forked from OSchip/llvm-project
math: Add ilogb ported from amd-builtins
The scalar float/double function bodies are a direct copy/paste with usage of the CLC wrappers to vectorize them. This commit also adds in the FP_ILOGB0 and FP_ILOGBNAN macros which are equal to the results of ilogb(0.0f) and ilogb(float nan) respectively. v2: Add FP_ILOGB0 and FP_ILOGBNAN definitions Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Jan Vesely <jan.vesely@rutgers.edu> v1 Reviewed-by: Tom Stellard <thomas.stellard@amd.com> llvm-svn: 261639
This commit is contained in:
parent
4924c7a2b5
commit
d6d0454231
|
@ -62,6 +62,7 @@
|
|||
#include <clc/math/half_rsqrt.h>
|
||||
#include <clc/math/half_sqrt.h>
|
||||
#include <clc/math/hypot.h>
|
||||
#include <clc/math/ilogb.h>
|
||||
#include <clc/math/ldexp.h>
|
||||
#include <clc/math/log.h>
|
||||
#include <clc/math/log10.h>
|
||||
|
|
|
@ -14,6 +14,9 @@
|
|||
#define FLT_MIN 0x1.0p-126f
|
||||
#define FLT_EPSILON 0x1.0p-23f
|
||||
|
||||
#define FP_ILOGB0 (-2147483647 - 1)
|
||||
#define FP_ILOGBNAN (-2147483647 - 1)
|
||||
|
||||
#define M_E_F 0x1.5bf0a8p+1f
|
||||
#define M_LOG2E_F 0x1.715476p+0f
|
||||
#define M_LOG10E_F 0x1.bcb7b2p-2f
|
||||
|
|
|
@ -0,0 +1,5 @@
|
|||
#define __CLC_BODY <clc/math/ilogb.inc>
|
||||
|
||||
#include <clc/math/gentype.inc>
|
||||
|
||||
#undef __CLC_BODY
|
|
@ -0,0 +1 @@
|
|||
_CLC_OVERLOAD _CLC_DECL __CLC_INTN ilogb(__CLC_GENTYPE x);
|
|
@ -90,6 +90,7 @@ math/frexp.cl
|
|||
math/half_rsqrt.cl
|
||||
math/half_sqrt.cl
|
||||
math/hypot.cl
|
||||
math/ilogb.cl
|
||||
math/clc_ldexp.cl
|
||||
math/ldexp.cl
|
||||
math/log.cl
|
||||
|
|
|
@ -0,0 +1,57 @@
|
|||
/*
|
||||
* Copyright (c) 2015 Advanced Micro Devices, Inc.
|
||||
* Copyright (c) 2016 Aaron Watry
|
||||
*
|
||||
* 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.
|
||||
*/
|
||||
|
||||
#include <clc/clc.h>
|
||||
#include "../clcmacro.h"
|
||||
#include "math.h"
|
||||
|
||||
_CLC_OVERLOAD _CLC_DEF int ilogb(float x) {
|
||||
uint ux = as_uint(x);
|
||||
uint ax = ux & EXSIGNBIT_SP32;
|
||||
int rs = -118 - (int) clz(ux & MANTBITS_SP32);
|
||||
int r = (int) (ax >> EXPSHIFTBITS_SP32) - EXPBIAS_SP32;
|
||||
r = ax < 0x00800000U ? rs : r;
|
||||
r = ax > EXPBITS_SP32 | ax == 0 ? 0x80000000 : r;
|
||||
r = ax == EXPBITS_SP32 ? 0x7fffffff : r;
|
||||
return r;
|
||||
}
|
||||
|
||||
_CLC_UNARY_VECTORIZE(_CLC_OVERLOAD _CLC_DEF, int, ilogb, float);
|
||||
|
||||
#ifdef cl_khr_fp64
|
||||
#pragma OPENCL EXTENSION cl_khr_fp64 : enable
|
||||
|
||||
_CLC_OVERLOAD _CLC_DEF ilogb(double x) {
|
||||
ulong ux = as_ulong(x);
|
||||
ulong ax = ux & ~SIGNBIT_DP64;
|
||||
int r = (int) (ax >> EXPSHIFTBITS_DP64) - EXPBIAS_DP64;
|
||||
int rs = -1011 - (int) clz(ax & MANTBITS_DP64);
|
||||
r = ax < 0x0010000000000000UL ? rs : r;
|
||||
r = ax > 0x7ff0000000000000UL | ax == 0UL ? 0x80000000 : r;
|
||||
r = ax == 0x7ff0000000000000UL ? 0x7fffffff : r;
|
||||
return r;
|
||||
}
|
||||
|
||||
_CLC_UNARY_VECTORIZE(_CLC_OVERLOAD _CLC_DEF, int, ilogb, double);
|
||||
|
||||
#endif // cl_khr_fp64
|
Loading…
Reference in New Issue