quokka-astro/quokka

NSCBC::setOutflowBoundaryLowOrder<...>(...): in the <code>DIR == FluxDir::X3</code> reflecting fallback branch, <code>Q_im3</code> is assigned three times and <code>Q_im4</code>/<code>Q_im5</code> are never populated (<code>src/hydro/NSCBC_outflow.hpp:427-432</code>; repeated assignments at <code>:430-432</code>)

#1702

· BenWibking opened

on Feb 24, 2026

NSCBC::detail::transverse_xdir_dQ_data(...): in the <code>AMREX_SPACEDIM == 3</code> z-derivative branch, the computed derivative is assigned to <code>dQ_dy_data</code> instead of <code>dQ_dz_data</code> (<code>src/hydro/NSCBC_outflow.hpp:132-137</code>, assignment at <code>:136</code>)

bug: wrong answer/failure/crash

code-audit

codex

#1701

· BenWibking opened

on Feb 24, 2026

rk_adaptive_integrate(...): ignores the return code of the initial <code>rhs(t0, y0, ydot0)</code> call used for timestep estimation (<code>src/math/ODEIntegrate.hpp:137</code>), so an RHS failure can leave <code>ydot0</code> invalid and contaminate <code>dt_guess</code> / subsequent integration control

bug: wrong answer/failure/crash

code-audit

codex

#1700

· BenWibking opened

on Feb 24, 2026

error_norm(...): weighting uses <code>reltol * y0[i] + abstol[i]</code> (<code>src/math/ODEIntegrate.hpp:118</code>) instead of <code>reltol * abs(y0[i]) + abstol[i]</code> (as in standard weighted RMS norms)

bug: wrong answer/failure/crash

code-audit

codex

#1699

· BenWibking opened

on Feb 24, 2026

interpolate2d(...): boundary-degenerate branch conditions mistakenly compare coordinate <code>yi</code> (double) instead of index <code>iy</code> (int) (<code>src/math/Interpolate2D.hpp:58</code>, <code>src/math/Interpolate2D.hpp:63</code>)

bug: wrong answer/failure/crash

code-audit

codex

#1698

· BenWibking opened

on Feb 24, 2026

SinkAccretionUtils::ComputeScaleDown<problem_t>(...): Jeans-density limiter branch is guarded by <code>if (accretion_rate_cell > std::numeric_limits<double>::min())</code> (<code>src/particles/particle_accretion.hpp:289</code>), but accretion-zone rates are accumulated as non-positive values (<code>src/particles/particle_accretion.hpp:237-239</code>)

bug: wrong answer/failure/crash

code-audit

codex

particles

#1693

· BenWibking opened

on Feb 24, 2026

particle_io::saveParticleDataToTxtFile<ContainerType>(...): gathers <code>particle_ids</code> (<code>src/particles/particle_IO.hpp:376</code>) but never writes them, and integer-component output starts at index <code>1</code> (<code>src/particles/particle_IO.hpp:398</code>) instead of <code>0</code>

code-audit

codex

particles

#1692

· BenWibking opened

on Feb 24, 2026

read_dataset(...): reads <code>ndims</code> but unconditionally indexes <code>dims[0..2]</code> when constructing the 3D table (<code>src/turbulence/TurbDataReader.cpp:27-30</code>, <code>:49</code>)

bug: wrong answer/failure/crash

code-audit

codex

#1690

· BenWibking opened

on Feb 24, 2026

PhysicsParticleDescriptor<...>::splitParticles(...): no validation that <code>splitFactor > 0</code> (<code>src/particles/PhysicsParticles.hpp:408-469</code>)

bug: wrong answer/failure/crash

code-audit

codex

particles

#1689

· BenWibking opened

on Feb 24, 2026

RadSystem::AddSourceTermsSingleGroup(...): in the <code>beta_order_ > 1</code> branch, <code>gasVel</code> is value-initialized but never populated before building the 3x3 flux-update matrix (<code>src/radiation/source_terms_single_group.hpp:399</code>, <code>:453-463</code>)

bug: wrong answer/failure/crash

code-audit

codex

#1688

· BenWibking opened

on Feb 24, 2026

HydroSystem::ComputeFluxes<...>(...): face-centered normal velocity uses the opposite-side density when deriving <code>v_norm</code> from mass flux (<code>F[rho] >= 0</code> divides by <code>rho_R</code>, <code>F[rho] < 0</code> divides by <code>rho_L</code>) at <code>src/hydro/hydro_system.hpp:1514-1521</code>

bug: wrong answer/failure/crash

code-audit

codex

#1687

· BenWibking opened

on Feb 24, 2026

AMRSimulation::ReadMetadataFile(...): uses <code>feholdexcept(...)</code> / <code>fesetenv(...)</code> for temporary FPE masking (<code>src/simulation.hpp:4164-4165</code>, <code>:4192</code>), but <code>YAML::LoadFile(...)</code> or subsequent parsing can throw before <code>fesetenv(...)</code> runs, leaving the process FPE environment altered

bug: wrong answer/failure/crash

code-audit

codex

#1684

· BenWibking opened

on Feb 24, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

NSCBC::detail::transverse_xdir_dQ_data(...): in the <code>AMREX_SPACEDIM == 3</code> z-derivative branch, the computed derivative is assigned to <code>dQ_dy_data</code> instead of <code>dQ_dz_data</code> (<code>src/hydro/NSCBC_outflow.hpp:132-137</code>, assignment at <code>:136</code>)

error_norm(...): weighting uses <code>reltol * y0[i] + abstol[i]</code> (<code>src/math/ODEIntegrate.hpp:118</code>) instead of <code>reltol * abs(y0[i]) + abstol[i]</code> (as in standard weighted RMS norms)

interpolate2d(...): boundary-degenerate branch conditions mistakenly compare coordinate <code>yi</code> (double) instead of index <code>iy</code> (int) (<code>src/math/Interpolate2D.hpp:58</code>, <code>src/math/Interpolate2D.hpp:63</code>)

particle_io::saveParticleDataToTxtFile<ContainerType>(...): gathers <code>particle_ids</code> (<code>src/particles/particle_IO.hpp:376</code>) but never writes them, and integer-component output starts at index <code>1</code> (<code>src/particles/particle_IO.hpp:398</code>) instead of <code>0</code>

read_dataset(...): reads <code>ndims</code> but unconditionally indexes <code>dims[0..2]</code> when constructing the 3D table (<code>src/turbulence/TurbDataReader.cpp:27-30</code>, <code>:49</code>)

PhysicsParticleDescriptor<...>::splitParticles(...): no validation that <code>splitFactor > 0</code> (<code>src/particles/PhysicsParticles.hpp:408-469</code>)

RadSystem::AddSourceTermsSingleGroup(...): in the <code>beta_order_ > 1</code> branch, <code>gasVel</code> is value-initialized but never populated before building the 3x3 flux-update matrix (<code>src/radiation/source_terms_single_group.hpp:399</code>, <code>:453-463</code>)

Issues

Search results