Hi Cyril, On Wed, May 19, 2010 at 20:35:58, Chemparathy, Cyril wrote: > Hi Sekhar, > > [...] > >>> Also, can you keep this as a platform variable (with > >>> a 2.2 MHz default)? The frequency depends on the board, > >>> and although most boards work at 2.2 MHz, not having it > >>> as a platform variable will make adding a board with a > >>> different frequency requirement difficult. > >> > >> I am not quite convinced that you'll have too many boards deviate from > >> the 2MHz ballpark. That seems to be a nice and safe frequency that > >> works well across phys, socs, and boards. > >> > >> That said, if we see the need to override the bus frequency in future, I > >> am all for a patch at that time. As it stands, I don't see the value in > >> adding platform data definitions for a capability that is not going to > >> be used at present. > > > > Okay, but why regress on existing functionality? > > I don't quite think that the prior existence of a "knob" is reason > enough to keep it around, considering that the knob is essentially > unused (below).
AFAIR, some boards in mach-omap2 use this driver as well. Hope you searched them as well before concluding that it is unused. > In short, we currently have a knob for frequency control, but we don't > use board-specific numbers from characterization data (if available). > That said, what good is the knob? All that it does at present, is muddy > up the code with identical definitions on every board. You can decide to default to 2.2 MHz if there is no platform data that way there wouldn't be any additional code in any of the board files. Thanks, Sekhar _______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
