//===-- DecodedThread.cpp -------------------------------------------------===// // // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // See https://llvm.org/LICENSE.txt for license information. // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // //===----------------------------------------------------------------------===// #include "DecodedThread.h" #include "TraceCursorIntelPT.h" #include #include #include using namespace lldb; using namespace lldb_private; using namespace lldb_private::trace_intel_pt; using namespace llvm; char IntelPTError::ID; IntelPTError::IntelPTError(int libipt_error_code, lldb::addr_t address) : m_libipt_error_code(libipt_error_code), m_address(address) { assert(libipt_error_code < 0); } void IntelPTError::log(llvm::raw_ostream &OS) const { OS << pt_errstr(pt_errcode(m_libipt_error_code)); if (m_address != LLDB_INVALID_ADDRESS && m_address > 0) OS << formatv(": {0:x+16}", m_address); } bool DecodedThread::TSCRange::InRange(uint64_t item_index) const { return item_index >= first_item_index && item_index < first_item_index + items_count; } bool DecodedThread::NanosecondsRange::InRange(uint64_t item_index) const { return item_index >= first_item_index && item_index < first_item_index + items_count; } double DecodedThread::NanosecondsRange::GetInterpolatedTime( uint64_t item_index, uint64_t begin_of_time_nanos, const LinuxPerfZeroTscConversion &tsc_conversion) const { uint64_t items_since_last_tsc = item_index - first_item_index; auto interpolate = [&](uint64_t next_range_start_ns) { if (next_range_start_ns == nanos) { // If the resolution of the conversion formula is bad enough to consider // these two timestamps as equal, then we just increase the next one by 1 // for correction next_range_start_ns++; } long double item_duration = static_cast(items_count) / (next_range_start_ns - nanos); return (nanos - begin_of_time_nanos) + items_since_last_tsc * item_duration; }; if (!next_range) { // If this is the last TSC range, so we have to extrapolate. In this case, // we assume that each instruction took one TSC, which is what an // instruction would take if no parallelism is achieved and the frequency // multiplier is 1. return interpolate(tsc_conversion.ToNanos(tsc + items_count)); } if (items_count < (next_range->tsc - tsc)) { // If the numbers of items in this range is less than the total TSC duration // of this range, i.e. each instruction taking longer than 1 TSC, then we // can assume that something else happened between these TSCs (e.g. a // context switch, change to kernel, decoding errors, etc). In this case, we // also assume that each instruction took 1 TSC. A proper way to improve // this would be to analize the next events in the trace looking for context // switches or trace disablement events, but for now, as we only want an // approximation, we keep it simple. We are also guaranteed that the time in // nanos of the next range is different to the current one, just because of // the definition of a NanosecondsRange. return interpolate( std::min(tsc_conversion.ToNanos(tsc + items_count), next_range->nanos)); } // In this case, each item took less than 1 TSC, so some parallelism was // achieved, which is an indication that we didn't suffered of any kind of // interruption. return interpolate(next_range->nanos); } uint64_t DecodedThread::GetItemsCount() const { return m_item_data.size(); } lldb::addr_t DecodedThread::GetInstructionLoadAddress(uint64_t item_index) const { return std::get(m_item_data[item_index]); } lldb::addr_t DecodedThread::GetSyncPointOffsetByIndex(uint64_t item_index) const { return m_psb_offsets.find(item_index)->second; } ThreadSP DecodedThread::GetThread() { return m_thread_sp; } template DecodedThread::TraceItemStorage & DecodedThread::CreateNewTraceItem(lldb::TraceItemKind kind, Data &&data) { m_item_data.emplace_back(data); if (m_last_tsc) (*m_last_tsc)->second.items_count++; if (m_last_nanoseconds) (*m_last_nanoseconds)->second.items_count++; return m_item_data.back(); } void DecodedThread::NotifySyncPoint(lldb::addr_t psb_offset) { m_psb_offsets.try_emplace(GetItemsCount(), psb_offset); AppendEvent(lldb::eTraceEventSyncPoint); } void DecodedThread::NotifyTsc(TSC tsc) { if (m_last_tsc && (*m_last_tsc)->second.tsc == tsc) return; if (m_last_tsc) assert(tsc >= (*m_last_tsc)->second.tsc && "We can't have decreasing times"); m_last_tsc = m_tscs.emplace(GetItemsCount(), TSCRange{tsc, 0, GetItemsCount()}).first; if (m_tsc_conversion) { uint64_t nanos = m_tsc_conversion->ToNanos(tsc); if (!m_last_nanoseconds || (*m_last_nanoseconds)->second.nanos != nanos) { m_last_nanoseconds = m_nanoseconds .emplace(GetItemsCount(), NanosecondsRange{nanos, tsc, nullptr, 0, GetItemsCount()}) .first; if (*m_last_nanoseconds != m_nanoseconds.begin()) { auto prev_range = prev(*m_last_nanoseconds); prev_range->second.next_range = &(*m_last_nanoseconds)->second; } } } AppendEvent(lldb::eTraceEventHWClockTick); } void DecodedThread::NotifyCPU(lldb::cpu_id_t cpu_id) { if (!m_last_cpu || *m_last_cpu != cpu_id) { m_cpus.emplace(GetItemsCount(), cpu_id); m_last_cpu = cpu_id; AppendEvent(lldb::eTraceEventCPUChanged); } } lldb::cpu_id_t DecodedThread::GetCPUByIndex(uint64_t item_index) const { auto it = m_cpus.upper_bound(item_index); return it == m_cpus.begin() ? LLDB_INVALID_CPU_ID : prev(it)->second; } std::optional DecodedThread::GetTSCRangeByIndex(uint64_t item_index) const { auto next_it = m_tscs.upper_bound(item_index); if (next_it == m_tscs.begin()) return std::nullopt; return prev(next_it)->second; } std::optional DecodedThread::GetNanosecondsRangeByIndex(uint64_t item_index) { auto next_it = m_nanoseconds.upper_bound(item_index); if (next_it == m_nanoseconds.begin()) return std::nullopt; return prev(next_it)->second; } uint64_t DecodedThread::GetTotalInstructionCount() const { return m_insn_count; } void DecodedThread::AppendEvent(lldb::TraceEvent event) { CreateNewTraceItem(lldb::eTraceItemKindEvent, event); m_events_stats.RecordEvent(event); } void DecodedThread::AppendInstruction(const pt_insn &insn) { CreateNewTraceItem(lldb::eTraceItemKindInstruction, insn.ip); m_insn_count++; } void DecodedThread::AppendError(const IntelPTError &error) { CreateNewTraceItem(lldb::eTraceItemKindError, error.message()); m_error_stats.RecordError(/*fatal=*/false); } void DecodedThread::AppendCustomError(StringRef err, bool fatal) { CreateNewTraceItem(lldb::eTraceItemKindError, err.str()); m_error_stats.RecordError(fatal); } lldb::TraceEvent DecodedThread::GetEventByIndex(int item_index) const { return std::get(m_item_data[item_index]); } const DecodedThread::EventsStats &DecodedThread::GetEventsStats() const { return m_events_stats; } void DecodedThread::EventsStats::RecordEvent(lldb::TraceEvent event) { events_counts[event]++; total_count++; } uint64_t DecodedThread::ErrorStats::GetTotalCount() const { uint64_t total = 0; for (const auto &[kind, count] : libipt_errors) total += count; return total + other_errors + fatal_errors; } void DecodedThread::ErrorStats::RecordError(bool fatal) { if (fatal) fatal_errors++; else other_errors++; } void DecodedThread::ErrorStats::RecordError(int libipt_error_code) { libipt_errors[pt_errstr(pt_errcode(libipt_error_code))]++; } const DecodedThread::ErrorStats &DecodedThread::GetErrorStats() const { return m_error_stats; } lldb::TraceItemKind DecodedThread::GetItemKindByIndex(uint64_t item_index) const { return std::visit( llvm::makeVisitor( [](const std::string &) { return lldb::eTraceItemKindError; }, [](lldb::TraceEvent) { return lldb::eTraceItemKindEvent; }, [](lldb::addr_t) { return lldb::eTraceItemKindInstruction; }), m_item_data[item_index]); } llvm::StringRef DecodedThread::GetErrorByIndex(uint64_t item_index) const { if (item_index >= m_item_data.size()) return llvm::StringRef(); return std::get(m_item_data[item_index]); } DecodedThread::DecodedThread( ThreadSP thread_sp, const std::optional &tsc_conversion) : m_thread_sp(thread_sp), m_tsc_conversion(tsc_conversion) {} size_t DecodedThread::CalculateApproximateMemoryUsage() const { return sizeof(TraceItemStorage) * m_item_data.size() + (sizeof(uint64_t) + sizeof(TSC)) * m_tscs.size() + (sizeof(uint64_t) + sizeof(uint64_t)) * m_nanoseconds.size() + (sizeof(uint64_t) + sizeof(lldb::cpu_id_t)) * m_cpus.size(); }