mirror of
https://github.com/VectorCamp/vectorscan.git
synced 2025-09-30 03:34:25 +03:00
rose: add new instruction CHECK_MASK_32
This is a specialisation of the "lookaround" code.
This commit is contained in:
86
src/util/copybytes.h
Normal file
86
src/util/copybytes.h
Normal file
@@ -0,0 +1,86 @@
|
||||
/*
|
||||
* Copyright (c) 2016, Intel Corporation
|
||||
*
|
||||
* 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.
|
||||
*/
|
||||
|
||||
#ifndef COPY_BYTES_H
|
||||
#define COPY_BYTES_H
|
||||
|
||||
#include "unaligned.h"
|
||||
#include "simd_utils.h"
|
||||
|
||||
static really_inline
|
||||
void copy_upto_32_bytes(u8 *dst, const u8 *src, unsigned int len) {
|
||||
switch (len) {
|
||||
case 0:
|
||||
break;
|
||||
case 1:
|
||||
*dst = *src;
|
||||
break;
|
||||
case 2:
|
||||
unaligned_store_u16(dst, unaligned_load_u16(src));
|
||||
break;
|
||||
case 3:
|
||||
unaligned_store_u16(dst, unaligned_load_u16(src));
|
||||
dst[2] = src[2];
|
||||
break;
|
||||
case 4:
|
||||
unaligned_store_u32(dst, unaligned_load_u32(src));
|
||||
break;
|
||||
case 5:
|
||||
case 6:
|
||||
case 7:
|
||||
unaligned_store_u32(dst + len - 4, unaligned_load_u32(src + len - 4));
|
||||
unaligned_store_u32(dst, unaligned_load_u32(src));
|
||||
break;
|
||||
case 8:
|
||||
unaligned_store_u64a(dst, unaligned_load_u64a(src));
|
||||
break;
|
||||
case 9:
|
||||
case 10:
|
||||
case 11:
|
||||
case 12:
|
||||
case 13:
|
||||
case 14:
|
||||
case 15:
|
||||
unaligned_store_u64a(dst + len - 8, unaligned_load_u64a(src + len - 8));
|
||||
unaligned_store_u64a(dst, unaligned_load_u64a(src));
|
||||
break;
|
||||
case 16:
|
||||
storeu128(dst, loadu128(src));
|
||||
break;
|
||||
case 32:
|
||||
storeu256(dst, loadu256(src));
|
||||
break;
|
||||
default:
|
||||
assert(len < 32);
|
||||
storeu128(dst + len - 16, loadu128(src + len - 16));
|
||||
storeu128(dst, loadu128(src));
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
#endif
|
@@ -71,6 +71,7 @@
|
||||
|
||||
#include "ue2common.h"
|
||||
#include "simd_types.h"
|
||||
#include "unaligned.h"
|
||||
|
||||
// Define a common assume_aligned using an appropriate compiler built-in, if
|
||||
// it's available. Note that we need to handle C or C++ compilation.
|
||||
@@ -354,6 +355,21 @@ m256 set32x8(u32 in) {
|
||||
return rv;
|
||||
}
|
||||
|
||||
static really_inline
|
||||
m256 eq256(m256 a, m256 b) {
|
||||
m256 rv;
|
||||
rv.lo = eq128(a.lo, b.lo);
|
||||
rv.hi = eq128(a.hi, b.hi);
|
||||
return rv;
|
||||
}
|
||||
|
||||
static really_inline
|
||||
u32 movemask256(m256 a) {
|
||||
u32 lo_mask = movemask128(a.lo);
|
||||
u32 hi_mask = movemask128(a.hi);
|
||||
return lo_mask | (hi_mask << 16);
|
||||
}
|
||||
|
||||
#endif
|
||||
|
||||
static really_inline m256 zeroes256(void) {
|
||||
@@ -525,6 +541,16 @@ static really_inline m256 loadu256(const void *ptr) {
|
||||
#endif
|
||||
}
|
||||
|
||||
// unaligned store
|
||||
static really_inline void storeu256(void *ptr, m256 a) {
|
||||
#if defined(__AVX2__)
|
||||
_mm256_storeu_si256((m256 *)ptr, a);
|
||||
#else
|
||||
storeu128(ptr, a.lo);
|
||||
storeu128((char *)ptr + 16, a.hi);
|
||||
#endif
|
||||
}
|
||||
|
||||
// packed unaligned store of first N bytes
|
||||
static really_inline
|
||||
void storebytes256(void *ptr, m256 a, unsigned int n) {
|
||||
|
Reference in New Issue
Block a user