0

TL;DR: How many ways can the bootm command fail with this error in version 1.1.4?:

# bootm
Booting image at 31000000
Bad Magic Number

Background:

I've bricked a device whose U-Boot is deeply crippled as far as its ability to load bytes from external sources, so I've written a quick script to automatically generate mm commands details here (including healthy boot messages) to populate the RAM and write that back to the flash with nandw and that seems to be working but the bootm command is still grumpy.

The binary data is a valid uImage file and is written into address 0x31000000 and it does have the magic number of 0x27051956 at that memory address but the system is cowardly refusing to boot. Here's the whole header from hexdump -C of the uImage file:

00000000  27 05 19 56 f4 91 0f 3f  4c 11 3c 57 00 17 a3 40  |'..V...?L.<W...@|
00000010  30 00 80 00 30 00 80 00  f4 67 23 80 05 02 02 00  |0...0....g#.....|
00000020  4c 69 6e 75 78 2d 32 2e  36 2e 31 38 2e 32 2d 6e  |Linux-2.6.18.2-n|
00000030  74 78 36 30 30 2d 76 30  2e 38 32 00 00 00 00 00  |tx600-v0.82.....|
00000040  00 00 a0 e1 00 00 a0 e1  00 00 a0 e1 00 00 a0 e1  |................|

I'm not sure what I'm doing wrong here, but I'm expecting to at least get to the CRC check.

  • It looks like I messed up the endianess of the data I was storing so that's the answer here. In looking at the source code, it looks like it's only the magic number in the header. – RandomInsano Feb 18 at 4:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.