Re: [RFC PATCH 1/3] kbuild: Warn on selecting symbols with unmet direct dependencies

[Andres Salomon <dilinger@queued.net>]

On Fri, 11 Jun 2010 13:31:19 -0700
Randy Dunlap <randy.dunlap@oracle.com> wrote:

> On Tue, 08 Jun 2010 17:25:57 +0100 Catalin Marinas wrote:
> 
> > The "select" statement in Kconfig files allows the enabling of options
> > even if they have unmet direct dependencies (i.e. "depends on" expands
> > to "no"). Currently, the "depends on" clauses are used in calculating
> > the visibility but they do not affect the reverse dependencies in any
> > way.
> > 
> > The patch introduces additional tracking of the "depends on" statements
> > and prints a warning on selecting an option if its direct dependencies
> > are not met.
> 
> 
> Hi Catalin,
> 
> Can these messages (on linux-next-20100611) be modified to include the
> kconfig symbol that is causing them?
> 
> warning: (IP_VS_PROTO_ESP && NET && NETFILTER && IP_VS || IP_VS_PROTO_AH && NET && NETFILTER && IP_VS) selects IP_VS_PROTO_AH_ESP which has unmet direct dependencies (UNDEFINED)
> warning: (SCx200_GPIO && SCx200 || PC8736x_GPIO && X86) selects NSC_GPIO which has unmet direct dependencies (X86_32)
> 
> Ah!  It's the first symbol listed in each "phrase":
> 
> SCx200_GPIO and PC8736x_GPIO both have this problem.
> (I haven't looked at IP_VS yet).

It's unclear why NSC_GPIO depends upon X86_32 (davej sent the patch,
699352c30da8525a).  While geode stuff will probably only ever exist on x86,
there's nothing in the driver that's X86_32-specific.


> 
> Jordan, is GEODE always 32-bit, so that some of these dependencies could be
> cleaned up by using X86_32?

Sadly, I don't think Jordan tracks this list any more.