Fix #425: irold not updated in electron step loop

Assign the current particle region, ir(np), to the old region index,
irold, within the ustep loop in electron transport. Previously, irold
was not updated, so the irnew ~= irold condition added to fix magnetic
field transport in vacuum caused zero-size steps to skip the region
change macro. Remove the now redundant irold = ir(np) assignment in the
non-zero step size block.

HEN_HOUSE/src/egsnrc.mortran

@@ -1342,7 +1342,8 @@ $start_new_particle;
 
  $SET-USTEP-EM-FIELD; "additional ustep restriction in em field
  "default for $SET-USTEP-EM-FIELD; is ;(null)
- irnew = ir(np); "default new region is old region
+ irold = ir(np); "save current region
+ irnew = ir(np); "default new region is current region
  idisc = 0; "default is no discard (this flag is initialized here)
  ustep0 = ustep; "Save the intended ustep."
 
@@ -1381,7 +1382,6 @@ $start_new_particle;
  "(dnear is distance to the nearest boundary
  " that goes along with particle stack and
  " which the user's howfar can supply (option)
- irold = ir(np); "save previous region
  $SET-ANGLES-EM-FIELD;
  "default for $SET-ANGLES-EM-FIELD; is ; (null)
  "(allows for EM field deflection