On 03/02/2012 12:37 PM, Troy Kisky wrote: > On 3/2/2012 10:45 AM, Stephen Warren wrote: >> On 03/02/2012 12:00 AM, Wolfgang Denk wrote: >>> Dear Troy Kisky, >>> >>> In message<4f4ffd1f.4080...@boundarydevices.com> you wrote: >>>>> Not sure I understand why we have a "machine_ix_xxx()" here. Could you> >>>>> explain? >>>>> >>>>> Amicalement, >>>> Because board/nvidia/seaboard/seaboard.c uses machine_is_ventana but> >>>> that machine is no longer >>>> in mach-types. >>> Then fix the seaboard.c code, please, instead of adding dead code. >> I think that should actually be very easy. Can't the following: >> >> if (machine_is_foo()) { >> ... >> } >> >> be converted to: >> >> #if CONFIG_MACH_TYPE == MACH_TYPE_FOO >> ... >> #endif >> >> As a related note, I expect you'll see more and more boards that don't >> exist in the kernel's mach-type.h since people won't register machine >> IDs for boards that only support device tree. >> > How about reversing the test > > > #if CONFIG_MACH_TYPE == MACH_TYPE_SEABORD > > instead of > > #if CONFIG_MACH_TYPE == MACH_TYPE_VENTANA > > so that I don't care that ventana is no longer defined?
That's probably fine. A comment right above that states that Ventana is the other possibility would be useful. Of course, this isn't going to solve anything when we have 3 variants of Seaboard, and only Seaboard is in mach-types.h but not the other 2, but I suppose we can deal with that when it happens. -- nvpublic _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot