[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Bad pwd detection in generated 'configure'
From: |
Pavel Raiskup |
Subject: |
Re: Bad pwd detection in generated 'configure' |
Date: |
Tue, 14 May 2013 13:01:52 +0200 |
User-agent: |
KMail/4.10.2 (Linux/3.8.11-200.fc18.x86_64; KDE/4.10.2; x86_64; ; ) |
> > Would there be possible to use something like this? It would pretty
> > easily work-around the issue:
> >
> > # When building in place, set srcdir=.
> > -if test "$ac_abs_confdir" = "$ac_pwd"; then
> > +if test "`$readlink -f $ac_abs_confdir`" = "`$readlink -f $ac_pwd`"; then
>
> Sorry, but 'readlink' is not portable.
>
> That said, if we can come up with a PORTABLE construct that checks
> whether the current directory matches srcdir, with minimal subprocesses,
> I'm not opposed to a patch.
- But still, e.g. $readlink could fall-back to empty operation if no
'readlink' binary is found.
- I was also thinking about `pwd -P`, or `cd -P`. These seem to be
defined in POSIX.
> Best may be just fixing your build environment to never try and set PWD,
> since that is already a non-portable operation.
Do you think that it is worth to fix this for the symlink case? I did not
described it properly before:
$ ls /tmp/ -l | grep source
drwxrwxr-x. 2 user user 40 May 14 12:47 source
lrwxrwxrwx. 1 user user 6 May 14 12:47 source_link -> source
$ cd /tmp/source_link
$ /tmp/source/configure
In such that case, the same issue occurs .. configure fails and the
'TESTFILE' used by 'AC_CONFIG_LINKS(TESTFILE:TESTFILE)' is lost.. It is
quite unpleasant and hard to recover situation (but unlikely, though).
If we found a portable enough solution, it would be nice to have it fixed,
imo. I could give it a and write a patch if you don't want to (if it was
considered as a bit useful).