[distcc] Problem with zeroconf daemonization

[Benjamin R. Haskell]

On Mon, 18 Jan 2010, Ihar `Philips` Filipau wrote:

> Well, you can't satisfy all.  And the problem (in the particular case) 
> isn't even performance critical: everything what would do the work in 
> less than half second is OK.  Most fool-proof solution wins.

But that's contradictory: most fool-proof is looping over all possible 
values for fd, but on many systems thats >> 5sec.  (10billion on mine = 
~1min)

> > > Either way, though, there seems to be agreement that this is a 
> > > 'distcc' problem and not a 'paludis' problem?
> >
> > [still interested in the consensus here]
> >
> IMO. Implement fool-prof generic method. Add /proc/*/fd trick for 
> Linux (as most commonly used platform).

Sounds like the plan.


> Code from Lennart's hint satisfies most of the requirements.  
> closefrom() syscall: let somebody else who has access to the 
> supporting platform add a new define and code path for it. Ditto 
> /dev/fd.

Also sounds reasonable.


> I'm no distcc maintainer, but IMHO there is no need consensus here for 
> a pure technical question on how to close fds.  Do not plan for 
> problem - solve the problems as they come. Especially considering that 
> in the case they are going to be easy to solve.

Sorry, I think my referent got lost...  I meant consensus on the fact 
that this is distcc's problem (closing fd's it didn't open), and not 
paludis's problem.  (Seems like people agree it's fixable easily-enough 
in distcc, but I wasn't sure if this was usually considered a problem 
that the daemonizing process was responsible for.)

Best,
Ben