305 lines
6.2 KiB
NASM
305 lines
6.2 KiB
NASM
; AMD64 mpn_mullow_n_basecase
|
|
;
|
|
; Copyright 2015 Free Software Foundation, Inc.
|
|
;
|
|
; This file is part of the GNU MP Library.
|
|
;
|
|
; The GNU MP Library is free software; you can redistribute it and/or modify
|
|
; it under the terms of either:
|
|
;
|
|
; * the GNU Lesser General Public License as published by the Free
|
|
; Software Foundation; either version 3 of the License, or (at your
|
|
; option) any later version.
|
|
;
|
|
; or
|
|
;
|
|
; * 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.
|
|
;
|
|
; or both in parallel, as here.
|
|
;
|
|
; The GNU MP 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 General Public License
|
|
; for more details.
|
|
;
|
|
; You should have received copies of the GNU General Public License and the
|
|
; GNU Lesser General Public License along with the GNU MP Library. If not,
|
|
; see https://www.gnu.org/licenses/.
|
|
;
|
|
;void mpn_mullow_n_basecase (mp_ptr rp, mp_srcptr up, mp_srcptr vp, mp_size_t n)
|
|
; rdi rsi rdx rcx
|
|
; rcx rdx r8 r9
|
|
|
|
%include 'yasm_mac.inc'
|
|
|
|
text
|
|
align 32
|
|
LEAF_PROC mpn_mullow_n_basecase
|
|
cmp r9, 3
|
|
je asm_sym(?mpn_mullow1)
|
|
ja asm_sym(?mpn_mullow2)
|
|
mov r10, rdx
|
|
mov rax, [rdx]
|
|
|
|
align 16
|
|
cmp r9, 2
|
|
jae .1
|
|
imul rax, [r8]
|
|
mov [rcx], rax
|
|
ret
|
|
|
|
.1: mov r9, [r8]
|
|
mul r9
|
|
mov [rcx], rax
|
|
mov rax, [r10+8]
|
|
imul rax, r9
|
|
add rdx, rax
|
|
mov r9, [r8+8]
|
|
mov r11, [r10]
|
|
imul r11, r9
|
|
add rdx, r11
|
|
mov [rcx+8], rdx
|
|
ret
|
|
|
|
%define reg_save_list rsi, rdi
|
|
|
|
align 32
|
|
FRAME_PROC ?mpn_mullow1, 0, reg_save_list
|
|
mov rdi, rcx
|
|
mov rsi, rdx
|
|
mov rcx, r9
|
|
mov rax, [rsi]
|
|
mov r9, [r8]
|
|
mul r9
|
|
mov [rdi], rax
|
|
mov r10, rdx
|
|
mov rax, [rsi+8]
|
|
mul r9
|
|
imul r9, [rsi+16]
|
|
add r10, rax
|
|
adc r9, rdx
|
|
mov r11, [r8+8]
|
|
mov rax, [rsi]
|
|
mul r11
|
|
add r10, rax
|
|
adc r9, rdx
|
|
imul r11, [rsi+8]
|
|
add r9, r11
|
|
mov [rdi+8], r10
|
|
mov r10, [r8+16]
|
|
mov rax, [rsi]
|
|
imul r10, rax
|
|
add r9, r10
|
|
mov [rdi+16], r9
|
|
END_PROC reg_save_list
|
|
|
|
%define reg_save_list rsi, rdi, rbx, rbp, r12, r13, r14, r15
|
|
|
|
align 32
|
|
FRAME_PROC ?mpn_mullow2, 0, reg_save_list
|
|
mov rdi, rcx
|
|
mov rsi, rdx
|
|
mov rcx, r9
|
|
|
|
mov rax, [rsi]
|
|
mov r9, [r8]
|
|
lea rdi, [rdi+rcx*8]
|
|
lea rsi, [rsi+rcx*8]
|
|
neg rcx
|
|
mul r9
|
|
mov rbx, [r8+8]
|
|
test cl, 1
|
|
jnz .2
|
|
.1: lea rbp, [rcx]
|
|
xor r10, r10
|
|
mov r12, rax
|
|
mov r11, rdx
|
|
jmp .5
|
|
.2: lea rbp, [rcx+1]
|
|
xor r11, r11
|
|
xor r12, r12
|
|
mov r10, rax
|
|
mov r13, rdx
|
|
jmp .4
|
|
align 32
|
|
.3: mul r9
|
|
add r10, rax
|
|
mov r13, rdx
|
|
adc r13, 0
|
|
.4: mov rax, [rsi+rbp*8-8]
|
|
mul rbx
|
|
add r10, r11
|
|
adc r13, 0
|
|
add r12, rax
|
|
mov [rdi+rbp*8-8], r10
|
|
mov r10, rdx
|
|
adc r10, 0
|
|
mov rax, [rsi+rbp*8]
|
|
mul r9
|
|
add r12, rax
|
|
mov r11, rdx
|
|
adc r11, 0
|
|
add r12, r13
|
|
.5: mov rax, [rsi+rbp*8]
|
|
adc r11, 0
|
|
mul rbx
|
|
mov [rdi+rbp*8], r12
|
|
add r10, rax
|
|
mov r12, rdx
|
|
mov rax, [rsi+rbp*8+8]
|
|
adc r12, 0
|
|
add rbp, 2
|
|
jnc .3
|
|
.6: imul rax, r9
|
|
add rax, r10
|
|
add rax, r11
|
|
mov [rdi-8], rax
|
|
add rcx, 2
|
|
lea r8, [r8+16]
|
|
lea rsi, [rsi-16]
|
|
cmp rcx, -2
|
|
jge .19
|
|
.7: mov r9, [r8]
|
|
mov rbx, [r8+8]
|
|
mov rax, [rsi+rcx*8]
|
|
mul r9
|
|
test cl, 1
|
|
jnz .11
|
|
.8: mov r15, [rdi+rcx*8]
|
|
xor r12, r12
|
|
xor r11, r11
|
|
test cl, 2
|
|
jnz .10
|
|
.9: lea rbp, [rcx+1]
|
|
jmp .16
|
|
.10:lea rbp, [rcx+3]
|
|
mov r13, rdx
|
|
add r15, rax
|
|
mov rax, [rsi+rcx*8]
|
|
mov r14, [rdi+rcx*8+8]
|
|
adc r13, 0
|
|
jmp .14
|
|
.11:mov r14, [rdi+rcx*8]
|
|
xor r10, r10
|
|
mov r11, rdx
|
|
test cl, 2
|
|
jz .13
|
|
.12:lea rbp, [rcx+2]
|
|
add r14, rax
|
|
adc r11, 0
|
|
mov rax, [rsi+rcx*8]
|
|
mul rbx
|
|
mov r15, [rdi+rcx*8+8]
|
|
jmp .15
|
|
.13:lea rbp, [rcx]
|
|
xor r13, r13
|
|
jmp .17
|
|
|
|
align 32
|
|
.14:mul rbx
|
|
mov r10, rdx
|
|
add r14, rax
|
|
adc r10, 0
|
|
add r15, r11
|
|
adc r13, 0
|
|
add r14, r12
|
|
adc r10, 0
|
|
mov rax, [rsi+rbp*8-16]
|
|
mul r9
|
|
add r14, rax
|
|
mov r11, rdx
|
|
adc r11, 0
|
|
mov rax, [rsi+rbp*8-16]
|
|
mul rbx
|
|
mov [rdi+rbp*8-24], r15
|
|
mov r15, [rdi+rbp*8-8]
|
|
add r14, r13
|
|
adc r11, 0
|
|
.15:mov r12, rdx
|
|
mov [rdi+rbp*8-16], r14
|
|
add r15, rax
|
|
adc r12, 0
|
|
mov rax, [rsi+rbp*8-8]
|
|
add r15, r10
|
|
adc r12, 0
|
|
mul r9
|
|
.16:add r15, rax
|
|
mov r13, rdx
|
|
adc r13, 0
|
|
mov rax, [rsi+rbp*8-8]
|
|
mul rbx
|
|
add r15, r11
|
|
mov r14, [rdi+rbp*8]
|
|
adc r13, 0
|
|
mov r10, rdx
|
|
add r14, rax
|
|
adc r10, 0
|
|
mov rax, [rsi+rbp*8]
|
|
mul r9
|
|
add r14, r12
|
|
mov [rdi+rbp*8-8], r15
|
|
mov r11, rdx
|
|
adc r10, 0
|
|
.17:add r14, rax
|
|
adc r11, 0
|
|
mov rax, [rsi+rbp*8]
|
|
add r14, r13
|
|
adc r11, 0
|
|
mul rbx
|
|
mov r15, [rdi+rbp*8+8]
|
|
add r15, rax
|
|
mov r12, rdx
|
|
adc r12, 0
|
|
mov rax, [rsi+rbp*8+8]
|
|
mov [rdi+rbp*8], r14
|
|
mul r9
|
|
add r15, r10
|
|
mov r13, rdx
|
|
adc r12, 0
|
|
add r15, rax
|
|
mov rax, [rsi+rbp*8+8]
|
|
mov r14, [rdi+rbp*8+16]
|
|
adc r13, 0
|
|
add rbp, 4
|
|
jnc .14
|
|
.18:imul rax, rbx
|
|
add r14, rax
|
|
add r15, r11
|
|
adc r13, 0
|
|
add r14, r12
|
|
mov rax, [rsi-8]
|
|
imul rax, r9
|
|
add rax, r14
|
|
mov [rdi-16], r15
|
|
add rax, r13
|
|
mov [rdi-8], rax
|
|
add rcx, 2
|
|
lea r8, [r8+16]
|
|
lea rsi, [rsi-16]
|
|
cmp rcx, -2
|
|
jl .7
|
|
jnz .20
|
|
.19:mov r9, [r8]
|
|
mov rbx, [r8+8]
|
|
mov rax, [rsi-16]
|
|
mul r9
|
|
add rax, [rdi-16]
|
|
adc rdx, [rdi-8]
|
|
mov r10, [rsi-8]
|
|
imul r10, r9
|
|
mov r11, [rsi-16]
|
|
imul r11, rbx
|
|
mov [rdi-16], rax
|
|
add r11, r10
|
|
add r11, rdx
|
|
mov [rdi-8], r11
|
|
EXIT_PROC reg_save_list
|
|
|
|
.20:mov r11, [r8]
|
|
imul r11, [rsi-8]
|
|
add r11, rax
|
|
mov [rdi-8], r11
|
|
END_PROC reg_save_list
|
|
|