And that causes m4's internal macros like "esyscmd" to fail at starting subprocesses, it shows up while trying to emerge subversion with this version of m4. A patch exists on freebsd ports: http://www.freebsd.org/cgi/cvsweb.cgi/ports/devel/m4/files/patch-lib_spawn.in.h
Just for tracking, will drop -fbsd keywords on this version, 1.4.13 works okay. @base-system: up to you to apply the patch or not.
Thanks for the heads up, I just masked this version in Prefix' freebsd profile.
anything that is in lib/ isnt specific to m4; it comes from gnulib. if the change is already in the upstream gnulib code, feel free to commit whatever you need.
Just curious, inheriting eutils to patch stuff here wouldn't mess bootstrapping? (or something else like a circular dep?)
As far as I'm concerned, usage of epatch early in the process is fine.
`patch` should be available in stage1, so there should never be a bootstrapping issue with using `epatch`. if there is, that's a bug in the release process.
Created attachment 225067 [details, diff] Patch to the ebuild. Already on upstream gnulib: http://git.savannah.gnu.org/gitweb/?p=gnulib.git;a=commit;h=88baa1ed437db0f062e49b87cff59002ec8f7ab2 And please check the patch to the ebuild, will commit it later if no objections. Thanks.
Fixed in CVS.