When I am working on our RHEL machines, I use sudo su - to switch to being root. One day, a typo meant I typed sudo su -- instead - it seems to me that everything was the same as with a single hyphen, except that I was in the same folder as before I issued the command (with a single hyphen I find myself in /root).
Are there other differences? Is it safe to use this in scenarios where I know I want to work in the same directory?
When you provide a double-hyphen the experience you will have is identical to if you had just executed sudo su without any hyphen.
Passing a single hyphen is identical to passing -l or --login. The man page for su describes the behavior as:
Provide an environment similar to what the user would expect had the user logged in directly.
This includes setting your directory to your home directory and setting a bunch of other environment variables.
Passing a double-hyphen to a command is typically used to mark the end of command-line flags and the beginning of non-flag arguments. For example, if you run touch -R you'll receive an error saying that -R is not an option to touch, but if you run touch -- -R it will create a file named -R. This is true of many command-line tools (ls -R will do a recursive ls whereas ls -- -R will perform an ls on a file or directory named -R.
So, to wrap this up, when you pass only -- to su it is basically ignoring the -- and acting like you did not pass any option at all.
sudo su - and sudo su -l differs. On an account where I ssh into with a key only, and I have sudo rights to change to a specific other user, sudo su - changes the user without problems but sudo su -l or sudo su --login will ask for a sudo password, which I don't have (although by examining sudo -l I could execute /bin/su with NOPASSWD) su man page.