This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: Vectorizing HIRLAM 4: complicated access patterns examined.
- From: Tobias Schlüter <tobias dot schlueter at physik dot uni-muenchen dot de>
- To: Toon Moene <toon at moene dot indiv dot nluug dot nl>
- Cc: jakub at redhat dot com, fortran at gcc dot gnu dot org
- Date: Wed, 02 Nov 2005 20:24:13 +0100
- Subject: Re: Vectorizing HIRLAM 4: complicated access patterns examined.
- References: <E1EXNlR-00013r-SG@laptop.moene.indiv.nluug.nl>
Toon Moene wrote:
> Nope, I checked. This has nothing to do with equivalences.
The issue in PR18833 was primarily the correct identification whether a(1:2)
is a substring or an array reference, so this might still be related. The
patch only worked around this issue in the special case of equivalence
statements, so it might be the case that we hit another case where this same
issue arises, but I'm not sure as I haven't looked in the debugger.
This is a slightly further reduced testcase:
PROGRAM TABLES
IMPLICIT CHARACTER*8(Y)
!WRITE(YLOCAL,'(I2.2)') ILOCAL
!ybtable='b'
YBTABLE=Ylocal(1:2)
END
Try some combination of removing comments and removing the substring
specifier, and you'll get different error messages or success rates.
Can you open a PR please?
- Tobi