ACK/Cmnt: [SRU][F:linux-bluefield][PATCH v2 0/1] UBUNTU: SAUCE: mlxbf_gige: clear valid_polarity upon open

Tim Gardner tim.gardner at canonical.com
Wed Sep 8 16:59:07 UTC 2021


Acked-by: Tim Gardner <tim.gardner at canonical.com>

The '[Other]' info in the bug report is no longer accurate. The version 
will remain at 1.24.

On 9/8/21 9:57 AM, David Thompson wrote:
> BugLink: https://bugs.launchpad.net/bugs/1942932
> 
> SRU Justification:
> 
> [Impact]
> * The oob_net0 interface, managed by the mlxbf_gige driver, will be
>    up and enabled but can stop processing received packets if three
>    specific conditions occur:
>    1) driver has received more than (N * RxRingSize) packets but less
>       than (N+1 * RxRingSize) packets where N is an odd number.
>       Note: the command "ethtool -g oob_net0" will display the current
>       receive ring size, which currently defaults to 128.
>    2) the driver's interface was disabled via "ifconfig oob_net0 down"
>       during the window described in #1.
>    3) the driver's interface is re-enabled via "ifconfig oob_net0 up"
> 
>    Without this fix, the driver needs to be unloaded and re-loaded
>    to correct this condition.
> 
> [Fix]
> * The fix is to ensure that the driver's "valid_polarity" field
>    is cleared during the open() method so that it always matches
>    the receive polarity used by hardware.
> 
> [Test Case]
> * To test this fix, need to send in a specific number of packets to
>    the "oob_net0" interface, then issue "ifconfig oob_net0 down" to
>    disable the interface.  A subsequent "ifconfig oob_net0 up" command
>    will succesfully bring up the "oob_net0" interface and start the
>    processing of received packets.
> 
> [Regression Potential]
> * There is low risk of this fix causing any problems.
> 
> [Other]
> * n/a
> 
> David Thompson (1):
>    UBUNTU: SAUCE: mlxbf_gige: clear valid_polarity upon open
> 
>   drivers/net/ethernet/mellanox/mlxbf_gige/mlxbf_gige_main.c | 7 +++++++
>   1 file changed, 7 insertions(+)
> 

-- 
-----------
Tim Gardner
Canonical, Inc



More information about the kernel-team mailing list