This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [patch, libstdc++, testsuite] Modify tests to use tmpnam()
- From: Paolo Carlini <paolo dot carlini at oracle dot com>
- To: sje at cup dot hp dot com
- Cc: gcc-patches at gcc dot gnu dot org, libstdc++ at gcc dot gnu dot org
- Date: Wed, 03 Mar 2010 23:15:21 +0100
- Subject: Re: [patch, libstdc++, testsuite] Modify tests to use tmpnam()
- References: <201003032135.o23LZur08153@lucas.cup.hp.com>
Hi,
> I took a look at two libstdc++ tests that are failing for me on IA64
> HP-UX and found that the problem was related to running the tests over
> NFS. Both tests are using mkfifo to create pipe's in the current
> working directory and that works for me on a local filesystem but fails
> over NFS. I am not sure if that is an HP-UX bug or not, but I would
> like to change the tests to use tmpnam to create the pipes, this fixes
> the problem on IA64 HP-UX and causes no regressions. There are already
> tests (27_io/headers/cstdio/macros.cc and
> 27_io/headers/cstdio/functions_std.cc) that check for the existence of
> the L_tmpnam macro and the tmpnam function so using them in these
> tests shouldn't be a problem.
>
> OK for checkin?
>
Ok. Actually, it seems to me we should use tmpnam more consistently -
after all it's a perfectly standard C89 facility - instead of inventing
names here and there for temporary files. Can you add a one-line comment
about that before the arrays?
Thanks,
Paolo.