So there is only one mount unit and that is for /boot/efi.
What gives systemctl show boot.efi.mount ?
Just for fun 
the output is long, that’s why i used eos-sendlog
1 Like
Mine?
The one I am currently on ( the AMD one) was installed: 2022-10-14
The other one may be a couple of months older.
1 Like
I just saw that there is a typo in the command I posted before.
It should be: systemctl show boot-efi.mount
Sorry for that @pycrk 
1 Like
That’s interesting!
Between the
GID=[not set]
and
Slice=system.slice
I have the following line in the output of the same command:
ExecMount={ path=/usr/bin/mount ; argv[]=/usr/bin/mount /dev/disk/by-uuid/4715-C7E2 /boot/efi -t vfat -o rw,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=ascii,shortname=mixed,utf8,errors=remount-ro ; ignore_errors=no ; start_time>
It may be totally irrelevant to the issue your describing here but it is just an observation.
1 Like
To help devs investigate this, anyone that has this issue (ESP on 2 mount points), please, post more info:
grep efi /etc/fstab
stat /efi /boot/efi
cat /proc/cmdline
journalctl -b | grep -E "efi\.mount|/efi"
1 Like
Can it be due to manual partitioning? Just a hunch as I did it.
I don’t think so, because I did a test with an automatic installation in the VM and the same behavior still occurred.
I guess it has something to do with the last systemd update, shortly after that I discovered the efi folder.
How about if you you removed /efi from your VM installation. Reboot and re-install systemd (packages) and check?
I’m in the same situation, during installation I did a manual partitioning with /boot/efi as mount point.
I have /boot/efi in /etc/fstab
Only /boot/efi is shown in /etc/mtab
stat /efi /boot/efi
File: /efi
Size: 0 Blocks: 0 IO Block: 1024 directory
Device: 0,33 Inode: 21063 Links: 2
Access: (0755/drwxr-xr-x) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2023-08-04 19:57:15.248987754 +0200
Modify: 2023-08-04 19:57:12.005654464 +0200
Change: 2023-08-04 19:57:12.005654464 +0200
Birth: -
File: /boot/efi
Size: 4096 Blocks: 8 IO Block: 4096 directory
Device: 8,1 Inode: 1 Links: 3
Access: (0755/drwxr-xr-x) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 1970-01-01 01:00:00.000000000 +0100
Modify: 1970-01-01 01:00:00.000000000 +0100
Change: 1970-01-01 01:00:00.000000000 +0100
Birth: 1970-01-01 01:00:00.000000000 +0100
When I check the logs, after /boot/efi was mounted, systemd-logind and thunar tried to mount /efi
août 04 19:57:15 linuxette systemd[1]: efi.automount: Got automount request for /efi, triggered by 749 (systemd-logind)
août 04 19:59:56 linuxette systemd[1]: efi.mount: Deactivated successfully.
août 04 20:33:36 linuxette systemd[1]: efi.automount: Got automount request for /efi, triggered by 1673 (Thunar)
août 04 20:37:37 linuxette systemd[1]: efi.mount: Deactivated successfully.
1 Like
any idea how i can rome the folder?
$ LANG=C sudo rm -rf efi
[sudo] password for pycrk:
rm: cannot remove 'efi': Device or resource busy
VM crashes and i can’t boot 
I’ll try something else.
If you want to remove it, you need to unmount it first.
Although, since it is probably being create by systemd, it will likely just get recreated.
As a side note, why is this a problem?
A partition mounted in two places shouldn’t cause any problems.
I can confirm :
systemctl stop efi.mount
systemctl stop efi.automount
sudo mv efi efi_bak
after reboot the /efi directory is re-created
ls / | grep efi
efi
efi_bak
1 Like
Since this is done by systemd, how come this happens only on some systems?
What version of systemd do you have on your installs?
Looks like it’s caused by systemd-gpt-auto-generator, according to it’s man page there’s a partition flag to not mount the efi partition automatically.
/boot/efi is mounted to through /etc/fstab first, then I guess /efi is created by systemd-gpt-auto-generator.
254-1
pacman -Qi systemd
Name : systemd
Version : 254-1