[SRU][N:linux-gcp][PATCH 0/2] Support larger gVNIC queue depth on Gen3+ GCE VMs

Rickey Castillo rickey.castillo.valenzuela at canonical.com
Wed Aug 19 00:15:30 UTC 2026


BugLink: https://bugs.launchpad.net/bugs/2153950

SRU Justification

[Impact]

Currently, the maximum queue depth supported for gVNIC on overcommitted Gen3+ 
VMs such as N4 is 1K, whereas the maximum queue depth supported on Gen1/Gen2 VMs 
is 2K. Customers who are migrating their workloads from N2 to N4 have requested 
higher queue depth support on N4 VMs. This patchset was already applied on GCP
kernel, but there are two commit fixes for one of the prerequisites. Two 
additional patches were included to fix the commit 5e37d8254e7f ("gve: Add header 
split data path") that were added to upstream. Details on the backport are in each 
commit message.

[Fix]

Target patches:
d676c9a73bdc ("gve: fix header buffer corruption with header-split and HW-GRO")
6bf14575c655 ("gve: fix zero-length skb frag with header-split")

[Test Plan]

* Compile Tested
* The following output indicates the feature is missing:
    $ sudo ethtool -G ens3 rx 4096
    netlink error: Operation not supported
  So the test plan for Canonical was to confirm that `sudo ethtool -G enp0s0 rx
  4096` exits successfully without an error.
* As well as this command, it should return 4096 as max no 1024:
  ethtool -g enp0s0

[Regression potential]

Regressions should only be possible when using the gVNIC device for 
networking (which depends on the gve driver), and could most likely result in 
networking failures on those configurations.

[Other]

SF: 00433236 (for GCP variants)
PIT: 509371317 (for GKE variants)

Ankit Garg (1):
  gve: fix header buffer corruption with header-split and HW-GRO

Jordan Rhee (1):
  gve: fix zero-length skb frag with header-split

 drivers/net/ethernet/google/gve/gve_rx_dqo.c | 32 ++++++++++++++------
 1 file changed, 22 insertions(+), 10 deletions(-)

-- 
2.53.0




More information about the kernel-team mailing list