[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: AC_CHECK_TARGET_TOOL behavior
From: |
Eric Blake |
Subject: |
Re: AC_CHECK_TARGET_TOOL behavior |
Date: |
Thu, 14 Aug 2008 21:15:19 +0000 (UTC) |
User-agent: |
Loom/3.14 (http://gmane.org/) |
Ralf Wildenhues <Ralf.Wildenhues <at> gmx.de> writes:
> The bit fixing most of the issues you reported is the missing '$' in
> AC_CHECK_TARGET_TOOL. I think the macros are wrong in warning about a
> previous AC_CANONICAL_TARGET, however. Shouldn't it rather be that they
> should require it?
Thanks for the analysis, and yes, I agree that an AC_REQUIRE is correct
(AC_BEFORE implies that the use of AC_CANONICAL_TARGET is optional albeit
ordering-constrained; but in the case of checking target tools,
AC_CANONICAL_TARGET is not optional). Wow, that missing $ has gone undetected
for quite some time: since its intruduction in 2004 (per commit e09a76, as
found in less than half a second by
$ git log -S'=ac_c' lib/autoconf/programs.m4
Have I ever mentioned how much I like git? That sort of historical analysis is
very time-consuming in CVS).
>
> I'm still looking for a good testsuite addition for these macros.
>
> OK to apply?
Yes; even without a testsuite addition, the change looks obviously correct by
inspection. If you later come up with a test, more power to you.
--
Eric Blake