mammos | Bootloader with minix fs implementation | File Utils library

mammos is a C library typically used in Utilities, File Utils applications. mammos has no bugs, it has no vulnerabilities and it has low support. You can download it from GitHub.

the minix filesystem uses blocks as a method of measuring. a block is 1024 bytes (default,) which means 2 sectors of 512 bytes. #i represents the number of inodes in the filesystem and #z the number of datazones. see below for more information about those. we are talking about the v1 filesystem so inodes are 16 bytes long. we'll take the example of a 16MB partition. this is 5120 inodes (as can be seen when formatting) for a total of 15360 blocks. an inode is composed of a mode (directory, block device??), permissions etc, but mainly a zones array. this array contains offset of zones composing the file. in the case of a directory, the inode contains offset to zones containing directory entries. a directory entry represents another file or directory contained in this directory. it's composed of the offset of the file/directory's inode in the inode table, as well as its name. for example, the path /boot/file1. here, the INODE TABLE column represents three inodes of 16 bytes each.contained in the first block of the ~150 composing the whole inode table. each entry in the datazone column (directory entries of the root folder, directory entries of the boot folder, datazone1 of big file1, datazone2 of big file1 as well as datazone3 of big file1) are all different blocks. each entry in the directory entries blocks is like 32 or 16 bytes (depending on the filename length, it's usually 32 now, 30 bytes for filename and 2 bytes for inode table offset.). +-------------------------------------------------------------------------+ | | | INODE TABLE DATA ZONES | | | | +--------------------> directory entries of the root folder: | | | | | | [root inode offset] ["."] ---------+ +-> root folder inode [boot inode offset] ["boot"] | | | +-----------------------------------------------+ | | | | +--------------------> directory entries of the boot folder: | | | | | | [root inode offset] [".."] ---------+ +-> boot folder inode [boot inode offset] ["."] ----------+ | [file1 inode offset] ["file1"] | | | | | +---------------------------------------------+ | | | | | +-> file1 inode ------------------> datazone1 of big file1 | | | | | | | +----------------------> datazone2 of big file1 | | | | | +---------------------------> datazone3 of big file1 | | | +-------------------------------------------------------------------------+. as you may already have understood, inode table offsets being composed of 2 bytes, there are maximum 2^16 different inodes possible, which means 2^16 files/directories/symbolic links/hard links... however, the limitating number of files may also be the number of datazones. in this case, there are like 15k blocks so ~15k datazones, which means maximum 15k folders/files/... you can see an implementation of a basic reading procedure in boot/stage2/fs/minix.{c,h}.