| 1 | ; $Id: fma-asm.asm 96108 2022-08-08 11:16:23Z vboxsync $
|
|---|
| 2 | ;; @file
|
|---|
| 3 | ; IPRT - No-CRT fma alternatives - AMD64 & X86.
|
|---|
| 4 | ;
|
|---|
| 5 |
|
|---|
| 6 | ;
|
|---|
| 7 | ; Copyright (C) 2006-2022 Oracle Corporation
|
|---|
| 8 | ;
|
|---|
| 9 | ; This file is part of VirtualBox Open Source Edition (OSE), as
|
|---|
| 10 | ; available from http://www.virtualbox.org. This file is free software;
|
|---|
| 11 | ; you can redistribute it and/or modify it under the terms of the GNU
|
|---|
| 12 | ; General Public License (GPL) as published by the Free Software
|
|---|
| 13 | ; Foundation, in version 2 as it comes in the "COPYING" file of the
|
|---|
| 14 | ; VirtualBox OSE distribution. VirtualBox OSE is distributed in the
|
|---|
| 15 | ; hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
|
|---|
| 16 | ;
|
|---|
| 17 | ; The contents of this file may alternatively be used under the terms
|
|---|
| 18 | ; of the Common Development and Distribution License Version 1.0
|
|---|
| 19 | ; (CDDL) only, as it comes in the "COPYING.CDDL" file of the
|
|---|
| 20 | ; VirtualBox OSE distribution, in which case the provisions of the
|
|---|
| 21 | ; CDDL are applicable instead of those of the GPL.
|
|---|
| 22 | ;
|
|---|
| 23 | ; You may elect to license modified versions of this file under the
|
|---|
| 24 | ; terms and conditions of either the GPL or the CDDL or both.
|
|---|
| 25 | ;
|
|---|
| 26 |
|
|---|
| 27 | %define RT_ASM_WITH_SEH64
|
|---|
| 28 | %include "iprt/asmdefs.mac"
|
|---|
| 29 |
|
|---|
| 30 | BEGINCODE
|
|---|
| 31 |
|
|---|
| 32 | ;;
|
|---|
| 33 | ; Fused multiplication and add, intel version.
|
|---|
| 34 | ;
|
|---|
| 35 | ; @returns st(0) / xmm0
|
|---|
| 36 | ; @param rdFactor1 [rbp + 08h] / xmm0
|
|---|
| 37 | ; @param rdFactor2 [rbp + 10h] / xmm1
|
|---|
| 38 | ; @param rdAddend [rbp + 18h] / xmm2
|
|---|
| 39 | BEGINPROC rtNoCrtMathFma3
|
|---|
| 40 | push xBP
|
|---|
| 41 | SEH64_PUSH_xBP
|
|---|
| 42 | mov xBP, xSP
|
|---|
| 43 | SEH64_SET_FRAME_xBP 0
|
|---|
| 44 | SEH64_END_PROLOGUE
|
|---|
| 45 |
|
|---|
| 46 | %ifdef RT_ARCH_X86
|
|---|
| 47 | movsd xmm0, qword [xBP + xCB*2 + 00h]
|
|---|
| 48 | movsd xmm1, qword [xBP + xCB*2 + 08h]
|
|---|
| 49 | movsd xmm2, qword [xBP + xCB*2 + 10h]
|
|---|
| 50 | %endif
|
|---|
| 51 |
|
|---|
| 52 | vfmadd132sd xmm0, xmm2, xmm1 ; xmm0 = xmm0 * xmm1 + xmm2 (132 = multiply op1 with op3 and add op2)
|
|---|
| 53 |
|
|---|
| 54 | %ifdef RT_ARCH_X86
|
|---|
| 55 | sub xSP, 10h
|
|---|
| 56 | movsd [xSP], xmm0
|
|---|
| 57 | fld qword [xSP]
|
|---|
| 58 | %endif
|
|---|
| 59 | leave
|
|---|
| 60 | ret
|
|---|
| 61 | ENDPROC rtNoCrtMathFma3
|
|---|
| 62 |
|
|---|
| 63 |
|
|---|
| 64 | ;;
|
|---|
| 65 | ; Fused multiplication and add, amd version.
|
|---|
| 66 | ;
|
|---|
| 67 | ; @returns st(0) / xmm0
|
|---|
| 68 | ; @param rdFactor1 [rbp + 08h] / xmm0
|
|---|
| 69 | ; @param rdFactor2 [rbp + 10h] / xmm1
|
|---|
| 70 | ; @param rdAddend [rbp + 18h] / xmm2
|
|---|
| 71 | BEGINPROC rtNoCrtMathFma4
|
|---|
| 72 | push xBP
|
|---|
| 73 | SEH64_PUSH_xBP
|
|---|
| 74 | mov xBP, xSP
|
|---|
| 75 | SEH64_SET_FRAME_xBP 0
|
|---|
| 76 | SEH64_END_PROLOGUE
|
|---|
| 77 |
|
|---|
| 78 | %ifdef RT_ARCH_X86
|
|---|
| 79 | movsd xmm0, qword [xBP + xCB*2 + 00h]
|
|---|
| 80 | movsd xmm1, qword [xBP + xCB*2 + 08h]
|
|---|
| 81 | movsd xmm2, qword [xBP + xCB*2 + 10h]
|
|---|
| 82 | %endif
|
|---|
| 83 |
|
|---|
| 84 | vfmaddsd xmm0, xmm0, xmm1, xmm2 ; xmm0 = xmm0 * xmm1 + xmm2
|
|---|
| 85 |
|
|---|
| 86 | %ifdef RT_ARCH_X86
|
|---|
| 87 | sub xSP, 10h
|
|---|
| 88 | movsd [xSP], xmm0
|
|---|
| 89 | fld qword [xSP]
|
|---|
| 90 | %endif
|
|---|
| 91 | leave
|
|---|
| 92 | ret
|
|---|
| 93 | ENDPROC rtNoCrtMathFma4
|
|---|
| 94 |
|
|---|
