The last few months I have been working on the Penumbra project. I started off patching wireless drivers in and out of the kernel tree to achieve the anonymous broadcast action that the project needs, but it became clear this would be completely unworkable for general use… getting wireless up in Linux can still be a struggle and hoping people will patch their driver or kernel in addition isn’t going to happen. After trying a couple of other methods in the end I created a radiotap-based packet injection patch for the mac80211 stack (formerly dscape / d80211), and bound it together with a patch from Michael Wu that provides radiotap-based Monitor mode. At the moment it is still in front of the linux-wireless folks and it’s not clear what the result will be. If the patch is accepted, then the code should make it into the mainline kernel and all mac80211-based wireless drivers will work with Penumbra out of the box in the future. The patchset provides generic radiotap monitoring injection that “just works” with libpcap both ways, so I am hoping it will get accepted without people having to form an opinion about Penumbra.
But one of the biggest hurdles in creating the patch was not technical, since I already had the core functionality working, but in fact the Linux kernel coding style. In some ways the coding style fits well with my own personal style (formed over 20 years of writing C and C++), we basically use the same K&R style. There are some spacing and commenting rules that are actually better than my style and I will adopt them wholesale. But that’s where the fun stops and the recrimination begins!
The willingness of the kernel devs to refactor stuff is both a huge strength and weakness for the kernel. The strength is in the extraordinary continual optimization and improvement in the codebase, not just locally to an area of code but for cross-kernel concepts, like the recent workqueue changes.