isa-l/raid/pq_gen_avx2.asm
Greg Tucker d6c5e9620d Add raid unit
New raid unit adds source for optimized xor and P+Q functions.

Signed-off-by: Greg Tucker <greg.b.tucker@intel.com>
2016-04-26 15:58:49 -07:00

257 lines
7.5 KiB
NASM

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Copyright(c) 2011-2015 Intel Corporation All rights reserved.
;
; Redistribution and use in source and binary forms, with or without
; modification, are permitted provided that the following conditions
; are met:
; * Redistributions of source code must retain the above copyright
; notice, this list of conditions and the following disclaimer.
; * Redistributions in binary form must reproduce the above copyright
; notice, this list of conditions and the following disclaimer in
; the documentation and/or other materials provided with the
; distribution.
; * Neither the name of Intel Corporation nor the names of its
; contributors may be used to endorse or promote products derived
; from this software without specific prior written permission.
;
; THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
; "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
; LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
; A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
; OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
; SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
; LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
; DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
; THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
; (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
; OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;; Optimized pq of N source vectors using AVX
;;; int pq_gen_avx(int vects, int len, void **array)
;;; Generates P+Q parity vector from N (vects-2) sources in array of pointers
;;; (**array). Last two pointers are the P and Q destinations respectively.
;;; Vectors must be aligned to 32 bytes. Length must be 32 byte aligned.
%include "reg_sizes.asm"
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
%define arg2 rdx
%define arg3 rcx
%define arg4 r8
%define arg5 r9
%define tmp r11
%define tmp3 arg4
%define return rax
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%endif
%ifidn __OUTPUT_FORMAT__, win64
%define arg0 rcx
%define arg1 rdx
%define arg2 r8
%define arg3 r9
%define tmp r11
%define tmp3 r10
%define return rax
%define stack_size 8*32 + 8 ; must be an odd multiple of 8
%define func(x) proc_frame x
%macro FUNC_SAVE 0
alloc_stack stack_size
;; Until a sav_ymm256 is defined
vmovdqu [rsp + 0*32], ymm6
vmovdqu [rsp + 1*32], ymm7
vmovdqu [rsp + 2*32], ymm8
vmovdqu [rsp + 3*32], ymm9
vmovdqu [rsp + 4*32], ymm10
vmovdqu [rsp + 5*32], ymm11
vmovdqu [rsp + 6*32], ymm14
vmovdqu [rsp + 7*32], ymm15
end_prolog
%endmacro
%macro FUNC_RESTORE 0
vmovdqu ymm6, [rsp + 0*32]
vmovdqu ymm7, [rsp + 1*32]
vmovdqu ymm8, [rsp + 2*32]
vmovdqu ymm9, [rsp + 3*32]
vmovdqu ymm10, [rsp + 4*32]
vmovdqu ymm11, [rsp + 5*32]
vmovdqu ymm14, [rsp + 6*32]
vmovdqu ymm15, [rsp + 7*32]
add rsp, stack_size
%endmacro
%endif
%define vec arg0
%define len arg1
%define ptr arg3
%define pos rax
%define xp1 ymm0
%define xq1 ymm1
%define xtmp1 ymm2
%define xs1 ymm3
%define xp2 ymm4
%define xq2 ymm5
%define xtmp2 ymm6
%define xs2 ymm7
%define xp3 ymm8
%define xq3 ymm9
%define xtmp3 ymm10
%define xs3 ymm11
%define xzero ymm14
%define xpoly ymm15
;;; Use Non-temporal load/stor
%ifdef NO_NT_LDST
%define XLDR vmovdqa
%define XSTR vmovdqa
%else
%define XLDR vmovntdqa
%define XSTR vmovntdq
%endif
default rel
[bits 64]
section .text
align 16
global pq_gen_avx2:function
func(pq_gen_avx2)
FUNC_SAVE
sub vec, 3 ;Keep as offset to last source
jng return_fail ;Must have at least 2 sources
cmp len, 0
je return_pass
test len, (32-1) ;Check alignment of length
jnz return_fail
mov pos, 0
vmovdqa xpoly, [poly]
vpxor xzero, xzero, xzero
cmp len, 96
jl loop32
len_aligned_32bytes:
sub len, 3*32 ;Len points to last block
loop96:
mov ptr, [arg2+vec*8] ;Fetch last source pointer
mov tmp, vec ;Set tmp to point back to last vector
XLDR xs1, [ptr+pos] ;Preload last vector (source)
XLDR xs2, [ptr+pos+32] ;Preload last vector (source)
XLDR xs3, [ptr+pos+64] ;Preload last vector (source)
vpxor xp1, xp1, xp1 ;p1 = 0
vpxor xp2, xp2, xp2 ;p2 = 0
vpxor xp3, xp3, xp3 ;p3 = 0
vpxor xq1, xq1, xq1 ;q1 = 0
vpxor xq2, xq2, xq2 ;q2 = 0
vpxor xq3, xq3, xq3 ;q3 = 0
next_vect:
sub tmp, 1 ;Inner loop for each source vector
mov ptr, [arg2+tmp*8] ; get pointer to next vect
vpxor xq1, xq1, xs1 ; q1 ^= s1
vpxor xq2, xq2, xs2 ; q2 ^= s2
vpxor xq3, xq3, xs3 ; q3 ^= s3
vpxor xp1, xp1, xs1 ; p1 ^= s1
vpxor xp2, xp2, xs2 ; p2 ^= s2
vpxor xp3, xp3, xs3 ; p3 ^= s2
vpblendvb xtmp1, xzero, xpoly, xq1 ; xtmp1 = poly or 0x00
vpblendvb xtmp2, xzero, xpoly, xq2 ; xtmp2 = poly or 0x00
vpblendvb xtmp3, xzero, xpoly, xq3 ; xtmp3 = poly or 0x00
XLDR xs1, [ptr+pos] ; Get next vector (source data1)
XLDR xs2, [ptr+pos+32] ; Get next vector (source data2)
XLDR xs3, [ptr+pos+64] ; Get next vector (source data3)
vpaddb xq1, xq1, xq1 ; q1 = q1<<1
vpaddb xq2, xq2, xq2 ; q2 = q2<<1
vpaddb xq3, xq3, xq3 ; q3 = q3<<1
vpxor xq1, xq1, xtmp1 ; q1 = q1<<1 ^ poly_masked
vpxor xq2, xq2, xtmp2 ; q2 = q2<<1 ^ poly_masked
vpxor xq3, xq3, xtmp3 ; q3 = q3<<1 ^ poly_masked
jg next_vect ; Loop for each vect except 0
mov ptr, [arg2+8+vec*8] ;Get address of P parity vector
mov tmp, [arg2+(2*8)+vec*8] ;Get address of Q parity vector
vpxor xp1, xp1, xs1 ;p1 ^= s1[0] - last source is already loaded
vpxor xq1, xq1, xs1 ;q1 ^= 1 * s1[0]
vpxor xp2, xp2, xs2 ;p2 ^= s2[0]
vpxor xq2, xq2, xs2 ;q2 ^= 1 * s2[0]
vpxor xp3, xp3, xs3 ;p3 ^= s3[0]
vpxor xq3, xq3, xs3 ;q3 ^= 1 * s3[0]
XSTR [ptr+pos], xp1 ;Write parity P1 vector
XSTR [ptr+pos+32], xp2 ;Write parity P2 vector
XSTR [ptr+pos+64], xp3 ;Write parity P3 vector
XSTR [tmp+pos], xq1 ;Write parity Q1 vector
XSTR [tmp+pos+32], xq2 ;Write parity Q2 vector
XSTR [tmp+pos+64], xq3 ;Write parity Q3 vector
add pos, 3*32
cmp pos, len
jle loop96
;; ------------------------------
;; Do last 16 or 32 Bytes remaining
add len, 3*32
cmp pos, len
je return_pass
loop32:
mov ptr, [arg2+vec*8] ;Fetch last source pointer
mov tmp, vec ;Set tmp to point back to last vector
XLDR xs1, [ptr+pos] ;Preload last vector (source)
vpxor xp1, xp1, xp1 ;p = 0
vpxor xq1, xq1, xq1 ;q = 0
next_vect32:
sub tmp, 1 ;Inner loop for each source vector
mov ptr, [arg2+tmp*8] ; get pointer to next vect
vpxor xq1, xq1, xs1 ; q1 ^= s1
vpblendvb xtmp1, xzero, xpoly, xq1 ; xtmp1 = poly or 0x00
vpxor xp1, xp1, xs1 ; p ^= s
vpaddb xq1, xq1, xq1 ; q = q<<1
vpxor xq1, xq1, xtmp1 ; q = q<<1 ^ poly_masked
XLDR xs1, [ptr+pos] ; Get next vector (source data)
jg next_vect32 ; Loop for each vect except 0
mov ptr, [arg2+8+vec*8] ;Get address of P parity vector
mov tmp, [arg2+(2*8)+vec*8] ;Get address of Q parity vector
vpxor xp1, xp1, xs1 ;p ^= s[0] - last source is already loaded
vpxor xq1, xq1, xs1 ;q ^= 1 * s[0]
XSTR [ptr+pos], xp1 ;Write parity P vector
XSTR [tmp+pos], xq1 ;Write parity Q vector
add pos, 32
cmp pos, len
jl loop32
return_pass:
mov return, 0
FUNC_RESTORE
ret
return_fail:
mov return, 1
FUNC_RESTORE
ret
endproc_frame
section .data
align 32
poly:
dq 0x1d1d1d1d1d1d1d1d, 0x1d1d1d1d1d1d1d1d
dq 0x1d1d1d1d1d1d1d1d, 0x1d1d1d1d1d1d1d1d
;;; func core, ver, snum
slversion pq_gen_avx2, 04, 03, 0041