__udivdi3.S revision 256281
1.file "__udivdi3.s" 2 3// 4// Copyright (c) 2000, Intel Corporation 5// All rights reserved. 6// 7// Contributed 2/15/2000 by Marius Cornea, John Harrison, Cristina Iordache, 8// Ted Kubaska, Bob Norin, and Shane Story of the Computational Software Lab, 9// Intel Corporation. 10// 11// WARRANTY DISCLAIMER 12// 13// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 14// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 15// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 16// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS 17// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 18// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 19// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR 20// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY 21// OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING 22// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 23// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 24// 25// Intel Corporation is the author of this code, and requests that all 26// problem reports or change requests be submitted to it directly at 27// http://developer.intel.com/opensource. 28// 29 30#include <machine/asm.h> 31__FBSDID("$FreeBSD: stable/10/lib/libc/ia64/gen/__udivdi3.S 111777 2003-03-03 01:09:46Z obrien $"); 32 33.section .text 34.proc __udivdi3# 35.align 32 36.global __udivdi3# 37.align 32 38 39// 64-bit unsigned integer divide 40 41__udivdi3: 42 43{ .mii 44 alloc r31=ar.pfs,2,0,0,0 45 nop.i 0 46 nop.i 0;; 47} 48 49{ .mmi 50 51 // 64-BIT UNSIGNED INTEGER DIVIDE BEGINS HERE 52 53 setf.sig f8=r32 54 setf.sig f9=r33 55 nop.i 0;; 56} { .mfb 57 nop.m 0 58 fma.s1 f6=f8,f1,f0 59 nop.b 0 60} { .mfb 61 nop.m 0 62 fma.s1 f7=f9,f1,f0 63 nop.b 0;; 64} { .mfi 65 nop.m 0 66 // Step (1) 67 // y0 = 1 / b in f8 68 frcpa.s1 f8,p6=f6,f7 69 nop.i 0;; 70} { .mfi 71 nop.m 0 72 // Step (2) 73 // e0 = 1 - b * y0 in f9 74 (p6) fnma.s1 f9=f7,f8,f1 75 nop.i 0 76} { .mfi 77 nop.m 0 78 // Step (3) 79 // q0 = a * y0 in f10 80 (p6) fma.s1 f10=f6,f8,f0 81 nop.i 0;; 82} { .mfi 83 nop.m 0 84 // Step (4) 85 // e1 = e0 * e0 in f11 86 (p6) fma.s1 f11=f9,f9,f0 87 nop.i 0 88} { .mfi 89 nop.m 0 90 // Step (5) 91 // q1 = q0 + e0 * q0 in f10 92 (p6) fma.s1 f10=f9,f10,f10 93 nop.i 0;; 94} { .mfi 95 nop.m 0 96 // Step (6) 97 // y1 = y0 + e0 * y0 in f8 98 (p6) fma.s1 f8=f9,f8,f8 99 nop.i 0;; 100} { .mfi 101 nop.m 0 102 // Step (7) 103 // q2 = q1 + e1 * q1 in f9 104 (p6) fma.s1 f9=f11,f10,f10 105 nop.i 0;; 106} { .mfi 107 nop.m 0 108 // Step (8) 109 // y2 = y1 + e1 * y1 in f8 110 (p6) fma.s1 f8=f11,f8,f8 111 nop.i 0;; 112} { .mfi 113 nop.m 0 114 // Step (9) 115 // r2 = a - b * q2 in f10 116 (p6) fnma.s1 f10=f7,f9,f6 117 nop.i 0;; 118} { .mfi 119 nop.m 0 120 // Step (10) 121 // q3 = q2 + r2 * y2 in f8 122 (p6) fma.s1 f8=f10,f8,f9 123 nop.i 0;; 124} { .mfb 125 nop.m 0 126 // (11) q = trunc(q3) 127 fcvt.fxu.trunc.s1 f8=f8 128 nop.b 0;; 129} { .mmi 130 // quotient will be in r8 (if b != 0) 131 getf.sig r8=f8 132 nop.m 0 133 nop.i 0;; 134} 135 136 // 64-BIT UNSIGNED INTEGER DIVIDE ENDS HERE 137 138{ .mmb 139 nop.m 0 140 nop.m 0 141 br.ret.sptk b0;; 142} 143 144.endp __udivdi3 145