Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 12 Mar 2008 11:46:15 +0100
From:      Alexander Leidinger <Alexander@Leidinger.net>
To:        pav@FreeBSD.org
Cc:        emulation@FreeBSD.org, cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, ports-committers@FreeBSD.org
Subject:   Re: cvs commit: ports/Mk bsd.port.mk bsd.port.subdir.mk bsd.xorg.mk
Message-ID:  <20080312114615.tcssgxahkck04c8s@webmail.leidinger.net>
In-Reply-To: <1205316953.51779.13.camel@pav.hide.vol.cz>
References:  <200803112345.m2BNj4Hj051138@repoman.freebsd.org> <20080312110416.rvewfpfrqcgswo0c@webmail.leidinger.net> <1205316953.51779.13.camel@pav.hide.vol.cz>

next in thread | previous in thread | raw e-mail | index | archive | help
Quoting Pav Lucistnik <pav@FreeBSD.org> (from Wed, 12 Mar 2008 =20
11:15:53 +0100):

> Alexander Leidinger p=C3=AD=C5=A1e v st 12. 03. 2008 v 11:04 +0100:
>> Quoting Pav Lucistnik <pav@FreeBSD.org> (from Tue, 11 Mar 2008
>> 23:45:04 +0000 (UTC)):
>>
>> >   - Teach USE_LDCONFIG to do the right thing when used with  =20
>> USE_LINUX_PREFIX
>> >
>> >   PR:             ports/118212
>> > http://www.FreeBSD.org/cgi/query-pr.cgi?pr=3D118212
>> >   Original patch: vd
>> >   Patch by:       pav
>>
>> I would like to know why emulation@ was not in the loop for this.
>>
>> At
>> http://www.leidinger.net/FreeBSD/port-patches/Mk:bsd.port.mk.diff
>
> I wasn't aware of this patch. Is it mentioned in a PR assigned to
> portmgr anywhere?

No PR. I mentioned it once on emulation@ I think. But it's been a =20
while. I would have expected that a linux related patch would be =20
presented for review on emulation@.

> If I ignore not-ldconfig related changes in your patch, the
> functionality should be equivalent. I do not need to set
> NO_LDCONFIG_MTREE, because that code branch is never reached.

Good.

> My code goes like if (linux) { do linux stuff } else { do native stuff }
> Your code went like if (linux) { setup some variables } else { setup
> some variables} endif; do { common stuff }.

Yes (regarding the behavior of my patch).

> Only difference I see is that you condition the use of linux mdconfig by
> presence of USE_LINUX, I need USE_LINUX_PREFIX, which is more strict, I
> think.

It depends upon your point of view:
  - we have ports which use USE_LINUX but not USE_LINUX_PREFIX
  - every port (except the base ports) needs to use USE_LINUX
    (the base ports are very special, we use ldconfig by hand
    there anyway)
  - Ports which don't install into LINUXBASE have no business
    using USE_LDCONFIG (libs which need to be cached for
    the run-time linker need to go into LINUXBASE)

So it would be stricter in my point of view to use USE_LINUX. I can =20
understand when you say, that it is stricter by using USE_LINUX_PREFIX =20
(a different point of view).

The question is now, if USE_LINUX_PREFIX is sufficient. Do we have =20
ports which install into LOCALBASE and LINUXBASE, and if yes, do they =20
install libs into LINUXBASE? emulation@ does not own a port which =20
installs into both, but emulation@ does not own all linux ports. The =20
nvidia driver installs into both, but does it _need_ (which is !=3D =20
"use") USE_LDCONFIG (I haven't checked)?

Every port which installs into both locations, needs to use USE_LINUX, =20
but may not use USE_LINUX_PREFIX. So I would say it is at least =20
failsave to make this dependent upon USE_LINUX. It may be beneficial =20
to also check the argument for USE_LDCONFIG in the linux case. Only =20
"yes" (or similar) makes sense here.

BTW: Thanks for working on this, it's on my TODO list for too long.

Bye,
Alexander.

--=20
Some people cause happiness wherever
they go; others, whenever they go.

http://www.Leidinger.net    Alexander @ Leidinger.net: PGP ID =3D B0063FE7
http://www.FreeBSD.org       netchild @ FreeBSD.org  : PGP ID =3D 72077137



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080312114615.tcssgxahkck04c8s>