Files
Jungfraujoch/tests/StillsGPUAnalysisTest.cpp
leonarski_f 4809c1622c
Build Packages / build:portable:linux-x86_64:nocuda (push) Successful in 16m43s
Build Packages / build:jfjoch:ubuntu2404:nocuda (push) Successful in 28m46s
Build Packages / build:jfjoch:rocky8:cuda-sls9 (push) Successful in 32m5s
Build Packages / build:jfjoch:rocky9:cuda (push) Successful in 39m44s
Build Packages / build:portable:macos-arm64:nocuda (push) Successful in 5m7s
Build Packages / Create release (push) Successful in 8s
Build Packages / Build documentation (push) Successful in 49s
Build Packages / build:portable:linux-x86_64:cuda (push) Successful in 22m30s
Build Packages / build:jfjoch:rocky8:nocuda (push) Successful in 31m54s
Build Packages / build:rugnux:linux-aarch64:cuda (push) Successful in 11m29s
Build Packages / build:portable:windows-x86_64:nocuda (push) Successful in 23m51s
Build Packages / build:jfjoch:ubuntu2204:nocuda (push) Successful in 26m7s
Build Packages / build:jfjoch:ubuntu2204:cuda (push) Successful in 29m38s
Build Packages / build:portable:windows-x86_64:cuda (push) Successful in 28m6s
Build Packages / HDF5 consumer tests (DIALS, XDS) (push) Successful in 29m44s
Build Packages / build:jfjoch:rocky9:nocuda (push) Successful in 33m22s
Build Packages / Generate python client (push) Successful in 47s
Build Packages / build:jfjoch:rocky9:cuda-sls9 (push) Successful in 34m6s
Build Packages / Unit tests (push) Successful in 1h2m47s
Build Packages / build:jfjoch:rocky8:cuda (push) Successful in 35m59s
Build Packages / build:jfjoch:ubuntu2404:cuda (push) Successful in 34m8s
Build Packages / Validation Battery (push) Successful in 38m31s
v1.0.0-rc.187 (#98)
* Rugnux: Refinement several times faster, one protocol on the CPU and the GPU, parts of it on the Apple GPU; its random draws are the same on every platform.
* Rugnux: Refinement follows common practice: the free set decides only the restraint weights and the displacement model (isotropic, TLS or anisotropic, by Hamilton's test), everything else is decided on the working set; lower R-free for placed and predicted models.
* Rugnux: Restraint dictionaries via `--restraints` (or among the inputs) and more built-in ligands; a model whose space group is a subgroup of the data's is refined in its own group; reflections plus a model select `--mode refine` on their own.
* Rugnux: `-C` fixes the crystal system on the axes as typed; `-S` alone fixes the space group only, the cell is still found and reduced; a space group or cell stored in the input file is ignored, with a line in the log.
* Rugnux: More reliable multi-sweep merging; `--mode scale` merges several stills runs as one; `_plot.txt` and the report name each input; a second crystal form is reported for stills given a cell.
* jfjoch_broker (experimental): Live Rugnux on FPGA detectors (JUNGFRAU) takes the same path as a DECTRIS stream, rotation and serial alike, model step included, with analysis threads per GPU; frames waiting for the analysis can be viewed; no host stage drops a frame - a source faster than the analysis is held at the card.
* jfjoch_broker: Summation of more than 256 frames per image is split between the card and the host (`summation_mode` is then `hybrid`), and live Rugnux runs on such collections; pedestal runs take a receiver of their own.
* jfjoch_broker (experimental): `stills_downsample` analyses 1 of every N serial images (chosen by a hash; the first 2000 all; the others written without analysis; reset to 1 by `/initialize`); `max_spot_count` left out means 250 for per-image indexing and up to 1000 for live stills; the live status gives the merge before the model work, the fitted resolution and the anomalous peak height of each unexplained blob.
* jfjoch_broker: Without a `pcie` entry the broker runs on a mock of the card instead of the HLS simulation; its spot, ROI and azimuthal results are zero then.
* Web frontend: Redesigned Rugnux live page; Vite 8, TypeScript 6, Mol* 5.13 and dependency security updates.
* Tools: `jfjoch_fpga_stills_test` measures live stills on the cards with a stored dataset; `jfjoch_fpga_test` loses the HLS-simulation option.
* Other: FFTW no longer fetched or linked (the CPU FFT indexer keeps the name `FFTW`); PCIe driver now GPL-2.0-or-later and third-party notices complete; several crash fixes.

Reviewed-on: #98
Co-authored-by: Filip Leonarski <filip.leonarski@psi.ch>
2026-11-19 20:49:28 +01:00

196 lines
7.8 KiB
C++

// SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
// SPDX-License-Identifier: GPL-3.0-only
#include <catch2/catch_all.hpp>
#include <atomic>
#include <functional>
#include <latch>
#include "../common/ImageBuffer.h"
#include "../common/time_utc.h"
#include "../image_analysis/structure_refinement/RugnuxModel.h"
#include "../receiver/StillsGPUAnalysis.h"
#include "../receiver/StillsSelection.h"
namespace {
struct TestSink : public StillsSink {
std::function<void(const StillsSetup &)> ready = [](const StillsSetup &) {};
std::function<void(DataMessage &)> analysed = [](DataMessage &) {};
std::function<void()> failed = [] {};
void StillsReady(const StillsSetup &setup) override { ready(setup); }
void SendAnalysedImage(DataMessage &message, const AzimuthalIntegrationProfile &) override { analysed(message); }
SpotFindingSettings AnalysisSettings() override { return DiffractionExperiment::DefaultDataProcessingSettings(); }
void StillsSetupFailed(const JFJochException &) override { failed(); }
};
DiffractionExperiment StillsExperiment() {
DiffractionExperiment x(DetJF(1));
x.ImagesPerTrigger(STILLS_PREPASS_IMAGES + 200).NumTriggers(1).DetectorDistance_mm(100).BeamX_pxl(500).BeamY_pxl(250)
.IncidentEnergy_keV(12.4);
return x;
}
// A frame after the pre-pass's sample (numbers from PREPASS_IMAGES on are never in it), in a slot of `buffer`.
std::shared_ptr<HeldImage> Frame(ImageBuffer &buffer, int64_t number) {
auto held = std::make_shared<HeldImage>();
held->slot = buffer.GetImageSlot();
REQUIRE(held->slot != nullptr);
held->message.number = STILLS_PREPASS_IMAGES + number;
held->message.image_collection_efficiency = 1.0f;
return held;
}
StartMessage Start(const DiffractionExperiment &x) {
StartMessage start;
x.FillMessage(start);
return start;
}
}
TEST_CASE("StillsGPUAnalysis_StartFailureReleasesThreads", "[StillsGPUAnalysis]") {
Logger logger(Catch::getResultCapture().getCurrentTestName());
const DiffractionExperiment x = StillsExperiment();
const PixelMask mask(x);
ImageBuffer buffer(64 * 1024 * 1024);
buffer.StartMeasurement(x);
RugnuxLiveStatus status;
std::atomic<int> failed = 0, sunk = 0;
TestSink sink;
sink.ready = [](const StillsSetup &) {
throw JFJochException(JFJochExceptionCategory::ZeroMQ, "Timeout on pushing start message");
};
sink.analysed = [&](DataMessage &) { ++sunk; };
sink.failed = [&] { ++failed; };
{
StillsGPUAnalysis gpu(x, mask, Start(x), nullptr, 4, 100, logger, sink, status);
for (int i = 0; i < 3; i++)
gpu.Hold(Frame(buffer, i));
REQUIRE_THROWS_AS(gpu.Finish(), JFJochException);
}
CHECK(failed == 1);
CHECK(sunk == 0);
CHECK(status.Get().state == RugnuxLiveState::Failed);
// Every slot the frames were held in is back.
CHECK(buffer.GetAvailSlots() == buffer.GetStatus().total_slots);
}
TEST_CASE("StillsGPUAnalysis_HoldIsBounded", "[StillsGPUAnalysis]") {
Logger logger(Catch::getResultCapture().getCurrentTestName());
const DiffractionExperiment x = StillsExperiment();
const PixelMask mask(x);
ImageBuffer buffer(64 * 1024 * 1024);
buffer.StartMeasurement(x);
REQUIRE(buffer.GetStatus().total_slots >= 4);
RugnuxLiveStatus status;
TestSink sink;
sink.ready = [](const StillsSetup &) { throw JFJochException(JFJochExceptionCategory::ZeroMQ, "no writer"); };
StillsGPUAnalysis gpu(x, mask, Start(x), nullptr, 2, 3, logger, sink, status);
// Past what the pre-pass holds, frames are still taken: only the image buffer bounds them.
for (int i = 0; i < 4; i++)
gpu.Hold(Frame(buffer, i));
CHECK(buffer.GetStatus().preparation_slots == 4);
CHECK(buffer.GetImageSlot(buffer.GetStatus().available_slots) == nullptr);
REQUIRE_THROWS(gpu.Finish());
CHECK(buffer.GetStatus().preparation_slots == 0);
}
TEST_CASE("StillsGPUAnalysis_FailingFrameDoesNotStopAnalysis", "[StillsGPUAnalysis]") {
Logger logger(Catch::getResultCapture().getCurrentTestName());
const DiffractionExperiment x = StillsExperiment();
const PixelMask mask(x);
ImageBuffer buffer(64 * 1024 * 1024);
buffer.StartMeasurement(x);
RugnuxLiveStatus status;
std::atomic<int> sunk = 0;
TestSink sink;
sink.analysed = [&](DataMessage &m) {
++sunk;
if (m.number == STILLS_PREPASS_IMAGES)
throw std::runtime_error("not a JFJochException");
};
StillsGPUAnalysis gpu(x, mask, Start(x), nullptr, 1, 100, logger, sink, status);
// The frames carry no image, so their analysis fails too - they are passed on all the same.
for (int i = 0; i < 3; i++) {
status.ImageReceived(std::chrono::steady_clock::now()); // as the receiver does before it holds a frame
gpu.Hold(Frame(buffer, i));
}
REQUIRE_NOTHROW(gpu.Finish());
CHECK(sunk == 3);
// The outcome waits for everything to be written, and comes after the latency.
CHECK(status.Get().state == RugnuxLiveState::Running);
CHECK(!status.Get().last_image_to_written_s);
status.Latency(SecondsSince(status.LastImage().value()));
gpu.ReportOutcome();
const auto s = status.Get();
CHECK(s.state == RugnuxLiveState::Failed);
CHECK(!s.message.empty());
REQUIRE(s.last_image_to_written_s);
REQUIRE(s.last_image_s);
CHECK(s.elapsed_s >= s.last_image_s.value() + s.last_image_to_written_s.value() - 1e-6);
REQUIRE(!s.steps.empty());
CHECK(s.steps.back().title == "Writing");
CHECK(s.steps.back().duration_s);
CHECK(buffer.GetStatus().preparation_slots == 0);
}
TEST_CASE("StillsGPUAnalysis_CancelCutsTheModelStep", "[StillsGPUAnalysis]") {
Logger logger(Catch::getResultCapture().getCurrentTestName());
const DiffractionExperiment x = StillsExperiment();
const PixelMask mask(x);
TestSink sink;
const auto run = [&](RugnuxLiveStatus &status, std::shared_ptr<const RugnuxModel> model) {
return std::make_unique<StillsGPUAnalysis>(x, mask, Start(x), nullptr, 1, 100, logger, sink, status, true,
std::move(model), 60.0);
};
RugnuxLiveStatus during;
auto a = run(during, std::make_shared<RugnuxModel>());
a->Cut("cancel");
CHECK(during.Get().cut == "cancel");
REQUIRE_NOTHROW(a->Finish());
// Nothing is left to cut once the model step has ended, and nothing runs without a model.
RugnuxLiveStatus after;
auto b = run(after, std::make_shared<RugnuxModel>());
b->Finish();
b->Cut("cancel");
CHECK(after.Get().cut.empty());
RugnuxLiveStatus no_model;
auto c = run(no_model, nullptr);
c->Cut("cancel");
CHECK(no_model.Get().cut.empty());
c->Finish();
}
TEST_CASE("StillsGPUAnalysis_FramesDuringThePrePassAreAnalysed", "[StillsGPUAnalysis]") {
Logger logger(Catch::getResultCapture().getCurrentTestName());
const DiffractionExperiment x = StillsExperiment();
const PixelMask mask(x);
ImageBuffer buffer(64 * 1024 * 1024);
buffer.StartMeasurement(x);
const int64_t frames = 3 + 5;
REQUIRE(buffer.GetStatus().total_slots >= frames);
std::atomic<int> sunk = 0;
std::latch offered(1);
RugnuxLiveStatus status;
TestSink sink;
// The pre-pass is not over until every frame has been offered.
sink.ready = [&](const StillsSetup &) { offered.wait(); };
sink.analysed = [&](DataMessage &) { ++sunk; };
StillsGPUAnalysis gpu(x, mask, Start(x), nullptr, 2, 3, logger, sink, status);
CHECK(status.Get().stills->images_needed == 3);
for (int i = 0; i < frames; i++)
gpu.Hold(Frame(buffer, i));
offered.count_down();
REQUIRE_NOTHROW(gpu.Finish());
CHECK(sunk == frames);
CHECK(buffer.GetStatus().preparation_slots == 0);
}