svn commit: r289710 - in head/sys/modules: i2c/iicbb uart

[Warner Losh]

> On Oct 21, 2015, at 1:16 PM, Ed Maste <emaste at FreeBSD.org> wrote:
> 
> On 21 October 2015 at 14:30, Ed Maste <emaste at freebsd.org> wrote:
>> Author: emaste
>> Date: Wed Oct 21 18:30:42 2015
>> New Revision: 289710
>> URL: https://svnweb.freebsd.org/changeset/base/289710
>> 
>> Log:
>>  Build ofw_bus_if.h for modules that need it on arm64
> 
> Module Makefile build rules for ofw_bus_if.h are a bit of a mess; some
> modules build it unconditionally while others build it only on certain
> archs. There are a large number of different combinations of sparc64,
> mips, powerpc, arm, and aarch64. This is true for a number of other
> arch-specific module Makefile support. Should we have MK_* knobs to
> control these cases -- e.g. MK_OFW (MK_FDT?) here? That would at least
> move the arch-specific tests to a common location.

How much of the swamp do you want to drain?

None: Create MK_OFW (not MK_FDT: it already means something else). Add
a few lines to src.opts.mk. Watch me complain :)

A little: Add DEV_OFW to options. Add DEV_OFW to the relevant KERN_OPTS
clauses in sys/conf/config.mk and then test for it like so in the modules makefiles:
.if (${KERN_OPTS:MDEV_OFW})
…
.endif
Maybe add it in config.mk if MK_FDT is defined, but I’d prefer not
That’s what we’re really testing for here: is this kernel likely to have the OFW
routines being called.

A lot: rewrite config(8).

I’m hoping you’ll pick “a little” to not make things worse.

Warner


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 842 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.freebsd.org/pipermail/freebsd-arch/attachments/20151021/87ec7d40/attachment.bin>