I'd like first to thank you all for this fantastic website.
I'm planning to implement an xHCI for my OS; but I don't want to implement previous OHCI, UHCI and EHCI specifications.
I'm currently reading the specs. If I understood the thing correctly, I can have an USBD with only an xHCI behind it. Am I correct?
I'd like to have support for USB3.0 only, but I read that implementing previous versions is necessary for full support of this specification...
Thanks for your advice
Implementing an xHCI that supports USB3.0 only
Re: Implementing an xHCI that supports USB3.0 only
I haven't been took in-depth look of xhci.
but form what i know - yes, the xhci backward support is now in-chip (instead of companion chips)
i have no ideal what does it mean from implementation point of view.
but form what i know - yes, the xhci backward support is now in-chip (instead of companion chips)
i have no ideal what does it mean from implementation point of view.
Re: Implementing an xHCI that supports USB3.0 only
I'm on the way to read the whole USB/xHCI specs, so I'll post what I find out later.
Re: Implementing an xHCI that supports USB3.0 only
At http://prettyos.de we are just developing an xHCI driver. We are going to try to integrate it into the general usb interface already working with UHCI, OHCI, EHCI.
http://www.henkessoft.de/OS_Dev/OS_Dev3.htm (OSDEV)
http://www.c-plusplus.de/forum/viewforu ... is-62.html
irc.euirc.net #PrettyOS
http://www.c-plusplus.de/forum/viewforu ... is-62.html
irc.euirc.net #PrettyOS
Re: Implementing an xHCI that supports USB3.0 only
update xhci/usb3:
PrettyOS has now a testwise working xhci module. With vmware we observed correct usb enumeration (slot state: enabled (0) ==> addressed (0)). With real hardware there are still "parameter errors". Perhaps anybody could help to sort that out.
PS: For a working xhci we had to implement acpi, apic, msi (for interrupts).
http://prettyos.de
PrettyOS has now a testwise working xhci module. With vmware we observed correct usb enumeration (slot state: enabled (0) ==> addressed (0)). With real hardware there are still "parameter errors". Perhaps anybody could help to sort that out.
PS: For a working xhci we had to implement acpi, apic, msi (for interrupts).
http://prettyos.de
http://www.henkessoft.de/OS_Dev/OS_Dev3.htm (OSDEV)
http://www.c-plusplus.de/forum/viewforu ... is-62.html
irc.euirc.net #PrettyOS
http://www.c-plusplus.de/forum/viewforu ... is-62.html
irc.euirc.net #PrettyOS