^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) /* SPDX-License-Identifier: GPL-2.0 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (C) 1999-2000 Hewlett-Packard Co
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) * Copyright (C) 1999-2000 David Mosberger-Tang <davidm@hpl.hp.com>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * 64-bit integer division.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) * This code is based on the application note entitled "Divide, Square Root
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) * and Remainder Algorithms for the IA-64 Architecture". This document
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * is available as Intel document number 248725-002 or via the web at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) * http://developer.intel.com/software/opensource/numerics/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) * For more details on the theory behind these algorithms, see "IA-64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * and Elementary Functions" by Peter Markstein; HP Professional Books
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * (http://www.goodreads.com/book/show/2019887.Ia_64_and_Elementary_Functions)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include <asm/asmmacro.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) #include <asm/export.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) #ifdef MODULO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) # define OP mod
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) # define OP div
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) #ifdef UNSIGNED
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) # define SGN u
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) # define INT_TO_FP(a,b) fcvt.xuf.s1 a=b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) # define FP_TO_INT(a,b) fcvt.fxu.trunc.s1 a=b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) # define SGN
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) # define INT_TO_FP(a,b) fcvt.xf a=b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) # define FP_TO_INT(a,b) fcvt.fx.trunc.s1 a=b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) #define PASTE1(a,b) a##b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) #define PASTE(a,b) PASTE1(a,b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) #define NAME PASTE(PASTE(__,SGN),PASTE(OP,di3))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) GLOBAL_ENTRY(NAME)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) .regstk 2,0,0,0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) // Transfer inputs to FP registers.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) setf.sig f8 = in0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) setf.sig f9 = in1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) // Convert the inputs to FP, to avoid FP software-assist faults.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) INT_TO_FP(f8, f8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) INT_TO_FP(f9, f9)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) frcpa.s1 f11, p6 = f8, f9 // y0 = frcpa(b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) (p6) fmpy.s1 f7 = f8, f11 // q0 = a*y0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) (p6) fnma.s1 f6 = f9, f11, f1 // e0 = -b*y0 + 1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) (p6) fma.s1 f10 = f7, f6, f7 // q1 = q0*e0 + q0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) (p6) fmpy.s1 f7 = f6, f6 // e1 = e0*e0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) #ifdef MODULO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) sub in1 = r0, in1 // in1 = -b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) (p6) fma.s1 f10 = f10, f7, f10 // q2 = q1*e1 + q1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) (p6) fma.s1 f6 = f11, f6, f11 // y1 = y0*e0 + y0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) (p6) fma.s1 f6 = f6, f7, f6 // y2 = y1*e1 + y1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) (p6) fnma.s1 f7 = f9, f10, f8 // r = -b*q2 + a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) #ifdef MODULO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) setf.sig f8 = in0 // f8 = a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) setf.sig f9 = in1 // f9 = -b
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) (p6) fma.s1 f11 = f7, f6, f10 // q3 = r*y2 + q2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) FP_TO_INT(f11, f11) // q = trunc(q3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) #ifdef MODULO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) xma.l f11 = f11, f9, f8 // r = q*(-b) + a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) ;;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) getf.sig r8 = f11 // transfer result to result register
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) br.ret.sptk.many rp
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) END(NAME)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) EXPORT_SYMBOL(NAME)
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags