glibc/sysdeps/i386/fpu/s_expm1.S
Joseph Myers 554edb23ff Fix expm1 missing underflows (bug 16353).
Similar to various other bugs in this area, some expm1 implementations
do not raise the underflow exception for subnormal arguments, when the
result is tiny and inexact.  This patch forces the exception in a
similar way to previous fixes.

(The issue does not apply to the ldbl-* implementations or to those
for x86 / x86_64 long double.  The change to
sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c is one I missed when
previously fixing bug 16354; the bug in that implementation was
previously latent, but the expm1 fixes stopped it being latent and so
required it to be fixed to avoid spurious underflows from cosh.)

Tested for x86_64 and x86.

	[BZ #16353]
	* sysdeps/i386/fpu/s_expm1.S (dbl_min): New object.
	(__expm1): Force underflow exception for arguments with small
	absolute value.
	* sysdeps/i386/fpu/s_expm1f.S (flt_min): New object.
	(__expm1f): Force underflow exception for arguments with small
	absolute value.
	* sysdeps/ieee754/dbl-64/s_expm1.c: Include <float.h>.
	(__expm1): Force underflow exception for arguments with small
	absolute value.
	* sysdeps/ieee754/flt-32/s_expm1f.c: Include <float.h>.
	(__expm1f): Force underflow exception for arguments with small
	absolute value.
	* sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c (__ieee754_cosh):
	Check for small arguments before calling __expm1.
	* math/auto-libm-test-in: Do not mark underflow exceptions as
	possibly missing for bug 16353.
	* math/auto-libm-test-out: Regenerated.
2015-06-22 21:06:19 +00:00

132 lines
3.4 KiB
ArmAsm

/* ix87 specific implementation of exp(x)-1.
Copyright (C) 1996-2015 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Ulrich Drepper <drepper@cygnus.com>, 1996.
Based on code by John C. Bowman <bowman@ipp-garching.mpg.de>.
Corrections by H.J. Lu (hjl@gnu.ai.mit.edu), 1997.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
/* Using: e^x - 1 = 2^(x * log2(e)) - 1 */
#include <sysdep.h>
#include <machine/asm.h>
.section .rodata
.align ALIGNARG(4)
.type minus1,@object
minus1: .double -1.0
ASM_SIZE_DIRECTIVE(minus1)
.type one,@object
one: .double 1.0
ASM_SIZE_DIRECTIVE(one)
.type l2e,@object
l2e: .tfloat 1.442695040888963407359924681002
ASM_SIZE_DIRECTIVE(l2e)
.section .rodata.cst8,"aM",@progbits,8
.p2align 3
.type dbl_min,@object
dbl_min: .byte 0, 0, 0, 0, 0, 0, 0x10, 0
ASM_SIZE_DIRECTIVE(dbl_min)
#ifdef PIC
#define MO(op) op##@GOTOFF(%edx)
#else
#define MO(op) op
#endif
.text
ENTRY(__expm1)
movzwl 4+6(%esp), %eax
xorb $0x80, %ah // invert sign bit (now 1 is "positive")
cmpl $0xc086, %eax // is num >= 704?
jae HIDDEN_JUMPTARGET (__exp)
fldl 4(%esp) // x
fxam // Is NaN, +-Inf or +-0?
xorb $0x80, %ah
cmpl $0xc043, %eax // is num <= -38.0?
fstsw %ax
movb $0x45, %ch
jb 4f
// Below -38.0 (may be -NaN or -Inf).
andb %ah, %ch
#ifdef PIC
LOAD_PIC_REG (dx)
#endif
cmpb $0x01, %ch
je 5f // If -NaN, jump.
jmp 2f // -large, possibly -Inf.
4: // In range -38.0 to 704.0 (may be +-0 but not NaN or +-Inf).
andb %ah, %ch
cmpb $0x40, %ch
je 3f // If +-0, jump.
#ifdef PIC
LOAD_PIC_REG (dx)
#endif
fld %st
fabs
fcoml MO(dbl_min)
fstp %st
fnstsw
sahf
jae 5f
subl $8, %esp
cfi_adjust_cfa_offset (8)
fld %st(0)
fmul %st(0)
fstpl (%esp)
addl $8, %esp
cfi_adjust_cfa_offset (-8)
ret
5: fldt MO(l2e) // log2(e) : x
fmulp // log2(e)*x
fld %st // log2(e)*x : log2(e)*x
// Set round-to-nearest temporarily.
subl $8, %esp
cfi_adjust_cfa_offset (8)
fstcw 4(%esp)
movl $0xf3ff, %ecx
andl 4(%esp), %ecx
movl %ecx, (%esp)
fldcw (%esp)
frndint // int(log2(e)*x) : log2(e)*x
fldcw 4(%esp)
addl $8, %esp
cfi_adjust_cfa_offset (-8)
fsubr %st, %st(1) // int(log2(e)*x) : fract(log2(e)*x)
fxch // fract(log2(e)*x) : int(log2(e)*x)
f2xm1 // 2^fract(log2(e)*x)-1 : int(log2(e)*x)
fscale // 2^(log2(e)*x)-2^int(log2(e)*x) : int(log2(e)*x)
fxch // int(log2(e)*x) : 2^(log2(e)*x)-2^int(log2(e)*x)
fldl MO(one) // 1 : int(log2(e)*x) : 2^(log2(e)*x)-2^int(log2(e)*x)
fscale // 2^int(log2(e)*x) : int(log2(e)*x) : 2^(log2(e)*x)-2^int(log2(e)*x)
fsubrl MO(one) // 1-2^int(log2(e)*x) : int(log2(e)*x) : 2^(log2(e)*x)-2^int(log2(e)*x)
fstp %st(1) // 1-2^int(log2(e)*x) : 2^(log2(e)*x)-2^int(log2(e)*x)
fsubrp %st, %st(1) // 2^(log2(e)*x)
ret
2: fstp %st
fldl MO(minus1) // Set result to -1.0.
3: ret
END(__expm1)
weak_alias (__expm1, expm1)