1dnl  SPARC mpn_add_n -- Add two limb vectors of the same length > 0 and store
2dnl  sum in a third limb vector.
3
4dnl  Copyright 2001 Free Software Foundation, Inc.
5
6dnl  This file is part of the GNU MP Library.
7dnl
8dnl  The GNU MP Library is free software; you can redistribute it and/or modify
9dnl  it under the terms of either:
10dnl
11dnl    * the GNU Lesser General Public License as published by the Free
12dnl      Software Foundation; either version 3 of the License, or (at your
13dnl      option) any later version.
14dnl
15dnl  or
16dnl
17dnl    * the GNU General Public License as published by the Free Software
18dnl      Foundation; either version 2 of the License, or (at your option) any
19dnl      later version.
20dnl
21dnl  or both in parallel, as here.
22dnl
23dnl  The GNU MP Library is distributed in the hope that it will be useful, but
24dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
25dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
26dnl  for more details.
27dnl
28dnl  You should have received copies of the GNU General Public License and the
29dnl  GNU Lesser General Public License along with the GNU MP Library.  If not,
30dnl  see https://www.gnu.org/licenses/.
31
32
33include(`../config.m4')
34
35C INPUT PARAMETERS
36define(rp,%o0)
37define(s1p,%o1)
38define(s2p,%o2)
39define(n,%o3)
40define(cy,%g1)
41
42C This code uses 64-bit operations on `o' and `g' registers.  It doesn't
43C require that `o' registers' upper 32 bits are preserved by the operating
44C system, but if they are not, they must be zeroed.  That is indeed what
45C happens at least on Slowaris 2.5 and 2.6.
46
47C On UltraSPARC 1 and 2, this code runs at 3 cycles/limb from the Dcache and at
48C about 10 cycles/limb from the Ecache.
49
50ASM_START()
51PROLOGUE(mpn_add_n)
52          lduw      [s1p+0],%o4
53          lduw      [s2p+0],%o5
54          addcc     n,-2,n
55          bl,pn     %icc,L(end1)
56          lduw      [s1p+4],%g2
57          lduw      [s2p+4],%g3
58          be,pn     %icc,L(end2)
59          mov       0,cy
60
61          .align    16
62L(loop):
63          add       %o4,%o5,%g4
64          add       rp,8,rp
65          lduw      [s1p+8],%o4
66          fitod     %f0,%f2
67C ---
68          add       cy,%g4,%g4
69          addcc     n,-1,n
70          lduw      [s2p+8],%o5
71          fitod     %f0,%f2
72C ---
73          srlx      %g4,32,cy
74          add       s2p,8,s2p
75          stw       %g4,[rp-8]
76          be,pn     %icc,L(exito)+4
77C ---
78          add       %g2,%g3,%g4
79          addcc     n,-1,n
80          lduw      [s1p+12],%g2
81          fitod     %f0,%f2
82C ---
83          add       cy,%g4,%g4
84          add       s1p,8,s1p
85          lduw      [s2p+4],%g3
86          fitod     %f0,%f2
87C ---
88          srlx      %g4,32,cy
89          bne,pt    %icc,L(loop)
90          stw       %g4,[rp-4]
91C ---
92L(exite):
93          add       %o4,%o5,%g4
94          add       cy,%g4,%g4
95          srlx      %g4,32,cy
96          stw       %g4,[rp+0]
97          add       %g2,%g3,%g4
98          add       cy,%g4,%g4
99          stw       %g4,[rp+4]
100          retl
101          srlx      %g4,32,%o0
102
103L(exito):
104          add       %g2,%g3,%g4
105          add       cy,%g4,%g4
106          srlx      %g4,32,cy
107          stw       %g4,[rp-4]
108          add       %o4,%o5,%g4
109          add       cy,%g4,%g4
110          stw       %g4,[rp+0]
111          retl
112          srlx      %g4,32,%o0
113
114L(end1):
115          add       %o4,%o5,%g4
116          stw       %g4,[rp+0]
117          retl
118          srlx      %g4,32,%o0
119
120L(end2):
121          add       %o4,%o5,%g4
122          srlx      %g4,32,cy
123          stw       %g4,[rp+0]
124          add       %g2,%g3,%g4
125          add       cy,%g4,%g4
126          stw       %g4,[rp+4]
127          retl
128          srlx      %g4,32,%o0
129EPILOGUE(mpn_add_n)
130