linux/arch/x86/crypto/sha512-mb/sha512_mb_mgr_submit_avx2.S
<<
>>
Prefs
   1/*
   2 * Buffer submit code for multi buffer SHA512 algorithm
   3 *
   4 * This file is provided under a dual BSD/GPLv2 license.  When using or
   5 * redistributing this file, you may do so under either license.
   6 *
   7 * GPL LICENSE SUMMARY
   8 *
   9 * Copyright(c) 2016 Intel Corporation.
  10 *
  11 * This program is free software; you can redistribute it and/or modify
  12 * it under the terms of version 2 of the GNU General Public License as
  13 * published by the Free Software Foundation.
  14 *
  15 * This program is distributed in the hope that it will be useful, but
  16 * WITHOUT ANY WARRANTY; without even the implied warranty of
  17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  18 * General Public License for more details.
  19 *
  20 * Contact Information:
  21 *     Megha Dey <megha.dey@linux.intel.com>
  22 *
  23 * BSD LICENSE
  24 *
  25 * Copyright(c) 2016 Intel Corporation.
  26 *
  27 * Redistribution and use in source and binary forms, with or without
  28 * modification, are permitted provided that the following conditions
  29 * are met:
  30 *
  31 *   * Redistributions of source code must retain the above copyright
  32 *     notice, this list of conditions and the following disclaimer.
  33 *   * Redistributions in binary form must reproduce the above copyright
  34 *     notice, this list of conditions and the following disclaimer in
  35 *     the documentation and/or other materials provided with the
  36 *     distribution.
  37 *   * Neither the name of Intel Corporation nor the names of its
  38 *     contributors may be used to endorse or promote products derived
  39 *     from this software without specific prior written permission.
  40 *
  41 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  42 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  43 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
  44 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
  45 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  46 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
  47 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
  48 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
  49 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  50 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  51 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  52 */
  53
  54#include <linux/linkage.h>
  55#include <asm/frame.h>
  56#include "sha512_mb_mgr_datastruct.S"
  57
  58.extern sha512_x4_avx2
  59
  60#define arg1    %rdi
  61#define arg2    %rsi
  62
  63#define idx             %rdx
  64#define last_len        %rdx
  65
  66#define size_offset     %rcx
  67#define tmp2            %rcx
  68
  69# Common definitions
  70#define state   arg1
  71#define job     arg2
  72#define len2    arg2
  73#define p2      arg2
  74
  75#define p               %r11
  76#define start_offset    %r11
  77
  78#define unused_lanes    %rbx
  79
  80#define job_rax         %rax
  81#define len             %rax
  82
  83#define lane            %r12
  84#define tmp3            %r12
  85#define lens3           %r12
  86
  87#define extra_blocks    %r8
  88#define lens0           %r8
  89
  90#define tmp             %r9
  91#define lens1           %r9
  92
  93#define lane_data       %r10
  94#define lens2           %r10
  95
  96#define DWORD_len %eax
  97
  98# JOB* sha512_mb_mgr_submit_avx2(MB_MGR *state, JOB *job)
  99# arg 1 : rcx : state
 100# arg 2 : rdx : job
 101ENTRY(sha512_mb_mgr_submit_avx2)
 102        FRAME_BEGIN
 103        push    %rbx
 104        push    %r12
 105
 106        mov     _unused_lanes(state), unused_lanes
 107        movzb     %bl,lane
 108        shr     $8, unused_lanes
 109        imul    $_LANE_DATA_size, lane,lane_data
 110        movl    $STS_BEING_PROCESSED, _status(job)
 111        lea     _ldata(state, lane_data), lane_data
 112        mov     unused_lanes, _unused_lanes(state)
 113        movl    _len(job),  DWORD_len
 114
 115        mov     job, _job_in_lane(lane_data)
 116        movl    DWORD_len,_lens+4(state , lane, 8)
 117
 118        # Load digest words from result_digest
 119        vmovdqu _result_digest+0*16(job), %xmm0
 120        vmovdqu _result_digest+1*16(job), %xmm1
 121        vmovdqu _result_digest+2*16(job), %xmm2
 122        vmovdqu _result_digest+3*16(job), %xmm3
 123
 124        vmovq    %xmm0, _args_digest(state, lane, 8)
 125        vpextrq  $1, %xmm0, _args_digest+1*32(state , lane, 8)
 126        vmovq    %xmm1, _args_digest+2*32(state , lane, 8)
 127        vpextrq  $1, %xmm1, _args_digest+3*32(state , lane, 8)
 128        vmovq    %xmm2, _args_digest+4*32(state , lane, 8)
 129        vpextrq  $1, %xmm2, _args_digest+5*32(state , lane, 8)
 130        vmovq    %xmm3, _args_digest+6*32(state , lane, 8)
 131        vpextrq  $1, %xmm3, _args_digest+7*32(state , lane, 8)
 132
 133        mov     _buffer(job), p
 134        mov     p, _args_data_ptr(state, lane, 8)
 135
 136        cmp     $0xFF, unused_lanes
 137        jne     return_null
 138
 139start_loop:
 140
 141        # Find min length
 142        mov     _lens+0*8(state),lens0
 143        mov     lens0,idx
 144        mov     _lens+1*8(state),lens1
 145        cmp     idx,lens1
 146        cmovb   lens1, idx
 147        mov     _lens+2*8(state),lens2
 148        cmp     idx,lens2
 149        cmovb   lens2,idx
 150        mov     _lens+3*8(state),lens3
 151        cmp     idx,lens3
 152        cmovb   lens3,idx
 153        mov     idx,len2
 154        and     $0xF,idx
 155        and     $~0xFF,len2
 156        jz      len_is_0
 157
 158        sub     len2,lens0
 159        sub     len2,lens1
 160        sub     len2,lens2
 161        sub     len2,lens3
 162        shr     $32,len2
 163        mov     lens0, _lens + 0*8(state)
 164        mov     lens1, _lens + 1*8(state)
 165        mov     lens2, _lens + 2*8(state)
 166        mov     lens3, _lens + 3*8(state)
 167
 168        # "state" and "args" are the same address, arg1
 169        # len is arg2
 170        call    sha512_x4_avx2
 171        # state and idx are intact
 172
 173len_is_0:
 174
 175        # process completed job "idx"
 176        imul    $_LANE_DATA_size, idx, lane_data
 177        lea     _ldata(state, lane_data), lane_data
 178
 179        mov     _job_in_lane(lane_data), job_rax
 180        mov     _unused_lanes(state), unused_lanes
 181        movq    $0, _job_in_lane(lane_data)
 182        movl    $STS_COMPLETED, _status(job_rax)
 183        shl     $8, unused_lanes
 184        or      idx, unused_lanes
 185        mov     unused_lanes, _unused_lanes(state)
 186
 187        movl    $0xFFFFFFFF,_lens+4(state,idx,8)
 188        vmovq    _args_digest+0*32(state , idx, 8), %xmm0
 189        vpinsrq  $1, _args_digest+1*32(state , idx, 8), %xmm0, %xmm0
 190        vmovq    _args_digest+2*32(state , idx, 8), %xmm1
 191        vpinsrq  $1, _args_digest+3*32(state , idx, 8), %xmm1, %xmm1
 192        vmovq    _args_digest+4*32(state , idx, 8), %xmm2
 193        vpinsrq  $1, _args_digest+5*32(state , idx, 8), %xmm2, %xmm2
 194        vmovq    _args_digest+6*32(state , idx, 8), %xmm3
 195        vpinsrq  $1, _args_digest+7*32(state , idx, 8), %xmm3, %xmm3
 196
 197        vmovdqu  %xmm0, _result_digest + 0*16(job_rax)
 198        vmovdqu  %xmm1, _result_digest + 1*16(job_rax)
 199        vmovdqu  %xmm2, _result_digest + 2*16(job_rax)
 200        vmovdqu  %xmm3, _result_digest + 3*16(job_rax)
 201
 202return:
 203        pop     %r12
 204        pop     %rbx
 205        FRAME_END
 206        ret
 207
 208return_null:
 209        xor     job_rax, job_rax
 210        jmp     return
 211ENDPROC(sha512_mb_mgr_submit_avx2)
 212.data
 213
 214.align 16
 215H0:     .int  0x6a09e667
 216H1:     .int  0xbb67ae85
 217H2:     .int  0x3c6ef372
 218H3:     .int  0xa54ff53a
 219H4:     .int  0x510e527f
 220H5:     .int  0x9b05688c
 221H6:     .int  0x1f83d9ab
 222H7:     .int  0x5be0cd19
 223