On Fri, Dec 16, 2011 at 9:18 AM, Stefano Babic sbabic@denx.de wrote:
On 16/12/2011 16:59, Tom Rini wrote:
On Fri, Dec 16, 2011 at 8:37 AM, Stefano Babic sbabic@denx.de wrote:
The SPL is developped first for TI-OMAPx. The patch move OMAP specific function into OMAP directory.
I wonder if we should fold this into the mv'ing patch as well.
I can squash the two patches, but I thought the review is easier as I did. The previous patch shows clearly that the patch was only moved, and the changes in Makefile makes possible to build the boards making the patch bisectable. This patch then
Yeah, I guess that works too then, nevermind.
[snip]
diff --git a/include/spl.h b/include/spl.h +/* Boot type */ +#define MMCSD_MODE_UNDEFINED 0
Not your bad spacing of course, but please fix.
Thanks to have found, I will fix it.
+/* NAND SPL functions */ +void spl_nand_load_image(void); +void spl_arch_nand_init(void);
+/* MMC SPL functions */ +void spl_mmc_load_image(void);
These should be covered by #ifdef CONFIG_SPL_(NAND||MMC)_SUPPORT
Do we need in the header file for the prototypes ? We have never done, and it should be not necessary.
I would swear there's other examples like this (and someone brought this to my attention when i first posted spl_board_init internally).