RE: Sensitivity to code changes in CW

I've never ran from external FLASH before - presumably you're
programming that correctly? CW has an option under debugger settings to
'verify memory writes' - I was going to suggest you use this to ensure
the FLASH has been programmed correctly, but if you're runnng from
FLASH, I expect the program download options for executable are
unchecked, so the tool wouldn't be writing and therefor wouldn't check,
so that probably doesn't help you.  At least run a verify from your
FLASH programmer.

I've had some problems with FLASH before when I had initialised strings
in the wrong place - make sure all your constant and initialised data
are in the correct place i.e. FLASH or RAM as appropriate.

You're not changing the chip select registers are you? You're not
supposed to change the CS registers while you're running from external
memory which is using that same CS region. If you want to do this, you
copy the instructions to internal RAM and execute from there.

Last suggestion as this one gave me similar types of issues. If you
write the SCI baud rate register, the upper 3 bits should be 0. I had
code that was writing non-0 to those bits and my program would crash at
different places depending what other code around it was doing. 

- Allan

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf
Of [email protected]
Sent: Thursday, December 22, 2005 9:40 AM
To: [email protected]
Subject: [MPC500] Sensitivity to code changes in CW


I'm having a heck of a time debugging a problem in which I'm seeing
extreme sensitivity to code changes in an application developed under
CW8.1. Starting from a fully functional build, the addition of a single
new line of application code can sometimes cause the program to
immediately exit. My inexperience with the CW debugger and the fact that
I'm running from external flash are combining to make this a tough
problem to solve. 
 
It appears that the program exits before main is called, i.e. the
programs exits during the startup phase. It does not appear to be
getting to the newly added line and the behavior is not sensitive to
what operations are done on the new line. It does not appear that code
size is an issue since I have experienced the problem with various major
code elements removed and present.  The problem can usually be relieved
by making even a minor change in some other area of the program. 
 
The fact that its repeatable, not associated with the details of the
particular change, and fixable by making some other change raises the
question of some type of alignment issue but I'm interested in comments
or ideas to try. I realize that asking for debugging help without many
details like this is not bound to net concrete answers but I wonder if
anyone has experienced this sort of issue either with CW or any other
toolset. 
TIA
 
 


[Non-text portions of this message have been removed]




-----------------------------------------------------------
To learn more about Freescale Microcontrollers, please visit
http://www.freescale.com/mcu


 
Yahoo! Groups Links



 




------------------------ Yahoo! Groups Sponsor --------------------~--> 
Fair play? Video games influencing politics. Click and talk back!
http://us.click.yahoo.com/u8TY5A/tzNLAA/yQLSAA/dN_tlB/TM
--------------------------------------------------------------------~-> 

-----------------------------------------------------------
To learn more about Freescale Microcontrollers, please visit
http://www.freescale.com/mcu


 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/MPC500/

<*> To unsubscribe from this group, send an email to:
    [email protected]

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/