2021-10-14 22:24:51 |
dann frazier |
bug |
|
|
added bug |
2021-10-14 22:24:51 |
dann frazier |
attachment added |
|
nvidia-peermem-test.sh https://bugs.launchpad.net/bugs/1947206/+attachment/5533001/+files/nvidia-peermem-test.sh |
|
2021-10-14 22:25:02 |
dann frazier |
nominated for series |
|
Ubuntu Hirsute |
|
2021-10-14 22:25:02 |
dann frazier |
bug task added |
|
linux (Ubuntu Hirsute) |
|
2021-10-14 22:25:10 |
dann frazier |
linux (Ubuntu Hirsute): status |
New |
In Progress |
|
2021-10-14 22:25:13 |
dann frazier |
linux (Ubuntu Hirsute): assignee |
|
dann frazier (dannf) |
|
2021-10-14 22:27:14 |
dann frazier |
nominated for series |
|
Ubuntu Impish |
|
2021-10-14 22:27:14 |
dann frazier |
bug task added |
|
linux (Ubuntu Impish) |
|
2021-10-14 22:27:14 |
dann frazier |
nominated for series |
|
Ubuntu Focal |
|
2021-10-14 22:27:14 |
dann frazier |
bug task added |
|
linux (Ubuntu Focal) |
|
2021-10-14 22:27:20 |
dann frazier |
linux (Ubuntu Focal): status |
New |
In Progress |
|
2021-10-14 22:27:23 |
dann frazier |
linux (Ubuntu Focal): assignee |
|
dann frazier (dannf) |
|
2021-10-14 22:28:18 |
dann frazier |
description |
[Impact]
Nvidia notified me via private email that they'd discovered some issues with the ib_peer_memory patch we are carrying in hirsute/impish and sent me a patch intended to resolve them. My knowledge of these changes is limited to what is mentioned in the commit message:
- Allow clients to opt out of unmap during invalidation
- Fix some bugs in the sequencing of mlx5 MRs
- Enable ATS for peer memory
[Test Case]
ib_write_bw from the perftest package, rebuilt with CUDA support, can be used as a smoke test of this feature. I'll attach a sample test script here. I've verified this test passes with the kernels in the archive, and continues to pass with the provided patch applied.
[Fix]
Nvidia has emailed me fixes for both trees. They are not currently available in a public tree elsewhere, though I'm told at some point they should end up in a branch here:
https://git.kernel.org/pub/scm/linux/kernel/git/leon/linux-rdma.git/
[What could go wrong]
The only known use case for ib_peer_memory are Nvidia GPU users making use of the GPU PeerDirect feature where GPUs can share memory with one another over an Infiniband network. Bugs here could cause problems (hangs, crashes, corruption) with such workloads. |
[Impact]
Nvidia notified me via private email that they'd discovered some issues with the ib_peer_memory patch we are carrying in hirsute/impish and sent me a patch intended to resolve them. My knowledge of these changes is limited to what is mentioned in the commit message:
- Allow clients to opt out of unmap during invalidation
- Fix some bugs in the sequencing of mlx5 MRs
- Enable ATS for peer memory
[Test Case]
ib_write_bw from the perftest package, rebuilt with CUDA support, can be used as a smoke test of this feature. I'll attach a sample test script here. I've verified this test passes with the kernels in the archive, and continues to pass with the provided patch applied.
[Fix]
Nvidia has emailed me fixes for both trees. They are not currently available in a public tree elsewhere, though I'm told at some point they should end up in a branch here:
https://git.kernel.org/pub/scm/linux/kernel/git/leon/linux-rdma.git/
[What could go wrong]
The only known use case for ib_peer_memory are Nvidia GPU users making use of the GPU PeerDirect feature where GPUs can share memory with one another over an Infiniband network. Bugs here could cause problems (hangs, crashes, corruption) with such workloads. |
|
2021-11-22 17:09:52 |
Stefan Bader |
linux (Ubuntu Impish): importance |
Undecided |
Medium |
|
2021-11-22 17:09:56 |
Stefan Bader |
linux (Ubuntu Hirsute): importance |
Undecided |
Medium |
|
2021-11-22 17:10:21 |
Stefan Bader |
linux (Ubuntu Hirsute): status |
In Progress |
Fix Committed |
|
2021-11-22 17:11:11 |
Stefan Bader |
linux (Ubuntu Impish): status |
In Progress |
Fix Committed |
|
2021-11-29 18:46:13 |
Ubuntu Kernel Bot |
tags |
|
verification-needed-impish |
|
2021-11-29 18:47:43 |
Ubuntu Kernel Bot |
tags |
verification-needed-impish |
verification-needed-hirsute verification-needed-impish |
|
2021-12-02 23:56:46 |
dann frazier |
tags |
verification-needed-hirsute verification-needed-impish |
verification-done-impish verification-needed-hirsute |
|
2021-12-03 16:25:50 |
dann frazier |
tags |
verification-done-impish verification-needed-hirsute |
verification-done-hirsute verification-done-impish |
|
2021-12-09 09:42:19 |
Stefan Bader |
linux (Ubuntu Focal): importance |
Undecided |
Medium |
|
2021-12-15 13:04:26 |
Kleber Sacilotto de Souza |
linux (Ubuntu Focal): status |
In Progress |
Fix Committed |
|
2022-01-04 14:34:32 |
Launchpad Janitor |
linux (Ubuntu Hirsute): status |
Fix Committed |
Fix Released |
|
2022-01-04 14:34:32 |
Launchpad Janitor |
cve linked |
|
2021-4002 |
|
2022-01-04 14:34:32 |
Launchpad Janitor |
cve linked |
|
2021-43267 |
|
2022-01-04 14:36:58 |
Launchpad Janitor |
linux (Ubuntu Impish): status |
Fix Committed |
Fix Released |
|
2022-01-07 06:47:48 |
Timo Aaltonen |
tags |
verification-done-hirsute verification-done-impish |
verification-done verification-done-hirsute verification-done-impish |
|
2022-01-11 23:35:57 |
Ubuntu Kernel Bot |
tags |
verification-done verification-done-hirsute verification-done-impish |
verification-done verification-done-hirsute verification-done-impish verification-needed-focal |
|
2022-01-17 16:49:01 |
Launchpad Janitor |
linux (Ubuntu): status |
In Progress |
Fix Released |
|
2022-01-25 15:41:52 |
dann frazier |
tags |
verification-done verification-done-hirsute verification-done-impish verification-needed-focal |
verification-done verification-done-focal verification-done-hirsute verification-done-impish |
|
2022-01-31 12:27:53 |
Launchpad Janitor |
linux (Ubuntu Focal): status |
Fix Committed |
Fix Released |
|
2022-02-07 15:05:16 |
Ubuntu Kernel Bot |
tags |
verification-done verification-done-focal verification-done-hirsute verification-done-impish |
verification-done verification-done-focal verification-done-hirsute verification-done-impish verification-needed-bionic |
|