GTA01 shared UART vs. flow control (bug #788)

Mike Westerhof mwester at dls.net
Tue Jan 22 18:39:13 CET 2008


 On Tue 22/01/08  8:46 AM , Werner Almesberger werner at openmoko.org sent:
> Andy Green wrote:
> > This problem is GTA01 -only?
> 
> Yes.
> 
> > Why don't we pick the most complete solution and accept
> that we have to> carry it as our own patch until a better way is
> found?
> The discussion is about what constitutes a complete solution :)
> 
> What we have now is:
> 
> 1) a relatively simple but known to be incomplete solution, which
> may indeed trip up in real life,
> 2) a more complex one that we'll never be able to merge upstream
> in its present form,
> 3) and a proposal for an entirely different approach that should
> avoid either limitation.
> 
> To me, 2) seems to have too poor a cost/benefit ratio to be a candiate
> for a patch we want to maintain locally forever, with the cost being
> fragility and having unmergeable code in the first place, and the
> benefit being what it offers beyond 3).

I'm right now imagining the response from RMK to a patch submitted upstream that drops characters on the floor whenever hardware flow control is asserted on the console port, the justification being to solve a hardware problem in a single, limited-run, pre-production special-purpose device.

> I've proposed a simple solution that should solve the problem for
> all practical purposes. So unless a flaw is found in that approach,
> I wouldn't want anything that's worse and/or causes more problems.
> 
> (Of course, given the controvery around the other approaches towards
> solving this, I had to propose something different. How else could
> I make sure to equally offend everyone ? ;-)

Actually this issue is now being discussed, so I'm not offended :P

> - Werner

Mike (mwester)





More information about the openmoko-kernel mailing list