[Bug-gsl] Possible bug in gsl_multifit_wlinear_svd

[Marco Lombardi]

Dear all,

occasionally I find that gsl_multifit_wlinear_svd never returns.  This
happens with particular matrices, which however do not seem to have
anything special of their own.  I attach below a simple program that
builds a random matrix and calls gsl_multifit_wlinear_svd, and then
reads a matrix from a file and calls again gsl_multifit_wlinear_svd:
the second call never exit (at least on the computers where I tried
it).

Best regards,
Marco Lombardi

P.S. The binary file gsl-test.dat is written in little-endian format.

gsl-test.c
Description: Text document

gsl-test.dat
Description: Binary data