Issues with GTX960 on CentOS7 using bhyve PCI passthru (FreeBSD 11-RC2)

[Peter Grehan]

Hi Dom,

> Bhyve's ACPI table produces this error linux-side regardless of
> "pci=" setting:
>
> [    0.163085] acpi PNP0A03:00: host bridge window
> [0xd000000000-0xd0100fffff] (ignored, not CPU addressable)

  That one is most likely a bug in bhyve, where the space used for 64-bit
BAR placement isn't tested against the max physaddr width of the host CPU.

  To confirm, would you be able to report on this value on your system ?

# sudo pkg install cpuid
# cpuid | grep ^80000008
80000008 00003027 00000000 00000000 00000000
                ^^
This is the output from a Xeon E3-1220 v3: 0x27 == 39 bits of phys
address (0x8000000000 max)

  0xd000000000 requires >= 40 bits.

> 2. "can't derive routing for PCI INT"
>
> Linux-side dmesg related output:
>
> [    1.677168] nvidia 0000:00:06.0: can't derive routing for PCI INT
> A [    1.677600] nvidia 0000:00:06.0: PCI INT A: no GSI
  ...
> Host-side info (when GTX960 is NOT configured as a pass-thru dev):
...
> cap 05[68] = MSI supports 1 message, 64 bit enabled with 1 message

  The device has an MSI capability, but the nvidia driver may not use it.
bhyve PCI passthrough requires the use of MSI/MSI-x interrupts, and 
doesn't support using legacy interrupts.

  This could be confirmed from the output of /proc/interrupts when 
booting Linux on the system.

later,

Peter.