Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 5 Sep 2013 11:57:58 -0700
From:      "Simon J. Gerraty" <sjg@juniper.net>
To:        Alexey Dokuchaev <danfe@FreeBSD.ORG>
Cc:        freebsd-bugs@FreeBSD.ORG
Subject:   Re: bin/178819: bmake w/ WRKDIRPEFIX=/tmp breaks Ports Collection
Message-ID:  <20130905185758.7D04D5807E@chaos.jnpr.net>
In-Reply-To: <20130905175936.GB58718@FreeBSD.org>
References:  <201309030703.r8373BSU072280@freefall.freebsd.org> <20130905164552.3639E5807E@chaos.jnpr.net> <20130905175936.GB58718@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help

On Thu, 5 Sep 2013 17:59:36 +0000, Alexey Dokuchaev writes:
>Yes, I've seen the fresh import; will test it tomorrow.  I'm not sure what
>is the problem, but "ports should be able to control the value of MAKEFILE
>as desired" looks strange.  MAKEFILE is one of the standard bsd.port.mk's

bmake treats MAKEFILE the same way old sysV make did, it gets set to the 
name of the [mM]akefile read.
This is obviously something fmake never did, and ports relies on being
able to set the variable.

The new version of bmake allows for both behaviors by using an internal
context for the sysV behavior.  If the makefiles explicitly set a value
for MAKEFILE, that overrides the one that bmake itself would set.

>something else (e.g. internal bmake(1)'s variable)?  And how does this all
>got affected by WRKDIRPREFIX=/tmp?

I don't know - the PR body described a different issue - which clearly
needed fixing.





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20130905185758.7D04D5807E>