Well, that's all. The big big big advantage of this scheme is that there is almost nothing to change in the file system. We just have to hack a little bit the read and the write routine, in order to compress or decompress the data if it turns out that the block we are working with belongs to a compressed cluster. This is just a matter of a few lines. [pjm: I think Antoine's understating things here a little.]
See `linux/fs/ext2/Readme.e2compr' of the kernel source tree, for a more detailed description and a map of the source code.
Go to the first, previous, next, last section, table of contents.