Re: [PATCH 5/12] OMAP3430 support in mach-omap2 folder

[Tony Lindgren <tony@atomide.com>]

* Syed Mohammed, Khasim <x0khasim@ti.com> [070612 20:48]:
> >> > We have now common debug card interface. Check
> >> > arch/arm/plat-omap/debug-devices.c and do similar calls like
> >> > board-h4.c {debug_card_init}.
> >>
> >> For now, I would like to continue with this. I will incorporate this
> >comment as a separate clean up patch.
> >
> >Can you please fix what Trilok suggested? Remember, we want to use
> >the same patch for sending it upstream eventually.
> >
> I think it's not a correct approach to use debug_card_init for 2430 SDP. I have following view point.
> 
> 1. There is no debug card on 3430 or 2430 SDP as it was on H4.
> 
> 2. There will be #if's for 2430 and 3430 because many statements doesn't hold correct for 2430 / 3430
> 
>         smc91x_resources[0].start = addr + 0x300;
>         smc91x_resources[0].end   = addr + 0x30f;
> 
> The offset address for 3430 is 0x0 and 0x0f so, one if here.
> 
>         led_resources[0].start = addr;
>         led_resources[0].end   = addr + SZ_4K - 1;
> 
> These things are not required. If tomorrow we add more for 2430 / 3430 then this function has to be split anyway.
> 
> 3. The function is exported as an extern from board.h. Which is mainly meant for Nokia N series, this should be corrected. Why should we include board.h for other boards that might not use these structures.
> 
> 4. The Configuration doesn't show in Menuconfig as there is no statement next to bool.
> 	config OMAP_DEBUG_DEVICES
>         bool
>         help
>           For debug cards on TI reference boards.
> 
> The macro CONFIG_OMAP_DEBUG_DEVICES doesn't get enabled for 2430 / 3430.
> 
> 5. Having a complete debug_devices and corresponding file for 2430 / 3430 doesn't make sense, there are no LEDs or such debug devices on these boards. Unfortuanely SMSC is same, so its bit confusing I believe. But this Ethernet controller can be handled in a much simpler way, why to unnecessarily complex it?
> 
> Please don't take me in a defensive mood. I am trying to convince that this method doesn't hold good for 2430 / 3430. And what ever we were doing in posted patch is a more optimal approach. But if still you feel that this is what has to be done, then fine, I can add more ifs and elses and get the code working for 2430 / 3430.
> 
> Let me know your views at the earliest.

OK, thanks for explaining. Let's keep what you had, then later on we
can use the sdp-common.c for 2430-sdp and 3430-sdp.

Regards,

Tony