IVT mismatch on i.MX 8MM EVK

取消
显示结果 
显示  仅  | 搜索替代 
您的意思是: 
已解决

IVT mismatch on i.MX 8MM EVK

跳至解决方案
731 次查看
kanimozhi_t
Contributor V

Hi,

    We've recently built Gatesgarth BSP for i.MX 8MM EVK. We can able to build the Yocto Linux image successfully.

    However when we try to sign the bootloader we found that the u-boot header is mismatching between gatesgarth and Zeus (or Warrior for that matter).

    The "Reserved 1" in IVT usually point the secondary loader's (SPL) header offset (load address) but in Gatesgarth it is 0.

 

So I'll summarise my questions as follows:

1. Is it intentional to return 0 on reserved 1 field instead of SPL load address?

2. Can we change the header data while building the u-boot? If so, how?

 

Feel free to revert for more details. Thanks in advance and any response would be highly appreciable.

标记 (5)
0 项奖励
1 解答
719 次查看
kanimozhi_t
Contributor V

I'm answering my own question here, for the future readers:

1. Is it intentional to return 0 on reserved 1 field instead of SPL load address?

    Yes, the reserved fields on U-Boot IVT are supposed to be 0. Refer AN4581 or this patch note

2. Can we change the header data while building the u-boot? If so, how?

    Yes, we can change the IVT while building the FIT image and it does not affect the FIT image structure. For updating the IVT, change the relevant fields on mkimage_imx8.c

在原帖中查看解决方案

0 项奖励
1 回复
720 次查看
kanimozhi_t
Contributor V

I'm answering my own question here, for the future readers:

1. Is it intentional to return 0 on reserved 1 field instead of SPL load address?

    Yes, the reserved fields on U-Boot IVT are supposed to be 0. Refer AN4581 or this patch note

2. Can we change the header data while building the u-boot? If so, how?

    Yes, we can change the IVT while building the FIT image and it does not affect the FIT image structure. For updating the IVT, change the relevant fields on mkimage_imx8.c

0 项奖励