Re: bootstrap failure (libstdc++ testsuite_hooks)

[Andreas Jaeger <aj at suse dot de>]

Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at> writes:

> On Fri, 20 Jun 2003, Andreas Jaeger wrote:
>> This is on mainline?  I successfully bootstrapped and run the
>> testsuite yesterday several times on both i686-linux and x86_64-linux.
>
> Yes.
>
>> What's the output of:
>> /tmp/OBJ-0619-2203/i686-pc-linux-gnu/libstdc++-v3/scripts/testsuite_flags --build-cxx
>> ?
>
>   % /files/pfeifer/OBJ-0620-1125/i386-unknown-freebsd4.8/libstdc++-v3/scripts/testsuite_flags --build-cxx
>   %
>
> No output at all, because -- as I just realized -- the script is
> completely empty -- filesize zero.  Checking my build log, I also
> realize why!
>
>   creating scripts/testsuite_flags
>   sed: /sw/test/gcc/cvs/libstdc++-v3/scripts/testsuite_flags.in: No such file or directory
>
> There seem to be two problems here:
>
>   1) The scripts Gaby, me (and surely several others) use to build GCC
>      failed to update the new directories from CVS; this is most probably
>      related to us not using the -d option of `cvs update` globally to
>      avoid getting Ada.

Ok, problem solved.  You should really use scripts/gcc_update since it
does use -dP

>   2) Still, failure to locate this file should have aborted bootstrap
>      immediately; would you mind fixing that?

This file is created as part of a configure script and we do not make
anywhere in GCC (at least AFAIK) a check that all files do exist, it's
expected that they're there.

Andreas
-- 
 Andreas Jaeger, SuSE Linux AG, aj@suse.de, http://www.suse.de/~aj
  GPG fingerprint = 93A3 365E CE47 B889 DF7F  FED1 389A 563C C272 A126

Attachment: pgp00000.pgp
Description: PGP signature