[Morphix-devel] Unforeseen Opportunity for Improvement in morphixinstaller?

[Mark G. <ma...@tr...>]

Hello,

I've been experiencing a segmentation fault when I try and run=20
morphixinstaller on my system.  After carefully examining the code, the=20
segmentation fault appeared to occur in the BuildRoot function of instlib.c=
=20
where the code is examining the existing fstab file.  I was able to confirm=
=20
that was the section with the problem by deleting the existing /etc/fstab o=
n=20
my system.  With no /etc/fstab, the code skips the section and doesn't=20
trigger the segmentation fault.

While I'm far from a competent C programmer, I did notice that the int used=
 as=20
a loop counter (gint j) is not initialized in the code.  When I changed the=
=20
line 517 of instlib.c from:

gint i, j;

to:

gint i =3D 0;
gint j =3D 0;

the problem went away.

I know I saw at least one other user on the forums with this issue.  Are th=
ere=20
any reasons we shouldn't make this change?

Regards,

Mark Gannon

=2D-=20
"I am an idealist. I don't know where I am going,=20
but I am on my way."=20

=2D Carl Sandburg=20