Fix an array bound mismatch.

Description

Fix an array bound mismatch.

In the case of k-point symmetries, together with a k-point path, the copy routine was leading to an array bound mismatch, as the arrays of the original grids are allocated to the number of points in the full Brillouin zone, and not to the irreducible one.

The range of the DOS was also computed taking into account k-points in the path, which is now solved.

News snippet

Fix an array bound mismatch. The range of the DOS is now independent of the k-points in the path.

Checklist

• I have checked that my code follows the Octopus coding standards
• I have added tests for all the new features added in this request.

src/system/dos.F90

@@ -25,6 +25,7 @@ module dos_oct_m
   use global_oct_m
   use hamiltonian_elec_oct_m
   use io_oct_m
+  use kpoints_oct_m
   use mesh_oct_m
   use messages_oct_m
   use mpi_oct_m
@@ -63,17 +64,22 @@ module dos_oct_m
 
 contains
 
-  subroutine dos_init(this, namespace, st)
+  subroutine dos_init(this, namespace, st, sb)
     type(dos_t),         intent(out)   :: this
     type(namespace_t),   intent(in)    :: namespace
     type(states_elec_t), intent(in)    :: st
+    type(simul_box_t),   intent(in)    :: sb
 
     FLOAT :: evalmin, evalmax, eextend
+    integer :: npath
 
     PUSH_SUB(dos_init)
 
-    evalmin = minval(st%eigenval)
-    evalmax = maxval(st%eigenval)
+    !The range of the dos is only calculated for physical points,
+    !without the one from a k-point path
+    npath = kpoints_nkpt_in_path(sb%kpoints)
+    evalmin = minval(st%eigenval(1:st%nst, 1:(st%d%nik-npath)))
+    evalmax = maxval(st%eigenval(1:st%nst, 1:(st%d%nik-npath)))
     ! we extend the energy mesh by this amount
     eextend  = (evalmax - evalmin) / M_FOUR