> On Dec 19, 2014, at 16:15, Ian Campbell <ijc@debian.org> wrote:
>
> On Fri, 2014-12-19 at 13:38 +0800, Chen Baozi wrote:
>> I’ve only booted the system by ‘bootz’ without initrd. If I load the raw
>> initrd image (rather than ‘uInitrd”), when I use bootz, it would output:
>>
>> Wrong Ramdisk Image Format
>> Ramdisk image is corrupt or invalid
>>
>> This is the reason that I’m still using bootm when initrd is needed.
>
> bootz requires you to give the filesize for the raw initrd. e.g.
> load $kernel_addr_r kernel
> load $fdt_addr_r
> load $ramdisk_addr_r
> bootz ${kernel_addr_r} ${ramdisk_addr_r}:${filesize} ${fdt_addr_r}
>
> The ":${filesize}" is what I mean, it is implicitly set by load (and
> similar commands) which is why initrd is loaded last in the above, so it
> doesn't get clobbered.
Aha, the CONFIG_SUPPORT_RAW_INITRD is not defined by default when building
u-boot with omap5_uevm_defconfig. That is why I used to fail booting with
‘bootz’.
With the latest u-boot enabling CONFIG_SUPPORT_RAW_INITRD, the following
db description works:
Machine: TI OMAP5 uEVM board
Method: generic
U-Boot-Script-Name: bootscr.uboot-generic
Boot-Script-Path: /boot/boot.scr
Required-Packages: u-boot-tools
DTB-Id: omap5-uevm.dtb
Baozi.
Attachment:
signature.asc
Description: Message signed with OpenPGP using GPGMail