[prev in list] [next in list] [prev in thread] [next in thread] 

List:       autoconf-bug
Subject:    Re: autotest problems with m4_defun and autoconf-2.64+
From:       Eric Blake <ebb9 () byu ! net>
Date:       2009-12-02 0:47:12
Message-ID: 4B15B910.9020000 () byu ! net
[Download RAW message or body]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to Mike Frysinger on 12/1/2009 5:12 PM:
> i have a package that uses autotest and after upgrading to 2.64, it started 
> issuing these errors:
> 
> $ autom4te --language=autotest -o testsuite.tmp testsuite.at
> testsuite.at:4: error: m4_divert_push: cannot change diversion to `GROW' 
> inside m4_expand
> testsuite.at:4: the top level
> autom4te-2.65: /usr/bin/m4 failed with exit status: 1
> make: *** [testsuite] Error 1
> 
> the testsuite.at code looks like this:
> 
> $ cat testsuite.at
> AT_INIT
> m4_defun([PCALC_FILESDIR], [$srcdir])
> AT_SETUP(001)
> AT_CHECK([pcalc @PCALC_FILESDIR/pcalc.001], [0], [ignore])
> AT_CLEANUP

Hmmm.  This is indeed a regression, caused by the fact that AT_CHECK now
wants to pre-expand its arguments.  Thanks for the report.

I'll have to think about this one.  Maybe it is still possible to get
m4_expand and m4_defun'd macros to play nicely, but it won't be trivial.

m4_defun exists mainly for cases where you want to guarantee ordering with
m4_require, as well as for ease in checking whether a macro has been
previously expanded via m4_provide_if.  For this simple example, you are
doing neither, so the lighter-weight m4_define is indeed the right
workaround.  But I don't know that I can (or want) to make the blanket
statement that m4_defun should blindly be replaced with m4_define in autotest.

> with a little more helpful error message ?  i was using m4_defun() in a 
> completely different file and erroring on the AT_CHECK() line didnt really 
> clue me into the actual source of the problem.

At least that task might be easier - m4_expand currently doesn't like
m4_defun'd macros attempting diversion magic, but even if I can't fix
m4_expand to play nicely with diversions, I might be able to still make it
be smarter and include the name of the m4_defun'd macro attempting the
diversion magic.  At any rate, I'll try to tackle that first.

- --
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9@byu.net
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAksVuRAACgkQ84KuGfSFAYACGgCgvt+I6LPH0nx7p5CXiRibZhsU
xDUAn2W+NtSrZdHl6BBufDlqcvmvTg0u
=SCNk
-----END PGP SIGNATURE-----



[prev in list] [next in list] [prev in thread] [next in thread]