IMF LA open-sourcing. FP64 acos.

Change-Id: Ied8a9b6eef4bad4566f0766347fd24b3d9e809ab

Author: grey-eminence

IGC/BiFModule/Implementation/IMF/FP64/acos_d_la.cl

@@ -0,0 +1,182 @@
+/*===================== begin_copyright_notice ==================================
+
+Copyright (c) 2017 Intel Corporation
+
+Permission is hereby granted, free of charge, to any person obtaining a
+copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be included
+in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+
+======================= end_copyright_notice ==================================*/
+
+#include "../imf.h"
+#pragma OPENCL FP_CONTRACT OFF
+typedef struct
+{
+    unsigned long SgnBit;
+    unsigned long OneHalf;
+    unsigned long SmallNorm;
+    unsigned long dRsqrtMsk;
+    unsigned long MOne;
+    unsigned long HalfMask;
+    unsigned long Two;
+    unsigned long sqrt_coeff[4];
+    unsigned long poly_coeff[12];
+    unsigned long PiL;
+    unsigned long PiH;
+    unsigned long Pi2L;
+    unsigned long Pi2H;
+    unsigned long Zero;
+    unsigned long SgnMask;
+    unsigned long NanMask;
+    unsigned long ep_coeff[6];
+
+    unsigned long dInfs[2];
+    unsigned long dOnes[2];
+    unsigned long dZeros[2];
+} __internal_dacos_la_data_t;
+static __constant __internal_dacos_la_data_t __internal_dacos_la_data = {
+
+    0x8000000000000000uL, 0x3fe0000000000000uL, 0x3000000000000000uL, 0xffffff0000000000uL, 0xbff0000000000000uL, 0xfffffffffc000000uL,
+        0x4000000000000000uL, {
+                               0xbf918000993B24C3uL, 0x3fa400006F70D42DuL, 0xbfb7FFFFFFFFFE97uL, 0x3fcFFFFFFFFFFF9DuL}
+
+    , {
+       0x3fa07520C70EB909uL, 0xbf90FB17F7DBB0EDuL, 0x3f943F44BFBC3BAEuL, 0x3f7A583395D45ED5uL, 0x3f88F8DC2AFCCAD6uL, 0x3f8C6DBBCB88BD57uL,
+       0x3f91C6DCF538AD2EuL, 0x3f96E89CEBDEFadduL, 0x3f9F1C72E13AD8BEuL, 0x3fa6DB6DB3B445F8uL, 0x3fb333333337E0DEuL, 0x3fc555555555529CuL}
+
+    , 0x3ca1a62633145c07uL, 0x400921fb54442d18uL, 0x3c91a62633145c07uL, 0x3ff921fb54442d18uL, 0x0000000000000000uL, 0x8000000000000000uL,
+        0xfffc000000000000uL, {
+                               0x3fa36C5AF645A11EuL, 0x3f8CE147EA9E9282uL, 0x3fa056B4151FA155uL, 0x3fa6C8ED2A4CCE54uL, 0x3fb33399EBF85B6AuL,
+                               0x3fc5555480C83A45uL}
+
+    , {0x7ff0000000000000uL, 0xfff0000000000000uL}
+
+    , {0x3ff0000000000000uL, 0xbff0000000000000uL}
+
+    , {0x0000000000000000uL, 0x8000000000000000uL}
+
+};
+static __constant int_double __dacos_la_c12 = { 0x3fa07520c559a401UL };
+static __constant int_double __dacos_la_c11 = { 0xbf90fb17f2824aa0UL };
+static __constant int_double __dacos_la_c10 = { 0x3f943f44bc0c21f0UL };
+static __constant int_double __dacos_la_c9 = { 0x3f7a58339bb6f20cUL };
+static __constant int_double __dacos_la_c8 = { 0x3f88f8dc2a3c76d7UL };
+static __constant int_double __dacos_la_c7 = { 0x3f8c6dbbcba98e67UL };
+static __constant int_double __dacos_la_c6 = { 0x3f91c6dcf536c796UL };
+static __constant int_double __dacos_la_c5 = { 0x3f96e89cebdf209cUL };
+static __constant int_double __dacos_la_c4 = { 0x3f9f1c72e13ad6d3UL };
+static __constant int_double __dacos_la_c3 = { 0x3fa6db6db3b44600UL };
+static __constant int_double __dacos_la_c2 = { 0x3fb333333337e0deUL };
+static __constant int_double __dacos_la_c1 = { 0x3fc555555555529cUL };
+static __constant int_double __dacos_la_c0 = { 0x3ff0000000000000UL };
+static __constant int_double __dacos_la_pi2h = { 0x3ff921fb54442d18UL };
+static __constant int_double __dacos_la_pi2l = { 0x3c91a62633145c07UL };
+
+static __constant int_double __dacos_la_pih = { 0x400921fb54442d18UL };
+static __constant int_double __dacos_la_pil = { 0x3ca1a62633145c07UL };
+
+static __constant int_float __dacos_la_small_float = { 0x01800000u };
+
+__attribute__((always_inline))
+inline int __internal_dacos_la_cout (double *pxin, double *pres)
+{
+    int nRet = 0;
+    double xin = *pxin;
+    int_double y, res;
+    {
+        int_double x, xa, RS, Shh2, High, R0;
+        double R, E, poly, Sh;
+        unsigned long sgn_x;
+        float yf;
+        int_float fcorr;
+
+        x.f = xin;
+
+        xa.f = __builtin_spirv_OpenCL_fabs_f64 (x.f);
+
+        sgn_x = x.w ^ xa.w;
+
+        y.f = __builtin_spirv_OpenCL_fma_f64_f64_f64 (-(0.5), xa.f, 0.5);
+
+        R = xin * xin;
+        R = __builtin_spirv_OpenCL_fmin_f64_f64 (R, y.f);
+
+        High.f = sgn_x ? __dacos_la_pih.f : 0;
+        High.f = (xa.f <= 0.5) ? __dacos_la_pi2h.f : High.f;
+
+        yf = (float) y.f;
+
+        yf += __dacos_la_small_float.f;
+        yf = 1.0f / __builtin_spirv_OpenCL_sqrt_f32 (yf);
+        RS.f = (double) (yf);
+
+        RS.w |= sgn_x;
+
+        Sh = (y.f * RS.f);
+
+        Shh2.f = -2.0 * Sh;
+
+        E = (__builtin_spirv_OpenCL_fma_f64_f64_f64 (-(RS.f), Sh, 1.0));
+
+        R0.f = __builtin_spirv_OpenCL_fma_f64_f64_f64 (0.375, E, 0.5);
+        R0.f *= E;
+
+        R0.f = __builtin_spirv_OpenCL_fma_f64_f64_f64 (R0.f, Shh2.f, Shh2.f);
+
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (__dacos_la_c12.f, R, __dacos_la_c11.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c10.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c9.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c8.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c7.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c6.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c5.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c4.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c3.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c2.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c1.f);
+        poly = __builtin_spirv_OpenCL_fma_f64_f64_f64 (poly, R, __dacos_la_c0.f);
+        R0.f = (xa.f <= 0.5) ? x.f : R0.f;
+
+        res.f = (__builtin_spirv_OpenCL_fma_f64_f64_f64 (-(poly), R0.f, High.f));
+
+    }
+
+    *pres = res.f;
+    nRet = (y.f >= 0) ? 0 : 1;
+
+    return nRet;
+}
+
+double __ocl_svml_acos (double a)
+{
+
+    double va1;
+    double vr1;
+    unsigned int vm;
+
+    double r;
+
+    va1 = a;;
+
+    __internal_dacos_la_cout (&va1, &vr1);
+    r = vr1;;
+
+    return r;
+
+}

IGC/BiFModule/Implementation/Math/acos.cl

@@ -30,17 +30,15 @@ SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 GENERATE_VECTOR_FUNCTIONS_1ARG( __builtin_spirv_OpenCL_acos, float, float, f32 )
 
 #if defined(cl_khr_fp64)
-
-    #include "../ExternalLibraries/libclc/doubles.cl"
-
+    #include "../IMF/FP64/acos_d_la.cl"
 #endif // defined(cl_khr_fp64)
 
 
 #if defined(cl_khr_fp64)
 
 INLINE double __builtin_spirv_OpenCL_acos_f64( double x )
 {
-        return libclc_acos_f64(x);
+    return __ocl_svml_acos(x);
 }
 
 GENERATE_VECTOR_FUNCTIONS_1ARG( __builtin_spirv_OpenCL_acos, double, double, f64 )