM_LGF4_3_08}; + vec_float4 c09a = (vec_float4) {SDM_LGF4_0_09, SDM_LGF4_1_09, SDM_LGF4_2_09, SDM_LGF4_3_09}; + vec_float4 c10a = (vec_float4) {SDM_LGF4_0_10, SDM_LGF4_1_10, SDM_LGF4_2_10, SDM_LGF4_3_10}; + vec_float4 c11a = (vec_float4) {SDM_LGF4_0_11, SDM_LGF4_1_11, SDM_LGF4_2_11, SDM_LGF4_3_11}; + + vec_float4 c00b = (vec_float4) {SDM_LGF4_4_00, SDM_LGF4_5_00, SDM_LGF4_6_00, SDM_LGF4_7_00}; + vec_float4 c01b = (vec_float4) {SDM_LGF4_4_01, SDM_LGF4_5_01, SDM_LGF4_6_01, SDM_LGF4_7_01}; + vec_float4 c02b = (vec_float4) {SDM_LGF4_4_02, SDM_LGF4_5_02, SDM_LGF4_6_02, SDM_LGF4_7_02}; + vec_float4 c03b = (vec_float4) {SDM_LGF4_4_03, SDM_LGF4_5_03, SDM_LGF4_6_03, SDM_LGF4_7_03}; + vec_float4 c04b = (vec_float4) {SDM_LGF4_4_04, SDM_LGF4_5_04, SDM_LGF4_6_04, SDM_LGF4_7_04}; + vec_float4 c05b = (vec_float4) {SDM_LGF4_4_05, SDM_LGF4_5_05, SDM_LGF4_6_05, SDM_LGF4_7_05}; + vec_float4 c06b = (vec_float4) {SDM_LGF4_4_06, SDM_LGF4_5_06, SDM_LGF4_6_06, SDM_LGF4_7_06}; + vec_float4 c07b = (vec_float4) {SDM_LGF4_4_07, SDM_LGF4_5_07, SDM_LGF4_6_07, SDM_LGF4_7_07}; + vec_float4 c08b = (vec_float4) {SDM_LGF4_4_08, SDM_LGF4_5_08, SDM_LGF4_6_08, SDM_LGF4_7_08}; + vec_float4 c09b = (vec_float4) {SDM_LGF4_4_09, SDM_LGF4_5_09, SDM_LGF4_6_09, SDM_LGF4_7_09}; + vec_float4 c10b = (vec_float4) {SDM_LGF4_4_10, SDM_LGF4_5_10, SDM_LGF4_6_10, SDM_LGF4_7_10}; + vec_float4 c11b = (vec_float4) {SDM_LGF4_4_11, SDM_LGF4_5_11, SDM_LGF4_6_11, SDM_LGF4_7_11}; + + + vec_uchar16 shuffle0 = (vec_uchar16) spu_splats(0x00010203); + vec_uchar16 shuffle1 = (vec_uchar16) spu_splats(0x04050607); + vec_uchar16 shuffle2 = (vec_uchar16) spu_splats(0x08090A0B); + vec_uchar16 shuffle3 = (vec_uchar16) spu_splats(0x0C0D0E0F); + vec_uchar16 shuffle4 = (vec_uchar16) spu_splats(0x10111213); + vec_uchar16 shuffle5 = (vec_uchar16) spu_splats(0x14151617); + vec_uchar16 shuffle6 = (vec_uchar16) spu_splats(0x18191A1B); + vec_uchar16 shuffle7 = (vec_uchar16) spu_splats(0x1C1D1E1F); + /* - * Adjust result if we shifted x into Stirling range. - * - * lgamma(x) = lgamma(x + n) - ln(x(x+1)(x+2)...(x+n-1) - * + * Determine the shuffle pattern based on which partition + * each element of x is in. */ - shresult = spu_mul(xabs, spu_add(xabs, spu_splats(1.0f))); - shresult = spu_mul(shresult, spu_add(xabs, spu_splats(2.0f))); - shresult = spu_mul(shresult, spu_add(xabs, spu_splats(3.0f))); - shresult = spu_mul(shresult, spu_add(xabs, spu_splats(4.0f))); - shresult = spu_mul(shresult, spu_add(xabs, spu_splats(5.0f))); - shresult = _logf4(shresult); - shresult = spu_sub(stresult, shresult); - stresult = spu_sel(stresult, shresult, isshifted); + vec_uchar16 gt_r1start = (vec_uchar16)spu_cmpgt(xabs, r1start); + vec_uchar16 gt_r2start = (vec_uchar16)spu_cmpgt(xabs, r2start); + vec_uchar16 gt_r3start = (vec_uchar16)spu_cmpgt(xabs, r3start); + vec_uchar16 gt_r4start = (vec_uchar16)spu_cmpgt(xabs, r4start); + vec_uchar16 gt_r5start = (vec_uchar16)spu_cmpgt(xabs, r5start); + vec_uchar16 gt_r6start = (vec_uchar16)spu_cmpgt(xabs, r6start); + vec_uchar16 gt_r7start = (vec_uchar16)spu_cmpgt(xabs, r7start); + + vec_uchar16 shufflepattern; + shufflepattern = spu_sel(shuffle0, shuffle1, gt_r1start); + shufflepattern = spu_sel(shufflepattern, shuffle2, gt_r2start); + shufflepattern = spu_sel(shufflepattern, shuffle3, gt_r3start); + shufflepattern = spu_sel(shufflepattern, shuffle4, gt_r4start); + shufflepattern = spu_sel(shufflepattern, shuffle5, gt_r5start); + shufflepattern = spu_sel(shufflepattern, shuffle6, gt_r6start); + shufflepattern = spu_sel(shufflepattern, shuffle7, gt_r7start); + + + + /* Use the shuffle pattern to select the coefficients */ + + vec_float4 coeff_00 = spu_shuffle(c00a, c00b, shufflepattern); + vec_float4 coeff_01 = spu_shuffle(c01a, c01b, shufflepattern); + vec_float4 coeff_02 = spu_shuffle(c02a, c02b, shufflepattern); + vec_float4 coeff_03 = spu_shuffle(c03a, c03b, shufflepattern); + vec_float4 coeff_04 = spu_shuffle(c04a, c04b, shufflepattern); + vec_float4 coeff_06 = spu_shuffle(c06a, c06b, shufflepattern); + vec_float4 coeff_07 = spu_shuffle(c07a, c07b, shufflepattern); + vec_float4 coeff_05 = spu_shuffle(c05a, c05b, shufflepattern); + vec_float4 coeff_08 = spu_shuffle(c08a, c08b, shufflepattern); + vec_float4 coeff_09 = spu_shuffle(c09a, c09b, shufflepattern); + vec_float4 coeff_10 = spu_shuffle(c10a, c10b, shufflepattern); + vec_float4 coeff_11 = spu_shuffle(c11a, c11b, shufflepattern); + + vec_float4 xoffset = spu_shuffle(xoffseta, xoffsetb, shufflepattern); + vec_uint4 tcorrection = spu_shuffle(tcorra, tcorrb, shufflepattern); + vec_uint4 isrational = spu_shuffle(israta, isratb, shufflepattern); /* - * Select either Maclaurin or Stirling result before Negative X calc. + * We've completed the coeff. setup. Now we actually do the + * approximation below. */ - vec_uint4 useStirlings = spu_cmpgt(xabs, mac_switch); - result = spu_sel(mresult, stresult, useStirlings); + + /* Adjust x value here (for approximations about a point) */ + vec_float4 xappr = spu_sub(xabs, xoffset); + + /* If in Stirling partition, do some setup before the madds */ + xappr = spu_sel(xappr, inv_xsqu, gt_r7start); + + + + /* Now we do the multiplies - either a big polynomial or + * a rational approximation. Use Horner's method. + */ + result = coeff_11; + result = spu_madd(xappr, result, coeff_10); + result = spu_madd(xappr, result, coeff_09); + result = spu_madd(xappr, result, coeff_08); + result = spu_madd(xappr, result, coeff_07); + result = spu_madd(xappr, result, coeff_06); + + /* For rational approximations, we save numerator. */ + vec_float4 resultn = result; + + /* For rational appr,, reset result for calculation of denominator. */ + result = spu_sel(result, spu_splats(0.0f), isrational); + + result = spu_madd(xappr, result, coeff_05); + result = spu_madd(xappr, result, coeff_04); + result = spu_madd(xappr, result, coeff_03); + result = spu_madd(xappr, result, coeff_02); + result = spu_madd(xappr, result, coeff_01); + result = spu_madd(xappr, result, coeff_00); + + /* Select either the polynomial or rational result */ + result = spu_sel(result, _divf4(resultn, result), isrational); + + /* + * Now we have to do a bit of additional calculations for + * partitions that weren't simple polynomial or rational + * approximations. + */ + + /* Finish the Near 0 formula */ + result = spu_sel(spu_sub(result, ln_x), result, gt_r1start); + + /* Finish Stirling's Approximation */ + vec_float4 resultstirling = spu_madd(spu_sub(xabs, spu_splats(0.5f)), ln_x, halflog2pi); + resultstirling = spu_sub(resultstirling, xabs); + resultstirling = spu_add(spu_mul(result,inv_x), resultstirling); + result = spu_sel(result, resultstirling, gt_r7start); + + + /* Adjust due to systematic truncation */ + result = (vec_float4)spu_add((vec_uint4)result, tcorrection); + /* * Approximation for Negative X @@ -202,29 +498,30 @@ static __inline vector float _lgammaf4(vector float x) * lgamma(x) = log(pi/(-x sin(pi x))) - lgamma(-x) * */ - nresult = spu_mul(x, _sinf4(spu_mul(x, pi))); + vec_float4 nresult = spu_mul(x, _sinf4(spu_mul(x, pi))); nresult = spu_andc(nresult, sign_maskf); - nresult = spu_sub(_logf4(pi), spu_add(result, _logf4(nresult))); - + nresult = spu_sub(logpi, spu_add(result, _logf4(nresult))); + nresult = (vec_float4)spu_add((vec_uint4)nresult, spu_splats(1u)); - /* - * Select between the negative or positive x approximations. - */ result = spu_sel(nresult, result, gt0); - /* - * Finally, special cases/errors. - */ /* - * x = non-positive integer, return infinity. + * Special Cases */ + + /* x = non-positive integer, return infinity */ + vec_uint4 isnonposint = spu_andc(spu_cmpeq(x, xtrunc), gt0); + result = spu_sel(result, inff, spu_or(isnonposint, spu_cmpgt(x, spu_splats(4.2e36f)))); result = spu_sel(result, inff, spu_andc(spu_cmpeq(x, xtrunc), gt0)); - /* x = +/- infinite or nan, return |x| */ - isnaninf = spu_cmpgt((vec_uint4)xabs, 0x7FEFFFFF); + /* Zeros of function */ + result = spu_sel(result, zerof, ret_zero); + + /* x = +/- infinity or nan, return |x| */ result = spu_sel(result, xabs, isnaninf); + return result; } diff --git a/newlib/libm/machine/spu/headers/logbf4.h b/newlib/libm/machine/spu/headers/logbf4.h index d3dbeeb93..aa3271b2e 100644 --- a/newlib/libm/machine/spu/headers/logbf4.h +++ b/newlib/libm/machine/spu/headers/logbf4.h @@ -57,19 +57,18 @@ /* * FUNCTION - * vector float _scalbnf4(vector float x, vector signed int exp) + * vector float _logbf4(vector float x) * * DESCRIPTION - * The _scalbnf4 function returns a vector containing each element of x - * multiplied by 2^n computed efficiently. This function is computed - * without the assistance of any floating point operations and as such - * does not set any floating point exceptions. + * The _logbf4 function returns a vector float that contains the exponent + * of the corresponding elements of the input vector x. The exponent is + * defined by: + * x = frac * FLT_RADIX^exp, with frac in [1, FLT_RADIX). * - * Special Cases: - * - if the exponent is 0, then x is either 0 or a subnormal, and the - * result will be returned as 0. - * - if the result if underflows, it will be returned as 0. - * - if the result overflows, it will be returned as FLT_MAX. + * Special Cases: + * x = 0, result is undefined. + * x = NaN, result is NaN. + * x = infinity, +infinity is returned. * */ static __inline vector float _logbf4(vector float x) diff --git a/newlib/libm/machine/spu/headers/nextafterd2.h b/newlib/libm/machine/spu/headers/nextafterd2.h index 78238487a..9a9e0247d 100644 --- a/newlib/libm/machine/spu/headers/nextafterd2.h +++ b/newlib/libm/machine/spu/headers/nextafterd2.h @@ -74,10 +74,13 @@ static __inline vector double _nextafterd2(vector double x, vector double y) { vec_double2 n1ulp = (vec_double2)spu_splats(0x8000000000000001ull); + vector unsigned char mov_carry = {0x80,0x80,0x80, 7, 0x80,0x80,0x80,0x80, + 0x80,0x80,0x80,15, 0x80,0x80,0x80,0x80}; vec_double2 zerod = spu_splats(0.0); vec_llong2 one = spu_splats(1ll); vec_ullong2 xlt0, xgty, xeqy, xeq0; vec_llong2 xllong; + vec_int4 carry; vec_llong2 delta, deltap1; vec_double2 result; @@ -102,10 +105,18 @@ static __inline vector double _nextafterd2(vector double x, vector double y) /* Determine value to add to x */ delta = (vec_llong2)spu_xor(xgty, xlt0); - deltap1 = delta + one; + + //deltap1 = delta + one; + carry = spu_genc((vec_int4)delta, (vec_int4)one); + carry = spu_shuffle(carry, carry, mov_carry); + deltap1 = (vec_llong2)spu_addx((vec_int4)delta, (vec_int4)one, (vec_int4)carry); + delta = spu_sel(deltap1, delta, (vec_ullong2)delta); - xllong = xllong + delta; + //xllong = xllong + delta; + carry = spu_genc((vec_int4)xllong, (vec_int4)delta); + carry = spu_shuffle(carry, carry, mov_carry); + xllong = (vec_llong2)spu_addx((vec_int4)xllong, (vec_int4)delta, (vec_int4)carry); /* Fix the case of x = 0, and answer should be -1 ulp */ result = spu_sel((vec_double2)xllong, n1ulp, spu_and((vec_ullong2)delta, xeq0)); diff --git a/newlib/libm/machine/spu/headers/nextafterf4.h b/newlib/libm/machine/spu/headers/nextafterf4.h index 15a5e3daf..3d2805ccf 100644 --- a/newlib/libm/machine/spu/headers/nextafterf4.h +++ b/newlib/libm/machine/spu/headers/nextafterf4.h @@ -97,10 +97,10 @@ static __inline vector float _nextafterf4(vector float x, vector float y) /* Determine value to add to x */ delta = (vec_int4)spu_xor(xgty, xlt0); - deltap1 = delta + one; + deltap1 = spu_add(delta,one); delta = spu_sel(deltap1, delta, (vec_uint4)delta); - xint = xint + delta; + xint = spu_add(xint, delta); /* Fix the case of x = 0, and answer should be -1 ulp */ result = spu_sel((vec_float4)xint, n1ulp, spu_and((vec_uint4)delta, xeq0)); diff --git a/newlib/libm/machine/spu/headers/recipd2.h b/newlib/libm/machine/spu/headers/recipd2.h index 463769db4..21e00c6be 100644 --- a/newlib/libm/machine/spu/headers/recipd2.h +++ b/newlib/libm/machine/spu/headers/recipd2.h @@ -80,9 +80,7 @@ */ static __inline vector double _recipd2(vector double value_d) { - vector unsigned long long zero = (vector unsigned long long) { 0x0000000000000000ULL, 0x0000000000000000ULL }; vector unsigned long long expmask = (vector unsigned long long) { 0x7FF0000000000000ULL, 0x7FF0000000000000ULL }; - vector unsigned long long signmask = (vector unsigned long long) { 0x8000000000000000ULL, 0x8000000000000000ULL }; vector float x0; vector float value; vector float two = spu_splats(2.0f); @@ -94,6 +92,7 @@ static __inline vector double _recipd2(vector double value_d) * point exponents that are out of single precision range. */ bias = spu_xor(spu_and(value_d, (vector double)expmask), (vector double)expmask); + value = spu_roundtf(spu_mul(value_d, bias)); x0 = spu_re(value); x1 = spu_extend(spu_mul(x0, spu_nmsub(value, x0, two))); @@ -102,10 +101,22 @@ static __inline vector double _recipd2(vector double value_d) x3 = spu_mul(x2, spu_nmsub(value_d, x2, two_d)); /* Handle input = +/- infinity or +/-0. */ - vec_double2 xabs = spu_andc(value_d, (vec_double2)signmask); - vec_ullong2 zeroinf = spu_or(spu_cmpeq(xabs, (vec_double2)expmask), - spu_cmpeq(xabs, (vec_double2)zero)); - x3 = spu_sel(x3, spu_xor(value_d, (vector double)expmask), zeroinf); + +#ifdef __SPU_EDP__ + vec_ullong2 is0inf = spu_testsv(value_d, SPU_SV_NEG_ZERO | SPU_SV_POS_ZERO | + SPU_SV_NEG_INFINITY | SPU_SV_POS_INFINITY); +#else + vec_double2 nzero = spu_splats(-0.0); + vec_double2 xabs = spu_andc(value_d, nzero); + vector unsigned char swap = (vector unsigned char) {4,5,6,7, 0,1,2,3, 12,13,14,15, 8,9,10,11}; + vec_uint4 isinf = spu_cmpeq((vec_uint4)xabs, (vec_uint4)expmask); + vec_uint4 iszero = spu_cmpeq((vec_uint4)xabs, 0); + isinf = spu_and(isinf, spu_shuffle(isinf, isinf, swap)); + iszero = spu_and(iszero, spu_shuffle(iszero, iszero, swap)); + vec_ullong2 is0inf = (vec_ullong2)spu_or(isinf, iszero); +#endif /* __SPU_EDP__ */ + + x3 = spu_sel(x3, spu_xor(value_d, (vector double)expmask), is0inf); return (x3); } diff --git a/newlib/libm/machine/spu/headers/simdmath.h b/newlib/libm/machine/spu/headers/simdmath.h index bba8debd1..bbfecf27f 100644 --- a/newlib/libm/machine/spu/headers/simdmath.h +++ b/newlib/libm/machine/spu/headers/simdmath.h @@ -212,9 +212,9 @@ vector double fmind2(vector double x, vector double y); vector double fmodd2(vector double x, vector double y); vector float fmodf4_fast(vector float x, vector float y); vector signed long long fpclassifyd2(vector double x); -vector double frexpd2(vector double x, vector signed int *pexp); +vector double frexpd2(vector double x, vector signed long long *pexp); vector double hypotd2(vector double x, vector double y); -vector signed int ilogbd2(vector double x); +vector signed long long ilogbd2(vector double x); vector unsigned long long is0denormd2(vector double x); vector unsigned long long isequald2(vector double x, vector double y); vector unsigned long long isfinited2(vector double x); @@ -227,7 +227,7 @@ vector unsigned long long islessgreaterd2(vector double x, vector double y); vector unsigned long long isnand2(vector double x); vector unsigned long long isnormald2(vector double x); vector unsigned long long isunorderedd2(vector double x, vector double y); -vector double ldexpd2(vector double x, vector signed int exp); +vector double ldexpd2(vector double x, vector signed long long exp); vector signed long long llabsi2(vector signed long long x); lldivi2_t lldivi2(vector signed long long x, vector signed long long y); lldivu2_t lldivu2(vector unsigned long long x, vector unsigned long long y); @@ -237,6 +237,7 @@ vector signed long long llroundd2(vector double x); vector double log10d2(vector double x); vector double log1pd2(vector double x); vector double log2d2(vector double x); +vector double logbd2(vector double x); vector double logd2(vector double x); vector double modfd2(vector double x, vector double* pint); vector double nearbyintd2(vector double x); @@ -247,7 +248,7 @@ vector double powd2(vector double x, vector double y); vector double recipd2(vector double value_d); vector float recipf4_fast(vector float a); vector double remainderd2(vector double x, vector double y); -vector double remquod2(vector double x, vector double y, vector signed int *quo); +vector double remquod2(vector double x, vector double y, vector signed long long *quo); vector double rintd2(vector double x); vector double roundd2(vector double x); vector double rsqrtd2(vector double x); |