Hi all,
TF-A recently introduced the FIP ("Firmware Image Package") which is a
format for embedding other bootloader images as payloads in a single
archive, which is then read and processed by TF-A.
In platforms such as the stm32mp1 where TF-A is used in conjunction with
U-Boot, U-Boot is now embedded in the FIP image as part of the TF-A package
build.
To achieve this we added a dependency in the TF-A rule files:
select U_BOOT if TF_A_FIP
The goal is to be able to build the fip command with a command similar to
this one:
make ARM_ARCH_MAJOR=7 ARCH=aarch32 PLAT=stm32mp1 \
BL33=<u-boot_path>/u-boot-nodtb.bin \
BL33_CFG=<u-boot_path>/u-boot.dtb \
BL32=<tfa_path>/bl32.bin \
FW_CONFIG=<tfa_path>/fw-config.dtb \
DTB_FILE_NAME=<fdt file name>.dtb \
fip
The problem is that the dependency set in the tf-a.in file only ensures
that the install target of u-boot will run before the install target of
tf-a, but at that time the images have not yet been copied to their final
destination (platform/images..). Which is the proper way to work around
this?
Thanks,
--
Guillermo Rodriguez Garcia
[email protected]
_______________________________________________
ptxdist mailing list
[email protected]
To unsubscribe, send a mail with subject "unsubscribe" to
[email protected]