For some reason, ./install.sh didn't finished (grub error), without any message

[xfim]

First of all, thank you very much for make such a contribution, that allows to combine the best of both worlds (asahi support with Gentoo's flexibility).

I have been able to solve the issue, but nevertheless I think I would write it here, so that maybe something can be done:

The first time runing the "./install.sh" script, during the Live session, the script finished and apparently all was well, since there was no error message and everything seemed well. I can't reproduce the exact message, but all seemed well. But it turns out that the machine was not starting properly. So then I went again throug the script and finished exactly there, but noticed that some messages that should appear (the ones regarding the firmware, at the end) were not appearing. Then, I realized that the grub-install part had not run, and that was the reason the machine was not working.

So I ended up manually running the grub part, and the firmware part too. This worked and made the machine bootable in Gentoo.

Since I can't really tell what was wrong, I am afraid I can not help debugging it. But at least, I believe that the script should throw an error message if it fails to complete some of the parts.

Thank you again,

[chadmed]

Are you sure it failed silently? The only reproducible way GRUB could fail to install is if the ESP is not mounted as expected. If there is some other failure mode, grub-install would throw its own error message and diagnostics before exiting.

[xfim]

You are right. I am not sure, since I could not see beyond the last, say, 50-80 lines. I was not using any terminal multiplexer and thereforth was not able to see the whole console output. But what I am saying is that the last 50-80 lines contained no indication of an error message (that I could read as such).

Definitively it was not the ESP, as it was mounted as expected.

Sorry for being so unspecific. Maybe it is better to close the issue and in case I ever have to reinstall again I will try to be more attentive.

In any case, it was a blessing to have it, thank you!

Missatge de James Calligeros @.***> del dia dc., 24 d’ag. 2022 a les 2:42:

Are you sure it failed silently? The only reproducible way GRUB could fail to install is if the ESP is not mounted as expected. If there is some other failure mode, grub-install would throw its own error message and diagnostics before exiting.

— Reply to this email directly, view it on GitHub https://github.com/chadmed/asahi-gentoosupport/issues/3#issuecomment-1225032495, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFEW4ZOOERWWUY7TLPEUOLV2VVX5ANCNFSM57GLVHBQ . You are receiving this because you authored the thread.Message ID: @.***>

-- Xavier

[chadmed]

I'll leave this open just in case anyone else experiences something similar themselves, but as you can imagine working out the kinks took me many many reinstalls (just look at the commit history lmao) and I did not see any instances of anything failing without providing at least some indication that it had in fact failed.

Regardless, I will put some more error handling in around the script just to remove any possibility of this happening at all.

[xfim]

Ok, thank you. I leave it to you.

Missatge de James Calligeros @.***> del dia dc., 24 d’ag. 2022 a les 7:20:

I'll leave this open just in case anyone else experiences something similar themselves, but as you can imagine working out the kinks took me many many reinstalls (just look at the commit history lmao) and I did not see any instances of anything failing without providing at least some indication that it had in fact failed.

Regardless, I will put some more error handling in around the script just to remove any possibility of this happening at all.

— Reply to this email directly, view it on GitHub https://github.com/chadmed/asahi-gentoosupport/issues/3#issuecomment-1225206075, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFEW42OZIAOBDDYNZKROX3V2WWLHANCNFSM57GLVHBQ . You are receiving this because you authored the thread.Message ID: @.***>

-- Xavier

[Cesura]

At least in my case, the reason for premature failure was a missing backslash escaping the newline during the genkernel execution:

genkernel \
   --kernel-config=/usr/src/linux/.config \
   --all-ramdisk-modules     <------
   initramfs

After fixing this, the script continued until completion.

[chadmed]

Whoops, thanks for picking that up. I've amended the bad commit.

[chadmed]

Closing this now since we are back to using dracut and whatever was causing this seems to have been fixed anyway.