I changed a lot in implementing and setting up testing stuff for netsplit, Including a new netsplit struct in netsukuku.h which may or may not be permanent. It was the only way I could find to preserve a variable from netsukuku.c to andns.c. Also, The netsukuku.c interface argument has, apparently, been just a break statement for a long time, I have reverted it. I still don't totally understand how it works, though. There is more to come!
I have done a LOT of work! I have changed the whole paradigm of the ntk console client and console bindings! I have changed it from peer to peer to server-client because there was no point doing it the other way, And that way was hard and I didn't really know what I was doing. I followed some IBM examples, And didn't try to extrapolate.
I have done a little bit of work trying to resolve some compilation errors, However, I cannot check if there are any compilation errors now because my root partition is full, again. My compiler needs to do something with /tmp so I can only see a portion of the error logs, All of which are warnings up until the no space error.
I will do some more testing and debugging on my other machine, Help is appreciated, Enjoy!
Wow! That was a lot of work! But, Netsukuku now has (or should have) a functioni
ng console with the bindings in ntkd to receive, process, and respond to command
s! Now, I just need to test stuff, And figure out how to deal with server_opt be
cause that is a type and not just a struct. Anyway, Enjoy!
Wow! I've done A LOT of work on the netsukuku console recently! I still haven't set up any way for ntkd to receive the commands, Or done much of anything in netsukuku itself recently except change some stuff in the exclude function to prevent possible errors if an interface had the same name up until the end of the one specified. I.E if it had tunl012, It wouldn't work because it had tunl0 in the beginning. Now, It will read from the entire length of the variable, And if it doesn't match it entirely, Then it is okay.
Anyway! The netsukuku console is almost ready! I just keep getting this weird issue around fgets() not working at all in console(). It causes the program to spam the > symbol. It functions in main()! However, After that, It has a segmentation fault. I've been trying everything I can think of, One thing made it not do a segmentation fault, However! The fgets() inside of console() still didn't work, So, It just spammed endlessly. Anyone who can, Please help! I've been working on this for a while! :/
Okay, There are some compilation errors, And, While the check_interfaces function will work, It will only work if the interface lists are already populated. Which, Doesn't seem to happen until after the argument section of netsukuku. I guess we will have to, Either, Call those functions somehow (In a way that won't cause them to be called twice, And, Thus, Create duplicates.), Or, Use the exclude_interfaces function. However, That function is probably sub-par, I don't know how well it will really work with the rest of netsukuku.
Much more work and testing is required, I will just comment this code out for now, I am going to look into the feasibility of porting netsukuku to windows.
I added a lot of stuff! I fully set up exclude_interface to ignore inactive interfaces, and optarg, Along with a small list of other interfaces. I did some bug fixing, And looked into how netsukuku sets it's interfaces by default, And added some stuff to deal with that too.
I, Also, Figured out that, The entire exclude_interface function is probably unnecessary, And can be fully replaced with just check_excluded, Which would be much simpler, Easier, And elegant. Along with the fact that, It will probably just work.
Added the beginnings of a function to check if a second exclude argument is excluding something that was already added into netsukuku's interface list, And removes it.
Oh, My, GOD! You may think I was just doing afk the past week or so, But, You'd be wrong, So, So, SO wrong. I've spent almost every day working as hard as I can on this commit here. It should work perfectly now, It should properly exclude interfaces, And, It should automatically exclude inactive ones. More testing is needed, Especially testing of running the -e argument multiple times, I.E -e eth0 -e wlan0. Anyway, Thank you for reading, And for downloading.
Here is a vagrant file I have been testing with, You will have to use vde_switch in order to run netsukuku without being disconnected from ssh. Also, I've done some work on the exclude argument, I'm testing some goto stuff right now, I'm primarily pushing this for testing purposes across multiple devices.
Everything should now be okay, Just, Don't leave any interfaces disabled. We still need to figure out how to handle excluding multiple interfaces, And for ignoring disabled ones.
I've done A LOT of work on an exclude interface argument. I.E us all interfaces except this one. It does not work yet, Any help anyone provides in fixing it will be appreciated.
Changed SIGKILL to SIGINT, So Netsukuku will clean and save everything properly. I, Also, Added more stuff to inet_sendto for packet division in the case of errno 90. Still working
Yann Weber
MissValeska