> > I have created a skeleton framework, and iPXE calls my uri open function, > > and then int13h calls my read capacity function, but I don't know what > > signals > > completion of the read capacity call, so int13h thinks it timed out, and > > fails. > > read_capacity() (and other block request methods) create a new > sub-interface to be used for that request. The request is completed by > calling intf_shutdown() to shut down that sub-interface.
Yep. Got all that working now. I can boot Win8 from my new SAN protocol (well... up to the point where it wants my windows SAN driver, which I haven't finished yet). > > The other thing that my project requires is a way to also read and write > > data to the primary disk (eg the original int13h disk before my san block > > device replaced it). I can probably achieve this goal one way or another, > > but if > > there is a clean way of doing it (or if a method is already exposed somehow) > > then that would be great! > > There's nothing that currently exposes this specifically for use within > iPXE. We do remap the underlying INT13 drive so that it remains > accessible (under a different drive number). > So calling INT13 with the new driver number would work. How can I tell what driver number this would be? But actually, what I need to to is to hide this drive from Windows too (the windows loader at least - I'll take care of hiding it from windows itself), so probably I want to call the original INT13 vector and avoid the remapping... fun times ahead! Thanks James _______________________________________________ ipxe-devel mailing list [email protected] https://lists.ipxe.org/mailman/listinfo.cgi/ipxe-devel
