vectorscan/src/nfa/shufti_sve.hpp
George Wort 00fff3f53c Use SVE for double shufti.
Change-Id: I09e0d57bb8a2f05b613f6225dea79ae823136268
2021-10-12 11:51:34 +03:00

242 lines
10 KiB
C++

/*
* Copyright (c) 2015-2017, Intel Corporation
* Copyright (c) 2021, Arm Limited
*
* 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.
*/
/** \file
* \brief Shufti: character class acceleration.
*
* Utilises the SVE tbl shuffle instruction
*/
static really_inline
svbool_t singleMatched(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf, svbool_t pg) {
svuint8_t vec = svld1_u8(pg, buf);
svuint8_t c_lo = svtbl(mask_lo, svand_z(svptrue_b8(), vec, (uint8_t)0xf));
svuint8_t c_hi = svtbl(mask_hi, svlsr_z(svptrue_b8(), vec, 4));
svuint8_t t = svand_z(svptrue_b8(), c_lo, c_hi);
return svcmpne(pg, t, (uint8_t)0);
}
static really_inline
const u8 *shuftiOnce(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf, const u8 *buf_end) {
DEBUG_PRINTF("start %p end %p\n", buf, buf_end);
assert(buf <= buf_end);
svbool_t pg = svwhilelt_b8_s64(0, buf_end - buf);
svbool_t matched = singleMatched(mask_lo, mask_hi, buf, pg);
return accelSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *shuftiLoopBody(svuint8_t mask_lo, svuint8_t mask_hi, const u8 *buf) {
DEBUG_PRINTF("start %p end %p\n", buf, buf + svcntb());
svbool_t matched = singleMatched(mask_lo, mask_hi, buf, svptrue_b8());
return accelSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *rshuftiOnce(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf, const u8 *buf_end) {
DEBUG_PRINTF("start %p end %p\n", buf, buf_end);
assert(buf <= buf_end);
DEBUG_PRINTF("l = %td\n", buf_end - buf);
svbool_t pg = svwhilelt_b8_s64(0, buf_end - buf);
svbool_t matched = singleMatched(mask_lo, mask_hi, buf, pg);
return accelRevSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *rshuftiLoopBody(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf) {
svbool_t matched = singleMatched(mask_lo, mask_hi, buf, svptrue_b8());
return accelRevSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *shuftiSearch(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf, const u8 *buf_end) {
assert(buf < buf_end);
size_t len = buf_end - buf;
if (len <= svcntb()) {
return shuftiOnce(mask_lo, mask_hi, buf, buf_end);
}
// peel off first part to align to the vector size
const u8 *aligned_buf = ROUNDUP_PTR(buf, svcntb_pat(SV_POW2));
assert(aligned_buf < buf_end);
if (buf != aligned_buf) {
const u8 *ptr = shuftiLoopBody(mask_lo, mask_hi, buf);
if (ptr) return ptr;
}
buf = aligned_buf;
size_t loops = (buf_end - buf) / svcntb();
DEBUG_PRINTF("loops %zu \n", loops);
for (size_t i = 0; i < loops; i++, buf += svcntb()) {
const u8 *ptr = shuftiLoopBody(mask_lo, mask_hi, buf);
if (ptr) return ptr;
}
DEBUG_PRINTF("buf %p buf_end %p \n", buf, buf_end);
return buf == buf_end ? NULL : shuftiLoopBody(mask_lo, mask_hi,
buf_end - svcntb());
}
static really_inline
const u8 *rshuftiSearch(svuint8_t mask_lo, svuint8_t mask_hi,
const u8 *buf, const u8 *buf_end) {
assert(buf < buf_end);
size_t len = buf_end - buf;
if (len <= svcntb()) {
return rshuftiOnce(mask_lo, mask_hi, buf, buf_end);
}
// peel off first part to align to the vector size
const u8 *aligned_buf_end = ROUNDDOWN_PTR(buf_end, svcntb_pat(SV_POW2));
assert(buf < aligned_buf_end);
if (buf_end != aligned_buf_end) {
const u8 *ptr = rshuftiLoopBody(mask_lo, mask_hi, buf_end - svcntb());
if (ptr) return ptr;
}
buf_end = aligned_buf_end;
size_t loops = (buf_end - buf) / svcntb();
DEBUG_PRINTF("loops %zu \n", loops);
for (size_t i = 0; i < loops; i++) {
buf_end -= svcntb();
const u8 *ptr = rshuftiLoopBody(mask_lo, mask_hi, buf_end);
if (ptr) return ptr;
}
DEBUG_PRINTF("buf %p buf_end %p \n", buf, buf_end);
return buf == buf_end ? NULL : rshuftiLoopBody(mask_lo, mask_hi, buf);
}
const u8 *shuftiExec(m128 mask_lo, m128 mask_hi, const u8 *buf,
const u8 *buf_end) {
DEBUG_PRINTF("shufti scan over %td bytes\n", buf_end - buf);
svuint8_t sve_mask_lo = getSVEMaskFrom128(mask_lo);
svuint8_t sve_mask_hi = getSVEMaskFrom128(mask_hi);
const u8 *ptr = shuftiSearch(sve_mask_lo, sve_mask_hi, buf, buf_end);
return ptr ? ptr : buf_end;
}
const u8 *rshuftiExec(m128 mask_lo, m128 mask_hi, const u8 *buf,
const u8 *buf_end) {
DEBUG_PRINTF("rshufti scan over %td bytes\n", buf_end - buf);
svuint8_t sve_mask_lo = getSVEMaskFrom128(mask_lo);
svuint8_t sve_mask_hi = getSVEMaskFrom128(mask_hi);
const u8 *ptr = rshuftiSearch(sve_mask_lo, sve_mask_hi, buf, buf_end);
return ptr ? ptr : buf - 1;
}
static really_inline
svbool_t doubleMatched(svuint8_t mask1_lo, svuint8_t mask1_hi,
svuint8_t mask2_lo, svuint8_t mask2_hi,
const u8 *buf, const svbool_t pg) {
svuint8_t vec = svld1_u8(pg, buf);
svuint8_t chars_lo = svand_x(svptrue_b8(), vec, (uint8_t)0xf);
svuint8_t chars_hi = svlsr_x(svptrue_b8(), vec, 4);
svuint8_t c1_lo = svtbl(mask1_lo, chars_lo);
svuint8_t c1_hi = svtbl(mask1_hi, chars_hi);
svuint8_t t1 = svorr_x(svptrue_b8(), c1_lo, c1_hi);
svuint8_t c2_lo = svtbl(mask2_lo, chars_lo);
svuint8_t c2_hi = svtbl(mask2_hi, chars_hi);
svuint8_t t2 = svext(svorr_z(pg, c2_lo, c2_hi), svdup_u8(0), 1);
svuint8_t t = svorr_x(svptrue_b8(), t1, t2);
return svnot_z(svptrue_b8(), svcmpeq(svptrue_b8(), t, (uint8_t)0xff));
}
static really_inline
const u8 *dshuftiOnce(svuint8_t mask1_lo, svuint8_t mask1_hi,
svuint8_t mask2_lo, svuint8_t mask2_hi,
const u8 *buf, const u8 *buf_end) {
DEBUG_PRINTF("start %p end %p\n", buf, buf_end);
assert(buf < buf_end);
DEBUG_PRINTF("l = %td\n", buf_end - buf);
svbool_t pg = svwhilelt_b8_s64(0, buf_end - buf);
svbool_t matched = doubleMatched(mask1_lo, mask1_hi, mask2_lo, mask2_hi,
buf, pg);
return accelSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *dshuftiLoopBody(svuint8_t mask1_lo, svuint8_t mask1_hi,
svuint8_t mask2_lo, svuint8_t mask2_hi,
const u8 *buf) {
DEBUG_PRINTF("start %p end %p\n", buf, buf + svcntb());
svbool_t matched = doubleMatched(mask1_lo, mask1_hi, mask2_lo, mask2_hi,
buf, svptrue_b8());
return accelSearchCheckMatched(buf, matched);
}
static really_inline
const u8 *dshuftiSearch(svuint8_t mask1_lo, svuint8_t mask1_hi,
svuint8_t mask2_lo, svuint8_t mask2_hi,
const u8 *buf, const u8 *buf_end) {
assert(buf < buf_end);
size_t len = buf_end - buf;
if (len <= svcntb()) {
return dshuftiOnce(mask1_lo, mask1_hi,
mask2_lo, mask2_hi, buf, buf_end);
}
// peel off first part to align to the vector size
const u8 *aligned_buf = ROUNDUP_PTR(buf, svcntb_pat(SV_POW2));
assert(aligned_buf < buf_end);
if (buf != aligned_buf) {
const u8 *ptr = dshuftiLoopBody(mask1_lo, mask1_hi,
mask2_lo, mask2_hi, buf);
if (ptr) return ptr;
}
buf = aligned_buf;
size_t loops = (buf_end - buf) / svcntb();
DEBUG_PRINTF("loops %zu \n", loops);
for (size_t i = 0; i < loops; i++, buf += svcntb()) {
const u8 *ptr = dshuftiLoopBody(mask1_lo, mask1_hi,
mask2_lo, mask2_hi, buf);
if (ptr) return ptr;
}
DEBUG_PRINTF("buf %p buf_end %p \n", buf, buf_end);
return buf == buf_end ? NULL : dshuftiLoopBody(mask1_lo, mask1_hi,
mask2_lo, mask2_hi,
buf_end - svcntb());
}
const u8 *shuftiDoubleExec(m128 mask1_lo, m128 mask1_hi,
m128 mask2_lo, m128 mask2_hi,
const u8 *buf, const u8 *buf_end) {
DEBUG_PRINTF("double shufti scan %td bytes\n", buf_end - buf);
DEBUG_PRINTF("buf %p buf_end %p \n", buf, buf_end);
svuint8_t sve_mask1_lo = getSVEMaskFrom128(mask1_lo);
svuint8_t sve_mask1_hi = getSVEMaskFrom128(mask1_hi);
svuint8_t sve_mask2_lo = getSVEMaskFrom128(mask2_lo);
svuint8_t sve_mask2_hi = getSVEMaskFrom128(mask2_hi);
const u8 *ptr = dshuftiSearch(sve_mask1_lo, sve_mask1_hi,
sve_mask2_lo, sve_mask2_hi, buf, buf_end);
return ptr ? ptr : buf_end;
}