Hi, > > I'll post again, when the repository access is available. > > Even googlepages or pbwiki is enough as long as you properly maintain > it. I did it sometime back for nngfs.pbwiki.com I skimmed through your pages - very nice. But as you wrote in docs-intro - most fs stuff is outdated and for so broken. The most up2date docs are in the kernel itself, but that's mostly api-docs - too small for a beginner. It's like having all parts of a puzzle but no image of the final result. I like most your random notes - that's the kind of information I'm searching for actual kernels. If you know an actual call flow or a way on how to get some, a hint is very welcome. But my idea is a bit different. I don't want to care about physical devices. There are a lot of people out there, that can do it better - so I like to create a tiny 'management'-layer in the middle, that creates appropiate events and routes the real work to existing fs-modules like ext3, jfs or so. >From toplevel view I'd like to create a thread at mounting time, that communicates with the mapped-module-functions on one side and with userspace on the other side. Depending on the userspace-actions, the module selects the appropiate 'real'-fs, that finally does the job. The comment of ramfs tells, that it serves mainly as a sample, but cloning ramfs does not work, as it uses internal kernel stuff and only works as 'compiled in', not as a loadable module. rkfs is a nice sample to start with, but is broken with current kernel. kind regards Santiago -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
