[U-Boot] [PATCH v2] [TESTING] da8xx: fixup ARM relocation support

[Sudhakar Rajashekhara <sudhakar.raj@ti.com>]

Hi Ben,

On Thu, Aug 19, 2010 at 18:32:48, Ben Gardiner wrote:
> Hi Sudhakar,
> 
> Thank you for reviewing this patch.
> 
> On Thu, Aug 19, 2010 at 2:25 AM, Sudhakar Rajashekhara
> <sudhakar.raj@ti.com> wrote:
> > Ben Gardiner <bengardiner <at> nanometrics.ca> writes:
> >
> > [...] snip
> >
> >> +
> >> +void dram_init_banksize (void)
> >> +{
> >> + ? ? gd->bd->bi_dram[0].start = CONFIG_SYS_SDRAM_BASE;
> >> + ? ? gd->bd->bi_dram[0].size = gd->ram_size;
> >> +}
> >> +#endif
> >>
> >
> > dram_init() gets called from arch/arm/lib/board.c and initializes gd->ram_size
> > but who is calling dram_init_banksize to initialize gd->bd->bi_dram[0].start
> > and gd->bd->bi_dram[0].size?
> 
> As of commit a716b323f10d4f2bce6b4ae01f6d1544e5781ed8 on the
> arm-reloc-and-cache-support branch of
> git://git.denx.de/u-boot-testing.git , dram_init_banksize() is called
> from board_init_f which is executed just before relocation.
> 

Though you have mentioned in your changelog, I just want to make sure
that gd->ramsize will be ZERO only when CONFIG_SYS_ARM_WITHOUT_RELOC is
defined. Is it right? Otherwise get_ram_size() needs to be called in both cases.

I was also about to comment about the line length in your patch but Heiko has
given the same comment. Please fix that and re-submit the patch. I can post
Acked-by/Reviewed-by on that version.

Regards,
Sudhakar