This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: Large number of new Fortran failures on AIX
- From: François-Xavier Coudert <fxcoudert at gmail dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: fortran at gcc dot gnu dot org
- Date: Thu, 28 Jul 2005 17:54:58 +0200
- Subject: Re: Large number of new Fortran failures on AIX
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=lp8Pg0F4shQYccQuB9Tg7NTOIYk/vBhvQUTdWKq3ej0tQa/Dl/he69b+NNjKBLSp/ESkdP6Mcb45HaD9ftXwHsy2h/5ezTjqWVEHEWxyVr/+yfHv/cs5xsxZNIyB+hSaY+OCiotKo4lbVonrK+HPrpCzlL9iZBnIAtabgffJ+v8=
- References: <200507281358.j6SDwAd29902@makai.watson.ibm.com> <fxcoudert@gmail.com> <19c433eb0507280839246d0774@mail.gmail.com> <200507281545.j6SFj3d31348@makai.watson.ibm.com>
- Reply-to: François-Xavier Coudert <fxcoudert at gmail dot com>
> The previous testsuite run was about 12 hour before. There were
> no fortran or libgfortran changes during that period, except possibly
> Volker's diagnostic patch.
> There were a number of C patches during that time, so it is
> possible that libgfortran is being miscompiled for AIX. libgfortran does
> not use mmap for I/O on AIX, unlike most of the other systems.
Oh, then there's not much help I can offer. You can always set a
breakpoint in the library (running a file that issues the "Record
number too large" error) to see what are the values of ioparm.rec and
current_unit->maxrec at line 1143 of libgfortran/io/transfer.c.
FX