I'm trying to install Centos 8 to an embedded computer fitted with onboard eMMC storage. The installation completes successfully but the system does not boot.

After reset, the grub2 boot menu is displayed showing the installed kernels but the boot process goes no further.

A few articles mention that Centos/anaconda does not include the proper drivers for accessing eMMC devices in the newly created initramfs file on the target machine:

• e_MMC install on intel NUC DE3815TYKHE
• Bug 0007453: Anaconda completes installation on MMC internal card on NUC and fails to include MMC kernel drivers in initramfs

Which says you need to add some additional drivers to the initramfs files:

I tried that by switching to a different console after a normal installation completed but dracut failed with:

This Red Hat article supposedly explains how to get around that problem and update the initramfs but I cannot access.

Anyone know how I can get the correct drivers into the initramfs during an automated kickstart installation?

You could add that command to the post-install instructions in your kickstaart script.

Thanks, I can try that. But as I understand, the %post section will operate with the same set of libs, bins and environment as switching to another console. Let's see.

If that doesn't work, then try these steps to add to the post-install section to run the command after it reboots.
https://bitraboy.wordpress.com/2018/...t-centos-rhel/

Oops, never mind. Forgot the system is not booting, but I will leave the link here for future reference.

Found better info to run scripts outside the chroot install environment:
https://serverfault.com/questions/97...ironment-and-t