mpir/mpn/x86_64w/haswell/lshift1.asm
2016-11-26 22:52:26 +00:00

267 lines
6.8 KiB
NASM

; Copyright 2016 Jens Nurmann and Alexander Kruppa
; 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.
; mp_limb_t mpn_lshift1(mp_ptr Op2, mp_srcptr Op1, mp_size_t Size1 )
; Linux RAX RDI RSI RDX
; Win7 RAX RCX RDX R8
;
; Description:
; The function shifts Op1 left by one bit, stores the result in Op2 (non-
; destructive shl) and hands back the shifted-out most significant bit of Op1.
; The function operates decreasing in memory supporting in-place operation.
;
; Caveats:
; - the AVX version uses mnemonics only available on Haswell, Broadwell and
; Skylake cores
; - the behaviour of cache prefetching in combination with AVX shifting seems
; somewhat erratic
; - slight (a few clock cycles) degradation for 1/2 LD1$ sizes
; - slight (a few percent) improvement for full LD1$ sizes
; - substantial (>10%) improvement for 1/2 LD2$ sizes
; - slight (a few percent) improvement for full LD2$ sizes
; - slight (a few percent) degradation for 1/2 LD3$ sizes
; - substantial (around 10%) degradation for full LD3$ sizes
;
; Comments:
; - implemented, tested and benched on 21.02.2016 by jn
; - includes cache prefetching
%define USE_WIN64
%include 'yasm_mac.inc'
BITS 64
%ifdef USE_WIN64
%define Op2 RCX
%define Op1 RDX
%define Size1 R8
%define Limb1 R9
%define Limb2 R10
%define Offs -512 ; used direct def. to stay in Win scratch regs
%define ShlDL0 XMM2 ; ATTN: this must match ShlQL0 definition
%define ShrDL0 XMM3 ; ATTN: this must match ShrQL0 definition
%define QLimb0 YMM0
%define QLimb1 YMM1
%define ShlQL0 YMM2
%define ShrQL0 YMM3
%define ShlQL1 YMM4
%define ShrQL1 YMM5
%else
%define Op2 RDI
%define Op1 RSI
%define Size1 RDX
%define Limb1 R8
%define Limb2 R9
%define Offs -512 ; used direct def. to stay in Win scratch regs
%define ShlDL0 XMM2 ; ATTN: this must match ShlQL0 definition
%define ShrDL0 XMM3 ; ATTN: this must match ShrQL0 definition
%define QLimb0 YMM0
%define QLimb1 YMM1
%define ShlQL0 YMM2
%define ShrQL0 YMM3
%define ShlQL1 YMM4
%define ShrQL1 YMM5
%endif
align 32
LEAF_PROC mpn_lshift1
xor EAX, EAX
sub Size1, 1
jc .Exit ;ajs:notshortform ; Size1=0 =>
lea Op1, [Op1+8*Size1]
lea Op2, [Op2+8*Size1]
mov Limb1, [Op1]
shld RAX, Limb1, 1
or Size1, Size1
je .lShl1EquPost ;ajs:notshortform ; Size1=1 =>
cmp Size1, 8
jc .lShl1EquFour ;ajs:notshortform ; AVX inefficient =>
; first align Op2 to 32 bytes
test Op2, 8
jne .lShl1EquA16
mov Limb2, [Op1-8]
shld Limb1, Limb2, 1
mov [Op2], Limb1
mov Limb1, Limb2
sub Op1, 8
sub Op2, 8
sub Size1, 1
.lShl1EquA16:
test Op2, 16
jne .lShl1EquAVX
mov Limb2, [Op1-8]
shld Limb1, Limb2, 1
mov [Op2], Limb1
mov Limb1, [Op1-16]
shld Limb2, Limb1, 1
mov [Op2-8], Limb2
sub Op1, 16
sub Op2, 16
sub Size1, 2
.lShl1EquAVX:
; pre-fetch first quad-limb
vmovdqu QLimb0, [Op1-24]
vpsrlq ShrQL0, QLimb0, 63
vpermq ShrQL0, ShrQL0, 147 ; 0b10010011
sub Op1, 32
sub Size1, 4
jmp .lShl1EquAVXCheck
; main loop requires on entry:
; - 0.60 cycles per limb in LD1$
; - 0.60-0.75 cycles per limb in LD2$
; - 0.75-1.00 cycles per limb in LD3$
align 16
.lShl1EquAVXLoop:
%ifdef USE_PREFETCH
prefetchnta [Op1+Offs]
%endif
vmovdqu QLimb1, [Op1-24]
vpsllq ShlQL0, QLimb0, 1
vmovdqu QLimb0, [Op1-56]
vpsrlq ShrQL1, QLimb1, 63
vpermq ShrQL1, ShrQL1, 147 ; 0b10010011
vpblendd ShrQL0, ShrQL0, ShrQL1, 3 ; 0b00000011
vpor ShlQL0, ShlQL0, ShrQL0
vpsllq ShlQL1, QLimb1, 1
vpsrlq ShrQL0, QLimb0, 63
vpermq ShrQL0, ShrQL0, 147 ; 0b10010011
vpblendd ShrQL1, ShrQL1, ShrQL0, 3 ; 0b00000011
vmovdqa [Op2-24], ShlQL0
vpor ShlQL1, ShlQL1, ShrQL1
vmovdqa [Op2-56], ShlQL1
sub Op1, 64
sub Op2, 64
.lShl1EquAVXCheck:
sub Size1, 8
jnc .lShl1EquAVXLoop
mov Limb2, [Op1]
mov Limb1, Limb2
shr Limb2, 63
%if 1
vmovq ShlDL0, Limb2
vpblendd ShrQL0, ShrQL0, ShlQL0, 3
%else
; I am mixing in a single SSE4.1 instruction into otherwise pure AVX2
; this is generating stalls on Haswell & Broadwell architecture (Agner Fog)
; but it is only executed once and there is no AVX2 based alternative
; Insert value of Limb2 into the 0-th qword of ShrDL0
pinsrq ShrDL0, Limb2, 0 ; SSE4.1
%endif
vpsllq ShlQL0, QLimb0, 1
vpor ShlQL0, ShlQL0, ShrQL0
vmovdqa [Op2-24], ShlQL0
sub Op2, 32
add Size1, 8
; shift remaining max. 7 limbs with SHLD mnemonic
.lShl1EquFour:
sub Op1, 8
test Size1, 4
je .lShl1EquTwo
mov Limb2, [Op1]
shld Limb1, Limb2, 1
mov [Op2], Limb1
mov Limb1, [Op1-8]
shld Limb2, Limb1, 1
mov [Op2-8], Limb2
mov Limb2, [Op1-16]
shld Limb1, Limb2, 1
mov [Op2-16], Limb1
mov Limb1, [Op1-24]
shld Limb2, Limb1, 1
mov [Op2-24], Limb2
sub Op1, 32
sub Op2, 32
.lShl1EquTwo:
test Size1, 2
je .lShl1EquOne
mov Limb2, [Op1]
shld Limb1, Limb2, 1
mov [Op2], Limb1
mov Limb1, [Op1-8]
shld Limb2, Limb1, 1
mov [Op2-8], Limb2
sub Op1, 16
sub Op2, 16
.lShl1EquOne:
test Size1, 1
je .lShl1EquPost
mov Limb2, [Op1]
shld Limb1, Limb2, 1
mov [Op2], Limb1
mov Limb1, Limb2
sub Op2, 8
.lShl1EquPost:
shl Limb1, 1
mov [Op2], Limb1
.Exit:
vzeroupper
ret
.end: