bug-autoconf
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: echo vs. printf regression (darwin8)


From: Ralf Wildenhues
Subject: Re: echo vs. printf regression (darwin8)
Date: Mon, 16 Aug 2010 21:53:18 +0200
User-agent: Mutt/1.5.20 (2010-04-22)

Hello,

* Paolo Bonzini wrote on Mon, Aug 16, 2010 at 09:05:40PM CEST:
> In order to avoid rewriting your makefiles, just do
> 
> AC_SUBST([ECHO], [echo])
> 
> for the time being.

I don't think that works currently.  Libtool sets $ECHO early in
configure and uses that when writing config.status.  When users
overwrite that, it may break the generated libtool script on some
systems.  Add to that that Libtool used to AC_SUBST([ECHO]).

I guess Libtool needs to be fixed to either not use ECHO, MV, and RM
any more in its configure macros, or claim ownership of these names.
They overlap with what other packages use and/or set.  One complication
is that $ECHO is used in some other libtool substitutions ($archive_cmds
etc) as well, which means that, when we rename to, say, $lt_ECHO, some
configure script trying to use $archive_cmds *at configure time* (after
LT_OUTPUT) will then break.  :-/

For $RM there has been a report recently, with another package using
  AC_SUBST([RM], [rm])

and Libtool expecting RM="rm -f" all over the place.

Whatever we do, NEWS needs a big warning note.

Cheers,
Ralf



reply via email to

[Prev in Thread] Current Thread [Next in Thread]