Fix QA: drop stale Test dep and resolve re-exported integrator accessors#152
Draft
ChrisRackauckas-Claude wants to merge 1 commit into
Draft
Conversation
…n_tests
The GPU.yml workflow runs the suite with GROUP="GPU", but the folder-discovery
run_tests() only accepts All/Core/QA plus groups declared in test_groups.toml
(Core, QA), so it errored: GROUP="GPU" is not a declared group. GPU is a
capability lane, not a folder of files — it runs the *same* Core files on a
CUDA-equipped runner so the CUDA.functional()-gated testsets in reflect.jl /
MCSample.jl self-activate. This cannot be expressed as a separate folder.
Switch to explicit-args run_tests (the DeepEquilibriumNetworks.jl convention):
core = the 9 Core files in their original runtests order; groups = {"GPU" =>
core_body}; qa = test/qa; all = ["Core"]. GROUP=GPU now routes to the Core
body, matching the pre-v1.2 dispatcher where GPU fell into the non-QA branch.
Verified locally on Julia 1.10 against SciMLTesting v1.2: GROUP=GPU -> Core,
GROUP=Core -> Core, GROUP=All -> Core, GROUP=QA -> QA. Runic-clean.
Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
ChrisRackauckas-Claude
force-pushed
the
qa-fix-stale-test-dep-undefined-exports
branch
from
b3d328d to
0f49cd5
Compare
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What (rebased onto current
main, 2026-07-05)Rebased onto
main(now at #156 "QA: run_qa v1.6 + ExplicitImports"). The QA parts of the original PR are already onmainvia #156 and were absorbed by the rebase:Testdep — QA: run_qa v1.6 form + ExplicitImports #156 already movedTestfrom[deps]to[extras]/[targets].test. Nothing left to do; the rebased branch'sProject.tomlequalsmain.du_cache/u_cache/user_cacheundefined-exports — QA: run_qa v1.6 form + ExplicitImports #156 handled this viaaqua_broken = (:undefined_exports,)(tracking QA: Aqua undefined_exports flags du_cache/u_cache/user_cache on Julia 1.12 (DiffEqBase @reexport binding-partition) #157). The rebased branch'ssrc/HighDimPDE.jlequalsmain(the originalusing DiffEqBase: du_cache, ...approach was dropped to avoid an Unexpected-Pass conflict with QA: run_qa v1.6 form + ExplicitImports #156'saqua_broken).The only surviving change is the GPU CI harness fix (
test/runtests.jl, +31/-1), which is still needed:GPU Testsis red onmainwithGPU is a capability lane (same Core files on a CUDA runner, where
CUDA.functional()-gated testsets self-activate), not a folder. Switched to explicit-argsrun_tests(core= the 9 Core files;groups = {"GPU" => core_body};qa = test/qa;all = ["Core"]). This is the documented SciMLTesting explicit-args API.Verified locally (Julia 1.12.6, released SciMLTesting 1.6): routing is correct —
GROUP=GPU→ Core body (no ArgumentError),GROUP=Core→ Core,GROUP=All→ Core,GROUP=QA→ QA.test/runtests.jlparses cleanly and is Runic-clean.Out of scope / pre-existing (NOT introduced here)
Core (julia 1 = 1.12, julia pre = 1.13), all OSes: Julia 1.12+ LLVM codegen segfault.
signal 11attest/DeepBSDE.jl:19, inemit_unbox/emit_unboxed_coercion(codegen.cpp) while JIT-compiling the Zygote adjoint of the DeepBSDE ensemble SDE solve. Reproduces byte-identically on unmodifiedmain(master run 28647091704, job 84956919471 — samesignal 11, sameDeepBSDE.jl:19). Core lts (1.10) is green with identical resolved deps, so this is purely a Julia-version codegen crash, not a HighDimPDE logic bug and not fixable in-package. Untouched by this PR.Please ignore until reviewed by @ChrisRackauckas.