Date: Sat, 28 Dec 2013 17:56:16 +0100 From: John Marino <freebsd.contact@marino.st> To: Tijl Coosemans <tijl@FreeBSD.org> Cc: svn-ports-head@freebsd.org, svn-ports-all@freebsd.org, John Marino <marino@FreeBSD.org>, Brendan Fabeny <bf@FreeBSD.org>, ports-committers@freebsd.org Subject: Re: svn commit: r337822 - head/math/parmetis Message-ID: <52BF02B0.2050807@marino.st> In-Reply-To: <20131228174341.77deb435@kalimero.tijl.coosemans.org> References: <201312272256.rBRMuojP063283@svn.freebsd.org> <20131228174341.77deb435@kalimero.tijl.coosemans.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 12/28/2013 17:43, Tijl Coosemans wrote:
> On Fri, 27 Dec 2013 22:56:50 +0000 (UTC) John Marino wrote:
>
> gmake does set MAKE correctly. You can test that with a simple
> Makefile like:
>
> all:
> @echo ${MAKE}
>
> The problem is that the code Makefile runs "make -C $(BUILDDIR) $@
> $(MAKEFLAGS)". MAKEFLAGS is not meant to be passed on the command
> line like that because it doesn't contain command line flags. It only
> contains single letters so "d" instead of "-d".
> Bmake sets this to "w" when it calls gmake like gmake does for its own
> submakes to get Entering/Leaving messages. This happens in do-build
> with USES=gmake.
>
> With the attached patch the port still builds for me.
Given that I have spent several hours on this port which isn't even
mine, I'm sick of looking at it. I didn't find out that ${MAKEFILES}
was the problem until the end. (I'm also now pretty confused when to
use $(MAKE) and when to use $(MAKE_CMD)).
In any case, this is bf@'s call. While your patch might be better, I
consider my fix "good enough" since the final result is the same. I am
CC'ing him, thanks.
John
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?52BF02B0.2050807>