ACK/Cmnt: [PATCH 00/12] [B][C] Update Hi1822 driver to mainline ver

[Stefan Bader]

On 18.03.19 09:27, Ike Panhc wrote:
> BugLink: https://bugs.launchpad.net/bugs/1820187
> 
> Current driver for Huawei NIC has poor performance. When connect to
> 10G switch it has only 1-2Gb/s throughput and not stable. 
> 
> Cherry-pick patches from mainline can solve this issue to have
> stable and satisfied performance
> 
> $ iperf -c 10.228.68.102 -P2 -t 180 # kernel with this patchset
> ------------------------------------------------------------
> Client connecting to 10.228.68.102, TCP port 5001
> TCP window size: 85.0 KByte (default)
> ------------------------------------------------------------
> [  3] local 10.228.68.135 port 57200 connected with 10.228.68.102 port 5001
> [  4] local 10.228.68.135 port 57198 connected with 10.228.68.102 port 5001
> [ ID] Interval       Transfer     Bandwidth
> [  3]  0.0-180.0 sec  84.4 GBytes  4.03 Gbits/sec
> [  4]  0.0-180.0 sec  84.1 GBytes  4.01 Gbits/sec
> [SUM]  0.0-180.0 sec   168 GBytes  8.04 Gbits/sec
> 
> 
> Colin Ian King (1):
>   net: hinic: fix null pointer dereference on pointer hwdev
> 
> Eric Dumazet (1):
>   hinic: remove ndo_poll_controller
> 
> Nathan Chancellor (1):
>   hinic: Fix l4_type parameter in hinic_task_set_tunnel_l4
> 
> Xue Chaojing (7):
>   net-next/hinic:replace multiply and division operators
>   net-next/hinic:add rx checksum offload for HiNIC
>   net-next/hinic:fix a bug in set mac address
>   net-next/hinic: fix a bug in rx data flow
>   hinic: optmize rx refill buffer mechanism
>   net-next/hinic:add shutdown callback
>   net-next/hinic: replace disable_irq_nosync/enable_irq
> 
> Zhao Chen (2):
>   net-next: hinic: fix a problem in free_tx_poll()
>   net-next/hinic: add checksum offload and TSO support
> 
>  .../net/ethernet/huawei/hinic/hinic_hw_dev.c  |  13 +
>  .../net/ethernet/huawei/hinic/hinic_hw_dev.h  |   7 +
>  .../net/ethernet/huawei/hinic/hinic_hw_if.c   |  28 ++
>  .../net/ethernet/huawei/hinic/hinic_hw_if.h   |  14 +
>  .../net/ethernet/huawei/hinic/hinic_hw_qp.c   | 157 +++++++--
>  .../net/ethernet/huawei/hinic/hinic_hw_qp.h   |  33 +-
>  .../net/ethernet/huawei/hinic/hinic_hw_wq.c   |  69 ++--
>  .../net/ethernet/huawei/hinic/hinic_hw_wq.h   |   5 +-
>  .../net/ethernet/huawei/hinic/hinic_hw_wqe.h  | 101 ++++--
>  .../net/ethernet/huawei/hinic/hinic_main.c    |  61 ++--
>  .../net/ethernet/huawei/hinic/hinic_port.c    |  62 ++++
>  .../net/ethernet/huawei/hinic/hinic_port.h    |  28 ++
>  drivers/net/ethernet/huawei/hinic/hinic_rx.c  |  58 ++--
>  drivers/net/ethernet/huawei/hinic/hinic_rx.h  |   6 +-
>  drivers/net/ethernet/huawei/hinic/hinic_tx.c  | 321 +++++++++++++++++-
>  15 files changed, 825 insertions(+), 138 deletions(-)
> 
Wholesale backports into a released LTS kernel will always earn you grumpy
points. Stable updates should be specific to fix a problem.

That said, since it is limited to that one specific driver which you certainly
have tested very well (right?), and we have been accepting similar things in the
past (though I wished we had not),

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/20190327/f1f23906/attachment-0001.sig>