1
0
Fork 0
mirror of https://gitlab.com/sortix/sortix.git synced 2023-02-13 20:55:38 -05:00
sortix--sortix/libm/arch/i387/fenv.c
Jonas 'Sortie' Termansen 5980be9b3c Add Sortix Math Library.
This work is based in part on code from NetBSD libm, libc and kernel.

The library is partly public domain and partly BSD-style licensed.
2013-12-17 14:30:39 +01:00

522 lines
12 KiB
C

/* $NetBSD: fenv.c,v 1.3.8.1.6.1 2013/06/14 02:43:36 msaitoh Exp $ */
/*-
* Copyright (c) 2004-2005 David Schultz <das@FreeBSD.ORG>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.
*/
#include <sys/cdefs.h>
__RCSID("$NetBSD: fenv.c,v 1.3.8.1.6.1 2013/06/14 02:43:36 msaitoh Exp $");
#include <assert.h>
#include <fenv.h>
#include <stddef.h>
#include <stdint.h>
#include <string.h>
/* Load x87 Control Word */
#define __fldcw(__cw) __asm__ __volatile__ \
("fldcw %0" : : "m" (__cw))
/* No-Wait Store Control Word */
#define __fnstcw(__cw) __asm__ __volatile__ \
("fnstcw %0" : "=m" (*(__cw)))
/* No-Wait Store Status Word */
#define __fnstsw(__sw) __asm__ __volatile__ \
("fnstsw %0" : "=am" (*(__sw)))
/* No-Wait Clear Exception Flags */
#define __fnclex() __asm__ __volatile__ \
("fnclex")
/* Load x87 Environment */
#define __fldenv(__env) __asm__ __volatile__ \
("fldenv %0" : : "m" (__env))
/* No-Wait Store x87 environment */
#define __fnstenv(__env) __asm__ __volatile__ \
("fnstenv %0" : "=m" (*(__env)))
/* Check for and handle pending unmasked x87 pending FPU exceptions */
#define __fwait(__env) __asm__ __volatile__ \
("fwait")
/* Load the MXCSR register */
#define __ldmxcsr(__mxcsr) __asm__ __volatile__ \
("ldmxcsr %0" : : "m" (__mxcsr))
/* Store the MXCSR register state */
#define __stmxcsr(__mxcsr) __asm__ __volatile__ \
("stmxcsr %0" : "=m" (*(__mxcsr)))
/*
* The following constant represents the default floating-point environment
* (that is, the one installed at program startup) and has type pointer to
* const-qualified fenv_t.
*
* It can be used as an argument to the functions within the <fenv.h> header
* that manage the floating-point environment, namely fesetenv() and
* feupdateenv().
*
* x87 fpu registers are 16bit wide. The upper bits, 31-16, are marked as
* RESERVED. We provide a partial floating-point environment, where we
* define only the lower bits. The reserved bits are extracted and set by the
* consumers of FE_DFL_ENV, during runtime.
*/
fenv_t __fe_dfl_env = {
{
__NetBSD_NPXCW__, /* Control word register */
0x0, /* Unused */
0x0000, /* Status word register */
0x0, /* Unused */
0x0000ffff, /* Tag word register */
0x0, /* Unused */
{
0x0000, 0x0000,
0x0000, 0xffff
}
},
__INITIAL_MXCSR__ /* MXCSR register */
};
/*
* Test for SSE support on this processor.
*
* We need to use ldmxcsr/stmxcsr to get correct results if any part
* of the program was compiled to use SSE floating-point, but we can't
* use SSE on older processors.
*
* In order to do so, we need to query the processor capabilities via the CPUID
* instruction. We can make it even simpler though, by querying the machdep.sse
* sysctl.
*/
static int __HAS_SSE = 0;
static void __test_sse(void) __attribute__ ((constructor));
static void __test_sse(void)
{
#if 0
size_t oldlen = sizeof(__HAS_SSE);
int rv;
/* TODO: Properly detect whether see is available! */
rv = sysctlbyname("machdep.sse", &__HAS_SSE, &oldlen, NULL, 0);
if (rv == -1)
__HAS_SSE = 0;
#elif defined(__SSE__)
__HAS_SSE = 1;
#elif defined(__sortix__)
__HAS_SSE = 1; /* Currently unconditionally enabled and needed to boot. */
#else
__HAS_SSE = 0;
#endif
}
/*
* The feclearexcept() function clears the supported floating-point exceptions
* represented by `excepts'.
*/
int
feclearexcept(int excepts)
{
fenv_t env;
uint32_t mxcsr;
int ex;
assert((excepts & ~FE_ALL_EXCEPT) == 0);
ex = excepts & FE_ALL_EXCEPT;
/* It's ~3x faster to call fnclex, than store/load fp env */
if (ex == FE_ALL_EXCEPT) {
__fnclex();
} else {
__fnstenv(&env);
env.x87.status &= ~ex;
__fldenv(env);
}
if (__HAS_SSE) {
__stmxcsr(&mxcsr);
mxcsr &= ~ex;
__ldmxcsr(mxcsr);
}
/* Success */
return (0);
}
/*
* The fegetexceptflag() function stores an implementation-defined
* representation of the states of the floating-point status flags indicated by
* the argument excepts in the object pointed to by the argument flagp.
*/
int
fegetexceptflag(fexcept_t *flagp, int excepts)
{
uint32_t mxcsr;
uint16_t status;
int ex;
assert(flagp != NULL);
assert((excepts & ~FE_ALL_EXCEPT) == 0);
ex = excepts & FE_ALL_EXCEPT;
__fnstsw(&status);
if (__HAS_SSE)
__stmxcsr(&mxcsr);
else
mxcsr = 0;
*flagp = (mxcsr | status) & ex;
/* Success */
return (0);
}
/*
* The feraiseexcept() function raises the supported floating-point exceptions
* represented by the argument `excepts'.
*
* The standard explicitly allows us to execute an instruction that has the
* exception as a side effect, but we choose to manipulate the status register
* directly.
*
* The validation of input is being deferred to fesetexceptflag().
*/
int
feraiseexcept(int excepts)
{
fexcept_t ex;
assert((excepts & ~FE_ALL_EXCEPT) == 0);
ex = excepts & FE_ALL_EXCEPT;
fesetexceptflag(&ex, excepts);
__fwait();
/* Success */
return (0);
}
/*
* This function sets the floating-point status flags indicated by the argument
* `excepts' to the states stored in the object pointed to by `flagp'. It does
* NOT raise any floating-point exceptions, but only sets the state of the flags.
*/
int
fesetexceptflag(const fexcept_t *flagp, int excepts)
{
fenv_t env;
uint32_t mxcsr;
int ex;
assert(flagp != NULL);
assert((excepts & ~FE_ALL_EXCEPT) == 0);
ex = excepts & FE_ALL_EXCEPT;
__fnstenv(&env);
env.x87.status &= ~ex;
env.x87.status |= *flagp & ex;
__fldenv(env);
if (__HAS_SSE) {
__stmxcsr(&mxcsr);
mxcsr &= ~ex;
mxcsr |= *flagp & ex;
__ldmxcsr(mxcsr);
}
/* Success */
return (0);
}
/*
* The fetestexcept() function determines which of a specified subset of the
* floating-point exception flags are currently set. The `excepts' argument
* specifies the floating-point status flags to be queried.
*/
int
fetestexcept(int excepts)
{
uint32_t mxcsr;
uint16_t status;
int ex;
assert((excepts & ~FE_ALL_EXCEPT) == 0);
ex = excepts & FE_ALL_EXCEPT;
__fnstsw(&status);
if (__HAS_SSE)
__stmxcsr(&mxcsr);
else
mxcsr = 0;
return ((status | mxcsr) & ex);
}
int
fegetround(void)
{
uint16_t control;
/*
* We assume that the x87 and the SSE unit agree on the
* rounding mode. Reading the control word on the x87 turns
* out to be about 5 times faster than reading it on the SSE
* unit on an Opteron 244.
*/
__fnstcw(&control);
return (control & __X87_ROUND_MASK);
}
/*
* The fesetround() function shall establish the rounding direction represented
* by its argument round. If the argument is not equal to the value of a
* rounding direction macro, the rounding direction is not changed.
*/
int
fesetround(int round)
{
uint32_t mxcsr;
uint16_t control;
if (round & ~__X87_ROUND_MASK) {
/* Failure */
return (-1);
}
__fnstcw(&control);
control &= ~__X87_ROUND_MASK;
control |= round;
__fldcw(control);
if (__HAS_SSE) {
__stmxcsr(&mxcsr);
mxcsr &= ~(__X87_ROUND_MASK << __SSE_ROUND_SHIFT);
mxcsr |= round << __SSE_ROUND_SHIFT;
__ldmxcsr(mxcsr);
}
/* Success */
return (0);
}
/*
* The fegetenv() function attempts to store the current floating-point
* environment in the object pointed to by envp.
*/
int
fegetenv(fenv_t *envp)
{
uint32_t mxcsr;
assert(envp != NULL);
/*
* fnstenv masks all exceptions, so we need to restore the old control
* word to avoid this side effect.
*/
__fnstenv(envp);
__fldcw(envp->x87.control);
if (__HAS_SSE) {
__stmxcsr(&mxcsr);
envp->mxcsr = mxcsr;
}
/* Success */
return (0);
}
/*
* The feholdexcept() function saves the current floating-point environment in
* the object pointed to by envp, clears the floating-point status flags, and
* then installs a non-stop (continue on floating-point exceptions) mode, if
* available, for all floating-point exceptions.
*/
int
feholdexcept(fenv_t *envp)
{
uint32_t mxcsr;
assert(envp != NULL);
__fnstenv(envp);
__fnclex();
if (__HAS_SSE) {
__stmxcsr(&mxcsr);
envp->mxcsr = mxcsr;
mxcsr &= ~FE_ALL_EXCEPT;
mxcsr |= FE_ALL_EXCEPT << __SSE_EMASK_SHIFT;
__ldmxcsr(mxcsr);
}
/* Success */
return (0);
}
/*
* The fesetenv() function attempts to establish the floating-point environment
* represented by the object pointed to by envp. The argument `envp' points
* to an object set by a call to fegetenv() or feholdexcept(), or equal a
* floating-point environment macro. The fesetenv() function does not raise
* floating-point exceptions, but only installs the state of the floating-point
* status flags represented through its argument.
*/
int
fesetenv(const fenv_t *envp)
{
fenv_t env;
assert(envp != NULL);
/* Store the x87 floating-point environment */
memset(&env, 0, sizeof(env));
__fnstenv(&env);
__fe_dfl_env.x87.unused1 = env.x87.unused1;
__fe_dfl_env.x87.unused2 = env.x87.unused2;
__fe_dfl_env.x87.unused3 = env.x87.unused3;
memcpy(__fe_dfl_env.x87.others,
env.x87.others,
sizeof(__fe_dfl_env.x87.others) / sizeof(uint32_t));
__fldenv(envp->x87);
if (__HAS_SSE)
__ldmxcsr(envp->mxcsr);
/* Success */
return (0);
}
/*
* The feupdateenv() function saves the currently raised floating-point
* exceptions in its automatic storage, installs the floating-point environment
* represented by the object pointed to by `envp', and then raises the saved
* floating-point exceptions. The argument `envp' shall point to an object set
* by a call to feholdexcept() or fegetenv(), or equal a floating-point
* environment macro.
*/
int
feupdateenv(const fenv_t *envp)
{
fenv_t env;
uint32_t mxcsr;
uint16_t status;
assert(envp != NULL);
/* Store the x87 floating-point environment */
memset(&env, 0, sizeof(env));
__fnstenv(&env);
__fe_dfl_env.x87.unused1 = env.x87.unused1;
__fe_dfl_env.x87.unused2 = env.x87.unused2;
__fe_dfl_env.x87.unused3 = env.x87.unused3;
memcpy(__fe_dfl_env.x87.others,
env.x87.others,
sizeof(__fe_dfl_env.x87.others) / sizeof(uint32_t));
__fnstsw(&status);
if (__HAS_SSE)
__stmxcsr(&mxcsr);
else
mxcsr = 0;
fesetenv(envp);
feraiseexcept((mxcsr | status) & FE_ALL_EXCEPT);
/* Success */
return (0);
}
/*
* The following functions are extentions to the standard
*/
int
feenableexcept(int mask)
{
uint32_t mxcsr, omask;
uint16_t control;
mask &= FE_ALL_EXCEPT;
__fnstcw(&control);
if (__HAS_SSE)
__stmxcsr(&mxcsr);
else
mxcsr = 0;
omask = (control | mxcsr >> __SSE_EMASK_SHIFT) & FE_ALL_EXCEPT;
control &= ~mask;
__fldcw(control);
if (__HAS_SSE) {
mxcsr &= ~(mask << __SSE_EMASK_SHIFT);
__ldmxcsr(mxcsr);
}
return (FE_ALL_EXCEPT & ~omask);
}
int
fedisableexcept(int mask)
{
uint32_t mxcsr, omask;
uint16_t control;
mask &= FE_ALL_EXCEPT;
__fnstcw(&control);
if (__HAS_SSE)
__stmxcsr(&mxcsr);
else
mxcsr = 0;
omask = (control | mxcsr >> __SSE_EMASK_SHIFT) & FE_ALL_EXCEPT;
control |= mask;
__fldcw(control);
if (__HAS_SSE) {
mxcsr |= mask << __SSE_EMASK_SHIFT;
__ldmxcsr(mxcsr);
}
return (FE_ALL_EXCEPT & ~omask);
}
int
fegetexcept(void)
{
uint16_t control;
/*
* We assume that the masks for the x87 and the SSE unit are
* the same.
*/
__fnstcw(&control);
return (~control & FE_ALL_EXCEPT);
}