>-----Original Message-----
>From: Mike Frysinger [mailto:[email protected]]
>Sent: Wednesday, January 20, 2010 11:27 AM
>To: Song, Barry
>Cc: [email protected]
>Subject: Re: [Linux-kernel-commits] [8199]
>trunk/arch/blackfin: blackfin xip: push the 64 byte header
>into the build system and let the xip version use arbitrary
>entry points too now that the mkimage util doesnt require a
>specific -e with the -x option
>
>On Tue, Jan 19, 2010 at 22:22, Song, Barry wrote:
>>From: Mike Frysinger [mailto:[email protected]]
>>>On Tue, Jan 19, 2010 at 21:54, Song, Barry wrote:
>>>> I think you should test before you check-in.
>>
>> Good. Thanks. Should be just a typo...
>>
>>>i guess it's a good thing i did:
>>>## Booting kernel from Legacy Image at 20040000 ...
>>> Image Name: bf537-2.6.32.4-ADI-2010R1-pre-gc
>>> Created: 2010-01-19 23:54:17 UTC
>>> Image Type: Blackfin Linux Kernel Image (uncompressed)
>>> Data Size: 2902096 Bytes = 2.8 MB
>>> Load Address: 20040040
>>> Entry Point: 200f8534
>>> Verifying Checksum ... OK
>>> Loading Kernel Image ... OK
>>>OK
>>>Starting Kernel at = 200f8534
>
>i went through the u-boot code some more as this output didnt look
>exactly right to me ... turns out there is some assumptions in the
>bootm code. if the entry point isnt at the bootm address + 64bytes,
>it'll assume a non-xip kernel. so when it says "loading", it actually
>means it's doing a memcpy. it should have said "XIP ...". i'll poke
>around u-boot to try and get this fixed.
Sounds good. So XIP should be just a flag in image header, not need offset
check :-)
>-mike
>
_______________________________________________
Linux-kernel-commits mailing list
[email protected]
https://blackfin.uclinux.org/mailman/listinfo/linux-kernel-commits
