Jo22 wrote on 2025-06-02, 08:28:

myne wrote on 2025-06-02, 03:42:

What is the fundamental difference between int13h translation to sata, and an int5h translation from sb# to hda?

Software interrupt vs hardware interrupt, I think.
MS-DOS uses int21h, for example.

Ah, okay. Now I understand the confusion here as well. Indeed, IRQ 5 (hardware interrupt typically used by sound cards) is not the same as INT 5h (software interrupt).

Both the IBM compatible PC BIOS API and the DOS API are accessed through software interrupts. Same with DOS device drivers. And those you can easily intercept by hooking into them, without having to result to virtualization and I/O port trapping and such, which can't be done in "pure" real mode.

Has anyone tried this directly on UEFI-only hardware yet? That's its purpose. I see it being used in VMs exclusively. Not sure what that proves. Installing XP 32-bit on a STCK1A8LFC Intel Compute Stick would be a modern miracle. Intel's red-headed stepchild. Lowest resource UEFI-only computer EVER created. 1GB RAM and 8GB storage. Perfect candidate. CSMWrap can be a potential goldmine if development continues and more versions are released. Old hardware is disappearing.

Did the older bioses allow scsi/lan modules?
Could a 64k module be inserted?

muefircate seems to have completely disappeared? codeberg.org's page for tkchia not showing it unless I'm missing something.

confirmed dead..not even on GitLab anymore

myne wrote on 2025-08-08, 02:16:

Did the older bioses allow scsi/lan modules?
Could a 64k module be inserted?

It allows you to substitute the video bios.

The rom scan routine just looks for magic header bytes.

Maybe, try concatonating a video bios rom with your desired option rom, with video rom first?

That way, video rom will start at c000, usually end at cc00 or cfff, and then your smashed on option rom would start at d000?

The rom scan routine would find it's magic byte sequence, then execute it?

Worthy of an experiment at least.

Might need padding between.

I gave it a try on real hardware a short moment ago. It got to the stage where it announces itself and then the text cursor blinks in an endless loop.
blinks in an endless loop.
blinks in an endless loop.

I am not quite sure whether it is waiting for something or it just got stuck. But I guess it's a bug.

Start me up wrote on 2025-08-09, 03:20:

I gave it a try on real hardware a short moment ago. It got to the stage where it announces itself and then the text cursor blinks in an endless loop.

It would be good to report your hardware details to Flygoat or Mintsuki...last guy is apparently more involved and is able to devote more time. This is still in its infancy. I want it to succeed.