ACK/Cmnt: [SRU][B/aws, F, G] EFA: add support for 0xefa1 devices

Stefan Bader stefan.bader at canonical.com
Thu Sep 24 13:46:35 UTC 2020


On 23.09.20 18:48, Kamal Mostafa wrote:
> BugLink: https://bugs.launchpad.net/bugs/1896791
> 
> AWS RDMA/efa driver: add support for new AWS EFA '0xefa1' devices.
> 
> These are cherry-picks and simple context backports of mainline commits
> to the AWS 'efa' driver, needed to support future AWS EFA Infiniband
> devices.
> 
>  -Kamal
> 
> 
> [Impact]
> 
> The following 4 mainline commits are required to support the new device
> features and ID:
> 
> d4f9cb5c5b22 RDMA/efa: Add EFA 0xefa1 PCI ID
> a5d87b698547 RDMA/efa: User/kernel compatibility handshake mechanism
> da2924bdca99 RDMA/efa: Expose minimum SQ size
> 556c811f24b3 RDMA/efa: Expose maximum TX doorbell batch
> 
> [Test Case]
> 
> New device is not yet generally available, but has been tested by AWS.
> 
> [Regression Potential]
> 
> Low regression potential; Affects only the EFA driver.
> 
> [Other Info]
> 
> Focal and Groovy generic kernels can easily support this patch set, so
> lets do that, to provide wider support and an updated driver source
> baseline for those.
> 
> Bionic can only support it in bionic/linux-aws (not generic), so lets do
> that.
> 
> -----
> 
> Gal Pressman (4):
>   RDMA/efa: Expose maximum TX doorbell batch
>   RDMA/efa: Expose minimum SQ size
>   RDMA/efa: User/kernel compatibility handshake mechanism
>   RDMA/efa: Add EFA 0xefa1 PCI ID
> 
>  .../infiniband/hw/efa/efa_admin_cmds_defs.h   | 15 ++++++-
>  drivers/infiniband/hw/efa/efa_com_cmd.c       |  2 +
>  drivers/infiniband/hw/efa/efa_com_cmd.h       |  2 +
>  drivers/infiniband/hw/efa/efa_main.c          |  6 ++-
>  drivers/infiniband/hw/efa/efa_verbs.c         | 42 +++++++++++++++++++
>  include/uapi/rdma/efa-abi.h                   | 15 ++++++-
>  6 files changed, 77 insertions(+), 5 deletions(-)
> 
While adding new drivers in SRU is always something I would frown upon, I guess
this is okay'ish. At least these changes are upstream so there is less danger of
things becoming stale and forgotten. And it is a very specific driver. And we
already <retracted> included the ENA stuff.

Acked-by: Stefan Bader <stefan.bader at canonical.com>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.ubuntu.com/archives/kernel-team/attachments/20200924/05eecab5/attachment.sig>


More information about the kernel-team mailing list