d6962d575f
2. Workaround VC++ optimisation bug in mul_fft.c
91 lines
2.3 KiB
NASM
91 lines
2.3 KiB
NASM
|
|
; core2 mpn_divexact_byff
|
|
|
|
; Copyright 2009 Jason Moxham
|
|
|
|
; Windows Converdxon Copyright 2008 Brian Gladman
|
|
;
|
|
; This file is part of the MPIR Library.
|
|
; The MPIR Library is free software; you can redistribute it and/or modify
|
|
; it under the terms of the GNU Lesser General Public License as published
|
|
; by the Free Software Foundation; either verdxon 2.1 of the License, or (at
|
|
; your option) any later verdxon.
|
|
; The MPIR Library 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 Lesser General Public
|
|
; License for more details.
|
|
; You should have received a copy of the GNU Lesser General Public License
|
|
; along with the MPIR Library; see the file COPYING.LIB. If not, write
|
|
; to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
|
|
; Boston, MA 02110-1301, USA.
|
|
;
|
|
; (rcx, r8) = (rdx, r8)/0xFFFFFFFFFFFFFFFF
|
|
; rax = "remainder"
|
|
; where (rdx, r8) = (rcx, r8)*(B - 1) -rax*B^r8 and 0 <= rax < B - 1 B = 0xFFFFFFFFFFFFFFFF
|
|
;
|
|
; this is good but suffers from alignment slowdown
|
|
; we dont seem to have much freedom to re-arrange the instructions to avoid
|
|
; it , I suppose we could detect alignment at the start and have different
|
|
; routines for different alignments
|
|
|
|
%include "..\yasm_mac.inc"
|
|
|
|
CPU Core2
|
|
BITS 64
|
|
|
|
LEAF_PROC mpn_divexact_byff
|
|
movsxd r8, r8d
|
|
|
|
xor eax, eax
|
|
mov r9, r8
|
|
and r9, 3
|
|
shr r8, 2
|
|
cmp r8, 0
|
|
; carry flag is clear here
|
|
jnz loop1
|
|
sbb rax, [rdx]
|
|
mov [rcx], rax
|
|
dec r9
|
|
jz end1
|
|
sbb rax, [rdx+8]
|
|
mov [rcx+8], rax
|
|
dec r9
|
|
jz end1
|
|
sbb rax, [rdx+16]
|
|
mov [rcx+16], rax
|
|
dec r9
|
|
end1:
|
|
sbb rax, 0
|
|
ret
|
|
xalign 16
|
|
loop1:
|
|
sbb rax, [rdx]
|
|
mov [rcx], rax
|
|
sbb rax, [rdx+8]
|
|
mov [rcx+8], rax
|
|
sbb rax, [rdx+16]
|
|
mov [rcx+16], rax
|
|
sbb rax, [rdx+24]
|
|
mov [rcx+24], rax
|
|
lea rdx, [rdx+32]
|
|
dec r8
|
|
lea rcx, [rcx+32]
|
|
jnz loop1
|
|
inc r9
|
|
dec r9
|
|
jz end
|
|
sbb rax, [rdx]
|
|
mov [rcx], rax
|
|
dec r9
|
|
jz end
|
|
sbb rax, [rdx+8]
|
|
mov [rcx+8], rax
|
|
dec r9
|
|
jz end
|
|
sbb rax, [rdx+16]
|
|
mov [rcx+16], rax
|
|
dec r9
|
|
end:
|
|
sbb rax, 0
|
|
ret
|