Subroutines or functions named as their containing module or program are 
reported as

subroutine abc()
             1
Error: MODULE attribute conflicts with PROCEDURE attribute at (1)

Attached patch checks whether a NAME is available. If yes, it is added to the 
message:

subroutine abc()
             1
Error: MODULE attribute of 'abc' conflicts with PROCEDURE attribute at (1)

A similar distinction as the one proposed is made in symbol.c:602.


:ADDPATCH fortran:

2007-04-30  Daniel Franke  <franke.daniel@gmail.com>

	* symbol.c (gfc_add_flavor): Add the NAME to error message if available.


As it is a minor clarification within a message only, no additional testcase 
is included.

Tested on i686-pc-linux-gnu, no regressions. Ok for mainline?

	Daniel


P.S. In general, should changes like this be backported to 4.2(.1) as well, 
although they are not actually regressions?