Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

Determinant and Inverse return wrong values when called from more than one thread

Aperta
  • Qualcosa non va?
#40 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub

@Rafael-SOWNet ci sta già lavorando.

Dal 9/9/2026.

  • #41 di @Rafael-SOWNet — aperta

Valutazione

Difficoltà
4/5
Tempo stimato
3-5 giorni
Idoneità per principianti
35/100
Tipo di issue
Bug
Chiarezza
Specificata chiaramente
Stato di attività
Ferma
Stack tecnologico
csharp
Ambito
performance

Direzione di ricerca

Inizia con SquareMatrixFactory<T, TWrapper>.GetMatrix e ExpressionCompiler<T, TWrapper>.storage, quindi riproduci i fallimenti riportati di determinante/inversa con Parallel.For, includendo una cache fredda. Il lavoro è completato quando le operazioni concorrenti sulle matrici restituiscono i valori verificati sequenzialmente senza eccezioni sulle raccolte né corruzione dei buffer condivisi.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

Two process-wide statics are read and written with no synchronisation, so matrix operations called from more than one thread return wrong numbers. Nothing throws, nothing is malformed, and no result looks suspicious.

I found this from the other side: AngouriMath uses GenTensor for its symbolic matrices, and a determinant computed on a background thread was coming back numerically wrong. AngouriMath's issue is #1219.

1. The scratch-matrix pool

SquareMatrixFactory<T, TWrapper>.GetMatrix hands every caller the same GenTensor for a given size, and the caller then writes into it — Inversion.GetCofactorMatrix(t, temp, ...) fills the matrix it is given. Two threads taking a determinant or an inverse of the same size get the same buffer and overwrite each other's minors between the write and the read.

Sixty 5×5 int matrices, each computed once sequentially for truth and then rebuilt and recomputed under Parallel.For:

DeterminantLaplace    53 of 60 disagree
Adjoint               60 of 60 disagree

Around 10 ms, every run. This is not a rare interleaving, it is the normal case.

DeterminantGaussianSafeDivision is unaffected — it works on its own copy and never asks the pool for anything.

The lock inside GetMatrix does not help and is itself incomplete: the list is indexed after the lock is released, so a concurrent Add can reallocate the backing array underneath the reader.

2. The compiled-operation cache

ExpressionCompiler<T, TWrapper>.storage is a plain Dictionary, written by whichever thread first asks for a given (operation, rank, parallel). Dictionary<,> is documented as safe for concurrent readers only while nobody is writing.

Reached cold from several threads it fails with the runtime's own diagnostic:

System.InvalidOperationException: Operations that change non-concurrent collections must
have exclusive access. A concurrent update was performed on this collection and corrupted
its state. The collection's state is no longer correct.
...
The given key '(Subtraction, 1, False)' was not present in the dictionary.

Worth noting that this one hides easily. My first test computed its expected values sequentially and only then went parallel, which populated every key before the threads started — the cache holds a handful of entries, so a warm-up makes the defect invisible and the test passes against the broken code.

Fix

I have opened a PR. Both are small: [ThreadStatic] for the pool, since a scratch buffer has nothing to share between threads, and ConcurrentDictionary for the cache. The Laplace determinant comes out about 5% faster, because the pool is now taken once at the top of the recursion instead of re-checked and re-indexed at every one of its n! nodes.

Lingua principale
C#
Stelle
52
Fork
6
Nessuna PR unita negli ultimi 30g

Preparare l'ambiente

Questo progetto non fornisce container di sviluppo, Dockerfile né guida per i contributori, quindi l'ambiente è a tuo carico: parti dal suo README e consulta la nostra guida al primo contributo per i passaggi generali.

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di ASC-Community/GenericTensor

Tutte le issue di ASC-Community/GenericTensor

Issue simili

Altre issue su C#

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.