beta-0.89.2
[luatex.git] / source / libs / gmp / gmp-src / mpn / x86 / pentium4 / sse2 / dive_1.asm
blob0ceef5bc8a2ff8e217cb72e14f467be197afae23
1 dnl Intel Pentium-4 mpn_divexact_1 -- mpn by limb exact division.
3 dnl Copyright 2001, 2002, 2007 Free Software Foundation, Inc.
5 dnl This file is part of the GNU MP Library.
6 dnl
7 dnl The GNU MP Library is free software; you can redistribute it and/or modify
8 dnl it under the terms of either:
9 dnl
10 dnl * the GNU Lesser General Public License as published by the Free
11 dnl Software Foundation; either version 3 of the License, or (at your
12 dnl option) any later version.
13 dnl
14 dnl or
15 dnl
16 dnl * the GNU General Public License as published by the Free Software
17 dnl Foundation; either version 2 of the License, or (at your option) any
18 dnl later version.
19 dnl
20 dnl or both in parallel, as here.
21 dnl
22 dnl The GNU MP Library is distributed in the hope that it will be useful, but
23 dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
24 dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
25 dnl for more details.
26 dnl
27 dnl You should have received copies of the GNU General Public License and the
28 dnl GNU Lesser General Public License along with the GNU MP Library. If not,
29 dnl see https://www.gnu.org/licenses/.
31 include(`../config.m4')
34 C P4: 19.0 cycles/limb
37 C void mpn_divexact_1 (mp_ptr dst, mp_srcptr src, mp_size_t size,
38 C mp_limb_t divisor);
40 C Pairs of movd's are used to avoid unaligned loads. Despite the loads not
41 C being on the dependent chain and there being plenty of cycles available,
42 C using an unaligned movq on every second iteration measured about 23 c/l.
44 C Using divl for size==1 seems a touch quicker than mul-by-inverse. The mul
45 C will be about 9+2*4+2*2+10*4+19+12 = 92 cycles latency, though some of
46 C that might be hidden by out-of-order execution, whereas divl is around 60.
47 C At size==2 an extra 19 for the mul versus 60 for the divl will see the mul
48 C faster.
50 defframe(PARAM_DIVISOR,16)
51 defframe(PARAM_SIZE, 12)
52 defframe(PARAM_SRC, 8)
53 defframe(PARAM_DST, 4)
55 TEXT
57 ALIGN(16)
58 PROLOGUE(mpn_divexact_1)
59 deflit(`FRAME',0)
61 movl PARAM_SIZE, %edx
63 movl PARAM_SRC, %eax
65 movl PARAM_DIVISOR, %ecx
66 subl $1, %edx
67 jnz L(two_or_more)
69 movl (%eax), %eax
70 xorl %edx, %edx
72 divl %ecx
73 movl PARAM_DST, %ecx
75 movl %eax, (%ecx)
76 ret
79 L(two_or_more):
80 C eax src
81 C ebx
82 C ecx divisor
83 C edx size-1
85 movl %ecx, %eax
86 bsfl %ecx, %ecx C trailing twos
88 shrl %cl, %eax C d = divisor without twos
89 movd %eax, %mm6
90 movd %ecx, %mm7 C shift
92 shrl %eax C d/2
94 andl $127, %eax C d/2, 7 bits
96 ifdef(`PIC',`
97 LEA( binvert_limb_table, %ecx)
98 movzbl (%eax,%ecx), %eax C inv 8 bits
99 ',`
100 movzbl binvert_limb_table(%eax), %eax C inv 8 bits
105 movd %eax, %mm5 C inv
107 movd %eax, %mm0 C inv
109 pmuludq %mm5, %mm5 C inv*inv
113 pmuludq %mm6, %mm5 C inv*inv*d
114 paddd %mm0, %mm0 C 2*inv
118 psubd %mm5, %mm0 C inv = 2*inv - inv*inv*d
119 pxor %mm5, %mm5
121 paddd %mm0, %mm5
122 pmuludq %mm0, %mm0 C inv*inv
124 pcmpeqd %mm4, %mm4
125 psrlq $32, %mm4 C 0x00000000FFFFFFFF
129 pmuludq %mm6, %mm0 C inv*inv*d
130 paddd %mm5, %mm5 C 2*inv
132 movl PARAM_SRC, %eax
133 movl PARAM_DST, %ecx
134 pxor %mm1, %mm1 C initial carry limb
138 psubd %mm0, %mm5 C inv = 2*inv - inv*inv*d
140 ASSERT(e,` C expect d*inv == 1 mod 2^GMP_LIMB_BITS
141 pushl %eax FRAME_pushl()
142 movq %mm6, %mm0
143 pmuludq %mm5, %mm0
144 movd %mm0, %eax
145 cmpl $1, %eax
146 popl %eax FRAME_popl()')
148 pxor %mm0, %mm0 C initial carry bit
151 C The dependent chain here is as follows.
153 C latency
154 C psubq s = (src-cbit) - climb 2
155 C pmuludq q = s*inverse 8
156 C pmuludq prod = q*divisor 8
157 C psrlq climb = high(prod) 2
158 C --
159 C 20
161 C Yet the loop measures 19.0 c/l, so obviously there's something gained
162 C there over a straight reading of the chip documentation.
164 L(top):
165 C eax src, incrementing
166 C ebx
167 C ecx dst, incrementing
168 C edx counter, size-1 iterations
170 C mm0 carry bit
171 C mm1 carry limb
172 C mm4 0x00000000FFFFFFFF
173 C mm5 inverse
174 C mm6 divisor
175 C mm7 shift
177 movd (%eax), %mm2
178 movd 4(%eax), %mm3
179 addl $4, %eax
180 punpckldq %mm3, %mm2
182 psrlq %mm7, %mm2
183 pand %mm4, %mm2 C src
184 psubq %mm0, %mm2 C src - cbit
186 psubq %mm1, %mm2 C src - cbit - climb
187 movq %mm2, %mm0
188 psrlq $63, %mm0 C new cbit
190 pmuludq %mm5, %mm2 C s*inverse
191 movd %mm2, (%ecx) C q
192 addl $4, %ecx
194 movq %mm6, %mm1
195 pmuludq %mm2, %mm1 C q*divisor
196 psrlq $32, %mm1 C new climb
198 subl $1, %edx
199 jnz L(top)
202 L(done):
203 movd (%eax), %mm2
204 psrlq %mm7, %mm2 C src
205 psubq %mm0, %mm2 C src - cbit
207 psubq %mm1, %mm2 C src - cbit - climb
209 pmuludq %mm5, %mm2 C s*inverse
210 movd %mm2, (%ecx) C q
212 emms
215 EPILOGUE()
216 ASM_END()