Date: Sun, 18 Jun 2017 04:50:26 +0200 From: Jan Beich <jbeich@FreeBSD.org> To: Gerald Pfeifer <gerald@pfeifer.com> Cc: ports-committers@freebsd.org, svn-ports-all@freebsd.org, svn-ports-head@freebsd.org Subject: Re: svn commit: r443793 - head/lang/gcc7-devel Message-ID: <shiy-2eml-wny@FreeBSD.org> In-Reply-To: <alpine.LSU.2.21.1706180321130.2749@anthias.pfeifer.com> (Gerald Pfeifer's message of "Sun, 18 Jun 2017 03:25:27 %2B0200 (CEST)") References: <201706172327.v5HNRbGE015821@repo.freebsd.org> <d1a2-9k83-wny@FreeBSD.org> <alpine.LSU.2.21.1706180321130.2749@anthias.pfeifer.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Gerald Pfeifer <gerald@pfeifer.com> writes: > On Sun, 18 Jun 2017, Jan Beich wrote: > >>> When we set STAGE1_CXXFLAGS / CXXFLAGS in MAKE_FLAGS for arm platforms, >>> do so incrementally (with += instead of =). >> Did you mean to use MAKE_ENV instead of +=? > > I'm not sure. ;-) In your example, += did not actually work in > the MAKE_ENV tests, did it? Correct, += syntax isn't supported by sh(1) and += syntax in gmake(1) acts differently between on command line and in a file. > Did you mean that the commit message should have referred to > MAKE_ARGS instead of MAKE_FLAGS? MAKE_ARGS usage looks wrong as it discards system CXXFLAGS i.e., -O2 -fno-strict-aliasing -pipe. Replacing = with += won't fix it. Whether there's a bug or not requires understanding the intent. With commit messages that state only "what" was done, but not "why", one can only guess the former. Testing can help deriving the intent but in my case poudriere fails to create armv6 jail (unlike aarch64). Or perhaps you can answer "Why do you need the incremental assignment?"
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?shiy-2eml-wny>