michaelrizzo

Thanks Aa617! This worked. I had to do it twice. The first time I added group 34, restarted the array, tried to start the VM, got an error on group 38. Added group 38, restarted the array, BOOM, VM works. Actually, scratch that, I had one other operator error, I had to shut down the server, go into BIOS, and make sure it was configured to use the onboard VGA graphics at startup so as to not conflict with the video card. Once I took care of those two things, my Mac VM and my Windows 10 VM (which I created to confirm I hadn't made an XML error and had also failed to boot) both worked great! Thanks for all your help! Now to get iMessages working...

Thanks for the help archedraft! The only qemu file I'm finding is in /flash/config/plugins/dynamix.kvm.manager I'm a bit of a novice, am I looking in the right place? Is the file even supposed to be on the UNRAID flash? The current contents of the qemu file don't look anything like the bit that's in jwegman's example:

Hello all, I've been trying to pass through a GPU on my new server for my OS X VM. I followed the guide: https://macosxvirtualmachinekvm.wordpress.com/guide-mac-os-x-10-11-el-capitan-vm-on-unraid/ And have been able to get the system setup and running via VNC, but have had no luck getting a GPU passed through. I only want a GPU, don't need USB passthrough. I have a Radeon 4870 installed (which, if I get this working nicely, I intend to replace with a newer Radeon that's fully supported by OSX. The first thing I did was use the QEMU code included in the guide itself. The only changes I made were to fix the ' that don't copy/paste properly and update the host to the correct ID for my GPU (which I got from SysDevs). When I do this and try to start the VM, I get the following error: Now, I've googled around that error and found this thread: https://lime-technology.com/forum/index.php?topic=36110.0 I tried editing my syslinux.cfg and rebooted into the new UnRaid configuration, but it didn't solve my problem. I'm also wary of PCIe ACS Override (Should I be?), and kind of want to avoid that if necessary. I have read through a couple of threads with the syslinux.cfg fix but since it didn't immediately solve my problem, I'm hoping there's another way. I saw this post on page 25: https://lime-technology.com/forum/index.php?topic=43816.msg478749#msg478749 Where another guy was using Hostdev to pass through his GPU, and archedraft seems to say that if you don't need USB passthrough that's better than QEMU. If so, where should I drop the hostdev code into my XML? And what should I plug in for my address info? The OTHER possible solution I'm looking at is maybe moving the GPU to a different PCI-E slot, as I'm in a server board with a bunch of them. I'm in the second topmost slot because I didn't want to be right up against my RAM, and apparently that's PCIE6 on this board because it counts up from the bottom rather than down from the top. All of this is fairly new to me, I'm just doing my best to learn as much as I can from existing forum posts. Here's my current unworking XML: