Intel® oneAPI Math Kernel Library
Ask questions and share information with other developers who use Intel® Math Kernel Library.
Announcements
This community is designed for sharing of public information. Please do not share Intel or third-party confidential information here.

lapack errror : access violation

elquin
Beginner
146 Views
Dear all,
I'm testing IMSL and MKL library.
When I tested the attached code, the error comes out like :
forrtl: severe (157) : Program Exception - access violation
What should I have to change??
0 Kudos
1 Solution
mecej4
Black Belt
146 Views
Line 38 of file mata.dat contains a number too large to be read into a single-precision real (0.743121979E+124). Likewise, in matb.dat you have -0.747209150E+127.

These errors abort the calculation before any MKL or IMSL routines get called.

I also note that dgerfs is called with argument arrays of type single precision real. That is probably incorrect.

At the refining stage, the pivot array generated during the factorization is needed. You are providing the uninitialized array ipiv for this purpose. That is an error. Since the values in this array are used as indices into the matrix, you are likely to run into access errors in the call to gerfs.

View solution in original post

3 Replies
mecej4
Black Belt
146 Views
Please provide the input files mata.dat and matb.dat, which are required to run your program until the access violation point is reached.

It helps in resolving such problems if you can enable traceback when compiling your program.
elquin
Beginner
146 Views
Sorry for my prev. answer but I cannot find how i delete thread.
Here I attach matrix.
My environments are :
MS Visual Studio 2008
Intel Visual Fortran 11.1.065
MKL 10.2 update release 4
mecej4
Black Belt
147 Views
Line 38 of file mata.dat contains a number too large to be read into a single-precision real (0.743121979E+124). Likewise, in matb.dat you have -0.747209150E+127.

These errors abort the calculation before any MKL or IMSL routines get called.

I also note that dgerfs is called with argument arrays of type single precision real. That is probably incorrect.

At the refining stage, the pivot array generated during the factorization is needed. You are providing the uninitialized array ipiv for this purpose. That is an error. Since the values in this array are used as indices into the matrix, you are likely to run into access errors in the call to gerfs.
Reply