On Sun, Mar 09, 2014 at 08:26:49AM +0000, Mark Brown wrote:
On Fri, Mar 07, 2014 at 04:34:20PM +0000, Charles Keepax wrote:
Currently the FLL configuration is calculated before it is known which FLL path the configuration will be applied to. Newer versions of the IP have differences in the configuration required for each FLL path, which makes it complicated to calculate the FLL configuration in advance.
This patch simply checks the validity of a requested input and output frequency before we know which FLL path they will be applied to and saves the actual calculation of the configuration until we know where the settings will be applied.
I'll apply this but are you sure that this validity check is actually accurate? For some of the devices there were input/output configurations that couldn't be reached due to constraints in the system even though both input and output were in range.
Only thing I can find directly is sub 5kHz as an input clock would be a problem on some parts but that is unlikely to ever happen in practice. Naturally there is the granularity of the multipliers as well but the step on that appears small enough that I wouldn't be overly concerned about the error, given there is ~46-bits worth of options there.
Also the error checking matches what was there before the patch, ie. anything that pass here would also have passed the old implementation. So even if I am missing something it should be as robust as the old solution.
Thanks, Charles