github_mirrors/vectorscan
3
0
Fork 1
mirror of https://github.com/VectorCamp/vectorscan.git synced 2025-06-28 16:41:01 +03:00
Code Issues Packages Projects Releases Wiki Activity
vectorscan/src/rose/rose_build_program.cpp
Hong, Yang A dea7c4dc2e lookaround: 
add 64x8 and 64x16 shufti models
add mask64 model
expand entry quantity
2021-01-25 14:13:13 +02:00

2601 lines
89 KiB
C++
Raw Blame History

/*
* Copyright (c) 2016-2020, 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 "rose_build_program.h"
#include "rose_build_engine_blob.h"
#include "rose_build_instructions.h"
#include "rose_build_lookaround.h"
#include "rose_build_resources.h"
#include "nfa/nfa_api_queue.h"
#include "nfa/nfa_build_util.h"
#include "nfa/tamaramacompile.h"
#include "nfagraph/ng_util.h"
#include "util/charreach_util.h"
#include "util/container.h"
#include "util/compile_context.h"
#include "util/compile_error.h"
#include "util/report_manager.h"
#include "util/unordered.h"
#include "util/verify_types.h"
#include <boost/range/adaptor/map.hpp>
#include <algorithm>
#include <cstring>
using namespace std;
using boost::adaptors::map_values;
using boost::adaptors::map_keys;
namespace ue2 {
engine_info::engine_info(const NFA *nfa, bool trans)
: type((NFAEngineType)nfa->type), accepts_eod(nfaAcceptsEod(nfa)),
stream_size(nfa->streamStateSize),
scratch_size(nfa->scratchStateSize),
scratch_align(state_alignment(*nfa)),
transient(trans) {
assert(scratch_align);
}
left_build_info::left_build_info(u32 q, u32 l, u32 t, rose_group sm,
const std::vector<u8> &stops, u32 max_ql,
u8 cm_count, const CharReach &cm_cr)
: queue(q), lag(l), transient(t), squash_mask(sm), stopAlphabet(stops),
max_queuelen(max_ql), countingMiracleCount(cm_count),
countingMiracleReach(cm_cr) {
}
left_build_info::left_build_info(const vector<vector<LookEntry>> &looks)
: has_lookaround(true), lookaround(looks) {
}
using OffsetMap = RoseInstruction::OffsetMap;
static
OffsetMap makeOffsetMap(const RoseProgram &program, u32 *total_len) {
OffsetMap offset_map;
u32 offset = 0;
for (const auto &ri : program) {
offset = ROUNDUP_N(offset, ROSE_INSTR_MIN_ALIGN);
DEBUG_PRINTF("instr %p (opcode %d) -> offset %u\n", ri.get(),
ri->code(), offset);
assert(!contains(offset_map, ri.get()));
offset_map.emplace(ri.get(), offset);
offset += ri->byte_length();
}
*total_len = offset;
return offset_map;
}
RoseProgram::RoseProgram() {
prog.push_back(make_unique<RoseInstrEnd>());
}
RoseProgram::~RoseProgram() = default;
RoseProgram::RoseProgram(RoseProgram &&) = default;
RoseProgram &RoseProgram::operator=(RoseProgram &&) = default;
bool RoseProgram::empty() const {
assert(!prog.empty());
assert(prog.back()->code() == ROSE_INSTR_END);
// Empty if we only have one element, the END instruction.
return next(prog.begin()) == prog.end();
}
const RoseInstruction *RoseProgram::end_instruction() const {
assert(!prog.empty());
assert(prog.back()->code() == ROSE_INSTR_END);
return prog.back().get();
}
void RoseProgram::update_targets(RoseProgram::iterator it,
RoseProgram::iterator it_end,
const RoseInstruction *old_target,
const RoseInstruction *new_target) {
assert(old_target && new_target && old_target != new_target);
for (; it != it_end; ++it) {
unique_ptr<RoseInstruction> &ri = *it;
assert(ri);
ri->update_target(old_target, new_target);
}
}
RoseProgram::iterator RoseProgram::insert(RoseProgram::iterator it,
unique_ptr<RoseInstruction> ri) {
assert(!prog.empty());
assert(it != end());
assert(prog.back()->code() == ROSE_INSTR_END);
return prog.insert(it, move(ri));
}
RoseProgram::iterator RoseProgram::insert(RoseProgram::iterator it,
RoseProgram &&block) {
assert(!prog.empty());
assert(it != end());
assert(prog.back()->code() == ROSE_INSTR_END);
if (block.empty()) {
return it;
}
const RoseInstruction *end_ptr = block.end_instruction();
assert(end_ptr->code() == ROSE_INSTR_END);
block.prog.pop_back();
const RoseInstruction *new_target = it->get();
update_targets(block.prog.begin(), block.prog.end(), end_ptr, new_target);
// Workaround: container insert() for ranges doesn't return an iterator
// in the version of the STL distributed with gcc 4.8.
auto dist = distance(prog.begin(), it);
prog.insert(it, make_move_iterator(block.prog.begin()),
make_move_iterator(block.prog.end()));
it = prog.begin();
advance(it, dist);
return it;
}
RoseProgram::iterator RoseProgram::erase(RoseProgram::iterator first,
RoseProgram::iterator last) {
return prog.erase(first, last);
}
void RoseProgram::add_before_end(std::unique_ptr<RoseInstruction> ri) {
assert(!prog.empty());
insert(std::prev(prog.end()), std::move(ri));
}
void RoseProgram::add_before_end(RoseProgram &&block) {
assert(!prog.empty());
assert(prog.back()->code() == ROSE_INSTR_END);
if (block.empty()) {
return;
}
insert(prev(prog.end()), move(block));
}
void RoseProgram::add_block(RoseProgram &&block) {
assert(!prog.empty());
assert(prog.back()->code() == ROSE_INSTR_END);
if (block.empty()) {
return;
}
// Replace pointers to the current END with pointers to the first
// instruction in the new sequence.
const RoseInstruction *end_ptr = end_instruction();
prog.pop_back();
update_targets(prog.begin(), prog.end(), end_ptr,
block.prog.front().get());
prog.insert(prog.end(), make_move_iterator(block.prog.begin()),
make_move_iterator(block.prog.end()));
}
bytecode_ptr<char> writeProgram(RoseEngineBlob &blob,
const RoseProgram &program) {
u32 total_len = 0;
const auto offset_map = makeOffsetMap(program, &total_len);
DEBUG_PRINTF("%zu instructions, len %u\n", program.size(), total_len);
auto bytecode = make_zeroed_bytecode_ptr<char>(total_len,
ROSE_INSTR_MIN_ALIGN);
char *ptr = bytecode.get();
for (const auto &ri : program) {
assert(contains(offset_map, ri.get()));
const u32 offset = offset_map.at(ri.get());
ri->write(ptr + offset, blob, offset_map);
}
return bytecode;
}
size_t RoseProgramHash::operator()(const RoseProgram &program) const {
size_t v = 0;
for (const auto &ri : program) {
assert(ri);
hash_combine(v, ri->hash());
}
return v;
}
bool RoseProgramEquivalence::operator()(const RoseProgram &prog1,
const RoseProgram &prog2) const {
if (prog1.size() != prog2.size()) {
return false;
}
u32 len_1 = 0, len_2 = 0;
const auto offset_map_1 = makeOffsetMap(prog1, &len_1);
const auto offset_map_2 = makeOffsetMap(prog2, &len_2);
if (len_1 != len_2) {
return false;
}
auto is_equiv = [&](const unique_ptr<RoseInstruction> &a,
const unique_ptr<RoseInstruction> &b) {
assert(a && b);
return a->equiv(*b, offset_map_1, offset_map_2);
};
return std::equal(prog1.begin(), prog1.end(), prog2.begin(), is_equiv);
}
/* Removes any CHECK_HANDLED instructions from the given program */
static
void stripCheckHandledInstruction(RoseProgram &prog) {
for (auto it = prog.begin(); it != prog.end();) {
auto ins = dynamic_cast<const RoseInstrCheckNotHandled *>(it->get());
if (!ins) {
++it;
continue;
}
auto next_it = next(it);
assert(next_it != prog.end()); /* there should always be an end ins */
auto next_ins = next_it->get();
/* update all earlier instructions which point to ins to instead point
* to the next instruction. Only need to look at earlier as we only ever
* jump forward. */
RoseProgram::update_targets(prog.begin(), it, ins, next_ins);
/* remove check handled instruction */
it = prog.erase(it, next_it);
}
}
/** Returns true if the program may read the interpreter's work_done flag */
static
bool reads_work_done_flag(const RoseProgram &prog) {
for (const auto &ri : prog) {
if (dynamic_cast<const RoseInstrSquashGroups *>(ri.get())) {
return true;
}
}
return false;
}
void addEnginesEodProgram(u32 eodNfaIterOffset, RoseProgram &program) {
if (!eodNfaIterOffset) {
return;
}
RoseProgram block;
block.add_before_end(make_unique<RoseInstrEnginesEod>(eodNfaIterOffset));
program.add_block(move(block));
}
void addSuffixesEodProgram(RoseProgram &program) {
RoseProgram block;
block.add_before_end(make_unique<RoseInstrSuffixesEod>());
program.add_block(move(block));
}
void addMatcherEodProgram(RoseProgram &program) {
RoseProgram block;
block.add_before_end(make_unique<RoseInstrMatcherEod>());
program.add_block(move(block));
}
void addFlushCombinationProgram(RoseProgram &program) {
program.add_before_end(make_unique<RoseInstrFlushCombination>());
}
void addLastFlushCombinationProgram(RoseProgram &program) {
program.add_before_end(make_unique<RoseInstrLastFlushCombination>());
}
static
void makeRoleCheckLeftfix(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
RoseVertex v, RoseProgram &program) {
auto it = leftfix_info.find(v);
if (it == end(leftfix_info)) {
return;
}
const left_build_info &lni = it->second;
if (lni.has_lookaround) {
return; // Leftfix completely implemented by lookaround.
}
assert(!build.cc.streaming ||
build.g[v].left.lag <= MAX_STORED_LEFTFIX_LAG);
bool is_prefix = build.isRootSuccessor(v);
const auto *end_inst = program.end_instruction();
unique_ptr<RoseInstruction> ri;
if (is_prefix) {
ri = make_unique<RoseInstrCheckPrefix>(lni.queue, build.g[v].left.lag,
build.g[v].left.leftfix_report,
end_inst);
} else {
ri = make_unique<RoseInstrCheckInfix>(lni.queue, build.g[v].left.lag,
build.g[v].left.leftfix_report,
end_inst);
}
program.add_before_end(move(ri));
}
static
void makeAnchoredLiteralDelay(const RoseBuildImpl &build,
const ProgramBuild &prog_build, u32 lit_id,
RoseProgram &program) {
// Only relevant for literals in the anchored table.
const rose_literal_id &lit = build.literals.at(lit_id);
if (lit.table != ROSE_ANCHORED) {
return;
}
// If this literal match cannot occur after floatingMinLiteralMatchOffset,
// we do not need this check.
bool all_too_early = true;
rose_group groups = 0;
const auto &lit_vertices = build.literal_info.at(lit_id).vertices;
for (RoseVertex v : lit_vertices) {
if (build.g[v].max_offset > prog_build.floatingMinLiteralMatchOffset) {
all_too_early = false;
}
groups |= build.g[v].groups;
}
if (all_too_early) {
return;
}
assert(contains(prog_build.anchored_programs, lit_id));
u32 anch_id = prog_build.anchored_programs.at(lit_id);
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrAnchoredDelay>(groups, anch_id, end_inst);
program.add_before_end(move(ri));
}
static
void makeDedupe(const ReportManager &rm, const Report &report,
RoseProgram &program) {
const auto *end_inst = program.end_instruction();
auto ri =
make_unique<RoseInstrDedupe>(report.quashSom, rm.getDkey(report),
report.offsetAdjust, end_inst);
program.add_before_end(move(ri));
}
static
void makeDedupeSom(const ReportManager &rm, const Report &report,
RoseProgram &program) {
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrDedupeSom>(report.quashSom,
rm.getDkey(report),
report.offsetAdjust, end_inst);
program.add_before_end(move(ri));
}
static
void makeCatchup(const ReportManager &rm, bool needs_catchup,
const flat_set<ReportID> &reports, RoseProgram &program) {
if (!needs_catchup) {
return;
}
// Everything except the INTERNAL_ROSE_CHAIN report needs catchup to run
// before reports are triggered.
auto report_needs_catchup = [&](const ReportID &id) {
const Report &report = rm.getReport(id);
return report.type != INTERNAL_ROSE_CHAIN;
};
if (!any_of(begin(reports), end(reports), report_needs_catchup)) {
DEBUG_PRINTF("none of the given reports needs catchup\n");
return;
}
program.add_before_end(make_unique<RoseInstrCatchUp>());
}
static
void writeSomOperation(const Report &report, som_operation *op) {
assert(op);
memset(op, 0, sizeof(*op));
switch (report.type) {
case EXTERNAL_CALLBACK_SOM_REL:
op->type = SOM_EXTERNAL_CALLBACK_REL;
break;
case INTERNAL_SOM_LOC_SET:
op->type = SOM_INTERNAL_LOC_SET;
break;
case INTERNAL_SOM_LOC_SET_IF_UNSET:
op->type = SOM_INTERNAL_LOC_SET_IF_UNSET;
break;
case INTERNAL_SOM_LOC_SET_IF_WRITABLE:
op->type = SOM_INTERNAL_LOC_SET_IF_WRITABLE;
break;
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA:
op->type = SOM_INTERNAL_LOC_SET_REV_NFA;
break;
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_UNSET:
op->type = SOM_INTERNAL_LOC_SET_REV_NFA_IF_UNSET;
break;
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_WRITABLE:
op->type = SOM_INTERNAL_LOC_SET_REV_NFA_IF_WRITABLE;
break;
case INTERNAL_SOM_LOC_COPY:
op->type = SOM_INTERNAL_LOC_COPY;
break;
case INTERNAL_SOM_LOC_COPY_IF_WRITABLE:
op->type = SOM_INTERNAL_LOC_COPY_IF_WRITABLE;
break;
case INTERNAL_SOM_LOC_MAKE_WRITABLE:
op->type = SOM_INTERNAL_LOC_MAKE_WRITABLE;
break;
case EXTERNAL_CALLBACK_SOM_STORED:
op->type = SOM_EXTERNAL_CALLBACK_STORED;
break;
case EXTERNAL_CALLBACK_SOM_ABS:
op->type = SOM_EXTERNAL_CALLBACK_ABS;
break;
case EXTERNAL_CALLBACK_SOM_REV_NFA:
op->type = SOM_EXTERNAL_CALLBACK_REV_NFA;
break;
case INTERNAL_SOM_LOC_SET_FROM:
op->type = SOM_INTERNAL_LOC_SET_FROM;
break;
case INTERNAL_SOM_LOC_SET_FROM_IF_WRITABLE:
op->type = SOM_INTERNAL_LOC_SET_FROM_IF_WRITABLE;
break;
default:
// This report doesn't correspond to a SOM operation.
assert(0);
throw CompileError("Unable to generate bytecode.");
}
op->onmatch = report.onmatch;
switch (report.type) {
case EXTERNAL_CALLBACK_SOM_REV_NFA:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_UNSET:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_WRITABLE:
op->aux.revNfaIndex = report.revNfaIndex;
break;
default:
op->aux.somDistance = report.somDistance;
break;
}
}
static
void addLogicalSetRequired(const Report &report, ReportManager &rm,
RoseProgram &program) {
if (report.lkey == INVALID_LKEY) {
return;
}
// set matching status of current lkey
auto risl = make_unique<RoseInstrSetLogical>(report.lkey,
report.offsetAdjust);
program.add_before_end(move(risl));
// set current lkey's corresponding ckeys active, pending to check
for (auto ckey : rm.getRelateCKeys(report.lkey)) {
auto risc = make_unique<RoseInstrSetCombination>(ckey);
program.add_before_end(move(risc));
}
}
static
void makeReport(const RoseBuildImpl &build, const ReportID id,
const bool has_som, RoseProgram &program) {
assert(id < build.rm.numReports());
const Report &report = build.rm.getReport(id);
RoseProgram report_block;
const RoseInstruction *end_inst = report_block.end_instruction();
// Handle min/max offset checks.
if (report.minOffset > 0 || report.maxOffset < MAX_OFFSET) {
auto ri = make_unique<RoseInstrCheckBounds>(report.minOffset,
report.maxOffset, end_inst);
report_block.add_before_end(move(ri));
}
// If this report has an exhaustion key, we can check it in the program
// rather than waiting until we're in the callback adaptor.
if (report.ekey != INVALID_EKEY) {
auto ri = make_unique<RoseInstrCheckExhausted>(report.ekey, end_inst);
report_block.add_before_end(move(ri));
}
// External SOM reports that aren't passthrough need their SOM value
// calculated.
if (isExternalSomReport(report) &&
report.type != EXTERNAL_CALLBACK_SOM_PASS) {
auto ri = make_unique<RoseInstrSomFromReport>();
writeSomOperation(report, &ri->som);
report_block.add_before_end(move(ri));
}
// Min length constraint.
if (report.minLength > 0) {
assert(build.hasSom);
auto ri = make_unique<RoseInstrCheckMinLength>(
report.offsetAdjust, report.minLength, end_inst);
report_block.add_before_end(move(ri));
}
if (report.quashSom) {
report_block.add_before_end(make_unique<RoseInstrSomZero>());
}
switch (report.type) {
case EXTERNAL_CALLBACK:
if (build.rm.numCkeys()) {
addFlushCombinationProgram(report_block);
}
if (!has_som) {
// Dedupe is only necessary if this report has a dkey, or if there
// are SOM reports to catch up.
bool needs_dedupe = build.rm.getDkey(report) != ~0U || build.hasSom;
if (report.ekey == INVALID_EKEY) {
if (needs_dedupe) {
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrDedupeAndReport>(
report.quashSom, build.rm.getDkey(report),
report.onmatch, report.offsetAdjust, end_inst));
} else {
makeDedupe(build.rm, report, report_block);
}
} else {
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrReport>(
report.onmatch, report.offsetAdjust));
}
}
} else {
if (needs_dedupe) {
makeDedupe(build.rm, report, report_block);
}
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrReportExhaust>(
report.onmatch, report.offsetAdjust, report.ekey));
} else {
report_block.add_before_end(
make_unique<RoseInstrSetExhaust>(report.ekey));
}
}
} else { // has_som
makeDedupeSom(build.rm, report, report_block);
if (report.ekey == INVALID_EKEY) {
if (!report.quiet) {
report_block.add_before_end(make_unique<RoseInstrReportSom>(
report.onmatch, report.offsetAdjust));
}
} else {
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrReportSomExhaust>(
report.onmatch, report.offsetAdjust, report.ekey));
} else {
report_block.add_before_end(
make_unique<RoseInstrSetExhaust>(report.ekey));
}
}
}
addLogicalSetRequired(report, build.rm, report_block);
break;
case INTERNAL_SOM_LOC_SET:
case INTERNAL_SOM_LOC_SET_IF_UNSET:
case INTERNAL_SOM_LOC_SET_IF_WRITABLE:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_UNSET:
case INTERNAL_SOM_LOC_SET_SOM_REV_NFA_IF_WRITABLE:
case INTERNAL_SOM_LOC_COPY:
case INTERNAL_SOM_LOC_COPY_IF_WRITABLE:
case INTERNAL_SOM_LOC_MAKE_WRITABLE:
case INTERNAL_SOM_LOC_SET_FROM:
case INTERNAL_SOM_LOC_SET_FROM_IF_WRITABLE:
if (build.rm.numCkeys()) {
addFlushCombinationProgram(report_block);
}
if (has_som) {
auto ri = make_unique<RoseInstrReportSomAware>();
writeSomOperation(report, &ri->som);
report_block.add_before_end(move(ri));
} else {
auto ri = make_unique<RoseInstrReportSomInt>();
writeSomOperation(report, &ri->som);
report_block.add_before_end(move(ri));
}
break;
case INTERNAL_ROSE_CHAIN: {
report_block.add_before_end(make_unique<RoseInstrReportChain>(
report.onmatch, report.topSquashDistance));
break;
}
case EXTERNAL_CALLBACK_SOM_REL:
case EXTERNAL_CALLBACK_SOM_STORED:
case EXTERNAL_CALLBACK_SOM_ABS:
case EXTERNAL_CALLBACK_SOM_REV_NFA:
if (build.rm.numCkeys()) {
addFlushCombinationProgram(report_block);
}
makeDedupeSom(build.rm, report, report_block);
if (report.ekey == INVALID_EKEY) {
if (!report.quiet) {
report_block.add_before_end(make_unique<RoseInstrReportSom>(
report.onmatch, report.offsetAdjust));
}
} else {
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrReportSomExhaust>(
report.onmatch, report.offsetAdjust, report.ekey));
} else {
report_block.add_before_end(
make_unique<RoseInstrSetExhaust>(report.ekey));
}
}
addLogicalSetRequired(report, build.rm, report_block);
break;
case EXTERNAL_CALLBACK_SOM_PASS:
if (build.rm.numCkeys()) {
addFlushCombinationProgram(report_block);
}
makeDedupeSom(build.rm, report, report_block);
if (report.ekey == INVALID_EKEY) {
if (!report.quiet) {
report_block.add_before_end(make_unique<RoseInstrReportSom>(
report.onmatch, report.offsetAdjust));
}
} else {
if (!report.quiet) {
report_block.add_before_end(
make_unique<RoseInstrReportSomExhaust>(
report.onmatch, report.offsetAdjust, report.ekey));
} else {
report_block.add_before_end(
make_unique<RoseInstrSetExhaust>(report.ekey));
}
}
addLogicalSetRequired(report, build.rm, report_block);
break;
default:
assert(0);
throw CompileError("Unable to generate bytecode.");
}
program.add_block(move(report_block));
}
static
void makeRoleReports(const RoseBuildImpl &build,
const std::map<RoseVertex, left_build_info> &leftfix_info,
bool needs_catchup, RoseVertex v, RoseProgram &program) {
const auto &g = build.g;
bool report_som = false;
if (g[v].left.tracksSom()) {
/* we are a suffaig - need to update role to provide som to the
* suffix. */
assert(contains(leftfix_info, v));
const left_build_info &lni = leftfix_info.at(v);
program.add_before_end(
make_unique<RoseInstrSomLeftfix>(lni.queue, g[v].left.lag));
report_som = true;
} else if (g[v].som_adjust) {
program.add_before_end(
make_unique<RoseInstrSomAdjust>(g[v].som_adjust));
report_som = true;
}
makeCatchup(build.rm, needs_catchup, g[v].reports, program);
RoseProgram report_block;
for (ReportID id : g[v].reports) {
makeReport(build, id, report_som, report_block);
}
program.add_before_end(move(report_block));
}
static
void makeRoleSetState(const unordered_map<RoseVertex, u32> &roleStateIndices,
RoseVertex v, RoseProgram &program) {
// We only need this instruction if a state index has been assigned to this
// vertex.
auto it = roleStateIndices.find(v);
if (it == end(roleStateIndices)) {
return;
}
program.add_before_end(make_unique<RoseInstrSetState>(it->second));
}
static
void makePushDelayedInstructions(const RoseLiteralMap &literals,
ProgramBuild &prog_build,
const flat_set<u32> &delayed_ids,
RoseProgram &program) {
vector<RoseInstrPushDelayed> delay_instructions;
for (const auto &delayed_lit_id : delayed_ids) {
DEBUG_PRINTF("delayed lit id %u\n", delayed_lit_id);
assert(contains(prog_build.delay_programs, delayed_lit_id));
u32 delay_id = prog_build.delay_programs.at(delayed_lit_id);
const auto &delay_lit = literals.at(delayed_lit_id);
delay_instructions.emplace_back(verify_u8(delay_lit.delay), delay_id);
}
sort_and_unique(delay_instructions, [](const RoseInstrPushDelayed &a,
const RoseInstrPushDelayed &b) {
return tie(a.delay, a.index) < tie(b.delay, b.index);
});
for (const auto &ri : delay_instructions) {
program.add_before_end(make_unique<RoseInstrPushDelayed>(ri));
}
}
static
void makeCheckLiteralInstruction(const rose_literal_id &lit,
size_t longLitLengthThreshold,
RoseProgram &program,
const CompileContext &cc) {
assert(longLitLengthThreshold > 0);
DEBUG_PRINTF("lit=%s, long lit threshold %zu\n", dumpString(lit.s).c_str(),
longLitLengthThreshold);
if (lit.s.length() <= ROSE_SHORT_LITERAL_LEN_MAX) {
DEBUG_PRINTF("lit short enough to not need confirm\n");
return;
}
// Check resource limits as well.
if (lit.s.length() > cc.grey.limitLiteralLength) {
throw ResourceLimitError();
}
if (lit.s.length() <= longLitLengthThreshold) {
DEBUG_PRINTF("is a medium-length literal\n");
const auto *end_inst = program.end_instruction();
unique_ptr<RoseInstruction> ri;
if (lit.s.any_nocase()) {
ri = make_unique<RoseInstrCheckMedLitNocase>(lit.s.get_string(),
end_inst);
} else {
ri = make_unique<RoseInstrCheckMedLit>(lit.s.get_string(),
end_inst);
}
program.add_before_end(move(ri));
return;
}
// Long literal support should only really be used for the floating table
// in streaming mode.
assert(lit.table == ROSE_FLOATING && cc.streaming);
DEBUG_PRINTF("is a long literal\n");
const auto *end_inst = program.end_instruction();
unique_ptr<RoseInstruction> ri;
if (lit.s.any_nocase()) {
ri = make_unique<RoseInstrCheckLongLitNocase>(lit.s.get_string(),
end_inst);
} else {
ri = make_unique<RoseInstrCheckLongLit>(lit.s.get_string(), end_inst);
}
program.add_before_end(move(ri));
}
static
void makeRoleCheckNotHandled(ProgramBuild &prog_build, RoseVertex v,
RoseProgram &program) {
u32 handled_key;
if (contains(prog_build.handledKeys, v)) {
handled_key = prog_build.handledKeys.at(v);
} else {
handled_key = verify_u32(prog_build.handledKeys.size());
prog_build.handledKeys.emplace(v, handled_key);
}
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrCheckNotHandled>(handled_key, end_inst);
program.add_before_end(move(ri));
}
static
void makeRoleCheckBounds(const RoseBuildImpl &build, RoseVertex v,
const RoseEdge &e, RoseProgram &program) {
const RoseGraph &g = build.g;
const RoseVertex u = source(e, g);
// We know that we can trust the anchored table (DFA) to always deliver us
// literals at the correct offset.
if (build.isAnchored(v)) {
DEBUG_PRINTF("literal in anchored table, skipping bounds check\n");
return;
}
// Use the minimum literal length.
u32 lit_length = g[v].eod_accept ? 0 : verify_u32(build.minLiteralLen(v));
u64a min_bound = g[e].minBound + lit_length;
u64a max_bound = g[e].maxBound == ROSE_BOUND_INF
? ROSE_BOUND_INF
: g[e].maxBound + lit_length;
if (g[e].history == ROSE_ROLE_HISTORY_ANCH) {
assert(g[u].fixedOffset());
// Make offsets absolute.
min_bound += g[u].max_offset;
if (max_bound != ROSE_BOUND_INF) {
max_bound += g[u].max_offset;
}
}
assert(max_bound <= ROSE_BOUND_INF);
assert(min_bound <= max_bound);
// CHECK_BOUNDS instruction uses 64-bit bounds, so we can use MAX_OFFSET
// (max value of a u64a) to represent ROSE_BOUND_INF.
if (max_bound == ROSE_BOUND_INF) {
max_bound = MAX_OFFSET;
}
// This instruction should be doing _something_ -- bounds should be tighter
// than just {length, inf}.
assert(min_bound > lit_length || max_bound < MAX_OFFSET);
const auto *end_inst = program.end_instruction();
program.add_before_end(
make_unique<RoseInstrCheckBounds>(min_bound, max_bound, end_inst));
}
static
void makeRoleGroups(const RoseGraph &g, ProgramBuild &prog_build,
RoseVertex v, RoseProgram &program) {
rose_group groups = g[v].groups;
if (!groups) {
return;
}
// The set of "already on" groups as we process this vertex is the
// intersection of the groups set by our predecessors.
assert(in_degree(v, g) > 0);
rose_group already_on = ~rose_group{0};
for (const auto &u : inv_adjacent_vertices_range(v, g)) {
already_on &= prog_build.vertex_group_map.at(u);
}
DEBUG_PRINTF("already_on=0x%llx\n", already_on);
DEBUG_PRINTF("squashable=0x%llx\n", prog_build.squashable_groups);
DEBUG_PRINTF("groups=0x%llx\n", groups);
already_on &= ~prog_build.squashable_groups;
DEBUG_PRINTF("squashed already_on=0x%llx\n", already_on);
// We don't *have* to mask off the groups that we know are already on, but
// this will make bugs more apparent.
groups &= ~already_on;
if (!groups) {
DEBUG_PRINTF("no new groups to set, skipping\n");
return;
}
program.add_before_end(make_unique<RoseInstrSetGroups>(groups));
}
static
bool checkReachMask(const CharReach &cr, u8 &andmask, u8 &cmpmask) {
size_t reach_size = cr.count();
assert(reach_size > 0);
// check whether entry_size is some power of 2.
if ((reach_size - 1) & reach_size) {
return false;
}
make_and_cmp_mask(cr, &andmask, &cmpmask);
if ((1 << popcount32((u8)(~andmask))) ^ reach_size) {
return false;
}
return true;
}
static
bool checkReachWithFlip(const CharReach &cr, u8 &andmask,
u8 &cmpmask, u8 &flip) {
if (checkReachMask(cr, andmask, cmpmask)) {
flip = 0;
return true;
}
if (checkReachMask(~cr, andmask, cmpmask)) {
flip = 1;
return true;
}
return false;
}
static
bool makeRoleByte(const vector<LookEntry> &look, RoseProgram &program) {
if (look.size() == 1) {
const auto &entry = look[0];
u8 andmask_u8, cmpmask_u8;
u8 flip;
if (!checkReachWithFlip(entry.reach, andmask_u8, cmpmask_u8, flip)) {
return false;
}
s32 checkbyte_offset = verify_s32(entry.offset);
DEBUG_PRINTF("CHECK BYTE offset=%d\n", checkbyte_offset);
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrCheckByte>(andmask_u8, cmpmask_u8, flip,
checkbyte_offset, end_inst);
program.add_before_end(move(ri));
return true;
}
return false;
}
static
bool makeRoleMask(const vector<LookEntry> &look, RoseProgram &program) {
if (look.back().offset < look.front().offset + 8) {
s32 base_offset = verify_s32(look.front().offset);
u64a and_mask = 0;
u64a cmp_mask = 0;
u64a neg_mask = 0;
for (const auto &entry : look) {
u8 andmask_u8, cmpmask_u8, flip;
if (!checkReachWithFlip(entry.reach, andmask_u8,
cmpmask_u8, flip)) {
return false;
}
DEBUG_PRINTF("entry offset %d\n", entry.offset);
u32 shift = (entry.offset - base_offset) << 3;
and_mask |= (u64a)andmask_u8 << shift;
cmp_mask |= (u64a)cmpmask_u8 << shift;
if (flip) {
neg_mask |= 0xffLLU << shift;
}
}
DEBUG_PRINTF("CHECK MASK and_mask=%llx cmp_mask=%llx\n",
and_mask, cmp_mask);
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrCheckMask>(and_mask, cmp_mask, neg_mask,
base_offset, end_inst);
program.add_before_end(move(ri));
return true;
}
return false;
}
static UNUSED
string convertMaskstoString(u8 *p, int byte_len) {
string s;
for (int i = 0; i < byte_len; i++) {
u8 hi = *p >> 4;
u8 lo = *p & 0xf;
s += (char)(hi + (hi < 10 ? 48 : 87));
s += (char)(lo + (lo < 10 ? 48 : 87));
p++;
}
return s;
}
static
bool makeRoleMask32(const vector<LookEntry> &look,
RoseProgram &program) {
if (look.back().offset >= look.front().offset + 32) {
return false;
}
s32 base_offset = verify_s32(look.front().offset);
array<u8, 32> and_mask, cmp_mask;
and_mask.fill(0);
cmp_mask.fill(0);
u32 neg_mask = 0;
for (const auto &entry : look) {
u8 andmask_u8, cmpmask_u8, flip;
if (!checkReachWithFlip(entry.reach, andmask_u8,
cmpmask_u8, flip)) {
return false;
}
u32 shift = entry.offset - base_offset;
assert(shift < 32);
and_mask[shift] = andmask_u8;
cmp_mask[shift] = cmpmask_u8;
if (flip) {
neg_mask |= 1 << shift;
}
}
DEBUG_PRINTF("and_mask %s\n",
convertMaskstoString(and_mask.data(), 32).c_str());
DEBUG_PRINTF("cmp_mask %s\n",
convertMaskstoString(cmp_mask.data(), 32).c_str());
DEBUG_PRINTF("neg_mask %08x\n", neg_mask);
DEBUG_PRINTF("base_offset %d\n", base_offset);
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrCheckMask32>(and_mask, cmp_mask, neg_mask,
base_offset, end_inst);
program.add_before_end(move(ri));
return true;
}
static
bool makeRoleMask64(const vector<LookEntry> &look,
RoseProgram &program, const target_t &target) {
if (!target.has_avx512()) {
return false;
}
if (look.back().offset >= look.front().offset + 64) {
return false;
}
s32 base_offset = verify_s32(look.front().offset);
array<u8, 64> and_mask, cmp_mask;
and_mask.fill(0);
cmp_mask.fill(0);
u64a neg_mask = 0;
for (const auto &entry : look) {
u8 andmask_u8, cmpmask_u8, flip;
if (!checkReachWithFlip(entry.reach, andmask_u8, cmpmask_u8, flip)) {
return false;
}
u32 shift = entry.offset - base_offset;
assert(shift < 64);
and_mask[shift] = andmask_u8;
cmp_mask[shift] = cmpmask_u8;
if (flip) {
neg_mask |= 1ULL << shift;
}
}
DEBUG_PRINTF("and_mask %s\n",
convertMaskstoString(and_mask.data(), 64).c_str());
DEBUG_PRINTF("cmp_mask %s\n",
convertMaskstoString(cmp_mask.data(), 64).c_str());
DEBUG_PRINTF("neg_mask %llx\n", neg_mask);
DEBUG_PRINTF("base_offset %d\n", base_offset);
const auto *end_inst = program.end_instruction();
auto ri = make_unique<RoseInstrCheckMask64>(and_mask, cmp_mask, neg_mask,
base_offset, end_inst);
program.add_before_end(move(ri));
return true;
}
// Sorting by the size of every bucket.
// Used in map<u32, vector<s8>, cmpNibble>.
struct cmpNibble {
bool operator()(const u32 data1, const u32 data2) const{
u32 size1 = popcount32(data1 >> 16) * popcount32(data1 << 16);
u32 size2 = popcount32(data2 >> 16) * popcount32(data2 << 16);
return std::tie(size1, data1) < std::tie(size2, data2);
}
};
// Insert all pairs of bucket and offset into buckets.
static really_inline
void getAllBuckets(const vector<LookEntry> &look,
map<u32, vector<s8>, cmpNibble> &buckets, u64a &neg_mask) {
s32 base_offset = verify_s32(look.front().offset);
for (const auto &entry : look) {
CharReach cr = entry.reach;
// Flip heavy character classes to save buckets.
if (cr.count() > 128 ) {
cr.flip();
} else {
neg_mask ^= 1ULL << (entry.offset - base_offset);
}
map <u16, u16> lo2hi;
// We treat Ascii Table as a 16x16 grid.
// Push every row in cr into lo2hi and mark the row number.
for (size_t i = cr.find_first(); i != CharReach::npos;) {
u8 it_hi = i >> 4;
u16 low_encode = 0;
while (i != CharReach::npos && (i >> 4) == it_hi) {
low_encode |= 1 << (i & 0xf);
i = cr.find_next(i);
}
lo2hi[low_encode] |= 1 << it_hi;
}
for (const auto &it : lo2hi) {
u32 hi_lo = (it.second << 16) | it.first;
buckets[hi_lo].push_back(entry.offset);
}
}
}
// Once we have a new bucket, we'll try to combine it with all old buckets.
static really_inline
void nibUpdate(map<u32, u16> &nib, u32 hi_lo) {
u16 hi = hi_lo >> 16;
u16 lo = hi_lo & 0xffff;
for (const auto pairs : nib) {
u32 old = pairs.first;
if ((old >> 16) == hi || (old & 0xffff) == lo) {
if (!nib[old | hi_lo]) {
nib[old | hi_lo] = nib[old] | nib[hi_lo];
}
}
}
}
static really_inline
void nibMaskUpdate(array<u8, 32> &mask, u32 data, u8 bit_index) {
for (u8 index = 0; data > 0; data >>= 1, index++) {
if (data & 1) {
// 0 ~ 7 bucket in first 16 bytes,
// 8 ~ 15 bucket in second 16 bytes.
if (bit_index >= 8) {
mask[index + 16] |= 1 << (bit_index - 8);
} else {
mask[index] |= 1 << bit_index;
}
}
}
}
static
bool getShuftiMasks(const vector<LookEntry> &look, array<u8, 32> &hi_mask,
array<u8, 32> &lo_mask, u8 *bucket_select_hi,
u8 *bucket_select_lo, u64a &neg_mask,
u8 &bit_idx, size_t len) {
map<u32, u16> nib; // map every bucket to its bucket number.
map<u32, vector<s8>, cmpNibble> bucket2offsets;
s32 base_offset = look.front().offset;
bit_idx = 0;
neg_mask = ~0ULL;
getAllBuckets(look, bucket2offsets, neg_mask);
for (const auto &it : bucket2offsets) {
u32 hi_lo = it.first;
// New bucket.
if (!nib[hi_lo]) {
if ((bit_idx >= 8 && len == 64) || bit_idx >= 16) {
return false;
}
nib[hi_lo] = 1 << bit_idx;
nibUpdate(nib, hi_lo);
nibMaskUpdate(hi_mask, hi_lo >> 16, bit_idx);
nibMaskUpdate(lo_mask, hi_lo & 0xffff, bit_idx);
bit_idx++;
}
DEBUG_PRINTF("hi_lo %x bucket %x\n", hi_lo, nib[hi_lo]);
// Update bucket_select_mask.
u8 nib_hi = nib[hi_lo] >> 8;
u8 nib_lo = nib[hi_lo] & 0xff;
for (const auto offset : it.second) {
bucket_select_hi[offset - base_offset] |= nib_hi;
bucket_select_lo[offset - base_offset] |= nib_lo;
}
}
return true;
}
static
unique_ptr<RoseInstruction>
makeCheckShufti16x8(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 32> &bucket_select_mask,
u32 neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 16 || bucket_idx > 8) {
return nullptr;
}
array<u8, 32> nib_mask;
array<u8, 16> bucket_select_mask_16;
copy(lo_mask.begin(), lo_mask.begin() + 16, nib_mask.begin());
copy(hi_mask.begin(), hi_mask.begin() + 16, nib_mask.begin() + 16);
copy(bucket_select_mask.begin(), bucket_select_mask.begin() + 16,
bucket_select_mask_16.begin());
return make_unique<RoseInstrCheckShufti16x8>
(nib_mask, bucket_select_mask_16,
neg_mask & 0xffff, base_offset, end_inst);
}
static
unique_ptr<RoseInstruction>
makeCheckShufti32x8(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 32> &bucket_select_mask,
u32 neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 32 || bucket_idx > 8) {
return nullptr;
}
array<u8, 16> hi_mask_16;
array<u8, 16> lo_mask_16;
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_16.begin());
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_16.begin());
return make_unique<RoseInstrCheckShufti32x8>
(hi_mask_16, lo_mask_16, bucket_select_mask,
neg_mask, base_offset, end_inst);
}
static
unique_ptr<RoseInstruction>
makeCheckShufti16x16(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 32> &bucket_select_mask_lo,
const array<u8, 32> &bucket_select_mask_hi,
u32 neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 16 || bucket_idx > 16) {
return nullptr;
}
array<u8, 32> bucket_select_mask_32;
copy(bucket_select_mask_lo.begin(), bucket_select_mask_lo.begin() + 16,
bucket_select_mask_32.begin());
copy(bucket_select_mask_hi.begin(), bucket_select_mask_hi.begin() + 16,
bucket_select_mask_32.begin() + 16);
return make_unique<RoseInstrCheckShufti16x16>
(hi_mask, lo_mask, bucket_select_mask_32,
neg_mask & 0xffff, base_offset, end_inst);
}
static
unique_ptr<RoseInstruction>
makeCheckShufti32x16(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 32> &bucket_select_mask_lo,
const array<u8, 32> &bucket_select_mask_hi,
u32 neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 32 || bucket_idx > 16) {
return nullptr;
}
return make_unique<RoseInstrCheckShufti32x16>
(hi_mask, lo_mask, bucket_select_mask_hi,
bucket_select_mask_lo, neg_mask, base_offset, end_inst);
}
static
unique_ptr<RoseInstruction>
makeCheckShufti64x8(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 64> &bucket_select_mask,
u64a neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 64 || bucket_idx > 8) {
return nullptr;
}
array<u8, 64> hi_mask_64;
array<u8, 64> lo_mask_64;
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_64.begin());
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_64.begin() + 16);
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_64.begin() + 32);
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_64.begin() + 48);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_64.begin());
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_64.begin() + 16);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_64.begin() + 32);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_64.begin() + 48);
return make_unique<RoseInstrCheckShufti64x8>
(hi_mask_64, lo_mask_64, bucket_select_mask,
neg_mask, base_offset, end_inst);
}
static
unique_ptr<RoseInstruction>
makeCheckShufti64x16(u32 offset_range, u8 bucket_idx,
const array<u8, 32> &hi_mask, const array<u8, 32> &lo_mask,
const array<u8, 64> &bucket_select_mask_lo,
const array<u8, 64> &bucket_select_mask_hi,
u64a neg_mask, s32 base_offset,
const RoseInstruction *end_inst) {
if (offset_range > 64 || bucket_idx > 16) {
return nullptr;
}
array<u8, 64> hi_mask_1;
array<u8, 64> hi_mask_2;
array<u8, 64> lo_mask_1;
array<u8, 64> lo_mask_2;
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_1.begin());
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_1.begin() + 16);
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_1.begin() + 32);
copy(hi_mask.begin(), hi_mask.begin() + 16, hi_mask_1.begin() + 48);
copy(hi_mask.begin() + 16, hi_mask.begin() + 32, hi_mask_2.begin());
copy(hi_mask.begin() + 16, hi_mask.begin() + 32, hi_mask_2.begin() + 16);
copy(hi_mask.begin() + 16, hi_mask.begin() + 32, hi_mask_2.begin() + 32);
copy(hi_mask.begin() + 16, hi_mask.begin() + 32, hi_mask_2.begin() + 48);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_1.begin());
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_1.begin() + 16);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_1.begin() + 32);
copy(lo_mask.begin(), lo_mask.begin() + 16, lo_mask_1.begin() + 48);
copy(lo_mask.begin() + 16, lo_mask.begin() + 32, lo_mask_2.begin());
copy(lo_mask.begin() + 16, lo_mask.begin() + 32, lo_mask_2.begin() + 16);
copy(lo_mask.begin() + 16, lo_mask.begin() + 32, lo_mask_2.begin() + 32);
copy(lo_mask.begin() + 16, lo_mask.begin() + 32, lo_mask_2.begin() + 48);
return make_unique<RoseInstrCheckShufti64x16>
(hi_mask_1, hi_mask_2, lo_mask_1, lo_mask_2, bucket_select_mask_hi,
bucket_select_mask_lo, neg_mask, base_offset, end_inst);
}
static
bool makeRoleShufti(const vector<LookEntry> &look, RoseProgram &program,
const target_t &target) {
s32 offset_limit;
if (target.has_avx512()) {
offset_limit = 64;
} else {
offset_limit = 32;
}
s32 base_offset = verify_s32(look.front().offset);
if (look.back().offset >= base_offset + offset_limit) {
return false;
}
u8 bucket_idx = 0; // number of buckets
u64a neg_mask_64;
array<u8, 32> hi_mask;
array<u8, 32> lo_mask;
array<u8, 64> bucket_select_hi_64; // for AVX512
array<u8, 64> bucket_select_lo_64; // for AVX512
array<u8, 32> bucket_select_hi;
array<u8, 32> bucket_select_lo;
hi_mask.fill(0);
lo_mask.fill(0);
bucket_select_hi_64.fill(0);
bucket_select_lo_64.fill(0);
bucket_select_hi.fill(0); // will not be used in 16x8 and 32x8.
bucket_select_lo.fill(0);
if (target.has_avx512()) {
if (!getShuftiMasks(look, hi_mask, lo_mask, bucket_select_hi_64.data(),
bucket_select_lo_64.data(), neg_mask_64, bucket_idx,
32)) {
return false;
}
copy(bucket_select_hi_64.begin(), bucket_select_hi_64.begin() + 32,
bucket_select_hi.begin());
copy(bucket_select_lo_64.begin(), bucket_select_lo_64.begin() + 32,
bucket_select_lo.begin());
DEBUG_PRINTF("bucket_select_hi_64 %s\n",
convertMaskstoString(bucket_select_hi_64.data(), 64).c_str());
DEBUG_PRINTF("bucket_select_lo_64 %s\n",
convertMaskstoString(bucket_select_lo_64.data(), 64).c_str());
} else {
if (!getShuftiMasks(look, hi_mask, lo_mask, bucket_select_hi.data(),
bucket_select_lo.data(), neg_mask_64, bucket_idx,
32)) {
return false;
}
}
u32 neg_mask = (u32)neg_mask_64;
DEBUG_PRINTF("hi_mask %s\n",
convertMaskstoString(hi_mask.data(), 32).c_str());
DEBUG_PRINTF("lo_mask %s\n",
convertMaskstoString(lo_mask.data(), 32).c_str());
DEBUG_PRINTF("bucket_select_hi %s\n",
convertMaskstoString(bucket_select_hi.data(), 32).c_str());
DEBUG_PRINTF("bucket_select_lo %s\n",
convertMaskstoString(bucket_select_lo.data(), 32).c_str());
const auto *end_inst = program.end_instruction();
s32 offset_range = look.back().offset - base_offset + 1;
auto ri = makeCheckShufti16x8(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo, neg_mask, base_offset,
end_inst);
if (!ri) {
ri = makeCheckShufti32x8(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo, neg_mask, base_offset,
end_inst);
}
if (target.has_avx512()) {
if (!ri) {
ri = makeCheckShufti64x8(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo_64, neg_mask_64,
base_offset, end_inst);
}
}
if (!ri) {
ri = makeCheckShufti16x16(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo, bucket_select_hi,
neg_mask, base_offset, end_inst);
}
if (!ri) {
ri = makeCheckShufti32x16(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo, bucket_select_hi,
neg_mask, base_offset, end_inst);
}
if (target.has_avx512()) {
if (!ri) {
ri = makeCheckShufti64x16(offset_range, bucket_idx, hi_mask, lo_mask,
bucket_select_lo_64, bucket_select_hi_64,
neg_mask_64, base_offset, end_inst);
}
}
assert(ri);
program.add_before_end(move(ri));
return true;
}
/**
* Builds a lookaround instruction, or an appropriate specialization if one is
* available.
*/
static
void makeLookaroundInstruction(const vector<LookEntry> &look,
RoseProgram &program, const target_t &target) {
assert(!look.empty());
if (makeRoleByte(look, program)) {
return;
}
if (look.size() == 1) {
s8 offset = look.begin()->offset;
const CharReach &reach = look.begin()->reach;
auto ri = make_unique<RoseInstrCheckSingleLookaround>(offset, reach,
program.end_instruction());
program.add_before_end(move(ri));
return;
}
if (makeRoleMask(look, program)) {
return;
}
if (makeRoleMask32(look, program)) {
return;
}
if (makeRoleMask64(look, program, target)) {
return;
}
if (makeRoleShufti(look, program, target)) {
return;
}
auto ri = make_unique<RoseInstrCheckLookaround>(look,
program.end_instruction());
program.add_before_end(move(ri));
}
static
void makeCheckLitMaskInstruction(const RoseBuildImpl &build, u32 lit_id,
RoseProgram &program) {
const auto &info = build.literal_info.at(lit_id);
if (!info.requires_benefits) {
return;
}
vector<LookEntry> look;
const auto &lit = build.literals.at(lit_id);
const ue2_literal &s = lit.s;
const auto &msk = lit.msk;
DEBUG_PRINTF("building mask for lit %u: %s\n", lit_id,
dumpString(s).c_str());
assert(s.length() <= MAX_MASK2_WIDTH);
// Note: the literal matcher will confirm the HWLM mask in lit.msk, so we
// do not include those entries in the lookaround.
auto it = s.begin();
for (s32 i = 0 - s.length(), i_end = 0 - msk.size(); i < i_end; ++i, ++it) {
if (!it->nocase) {
look.emplace_back(verify_s8(i), *it);
}
}
if (look.empty()) {
return; // all caseful chars handled by HWLM mask.
}
makeLookaroundInstruction(look, program, build.cc.target_info);
}
static
void makeCheckLitEarlyInstruction(const RoseBuildImpl &build, u32 lit_id,
const vector<RoseEdge> &lit_edges,
u32 floatingMinLiteralMatchOffset,
RoseProgram &prog) {
if (lit_edges.empty()) {
return;
}
if (floatingMinLiteralMatchOffset == 0) {
return;
}
RoseVertex v = target(lit_edges.front(), build.g);
if (!build.isFloating(v)) {
return;
}
const auto &lit = build.literals.at(lit_id);
size_t min_len = lit.elength();
u32 min_offset = findMinOffset(build, lit_id);
DEBUG_PRINTF("has min_len=%zu, min_offset=%u, global min is %u\n", min_len,
min_offset, floatingMinLiteralMatchOffset);
// If we can't match before the min offset, we don't need the check.
if (min_len >= floatingMinLiteralMatchOffset) {
DEBUG_PRINTF("no need for check, min is %u\n",
floatingMinLiteralMatchOffset);
return;
}
assert(min_offset >= floatingMinLiteralMatchOffset);
assert(min_offset < UINT32_MAX);
DEBUG_PRINTF("adding lit early check, min_offset=%u\n", min_offset);
const auto *end = prog.end_instruction();
prog.add_before_end(make_unique<RoseInstrCheckLitEarly>(min_offset, end));
}
static
void makeGroupCheckInstruction(const RoseBuildImpl &build, u32 lit_id,
RoseProgram &prog) {
const auto &info = build.literal_info.at(lit_id);
if (!info.group_mask) {
return;
}
prog.add_before_end(make_unique<RoseInstrCheckGroups>(info.group_mask));
}
static
bool hasDelayedLiteral(const RoseBuildImpl &build,
const vector<RoseEdge> &lit_edges) {
auto is_delayed = [&build](u32 lit_id) { return build.isDelayed(lit_id); };
for (const auto &e : lit_edges) {
auto v = target(e, build.g);
const auto &lits = build.g[v].literals;
if (any_of(begin(lits), end(lits), is_delayed)) {
return true;
}
}
return false;
}
static
RoseProgram makeLitInitialProgram(const RoseBuildImpl &build,
ProgramBuild &prog_build, u32 lit_id,
const vector<RoseEdge> &lit_edges,
bool is_anchored_replay_program) {
RoseProgram program;
// Check long literal info.
if (!build.isDelayed(lit_id)) {
makeCheckLiteralInstruction(build.literals.at(lit_id),
prog_build.longLitLengthThreshold,
program, build.cc);
}
// Check lit mask.
makeCheckLitMaskInstruction(build, lit_id, program);
// Check literal groups. This is an optimisation that we only perform for
// delayed literals, as their groups may be switched off; ordinarily, we
// can trust the HWLM matcher.
if (hasDelayedLiteral(build, lit_edges)) {
makeGroupCheckInstruction(build, lit_id, program);
}
// Add instructions for pushing delayed matches, if there are any.
makePushDelayedInstructions(build.literals, prog_build,
build.literal_info.at(lit_id).delayed_ids,
program);
// Add pre-check for early literals in the floating table.
makeCheckLitEarlyInstruction(build, lit_id, lit_edges,
prog_build.floatingMinLiteralMatchOffset,
program);
/* Check if we are able to deliever matches from the anchored table now */
if (!is_anchored_replay_program) {
makeAnchoredLiteralDelay(build, prog_build, lit_id, program);
}
return program;
}
static
bool makeRoleMultipathShufti(const vector<vector<LookEntry>> &multi_look,
RoseProgram &program) {
if (multi_look.empty()) {
return false;
}
// find the base offset
assert(!multi_look[0].empty());
s32 base_offset = multi_look[0].front().offset;
s32 last_start = base_offset;
s32 end_offset = multi_look[0].back().offset;
size_t multi_len = 0;
for (const auto &look : multi_look) {
assert(look.size() > 0);
multi_len += look.size();
LIMIT_TO_AT_MOST(&base_offset, look.front().offset);
ENSURE_AT_LEAST(&last_start, look.front().offset);
ENSURE_AT_LEAST(&end_offset, look.back().offset);
}
assert(last_start < 0);
if (end_offset - base_offset >= MULTIPATH_MAX_LEN) {
return false;
}
if (multi_len <= 16) {
multi_len = 16;
} else if (multi_len <= 32) {
multi_len = 32;
} else if (multi_len <= 64) {
multi_len = 64;
} else {
DEBUG_PRINTF("too long for multi-path\n");
return false;
}
vector<LookEntry> linear_look;
array<u8, 64> data_select_mask;
data_select_mask.fill(0);
u64a hi_bits_mask = 0;
u64a lo_bits_mask = 0;
for (const auto &look : multi_look) {
assert(linear_look.size() < 64);
lo_bits_mask |= 1LLU << linear_look.size();
for (const auto &entry : look) {
assert(entry.offset - base_offset < MULTIPATH_MAX_LEN);
data_select_mask[linear_look.size()] =
verify_u8(entry.offset - base_offset);
linear_look.emplace_back(verify_s8(linear_look.size()), entry.reach);
}
hi_bits_mask |= 1LLU << (linear_look.size() - 1);
}
u8 bit_index = 0; // number of buckets
u64a neg_mask;
array<u8, 32> hi_mask;
array<u8, 32> lo_mask;
array<u8, 64> bucket_select_hi;
array<u8, 64> bucket_select_lo;
hi_mask.fill(0);
lo_mask.fill(0);
bucket_select_hi.fill(0);
bucket_select_lo.fill(0);
if (!getShuftiMasks(linear_look, hi_mask, lo_mask, bucket_select_hi.data(),
bucket_select_lo.data(), neg_mask, bit_index,
multi_len)) {
return false;
}
DEBUG_PRINTF("hi_mask %s\n",
convertMaskstoString(hi_mask.data(), 16).c_str());
DEBUG_PRINTF("lo_mask %s\n",
convertMaskstoString(lo_mask.data(), 16).c_str());
DEBUG_PRINTF("bucket_select_hi %s\n",
convertMaskstoString(bucket_select_hi.data(), 64).c_str());
DEBUG_PRINTF("bucket_select_lo %s\n",
convertMaskstoString(bucket_select_lo.data(), 64).c_str());
DEBUG_PRINTF("data_select_mask %s\n",
convertMaskstoString(data_select_mask.data(), 64).c_str());
DEBUG_PRINTF("hi_bits_mask %llx\n", hi_bits_mask);
DEBUG_PRINTF("lo_bits_mask %llx\n", lo_bits_mask);
DEBUG_PRINTF("neg_mask %llx\n", neg_mask);
DEBUG_PRINTF("base_offset %d\n", base_offset);
DEBUG_PRINTF("last_start %d\n", last_start);
// Since we don't have 16x16 now, just call 32x16 instead.
if (bit_index > 8) {
assert(multi_len <= 32);
multi_len = 32;
}
const auto *end_inst = program.end_instruction();
assert(multi_len == 16 || multi_len == 32 || multi_len == 64);
if (multi_len == 16) {
neg_mask &= 0xffff;
assert(!(hi_bits_mask & ~0xffffULL));
assert(!(lo_bits_mask & ~0xffffULL));
assert(bit_index <=8);
array<u8, 32> nib_mask;
copy(begin(lo_mask), begin(lo_mask) + 16, nib_mask.begin());
copy(begin(hi_mask), begin(hi_mask) + 16, nib_mask.begin() + 16);
auto ri = make_unique<RoseInstrCheckMultipathShufti16x8>
(nib_mask, bucket_select_lo, data_select_mask, hi_bits_mask,
lo_bits_mask, neg_mask, base_offset, last_start, end_inst);
program.add_before_end(move(ri));
} else if (multi_len == 32) {
neg_mask &= 0xffffffff;
assert(!(hi_bits_mask & ~0xffffffffULL));
assert(!(lo_bits_mask & ~0xffffffffULL));
if (bit_index <= 8) {
auto ri = make_unique<RoseInstrCheckMultipathShufti32x8>
(hi_mask, lo_mask, bucket_select_lo, data_select_mask,
hi_bits_mask, lo_bits_mask, neg_mask, base_offset,
last_start, end_inst);
program.add_before_end(move(ri));
} else {
auto ri = make_unique<RoseInstrCheckMultipathShufti32x16>
(hi_mask, lo_mask, bucket_select_hi, bucket_select_lo,
data_select_mask, hi_bits_mask, lo_bits_mask, neg_mask,
base_offset, last_start, end_inst);
program.add_before_end(move(ri));
}
} else {
auto ri = make_unique<RoseInstrCheckMultipathShufti64>
(hi_mask, lo_mask, bucket_select_lo, data_select_mask,
hi_bits_mask, lo_bits_mask, neg_mask, base_offset,
last_start, end_inst);
program.add_before_end(move(ri));
}
return true;
}
static
void makeRoleMultipathLookaround(const vector<vector<LookEntry>> &multi_look,
RoseProgram &program) {
assert(!multi_look.empty());
assert(multi_look.size() <= MAX_LOOKAROUND_PATHS);
vector<vector<LookEntry>> ordered_look;
set<s32> look_offset;
assert(!multi_look[0].empty());
s32 last_start = multi_look[0][0].offset;
// build offset table.
for (const auto &look : multi_look) {
assert(look.size() > 0);
last_start = max(last_start, (s32)look.begin()->offset);
for (const auto &t : look) {
look_offset.insert(t.offset);
}
}
array<u8, MULTIPATH_MAX_LEN> start_mask;
if (multi_look.size() < MAX_LOOKAROUND_PATHS) {
start_mask.fill((1 << multi_look.size()) - 1);
} else {
start_mask.fill(0xff);
}
u32 path_idx = 0;
for (const auto &look : multi_look) {
for (const auto &t : look) {
assert(t.offset >= (int)*look_offset.begin());
size_t update_offset = t.offset - *look_offset.begin() + 1;
if (update_offset < start_mask.size()) {
start_mask[update_offset] &= ~(1 << path_idx);
}
}
path_idx++;
}
for (u32 i = 1; i < MULTIPATH_MAX_LEN; i++) {
start_mask[i] &= start_mask[i - 1];
DEBUG_PRINTF("start_mask[%u] = %x\n", i, start_mask[i]);
}
assert(look_offset.size() <= MULTIPATH_MAX_LEN);
assert(last_start < 0);
for (const auto &offset : look_offset) {
vector<LookEntry> multi_entry;
multi_entry.resize(MAX_LOOKAROUND_PATHS);
for (size_t i = 0; i < multi_look.size(); i++) {
for (const auto &t : multi_look[i]) {
if (t.offset == offset) {
multi_entry[i] = t;
}
}
}
ordered_look.emplace_back(multi_entry);
}
auto ri = make_unique<RoseInstrMultipathLookaround>(move(ordered_look),
last_start, start_mask,
program.end_instruction());
program.add_before_end(move(ri));
}
static
void makeRoleLookaround(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
RoseVertex v, RoseProgram &program) {
if (!build.cc.grey.roseLookaroundMasks) {
return;
}
vector<vector<LookEntry>> looks;
// Lookaround from leftfix (mandatory).
if (contains(leftfix_info, v) && leftfix_info.at(v).has_lookaround) {
DEBUG_PRINTF("using leftfix lookaround\n");
looks = leftfix_info.at(v).lookaround;
}
// We may be able to find more lookaround info (advisory) and merge it
// in.
if (looks.size() <= 1) {
vector<LookEntry> look;
vector<LookEntry> look_more;
if (!looks.empty()) {
look = move(looks.front());
}
findLookaroundMasks(build, v, look_more);
mergeLookaround(look, look_more);
if (!look.empty()) {
makeLookaroundInstruction(look, program, build.cc.target_info);
}
return;
}
if (!makeRoleMultipathShufti(looks, program)) {
assert(looks.size() <= 8);
makeRoleMultipathLookaround(looks, program);
}
}
static
void makeRoleSuffix(const RoseBuildImpl &build,
const map<suffix_id, u32> &suffixes,
const map<u32, engine_info> &engine_info_by_queue,
RoseVertex v, RoseProgram &prog) {
const auto &g = build.g;
if (!g[v].suffix) {
return;
}
assert(contains(suffixes, g[v].suffix));
u32 queue = suffixes.at(g[v].suffix);
u32 event;
assert(contains(engine_info_by_queue, queue));
const auto eng_info = engine_info_by_queue.at(queue);
if (isContainerType(eng_info.type)) {
auto tamaProto = g[v].suffix.tamarama.get();
assert(tamaProto);
event = (u32)MQE_TOP_FIRST +
tamaProto->top_remap.at(make_pair(g[v].index,
g[v].suffix.top));
assert(event < MQE_INVALID);
} else if (isMultiTopType(eng_info.type)) {
assert(!g[v].suffix.haig);
event = (u32)MQE_TOP_FIRST + g[v].suffix.top;
assert(event < MQE_INVALID);
} else {
// DFAs/Puffs have no MQE_TOP_N support, so they get a classic TOP
// event.
assert(!g[v].suffix.graph || onlyOneTop(*g[v].suffix.graph));
event = MQE_TOP;
}
prog.add_before_end(make_unique<RoseInstrTriggerSuffix>(queue, event));
}
static
void addInfixTriggerInstructions(vector<TriggerInfo> triggers,
RoseProgram &prog) {
// Order, de-dupe and add instructions to the end of program.
sort_and_unique(triggers, [](const TriggerInfo &a, const TriggerInfo &b) {
return tie(a.cancel, a.queue, a.event) <
tie(b.cancel, b.queue, b.event);
});
for (const auto &ti : triggers) {
prog.add_before_end(
make_unique<RoseInstrTriggerInfix>(ti.cancel, ti.queue, ti.event));
}
}
static
void makeRoleInfixTriggers(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
const map<u32, engine_info> &engine_info_by_queue,
RoseVertex u, RoseProgram &program) {
const auto &g = build.g;
vector<TriggerInfo> triggers;
for (const auto &e : out_edges_range(u, g)) {
RoseVertex v = target(e, g);
if (!g[v].left) {
continue;
}
assert(contains(leftfix_info, v));
const left_build_info &lbi = leftfix_info.at(v);
if (lbi.has_lookaround) {
continue;
}
assert(contains(engine_info_by_queue, lbi.queue));
const auto &eng_info = engine_info_by_queue.at(lbi.queue);
// DFAs have no TOP_N support, so they get a classic MQE_TOP event.
u32 top;
if (isContainerType(eng_info.type)) {
auto tamaProto = g[v].left.tamarama.get();
assert(tamaProto);
top = MQE_TOP_FIRST + tamaProto->top_remap.at(
make_pair(g[v].index, g[e].rose_top));
assert(top < MQE_INVALID);
} else if (!isMultiTopType(eng_info.type)) {
assert(num_tops(g[v].left) == 1);
top = MQE_TOP;
} else {
top = MQE_TOP_FIRST + g[e].rose_top;
assert(top < MQE_INVALID);
}
triggers.emplace_back(g[e].rose_cancel_prev_top, lbi.queue, top);
}
addInfixTriggerInstructions(move(triggers), program);
}
/**
* \brief True if the given vertex is a role that can only be switched on at
* EOD.
*/
static
bool onlyAtEod(const RoseBuildImpl &tbi, RoseVertex v) {
const RoseGraph &g = tbi.g;
// All such roles have only (0,0) edges to vertices with the eod_accept
// property, and no other effects (suffixes, ordinary reports, etc, etc).
if (isLeafNode(v, g) || !g[v].reports.empty() || g[v].suffix) {
return false;
}
for (const auto &e : out_edges_range(v, g)) {
RoseVertex w = target(e, g);
if (!g[w].eod_accept) {
return false;
}
assert(!g[w].reports.empty());
assert(g[w].literals.empty());
if (g[e].minBound || g[e].maxBound) {
return false;
}
}
/* There is no pointing enforcing this check at runtime if
* this role is only fired by the eod event literal */
if (tbi.eod_event_literal_id != MO_INVALID_IDX &&
g[v].literals.size() == 1 &&
*g[v].literals.begin() == tbi.eod_event_literal_id) {
return false;
}
return true;
}
static
void addCheckOnlyEodInstruction(RoseProgram &prog) {
DEBUG_PRINTF("only at eod\n");
const auto *end_inst = prog.end_instruction();
prog.add_before_end(make_unique<RoseInstrCheckOnlyEod>(end_inst));
}
static
void makeRoleEagerEodReports(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
bool needs_catchup, RoseVertex v,
RoseProgram &program) {
RoseProgram eod_program;
for (const auto &e : out_edges_range(v, build.g)) {
if (canEagerlyReportAtEod(build, e)) {
RoseProgram block;
makeRoleReports(build, leftfix_info, needs_catchup,
target(e, build.g), block);
eod_program.add_block(move(block));
}
}
if (eod_program.empty()) {
return;
}
if (!onlyAtEod(build, v)) {
// The rest of our program wasn't EOD anchored, so we need to guard
// these reports with a check.
addCheckOnlyEodInstruction(program);
}
program.add_before_end(move(eod_program));
}
/** Makes a program for a role/vertex given a specific pred/in_edge. */
static
RoseProgram makeRoleProgram(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
const map<suffix_id, u32> &suffixes,
const map<u32, engine_info> &engine_info_by_queue,
const unordered_map<RoseVertex, u32> &roleStateIndices,
ProgramBuild &prog_build, const RoseEdge &e) {
const RoseGraph &g = build.g;
auto v = target(e, g);
RoseProgram program;
// First, add program instructions that enforce preconditions without
// effects.
if (onlyAtEod(build, v)) {
addCheckOnlyEodInstruction(program);
}
if (g[e].history == ROSE_ROLE_HISTORY_ANCH) {
makeRoleCheckBounds(build, v, e, program);
}
// This role program may be triggered by different predecessors, with
// different offset bounds. We must ensure we put this check/set operation
// after the bounds check to deal with this case.
if (in_degree(v, g) > 1) {
assert(!build.isRootSuccessor(v));
makeRoleCheckNotHandled(prog_build, v, program);
}
makeRoleLookaround(build, leftfix_info, v, program);
makeRoleCheckLeftfix(build, leftfix_info, v, program);
// Next, we can add program instructions that have effects. This must be
// done as a series of blocks, as some of them (like reports) are
// escapable.
RoseProgram effects_block;
RoseProgram reports_block;
makeRoleReports(build, leftfix_info, prog_build.needs_catchup, v,
reports_block);
effects_block.add_block(move(reports_block));
RoseProgram infix_block;
makeRoleInfixTriggers(build, leftfix_info, engine_info_by_queue, v,
infix_block);
effects_block.add_block(move(infix_block));
// Note: SET_GROUPS instruction must be after infix triggers, as an infix
// going dead may switch off groups.
RoseProgram groups_block;
makeRoleGroups(build.g, prog_build, v, groups_block);
effects_block.add_block(move(groups_block));
RoseProgram suffix_block;
makeRoleSuffix(build, suffixes, engine_info_by_queue, v, suffix_block);
effects_block.add_block(move(suffix_block));
RoseProgram state_block;
makeRoleSetState(roleStateIndices, v, state_block);
effects_block.add_block(move(state_block));
// Note: EOD eager reports may generate a CHECK_ONLY_EOD instruction (if
// the program doesn't have one already).
RoseProgram eod_block;
makeRoleEagerEodReports(build, leftfix_info, prog_build.needs_catchup, v,
eod_block);
effects_block.add_block(move(eod_block));
/* a 'ghost role' may do nothing if we know that its groups are already set
* - in this case we can avoid producing a program at all. */
if (effects_block.empty()) {
return {};
}
program.add_before_end(move(effects_block));
return program;
}
static
void makeGroupSquashInstruction(const RoseBuildImpl &build, u32 lit_id,
RoseProgram &prog) {
const auto &info = build.literal_info.at(lit_id);
if (!info.squash_group) {
return;
}
DEBUG_PRINTF("squashes 0x%llx\n", info.group_mask);
assert(info.group_mask);
/* Note: group_mask is negated. */
prog.add_before_end(make_unique<RoseInstrSquashGroups>(~info.group_mask));
}
namespace {
struct ProgKey {
ProgKey(const RoseProgram &p) : prog(&p) {}
bool operator==(const ProgKey &b) const {
return RoseProgramEquivalence()(*prog, *b.prog);
}
size_t hash() const {
return RoseProgramHash()(*prog);
}
private:
const RoseProgram *prog;
};
}
RoseProgram assembleProgramBlocks(vector<RoseProgram> &&blocks_in) {
DEBUG_PRINTF("%zu blocks before dedupe\n", blocks_in.size());
vector<RoseProgram> blocks;
blocks.reserve(blocks_in.size()); /* to ensure stable reference for seen */
ue2_unordered_set<ProgKey> seen;
for (auto &block : blocks_in) {
if (contains(seen, block)) {
continue;
}
blocks.push_back(move(block));
seen.emplace(blocks.back());
}
DEBUG_PRINTF("%zu blocks after dedupe\n", blocks.size());
RoseProgram prog;
for (auto &block : blocks) {
/* If we have multiple blocks from different literals and any of them
* squash groups, we will have to add a CLEAR_WORK_DONE instruction to
* each literal program block to clear the work_done flags so that it's
* only set if a state has been. */
if (!prog.empty() && reads_work_done_flag(block)) {
RoseProgram clear_block;
clear_block.add_before_end(make_unique<RoseInstrClearWorkDone>());
prog.add_block(move(clear_block));
}
prog.add_block(move(block));
}
return prog;
}
RoseProgram makeLiteralProgram(const RoseBuildImpl &build,
const map<RoseVertex, left_build_info> &leftfix_info,
const map<suffix_id, u32> &suffixes,
const map<u32, engine_info> &engine_info_by_queue,
const unordered_map<RoseVertex, u32> &roleStateIndices,
ProgramBuild &prog_build, u32 lit_id,
const vector<RoseEdge> &lit_edges,
bool is_anchored_replay_program) {
const auto &g = build.g;
DEBUG_PRINTF("lit id=%u, %zu lit edges\n", lit_id, lit_edges.size());
// Construct initial program up front, as its early checks must be able
// to jump to end and terminate processing for this literal.
auto lit_program = makeLitInitialProgram(build, prog_build, lit_id,
lit_edges,
is_anchored_replay_program);
RoseProgram role_programs;
// Predecessor state id -> program block.
map<u32, RoseProgram> pred_blocks;
// Construct sparse iter sub-programs.
for (const auto &e : lit_edges) {
const auto &u = source(e, g);
if (build.isAnyStart(u)) {
continue; // Root roles are not handled with sparse iterator.
}
DEBUG_PRINTF("sparse iter edge (%zu,%zu)\n", g[u].index,
g[target(e, g)].index);
assert(contains(roleStateIndices, u));
u32 pred_state = roleStateIndices.at(u);
auto role_prog = makeRoleProgram(build, leftfix_info, suffixes,
engine_info_by_queue, roleStateIndices,
prog_build, e);
if (!role_prog.empty()) {
pred_blocks[pred_state].add_block(move(role_prog));
}
}
// Add blocks to deal with non-root edges (triggered by sparse iterator or
// mmbit_isset checks).
addPredBlocks(pred_blocks, roleStateIndices.size(), role_programs);
// Add blocks to handle root roles.
for (const auto &e : lit_edges) {
const auto &u = source(e, g);
if (!build.isAnyStart(u)) {
continue;
}
DEBUG_PRINTF("root edge (%zu,%zu)\n", g[u].index,
g[target(e, g)].index);
auto role_prog = makeRoleProgram(build, leftfix_info, suffixes,
engine_info_by_queue, roleStateIndices,
prog_build, e);
role_programs.add_block(move(role_prog));
}
if (lit_id == build.eod_event_literal_id) {
/* Note: does not require the lit initial program */
assert(build.eod_event_literal_id != MO_INVALID_IDX);
return role_programs;
}
/* Instructions to run even if a role program bails out */
RoseProgram unconditional_block;
// Literal may squash groups.
makeGroupSquashInstruction(build, lit_id, unconditional_block);
role_programs.add_block(move(unconditional_block));
lit_program.add_before_end(move(role_programs));
return lit_program;
}
RoseProgram makeDelayRebuildProgram(const RoseBuildImpl &build,
ProgramBuild &prog_build,
const vector<u32> &lit_ids) {
assert(!lit_ids.empty());
assert(build.cc.streaming);
vector<RoseProgram> blocks;
for (const auto &lit_id : lit_ids) {
DEBUG_PRINTF("lit_id=%u\n", lit_id);
const auto &info = build.literal_info.at(lit_id);
if (info.delayed_ids.empty()) {
continue; // No delayed IDs, no work to do.
}
RoseProgram prog;
if (!build.isDelayed(lit_id)) {
makeCheckLiteralInstruction(build.literals.at(lit_id),
prog_build.longLitLengthThreshold, prog,
build.cc);
}
makeCheckLitMaskInstruction(build, lit_id, prog);
makePushDelayedInstructions(build.literals, prog_build,
build.literal_info.at(lit_id).delayed_ids,
prog);
blocks.push_back(move(prog));
}
return assembleProgramBlocks(move(blocks));
}
RoseProgram makeEodAnchorProgram(const RoseBuildImpl &build,
ProgramBuild &prog_build, const RoseEdge &e,
const bool multiple_preds) {
const RoseGraph &g = build.g;
const RoseVertex v = target(e, g);
RoseProgram program;
if (g[e].history == ROSE_ROLE_HISTORY_ANCH) {
makeRoleCheckBounds(build, v, e, program);
}
if (multiple_preds) {
// Only necessary when there is more than one pred.
makeRoleCheckNotHandled(prog_build, v, program);
}
makeCatchup(build.rm, prog_build.needs_catchup, g[v].reports, program);
const bool has_som = false;
RoseProgram report_block;
for (const auto &id : g[v].reports) {
makeReport(build, id, has_som, report_block);
}
program.add_before_end(move(report_block));
return program;
}
static
void makeCatchupMpv(const ReportManager &rm, bool needs_mpv_catchup,
ReportID id, RoseProgram &program) {
if (!needs_mpv_catchup) {
return;
}
const Report &report = rm.getReport(id);
if (report.type == INTERNAL_ROSE_CHAIN) {
return;
}
program.add_before_end(make_unique<RoseInstrCatchUpMpv>());
}
RoseProgram makeReportProgram(const RoseBuildImpl &build,
bool needs_mpv_catchup, ReportID id) {
RoseProgram prog;
makeCatchupMpv(build.rm, needs_mpv_catchup, id, prog);
const bool has_som = false;
makeReport(build, id, has_som, prog);
return prog;
}
RoseProgram makeBoundaryProgram(const RoseBuildImpl &build,
const set<ReportID> &reports) {
// Note: no CATCHUP instruction is necessary in the boundary case, as we
// should always be caught up (and may not even have the resources in
// scratch to support it).
const bool has_som = false;
RoseProgram prog;
for (const auto &id : reports) {
makeReport(build, id, has_som, prog);
}
return prog;
}
void addIncludedJumpProgram(RoseProgram &program, u32 child_offset,
u8 squash) {
RoseProgram block;
block.add_before_end(make_unique<RoseInstrIncludedJump>(child_offset,
squash));
program.add_block(move(block));
}
static
void addPredBlockSingle(u32 pred_state, RoseProgram &pred_block,
RoseProgram &program) {
// Prepend an instruction to check the pred state is on.
const auto *end_inst = pred_block.end_instruction();
pred_block.insert(begin(pred_block),
make_unique<RoseInstrCheckState>(pred_state, end_inst));
program.add_block(move(pred_block));
}
static
void addPredBlocksAny(map<u32, RoseProgram> &pred_blocks, u32 num_states,
RoseProgram &program) {
RoseProgram sparse_program;
vector<u32> keys;
for (const u32 &key : pred_blocks | map_keys) {
keys.push_back(key);
}
const RoseInstruction *end_inst = sparse_program.end_instruction();
auto ri = make_unique<RoseInstrSparseIterAny>(num_states, keys, end_inst);
sparse_program.add_before_end(move(ri));
RoseProgram &block = pred_blocks.begin()->second;
/* we no longer need the check handled instruction as all the pred-role
* blocks are being collapsed together */
stripCheckHandledInstruction(block);
sparse_program.add_before_end(move(block));
program.add_block(move(sparse_program));
}
static
void addPredBlocksMulti(map<u32, RoseProgram> &pred_blocks,
u32 num_states, RoseProgram &program) {
assert(!pred_blocks.empty());
RoseProgram sparse_program;
const RoseInstruction *end_inst = sparse_program.end_instruction();
vector<pair<u32, const RoseInstruction *>> jump_table;
// BEGIN instruction.
auto ri_begin = make_unique<RoseInstrSparseIterBegin>(num_states, end_inst);
RoseInstrSparseIterBegin *begin_inst = ri_begin.get();
sparse_program.add_before_end(move(ri_begin));
// NEXT instructions, one per pred program.
u32 prev_key = pred_blocks.begin()->first;
for (auto it = next(begin(pred_blocks)); it != end(pred_blocks); ++it) {
auto ri = make_unique<RoseInstrSparseIterNext>(prev_key, begin_inst,
end_inst);
sparse_program.add_before_end(move(ri));
prev_key = it->first;
}
// Splice in each pred program after its BEGIN/NEXT.
auto out_it = begin(sparse_program);
for (auto &m : pred_blocks) {
u32 key = m.first;
RoseProgram &flat_prog = m.second;
assert(!flat_prog.empty());
const size_t block_len = flat_prog.size() - 1; // without INSTR_END.
assert(dynamic_cast<const RoseInstrSparseIterBegin *>(out_it->get()) ||
dynamic_cast<const RoseInstrSparseIterNext *>(out_it->get()));
out_it = sparse_program.insert(++out_it, move(flat_prog));
// Jump table target for this key is the beginning of the block we just
// spliced in.
jump_table.emplace_back(key, out_it->get());
assert(distance(begin(sparse_program), out_it) + block_len <=
sparse_program.size());
advance(out_it, block_len);
}
// Write the jump table back into the SPARSE_ITER_BEGIN instruction.
begin_inst->jump_table = move(jump_table);
program.add_block(move(sparse_program));
}
void addPredBlocks(map<u32, RoseProgram> &pred_blocks, u32 num_states,
RoseProgram &program) {
// Trim empty blocks, if any exist.
for (auto it = pred_blocks.begin(); it != pred_blocks.end();) {
if (it->second.empty()) {
it = pred_blocks.erase(it);
} else {
++it;
}
}
const size_t num_preds = pred_blocks.size();
if (num_preds == 0) {
return;
}
if (num_preds == 1) {
const auto head = pred_blocks.begin();
addPredBlockSingle(head->first, head->second, program);
return;
}
// First, see if all our blocks are equivalent, in which case we can
// collapse them down into one.
const auto &blocks = pred_blocks | map_values;
if (all_of(begin(blocks), end(blocks), [&](const RoseProgram &block) {
return RoseProgramEquivalence()(*begin(blocks), block);
})) {
DEBUG_PRINTF("all blocks equiv\n");
addPredBlocksAny(pred_blocks, num_states, program);
return;
}
addPredBlocksMulti(pred_blocks, num_states, program);
}
void applyFinalSpecialisation(RoseProgram &program) {
assert(!program.empty());
assert(program.back().code() == ROSE_INSTR_END);
if (program.size() < 2) {
return;
}
/* Replace the second-to-last instruction (before END) with a one-shot
* specialisation if available. */
auto it = next(program.rbegin());
if (auto *ri = dynamic_cast<const RoseInstrReport *>(it->get())) {
DEBUG_PRINTF("replacing REPORT with FINAL_REPORT\n");
program.replace(it, make_unique<RoseInstrFinalReport>(
ri->onmatch, ri->offset_adjust));
}
}
void recordLongLiterals(vector<ue2_case_string> &longLiterals,
const RoseProgram &program) {
for (const auto &ri : program) {
if (const auto *ri_check =
dynamic_cast<const RoseInstrCheckLongLit *>(ri.get())) {
DEBUG_PRINTF("found CHECK_LONG_LIT for string '%s'\n",
escapeString(ri_check->literal).c_str());
longLiterals.emplace_back(ri_check->literal, false);
continue;
}
if (const auto *ri_check =
dynamic_cast<const RoseInstrCheckLongLitNocase *>(ri.get())) {
DEBUG_PRINTF("found CHECK_LONG_LIT_NOCASE for string '%s'\n",
escapeString(ri_check->literal).c_str());
longLiterals.emplace_back(ri_check->literal, true);
}
}
}
void recordResources(RoseResources &resources, const RoseProgram &program) {
for (const auto &ri : program) {
switch (ri->code()) {
case ROSE_INSTR_TRIGGER_SUFFIX:
resources.has_suffixes = true;
break;
case ROSE_INSTR_TRIGGER_INFIX:
case ROSE_INSTR_CHECK_INFIX:
case ROSE_INSTR_CHECK_PREFIX:
case ROSE_INSTR_SOM_LEFTFIX:
resources.has_leftfixes = true;
break;
case ROSE_INSTR_SET_STATE:
case ROSE_INSTR_CHECK_STATE:
case ROSE_INSTR_SPARSE_ITER_BEGIN:
case ROSE_INSTR_SPARSE_ITER_NEXT:
resources.has_states = true;
break;
case ROSE_INSTR_CHECK_GROUPS:
resources.checks_groups = true;
break;
case ROSE_INSTR_PUSH_DELAYED:
resources.has_lit_delay = true;
break;
case ROSE_INSTR_CHECK_LONG_LIT:
case ROSE_INSTR_CHECK_LONG_LIT_NOCASE:
resources.has_lit_check = true;
break;
default:
break;
}
}
}
} // namespace ue2
Reference in New Issue View Git Blame Copy Permalink