[U-Boot] Using spi_alloc_slave() from SPL

[Simon Glass]

Hi,

On 6 August 2015 at 12:24, Fabio Estevam <festevam at gmail.com> wrote:
> On Thu, Aug 6, 2015 at 2:03 PM, Stefano Babic <sbabic at denx.de> wrote:
>
>> This is in RAM, of course. Increasing the size by 3KiB is not IMHO for
>> i.MX6 a problem, there is enough space in IRAM. But what is surprising
>> is that Fabio gets a Null pointer by malloc().
>
> Yes, so I did a simpler patch that shows the malloc() issue with SPL:
>
> --- a/board/freescale/mx6sabresd/mx6sabresd.c
> +++ b/board/freescale/mx6sabresd/mx6sabresd.c
> @@ -692,6 +692,7 @@ int checkboard(void)
>  #ifdef CONFIG_SPL_BUILD
>  #include <spl.h>
>  #include <libfdt.h>
> +#include <malloc.h>
>
>  const struct mx6dq_iomux_ddr_regs mx6_ddr_ioregs = {
>         .dram_sdclk_0 =  0x00020030,
> @@ -822,6 +823,7 @@ static void spl_dram_init(void)
>
>  void board_init_f(ulong dummy)
>  {
> +       void __iomem *ptr;
>         /* setup AIPS and disable watchdog */
>         arch_cpu_init();
>
> @@ -837,6 +839,10 @@ void board_init_f(ulong dummy)
>         /* UART clocks enabled and gd valid - init serial console */
>         preloader_console_init();
>
> +       ptr = malloc(64);
> +       if (!ptr)
> +               puts("*** malloc returned NULL\n");
> +
>         /* DDR initialization */
>         spl_dram_init();
>
> when I run it:
>
> U-Boot SPL 2015.07-08201-gfb44bcd-dirty (Aug 06 2015 - 15:19:54)
> *** malloc returned NULL
>
> Even if I put the malloc() after spl_dram_init() it still returns NULL.

Please check the README about the SPL flow. From what I can see
malloc() is not available before board_init_r() in SPL.

However, if you add a call to spl_init() from your board_init_f(),
then early malloc would be available. Enable this with
CONFIG_SYS_MALLOC_F_... as people explained.

It would be nice if someone could tidy this up so there is only one
generic board_init_f(), and it calls board-specific code from there.

Regards,
Simon