[Bug-gsl] Bugs in I/O for gsl_matrix_complex

[Matthias Sitte]

Hi,

I've discovered a bug in the I/O for complex matrices. If tda is larger
than the actual size of the matrix, the fwrite() function called
internally writes the wrong blocks of data to disk. In particular, the
code containing the address of the memory block ("m->data + i * tda")
does /not/ respect the MULTIPLICITY!

I've written a small program to demonstrate the bug with some small
complex-valued matrix that is stored and read immediately afterwards
from disk.

I've also included a patch proposal for "matrix/file_source.c".

I'd be happy if you could verify/falsify my findings.

Best,

  Matthias

matrix_fwrite_bug.c
Description: Text Data

file_source.patch
Description: Text Data

signature.asc
Description: OpenPGP digital signature