---- "Thomas D. Briglia" <briglia at stanford.edu> wrote: > > Thanks Ben, (thanks for your reply too Bernd) yet that was one of the > first things I tried! I tried every iteration of a Solaris device name > or mount point I could think of.
Note to self: This needs documentation. I hope to expand the Qemu project page with some really cool things I've done over the last couple of years so folks can get more use out of it. I have the week off during Christmas, so I'll probably get to it then. > I'll try one more time though just for sanity sake when I fire up that > enviromnent later today. :-) > > Based on the qemu docs though and the example usage of the "change > cdrom" verbiage in the FAQ, I assumed that qemu just did some kind of > mapping so that whatever your host OS was, /dev/cdrom under the guest > would map to whatever the true cdrom device was on the Host OS. No, because the cdrom that Solaris presents to Qemu acutally uses the raw device. I actually worked on part of this code with help from Juergen Keil and Mike Riley (from Sun), and it's been working AFAIK. > I also though the "[not inserted]" output from the "info block" command > might mean that qemu does not see the cdrom. it's supposed to mean it can't see the CDROM. In the past, I'm pretty sure this was working. > I'll try one more time using the Solaris device name just to make sure I > have tried everything . . . Yeah, volmgr really makes this stuff harder than it has to be, but once you get the Solaris way of doing things, volmgr becomes a non-issue. Let me know if you have any problems. Ben
