LAPACKE matrix inversion segmentation fault C

Question

I am trying to write a code using lapacke libraries to invert a complex matrix in C. However I am stuck with a segmentation fault which seems to depend on the size N of the matrix. What's more is that the size for which the segmentation fault happens varies each time I compile the program or I touch anything. This makes me think that somewhere the code is trying to access badly allocated or forbidden memory. Unfortunately, I dont' understand how this happens since it seems to be related with the LAPACKE functions themeselves. In fact, when the function /*MatrixComplexInv(invA,A,N);*/ ( in which the LAPACKE functions are called for the inversion) is commented the segmentation fault doesn't happen. Below there is the working code that can be compiled and run on its own.

 #include <stdio.h>
#include <lapacke.h>
#include <complex.h>
#include <stdlib.h>
#include <math.h>

void  Ctranspose( double complex *, double complex * ,int );
void MatrixComplexInv(double complex *, double complex *, int );


int main(int argc,  const char * * argv) {


 int  i,j,k,N = 4;/*if N> bigger than a small number 4,6,7.. it gives segmentation fault*/

double complex *A = calloc(N*N,sizeof(double complex)),
               *b = calloc(N*N,sizeof(double complex)),
               *Ap =calloc(N*N,sizeof(double complex));


    double complex *invA =calloc(N*N,sizeof(double complex));


        for(i=0;i<N;i++){
            for(j=0;j<N;j++){ 

            A[i*N+j] = 1+sin(i*j)*i+I*j;    
            Ap[i*N+j] = 1+sin(i*j)*i+I*j;   
            }
          }

    /*Segmentation fault in this function, due to 
     * 
    LAPACKE_zgetrf(LAPACK_ROW_MAJOR, n, n, tempA , n,&n);


    LAPACKE_zgetri(LAPACK_ROW_MAJOR, n, tempA , n, &n );
    * 
    * both.
    */
    MatrixComplexInv(invA,A,N);


    for(i=0;i<N;i++){
     for(j=0;j<N;j++){ 
         for(k = 0;k<N;k++){
            b[i*N+j]+=invA[i*N + k]*Ap[k*N + j];
         }


   printf("(%lf,%lf)\t", creal(b[i*N + j]),cimag(b[i*N + j]));/*tests that the result produces the inverse matrix A^{-1}A = 1*/


    }
     printf("\n");
    }

      return 0;
    }   


    void Ctranspose( double complex *Transposed, double complex *M ,int n)
    {

    int i,j;
    for(i=0;i<n;i++)

    for(j=0;j<n;j++) Transposed[i+n*j] = M[i*n+j];
    }


    void MatrixComplexInv(double complex *invA, double complex *A, int n)
    {

    double complex *tempA = (double complex*) malloc( n*n*sizeof(double complex) );


     Ctranspose(tempA,A,n);

    /*SEGMENTATION HAPPEN IN THESE TWO FUNCTIONS*/
    LAPACKE_zgetrf(LAPACK_ROW_MAJOR, n, n, tempA , n,&n);


    LAPACKE_zgetri(LAPACK_ROW_MAJOR, n, tempA , n, &n );


     Ctranspose(invA,tempA,n);


     free(tempA);

    }

Thank you for the correction, however this doesn't affect the outcome. — Fra
– Fra, Commented Feb 23, 2019 at 0:40
yeah this was a quick test program and the printf was sloppy. However the segmentation fault stands regardless. — Fra
– Fra, Commented Feb 23, 2019 at 0:41
Why are you allocating memory with malloc but freeing it with LAPACKE_free? — melpomene
– melpomene, Commented Feb 23, 2019 at 0:42
Where exactly does the segfault happen? What's the stack trace? — melpomene
– melpomene, Commented Feb 23, 2019 at 0:43
I have edited the question. The segmentation fault happens in both lapacke functions. The error message on terminal is simply Segmentation fault (core dumped) — Fra
– Fra, Commented Feb 23, 2019 at 0:45

francis · Accepted Answer · 2019-02-23 10:35:44Z

1

In LAPACKE_zgetrf(LAPACK_ROW_MAJOR, n, n, tempA , n,&n);, the last argument of LAPACKE_zgetrf points to n, a single integer. On the contrary, the argument ipiv should be a pointer to an array of integer of dimension max(m,n), to store pivot indices This could explain a segmentation fault.

The ipiv computed by LAPACKE_zgetrf() must also be provided to LAPACKE_zgetri() as input, to get the correct inverse matrix.

answered Feb 23, 2019 at 10:35

francis

9,8672 gold badges30 silver badges43 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Fra Over a year ago

I wish I could pay you a beer.

Collectives™ on Stack Overflow

LAPACKE matrix inversion segmentation fault C

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related