U-Boot SPL binary does not read BIN header arguments, so passing some dummy values 0000005b and 00000068 has no effect for U-Boot SPL code.
Probably these two values comes from old Marvell DDR training code which was separated from U-Boot and used it for some configuration.
Seems that two 32-bit values were specified here to ensure SPL code alignment to 128-bit boundary as it is required e.g. for A370 or AXP processors. Main kwbimage header is 64-byte long which is aligned to 128-bit boundary. Optional kwbheader is 32-bit long, number of BIN header arguments is stored in 32-bit number. So for alignment to 128-bit boundary is needed 64-bit padding which exactly these two 32-bit dummy arguments provided.
Now when mkimage correctly aligns start of executable code in BIN header to 128-bit boundary, there is no requirement to put dummy argument values into kwbimage. So remove them.
Signed-off-by: Pali Rohár pali@kernel.org --- arch/arm/mach-mvebu/kwbimage.cfg.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/arm/mach-mvebu/kwbimage.cfg.in b/arch/arm/mach-mvebu/kwbimage.cfg.in index 72e67d75c325..049d23c6ef08 100644 --- a/arch/arm/mach-mvebu/kwbimage.cfg.in +++ b/arch/arm/mach-mvebu/kwbimage.cfg.in @@ -9,4 +9,4 @@ VERSION 1 #@BOOT_FROM
# Binary Header (bin_hdr) with DDR3 training code -BINARY spl/u-boot-spl.bin 0000005b 00000068 +BINARY spl/u-boot-spl.bin