1.file "__udivdi3.s"
2
3// $NetBSD$
4
5//-
6// Copyright (c) 2000, Intel Corporation
7// All rights reserved.
8//
9// Contributed 2/15/2000 by Marius Cornea, John Harrison, Cristina Iordache,
10// Ted Kubaska, Bob Norin, and Shane Story of the Computational Software Lab,
11// Intel Corporation.
12//
13// WARRANTY DISCLAIMER
14//
15// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
16// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
17// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
18// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS
19// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
20// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
21// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
22// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
23// OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING
24// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
25// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26//
27// Intel Corporation is the author of this code, and requests that all
28// problem reports or change requests be submitted to it directly at
29// http://developer.intel.com/opensource.
30//
31
32.section .text
33.proc __udivdi3#
34.align 32
35.global __udivdi3#
36.align 32
37
38// 64-bit unsigned integer divide
39
40__udivdi3:
41
42{ .mii
43  alloc r31=ar.pfs,2,0,0,0
44  nop.i 0
45  nop.i 0;;
46}
47
48{ .mmi
49
50  // 64-BIT UNSIGNED INTEGER DIVIDE BEGINS HERE
51
52  setf.sig f8=r32
53  setf.sig f9=r33
54  nop.i 0;;
55} { .mfb
56  nop.m 0
57  fma.s1 f6=f8,f1,f0
58  nop.b 0
59} { .mfb
60  nop.m 0
61  fma.s1 f7=f9,f1,f0
62  nop.b 0;;
63} { .mfi
64  nop.m 0
65  // Step (1)
66  // y0 = 1 / b in f8
67  frcpa.s1 f8,p6=f6,f7
68  nop.i 0;;
69} { .mfi
70  nop.m 0
71  // Step (2)
72  // e0 = 1 - b * y0 in f9
73  (p6) fnma.s1 f9=f7,f8,f1
74  nop.i 0
75} { .mfi
76  nop.m 0
77  // Step (3)
78  // q0 = a * y0 in f10
79  (p6) fma.s1 f10=f6,f8,f0
80  nop.i 0;;
81} { .mfi
82  nop.m 0
83  // Step (4)
84  // e1 = e0 * e0 in f11
85  (p6) fma.s1 f11=f9,f9,f0
86  nop.i 0
87} { .mfi
88  nop.m 0
89  // Step (5)
90  // q1 = q0 + e0 * q0 in f10
91  (p6) fma.s1 f10=f9,f10,f10
92  nop.i 0;;
93} { .mfi
94  nop.m 0
95  // Step (6)
96  // y1 = y0 + e0 * y0 in f8
97  (p6) fma.s1 f8=f9,f8,f8
98  nop.i 0;;
99} { .mfi
100  nop.m 0
101  // Step (7)
102  // q2 = q1 + e1 * q1 in f9
103  (p6) fma.s1 f9=f11,f10,f10
104  nop.i 0;;
105} { .mfi
106  nop.m 0
107  // Step (8)
108  // y2 = y1 + e1 * y1 in f8
109  (p6) fma.s1 f8=f11,f8,f8
110  nop.i 0;;
111} { .mfi
112  nop.m 0
113  // Step (9)
114  // r2 = a - b * q2 in f10
115  (p6) fnma.s1 f10=f7,f9,f6
116  nop.i 0;;
117} { .mfi
118  nop.m 0
119  // Step (10)
120  // q3 = q2 + r2 * y2 in f8
121  (p6) fma.s1 f8=f10,f8,f9
122  nop.i 0;;
123} { .mfb
124  nop.m 0
125  // (11) q = trunc(q3)
126  fcvt.fxu.trunc.s1 f8=f8
127  nop.b 0;;
128} { .mmi
129  // quotient will be in r8 (if b != 0)
130  getf.sig r8=f8
131  nop.m 0
132  nop.i 0;;
133}
134
135  // 64-BIT UNSIGNED INTEGER DIVIDE ENDS HERE
136
137{ .mmb
138  nop.m 0
139  nop.m 0
140  br.ret.sptk b0;;
141}
142
143.endp __udivdi3
144