linux_dsm_epyc7002/arch/c6x/lib/divremu.S
Thomas Gleixner 1a59d1b8e0 treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 156
Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license as published by
  the free software foundation either version 2 of the license or at
  your option any later version this program is distributed in the
  hope that it will be useful but without any warranty without even
  the implied warranty of merchantability or fitness for a particular
  purpose see the gnu general public license for more details you
  should have received a copy of the gnu general public license along
  with this program if not write to the free software foundation inc
  59 temple place suite 330 boston ma 02111 1307 usa

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-or-later

has been chosen to replace the boilerplate/reference in 1334 file(s).

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Richard Fontana <rfontana@redhat.com>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070033.113240726@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-05-30 11:26:35 -07:00

76 lines
2.0 KiB
ArmAsm

;; SPDX-License-Identifier: GPL-2.0-or-later
;; Copyright 2011 Free Software Foundation, Inc.
;; Contributed by Bernd Schmidt <bernds@codesourcery.com>.
;;
#include <linux/linkage.h>
.text
ENTRY(__c6xabi_divremu)
;; We use a series of up to 31 subc instructions. First, we find
;; out how many leading zero bits there are in the divisor. This
;; gives us both a shift count for aligning (shifting) the divisor
;; to the, and the number of times we have to execute subc.
;; At the end, we have both the remainder and most of the quotient
;; in A4. The top bit of the quotient is computed first and is
;; placed in A2.
;; Return immediately if the dividend is zero. Setting B4 to 1
;; is a trick to allow us to leave the following insns in the jump
;; delay slot without affecting the result.
mv .s2x A4, B1
[b1] lmbd .l2 1, B4, B1
||[!b1] b .s2 B3 ; RETURN A
||[!b1] mvk .d2 1, B4
||[!b1] zero .s1 A5
mv .l1x B1, A6
|| shl .s2 B4, B1, B4
;; The loop performs a maximum of 28 steps, so we do the
;; first 3 here.
cmpltu .l1x A4, B4, A2
[!A2] sub .l1x A4, B4, A4
|| shru .s2 B4, 1, B4
|| xor .s1 1, A2, A2
shl .s1 A2, 31, A2
|| [b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
;; RETURN A may happen here (note: must happen before the next branch)
__divremu0:
cmpgt .l2 B1, 7, B0
|| [b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
|| [b0] b .s1 __divremu0
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
[b1] subc .l1x A4,B4,A4
|| [b1] add .s2 -1, B1, B1
;; loop backwards branch happens here
ret .s2 B3
|| mvk .s1 32, A1
sub .l1 A1, A6, A6
|| extu .s1 A4, A6, A5
shl .s1 A4, A6, A4
shru .s1 A4, 1, A4
|| sub .l1 A6, 1, A6
or .l1 A2, A4, A4
shru .s1 A4, A6, A4
nop
ENDPROC(__c6xabi_divremu)