Make ext4 filesystem explicitly on zram device

[damentz]

The command 'mke2fs' can have different results depending on the configuration of /etc/mke2fs.conf. If an ext2 or ext3 filesystem is created, mount with "-o discard" will fail.

Use standard mkfs.ext4 command available on all distributions to eliminate potential variation in resulting filesystem.

Per Arch's wiki, ext4 supports continuous fstrim while ext3 does not: https://wiki.archlinux.org/index.php/Solid_State_Drives#TRIM

Closes suaefar/ryzen-test#11

[damentz]

@suaefar you're correct, discard is supported on zram. After some local testing, the default settings on Debian prefer to make a featureless ext2 filesystem with mke2fs versus ext4, or whichever filesystem / feature set is picked on Ubuntu.

Specifically, this is in my /etc/mke2fs.conf file:

[defaults]
        base_features = sparse_super,large_file,filetype,resize_inode,dir_index,ext_attr
        default_mntopts = acl,user_xattr
        enable_periodic_fsck = 0
        blocksize = 4096
        inode_size = 256
        inode_ratio = 16384

[fs_types]
        ext3 = {
                features = has_journal
        }
        ext4 = {
                features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
                inode_size = 256
        }

Unless your script depends on the exact mount options you picked, I feel that this change is sufficient. Otherwise we can look at which exact feature we need to explicitly add for discard to work properly across multiple distributions.

[suaefar]

Okay, thank you.