Merge tag 'mlx5-updates-2019-04-22' of git://git.kernel.org/pub/scm/linux/kernel...

Merge tag 'mlx5-updates-2019-04-22' of git://git./linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2019-04-22

This series includes updates to mlx5e driver RX data path and some
significant XDP RX/TX improvements to overcome/mitigate HW and PCIE
bottlenecks.

From Tariq:
1) Some Enhancements in rq->flags
2) Stabilize RX packet rate (on Striding RQ) with
multiple outstanding UMR posts
In this patch, we add support for multiple outstanding UMR posts,
 to allow faster gap closure between consuming MPWQEs and reposting
them back into the WQ.

Performance test:
As expected, huge improvement in large-scale (48 cores).

xdp_redirect_map, 64B UDP multi-stream.
Redirect from ConnectX-5 100Gbps to ConnectX-6 100Gbps.
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz.

Before: Unstable, 7 to 30 Mpps
After:  Stable,   at 70.5 Mpps

From Shay:
3) XDP, Inline small packets into the TX MPWQE in XDP xmit flow

Upon high packet rate with multiple CPUs TX workloads, much of the HCA's
resources are spent on prefetching TX descriptors, thus affecting
transmission rates.
This patch comes to mitigate this problem by moving some workload to the
CPU and reducing the HW data prefetch overhead for small packets (<= 256B).

When forwarding packets with XDP, a packet that is smaller
than a certain size (set to ~256 bytes) would be sent inline within
its WQE TX descrptor (mem-copied), when the hardware tx queue is congested
beyond a pre-defined water-mark.

Performance:
    Tested packet rate for UDP 64Byte multi-stream
    over two dual port ConnectX-5 100Gbps NICs.
    CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz

    * Tested with hyper-threading disabled

    XDP_TX:

    |          | before | after   |       |
    | 24 rings | 51Mpps | 116Mpps | +126% |
    | 1 ring   | 12Mpps | 12Mpps  | same  |

    XDP_REDIRECT:

    ** Below is the transmit rate, not the redirection rate
    which might be larger, and is not affected by this patch.

    |          | before  | after   |      |
    | 32 rings | 64Mpps  | 92Mpps  | +43% |
    | 1 ring   | 6.4Mpps | 6.4Mpps | same |

As we can see, feature significantly improves scaling, without
hurting single ring performance.

From Maxim:
4) Some trivial refactoring and code improvements prior to a larger series
to support AF_XDP.
====================

Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>