Lpc2000

Hola~

1a) Anybody looked into boundschecking & test coverage tools, such as:

http://sourceforge.net/project/showfiles.php?group_id=88372

and

http://sourceforge.net/projects/covtool

by chance?

1b) Any GCC experts out there know if they would play with the embedded gcc
flavors as opposed to just the 'big machine' (linux/win32) flavors of gcc?

2) Anyone ever played with the NuMega, WindRiver or other professional stuff
for that matter on embedded compilers of any kind?  I think the level of
projects our community is dealing with is ready for this level
of verification - in my 'day job' I have to play with professional
verification tools for powerpc, and they @!#$%^ in terms of integration and
downright quirkiness - makes GCC internals & configuration look clean by
comparison, so needless to say I'd prefer something developed by a
'community', but I'm open to any sort of experiences y'all are willing to
share...(appologies if anyone out there works for Windriver - I'll
substantiate my flame with specifics if you want to talk)

Good evening,
Steve

--

Steve Franks
Staff Engineer
Franks Development, LLC
http://www.franks-development.com
(520) 312-0089


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

> 2) Anyone ever played with the NuMega, WindRiver or other professional
> stuff
> for that matter on embedded compilers of any kind?  I think the level of

The IAR compiler offers coverage analysis when debugging in simulation mode,
and code profiling when debugging target, or in simulation mode.  It would
be really useful to have coverage analysis when debugging target as well.  

Perhaps one of the JTAG vendors would implement some form of coverage
analysis that utilized the ETM.

Jerry care to comment?

Joel

(sorry for late reply)

> Steve Franks:

> 1a) Anybody looked into boundschecking & test coverage tools, such as:
> 
> http://sourceforge.net/project/showfiles.php?group_id=88372
> 
> and
> 
> http://sourceforge.net/projects/covtool
> 
> by chance?
> 
> 1b) Any GCC experts out there know if they would play with the embedded gcc
> flavors as opposed to just the 'big machine' (linux/win32) flavors of gcc?

i didn't try the GCC-patches or `covtool', but had a coursory glance
at the source. the GCC-patches seem to be of good quality, they add a
little overhead to the assembly-source to tally your array bounds. thus
your code gets bigger. you would have to patch up and compile GCC from
source to have an all new cross GCC for your platform. the documentation
is quite extensive. it lists limitations of the implementation (using
`goto', (set|long)jump(), function-pointers), but also specifically
lists ports to other architectures. i guess as long as there's a working
assembler, you get bounds-checking, and thassit!

seems you should try that stuff (and tell us about is :).

  clemens