Tim Rice
Mon, 18 Jan 2010 21:21:58 -0800
On Mon, 18 Jan 2010, Ralf Wildenhues wrote: > [ let's drop bug-autoconf ] > > Hi Tim, > > * Tim Rice wrote on Mon, Jan 18, 2010 at 09:46:07PM CET: > > On Mon, 18 Jan 2010, Ralf Wildenhues wrote: > > > > > Anyway, the failure of 211 is easy, and due to 'false' exiting 255 on > > > your system rather than 1 (which is documented in autoconf.texi). > > > Please confirm that the patch below fixes the issue; the test addition > > > is to ensure the respective code path for $srcdir != '.' is covered as > > > well. > > > Not quite there yet. > > Now 211 & 371 fail. > > log attached. > > 371 is another issue: your shell leaks temporary files and forgets to > clean them up. Please clean them up yourself at a convenient point. > If you want to help debugging your shell to find out whether it is > certain constructs that cause it to do so, it might be helpful if you > could look at the time stamps and contents of the leftover temp files. > Presumably most of them come from Autoconf or Libtool scripts.
Ah, I see what happened. Running "gmake check CONFIG_SHELL=/opt/bin/bash"
didn't use bash for everything. Some tests used ksh.
>
> Anyway, 211 should be helped with this incremental patch.
Yes it does the job.
I ran "gmake check TESTSUITEFLAGS=CONFIG_SHELL=/opt/bin/bash" and
all tests pass.
Thanks
> * lib/autoconf/status.m4 (AC_OUTPUT): Ensure to exit 1 in case
> of config.status failure.
>
P.S. What list do I need to be on to be notified prior to release so I
can test?
--
Tim Rice Multitalents (707) 887-1469
t...@multitalents.net