I had some problems recently in Manjaro with newer kernels not getting properly listed in my multi-boot Mac Pro, such that only an older kernel that would not revive the display from suspend in Manjaro, was showing. Tumbleweed was the system that I had grub/osprober installed in and the forum here said, “You need to change grub to Manjaro.” At the time I didn’t want to, but seems like TW’s grub is also not finding my other systems kernels too well, so yesterday I installed grub and turned osprober on in Manjaro and ran the “grub-mkconfig” command, which went through and found the other 6 installs.
I shut down and on cold boot grub was still being handled by TW. I turned off osprober there and shut down. On cold boot now grub was only showing TW. I then removed grub from TW and shut down.
I got my SuperGrub2 disk out and booted to Manjaro, ran the grub-mkconfig command again, but still grub is showing only TW.
Any thoughts on how I could get Manjaro to be the grub handler?? Right now no updates in pacman to run to see if that would find grub powers within.
Then you need to either install Manjaro’s grub to the BIOS/EFI, or if you already did, you need to select it as the default boot option in your BIOS settings.
OK, thanks for the replies . . . google search is always an option . . . .
The Alt/option key method just shows OSX options, or in my case now, using that to get to my usb flash drive with SG2 to boot the system of the day. The “EFI” disks that show up there usually do not boot the Linux options, only grub will do that.
I guess I’ll post back if and when I figure out how to get Manjaro’s grub to show up . . . or if I have to fall back to TW . . . for internal grub handling.
[edit] Took some time to read through some forum results that showed up in Google search . . . some of it “complicated,” but in looking at:
sudo cat /etc/fstab
[sudo] password for noone00:
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a device; this may
# be used with UUID= as a more robust way to name devices that works even if
# disks are added and removed. See fstab(5).
#
# <file system> <mount point> <type> <options> <dump> <pass>
UUID=4b916772-18f6-4b7c-b47c-eb0fa3796eff / ext4 defaults,noatime 0 1
UUID=64c5dba2-bacd-4459-9904-6bda0fbe24e7 /home ext4 defaults,noatime 0 2
It is clear that in the recent install of grub from the GUI software app in Manjaro, the /boot/efi UUID data is not included . . . . So, at some point I will have to add that data in so that perhaps the system could know where to install grub. [/edit]
Noticing that Manjaro doesn’t show a swap partition in the /etc/fstab data?? Does that matter in the basic operation of the system?? Or does Manjaro do what ubuntu started doing where they just have a swap file?? Partition is not necessary with UUID number??
And then, if I edit the /etc/fstab data to add in the /boot/efi partition what are the suggested numbers at the end of each line about?? the “01” “02” ?? What would the best choice be for the /boot/efi data??
Thanks for the question . . . if you look up at the post where I have the data from /etc/fstab showing, it has two lines . . . one for / and one for /home.
Well, actually in each of the three drives I have now in that machine, there IS swap space . . . but it isn’t mentioned in the Manjaro data . . . .
Still trying to get some response on the “0” numbers for what would be the added /boot/efi UUID data to try to get Manjaro’s grub function working, which now it is not.
The swap space doesn’t need to be in the fstab because systemd will automatically activate any swap partition on the same physical disk as the EFI system partition.
fstab - ArchWiki does an excellent job of explaining what the numbers are at the end of lines in /etc/fstab.
Latest update: I added the /boot/efi UUID info into the /etc/fstab data . . . added a few kernels to test out for function. I can see that each time a kernel is added/subtracted the system is going through the grub/osprober details . . . .
But, so far Manjaro’s grub menu has not “taken the helm” of the grub duties. Perhaps at some point after a pacman -Syu there will be some new grub package that might take over from the TW hander???
And then, if I edit the /etc/fstab data to add in the /boot/efi partition what are the suggested numbers at the end of each line about?? the “01” “02” ?? What would the best choice be for the /boot/efi data??
Yes … like @ben75 earlier said
OP seems to have missed that these are two fields, not one number, there is a space between them
and these represent the fifth (5) and sixth (6) field in fstab
Alrighty, thanks for the link, I think somebody on this thread had already posted that data . . . . I have used chroot to get into a system in the past . . . in this present case grub IS installed in Manjaro, and running “grub-mkconfig xxxx” does show that osprober is finding the kernels and the other systems within Manjaro. But, on cold boot or reboot openSUSE’s version of grub shows up . . . .
@et al: Thanks for the thoughts on the two numbers, that was also previously linked, so yes, “two different functions” . . . . That has been utilized.
For now I will continue to monitor the situation, as there is grub function for the linux installs.
This perhaps just inaccurately worded comment leaves open whether the difference between “Installed in Manjaro” and “boot loader restored” is really understood.
Installing grub in Manjaro will never update grub to ESP or MBR level, these instances must always be manually installed using the commands as per wiki to then make any updates to grub take effect.
If you want to have Manjaro controlling this you need to restore the bootloader. Also, if you have an UEFI system make sure the firmware boots the right entry, could be verified via firmware (BIOS) setup for example but there are other ways, too (efibootmgr).