2009-08-27 10:46:36     u-boot hangs on a custom BF527' board

Document created by Aaronwu Employee on Sep 24, 2013
Version 1Show Document
  • View in full screen mode

2009-08-27 10:46:36     u-boot hangs on a custom BF527' board

Nikolay Chokoev (IRELAND)

Message: 79304   

 

Hello,

 

 

 

I've downloaded the latest u-boot release and tried to run it on our BF527 based bard.

 

The output is:

 

bfin> loadb

## Ready for binary (kermit) download to 0x01000000 at 57600 bps...

## Total Size      = 0x000303a8 = 197544 Bytes

## Start Addr      = 0x01000000

bfin> go 0x1000000LMNOPQST>

 

and then hangs.

 

I verified that the SDRAM is initialized with the correct parameters (64M), and the PLL settings seems to be correct.

 

The toolchain is 2009R1-RC10

 

The u-boot is 2009R1-RC3

 

The config settings are like bf527ezkit...

 

... what could be wrong...

 

Regards,

 

Nikolay

QuoteReplyEditDelete

 

 

2009-08-27 11:31:14     Re: u-boot hangs on a custom BF527' board

Nikolay Chokoev (IRELAND)

Message: 79309   

 

.... sometimes the output is:

 

bfin> go 0x1000000

## Starting application at 0x01000000 ...

Early:start.S: Init Registers

Early:start.S: Find ourselves

ABCDEGHIJKLMNOPQST>ram Clocks

QuoteReplyEditDelete

 

 

2009-08-27 16:10:48     Re: u-boot hangs on a custom BF527' board

Mike Frysinger (UNITED STATES)

Message: 79320   

 

are you using a LCD splash screen or anything ?

 

that sequence indicates the reprogramming worked fine, but when returning to external memory, things crashed.  you might want to double check your memory settings.  use the documentation:

http://docs.blackfin.uclinux.org/doku.php?id=bfin:sdram

 

do you have jtag to confirm ?  that should let you read external memory before reprogramming things and then read it again after, but before executing out of it.

QuoteReplyEditDelete

 

 

2009-08-27 16:28:45     Re: u-boot hangs on a custom BF527' board

Nikolay Chokoev (IRELAND)

Message: 79321   

 

I'm using the LCD splash screen, but I tried with "#define CONFIG_VIDEO" commented out (that should not build the LCD support I think?) and it's the same...

 

The memory is the same as this in the BF527-EZKIT (well, the clock is a little bit faster (131.25MHz in EZKIT vs. 132MHz in my case), but that shouldn't affect much...).

 

The only JTAG I have is HPUSB-ICE with the VDSP. I've checked when the u-boot hangs, that the 'new' PLLs and SDRAM settings are OK, and I can access the whole 64MB RAM space from the JTAG...

QuoteReplyEditDelete

 

 

2009-08-27 16:39:45     Re: u-boot hangs on a custom BF527' board

Mike Frysinger (UNITED STATES)

Message: 79323   

 

you need to `make clean` after changing your board configuration

 

with the JTAG, you should do a dump of where the new u-boot is loaded before executing the initcode and then a dump after executing the initcode and make sure they're the same

QuoteReplyEditDelete

 

 

2009-08-27 17:49:18     Re: u-boot hangs on a custom BF527' board

Nikolay Chokoev (IRELAND)

Message: 79324   

 

This with the memory dump is a good idea. As I understand I have to dump the 0x1000000, run the initcode (with 'go 0x1000000') and then dump 0x1000000 again? Is that correct?

QuoteReplyEditDelete

 

 

2009-08-27 21:51:47     Re: u-boot hangs on a custom BF527' board

Mike Frysinger (UNITED STATES)

Message: 79327   

 

if you loaded u-boot to that address, that'd be the place to dump

 

the initcode executes in L1, so you might want to edit initcode.c and add:

asm volatile ("emuexcpt;");

 

to the end of the function so it'll trigger jtag before returning and (presumably) crashing

QuoteReplyEditDelete

 

 

2009-08-28 04:25:03     Re: u-boot hangs on a custom BF527' board

Nikolay Chokoev (IRELAND)

Message: 79346   

 

Hi Mike,

 

Thanks for your help.

 

The problem was - in the 'new' u-boot I'm changing the SDRAM settings and the SDRAM data gets corrupted...

 

Regards,

 

Nikolay

QuoteReplyEditDelete

 

 

2009-08-28 06:21:34     Re: u-boot hangs on a custom BF527' board

Mike Frysinger (UNITED STATES)

Message: 79354   

 

i guess it'd help to flesh out the initcode a bit more so it could handle/report exceptions.  hopefully i can get that done in this release cycle.

QuoteReplyEditDelete

Attachments

    Outcomes