]>
Commit | Line | Data |
---|---|---|
5ae3e846 UD |
1 | /* SPARC v9 __mpn_mul_1 -- Multiply a limb vector with a single limb and |
2 | store the product in a second limb vector. | |
6b628d36 | 3 | |
f7a9f785 | 4 | Copyright (C) 1995-2016 Free Software Foundation, Inc. |
6b628d36 | 5 | |
5ae3e846 | 6 | This file is part of the GNU MP Library. |
6b628d36 | 7 | |
5ae3e846 | 8 | The GNU MP Library is free software; you can redistribute it and/or modify |
6d84f89a AJ |
9 | it under the terms of the GNU Lesser General Public License as published by |
10 | the Free Software Foundation; either version 2.1 of the License, or (at your | |
5ae3e846 | 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 | |
6d84f89a | 15 | or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public |
5ae3e846 | 16 | License for more details. |
6b628d36 | 17 | |
6d84f89a | 18 | You should have received a copy of the GNU Lesser General Public License |
59ba27a6 PE |
19 | along with the GNU MP Library; see the file COPYING.LIB. If not, |
20 | see <http://www.gnu.org/licenses/>. */ | |
6b628d36 | 21 | |
5ae3e846 | 22 | #include <sysdep.h> |
6b628d36 | 23 | |
6b628d36 | 24 | |
5ae3e846 UD |
25 | /* INPUT PARAMETERS |
26 | res_ptr o0 | |
27 | s1_ptr o1 | |
28 | size o2 | |
29 | s2_limb o3 */ | |
30 | ||
31 | ENTRY(__mpn_mul_1) | |
6b628d36 | 32 | !#PROLOGUE# 0 |
9c4c0024 | 33 | save %sp,-192,%sp |
6b628d36 | 34 | !#PROLOGUE# 1 |
5ae3e846 | 35 | |
6b628d36 RM |
36 | sub %g0,%i2,%o7 |
37 | sllx %o7,3,%g5 | |
38 | sub %i1,%g5,%o3 | |
39 | sub %i0,%g5,%o4 | |
40 | mov 0,%o0 ! zero cy_limb | |
41 | ||
42 | srl %i3,0,%o1 ! extract low 32 bits of s2_limb | |
43 | srlx %i3,32,%i3 ! extract high 32 bits of s2_limb | |
44 | mov 1,%o2 | |
45 | sllx %o2,32,%o2 ! o2 = 0x100000000 | |
46 | ||
47 | ! hi ! | |
48 | ! mid-1 ! | |
49 | ! mid-2 ! | |
50 | ! lo ! | |
51 | .Loop: | |
52 | sllx %o7,3,%g1 | |
53 | ldx [%o3+%g1],%g5 | |
54 | srl %g5,0,%i0 ! zero hi bits | |
55 | srlx %g5,32,%g5 | |
56 | mulx %o1,%i0,%i4 ! lo product | |
57 | mulx %i3,%i0,%i1 ! mid-1 product | |
58 | mulx %o1,%g5,%l2 ! mid-2 product | |
59 | mulx %i3,%g5,%i5 ! hi product | |
60 | srlx %i4,32,%i0 ! extract high 32 bits of lo product... | |
61 | add %i1,%i0,%i1 ! ...and add it to the mid-1 product | |
62 | addcc %i1,%l2,%i1 ! add mid products | |
63 | mov 0,%l0 ! we need the carry from that add... | |
64 | movcs %xcc,%o2,%l0 ! ...compute it and... | |
65 | add %i5,%l0,%i5 ! ...add to bit 32 of the hi product | |
66 | sllx %i1,32,%i0 ! align low bits of mid product | |
67 | srl %i4,0,%g5 ! zero high 32 bits of lo product | |
68 | add %i0,%g5,%i0 ! combine into low 64 bits of result | |
69 | srlx %i1,32,%i1 ! extract high bits of mid product... | |
70 | add %i5,%i1,%i1 ! ...and add them to the high result | |
71 | addcc %i0,%o0,%i0 ! add cy_limb to low 64 bits of result | |
72 | mov 0,%g5 | |
73 | movcs %xcc,1,%g5 | |
2e65ca2b | 74 | addcc %o7,1,%o7 |
6b628d36 | 75 | stx %i0,[%o4+%g1] |
2e65ca2b | 76 | bne,pt %xcc,.Loop |
5ae3e846 | 77 | add %i1,%g5,%o0 ! compute new cy_limb |
6b628d36 | 78 | |
5ae3e846 | 79 | jmpl %i7+8,%g0 |
2e65ca2b | 80 | restore %o0,%g0,%o0 |
5ae3e846 UD |
81 | |
82 | END(__mpn_mul_1) |