В Sat, 20 Nov 2010 17:27:18 +0100
Rodolfo Giometti <giometti@enneenne.com> пишет:

> On Thu, Nov 18, 2010 at 07:01:06PM +0300, Alexander Gordeev wrote:
> > This commit adds hardpps() implementation based upon the original one
> > from the NTPv4 reference kernel code from David Mills. However, it is
> > highly optimized towards very fast syncronization and maximum stickness
> > to PPS signal. The typical error is less then a microsecond.
> > To make it sync faster I had to throw away exponential phase filter so
> > that the full phase offset is corrected immediately. Then I also had to
> > throw away median phase filter because it gives a bigger error itself
> > if used without exponential filter.
> > Maybe we will find an appropriate filtering scheme in the future but
> > it's not necessary if the signal quality is ok.
> 
> This patch (and follows) adds a new functionality to PPS subsystem, so
> let me suggest to you splitting in two parts this patch set: a first
> patch set who fixes up the current PPS implementation, and a second
> patch set who adds kernel consumer (and follows).

But the patches that add hardpps and parport client+generator depend on
the previous ones because the latter change lots of things in PPS
subsystem. I don't want to maintain them separately because we use all
of them in the production.

However I can tell here that it'll be ok for me if not all the patches
enter mainline at the same time. For example, patches 1-3 are already
ACKed and are actually bugfixes (patch 2 is trivial and patch 3 depends
on it) so I think they could be merged in the next rc. I placed them
in the beginning exactly for this reason. Whom should I mail them?

Same thing with other patches. Smaller patchset -> me (and probably
users) more happy. :)

-- 
  Alexander