Re: [TEST] make -j patch [take 2]

On Mon, 15 Nov 2004, Alexander Leidinger wrote:

AL>Zitat von Harti Brandt <harti_at_freebsd.org>:
AL>
AL>[knu CCed because he should know how portupgrade operates :-)]
AL>
AL>> Unless you reset MAKEFLAGS along the call path to the portupgrade's make
AL>> they'll see the -j, because the top-level make will stuff the -j into
AL>> MAKEFLAGS and that is probably inherited through portupgrade.
AL>
AL>I don't know how ruby handles exec()ing of external programs, but unless
AL>it inherits the environment by default, portupgrade doesn't seems to
AL>inherit MAKEFLAGS ("grep MAKEFLAGS /usr/local/lib/ruby/site_ruby/1.8/*
AL>/usr/local/sbin/port*" shows no hits).
AL>
AL>So if portupgrade inherits MAKEFLAGS somehow, phk's patch shouldn't
AL>cause unexpected harm in this szenario, if portupgrade doesn't inherit
AL>MAKEFLAGS, his patch violates POLA in this case.

At least portinstall doesn't touch MAKEFLAGS: insert something like
FOO!=echo -- ${MAKEFLAGS} >/tmp/A
into a port's makefile and call portinstall for than port:

MAKEFLAGS=-j2 portinstall ...

Would be strange if portupgrade would be different from portinstall. So 
nothing is broken that wasn't already.

harti