[GRASS-dev] Re: Probable bug fix

[Hamish]

Jyothish wrote:
> > I have a small problem in understanding how the vect
> > variable is being declared... in function tcholInv and
> > tcholSolveInv.
Hamish:
> i.e. array used but no space never allocated for it?
> looks like a bug to me.
Jyothish:
> And I think the initial author missed a 
> vect=G_alloc_vector(n);
> T=G_alloc_matrix(n, BW);

ok. ... proposed patch:


Index: vector/lidar/lidarlib/TcholBand.c
===================================================================
--- vector/lidar/lidarlib/TcholBand.c   (revision 37978)
+++ vector/lidar/lidarlib/TcholBand.c   (working copy)
@@ -1,8 +1,7 @@
 #include <stdlib.h>            /* imported libraries */
 #include <stdio.h>
 #include <math.h>
-#include <grass/gis.h>
-#include <grass/PolimiFunct.h>
+#include "PolimiFunct.h"
 
 /*--------------------------------------------------------------------------------------*/
 /* Tcholetsky decomposition -> T= Lower Triangular Matrix */
@@ -129,8 +128,8 @@
     double somma;
 
        /*--------------------------------------*/
-    G_alloc_matrix(n, BW);
-    G_alloc_vector(n);
+    T = G_alloc_matrix(n, BW);
+    vect = G_alloc_vector(n);
 
     /* T computation                */
     tcholDec(N, T, n, BW);
@@ -177,8 +176,8 @@
     double somma;
 
        /*--------------------------------------*/
-    G_alloc_matrix(n, BW);
-    G_alloc_vector(n);
+    T = G_alloc_matrix(n, BW);
+    vect = G_alloc_vector(n);
 
     /* T computation                */
     tcholDec(N, T, n, BW);



Hamish