> On May 1, 2023, at 6:59 PM, Alberto Bursi <bobafetthotm...@gmail.com> wrote:
>
>
>
> On 01/05/23 06:40, Philip Prindeville wrote:
>>> On Apr 28, 2023, at 11:18 PM, Elliott Mitchell <ehem+open...@m5p.com> wrote:
>>>
>>> On Fri, Apr 28, 2023 at 12:04:15PM -0600, Philip Prindeville wrote:
>>>>
>>>>> [snip]
>>>
>>>> See above: the radios and antennae I can get as add-ons for a Xeon-D 1U
>>>> pizza box or even an APU6 mPCIe slot are vastly inferior to what ODM
>>>> purpose-built hardware like an U6-LR can do in terms of cost and
>>>> performance.
>>>>
>>>> Um... you can't "virtualize" WiFi in any VM I've ever seen.
>>>
>>> You can though pass PCIe devices to a VM. The hardware will physically
>>> attach to the control host, but a VM will be able to do anything it wants
>>> with it.
>> So the guest has the potential to crash or hang the host?
>
> Passthrough devices aren't really handled by the host OS (it's commonly
> required to blacklist their kernel drivers to make sure the host does NOT
> touch them) so in most cases if something hard-crashes and a kernel panics
> it's still all drama that happens inside the VM. This is how it usually goes
> in my experience.
>
> GPU passthrough can cause a whole host to lock up and require power cycle.
> I've seen that happen in the wild only once (it was repeatable, fun times),
> but imho it's a IOMMU or GPU issue not a "normal" thing that should happen.
>
> -Alberto
You can also lock up the PCIe bus so that the CPU can't access the bus or
bus-attached devices like disk controllers, network interfaces, etc.
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
