[Patch, Fortran] assumed-rank some bound intrinsics support, fix failures and improve diagnostcs

[Tobias Burnus]

Dear all,

this patch is a cleanup follow up to the assumed-rank patch. Changes:
* Fix pattern matching for the dump. for some reason, the DTYPE differes 
between i686 and x86-64.
* There was a proper check missing that assumed-rank arrays may not have 
a codimension. (There are now checks in array.c, decl.c and - 
preexisting - resolve.c, all are triggered by the test case _11.)
* I re-added my lbound/ubound patch. It still only works with dim=.

For lbound/ubound without dim= and for shape, one has to modify the 
scalarizer a bit. See discussion at 
http://gcc.gnu.org/ml/fortran/2012-07/msg00032.html


Mikael: I wouldn't mind if you could have a look at the scalarizer - you 
seem to have an idea how one can implement it with minimal effort/code 
cluttering.


Build and regtested on x86-64-linux.
OK for the trunk?

Tobias

PS: Sorry for missing the regression failure before the committal.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: assumed-rank-bounds.diff
Type: text/x-patch
Size: 17166 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20120720/da39a205/attachment.bin>