Incomplete Product details of NVIDIA GPU after pass-through.

Bug #1826103 reported by bugproxy
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Ubuntu
New
Undecided
Unassigned

Bug Description

== Comment: #0 - SANTWANA SAMANTRAY <email address hidden> - 2019-03-15 07:42:57 ==
---Problem Description---
Incomplete Product details of the NVIDIA GPU in the guest , after pass-through.
The below NVIDIA GPU is present on the host which is later pass-through'ed to guest.
On Host:
0004:04:00.0 3D controller [0302]: NVIDIA Corporation GV100GL [Tesla V100 SXM2 32GB] [10de:1db5] (rev a1)
0004:05:00.0 3D controller [0302]: NVIDIA Corporation GV100GL [Tesla V100 SXM2 32GB] [10de:1db5] (rev a1)

After the passthrough the adapter is listed as below in the guest:
0001:00:10.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)
0001:00:11.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)

The complete product information : GV100GL [Tesla V100 SXM2 32GB is missing in the guest pci details.

The complete product information is essential for user awareness of the device model.

The expected output in the guest should be:
0001:00:10.0 3D controller [0302]: NVIDIA Corporation GV100GL [Tesla V100 SXM2 32GB] [10de:1db5] (rev a1)
0001:00:11.0 3D controller [0302]: NVIDIA Corporation GV100GL [Tesla V100 SXM2 32GB] [10de:1db5] (rev a1)

Snip of hardware details of 1 GPU on host and guest:

# lshw [Host]
-display
                   description: 3D controller
                   product: GV100GL [Tesla V100 SXM2 32GB]
                   vendor: NVIDIA Corporation
                   physical id: 0
                   bus info: pci@0035:03:00.0
                   version: a1
                   slot: GPU3
                   width: 64 bits
                   clock: 33MHz
                   capabilities: pm msi pciexpress bus_master cap_list
                   configuration: driver=vfio-pci latency=0
                   resources: iomemory:622800-6227ff iomemory:622880-62287f irq:735 memory:620c280000000-620c280ffffff memory:6228000000000-62287ffffffff memory:6228800000000-6228801ffffff

# lshw [Guest]
display:0
          description: 3D controller
          product: NVIDIA Corporation <--------------- Complete Details Missing
          vendor: NVIDIA Corporation
          physical id: 10
          bus info: pci@0001:00:10.0
          version: a1
          slot: vfio_vfio-pci:0001:00:10.0
          width: 64 bits
          clock: 33MHz
          capabilities: pm msi pciexpress bus_master cap_list
          configuration: driver=nvidia latency=0
          resources: iomemory:22000-21fff iomemory:22080-2207f irq:47 memory:200100000000-200100ffffff memory:220000000000-2207ffffffff memory:220800000000-220801ffffff

Machine Type = Witherspoon

---Debugger---
A debugger is not configured

---Steps to Reproduce---
1. Configure the guest with GPU pass-through.
2. After the guest starts, check the pass-through'ed device details:
lspci -nn
0000:00:01.0 Ethernet controller [0200]: Red Hat, Inc Virtio network device [1af4:1000]
0000:00:02.0 SCSI storage controller [0100]: Red Hat, Inc Virtio SCSI [1af4:1004]
0000:00:03.0 USB controller [0c03]: Red Hat, Inc. QEMU XHCI Host Controller [1b36:000d] (rev 01)
0000:00:04.0 Unclassified device [00ff]: Red Hat, Inc Virtio memory balloon [1af4:1002]
0001:00:10.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)
0001:00:11.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)
0001:00:12.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0001:00:13.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0001:00:14.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0001:00:15.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0001:00:16.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0001:00:17.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:01.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)
0002:00:02.0 3D controller [0302]: NVIDIA Corporation Device [10de:1db5] (rev a1)
0002:00:03.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:04.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:05.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:06.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:07.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)
0002:00:08.0 Bridge [0680]: IBM Device [1014:04ea] (rev 01)

Contact Information = Santwana <email address hidden>

== Comment: #1 - SANTWANA SAMANTRAY <email address hidden> - 2019-03-15 07:44:30 ==

== Comment: #2 - SANTWANA SAMANTRAY <email address hidden> - 2019-03-15 07:48:54 ==

== Comment: #3 - SANTWANA SAMANTRAY <email address hidden> - 2019-03-15 07:52:34 ==

== Comment: #9 - SANTWANA SAMANTRAY <email address hidden> - 2019-04-01 04:27:39 ==

Santwana

== Comment: #10 - Maxiwell Salvador Garcia <email address hidden> - 2019-04-05 10:30:46 ==
Hi,

The problem is the guest's pci ids database, located in /usr/share/misc/pci.ids. The GPU model "Tesla V100 SXM2 32GB", PCI ID 1db5, was added in the database on 2018-03-30:
https://github.com/pciutils/pciids/commit/d8c893dfff4d

The pci.ids version in the guests with Ubuntu 18.04 and 18.04.1 that I tested has an older version:

vm1:
user@ubuntu:~$ cat /etc/os-release | grep VERSION=
VERSION="18.04 LTS (Bionic Beaver)"
user@ubuntu:~$ cat /usr/share/misc/pci.ids | grep Version:
# Version: 2017.03.16

vm2:
user@ubuntu:~$ cat /etc/os-release | grep VERSION=
VERSION="18.04.1 LTS (Bionic Beaver)"
user@ubuntu:~$ cat /usr/share/misc/pci.ids | grep Version:
# Version: 2017.03.16

So, to lspci or lshw show this GPU model correctly, the pci.ids must be a version greater than or equal to 2018-03-30. A guest with Ubuntu 18.04.2 worked correctly, because:

vm3:
root@ubuntu:~# cat /etc/os-release | grep VERSION=
VERSION="18.04.2 LTS (Bionic Beaver)"
root@ubuntu:~# cat /usr/share/misc/pci.ids | grep Version:
# Version: 2018.07.21

To update this database manually, you can run "update-pciids".

== Comment: #13 - Leonardo Augusto Guimaraes Garcia <email address hidden> - 2019-04-23 11:48:28 ==
Canonical,

This is an issue with stock Ubuntu 18.04.1. Is it possible to update the PCI ID database on this version?

Revision history for this message
Ubuntu Foundations Team Bug Bot (crichton) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. It seems that your bug report is not filed about a specific source package though, rather it is just filed against Ubuntu in general. It is important that bug reports be filed about source packages so that people interested in the package can find the bugs about it. You can find some hints about determining what package your bug might be about at https://wiki.ubuntu.com/Bugs/FindRightPackage. You might also ask for help in the #ubuntu-bugs irc channel on Freenode.

To change the source package that this bug is filed about visit https://bugs.launchpad.net/ubuntu/+bug/1826103/+editstatus and add the package name in the text box next to the word Package.

[This is an automated message. I apologize if it reached you inappropriately; please just reply to this message indicating so.]

tags: added: bot-comment
Revision history for this message
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla

------- Comment From <email address hidden> 2019-04-24 09:41 EDT-------
Changed Target milestone to 18.04.1

tags: added: architecture-ppc64le bugnameltc-176140 severity-medium targetmilestone-inin18041
Revision history for this message
bugproxy (bugproxy) wrote : Guest sosreport

Default Comment by Bridge

Revision history for this message
bugproxy (bugproxy) wrote : Host sosreport

Default Comment by Bridge

Revision history for this message
bugproxy (bugproxy) wrote : Guest XML

Default Comment by Bridge

bugproxy (bugproxy)
tags: added: targetmilestone-inin18041genesis2
removed: targetmilestone-inin18041
bugproxy (bugproxy)
tags: added: targetmilestone-inin18041
removed: targetmilestone-inin18041genesis2
Revision history for this message
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla

------- Comment From <email address hidden> 2019-10-29 18:36 EDT-------
Dropping to P3 to match S3, as there seems to be no reason for a P2 designation.

Revision history for this message
bugproxy (bugproxy) wrote :

------- Comment From <email address hidden> 2020-07-03 15:41 EDT-------
All Ubuntu version after the 18.04.2 version (18.04.4 is the newest image) has an updated version of PCI IDS database. So, we can close this issue.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.