@peteorrall

That enlarges the locus of the problem, unfortunately; as the firmware's CSM could be faulty given that behaviour.

That said, EFI bootstrap instead of PC/AT MBR is the way to go if one can.

@rl_dane has posted some recent war stories when it comes to #FreeBSD 15 and #WiFi. They may or may not help.

#BootLoaders

@peteorrall

Don't get hung up on the message just because it is something that you can see. As I said, it is very likely a red herring.

In fact the system *has* proceeded beyond it. It has done other stuff, printed that bootstrap loader version line, and is now doing something else.

That something else could be reading the firmware memory map using INT 15/E820 or reading ZFS boot information or loading fonts; all three of which happen without printing anything.

#FreeBSD #BootLoaders

@peteorrall

The actual message is not, strictly speaking, an error. It just means that you're bootstrapping the non-EFI way, and your machine's firmware does not adhere to a convention (I believe from DOS+Windows 95 and the 1990s) for marking "not an MBR virus and safe" firmware add-on disc device I/O handlers.

It's unlikely that F000:F0BF is actually unsafe, as that's in (notional) ROM.

This message is very likely a red herring.

#RealMode #IA32 #BootLoaders #FreeBSD

@peteorrall

The odd thing is that the linear address does not match the segmented address.

It's likely not the true cause of your problems, though; but the mis-match might be symptomatic.

#RealMode #80386 #IA32 #BootLoaders