On May 4, 2014, at 12:12 PM, Andreas Fritiofson <[email protected]>
wrote:
> Well, actually I think that it's common to put the minimum sram size for the
> family in the config file. Look at the stm32 configs for example.
>
Stated another way, the openocd configuration file states
MEM_RANGE( START, END ) = CLASS( R ) memory
There may be multiple ranges of CLASS(R)
Where R some type or class of memory
However in this case, we are demanding CLASS(F) be probed .
Rhetorical question: Why can’t CLASS(R) be probed let us treat all of these
classes as identical.
Do not forget, there is other classes of memory CLASS(P), and CLASS(N) and
Class(little-r) why can they not also be probed?
The answer is:
It is not always practical, nor is it possible.
Some chips - the memory is probable, and in others they cannot be
probed.
Some chips do not have class (F) maybe they only have class (R) and
class(little-r)
Or perhaps the sizes cannot be determined, or maybe each chip has to be
hard coded in C code.
I would argue that moving that “chip-specific” hard coding to a config
file is better, it makes C code more generic.
I’m being generic here my letters on purpose.
R = RAM
little-r = MASKED ROM (ie: a boot loader)
F = Flash
N = No mans land, i.e.: Nothing is there
P = Peripheral registers
E = External memory window
I also know that on some chips, certain memory ranges are only accessible as
32bit, or 16 bit modes.
How do we specify that type of of information in a memory class?
Hence, perhaps what is really required is an array of TARGET memory ranges
Where each range has a set of attributes …. i.e.: HW BREAKPOINT and RAM
BREAKPOINTS
And some attributes may be overlapping (i.e.: a bit field of types, each bit
identifies a class)
-Duane.
------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos. Get
unparalleled scalability from the best Selenium testing platform available.
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
OpenOCD-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openocd-devel
