blob: 8d5a904e416710fe8c94a29477e82cf6bcbd2753 [file] [log] [blame]
/*
* Copyright (c) 2016 The University of Virginia
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met: redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer;
* redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution;
* neither the name of the copyright holders nor the names of its
* contributors may be used to endorse or promote products derived from
* this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#pragma once
#include <cstdint>
#include <limits>
#include "insttest.h"
namespace D
{
constexpr inline uint64_t
bits(double d)
{
return reinterpret_cast<uint64_t&>(d);
}
constexpr inline double
number(uint64_t b)
{
return reinterpret_cast<double&>(b);
}
inline bool
isquietnan(double f)
{
return std::isnan(f) && (bits(f)&0x0008000000000000ULL) != 0;
}
inline bool
issignalingnan(double f)
{
return std::isnan(f) && (bits(f)&0x0008000000000000ULL) == 0;
}
inline double
load(double mem)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fld %0,%1"
: "=f" (fd)
: "m" (mem));
return fd;
}
inline double
store(double fs)
{
double mem = std::numeric_limits<double>::signaling_NaN();
asm volatile("fsd %1,%0" : "=m" (mem) : "f" (fs));
return mem;
}
inline double
fmadd_d(double fs1, double fs2, double fs3)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FR4OP("fmadd.d", fd, fs1, fs2, fs3);
return fd;
}
inline double
fmsub_d(double fs1, double fs2, double fs3)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FR4OP("fmsub.d", fd, fs1, fs2, fs3);
return fd;
}
inline double
fnmsub_d(double fs1, double fs2, double fs3)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FR4OP("fnmsub.d", fd, fs1, fs2, fs3);
return fd;
}
inline double
fnmadd_d(double fs1, double fs2, double fs3)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FR4OP("fnmadd.d", fd, fs1, fs2, fs3);
return fd;
}
inline double
fadd_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fadd.d", fd, fs1, fs2);
return fd;
}
inline double
fsub_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fsub.d", fd, fs1, fs2);
return fd;
}
inline double
fmul_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fmul.d", fd, fs1, fs2);
return fd;
}
inline double
fdiv_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fdiv.d", fd, fs1, fs2);
return fd;
}
inline double
fsqrt_d(double fs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fsqrt.d %0,%1" : "=f" (fd) : "f" (fs1));
return fd;
}
inline double
fsgnj_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fsgnj.d", fd, fs1, fs2);
return fd;
}
inline double
fsgnjn_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fsgnjn.d", fd, fs1, fs2);
return fd;
}
inline double
fsgnjx_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fsgnjx.d", fd, fs1, fs2);
return fd;
}
inline double
fmin_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fmin.d", fd, fs1, fs2);
return fd;
}
inline double
fmax_d(double fs1, double fs2)
{
double fd = std::numeric_limits<double>::signaling_NaN();
FROP("fmax.d", fd, fs1, fs2);
return fd;
}
inline float
fcvt_s_d(double fs1)
{
float fd = std::numeric_limits<float>::signaling_NaN();
asm volatile("fcvt.s.d %0,%1" : "=f" (fd) : "f" (fs1));
return fd;
}
inline double
fcvt_d_s(float fs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fcvt.d.s %0,%1" : "=f" (fd) : "f" (fs1));
return fd;
}
inline bool
feq_d(double fs1, double fs2)
{
bool rd = false;
asm volatile("feq.d %0,%1,%2" : "=r" (rd) : "f" (fs1), "f" (fs2));
return rd;
}
inline bool
flt_d(double fs1, double fs2)
{
bool rd = false;
asm volatile("flt.d %0,%1,%2" : "=r" (rd) : "f" (fs1), "f" (fs2));
return rd;
}
inline bool
fle_d(double fs1, double fs2)
{
bool rd = false;
asm volatile("fle.d %0,%1,%2" : "=r" (rd) : "f" (fs1), "f" (fs2));
return rd;
}
inline uint64_t
fclass_d(double fs1)
{
uint64_t rd = -1;
asm volatile("fclass.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline int64_t
fcvt_w_d(double fs1)
{
int64_t rd = 0;
asm volatile("fcvt.w.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline uint64_t
fcvt_wu_d(double fs1)
{
uint64_t rd = 0;
asm volatile("fcvt.wu.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline float
fcvt_d_w(int64_t rs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fcvt.d.w %0,%1" : "=f" (fd) : "r" (rs1));
return fd;
}
inline double
fcvt_d_wu(uint64_t rs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fcvt.d.wu %0,%1" : "=f" (fd) : "r" (rs1));
return fd;
}
inline int64_t
fcvt_l_d(double fs1)
{
int64_t rd = 0;
asm volatile("fcvt.l.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline uint64_t
fcvt_lu_d(double fs1)
{
uint64_t rd = 0;
asm volatile("fcvt.lu.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline uint64_t
fmv_x_d(double fs1)
{
uint64_t rd = 0;
asm volatile("fmv.x.d %0,%1" : "=r" (rd) : "f" (fs1));
return rd;
}
inline double
fcvt_d_l(int64_t rs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fcvt.d.l %0,%1" : "=f" (fd) : "r" (rs1));
return fd;
}
inline double
fcvt_d_lu(uint64_t rs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fcvt.d.lu %0,%1" : "=f" (fd) : "r" (rs1));
return fd;
}
inline double
fmv_d_x(uint64_t rs1)
{
double fd = std::numeric_limits<double>::signaling_NaN();
asm volatile("fmv.d.x %0,%1" : "=f" (fd) : "r" (rs1));
return fd;
}
} // namespace D