question with redboot

Hi all,

I'm now porting the redboot from a EVM board to my target board. However,
I've changed the flash and it seems doesn't work with my old redboot
configuration. I can only see the target board send some rubbish out from
the serial port. I don't know what to do next. I'm actually loading the
redboot thru a JTAG. In normal case, how does people solve this problem
since it is tedious to trace the assembly from my windows based debugger?

Thanks,
Eric

In article <c9hc5n$mq1$1@news.ust.hk>, eric <ericlist@hotmail.com> wrote:
>I'm now porting the redboot from a EVM board to my target board. However,
>I've changed the flash and it seems doesn't work with my old redboot
>configuration. I can only see the target board send some rubbish out from
>the serial port. I don't know what to do next. I'm actually loading the
>redboot thru a JTAG. In normal case, how does people solve this problem
>since it is tedious to trace the assembly from my windows based debugger?

Probably a baud rate problem.  Try various settings; 115k is now common,
as are 57600 and 9600.

If fiddling around with the settings in your terminal program doesn't
reveal anything, it's possible the serial port init is busted for
your board, and you get to debug that.
-- 
Steve Watt KD6GGD  PP-ASEL-IA          ICBM: 121W 56' 57.8" / 37N 20' 14.9"
 Internet: steve @ Watt.COM                         Whois: SW32
   Free time?  There's no such thing.  It just comes in varying prices...

Hi Watt,

Sorry, I did not describe the problem very clear. In fact, I know it is the
flash problem because I can actually boot up the board when I remove the
configuration of the flash I am using. When I just add the configuration of
my flash, the output skew up and I don't know how to debug it. Do you know
how can I check the problem in this stage?

Eric

"Steve Watt" <steve@nospam.Watt.COM> wrote in message
news:HynnKs.n8w@Watt.COM...
> In article <c9hc5n$mq1$1@news.ust.hk>, eric <ericlist@hotmail.com> wrote:
> >I'm now porting the redboot from a EVM board to my target board. However,
> >I've changed the flash and it seems doesn't work with my old redboot
> >configuration. I can only see the target board send some rubbish out from
> >the serial port. I don't know what to do next. I'm actually loading the
> >redboot thru a JTAG. In normal case, how does people solve this problem
> >since it is tedious to trace the assembly from my windows based debugger?
>
> Probably a baud rate problem.  Try various settings; 115k is now common,
> as are 57600 and 9600.
>
> If fiddling around with the settings in your terminal program doesn't
> reveal anything, it's possible the serial port init is busted for
> your board, and you get to debug that.
> -- 
> Steve Watt KD6GGD  PP-ASEL-IA          ICBM: 121W 56' 57.8" / 37N 20'
14.9"
>  Internet: steve @ Watt.COM                         Whois: SW32
>    Free time?  There's no such thing.  It just comes in varying prices...

In article <c9jf2h$j67$1@news.ust.hk>, eric <ericlist@hotmail.com> wrote:
[ top-post repaired ]
>"Steve Watt" <steve@nospam.Watt.COM> wrote in message
>   news:HynnKs.n8w@Watt.COM...
>> In article <c9hc5n$mq1$1@news.ust.hk>, eric <ericlist@hotmail.com> wrote:
>> >I'm now porting the redboot from a EVM board to my target board. However,
>> >I've changed the flash and it seems doesn't work with my old redboot
>> >configuration. I can only see the target board send some rubbish out from
>> >the serial port. I don't know what to do next. I'm actually loading the
>> >redboot thru a JTAG. In normal case, how does people solve this problem
>> >since it is tedious to trace the assembly from my windows based debugger?
>>
>> Probably a baud rate problem.  Try various settings; 115k is now common,
>> as are 57600 and 9600.
>>
>> If fiddling around with the settings in your terminal program doesn't
>> reveal anything, it's possible the serial port init is busted for
>> your board, and you get to debug that.
>
>Sorry, I did not describe the problem very clear. In fact, I know it is the
>flash problem because I can actually boot up the board when I remove the
>configuration of the flash I am using. When I just add the configuration of
>my flash, the output skew up and I don't know how to debug it. Do you know
>how can I check the problem in this stage?

I don't know what you mean by "remove the configuration of the flash".
Does that mean that you're restoring the old version of RedBoot?  If so,
it *still* sounds like a baud rate problem, since you probably have a
different number compiled in.  Or if you're restoring the RedBoot
configuration partition, the one you program in isn't set up correctly
for the RedBoot you're building.

If it's truly a problem with the flash load, I would be extremely
surprised to see any noise at all out the serial port, as such
fundamental problems tend to lock the processor up immediately.
-- 
Steve Watt KD6GGD  PP-ASEL-IA          ICBM: 121W 56' 57.8" / 37N 20' 14.9"
 Internet: steve @ Watt.COM                         Whois: SW32
   Free time?  There's no such thing.  It just comes in varying prices...

eric <ericlist@hotmail.com> enlightened us with:
>
>I'm now porting the redboot from a EVM board to my target board. However,
>I've changed the flash and it seems doesn't work with my old redboot
>configuration. I can only see the target board send some rubbish out from
>the serial port. I don't know what to do next. I'm actually loading the
>redboot thru a JTAG. In normal case, how does people solve this problem
>since it is tedious to trace the assembly from my windows based debugger?

I know this post is a few weeks old but I've only just seen it. In case
you didn't work it out, there are two possibilities: redboot stores 
baud rate configuration information in the flash. If you can erase the
flash entirely first, then RedBoot will pick up that it isn't initialised
(a checksum failure) and will fall back to its default baud rate, whatever
that is for your board.

The second possibility is that the data you are seeing goes something
like $T02:12345678;fedcba98;656536565654#ab or whatever numbers. That's GDB
remote protocol and means redboot has hit a processor exception for
whatever reason and bombed out. One of the numbers will be the PC
register in hex which you can look up in your redboot image or linker
map, or you can attach with GDB to find out in a more friendly way where
it stopped, assuming you have the source tree still around.

Jifl
-- 
--["No sense being pessimistic, it wouldn't work anyway"]-- Opinions==mine