June 30, 2010
11:46 a.m.
> Seems best of all worlds to me :-) > Can we get an ack and a commit? Anyone can ack. If you've reviewed it and tested it on your board, you'd be in the best position for that. Thanks, Myles > On 25 Jun 2010, at 20:54, Myles Watson wrote: > > > On Thu, Jun 24, 2010 at 7:08 AM, Edwin Beasant > > <edwin_beasant@virtensys.com> wrote: > >> Re-integrated the "USE_OPTION_TABLE" code so that it will compile > >> for all > >> boards that define it at configure time. > >> > >> Made USE_OPTION_TABLE dependant on HAVE_OPTION_TABLE > >> > >> Signed-off-by: Edwin Beasant edwin_beasant@virtensys.com > > I updated your patch to: > > - Have a default value for USE_OPTION_TABLE even when > > HAVE_OPTION_TABLE isn't defined > > - Pass abuild with USE_OPTION_TABLE defaulting to y or n > > > > use_options.diff: Your patch, slightly modified > > options_fixup.diff: Fix boards to include the .h file instead of > > the .c file > > have_option_polarity.diff: Default to no cmos.layout file so it can be > > selected (like most other options) > > > > Patch for testing with USE_OPTION_TABLE defaulting to y > > use_option_default.diff > > > > I think it would make sense to default to using the option table if it > > is present. > > > > A next step might be to remove the "default" parameter from > > read_option and make it get the default from option_table.h. > > > > Signed-off-by: Myles Watson <mylesgw@gmail.com> > > > > Thanks, > > Myles > > < > > use_options > > .diff > > > > > < > > options_fixup > > .diff><have_option_polarity.diff><use_option_default.diff>