A swap file created with dd is only a regular file until it is formatted with mkswap, which writes the swap header that swapon expects. After the header exists, the entry already placed in /etc/fstab can be activated with a single command that reads all swap definitions: swapon -a (--all). Without mkswap the header is missing (“read swap header failed”); without swapon -a the new file is not enabled.

C. Changing mount options in /etc/fstab does not create a swap header or activate the file; it has no effect on the reported errors.

D. swaplabel is optional and only needed when you intend to reference the area by label.

E. swapon -L requires a label first (option D) and is unnecessary when the file is already listed in /etc/fstab.

F. Partition type 82 applies to disk partitions, not to swap files created inside an existing filesystem.

1. Oracle Linux 8: Performing Common Tasks

“2.4.3 Creating a Swap File”

steps 3-4 (mkswap

swapon -a).

https://docs.oracle.com/en/operating-systems/oracle-linux/8/olperf/creating-swap-space.html

2. util-linux Project

man pages:

– mkswap(8)

“mkswap prepares an area to be used as swap space.”

– swapon(8)

“-a

--all: All devices marked as ‘swap’ in /etc/fstab are made available.”

3. MIT-OCW

Operating System Engineering (6.828)

Lecture notes “Virtual Memory: swapping”

p.4 – distinguishes formatting swap area (mkswap) from enabling it (swapon).