Skip to content

Define PmSamplingEngine::drainData outside the PerfWorks guard so Per… #151

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
CodingInAVan merged 1 commit into from
Aug 7, 2026
Merged

Define PmSamplingEngine::drainData outside the PerfWorks guard so Per… #151

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 19 additions & 14 deletions include/gpufl/backends/nvidia/engine/pm_sampling_engine.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,6 +108,25 @@ void PmSamplingEngine::onScopeStop(const char*) {
#endif
}

// Declared unconditionally in the header, so the vtable carries this slot in
// every build: the definition stays outside the GPUFL_HAS_PERFWORKS block and
// guards its body instead, or a PerfWorks-less link fails on the vtable.
void PmSamplingEngine::drainData() {
#if GPUFL_HAS_PERFWORKS
if (!running_) return;
// Half the buffer span: enough headroom that a late collector tick still
// lands before the hardware wraps.
const int64_t now = detail::GetTimestampNs();
const int64_t due = last_drain_ns_ + BufferSpanNs_() / 2;
if (last_drain_ns_ != 0 && now < due) return;

std::lock_guard<std::mutex> lk(pm_mu_);
if (!running_) return;
last_drain_ns_ = now;
DecodeAndEmit_();
#endif
}

std::vector<std::string> PmSamplingEngine::ResolveMetrics_() const {
if (!opts_.pm_sampling_metrics.empty()) return opts_.pm_sampling_metrics;
return OverviewMetrics();
Expand Down Expand Up @@ -421,20 +440,6 @@ int64_t PmSamplingEngine::BufferSpanNs_() const {
return samples * interval_us * 1000;
}

void PmSamplingEngine::drainData() {
if (!running_) return;
// Half the buffer span: enough headroom that a late collector tick still
// lands before the hardware wraps.
const int64_t now = detail::GetTimestampNs();
const int64_t due = last_drain_ns_ + BufferSpanNs_() / 2;
if (last_drain_ns_ != 0 && now < due) return;

std::lock_guard<std::mutex> lk(pm_mu_);
if (!running_) return;
last_drain_ns_ = now;
DecodeAndEmit_();
}

void PmSamplingEngine::DecodeAndEmit_() {
if (!pm_object_ || !host_object_ || counter_data_image_.empty()) return;

Expand Down
Loading