[v2] selftests: drv-net: Fix issues in devlink_rate_tc_bw.py

[PATCH v2 net 0/2] selftests: drv-net: Fix issues in devlink_rate_tc_bw.py

Posted by Carolina Jubran 3 weeks, 2 days ago

This series fixes issues in devlink_rate_tc_bw.py selftest that made
its checks unreliable and its documentation inconsistent with the
actual configuration.

V2:
- Dropped the patch that relaxed the total bandwidth check. Jakub
  suggested addressing the instability with interval-based measurement
  and by migrating to load.py. That will be handled in a follow-up.
- Link to V1: https://lore.kernel.org/netdev/20250831080641.1828455-1-cjubran@nvidia.com/

Thanks

Carolina Jubran (2):
  selftests: drv-net: Fix and clarify TC bandwidth split in
    devlink_rate_tc_bw.py
  selftests: drv-net: Fix tolerance calculation in devlink_rate_tc_bw.py

 .../drivers/net/hw/devlink_rate_tc_bw.py      | 100 ++++++++----------
 1 file changed, 43 insertions(+), 57 deletions(-)

-- 
2.38.1

Re: [PATCH v2 net 0/2] selftests: drv-net: Fix issues in devlink_rate_tc_bw.py

Posted by Simon Horman 3 weeks, 1 day ago

On Tue, Sep 09, 2025 at 01:13:51PM +0300, Carolina Jubran wrote:
> This series fixes issues in devlink_rate_tc_bw.py selftest that made
> its checks unreliable and its documentation inconsistent with the
> actual configuration.
> 
> V2:
> - Dropped the patch that relaxed the total bandwidth check. Jakub
>   suggested addressing the instability with interval-based measurement
>   and by migrating to load.py. That will be handled in a follow-up.
> - Link to V1: https://lore.kernel.org/netdev/20250831080641.1828455-1-cjubran@nvidia.com/
> 
> Thanks
> 
> Carolina Jubran (2):
>   selftests: drv-net: Fix and clarify TC bandwidth split in
>     devlink_rate_tc_bw.py
>   selftests: drv-net: Fix tolerance calculation in devlink_rate_tc_bw.py
> 
>  .../drivers/net/hw/devlink_rate_tc_bw.py      | 100 ++++++++----------
>  1 file changed, 43 insertions(+), 57 deletions(-)

Hi Carolina,

It isn't strictly related to these changes.
But CI flaggs that devlink_rate_tc_bw.py should
be present in the Makefile for in the same directory.

Given the wildcard in the Makefile I'm unsure if that
is true or not. Could you take a look?