Re: [Openocd-development] 0.4.0-RC1 status ...

Mon, 11 Jan 2010 18:19:30 -0800

David Brownell a écrit : 
On Sunday 10 January 2010, Michel Catudal wrote:
  
I still have a problem with STM32 where the first load doesn't work. 
    

I don't remember seeing a description of such an issue ... can
you post one, or an URL to a description in the mailing list?

When you say "still", that suggests it's a problem that's been
around for a long time, not a recently introduced bug.  It might
be practical to use "git bisect" to track it down, if you know
some version where it once worked.  (Look at "git log" history
for the old SVN commit IDs, if you need to go back that far.)

That problem has existed as long as I can remember. I have seen no improvement in the latest versions.
Since no one seemed to have that problem, Like Freddie I assumed the problem was with me.
It has never worked correctly so I can't give you a version where it worked.

The problem is from a power up. It could be a problem with the clock. Talking about problems with the clock
it seems like a new problem has occured but I won't say for sure until I do more tests. My clock seems to
come to a crawl when debugging. I start at 72Mhz, does the debugger switch the speed to a lower one?


  

    
There is also issues with interrupt tracing without disabling interrupts 
completely and watchdog.
    

  

  

I believe I've heard about the former (see below); not the latter.
More info?


  

  

    
Have these issues been covered recently? Maybe just some parameters to 
pass to gdb that I am not aware of.
    

  

  

Someone posted GDB event handlers they successfully used to step
through some code with IRQs active, ISTR "cortex_m3 maskisr" did
the job, if a bit unnaturally.  Look at list archives for December,
in the "Cross Platform Debugger" thread.

  





Then the interrupts no longer work. I have to read a bit more in gdb to
see how to go around this.

I guess it would be too much a burden to put that into OpenOCD. I need
to find a clean way.





  
Was this the IRQ tracing issue you refer to?  With "gdb -x" used to
load the definitions below.  I was planning to look at this and
maybe transfer this information into the User's Guide.

- Dave

  



That is correct. As for the watchdog, once you enable it you can no
longer debug as it keep reseting.

I will test this on the weekend to see if this is still true. Right now
for debugging I have conditional that disables the watchdog if I want
to debug.

I have the same problem with IAR so that is not an openocd problem.





  
  

  

    
Am 15.12.2009 16:51, schrieb Alain Mouette:
    

    

      
Can anyone help me with this?

To make single-step work better with IRQs and FreeRTOS in Eclipse, it
seems that I need to send this do GDB:

define hook-step
mon cortex_m3 maskisr on
end
define hookpost-step
mon cortex_m3 maskisr off
end
      

    

  



What calls these hooks? gdb?






-- 
Tired of Microsoft's rebootive multitasking?
then it's time to upgrade to Linux.
http://home.comcast.net/~mcatudal





_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development






















					Reply via email to