beta-0.89.2
[luatex.git] / source / libs / gmp / gmp-src / mpn / sparc64 / ultrasparct3 / sub_n.asm
blob0e4bc939e30dbe95bf82efb9aa18279b6bd1650d
1 dnl SPARC v9 mpn_sub_n for T3/T4.
3 dnl Contributed to the GNU project by David Miller.
5 dnl Copyright 2013 Free Software Foundation, Inc.
7 dnl This file is part of the GNU MP Library.
8 dnl
9 dnl The GNU MP Library is free software; you can redistribute it and/or modify
10 dnl it under the terms of either:
11 dnl
12 dnl * the GNU Lesser General Public License as published by the Free
13 dnl Software Foundation; either version 3 of the License, or (at your
14 dnl option) any later version.
15 dnl
16 dnl or
17 dnl
18 dnl * the GNU General Public License as published by the Free Software
19 dnl Foundation; either version 2 of the License, or (at your option) any
20 dnl later version.
21 dnl
22 dnl or both in parallel, as here.
23 dnl
24 dnl The GNU MP Library is distributed in the hope that it will be useful, but
25 dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
26 dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
27 dnl for more details.
28 dnl
29 dnl You should have received copies of the GNU General Public License and the
30 dnl GNU Lesser General Public License along with the GNU MP Library. If not,
31 dnl see https://www.gnu.org/licenses/.
33 include(`../config.m4')
35 C cycles/limb
36 C UltraSPARC T3: 8
37 C UltraSPARC T4: 3
39 C INPUT PARAMETERS
40 define(`rp', `%i0')
41 define(`up', `%i1')
42 define(`vp', `%i2')
43 define(`n', `%i3')
44 define(`cy', `%i4')
46 define(`u0_off', `%l0')
47 define(`u1_off', `%l1')
48 define(`v0_off', `%l2')
49 define(`v1_off', `%l3')
50 define(`r0_off', `%l4')
51 define(`r1_off', `%l5')
52 define(`loop_n', `%l6')
53 define(`tmp', `%l7')
55 ASM_START()
56 REGISTER(%g2,#scratch)
57 REGISTER(%g3,#scratch)
58 PROLOGUE(mpn_sub_nc)
59 save %sp, -176, %sp
60 ba,pt %xcc, L(ent)
61 xor cy, 1, cy
62 EPILOGUE()
63 PROLOGUE(mpn_sub_n)
64 save %sp, -176, %sp
65 mov 1, cy
66 L(ent):
67 subcc n, 1, n
68 be L(final_one)
69 cmp %g0, cy
71 ldx [up + 0], %o4
72 sllx n, 3, tmp
74 ldx [vp + 0], %o5
75 add up, tmp, u0_off
77 ldx [up + 8], %g5
78 add vp, tmp, v0_off
80 ldx [vp + 8], %g1
81 add rp, tmp, r0_off
83 neg tmp, loop_n
84 add u0_off, 8, u1_off
86 add v0_off, 8, v1_off
87 sub loop_n, -(2 * 8), loop_n
89 sub r0_off, 16, r0_off
90 brgez,pn loop_n, L(loop_tail)
91 sub r0_off, 8, r1_off
93 b,a L(top)
94 ALIGN(16)
95 L(top):
96 xnor %o5, 0, tmp
97 ldx [loop_n + v0_off], %o5
99 addxccc(%o4, tmp, %g3)
100 ldx [loop_n + u0_off], %o4
102 xnor %g1, 0, %g1
103 stx %g3, [loop_n + r0_off]
105 addxccc(%g5, %g1, tmp)
106 ldx [loop_n + v1_off], %g1
108 ldx [loop_n + u1_off], %g5
109 sub loop_n, -(2 * 8), loop_n
111 brlz loop_n, L(top)
112 stx tmp, [loop_n + r1_off]
114 L(loop_tail):
115 xnor %o5, 0, tmp
116 xnor %g1, 0, %g1
118 addxccc(%o4, tmp, %g3)
119 add loop_n, u0_off, up
121 addxccc(%g5, %g1, %g5)
122 add loop_n, r0_off, rp
124 stx %g3, [rp + 0]
125 add loop_n, v0_off, vp
127 brgz,pt loop_n, L(done)
128 stx %g5, [rp + 8]
130 add rp, (2 * 8), rp
132 L(final_one):
133 ldx [up+0], %o4
134 ldx [vp+0], %o5
135 xnor %o5, %g0, %o5
136 addxccc(%o4, %o5, %g3)
137 stx %g3, [rp+0]
139 L(done):
140 clr %i0
141 movcc %xcc, 1, %i0
143 restore
144 EPILOGUE()