This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Patch, Fortran] PR 44929: Parsing error of derived type name starting with 'REAL'


Hi all,

here is a simple patch which fixes the PR in the subject line. It is
very similar in concept to the patch that has lately been committed by
Steve as r162325 (and has been withdrawn later), i.e. it moves the
parsing of derived types before the parsing of intrinsic types, to
avoid confusion when derived type names start with intrinsic type
strings (such as "real_type").

My patch improves Steve's earlier patch in two details: It fixes a
whitespace issue, which was the cause of the failure reported by
Tobias in comment #2. And it ensures to correctly error out, avoiding
false error messages (and the need to modify allocate_derived_1.f90).

I think the regressions reported in PR 45005 after Steve's commit were
an unrelated OOP issue, which has probably been fixed by Mikael's
recent patch for PR 42051/44064 (thanks for that).

The patch has been regtested on x86_64-unknown-linux-gnu. Ok for
trunk? (And 4.5?)

Cheers,
Janus



2010-07-30  Janus Weil  <janus@gcc.gnu.org>

	PR fortran/44929
	* match.c (match_type_spec): Try to parse derived types before
	intrinsic types.


2010-07-30  Janus Weil  <janus@gcc.gnu.org>

	PR fortran/44929
	* gfortran.dg/allocate_derived_3.f90: New.

Attachment: pr44929_v2.diff
Description: Binary data

Attachment: allocate_derived_3.f90
Description: Binary data


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]