Re: [dpdk-dev] [PATCH v14 00/23] Add DLB PMD

[David Marchand <david.marchand@redhat.com>]

On Sat, Oct 31, 2020 at 7:16 PM Timothy McDaniel
<timothy.mcdaniel@intel.com> wrote:
>
> The following patch series adds support for a new eventdev PMD. The DLB
> PMD adds support for the Intel Dynamic Load Balancer (DLB) hardware.
> The DLB is a PCIe device that provides load-balanced, prioritized
> scheduling of core-to-core communication. The device consists of
> queues and arbiters that connect producer and consumer cores, and
> implements load-balanced queueing features including:
> - Lock-free multi-producer/multi-consumer operation.
> - Multiple priority levels for varying traffic types.
> - 'Direct' traffic (i.e. multi-producer/single-consumer)
> - Simple unordered load-balanced distribution.
> - Atomic lock-free load balancing across multiple consumers.
> - Queue element reordering feature allowing ordered load-balanced
>   distribution.
>
> The DLB hardware supports both load balanced and directed ports and
> queues. Unlike other eventdev devices already in the repo,  not all
> DLB ports and queues are equally capable. In particular, directed
> ports are limited to a single link, and must be connected to a directed
> queue.
> Additionally, even though LDB ports may link multiple queues, the
> number of queues that may be linked is limited by hardware. Another
> difference is that DLB does not have a straightforward way of carrying
> the flow_id in the queue elements (QE) that the hardware operates on.
>
> While reviewing the code, please be aware that this PMD has full
> control over the DLB hardware. Intel will be extending the DLB PMD
> in the future (not as part of this first series) with a mode that we
> refer to as the bifurcated PMD. The bifurcated PMD communicates with a
> kernel driver to configure the device, ports, and queues, and memory
> maps device MMIO so datapath operations occur purely in user-space.
>
> The framework to support both the PF PMD and bifurcated PMD exists in
> this patchset, and is why the iface.[ch] layer is present.
>
> Major changes in V14
> ====================
> - Fixed format errors in doc/api/doxy-api-index.md
> - Delayed introduction of dlb2_consume_qe_immediate until
>   add-dequeue-and-its-burst-variants.patch
> - Delayed introduction of dlb2_construct_token_pop_qe until
>   add-PMD-s-token-pop-public-interface.patch
> - Delayed introduction of dlb_equeue_*_delayed until
>   add dequeue and its burst variants.patch

I just sent a bunch of comments.
I still see a build error with clang for unused stuff.

There is no point in sending a new series unless the clang build is
fixed once and for all.


I compared dlb and dlb2 code.
I presume fixing bugs in the future will amount to double patches every time.


But, on a positive note, France won against Ireland.


-- 
David Marchand