vendor-reset

Coincidentally, the first issue (referencing Navi 21) was the one I started these experiments with, and this turned out to be pretty informative. Our Navi 21 would almost always go AWOL after a test run had been completed, requiring a full reboot. At some point, I noticed that this only happened when our test runner was driving the test; I never had an issue when testing interactively. I eventually realized that our test driver was simply killing the VM when the test was done, which is fine for a CPU-based test, but this messed with the GPU's state. When working interactively, I was always shutting down the host cleanly, which apparently resolved this. A patch to our test runner to cleanly shut down VMs fixed this. And I've had no luck with iGPUs, as referenced by the second issue. From what I understand, I don't think that consumer AMD G

I use the amdgpu driver and my luck has not been as good as yours. Can't sleep my PC without having it wake up to fill my logs with spam [0] and eventually crash. Then there is the (in)famous AMD reset bug that makes AMD a real headache to use with GPU passthrough. The card can't be properly reset when the VM shuts down so you have to reboot the PC to start the VM a second time. There are workarounds but they only work on some cards & scenarios [1] [2]. This problem goes back to around the 390 series cards so they've had forever to properly implement reset according to the pci spec but haven't. nvidia handles this flawlessly [0] https://gitlab.freedesktop.org/drm/amd/-/issues/3911 [1] https://github.com/gnif/vendor-reset [2] https://github.com/inga-lovinde/RadeonResetBugFix

It depends on the use case. AMD is notorious for their "AMD reset bug" when passing a GPU through to a VM using VFIO. Restart the guest and most cards will lock up because AMD doesn't handle PCI resets properly. You then need to reboot the host to fix it (!). This has been a problem since Polaris if not before and AMD hasn't fixed it, despite knowing full well that the problem exists. At least in this regard, NVIDIA (and Intel as far as I know) work fine. The community has been able to come up with a workaround for some older cards but the problem persists even in their current cards. https://github.com/gnif/vendor-reset

I’m not sure what you mean by forwarding, but if you mean regular gpu passthrough and the reset bugs with amd gpus, then for the rx580 vendor-reset should typically work: https://github.com/gnif/vendor-reset Does not work with the lower 6000 (below 6800) and 7000 series that can also have reset issues.

Unfortunately AMD cards suffer from a reset bug, still, when used with passthrough. The reset bug being that you can pass through the card fine, once. But if you try to pass it through again (or the card experiences an issue and needs to reset), they get caught in some kind of bad state and won’t work until power is removed and restored. Which requires a reboot or a only slightly less disruptive dance with system power states. For vega and 5000 series gpu’s, there’s https://github.com/gnif/vendor-reset Incidentally, nvidia gpus are so good at resetting, they’ve probably done so without you noticing. If the screen ever goes black for a fraction of a second and returns in normal usage, it was probably because it reset itself. The lower 6000 series lower than the 6800’s for example may or may not have the issue. It seems most “reference” car

It works with some cards, not with others. Eg. for Radeon Pro W5500 there's no known card reset method that works (no method from https://github.com/gnif/vendor-reset works) so I had to do S3 suspend before running a VM with systemctl suspend or with rtcwake -m mem -s 2 Now I have additional RTX 2070 and it works ok.