
On 12/07/2011 04:39 PM, Stefano Babic wrote:
On 06/12/2011 19:34, Simon Schwarz wrote:
From: Simon Schwarzsimonschwarzcor@googlemail.com
This implements booting of Linux from NAND in SPL
Related config parameters: CONFIG_SYS_NAND_SPL_KERNEL_OFFS Offset in NAND of direct boot kernel image to use in SPL CONFIG_SYS_SPL_ARGS_ADDR Address where the kernel boot arguments are expected - this is normally RAM-start + 0x100 (on ARM)
Signed-off-by: Simon Schwarzsimonschwarzcor@gmail.com
Hi Simon,
diff --git a/arch/arm/cpu/armv7/omap-common/spl_nand.c b/arch/arm/cpu/armv7/omap-common/spl_nand.c index 38d06b1..db52879 100644 --- a/arch/arm/cpu/armv7/omap-common/spl_nand.c +++ b/arch/arm/cpu/armv7/omap-common/spl_nand.c @@ -24,6 +24,7 @@ #include<asm/u-boot.h> #include<asm/utils.h> #include<asm/arch/sys_proto.h> +#include<asm/io.h> #include<nand.h> #include<version.h> #include<asm/omap_common.h> @@ -32,6 +33,7 @@ void spl_nand_load_image(void) { struct image_header *header;
- int *src, *dst; switch (omap_boot_mode()) { case NAND_MODE_HW_ECC: debug("spl: nand - using hw ecc\n");
@@ -45,26 +47,56 @@ void spl_nand_load_image(void)
/*use CONFIG_SYS_TEXT_BASE as temporary storage area */ header = (struct image_header *)(CONFIG_SYS_TEXT_BASE); +#ifdef CONFIG_SPL_OS_BOOT
- if (!spl_uboot_key()) {
Using a gpio or whatever mechanis is baord specific. What about to set a weak function and let that the board decides itself what to do ?
Sounds reasonable to me. So I'am ok with this.
Not always a GPIO can be used for this purpose.
I have noted another issue. If the kernel image is not found, the fallback is to load u-boot. However, u-boot is searched at the address in NAND of the kernel, not at CONFIG_SYS_NAND_U_BOOT_OFFS.
This means that if the kernel is not loaded at all, the SPL starts to start a u-boot.bin instead of u-boot.img loaded at CONFIG_CMD_SPL_NAND_OFS, and this fails.
/*
* load parameter image
* load to temp position since nand_spl_load_image reads
* a whole block which is typically larger than
* CONFIG_CMD_SAVEBP_WRITE_SIZE therefore may overwrite
* following sections like BSS
*/
nand_spl_load_image(CONFIG_CMD_SPL_NAND_OFS,
CONFIG_CMD_SPL_WRITE_SIZE,
(void *)CONFIG_SYS_TEXT_BASE);
/* copy to destintion */
for (dst = (int *)CONFIG_SYS_SPL_ARGS_ADDR,
src = (int *)CONFIG_SYS_TEXT_BASE;
src< (int *)(CONFIG_SYS_TEXT_BASE +
CONFIG_CMD_SPL_WRITE_SIZE);
src++, dst++) {
writel(readl(src), dst);
}
/* load linux */
nand_spl_load_image(CONFIG_SYS_NAND_SPL_KERNEL_OFFS,
CONFIG_SYS_NAND_PAGE_SIZE, (void *)header);
spl_parse_image_header(header);
So if the parse function fails, we should have the fallback to u-boot, exactly as we have now the fallback from u-boot.img to u-boot.bin.
Hm, I don't think that we want any fallback here. I would prefere an error message and hang. The direct boot is designed to be used in the field - so IMHO we don't want to start the bootloader automatically if a normal startup fails.
The function you write are in omap-common/spl_nand.c, but they are not specific for OMAP, and any architecture can profit from your work. What about to move common functions away from TI related directory ? I do not know which is the best place, maybe a new directory under the root as spllib ?
Hm thats generally right. I would prefer lib_spl. Somehow it feels like going back to nand_spl times. Duno if there are better places?
The first two points I will fix/rework hopfully on monday (I'am not in the next two days)
Regards Simon