4 // Copyright (c) 2000, Intel Corporation
5 // All rights reserved.
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,
11 // WARRANTY DISCLAIMER
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.
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.
30 #include <machine/asm.h>
31 __FBSDID("$FreeBSD$");
35 // 32-bit signed integer remainder
45 alloc r31=ar.pfs,2,0,0,0
51 // 32-BIT SIGNED INTEGER REMAINDER BEGINS HERE
53 // general register used:
54 // r32 - 32-bit signed integer dividend
55 // r33 - 32-bit signed integer divisor
56 // r8 - 32-bit signed integer result
57 // r2 - scratch register
58 // floating-point registers used: f6, f7, f8, f9, f10, f11
59 // predicate registers used: p6
68 // get 2's complement of b
84 (p6) fma.s1 f10=f6,f8,f0
88 // (3) e0 = 1 - b * y0
89 (p6) fnma.s1 f8=f7,f8,f1
92 // 2's complement of b
94 // (4) q1 = q0 + e0 * q0
95 (p6) fma.s1 f10=f8,f10,f10
99 // (5) e1 = e0 * e0 + 2^-34
100 (p6) fma.s1 f8=f8,f8,f9
104 // (6) q2 = q1 + e1 * q1
105 (p6) fma.s1 f8=f8,f10,f10
110 fcvt.fx.trunc.s1 f8=f8
114 // (8) r = a + (-b) * q
118 // remainder will be in the least significant 32 bits of r8 (if b != 0)
124 // 32-BIT SIGNED INTEGER REMAINDER ENDS HERE