gdalle
commented
3 weeks ago Since I'm a good person, here's the guilty matrix for which bipartite partial coloring failed so badly. GitHub doesn't support .mtx so I just copied the file contents below
```
%%MatrixMarket matrix coordinate pattern general
20 10 74
2 1
8 1
10 1
11 1
12 1
13 1
18 1
2 2
6 2
9 2
10 2
12 2
13 2
17 2
19 2
7 3
12 3
13 3
15 3
3 4
4 4
5 4
6 4
14 4
16 4
18 4
19 4
3 5
7 5
8 5
11 5
14 5
17 5
18 5
20 5
8 6
9 6
10 6
11 6
12 6
17 6
19 6
20 6
1 7
5 7
6 7
8 7
10 7
18 7
2 8
4 8
5 8
8 8
10 8
11 8
14 8
20 8
2 9
4 9
6 9
10 9
13 9
14 9
17 9
18 9
19 9
2 10
3 10
4 10
7 10
10 10
11 10
14 10
15 10
```
amontoison
To diagnose the nondeterministic behavior of #21, I added some coloring correctness tests from SparseMatrixColorings.jl.
On my machine, the test file with this random seed causes a reproducible segfault, which is clearly not on me because I don't use
@inboundsanywhere in my Julia code. With other random seeds, I observe other kinds of failures, like colorings that don't start with1(which should be impossible).I temporarily disabled the tests on 1.6 because they failed for other reasons before the segfault could occur on 1.10.