AT%N0187 openmoko echo patch?

Sarton O'Brien roguemoko at roguewrt.org
Tue Sep 2 01:46:05 CEST 2008


On Tuesday 02 September 2008 08:33:28 Jim Morris wrote:
> roguemoko at roguewrt.org wrote:
> > Cesar Eduardo Barros wrote:
> >> NeilBrown escreveu:
> >>> On Mon, September 1, 2008 6:59 pm, Yorick Moko wrote:
> >>>> On Mon, Sep 1, 2008 at 10:43 AM, Lorn Potter <lpotter at trolltech.com>
> >>>>
> >>>> wrote:
> >>>>> They certainly are not in our copy of the calypso spec's that we got
> >>>>> directly from TI. If it is, I must be missing it, and haven't found
> >>>>> it.
> >
> > But couldn't both of those have been documented? And if they have not
> > been, might be classed as incompetent. They are after all just part of
> > the modem instruction set.
> >
> > To not release the entire lot is a bit lame.
>
> Using an undocumented feature in a chip is very dangerous. Minor changes to
> the fab, even though the chip has the same number may change or remove that
> undocumented feature that the phones now rely on.

I don't know if it's so much as an undocumented feature as an undocumented 
command. I think the response and a subsequent lookup provides enough relevant 
info as to what the 'feature' is.

> If this is going to be adopted in builds, then someone at OM needs to get
> TI to officially support that feature so that it does not disappear later.

I tend to agree but I'd extend that to, OM should ensure that we are able to 
utilise the hardware effectively. A complete list of so-called documented and 
undocumented commands needs to be compiled or obtained by somebody, whether 
they are for development or not. After all, what the hell are we doing here?

Sarton




More information about the community mailing list