[mythtv] firewire branch status?

[Steven Adeff]

On 4/24/07, Daniel Kristjansson <danielk at cuymedia.net> wrote:
> On Tue, 2007-04-24 at 07:45 -0400, Steven Adeff wrote:
> > Which got me thinking, this reset doesn't need to occur that often (at
> > least in my case), and I don't know how things will behave when I've
> > got both boxes daisy chained to one machine, but... What would the
> > effect be of reseting the bus when no firewire devices are recording
> > right before one will begin recording? Sort of a "lets do this just in
> > case" sorta thing?
>
> Well it would have to be testing for data since resets are just
> as likely to stop the FireWire from working as to make it work.
> Really someone should look closely at the Linux driver and
> figure out why the first reset usually doesn't work, on OS X
> a reset usually works (they do one whenever a program using
> firewire exits, so their implementation has its own problems.)

so would implementing something along the lines of what
firewire_tester does work? ie. check if data can be properly read, run
a fix if not, and run a bus
reset if required? Obviously from what you said this wouldn't
necessarily be as ideal as "fixing" the linux driver, but may be a
good work around until then?

-- 
Steve