[thirdparty/glibc.git] / sysdeps / sparc / sparc64 / mul_1.S

/* SPARC v9 __mpn_mul_1 -- Multiply a limb vector with a single limb and
   store the product in a second limb vector.

   Copyright (C) 1995, 1996 Free Software Foundation, Inc.

   This file is part of the GNU MP Library.

   The GNU MP Library is free software; you can redistribute it and/or modify
   it under the terms of the GNU Library General Public License as published by
   the Free Software Foundation; either version 2 of the License, or (at your
   option) any later version.

   The GNU MP 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 Library General Public
   License for more details.

   You should have received a copy of the GNU Library General Public License
   along with the GNU MP Library; see the file COPYING.LIB.  If not, write to
   the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
   MA 02111-1307, USA.  */

#include <sysdep.h>


/* INPUT PARAMETERS
   res_ptr	o0
   s1_ptr	o1
   size		o2
   s2_limb	o3  */

ENTRY(__mpn_mul_1)
	!#PROLOGUE#	0
	save	%sp,-128,%sp
	!#PROLOGUE#	1

	sub	%g0,%i2,%o7
	sllx	%o7,3,%g5
	sub	%i1,%g5,%o3
	sub	%i0,%g5,%o4
	mov	0,%o0			! zero cy_limb

	srl	%i3,0,%o1		! extract low 32 bits of s2_limb
	srlx	%i3,32,%i3		! extract high 32 bits of s2_limb
	mov	1,%o2
	sllx	%o2,32,%o2		! o2 = 0x100000000

	!   hi   !
             !  mid-1 !
             !  mid-2 !
		 !   lo   !
.Loop:
	sllx	%o7,3,%g1
	ldx	[%o3+%g1],%g5
	srl	%g5,0,%i0		! zero hi bits
	srlx	%g5,32,%g5
	mulx	%o1,%i0,%i4		! lo product
	mulx	%i3,%i0,%i1		! mid-1 product
	mulx	%o1,%g5,%l2		! mid-2 product
	mulx	%i3,%g5,%i5		! hi product
	srlx	%i4,32,%i0		! extract high 32 bits of lo product...
	add	%i1,%i0,%i1		! ...and add it to the mid-1 product
	addcc	%i1,%l2,%i1		! add mid products
	mov	0,%l0			! we need the carry from that add...
	movcs	%xcc,%o2,%l0		! ...compute it and...
	add	%i5,%l0,%i5		! ...add to bit 32 of the hi product
	sllx	%i1,32,%i0		! align low bits of mid product
	srl	%i4,0,%g5		! zero high 32 bits of lo product
	add	%i0,%g5,%i0		! combine into low 64 bits of result
	srlx	%i1,32,%i1		! extract high bits of mid product...
	add	%i5,%i1,%i1		! ...and add them to the high result
	addcc	%i0,%o0,%i0		! add cy_limb to low 64 bits of result
	mov	0,%g5
	movcs	%xcc,1,%g5
	add	%o7,1,%o7
	stx	%i0,[%o4+%g1]
	brnz	%o7,.Loop
	 add	%i1,%g5,%o0		! compute new cy_limb

	mov	%o0,%i0
	jmpl	%i7+8,%g0
	 restore

END(__mpn_mul_1)
Commit	Line	Data
5ae3e846 UD	1	/* SPARC v9 __mpn_mul_1 -- Multiply a limb vector with a single limb and
5ae3e846 UD	2	store the product in a second limb vector.
6b628d36	3
5ae3e846	4	Copyright (C) 1995, 1996 Free Software Foundation, Inc.
6b628d36	5
5ae3e846	6	This file is part of the GNU MP Library.
6b628d36	7
5ae3e846 UD	8	The GNU MP Library is free software; you can redistribute it and/or modify
	9	it under the terms of the GNU Library General Public License as published by
	10	the Free Software Foundation; either version 2 of the License, or (at your
	11	option) any later version.
6b628d36	12
5ae3e846 UD	13	The GNU MP Library is distributed in the hope that it will be useful, but
	14	WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
	15	or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Library General Public
	16	License for more details.
6b628d36	17
5ae3e846 UD	18	You should have received a copy of the GNU Library General Public License
	19	along with the GNU MP Library; see the file COPYING.LIB. If not, write to
	20	the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
	21	MA 02111-1307, USA. */
6b628d36	22
5ae3e846	23	#include <sysdep.h>
6b628d36	24
6b628d36	25
5ae3e846 UD	26	/* INPUT PARAMETERS
	27	res_ptr o0
	28	s1_ptr o1
	29	size o2
	30	s2_limb o3 */
	31
	32	ENTRY(__mpn_mul_1)
6b628d36	33	!#PROLOGUE# 0
5ae3e846	34	save %sp,-128,%sp
6b628d36	35	!#PROLOGUE# 1
5ae3e846	36
6b628d36 RM	37	sub %g0,%i2,%o7
	38	sllx %o7,3,%g5
	39	sub %i1,%g5,%o3
	40	sub %i0,%g5,%o4
	41	mov 0,%o0 ! zero cy_limb
	42
	43	srl %i3,0,%o1 ! extract low 32 bits of s2_limb
	44	srlx %i3,32,%i3 ! extract high 32 bits of s2_limb
	45	mov 1,%o2
	46	sllx %o2,32,%o2 ! o2 = 0x100000000
	47
	48	! hi !
	49	! mid-1 !
	50	! mid-2 !
	51	! lo !
	52	.Loop:
	53	sllx %o7,3,%g1
	54	ldx [%o3+%g1],%g5
	55	srl %g5,0,%i0 ! zero hi bits
	56	srlx %g5,32,%g5
	57	mulx %o1,%i0,%i4 ! lo product
	58	mulx %i3,%i0,%i1 ! mid-1 product
	59	mulx %o1,%g5,%l2 ! mid-2 product
	60	mulx %i3,%g5,%i5 ! hi product
	61	srlx %i4,32,%i0 ! extract high 32 bits of lo product...
	62	add %i1,%i0,%i1 ! ...and add it to the mid-1 product
	63	addcc %i1,%l2,%i1 ! add mid products
	64	mov 0,%l0 ! we need the carry from that add...
	65	movcs %xcc,%o2,%l0 ! ...compute it and...
	66	add %i5,%l0,%i5 ! ...add to bit 32 of the hi product
	67	sllx %i1,32,%i0 ! align low bits of mid product
	68	srl %i4,0,%g5 ! zero high 32 bits of lo product
	69	add %i0,%g5,%i0 ! combine into low 64 bits of result
	70	srlx %i1,32,%i1 ! extract high bits of mid product...
	71	add %i5,%i1,%i1 ! ...and add them to the high result
	72	addcc %i0,%o0,%i0 ! add cy_limb to low 64 bits of result
	73	mov 0,%g5
	74	movcs %xcc,1,%g5
	75	add %o7,1,%o7
	76	stx %i0,[%o4+%g1]
	77	brnz %o7,.Loop
5ae3e846	78	add %i1,%g5,%o0 ! compute new cy_limb
6b628d36 RM	79
6b628d36 RM	80	mov %o0,%i0
5ae3e846 UD	81	jmpl %i7+8,%g0
	82	restore
	83
	84	END(__mpn_mul_1)