LookingGlass/common/include/common/memcpySSE.h

136 lines
4.1 KiB
C
Raw Normal View History

2017-11-17 03:08:45 +00:00
/*
KVMGFX Client - A KVM Client for VGA Passthrough
2019-02-22 11:16:14 +00:00
Copyright (C) 2017-2019 Geoffrey McRae <geoff@hostfission.com>
2017-12-11 17:30:47 +00:00
https://looking-glass.hostfission.com
2017-11-17 03:08:45 +00:00
This program is free software; you can redistribute it and/or modify it under
the terms of 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.
This program 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 a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place, Suite 330, Boston, MA 02111-1307 USA
*/
#pragma once
#include <stdbool.h>
#include <stdint.h>
#include <string.h>
#include <tmmintrin.h>
#include <immintrin.h>
#include "debug.h"
#if defined(NATIVE_MEMCPY)
#define memcpySSE memcpy
#elif defined(_MSC_VER)
extern "C" void * memcpySSE(void *dst, const void * src, size_t length);
#elif (defined(__GNUC__) || defined(__GNUG__)) && defined(__i386__)
inline static void * memcpySSE(void *dst, const void * src, size_t length)
2017-11-17 03:08:45 +00:00
{
if (length == 0 || dst == src)
return;
2017-11-17 03:08:45 +00:00
// copies under 1MB are faster with the inlined memcpy
// tell the dev to use that instead
if (length < 1048576)
2017-11-17 03:08:45 +00:00
{
static bool smallBufferWarn = false;
if (!smallBufferWarn)
2017-11-17 03:08:45 +00:00
{
DEBUG_WARN("Do not use memcpySSE for copies under 1MB in size!");
smallBufferWarn = true;
2017-11-17 03:08:45 +00:00
}
memcpy(dst, src, length);
return;
2017-11-17 03:08:45 +00:00
}
const void * end = dst + (length & ~0x7F);
const size_t off = (7 - ((length & 0x7F) >> 4)) * 9;
__asm__ __volatile__ (
"cmp %[dst],%[end] \n\t"
"je Remain_%= \n\t"
// perform SIMD block copy
"loop_%=: \n\t"
"movaps 0x00(%[src]),%%xmm0 \n\t"
"movaps 0x10(%[src]),%%xmm1 \n\t"
"movaps 0x20(%[src]),%%xmm2 \n\t"
"movaps 0x30(%[src]),%%xmm3 \n\t"
"movaps 0x40(%[src]),%%xmm4 \n\t"
"movaps 0x50(%[src]),%%xmm5 \n\t"
"movaps 0x60(%[src]),%%xmm6 \n\t"
"movaps 0x70(%[src]),%%xmm7 \n\t"
"movntdq %%xmm0 ,0x00(%[dst]) \n\t"
"movntdq %%xmm1 ,0x10(%[dst]) \n\t"
"movntdq %%xmm2 ,0x20(%[dst]) \n\t"
"movntdq %%xmm3 ,0x30(%[dst]) \n\t"
"movntdq %%xmm4 ,0x40(%[dst]) \n\t"
"movntdq %%xmm5 ,0x50(%[dst]) \n\t"
"movntdq %%xmm6 ,0x60(%[dst]) \n\t"
"movntdq %%xmm7 ,0x70(%[dst]) \n\t"
"add $0x80,%[dst] \n\t"
"add $0x80,%[src] \n\t"
"cmp %[dst],%[end] \n\t"
"jne loop_%= \n\t"
"Remain_%=: \n\t"
// copy any remaining 16 byte blocks
"call GetPC_%=\n\t"
"Offset_%=:\n\t"
"add $(BlockTable_%= - Offset_%=), %%eax \n\t"
"add %[off],%%eax \n\t"
"jmp *%%eax \n\t"
"GetPC_%=:\n\t"
"mov (%%esp), %%eax \n\t"
"ret \n\t"
"BlockTable_%=:\n\t"
"movaps 0x60(%[src]),%%xmm6 \n\t"
"movntdq %%xmm6 ,0x60(%[dst]) \n\t"
"movaps 0x50(%[src]),%%xmm5 \n\t"
"movntdq %%xmm5 ,0x50(%[dst]) \n\t"
"movaps 0x40(%[src]),%%xmm4 \n\t"
"movntdq %%xmm4 ,0x40(%[dst]) \n\t"
"movaps 0x30(%[src]),%%xmm3 \n\t"
"movntdq %%xmm3 ,0x30(%[dst]) \n\t"
"movaps 0x20(%[src]),%%xmm2 \n\t"
"movntdq %%xmm2 ,0x20(%[dst]) \n\t"
"movaps 0x10(%[src]),%%xmm1 \n\t"
"movntdq %%xmm1 ,0x10(%[dst]) \n\t"
"movaps 0x00(%[src]),%%xmm0 \n\t"
"movntdq %%xmm0 ,0x00(%[dst]) \n\t"
"nop\n\t"
"nop\n\t"
: [dst]"+r" (dst),
[src]"+r" (src)
: [off]"r" (off),
[end]"r" (end)
: "eax",
"xmm0",
"xmm1",
"xmm2",
"xmm3",
"xmm4",
"xmm5",
"xmm6",
"xmm7",
"memory"
);
//copy any remaining bytes
memcpy(dst, src, length & 0xF);
}
#else
#define memcpySSE memcpy
#endif