Add JFJochViewerROIList to the side panel: a list of the dataset's ROIs
with selection, add (box/circle/azimuthal), rename and delete, plus a
statistics readout (sum/mean/std/max/valid/masked/centre-of-mass) for the
selected ROI, taken from the analysis output for the current image. Edits
emit a full ROIDefinition, routed to the worker's SetROIDefinition.
Per-ROI statistics now live in this panel rather than the canvas labels;
the diffraction image's labels show only the ROI name, and the ad-hoc
ROIIntegrationCPU computation there is removed in favour of the analysis
pipeline. The result widget now reports std dev instead of variance.
The single-ROI scratch panel remains for now and will be retired once the
interactive canvas editing replaces it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The reading worker's experiment is the source of truth for ROIs. Add a
SetROIDefinition slot that updates curr_experiment.ROI(), rebuilds the
analysis ROI engine, mutates the dataset (so the canvas reflects the new
ROIs) and recomputes only the ROIs for the current image via RunROIOnly.
On image load when full re-analysis is off, ROIs are still computed via
AnalyzeROIOnly so the statistics stay current.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add three entry points so ROI statistics can be (re)computed without a
full re-analysis, in support of interactive ROI editing in the viewer:
- RebuildROI(): recreate the ROI engine after the ROI set changes (the
CudaStream is now kept as a member so the GPU engine can be rebuilt).
- AnalyzeROIOnly(): decompress + preprocess + ROI, skipping azimuthal
integration, spot finding and indexing (a new image when re-analysis
is off).
- RunROIOnly(): rerun only the ROI integration on the already-preprocessed
image (an interactive ROI move). A full Analyze() still computes ROIs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When ROI labels are shown, each ROI's label now also reports its sum,
max and pixel count for the current image. Rather than reimplementing the
accumulation, this reuses the existing ROIIntegrationCPU engine (the
software counterpart of the FPGA roi_calc), built from the experiment and
cached per dataset. A small adapter folds the viewer's gap sentinel
(INT32_MIN+1) onto the engine's masked sentinel (INT32_MIN) so masked and
saturated pixels are handled correctly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The full-ring annulus built the inner ring as a separate QPainterPath and
merged it with addPath, which stitched a spurious segment from (0,0) to
the ring. Build both concentric rings as subpaths of one path, each begun
with an explicit moveTo, so there is no connecting segment.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Azimuthal ROIs now render on the diffraction image as annular sectors
(or full-ring annuli), sampled through DiffractionGeometry::ResPhiToPxl
so the outline follows the ROI footprint, including wrap-around sectors.
Add two side-panel toggles (both default off): a translucent fill for
every ROI (helpful when outline colours clash with the image colour map,
and with many ROIs) and ROI name labels (constant on-screen size). Wired
side panel -> window -> diffraction image like the existing feature
toggles.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When a file with ROI definitions is opened (read into experiment.ROI()
by the HDF5 reader), the diffraction image now overlays the configured
box and circle ROIs as distinct colour-coded outlines, alongside the
existing resolution rings and spots. This is the first step of the
ROI-map-based multi-ROI canvas: showing the ROIs loaded from a file.
Azimuthal ROIs (wedge rendering) and per-ROI statistics from the bitmap
follow in subsequent steps.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Footer "Paul Scherrer Institute" now links to https://www.psi.ch
- Reference drawer gains a "Source code" entry linking out to the Gitea
repository (new nav-item href that opens in a new tab)
- Reference drawer gains a "License" entry rendering the nicely formatted
docs/LICENSE.md inline (generalises the changelog renderer into MarkdownDoc)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The framed Sphinx changelog reused MUI with the same colours, so it read as
a picture-in-picture. Bundle docs/CHANGELOG.md at build time (?raw) and render
it with react-markdown, themed to match the app, for a quick frame-less view.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The Documentation tab pointed at jungfraujoch.readthedocs.io, which lags
behind the installed version. Build the Sphinx docs into frontend/dist/docs
and serve them at /frontend/docs so they always match the package.
- make_doc.sh: accept an output dir (default public), absolutize it, run
from the script dir, and allow $PYTHON override
- package.json: add "docs" script building into dist/docs
- CMakeLists: run "npm run docs" in the frontend target so CPack packs the
docs via the existing dist/ install
- App.tsx: point the Documentation DocFrame at /frontend/docs/index.html
Also clean up the Sphinx build (911 -> 0 warnings):
- fix typo myst_heading_anchor -> myst_heading_anchors so cross-doc
#anchor links actually resolve
- suppress myst.header/myst.xref_missing (auto-generated OpenAPI client
docs only) and the harmless sphinx_material config.cache note
- drop a dangling "---" transition at the end of CPU_DATA_ANALYSIS.md
- add REPOSITORIES and the generated python_client model pages to toctrees
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The phi bounds were stored as plain floats internally but passed through
std::optional in the constructor, which obscured the intent. Take them
as float arguments defaulting to 0 instead. The convention is explicit:
phi_min == phi_max means the full ring (all angles), a sector wraps
across 0 when phi_min > phi_max, and otherwise phi_min <= phi_max. The
"both bounds or none" rule belongs at the API/file boundary
(OpenAPIConvert, reader), not in the core class, so it lives there now.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
In VDS mode the per-image ROI results (max/sum/sum_sq/npixel/x/y) are
written into the data files but were not exposed in the master, so a VDS
master surfaced no ROI statistics. Add virtual datasets under
/entry/roi/<name> in LinkToData_VDS, one group per ROI, mirroring how the
spot-finding and azimuthal-integration arrays are linked. Integrated and
legacy formats are unaffected (the results are already reachable there).
Extended the reader round-trip test to write real ROI results and check
they read back from the master for both VDS and integrated formats.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Relocate the ROI definitions and the roi_map bitmap from /entry/roi to a
separate /entry/roi_defs group in the master file. /entry/roi keeps its
original meaning (per-image ROI results only), so an older reader that
iterates /entry/roi is not disturbed by the bitmap/definition entries
(which would otherwise make it try to read /entry/roi/roi_map/max and
fail to open the file at all). Back-compatibility for already-deployed
viewers is the reason for the split.
JFJochHDF5Reader now reads the logical definitions into experiment.ROI()
and the roi_map bitmask (+ a name->bit index) into the dataset, for all
file formats (the master always carries them). Reading both lets a later
viewer either re-derive ROIs with the current geometry (logical) or show
the exact written footprint (bitmap).
Added a write/read round-trip test over VDS and integrated formats.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
AddROI now renders azimuthal ROIs in addition to box and circle ROIs:
inner/outer arcs at the Q (d) bounds and, for a sector, the two radial
edges at the phi bounds. The per-pixel phi comes from the same
DiffractionGeometry::Phi_rad used to assign ROIs, so the overlay matches
the ROI footprint exactly (including wrap-around sectors). Full-ring
azimuthal ROIs draw as two concentric arcs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Persist regions of interest so an acquisition's ROI layout can be
recovered from the master (_master.h5 / _process.h5):
- /entry/roi/roi_map: uint16 per-pixel bitmask (bit i == ROI i).
- /entry/roi/<name>: logical definition (type + geometry params, incl.
azimuthal phi sector) plus bit_index tying it to the bitmap plane.
The bitmap rides along in the StartMessage (CBOR, mirroring az_int_map)
and is filled both online (JFJochReceiver::SendStartMessage) and offline
(jfjoch_process). Definitions come from the already-transmitted rois.
Both are dataset-wide metadata, so they are written by the NXmx master
writer (new NXmx::ROI), not the per-image data-file plugin.
Documented the /entry/roi layout in docs/HDF5.md and the master-vs-data
writer convention in CLAUDE.md; added a CBOR roi_map round-trip test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Expose the optional phi_min_deg / phi_max_deg azimuthal-sector bounds
added to ROIAzimuthal through the wire formats. Both are optional, so
older clients and streams (Q-only ROIs) remain valid.
- jfjoch_api.yaml: optional phi_min_deg / phi_max_deg on roi_azimuthal;
regenerated the cpp-pistache server model and the TS frontend client.
- OpenAPIConvert: map phi both directions (set only when present).
- CBOR start message: emit phi_min/phi_max only for a sector; on decode,
absent phi means a full ring.
- Documented the new optional CBOR fields; extended the CBOR round-trip
test with a sector and a full-ring azimuthal ROI.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ROIAzimuthal can now be restricted to an azimuthal-angle sector in
addition to its Q/d range, enabling STXM-style directional ROIs. phi
bounds are optional (both-or-neither); absent means a full 360 ring, so
existing Q-only azimuthal ROIs are unchanged.
- ROIElement::CheckROI gains a phi_deg argument; box/circle ignore it.
- MarkROI gains an optional phi_map; ROIMap builds it from Phi_rad
alongside the resolution map only when azimuthal ROIs are present.
- phi bounds are normalized to [0,360) and wrap-around sectors are
supported (phi_min > phi_max).
- ROIConfigAzim carries phi_min/phi_max (kept trivially copyable for the
union; phi_min == phi_max means full ring).
No phi crosses the wire yet (CBOR/API wiring follows separately).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DetectorSelection rendered the info-table rows and the detector-select
menu items without keys; ROI's grid action cells likewise. Add stable
keys to silence the warnings.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the stack of toggle-revealed sections with a left navigation
drawer (collapsible, expanded by default) clipped under a persistent
status bar. Dashboard (plots + measurement statistics + spot-finding)
is the default view; every other panel is its own drawer entry, so one
thing shows at a time.
- Group entries under Acquisition / Expert / Reference subheaders;
ROI folded into On-the-fly processing.
- Pale-indigo drawer with a lime accent on the selected entry.
- Gray out JUNGFRAU calibration / FPGA status when their data is empty.
- API reference and documentation open in-app via iframe.
- Fixed footer pinned to the page bottom.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Generate Zod schemas alongside the typed client (src/client/zod.gen.ts),
picking up the OpenAPI field defaults. Enables runtime validation and
default-filling for form payloads.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Match the unsaved-changes dot: the Upload button uses the warning color
while the panel has unedited changes, primary otherwise.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Generalize the expert-panel boilerplate (container + title + widgets +
upload button) and surface real upload errors via hey-api.
- SettingsPanel: consistent padded Paper + Typography title (one place to
size it) + optional Upload button + result snackbar; PanelTitle and
UploadSnackbarView exported for panels with bespoke layouts (DetectorSettings
grid, ROI)
- useUpload(mutation): wraps a generated hey-api mutation, shows the server's
error message on failure (errorMessage handles 400 string / 500 error_message)
- "unsaved changes" dot on panel titles, derived from the existing
last-downloaded snapshot (dirty = !isEqual(value, lastDownloaded))
- fix PixelMask: TIFF upload was swallowing errors; now shows them
- config panels (FileWriter, ImageFormat, Detector, AzInt, DarkMask, Indexing,
Instrument, ZeroMQ, ROI, DetectorSelection) upload through the typed SDK
instead of ButtonWithSnackbar's raw fetch; status/display panels share the
same shell for consistent spacing
- ButtonWithSnackbar now only powers bodyless action buttons (pedestal/cancel/
initialize, start/trigger, deactivate)
Build (tsc + vite) passes; dev server transforms all modules.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace all class components with function components and move server-state
polling to @tanstack/react-query.
- add @tanstack/react-query; generate query helpers via the hey-api
@tanstack/react-query plugin (src/client/@tanstack)
- QueryClientProvider + one-time client.setConfig({ baseUrl: '' }) in index.tsx
- App polls statistics with useQuery(getStatisticsOptions, refetchInterval),
DataProcessingPlot uses getPreviewPlotOptions; manual setInterval polling gone
- memo() on presentational children; with TanStack structural sharing this
re-renders a child only when its own statistics slice changes
- PreviewImage stays imperative (binary JPEG -> object URL) using useEffect +
a ref for the object-URL lifecycle
- fix AzIntSettings correction checkboxes that mutated state in place (relied on
the poll re-render); they now use setState
- drop dead code uncovered during the port (unused upload/deactivate handlers
and imports in ImageFormatSettings, DetectorSettings, ROI)
Build (tsc + vite) passes; dev server transforms all entry modules.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The raw stacked-module layout check compared the image width (cols) against
RAW_MODULE_SIZE (the per-module pixel count, 524288) instead of
RAW_MODULE_COLS (1024), so a raw-geometry TIFF could never be accepted.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
PUT /config/user_mask.tiff only accepted 32-bit unsigned TIFF, so masks
exported by tools like PyFAI (8-bit) failed. Route the upload through the
universal ReadTIFF reader and let PixelMask take a CompressedImage directly:
it validates the 2D shape against the detector's converted/raw layouts,
binarizes any 8/16/32-bit integer image (non-zero == masked), and rejects
float/multi-channel images.
Also dedupe the TIFF readers: ReadTIFFFromString16 is now a thin wrapper over
ReadTIFF, and the now-unused ReadTIFFFromString32 is removed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Nothing in HDF5DataFile.cpp uses sys/stat.h (no stat/mkdir/S_* etc.); it relies
on <filesystem>. Removing the dead POSIX include also avoids a needless
non-portable header for the eventual MSVC build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
AppendImage wrote the CBOR byte-string length (major type 2, 8-byte length,
which is defined big-endian) via __builtin_bswap64 under #ifdef LITTLE_ENDIAN.
That is broken two ways off GCC/Clang+glibc: __builtin_bswap64 is not available
on MSVC, and LITTLE_ENDIAN is not defined there, so the #else path skipped the
swap and emitted a little-endian length on a little-endian host -> corrupt
frames.
Write the 8 length bytes most-significant-first directly. This is big-endian by
construction on any host, needs no byte-swap intrinsic and no endianness macro,
and is byte-for-byte identical to the previous output on little-endian hosts.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
M_PI is a POSIX <math.h> extension that MSVC does not define without
_USE_MATH_DEFINES. std::numbers::pi (introduced in the viewer guard commit) is
C++20, but CUDA here is compiled as C++17 (CMAKE_CUDA_STANDARD 17) and several
common/ headers are pulled into .cu device translation units, so std::numbers is
not available there.
Add common/JFJochMath.h with a dependency-free `constexpr double PI` that works
in host code (including MSVC), in CUDA device code, and under C++17/20, and use
it everywhere:
- common/ and image_analysis/ (incl. CUDA .cu): 78 M_PI occurrences, 22 files
- broker/OpenAPIConvert.cpp
- viewer/: the 5 files that used std::numbers::pi now use PI, for one consistent
convention across the codebase
Verified to build: JFJochImageAnalysis (incl. CUDA), jfjoch_viewer, JFJochBroker.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
IndexerThreadPool dispatches on DiffractionExperiment::GetIndexingAlgorithm(),
which already resolves Auto to a concrete algorithm (FFTW/FFT/FFBIDX) or None;
the pool has no policy to resolve Auto itself. The worker handled a stray Auto
with a dead branch and silently produced no result when the resolved algorithm
had no matching indexer built.
- Document on GetIndexingAlgorithm() that it never returns Auto.
- Throw a clear internal error at the pool boundary if Auto ever arrives.
- In the worker, replace the dead Auto branch with a loud failure for any
resolved algorithm that has no matching indexer (e.g. a GPU algorithm on a
host without a GPU), instead of returning no result silently.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
PlotAzintUnit used D_A while MultiLinePlotUnits used d_A for the same
d-spacing-in-angstrom unit. Align PlotAzintUnit to the lowercase crystallographic
convention already used by MultiLinePlotUnits. Generated Plot_unit_x::D_A is
left as-is (controlled by the OpenAPI generator's C++ enum casing).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
gmtime() returns a pointer to a shared static buffer, so concurrent
deserialization of CborUnixTime_tTag values was a data race. Decode into a local
struct tm via gmtime_r (gmtime_s on Windows) instead.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
images_sent was incremented right after image_pusher.SendImage(*loc), but the
ZeroCopyReturnValue overload was void and, for the TCP pusher, asynchronous: it
silently drops the image (releases the slot and returns) when there is no live
connection or the 2 s enqueue deadline expires. So images_sent over-counted on a
broken/slow writer connection and disagreed with the ACK-based GetImagesWritten().
Make SendImage(ZeroCopyReturnValue&) return whether the image was accepted
(enqueued/handed off) and only increment images_sent on success. The slot is
still released on the drop path. The authoritative delivered count remains
GetImagesWritten() (total_data_acked_ok for TCP). File/ZMQ pushers return true on
accept, preserving their previous always-counted behaviour.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The receiver already computes and stores the compression ratio per image
(JFJochReceiverPlots Add/GetPlots/GetPlotRaw all handle PlotType::CompressionRatio),
but the type was unreachable from the API: it was missing from ConvertPlotType,
the OpenAPI plot_type enum, and the TS client, so requesting it returned 500 and
no UI menu entry existed.
Wire it through all layers:
- broker: add the compression_ratio -> PlotType::CompressionRatio case in
ConvertPlotType.
- spec: add compression_ratio to the plot_type parameter enum.
- frontend: add COMPRESSION_RATIO to the generated plot_type enum (matches what
regeneration would produce), a "Compression ratio" menu entry, and a Y-axis
label.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Step toward building jfjoch_viewer on Windows/MSVC. No change to the Linux
build: D-Bus stays on by default and the XCB plugin is still used.
- CMake: make Qt6::DBus optional via JFJOCH_VIEWER_DBUS (ON on Linux, OFF on
Windows/macOS where Qt6::DBus does not exist); compile/link/install the dbus/
adaptor + service file only when enabled. Select the platform integration
plugin per-OS (QXcb on Linux, QWindows on Windows, QCocoa on macOS) instead
of hard-coding QXcbIntegrationPlugin.
- JFJochViewerWindow: wrap the adaptor include and D-Bus registration in
#ifdef JFJOCH_VIEWER_DBUS.
- JFJochImageReadingWorker: the POSIX open()/fstat()/NFS-errno preflight is now
#ifndef _WIN32, with a portable QFileInfo fallback elsewhere; POSIX-only
headers are guarded too.
- Replace the POSIX M_PI/M_PI_2 extension with C++20 std::numbers::pi across the
viewer (not defined by MSVC without _USE_MATH_DEFINES).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Only gen/model/ is compiled (into JFJochAPI) and tracked; the REST server is
hand-written on cpp-httplib, so the generated Pistache server stubs (api/,
impl/, main-api-server.cpp, CMakeLists.txt, README.md, .openapi-generator*) are
regeneration debris that was previously left untracked and tended to carry a
newer spec version than the committed models. Ignore it so it can't be
committed by accident.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a status-bar cluster, shown when connected over HTTP, that surfaces
the broker state and live acquisition info:
- broker state box (QProgressBar) with progress drawn as the bar fill,
polled ~1 Hz on its own timer independent of image sync
- Live/Disconnected connection badge (host:port in tooltip)
- "+N new" badge and effective live-rate (Hz) readout
All widgets are fixed-width and only blanked in place, so the bar never
reflows when things appear/disappear.
Add a third autoload mode (HTTPSyncDataset): manually selecting an image
while following live now freezes the displayed image but keeps the dataset,
plots and image count updating. Surfaced via a tri-state HTTP Sync button
(off / live / data-only).
Also fix GetBrokerStatus() dropping the message field and a couple of
copy-pasted exception strings in JFJochHttpReader.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Runs ROIIntegrationGPU and ROIIntegrationCPU on identical input and asserts
every per-ROI field (sum, sum_square, max, pixels, weighted centre, masked
count) matches bit-for-bit. Uses overlapping ROI boxes (multi-bit masks),
negative pixel values (signed weighted-sum path), and an injected saturated
and masked pixel per ROI to cover the "max only" and "fully excluded" branches.
Guarded by JFJOCH_USE_CUDA and skips with a warning when no CUDA GPU is present,
mirroring ImageSpotFinderGPUTest.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Three robustness fixes for the writer-facing TCP stream, addressing the
spurious "wrong number of images" / connection failures seen under load.
1. Never MSG_ZEROCOPY a transient buffer. The synchronous SendImage path
passes a caller-owned buffer with z == nullptr and reuses it for the next
frame immediately. With MSG_ZEROCOPY the kernel still references that buffer
after send() returns, so the peer could receive corrupted frames and drop
the connection mid-stream, truncating the run. Zero-copy is now gated on a
ZeroCopyReturnValue that keeps the buffer alive until completion.
2. Make the SendAll watchdog measure lack of progress, not total wall-clock.
The previous absolute deadline tore down a healthy but back-pressured
connection (slow/starved writer) after a fixed window; the watchdog now
resets on every byte actually sent, so only a genuinely stuck socket is
closed. Dead peers are still caught by OS keepalive and POLLHUP/POLLERR.
3. Make the END-ack wait progress-based (WaitForEndAck). The writer may still
be draining a backlog of DATA frames when END is sent; each DATA ACK is
progress, so the timeout only arms once the writer falls silent rather than
firing while images are legitimately still being drained.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
MXAnalysisWithoutFPGA never filled DataMessage.roi, so ROI integrals were
only available on the FPGA path. Add a software ROI engine that mirrors the
FPGA roi_calc kernel: per-ROI sum, sum of squares, good-pixel count, max and
intensity-weighted centre of mass, with each pixel carrying a 16-bit mask so
it can contribute to any subset of up to 16 ROIs.
New image_analysis/roi/ library (JFJochROIIntegration), structured like azint:
a base that precomputes the per-pixel mask and names, a templated CPU engine
(generic over pixel type for a future 16-bit path), and a GPU kernel using
per-block shared-memory atomics for the STXM case (half-detector ROIs).
Masked pixels are excluded entirely; saturated pixels are excluded from the
sums but still count towards the max, matching roi_calc exactly. The engine is
only constructed when at least one ROI is defined. Downstream CBOR/HDF5 already
consume message.roi, so no further changes are needed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
NUMA CPU/memory pinning is no longer worthwhile: the FPGA DMA buffers are
placed device-local by the kernel (dma_alloc_coherent), the big RAM ring
buffer is random-access (first-touch handles placement), and GPU work is
already spread across all visible devices. So drop the pinning entirely
and with it libnuma.
- Delete NUMAHWPolicy; the only concern worth keeping - GPU selection -
is done directly via pin_gpu() (round-robin over visible GPUs) in the
indexer pool and the Lite analysis threads. CPU-only threads
(FPGA acquire/pedestal/summation/frame-transform) no longer bind
anything.
- Drop get_gpu_numa_node() (sysfs lookup) - only SelectGPUAndItsNUMA
used it.
- numa_policy broker setting is deprecated and ignored (kept in the API
for backward compatibility; warns once on startup).
- Remove NUMA_LIBRARY / numa.h / numaif.h detection from CMake.
- Docs: drop the NUMA dependency, remove the numa_policy config example,
and document running multiple brokers on disjoint GPUs via
CUDA_VISIBLE_DEVICES.
- Remove NUMA_GPU_REVIEW.md (the planning note; this work is now done).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Several items in the note had landed or were inaccurate. Update it to:
- mark DONE: G2 pin_gpu, ImageBuffer first-touch, acquisition_device
de-NUMA, CUDAWrapper sysfs node lookup;
- add the previously-missing FPGA DMA buffer section - placement is a
kernel concern (dma_alloc_coherent, device-local), not the userspace
mbind, which was only the simulator;
- record that libnuma is now down to a single file (NUMAHWPolicy.cpp);
- note that dependency removal (G3) and pinning behaviour (G4) are
separable axes.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
