mirror of
https://github.com/VectorCamp/vectorscan.git
synced 2025-06-28 16:41:01 +03:00
216 lines
7.0 KiB
C++
216 lines
7.0 KiB
C++
/*
|
|
* Copyright (c) 2015-2017, 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.
|
|
*/
|
|
|
|
#include "fdr_internal.h"
|
|
#include "fdr_compile_internal.h"
|
|
#include "fdr_confirm.h"
|
|
#include "ue2common.h"
|
|
#include "hs_internal.h"
|
|
#include "fdr_engine_description.h"
|
|
#include "teddy_internal.h"
|
|
#include "teddy_engine_description.h"
|
|
|
|
#include <cmath>
|
|
|
|
using namespace std;
|
|
|
|
namespace ue2 {
|
|
|
|
TeddyEngineDescription::TeddyEngineDescription(const TeddyEngineDef &def)
|
|
: EngineDescription(def.id, targetByArchFeatures(def.cpu_features),
|
|
def.numBuckets),
|
|
numMasks(def.numMasks), packed(def.packed) {}
|
|
|
|
u32 TeddyEngineDescription::getDefaultFloodSuffixLength() const {
|
|
return numMasks;
|
|
}
|
|
|
|
void getTeddyDescriptions(vector<TeddyEngineDescription> *out) {
|
|
static const TeddyEngineDef defns[] = {
|
|
{ 3, HS_CPU_FEATURES_AVX2, 1, 16, false },
|
|
{ 4, HS_CPU_FEATURES_AVX2, 1, 16, true },
|
|
{ 5, HS_CPU_FEATURES_AVX2, 2, 16, false },
|
|
{ 6, HS_CPU_FEATURES_AVX2, 2, 16, true },
|
|
{ 7, HS_CPU_FEATURES_AVX2, 3, 16, false },
|
|
{ 8, HS_CPU_FEATURES_AVX2, 3, 16, true },
|
|
{ 9, HS_CPU_FEATURES_AVX2, 4, 16, false },
|
|
{ 10, HS_CPU_FEATURES_AVX2, 4, 16, true },
|
|
{ 11, 0, 1, 8, false },
|
|
{ 12, 0, 1, 8, true },
|
|
{ 13, 0, 2, 8, false },
|
|
{ 14, 0, 2, 8, true },
|
|
{ 15, 0, 3, 8, false },
|
|
{ 16, 0, 3, 8, true },
|
|
{ 17, 0, 4, 8, false },
|
|
{ 18, 0, 4, 8, true },
|
|
};
|
|
out->clear();
|
|
for (const auto &def : defns) {
|
|
out->emplace_back(def);
|
|
}
|
|
}
|
|
|
|
static
|
|
size_t maxFloodTailLen(const vector<hwlmLiteral> &vl) {
|
|
size_t max_flood_tail = 0;
|
|
for (const auto &lit : vl) {
|
|
const string &s = lit.s;
|
|
assert(!s.empty());
|
|
size_t j;
|
|
for (j = 1; j < s.length(); j++) {
|
|
if (s[s.length() - j - 1] != s[s.length() - 1]) {
|
|
break;
|
|
}
|
|
}
|
|
max_flood_tail = max(max_flood_tail, j);
|
|
}
|
|
return max_flood_tail;
|
|
}
|
|
|
|
/**
|
|
* \brief True if this Teddy engine is qualified to handle this set of literals
|
|
* on this target.
|
|
*/
|
|
static
|
|
bool isAllowed(const vector<hwlmLiteral> &vl, const TeddyEngineDescription &eng,
|
|
const size_t max_lit_len, const target_t &target) {
|
|
if (!eng.isValidOnTarget(target)) {
|
|
DEBUG_PRINTF("%u disallowed: not valid on target\n", eng.getID());
|
|
return false;
|
|
}
|
|
if (eng.getNumBuckets() < vl.size() && !eng.packed) {
|
|
DEBUG_PRINTF("%u disallowed: num buckets < num lits and not packed\n",
|
|
eng.getID());
|
|
return false;
|
|
}
|
|
if (eng.getNumBuckets() * TEDDY_BUCKET_LOAD < vl.size()) {
|
|
DEBUG_PRINTF("%u disallowed: too many lits for num buckets\n",
|
|
eng.getID());
|
|
return false;
|
|
}
|
|
if (eng.numMasks > max_lit_len) {
|
|
DEBUG_PRINTF("%u disallowed: more masks than max lit len (%zu)\n",
|
|
eng.getID(), max_lit_len);
|
|
return false;
|
|
}
|
|
|
|
if (vl.size() > 40) {
|
|
u32 n_small_lits = 0;
|
|
for (const auto &lit : vl) {
|
|
if (lit.s.length() < eng.numMasks) {
|
|
n_small_lits++;
|
|
}
|
|
}
|
|
if (n_small_lits * 5 > vl.size()) {
|
|
DEBUG_PRINTF("too many short literals (%u)\n", n_small_lits);
|
|
return false;
|
|
}
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
unique_ptr<TeddyEngineDescription>
|
|
chooseTeddyEngine(const target_t &target, const vector<hwlmLiteral> &vl) {
|
|
vector<TeddyEngineDescription> descs;
|
|
getTeddyDescriptions(&descs);
|
|
const TeddyEngineDescription *best = nullptr;
|
|
|
|
const size_t max_lit_len = maxLen(vl);
|
|
const size_t max_flood_tail = maxFloodTailLen(vl);
|
|
DEBUG_PRINTF("%zu lits, max_lit_len=%zu, max_flood_tail=%zu\n", vl.size(),
|
|
max_lit_len, max_flood_tail);
|
|
|
|
u32 best_score = 0;
|
|
for (size_t engineID = 0; engineID < descs.size(); engineID++) {
|
|
const TeddyEngineDescription &eng = descs[engineID];
|
|
if (!isAllowed(vl, eng, max_lit_len, target)) {
|
|
continue;
|
|
}
|
|
|
|
u32 score = 0;
|
|
|
|
// We prefer unpacked Teddy models.
|
|
if (!eng.packed) {
|
|
score += 100;
|
|
}
|
|
|
|
// If we're heavily loaded, we prefer to have more masks.
|
|
if (vl.size() > 4 * eng.getNumBuckets()) {
|
|
score += eng.numMasks * 4;
|
|
} else {
|
|
// Lightly loaded cases are great.
|
|
score += 100;
|
|
}
|
|
|
|
// We want enough masks to avoid becoming flood-prone.
|
|
if (eng.numMasks > max_flood_tail) {
|
|
score += 50;
|
|
}
|
|
|
|
// We prefer having 3 masks. 3 is just right.
|
|
score += 6 / (abs(3 - (int)eng.numMasks) + 1);
|
|
|
|
// We prefer cheaper, smaller Teddy models.
|
|
score += 16 / eng.getNumBuckets();
|
|
|
|
DEBUG_PRINTF("teddy %u: masks=%u, buckets=%u, packed=%u "
|
|
"-> score=%u\n",
|
|
eng.getID(), eng.numMasks, eng.getNumBuckets(),
|
|
eng.packed ? 1U : 0U, score);
|
|
|
|
if (!best || score > best_score) {
|
|
best = ŋ
|
|
best_score = score;
|
|
}
|
|
}
|
|
|
|
if (!best) {
|
|
DEBUG_PRINTF("failed to find engine\n");
|
|
return nullptr;
|
|
}
|
|
|
|
DEBUG_PRINTF("using engine %u\n", best->getID());
|
|
return std::make_unique<TeddyEngineDescription>(*best);
|
|
}
|
|
|
|
unique_ptr<TeddyEngineDescription> getTeddyDescription(u32 engineID) {
|
|
vector<TeddyEngineDescription> descs;
|
|
getTeddyDescriptions(&descs);
|
|
|
|
for (const auto &desc : descs) {
|
|
if (desc.getID() == engineID) {
|
|
return std::make_unique<TeddyEngineDescription>(desc);
|
|
}
|
|
}
|
|
|
|
return nullptr;
|
|
}
|
|
|
|
} // namespace ue2
|