[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Chicken-hackers] LDFLAGS and Makefile.linux
From: |
Felix |
Subject: |
Re: [Chicken-hackers] LDFLAGS and Makefile.linux |
Date: |
Fri, 30 Dec 2011 13:01:22 +0100 (CET) |
Hello!
> I've taken over the care and feeding of the Chicken package for Gentoo
> Linux, and as part of my new duties I've been clearing out some of the
> bugs that have been accumulating. One of the bugs[1] involves
> Makefile not consulting the user's LDFLAGS environment variable.
>
> Gentoo best-practices dictate that the *FLAGS variables be respected
> by build scripts because giving this level of control over their
> packages is the reason they use Gentoo in the first place.
>
> However, I understand that Chicken's build process is unique among
> software, and wanted to ask the hackers what the consequences would be
> of including the user's LDFLAGS into Makefile.linux's
> LIBCHICKEN_SO_LINKER_OPTIONS variable. Are there some LDFLAGS which
> would result in a Chicken that can't cross the road? Or would cause
> it to cross too slowly for safety?
There are probably numerous LDFLAGS that could break the build. I would
suggest to run "make" like this:
make ... LINKER_OPTIONS="$LDFLAGS"
>
> I've noticed that some makefiles have a variable LINKER_OPTIONS (e.g.
> Makefile.macosx, Makefile.mingw, Makefile.cygwin), but not all of
> these incorporate its value into LIBCHICKEN_SO_LINKER_OPTIONS. Since
> I've only ever built Chicken on Linux systems I can't be sure what
> reasoning there is behind this design.
AFAIK, LINKER_OPTIONS is not inorporated into
LIBCHICKEN_SO_LINKER_OPTIONS anywhere. The intention for the latter is
to only hold options specific for linking libchicken.so.
cheers,
felix