Re: regression in autoconf-2.62 vs. 2.61

[Stepan Kasal]

Hello,

I'd like to propose a solution, please speak up if you are not
satisfied with it.  (If we reach an agreement, I'm willing to prepare
a patch later.)

First, let me state that, strictly speaking, this is not a
regression.  The Autoconf manual says that the #undef line cannot
contain anything after the symbol.   And in that case, all versions
tested by Ralf do the same.

Does anyone know whether comments on #undef lines and #define lines
are portable?  I always thought that #undef comments were not
portable and #define comments could at least became part of the body
of the macro definition.  I thought this were the reason why these
comments were generally frown upon.
If that were true, than I would prefer deleting the rest of the line
after #undef since it might make the resulting header more portable.

About the case of #undef OTHER_SYMBOL being commented out, I suggest
to preserve that behaviour, or someone will get trapped by that
change and report a regression.  ;-)

I suggest to add a note to the manual along these lines:

``
Generally speaking, all #define and #undef directives in the header
template may be modified by config.status, under some circumstances.
Consequently, if you need to define or undefine a symbol under some
circumstances (#ifdef THIS #define THAT, etc.), you should place
those directives outside the config headers.  If you absolutely need
to hook that to the config header itself, please put these directives
to a separate file, and #include it from the config header template.
(If you are using autoheader, you would probably use AH_TOP or
AH_BOTTOM to inject the #include directive.

For certain symbold it is important to avoid #undef ... [explanation
about _POSIX_WHATEVER].
To achive this, current Autoconf comments out _all_ remaining #undef
directives remaining in the config header at the end of the
instatiation, but this may change in future releases.
''

WDYT?

Stepan