![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to addmul_1.asm CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [local] / OpenXM_contrib / gmp / mpn / ia64|
Return to addmul_1.asm CVS log | Up to [local] / OpenXM_contrib / gmp / mpn / ia64 |
1.1 ohara 1: dnl IA-64 mpn_addmul_1 -- Multiply a limb vector with a limb and add the
2: dnl result to a second limb vector.
3:
4: dnl Copyright 2000, 2001, 2002 Free Software Foundation, Inc.
5:
6: dnl This file is part of the GNU MP Library.
7:
8: dnl The GNU MP Library is free software; you can redistribute it and/or modify
9: dnl it under the terms of the GNU Lesser General Public License as published
10: dnl by the Free Software Foundation; either version 2.1 of the License, or (at
11: dnl your option) any later version.
12:
13: dnl The GNU MP Library is distributed in the hope that it will be useful, but
14: dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
15: dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
16: dnl License for more details.
17:
18: dnl You should have received a copy of the GNU Lesser General Public License
19: dnl along with the GNU MP Library; see the file COPYING.LIB. If not, write to
20: dnl the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
21: dnl MA 02111-1307, USA.
22:
23: include(`../config.m4')
24:
25: C INPUT PARAMETERS
26: C rp = r32
27: C up = r33
28: C n = r34
29: C v = r35
30:
31: C cycles/limb
32: C Itanium: 4
33: C Itanium 2: ?
34:
35: C The full speed is reached C only for really huge operands. See README for
36: C possible improvements.
37:
38:
39: ASM_START()
40: PROLOGUE(mpn_addmul_1)
41: .prologue
42: .save ar.pfs, r21
43: alloc r21 = ar.pfs, 4, 12, 0, 16
44: .save ar.lc, r2
45: mov r2 = ar.lc
46: mov r20 = ar.ec
47: .save pr, r22
48: mov r22 = pr
49: .body
50: ifdef(`HAVE_ABI_32',
51: ` addp4 r32 = 0, r32
52: addp4 r33 = 0, r33
53: sxt4 r34 = r34
54: ;;
55: ')
56: { .mfi; setf.sig f6 = r35
57: nop.f 0
58: adds r19 = -1, r34 C n - 1
59: } { .mfi; cmp.ne p6, p7 = r0, r0
60: nop.f 0
61: mov r18 = r32 ;;
62: } { .mfi; mov r16 = r32
63: nop.f 0
64: mov ar.lc = r19
65: } { .mfi; mov r17 = r33
66: nop.f 0
67: mov ar.ec = 7
68: } { .mii; cmp.ne p6, p7 = r0, r0
69: mov pr.rot = 1<<16
70: mov r32 = 0 C clear "carry in"
71: } { .mib; mov r33 = 0 C clear for cmp
72: mov r34 = 0 C clear for cmp
73: nop.b 0
74: } { .mib; mov r35 = 0 C clear for cmp
75: mov r36 = 0 C clear for cmp
76: nop.b 0 ;;
77: }
78: .align 32
79: .Loop:
80: .pred.rel "mutex",p6,p7
81: { .mfi; (p16) ldf8 f32 = [r17], 8 C *0,3,6,9,12,15,18
82: (p19) xma.l f40 = f35, f6, f39 C 0,3,6,*9,12,15,18
83: (p6) add r14 = r33, r38, 1 C 0,3,6,9,12,15,*18
84: } { .mfi; (p16) ldf8 f36 = [r16], 8 C *0,3,6,9,12,15,18
85: (p19) xma.hu f44 = f35, f6, f39 C 0,3,6,*9,12,15,18
86: (p7) add r14 = r33, r38 ;; C 0,3,6,9,12,15,*18
87: } { .mii; (p21) getf.sig r32 = f42 C 1,4,7,10,13,*16,19
88: (p6) cmp.leu p8, p9 = r14, r33 C 1,4,7,10,13,16,*19
89: (p7) cmp.ltu p8, p9 = r14, r33;; C 1,4,7,10,13,16,*19
90: }
91: .pred.rel "mutex",p8,p9
92: { .mib; (p21) getf.sig r36 = f46 C 2,5,8,11,14,*17,20
93: (p8) cmp.eq p6, p7 = r0, r0
94: nop.b 0
95: } { .mib; (p22) st8 [r18] = r14, 8 C 2,5,8,11,14,17,*20
96: (p9) cmp.ne p6, p7 = r0, r0
97: br.ctop.sptk .Loop ;;
98: }
99: .pred.rel "mutex",p6,p7
100: (p6) add r8 = 1, r38
101: (p7) mov r8 = r38
102: mov pr = r22,0x1fffe
103: mov ar.lc = r2
104: mov ar.ec = r20
105: mov ar.pfs = r21;;
106: br.ret.sptk.many b0
107: EPILOGUE(mpn_addmul_1)
108: ASM_END()