The ext3 and ext2 filesystems preallocate a fixed number of inodes at the filesystem's make/creation time and do not generate them as needed. This means that the number of inodes in these filesystems is determined by the size of the partition and the inode ratio, which is usually 16 KB per inode. The inode ratio can be specified when creating the filesystem with the mke2fs command. The number of inodes in these filesystems cannot be changed dynamically, unless the filesystem is reformatted. This can lead to inode exhaustion, which occurs when all the available inodes are used up, even if there is still free disk space. To avoid this problem, it is advisable to estimate the number of files and directories that will be stored on the filesystem and adjust the inode ratio accordingly. The other options are incorrect because they do not preallocate a fixed number of inodes at the filesystem's make/creation time. The JFS and XFS filesystems use dynamic inode allocation, which means that they create inodes as needed, depending on the file and directory creation. The procfs filesystem is a virtual filesystem that does not store any data on the disk, but provides information about the processes and the kernel. Therefore, it does not have any inodes.
For more information on how inodes work and how to manage them, you can refer to the following articles:
* Inodes in Linux: limit, usage and helpful commands | Stackscale
* filesystems - What is an inode? - Unix & Linux Stack Exchange
* Allocation Methods | Linux.org