Intel® oneAPI Math Kernel Library
Ask questions and share information with other developers who use Intel® Math Kernel Library.

unhandled exception error while executing

john-royv
Beginner
635 Views
Hi there, I am new to the programming world and have been asked to investigate using the MKL. We have the libraries on a trial license at the moment and should we find them applicable we will purchase them. We have managed to compile test versions that use the cblas functions and they operate fine, but when compiling the lapack functions we get problems. The code builds without any errors or warnings but when trying to execute it, we get an unhandled exception error.... Hopefully this is a trivial problem yet we have not been unable to come up with a solution and hope you could help. The code exits without any problems when the size of the matrices is small but when larger matrices are used it gives the error. The functions we were trying to test were dgetrf and dgesv
Please find the offending file attached as a .txt.....
0 Kudos
2 Replies
john-royv
Beginner
635 Views
I forgot to mention that my OS is winxp prof and i am using VS .Net 2003
0 Kudos
Intel_C_Intel
Employee
635 Views
My colleague Greg and I have been looking over your code and most of it is correct. However we believe there is an error which is giving rise to the porgram failure.
You are required to provide a vector of size M for the pivot vector, which is needed both by dgetrf and later, at solve time, by dgetrs. You have assigned a variable "dim" to play this role, but it is a scalar set to the value of M but not to the size of M. DGETRF is treating this scalar as a vector and is trying to write where there is no allocation.
We think that if you fix that your test will run correctly.
Bruce
0 Kudos
Reply