[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20080618184124.GA24826@uranus.ravnborg.org>
Date: Wed, 18 Jun 2008 20:41:24 +0200
From: Sam Ravnborg <sam@...nborg.org>
To: Subbu Seetharaman <subbus@...verengines.com>
Cc: jeff@...zik.org, netdev@...r.kernel.org
Subject: Re: [PATCH 3/12] benet: net_object i/f functions, Makefile
On Wed, Jun 18, 2008 at 09:20:41AM -0700, Subbu Seetharaman wrote:
> Sam,
>
> We could not find any good way to eliminate the unsresolved
> symbol warnings with the message code separated as
> a module.
There should not be any when building in-kernel.
If you build the two modules as external modules you need
to use a few tricks to get access to the symbol.
See Documentation/kbuild/modules.txt - look
for KBUILD_EXTRA_SYMBOLS
> Also, is there a way to prevent
> configuration of "y" for benet and "m" for beclib ?
IIRC a similar case is already described in
Documentation/kbuild/kconfig-language.txt
But the simple solution is to let beclib be invisible
like this:
config BECLIB
tristate
config BENET
tristate "BE Module"
select BECLIB
depends on ...
Then BECLIB will have same value as BENET.
[Example only - use more appropriate names].
> What is the advantage of using benet-y instead
> of benet-objs ?
Lets say you have an optionla module that may
be selected in kconfig. Then you would do:
benet-y := benet-main.o benet-foo.o benet-bar.o ...
benet-$(OPTIONAL_FEATURE) += optional.o
> I do not see net driver makefile
> using *-y.
It is only recently I have started asking people to use -y for new stuff.
And no one has done any big effort to convert existing -objs users to -y
because that would mostly be code chrunch for little gain.
Sam
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists