Add memprobe diagnostic module for memory-growth tracking (#176)#179
Merged
Conversation
Closed
Contributor
There was a problem hiding this comment.
Pull request overview
Adds an opt-in uw.utilities.memprobe diagnostic module to help attribute memory growth in long MPI runs (motivated by issue #176), including lightweight RSS/KDTree counters, optional GC-based class counting, PETSc leak-dump option helper, solver-level instrumentation hooks, documentation, and smoke tests.
Changes:
- Introduces
underworld3.utilities.memprobewith snapshot/diff/probe/decorator APIs and a PETSc finalize leak-dump helper. - Instruments KDTree lifecycle counts in
ckdtree.pyxand decoratesStokes.solve()/NavierStokes.solve()with memprobe instrumentation. - Adds developer guide + pytest smoke tests for the new module.
Reviewed changes
Copilot reviewed 6 out of 6 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
src/underworld3/utilities/memprobe.py |
Implements snapshot/diff formatting, probe context manager, instrumentation decorator, and PETSc option helper. |
src/underworld3/ckdtree.pyx |
Adds module-level live/constructed counters and exposes them for diagnostics. |
src/underworld3/systems/solvers.py |
Decorates key solver solve() methods to emit memprobe diffs when enabled. |
src/underworld3/utilities/__init__.py |
Exposes memprobe under underworld3.utilities. |
tests/test_0780_memprobe.py |
Smoke tests for memprobe API, KDTree counters, and enabled/disabled behavior. |
docs/developer/guides/memory-diagnostics.md |
Documents how to use memprobe and PETSc leak-tracking flags. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+79
to
+89
| def _rss_mb() -> float: | ||
| """Current resident-set size in MiB. | ||
|
|
||
| On Linux ``ru_maxrss`` is in KiB; on macOS it is in bytes. We probe the | ||
| platform once to avoid getting it wrong silently. | ||
| """ | ||
| rss = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss | ||
| if sys.platform == "darwin": | ||
| return rss / (1024 * 1024) | ||
| return rss / 1024 | ||
|
|
| """ | ||
| delta: dict[str, Any] = {} | ||
|
|
||
| drss = after["rss_mb"] - before["rss_mb"] |
Comment on lines
+262
to
+265
| opts.setValue("-malloc_dump", "") | ||
| opts.setValue("-objects_dump", "") | ||
| if filename: | ||
| opts.setValue("-malloc_view", filename) |
| itself can verify. | ||
| """ | ||
| import gc | ||
| import os |
Comment on lines
+19
to
+22
| # Incremented in __cinit__ and decremented in __dealloc__ — Cython | ||
| # guarantees deterministic destruction so this stays accurate across | ||
| # normal use. Read via uw.utilities.memprobe.snapshot(), or directly | ||
| # via uw.kdtree.live_count(). |
|
|
||
| | Signal | Source | Cost | | ||
| |---|---|---| | ||
| | Process RSS (MiB) | `resource.getrusage` | free | |
lmoresi
added a commit
that referenced
this pull request
May 10, 2026
- _rss_mb(): switch from resource.ru_maxrss (peak/high-water RSS) to psutil.Process().memory_info().rss (current RSS), with /proc/self/statm Linux fallback and ru_maxrss kept only as a last resort. Current RSS is preferred so freed memory shows as a negative delta. (Copilot #1, #6) - diff(): apply 0.01 MiB threshold before including rss_mb in the delta. Smaller noise prints as "+0.00 MiB" and defeats the "no change" fast path. (Copilot #2) - dump_petsc_leaks_at_finalize(): drop leading "-" from PETSc option keys (codebase convention is no dash; "-malloc_dump" was a no-op), and align docstring with what the function actually sets. (Copilot #3) - Remove unused `os` import from test module. (Copilot #4) - Soften ckdtree.pyx comment claiming Cython "guarantees deterministic destruction" — that's CPython refcounting, which can lag for objects trapped in reference cycles until cyclic GC runs. (Copilot #5) - Test: assert KDTree count deltas relative to immediate before/after, never to absolute baselines. Earlier tests in the same pytest session can leave KDTree refs alive that the cyclic GC may collect at any time, shifting the absolute count. CI surfaced this with "assert 332 == 338" — six trees from earlier tests collected during our gc.collect(). Underworld development team with AI support from Claude Code
lmoresi
added a commit
that referenced
this pull request
May 10, 2026
CI on PR #179 surfaced a flaky failure in test_bc_accepts_raw_numbers: any(k[0] == name for k in UWexpression._ephemeral_expr_names) RuntimeError: dictionary changed size during iteration The dict is mutated asynchronously by weakref finalizers running during cyclic GC, so iterating it directly races against those callbacks. The fix is to snapshot the keys with list(...) before iterating — at most a few hundred entries, negligible cost. Pre-existing bug — surfaces depending on test ordering, GC pressure, and timing. The memprobe PR's added tests happen to shift teardown state enough to trigger it consistently. Worth landing here so #179 isn't blocked. Underworld development team with AI support from Claude Code
Long parallel runs occasionally OOM on HPC even when each step looks
small. memprobe gives a way to "light up" memory tracking on demand,
sample at regular intervals, and pin which subsystem is growing.
Components:
- src/underworld3/utilities/memprobe.py — snapshot/diff/probe/instrument
API. Snapshots capture process RSS (resource.getrusage), KDTree live +
total-constructed counts, and (with full=True) per-class Python
instance counts via gc.get_objects. Diffs filter out unchanged keys
and sort py_classes by absolute change so dominant suspects surface
first.
- ckdtree.pyx — module-level live-instance counter, accurate via
Cython's deterministic __cinit__/__dealloc__. Exposed as
uw.kdtree.live_count() and uw.kdtree.total_constructed().
- Stokes.solve() and NavierStokes.solve() decorated with
@memprobe.instrument(...). The decorator fast-returns when ENABLED
is False (one bool check, sub-microsecond), so it's safe on hot paths.
Activation:
- UW_MEMPROBE=1 env var → enables instrumentation hooks at import time.
- memprobe.enable() / disable() for runtime toggling.
- with memprobe.probe("label"): … for ad-hoc blocks.
Skipped: parsing PETSc.Log object tables from Python. The runtime
flags -log_view and -malloc_dump give the same information more
reliably; documented in the guide and exposed via
memprobe.dump_petsc_leaks_at_finalize().
Adds tests/test_0780_memprobe.py (9 smoke tests) and
docs/developer/guides/memory-diagnostics.md with debugging recipes.
Does not fix issue #176 (Ben's OOM on HPC) — provides the
instrumentation he needs to bisect it.
Underworld development team with AI support from Claude Code
- _rss_mb(): switch from resource.ru_maxrss (peak/high-water RSS) to psutil.Process().memory_info().rss (current RSS), with /proc/self/statm Linux fallback and ru_maxrss kept only as a last resort. Current RSS is preferred so freed memory shows as a negative delta. (Copilot #1, #6) - diff(): apply 0.01 MiB threshold before including rss_mb in the delta. Smaller noise prints as "+0.00 MiB" and defeats the "no change" fast path. (Copilot #2) - dump_petsc_leaks_at_finalize(): drop leading "-" from PETSc option keys (codebase convention is no dash; "-malloc_dump" was a no-op), and align docstring with what the function actually sets. (Copilot #3) - Remove unused `os` import from test module. (Copilot #4) - Soften ckdtree.pyx comment claiming Cython "guarantees deterministic destruction" — that's CPython refcounting, which can lag for objects trapped in reference cycles until cyclic GC runs. (Copilot #5) - Test: assert KDTree count deltas relative to immediate before/after, never to absolute baselines. Earlier tests in the same pytest session can leave KDTree refs alive that the cyclic GC may collect at any time, shifting the absolute count. CI surfaced this with "assert 332 == 338" — six trees from earlier tests collected during our gc.collect(). Underworld development team with AI support from Claude Code
CI on PR #179 surfaced a flaky failure in test_bc_accepts_raw_numbers: any(k[0] == name for k in UWexpression._ephemeral_expr_names) RuntimeError: dictionary changed size during iteration The dict is mutated asynchronously by weakref finalizers running during cyclic GC, so iterating it directly races against those callbacks. The fix is to snapshot the keys with list(...) before iterating — at most a few hundred entries, negligible cost. Pre-existing bug — surfaces depending on test ordering, GC pressure, and timing. The memprobe PR's added tests happen to shift teardown state enough to trigger it consistently. Worth landing here so #179 isn't blocked. Underworld development team with AI support from Claude Code
lmoresi
force-pushed
the
bugfix/memprobe-diagnostics
branch
from
d1a66a4 to
5f64de2
Compare
This was referenced May 11, 2026
4 tasks
lmoresi
added a commit
that referenced
this pull request
Jul 6, 2026
…og, value-first call-site sweep (WE-01..03,05,06,08,09,10) (#338) * docs(WE-01): adopt the one-governing-doc-per-topic authority map Repoint CLAUDE.md's Data Access 'Authoritative Reference' from the stale UW3_Style_and_Patterns_Guide.md to subsystems/data-access.md (the guide it crowned teaches patterns the code deprecates at runtime — DOC-04), and record the Style Charter §10 authority table in docs/developer/index.md as the master authority index. The Charter is added to the Getting Started toctree (removes a baseline 'not included in any toctree' warning). Finding: DOC-04 (docs/reviews/2026-07/DOCS-STANDARDS-COHERENCE.md). Underworld development team with AI support from Claude Code * docs(WE-02): de-drift the Style Guide's four stale normative sections Rewrites the sections DOC-01 verified as contradicting the settled standards: - Docstring format: the 'Markdown Docstrings for pdoc/pdoc3' section is replaced by the NumPy/Sphinx RST standard (worked example with :math: and Parameters/Returns/Examples/Notes; conversion tracked in docs/plans/docstring-conversion-plan.md), per Style Charter section 6. - Doc file format: Quarto .qmd prescription (zero .qmd files exist in the repo) replaced by MyST .md/Sphinx guidance matching CLAUDE.md; migration table row updated. - Data access examples: 'Preferred' coordinate examples now use the real, runnable API — mesh.X.coords (read), mesh.deform() (coordinate changes), and the swarm.coords getter/setter for particle positions. The previous 'Preferred' example swarm.data += displacement raises AttributeError (getter-only property — SWARM-13 evidence); mesh.data warns at runtime. The private-attribute migration advice (swarm._particle_coordinates, mesh._deform_mesh presented as the NEW pattern) is deleted. - Front matter: the 21-line Quarto YAML header is replaced by a minimal MyST title block, and the guide now states that the UW3 Style Charter is the normative contract and wins on conflict. All replacement examples verified against current source: Swarm.coords setter (swarm.py), Mesh.deform (discretisation_mesh.py:3133), uw.synchronised_array_update / NDArray_With_Callback.delay_callbacks_global. Findings: DOC-01, SWARM-13 (style-guide part). Underworld development team with AI support from Claude Code * docs(WE-03): regenerate the docstring review queue; add the sweep to the release checklist The queue (last generated 2026-01-13, cdf5bb2) misrepresented the codebase both ways: it flagged now-complete items (solve, SNES_Scalar) as missing and contained zero entries for the June 2026 API (DOC-02). Regenerated over src/underworld3/**/*.py + **/*.pyx at the current tip. Two bugs in scripts/docstring_sweep.py's regex-based Cython parser made the regenerated queue lie about .pyx docstrings and are fixed as part of making the regeneration meaningful: - the indent group '(\s*)' with re.MULTILINE consumed preceding blank lines, shifting the computed definition line so the docstring search started ON the def/class line and always missed; - the docstring search started at the definition line rather than after the (possibly multi-line) signature, so long signatures hid their docstrings; - raw-string docstrings (r""", the norm in the solver .pyx) were not recognised. DOC-02 cross-validation on the regenerated queue now passes: solve / SNES_Scalar in the solver pyx are no longer flagged 'none'; the queue contains the June API (add_nitsche_bc, add_rotated_freeslip_bc, boundary_flux, set_custom_fmg, consistent_jacobian: 13 mentions) and flags the DOC-05 targets (Swarm.advection x2, read_timestep, write_proxy) as undocumented. Also adds the sweep to the quarterly release checklist (guides/release-process.md) so the queue cannot go stale unnoticed again. Findings: DOC-02 (docs/reviews/2026-07/DOCS-STANDARDS-COHERENCE.md). Underworld development team with AI support from Claude Code * docs(WE-05): backfill the changelog for May - early July 2026; add the changelog sweep to the release checklist The changelog (the quarterly CIG/stakeholder record) ended in April 2026 while ~117 first-parent commits landed May through early July (DOC-03). Backfilled at the existing conceptual granularity — 14 grouped entries, grouped by subsystem rather than by PR, matching the established format (### Title (Month Year), bold lead sentence, hyphen bullets, inline PR references): - New '2026 Q3 (July - September)' section: the July 2026 quality campaign (#309-#313, #317, #322-#326, #329, #334 as grouped entries), rotated strong free-slip / boundary traction / dynamic topography (#293, #294, #298, #306), generalized geometric multigrid via custom prolongation (#290, #297), consistent Jacobian tangent (#258), swarm correctness (#216, #313, #323, #329), numpy 2 support (#301, #305). - Extended '2026 Q2' section with the May-June entries: mesh adaptation movers (#190, #209, #213, #228, #259, #264, #266), moving-mesh field transfer / deform() (#246, #249, #251), semi-Lagrangian accuracy controls (#164, #183, #185-#189, #208, #220), snapshot/checkpoint toolkit (#146, #195, #196, #198), Stokes_Constrained (#224, #229, #240, #265), local-h Nitsche + boundary-slip surfaces (#225, #241, #275), units interoperability (#277, #278, #283, #284), memory/evaluation/solver infrastructure (#161, #177-#179, #181, #182, #222, #237, #250, ...). Every entry is backed by a merged commit on development (verified against git log --first-parent aed517f..3184a40). Also adds a quarterly-changelog sweep step beside the docstring sweep in the release checklist (guides/release-process.md) per DOC-03's proposed fix. Findings: DOC-03 (docs/reviews/2026-07/DOCS-STANDARDS-COHERENCE.md). Underworld development team with AI support from Claude Code * docs(WE-06): status headers on the unmarked design docs (per-doc git verification) Adds one-to-three-line Status markers to the 13 design docs that lacked one, following the directory's existing conventions (**Status**: line under the title; status: key inside existing YAML frontmatter for the three frontmatter-only docs), and corrects the stale 'Design Phase' marker on MATHEMATICAL_MIXIN_DESIGN.md (the mixin ships in utilities/mathematical_mixin.py). Every stamp was verified against git history (git log --follow dates) and the current source tree before writing: - Implemented: jacobian-consistent-tangent (PR #258, c63cd70), fmg-checkpoint-hierarchy (3cd73cd), petsc-dmplex-checkpoint-reload-plan (PR #146, write_timestep(petsc_reload=True) in tree), fault-refinement-simplification (smooth_mesh_interior / metric_density_from_gradient / fault_comb_metric all in tree), MATHEMATICAL_MIXIN_DESIGN. - Current reference/contract: mesh-adaptation-formulation, ND_UNITS_BOUNDARY_CONTRACT (PR #278, e0ece9a). - Investigation records (preserved via PR #245, 34a9dd4; production geometric-MG is custom prolongation, PR #290): snesfas-feasibility, snesfas-vanka-feasibility-study. - Design notes / prototypes with honest gaps: in_memory_checkpoint_design (not implemented, per its own trailing Status section), submesh-solver-architecture (extract_region/extract_surface exist; coarsened_companion does not). - Historical: ARCHITECTURE_ANALYSIS (persistence.py layout superseded), COORDINATE_MIGRATION_GUIDE (transition shipped), WHY_UNITS_NOT_DIMENSIONALITY (decision record). The audit's ~16 estimate over-counted: re-derived at this tip, 13 docs were unmarked plus one marked-but-stale (DOC-07). Findings: DOC-07 (docs/reviews/2026-07/DOCS-STANDARDS-COHERENCE.md). Underworld development team with AI support from Claude Code * docs(WE-08): convert units.py public docstrings Google -> NumPy style Docstring-only conversion of the 18 public module-level functions that carried Google-style Args:/Returns:/Raises:/Examples: labels (check_units_consistency, get_dimensionality, get_units, non_dimensionalise, show_nondimensional_form, simplify_units, create_quantity, convert_units, to_base_units, to_reduced_units, to_compact, get_scaling_coefficients, set_scaling_coefficients, validate_expression_units, assert_dimensionality, validate_coordinates_dimensionality, enforce_units_consistency, require_units_if_active, convert_angle_to_degrees) to the NumPy/Sphinx standard (Style Charter section 6). dimensionalise was already NumPy style; one-line docstrings and private helpers are untouched. No code, signature, or behaviour changes (verified: every diff hunk is inside a docstring; ast.parse clean). Finding: API-12 (docs/reviews/2026-07/API-CONSISTENCY-REVIEW.md). Underworld development team with AI support from Claude Code * docs(WE-09): sweep call sites of the newer BC methods to value-first (conds, boundary, ...) order Wave C (#334) made the ORIGINAL value-first order canonical for add_nitsche_bc / add_rotated_freeslip_bc / add_constraint_bc (maintainer decisions D2/D3; Style Charter section 6) with deprecation shims for the legacy boundary-first and g= spellings. This sweep updates every call site of those THREE methods to the canonical order so nothing in the repository exercises the shims — 74 sites total: - tests/: 63 call sites across 12 files (test_1017, test_1018, test_1060, test_1061, test_1062, test_1064, test_1065 x2 serial; parallel test_1017, test_1062, test_1063, test_1064). tests/test_0641_wave_c_api_shims.py is deliberately untouched — its legacy-order calls ARE the deprecation contract. - docs/: 7 sites (curved-boundary-conditions.md x4, CONSTRAINED_FREESLIP_MULTIPLIER.md call + signature line, examples/submesh_investigation/test_region_ds_nitsche.py). - .claude/skills/: 3 sites (adapt-on-top-faults x2, free-surface-convection x1). - CLAUDE.md: 1 signature reference (free-slip BC preference section). The ~1,370 legacy-trio (add_dirichlet_bc/add_natural_bc/add_essential_bc) sites already conform and are untouched per the D2 decision. The audit review documents under docs/reviews/2026-07/ record the pre-decision state as evidence and are not swept. Discovered while verifying the swept tests run warning-free: the Wave C zero-datum guard in add_rotated_freeslip_bc rejects FLOAT zero (sympy.sympify(0.0) != 0 is structurally True), so the canonical add_rotated_freeslip_bc(0.0, boundary) raises NotImplementedError while conds=0 works. Filed as issue #336 with a TODO(BUG) marker at the guard (comment-only src touch); the swept call sites use the working integer form add_rotated_freeslip_bc(0, boundary). No fix applied here (Charter section 9 scope discipline). Findings: API-01/API-02 sweep (WE-09, REMEDIATION-WORKLIST.md). Underworld development team with AI support from Claude Code
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds
uw.utilities.memprobe, an opt-in diagnostic module for surfacing memory growth in long parallel runs. Complementary to #178 (the fix for #176): #178 plugged the specific leak holes, this PR is the leak-detector that survives for the next time someone reports OOM.What it tracks
psutil→/proc/self/statm(Linux) →resource.ru_maxrss(last resort, peak only)KDTreelive + total-constructed__cinit__/__dealloc__gc.get_objects()walkfull=TrueKDTree counts are the signal neither PETSc's tracker nor #178's regression test covers — UW's nanoflann wrapper is invisible to PETSc.
PETSc-side object/allocation tracking is not parsed from Python. PETSc's
-log_viewand-malloc_dumpruntime flags give the same information more reliably and are documented in the guide.memprobe.dump_petsc_leaks_at_finalize()is a small helper that turns those on programmatically.Activation
UW_MEMPROBE=1env var → enables at import time, decorated solvers start emitting per-call diffs.memprobe.enable()/disable()for runtime toggling.with memprobe.probe("step 42"):for ad-hoc blocks.@memprobe.instrument("label")decorator — fast-returns when disabled (one bool check), safe on hot paths. Pre-applied toStokes.solve()andNavierStokes.solve().Sample output
Bonus: dictionary-iteration race fix
UWexpression._ephemeral_expr_nameswas iterated directly while weakref finalizers mutated it under cyclic GC, raisingRuntimeError: dictionary changed size during iteration. Pre-existing bug — surfaced flakily depending on test ordering. Fixed by snapshotting keys vialist(...)before iterating. Independent of memprobe but landed here since the new tests reliably triggered it.Files
src/underworld3/utilities/memprobe.py— modulesrc/underworld3/ckdtree.pyx— KDTree counter instrumentationsrc/underworld3/systems/solvers.py— decorate Stokes + NavierStokes solvesrc/underworld3/function/expressions.py— race fixtests/test_0780_memprobe.py— 9 smoke testsdocs/developer/guides/memory-diagnostics.md— usage + debugging recipesValidation against #178
Once merged, a useful cross-check: run any earlier-leaking workload with
UW_MEMPROBE=1and confirm flat per-step RSS deltas where they used to climb. That validates both PRs at once.Test plan
pytest tests/test_0780_memprobe.py— 9 passed locally and on CIUW_MEMPROBE=1emits expected diffslivereturns to baseline afterdel tree; gc.collect()test_0006_memory_leak.py(Ben's Fix comprehensive memory leaks in solver setup, interpolation caching, and SubDM syncing #178 regression test) — passesNote on Copilot review
Copilot's two outstanding inline comments on
_rss_mb()and the RSS threshold referenceresource.getrusage/ a missing threshold. Both were addressed inb4665d1— psutil is now the primary source (resource is a last-resort fallback only), and the threshold isabs(drss) >= 0.01 MiB. The comments are pinned to the same line numbers in the rebased file and so re-surfaced after the rebase.Underworld development team with AI support from Claude Code