public inbox for libc-hacker@sourceware.org
 help / color / mirror / Atom feed
* AUTOCONF not set when --with-cvs
@ 2003-03-21 17:58 Andreas Schwab
  2003-03-26  3:43 ` Roland McGrath
  0 siblings, 1 reply; 2+ messages in thread
From: Andreas Schwab @ 2003-03-21 17:58 UTC (permalink / raw)
  To: Roland McGrath; +Cc: libc-hacker

Why is the autoconf check not supposed to be run when --without-cvs?
Apart from the fact that the test is backwards the autoconf rule in
Makefile is always active even when --without-cvs.  Or perhaps you had
this in mind?

--- configure.in.~1.385.~	2003-03-19 10:48:20.000000000 +0100
+++ configure.in	2003-03-21 16:26:45.000000000 +0100
@@ -670,6 +670,8 @@ AC_CHECK_PROG_VER(SED, sed, --version,
   SED=: aux_missing="$aux_missing sed")
 
 if test "x$with_cvs" != xyes; then
+  AUTOCONF=no
+else
   AC_CHECK_PROGS(AUTOCONF, autoconf, no)
   case "x$AUTOCONF" in
   xno|x|x:) AUTOCONF=no ;;


But I would prefer to keep the autoconf test independent of --without-cvs.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: AUTOCONF not set when --with-cvs
  2003-03-21 17:58 AUTOCONF not set when --with-cvs Andreas Schwab
@ 2003-03-26  3:43 ` Roland McGrath
  0 siblings, 0 replies; 2+ messages in thread
From: Roland McGrath @ 2003-03-26  3:43 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: libc-hacker

> Why is the autoconf check not supposed to be run when --without-cvs?
> Apart from the fact that the test is backwards the autoconf rule in
> Makefile is always active even when --without-cvs.  Or perhaps you had
> this in mind?

That is obviously what I had in mind when I made that change.  I was
evidently not thinking too clearly.  I never ever use --with-cvs, because I
find the notion of automatic implicit commits just too wrong.  But I do
still want the autoconf runs in my builds (I notice when they happen, check
that they are sane, and commit them by hand).

The motivation for the tweak to configure was that the autoconf check takes
quite a while relative to other tests (it takes a few seconds on my ppc
box), because it simply tries autoconf on configure.in rather than any
simpler test.  Now that I have autoconf-2.5x installed as the first
autoconf in my path, I would be happy with getting rid of this check and
letting people lose when they touch configure.in files without the right
autoconf installed or the right AUTOCONF= value in their configure command.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2003-03-26  3:42 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-03-21 17:58 AUTOCONF not set when --with-cvs Andreas Schwab
2003-03-26  3:43 ` Roland McGrath

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).