[Splashy-devel] My Tests.

[Pat Suwalski]

Tim Dijkstra wrote:
>> If we remove S05keymap, it still gets started later on in init, on my
>> clean install it happens right between sysklog and klog in rc2. 
> 
> How do you know? 

Easy. Put a script that calls /bin/bash between the two scripts. I tried 
it iteratively until I found the spot.

>> Now, I have commented before about how splashy seems to use tty0 and
>> tty2. The latter is used for apparently no reason. This has struck me as
>> odd before, about a year ago, iirc.
> 
> Maybe this is a directfb feature where it redirects part of its output to
> another vt?

Possibly. But it's still interesting that there is a gap from tty0 to tty2.

>> Long story short, what I witnessed tonight is exactly what I witnessed a
>> year ago: resource contention. Splashy takes two ttys, stays on one of
>> them, init is forced to take another, things expect to be on the tty
>> that init is on, there is a lock.
> 
> Well, this sounds logical, but it doesn't tell me anything. Why is it a
> problem we're taking two tty's? Also if we're started after init, we do
> the same, I think.

Init likes to be on tty0. Other things like to use /dev/console which is 
supposed to point to the current tty. Using fgconsole further confuses 
things. I feel there is a lot of assumption around being on tty0 at bootup.

>> Something I had meant to experiment with last year (never got the
>> chance), was to make splashy open a higher tty by default, say tty6 or
>> tty9. Perhaps if init is on tty0 it will be happier. Maybe there could
>> be some experimentation whether a tty even needs to be opened?
>  
> I tried that, but it didn't help. What I did is chvt to 12 before we init
> the directfb.

When I tried it, splashy (or directfb) still moved the lower console 
over. So, there is something "special" happening there, and I'll bet 
dollars to donuts that's the problem.

--Pat