[SqNOS] Monticello update: USB first version sort of coming alive :)

Gerardo Richarte gera at corest.com
Sun Sep 14 18:34:15 UTC 2008


Name: SqueakNOS-gera.47
Author: gera
Time: 14 September 2008, 3:30 pm
UUID: 37560d3f-2be9-447d-aa75-2fb7ba966a1c
Ancestors: SqueakNOS-gera.46

Commented out mouse scrolling, because although it works very nicely in
VMWare, it hung the keyboard when booted for real on my Laptop.

USB is kind of working, sort of, maybe :)

The scheduling for talking to USB devices is somehow complex, right now
what we have is more than very basic, and the only thing that works is
USBUHCIController >> #testGetDeviceDescriptor

In a VMWare, I booted, connected a USB device (only one, from the VMWare
menu), and did:

Computer current usbUHCIController testGetDeviceDescriptor

Inspecting this brings back a 2 elements array, the second is the
buffer, if in the inspector you do:

(self second stringAt: 1 size: 1000) asByteArray

That's going to be the answer from the device. you can compare the bytes
from the device, 4 of the first few are the device ID. You can compare
that to what lsusb returns (in linux), or something similar on windows
(don't really remember right now how to get the USB device IDs on windows).

When booting for real, this works much better than previous versions,
but I had not finished testing it, as my laptop has 4 UHCI controllers,
and I've always been testing in the first (and apparently, that's not
where I'm plugging my devices).

Good luck! I'll keep working a little bit more :)


More information about the SqueakNOS mailing list