FlyGoat
Stuck on "Press ESC for boot menu" on ASRock N100DC-ITX
Hi
I tried using CSMWrap to boot a 32-bit linux on the (64-bit UEFI) ASRock N100DC-ITX board. The boot starts ok but stops at "Press ESC for boot menu"
https://github.com/user-attachments/assets/c0317240-6a09-48df-b45d-71f2cff552ab
I tried running from shell and adding PciSioSerialDxe and TerminalDxe (as others have suggested) with a null modem attached, but I got no output at all on the serial device. Is there an additional step I need? What baud rate should it be?
Or is the binary release not compiled with debug output?
Any suggestions would be appreciated
Thanks
Geoff
I tried running from shell and adding PciSioSerialDxe and TerminalDxe
Drivers need to be loaded, not added. Baudrate in Putty 115200 https://github.com/FlyGoat/CSMWrap/issues/14#issuecomment-2889000170
I tried running from shell and adding PciSioSerialDxe and TerminalDxe
Drivers need to be loaded, not added. Baudrate in Putty 115200 #14 (comment)
Apologies, I should have been clearer in the syntax of my message. I added them by typing "load", as you did.
So the precompiled release version 1.2.1 does have debug enabled?
Geoff
Apologies: I realised that I'd disabled the COM port in the UEFI setup in case that was causing the boot problem (I think I'd seen some suggestion that it might).
The output from the COM port sadly only matches what we see on the screen before seabios loads:
Is there a way to get any further debug output from seabios itself?
Thanks
Geoff
There is a debug version but not 1.2.1 but older - probably 1.1.0 and advanced debug but both don't have AHCI fixes so they won't work with a SATA drive on some PC: 1.1.0 - https://github.com/FlyGoat/CSMWrap/releases/tag/1.1.0 Advanced Debug - https://github.com/FlyGoat/CSMWrap/issues/14#issuecomment-2889051446 Also 1.2.0-4 have normal debug + AHCI fixes (probably advanced AHCI debug) + OpROM usage:
@FlyGoat Please compile normal debug (not advanced) 1.2.1 CSMWrap like in 1.1.0 Thx.
There is a debug version but not 1.2.1 but older - probably 1.1.0 and advanced debug but both don't have AHCI fixes so they won't work with a SATA drive on some PC: 1.1.0 - https://github.com/FlyGoat/CSMWrap/releases/tag/1.1.0 Advanced Debug - #14 (comment) Also 1.2.0-4 have normal debug + AHCI fixes (probably advanced AHCI debug) + OpROM usage:
@FlyGoat Please compile normal debug (not advanced) 1.2.1 CSMWrap like in 1.1.0 Thx.
Thanks, that's much more useful. Log attached below.
I can see it's finding the bootable drives:
|534ef000| Registering bootable: AHCI/0: SATADOM-SV 3SE ATA-8 Hard-Disk (964 MiBytes) (type:2 prio:103 data:f5c90) ... |534f4000| Registering bootable: NVMe NS 1: 238475 MiB (488397168 512-byte blocks + 0-byte metadata) (type:2 prio:103 data:f5c40)
But then it seems to stick here:
|534f4000| NVMe initialization complete! \534f4000/ End thread phys_free 534f4000 (detail=0x534f59e0) All threads complete. Checking for bootsplash
Is the bootsplash code doing something odd, perhaps?
Edit: I notice the keyboard init is outputting an error:
|534f2000| i8042_command cmd=1060 |534f2000| i8042_wait_write |534f2000| i8042_wait_write |534f2000| ps2 command 1ff failed (aux=0)
Is it just that everything is working but it needs me to press escape, and the kb fails?
I note also a couple of xhci errors:
|534f3000| WARNING - Unable to allocate resource at configure_xhci:499! |534f4000| XHCI no devices found
Would I be better off trying to find a proper PS2 keyboard, since this is plugged into USB?
Enabling xhci handover in the uefi setup seems to remove the
|534f3000| WARNING - Unable to allocate resource at configure_xhci:499!
error; however I still get
|534f4000| XHCI no devices found
Also hitting esc at the serial console doesn't do anything, would you expect to be able to use that at that point?
I dug out a PS2 keyboard from the junk drawer, and hey presto! it lets me get to the boot menu. It even boots syslinux, although it then seems to get stuck, but that's a later issue, I guess.