So, the regression potential is wrong and the fallback code I was thinking of does not exist. This seems to have caused a regression in bug 1937115. This will affect some systems that contain garbage in default boot order entries that still gets parsed as a valid second stage loader.
This should only be a problem when booting installer media, and maybe when booting via fallback, though I guess fallback will fix that issue before loading grub?
So, the regression potential is wrong and the fallback code I was thinking of does not exist. This seems to have caused a regression in bug 1937115. This will affect some systems that contain garbage in default boot order entries that still gets parsed as a valid second stage loader.
This should only be a problem when booting installer media, and maybe when booting via fallback, though I guess fallback will fix that issue before loading grub?