mpir/mpn/x86_64/atom/rshift.as

167 lines
3.7 KiB
ActionScript

; AMD64 mpn_rshift
; Copyright 2009 Jason Moxham
; 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 version 2.1 of the License, or (at
; your option) any later version.
; 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.
; (rdi,rdx) = (rsi,rdx)>>rcx
; rax = carry
; decent assmeblers understand what movq means ,except
; microsofts/apple masm (what a suprise there) so for the broken old masm
; assembler. Needed for movq reg64,mediareg and movq mediareg,reg64
; only , where mediareg is xmm or mm
%define MOVQ movd
%include 'yasm_mac.inc'
BITS 64
GLOBAL_FUNC mpn_rshift
cmp rdx, 2
ja threeormore
jz two
one:
mov rdx, [rsi]
mov rax, rdx
shr rdx, cl
neg rcx
shl rax, cl
mov [rdi], rdx
ret
two:
mov r8, [rsi]
mov r9, [rsi+8]
mov rax, r8
mov r11, r9
shr r8, cl
shr r9, cl
neg rcx
shl r11, cl
shl rax, cl
or r8, r11
mov [rdi], r8
mov [rdi+8], r9
ret
threeormore:
mov eax, 64
lea r9, [rsi+8]
sub rax, rcx
and r9, -16
MOVQ xmm0, rcx
MOVQ xmm1, rax
movdqa xmm5, [r9]
movdqa xmm3, xmm5
psllq xmm5, xmm1
MOVQ rax, xmm5
cmp rsi, r9
lea rsi, [rsi+rdx*8-40]
je aligned
movq xmm2, [r9-8]
movq xmm4, xmm2
psllq xmm2, xmm1
psrlq xmm4, xmm0
por xmm4, xmm5
movq [rdi], xmm4
lea rdi, [rdi+8]
dec rdx
MOVQ rax, xmm2
aligned:
lea rdi, [rdi+rdx*8-40]
psrlq xmm3, xmm0
mov r8d, 5
sub r8, rdx
jnc skiploop
align 16
loop1:
movdqa xmm2, [rsi+r8*8+16]
movdqa xmm4, xmm2
psllq xmm2, xmm1
shufpd xmm5, xmm2, 1
por xmm3, xmm5
movq [rdi+r8*8], xmm3
movhpd [rdi+r8*8+8], xmm3
psrlq xmm4, xmm0
movdqa xmm5, [rsi+r8*8+32]
movdqa xmm3, xmm5
psllq xmm5, xmm1
shufpd xmm2, xmm5, 1
psrlq xmm3, xmm0
por xmm4, xmm2
movq [rdi+r8*8+16], xmm4
movhpd [rdi+r8*8+24], xmm4
add r8, 4
jnc loop1
skiploop:
cmp r8, 2
ja left0
jz left1
jp left2
left3:
movdqa xmm2, [rsi+r8*8+16]
movdqa xmm4, xmm2
psllq xmm2, xmm1
shufpd xmm5, xmm2, 1
por xmm3, xmm5
movq [rdi+r8*8], xmm3
movhpd [rdi+r8*8+8], xmm3
psrlq xmm4, xmm0
movq xmm5, [rsi+r8*8+32]
movq xmm3, xmm5
psllq xmm5, xmm1
shufpd xmm2, xmm5, 1
psrlq xmm3, xmm0
por xmm4, xmm2
movq [rdi+r8*8+16], xmm4
movhpd [rdi+r8*8+24], xmm4
psrldq xmm5, 8
por xmm3, xmm5
movq [rdi+r8*8+32], xmm3
ret
align 16
left2:
movdqa xmm2, [rsi+r8*8+16]
movdqa xmm4, xmm2
psllq xmm2, xmm1
shufpd xmm5, xmm2, 1
por xmm3, xmm5
movq [rdi+r8*8], xmm3
movhpd [rdi+r8*8+8], xmm3
psrlq xmm4, xmm0
psrldq xmm2, 8
por xmm4, xmm2
movq [rdi+r8*8+16], xmm4
movhpd [rdi+r8*8+24], xmm4
ret
align 16
left1:
movq xmm2, [rsi+r8*8+16]
movq xmm4, xmm2
psllq xmm2, xmm1
shufpd xmm5, xmm2, 1
por xmm3, xmm5
movq [rdi+r8*8], xmm3
movhpd [rdi+r8*8+8], xmm3
psrlq xmm4, xmm0
psrldq xmm2, 8
por xmm4, xmm2
movq [rdi+r8*8+16], xmm4
ret
align 16
left0:
psrldq xmm5, 8
por xmm3, xmm5
movq [rdi+r8*8], xmm3
movhpd [rdi+r8*8+8], xmm3
ret