Squashfs LZMA

LZMA is one of the best compression algorithms and Squashfs is one of the best compressed filesystems available nowadays. This project combines both to make something even better. LZMA can compress much better then gzip, which is used in squashfs now. Compressing to LZMA format is very slow, but this is not a problem as the squashfs filesystem needs to be compressed only once. Then it's only read million times; so what we care about is the speed of decompression. LZMA decompression is slower than gzip, but it's even very fast, about 10-20 MB/s on a 2GHz CPU.

method	block size	Slax data size	percent
uncompressed	-	668 MB	100%
mksquashfs+gzip	64KB	227 MB	34%
mksquashfs+gzip	1024KB	222 MB	33%
mksquashfs+lzma	64KB	191 MB	28%
mksquashfs+lzma	128KB	184 MB	27%
mksquashfs+lzma	512KB	172 MB	26%
mksquashfs+lzma	1024KB	167 MB	25%

Squashfs' author can't support LZMA directly in squashfs source code. He would like to have the code included in official Linux Kernel, but Linux Kernel doesn't support LZMA yet, so adding LZMA to squashfs would decrease the chance for its (squashfs's) inclusion. This is the reason we have to maintain LZMA support as a separate patch.

Author of sqlzma patches is Junjiro Okajima

Author of SquashFS is Phillip Lougher

Author of LZMA compression is Igor Pavlov

Author of Slax (the sponsor of this code) is Toms M, see www.slax.org

Download all the sources from this website, using the download links at the right side. Make sure to use a supported Linux Kernel, as only version 2.6.24.3 is supported. The info about the compilation is described in sqlzma.txt file in squashfs-lzma patches. Optionally apply Squashfs CVS fixes (link at the left) AFTER you applied sqlzma patches. Finally, make sure you have CONFIG_SQUASHFS*=n in your kernel config, because squashfs will be compiled separately as module. You may use precompiled Linux Kernel with Squashfs-LZMA as a module.