duane> Can we get confirmation about how the JRC works in OMAP?
kees> I am sure there must be exceptions to your rules but I guess that
having this 3rd state will make it very easy to implement what we need
right now
kees>Why are those assumptions important? Why is is hard to "insert" an
element in the chain?
Today - the items are sort of numbered in a linked list.
And there are things that refer to each other.
If we add the idea of a "3rd state" - then things work like this:
A tap can be in one of 3 states:
Inserted - and full operational, thus the length is >= 1 bit long.
Inserted - and in bypass, and thus is exactly 1 bit long
Or -
Disabled - [You refer to it as "removed" or "de-inserted"]
That alone can greatly simplify some things in OpenOCD - when
calculating prefix shift lengths (ie: for the Nth item in the chain - is
the sum of all previous item lengths in the chain) we add 0, 1, or N,
today - the idea of "ZERO" length was never considered. If the order
never changes - you can see the simplicity of the solution.
The problem becomes if we "re-sort" that list. yes, it could be done -
but that is a lot of muck to clean up.
Not sure if the American usage of "muck" translates well for all those
on the list - or maybe - it does...
Think of a large barn - with many horses - "mucking is not always a
pleasant task"
see Definition (1)(a)
http://www.merriam-webster.com/dictionary/mucking
_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development
