On 14.02.20 14:23, [email protected] wrote:
Hi all,
I am facing issues related to the integrated GPU (iGPU) on an Intel
consumer SoC.
First, some generic system information:
Product: Dell OptiPlex 3060
CPU: Intel(R) Core(TM) i5-8400 CPU @ 2.80GHz
Now, the Jailhouse stack I'm using:
GIT revision: 92db71f257fabd3c08fa4b99498fa61a41ea831d (0.12)
Root cell config: attached
After enabling Jailhouse using:
|
jailhouse enable dell-optiplex.cell
|
I observe logspam in the kernel logs (dmesg) like the following:
|
[drm:gen8_de_irq_handler [i915]]*ERROR*Faulterrors on pipe A:0x00000080
|
The same message is repeated numerous times each second.
Unloading Jailhouse stops the logspam.
Reading from the /dev/jailhouse device, I can see there are some VT-d
issues with the iGPU PCIe device at 00:02.0
|
<missed84 bytesofconsolelog>
Using x2APIC
Page pool usage after early setup: mem 59/975, remap 0/131072
Initializing processors:
CPU 1... (APIC ID 2) OK
CPU 4... (APIC ID 8) OK
CPU 2... (APIC ID 4) OK
CPU 3... (APIC ID 6) OK
CPU 0... (APIC ID 0) OK
CPU 5... (APIC ID 10) OK
Initializing unit: VT-d
DMAR unit @0xfed90000/0x1000
DMAR unit @0xfed91000/0x1000
Reserving 120 interrupt(s) for device 00:1e.7 at index 0
Initializing unit: IOAPIC
Initializing unit: Cache Allocation Technology
Initializing unit: PCI
Adding PCI device 00:00.0 to cell "RootCell"
Adding PCI device 00:02.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:02.0 at index 120
Adding PCI device 00:08.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:08.0 at index 121
Adding PCI device 00:12.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:12.0 at index 122
Adding PCI device 00:14.0 to cell "RootCell"
Reserving 8 interrupt(s) for device 00:14.0 at index 123
Adding PCI device 00:14.2 to cell "RootCell"
Adding PCI device 00:16.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:16.0 at index 131
Adding PCI device 00:17.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:17.0 at index 132
Adding PCI device 00:1c.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:1c.0 at index 133
Adding PCI device 00:1c.5 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:1c.5 at index 134
Adding PCI device 00:1f.0 to cell "RootCell"
Adding PCI device 00:1f.3 to cell "RootCell"
Reserving 1 interrupt(s) for device 00:1f.3 at index 135
Adding PCI device 00:1f.4 to cell "RootCell"
Adding PCI device 00:1f.5 to cell "RootCell"
Adding PCI device 01:00.0 to cell "RootCell"
Reserving 4 interrupt(s) for device 01:00.0 at index 136
Adding PCI device 02:00.0 to cell "RootCell"
Reserving 1 interrupt(s) for device 02:00.0 at index 140
Page pool usage after late setup: mem 279/975, remap 16395/131072
Activating hypervisor
VT-d fault event reported by IOMMU 0:
Source Identifier (bus:dev.func): 00:02.0
Fault Reason: 0x6 Fault Info: 6a1cea1 Type 1
|
This is a translation fault, read permission at physical address
0x6a1cea1000 was denied. Check if there is a memory region covering this
already. It might be an ACPI-reserved region that Linux is not reporting
but that has to be added due to the GPU accesses.
Jan
|
lspci -nnk
...
00:02.0VGA compatible controller [0300]:IntelCorporationUHD
Graphics630(Desktop)[8086:3e92]
DeviceName:Onboard-Video
Subsystem:DellDevice[1028:085c]
Kerneldriver inuse:i915
Kernelmodules:i915
|
Unfortunately I'm not really familiar with VT-d and the Intel
virtualization stack in general. Does "Fault Reason: 0x6" ring a bell to
anyone?
I'm happy to provide more information if needed.
Cheers
--
Siemens AG, Corporate Technology, CT RDA IOT SES-DE
Corporate Competence Center Embedded Linux
--
You received this message because you are subscribed to the Google Groups
"Jailhouse" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jailhouse-dev/9c349699-be0d-c112-041e-7f6485132b3f%40siemens.com.
