Dynamic IfDesc Table

[Uglymotha]

Fixes issue #62 and build errors on freebsd because of incorrect placement of includes.

[pali]

Ok, so you have already figured out that there is automatic compile check for every pull request and you have already fixed code as to be compatible with more compiler versions...

Some more comments:

• If diff there are lot of whitespace changes, which make only noise and make reviewing patches more complicated. If there are no reasons for whitespace changes, please remove them.
• Squash "fixup" commits into some relevant previous commit (see git rebase --interactive) and then do force update (see git push --force); every commit in history should be compile-able
• If it is not really necessary, do not try to change indentation of code as it just increase diff

[Uglymotha]

Over the last few days my rework of the code has become increasingly extensive. Igmpproxy was indeed a good project to get back into programming, it's definitely been fun.
In some places the code was just ancient, like using socket io to get interface info instead of getifaddrs(). In other places code was shabby, in some places bad and in a couple absolutely terrible, like the main loop and queue timing. The project seems to hav esuffered from repeated unfinished attempts to implement functionality. Leading to a bit of a mess. Outside of the actual igmp functionality it's all redone the way it should be. Lots a functions were altered to more readable and flexible code, some functionality split of into new functions and some removed.

I currently have the interface and configuration management working correctly as far as I am able to test, SIGHUP is functional and main loop timing is now implemented in just 12 simple lines of code, using 1 call to clock_gettime. instead of the some 60 lines it was before. This also fixes issue #58 without adding another shabby patch to a completely incorrectly implemented timing.

I am now reworking the callout queue, when that's finished I'll do some more testing of my own before I am confident in releasing the new code. Hopefully I'll be able to get a version of this all in my own repo today.

Leaves the question of how to proceed. Opening a new pull request for all the changes when I'm finished or maybe fork off completely. My preference would be the first.

[pali]

Please separate changes (ideally to more pull reaquests) as reviewing one big change is hard. Also please update this pull request as I mentioned in previous comment so I can look at it.

[Uglymotha]

After some more testing, I am now confident to release the new code base.

Closing this PR and opening a new one, is it has become much more than just dynamic ifdesc tables.