Re: [PATCH 00/14] IB/mad: Add support for OPA MAD processing.

[Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 2916 bytes --]

On 06/06/2015 02:38 PM, ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org wrote:
> From: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> 
> The following patch series modifies the kernel MAD processing (ib_mad/ib_umad)
> and related interfaces to send and receive Intel Omni-Path Architecture MADs on
> devices which support them.
> 
> OPA MADs share the same common header with IBTA MADs which allows us to share
> most of the MAD processing code.
> 
> In addition to supporting some IBTA management classes, OPA devices use MADs
> with lengths up to 2K.  These MADs increase the performance of management
> traffic on OPA fabrics.
> 
> Devices report their support of OPA MADs through the new immutable data
> capability flag and immutable max mad size.
> 
> Changes from V1:
> ================
> 
> Remove patch:
> 	IB/mad: Create an RMPP Base header
> 
> Add new patch:
> 	IB/mad cleanup: Clean up function params -- find_mad_agent
> 
> Address comments from Jason about the idea of a flex array for struct ib_mad:
> 	ib_mad does not really allocate struct ib_mads.  Rather it allocates
> 	ib_mad_private objects.  This is where the flex array was more
> 	appropriate.  So this series changes struct ib_mad_private to end in a
> 	flex array to store MAD data.  Casts are used where appropriate to
> 	IB/OPA mad structures or headers.
> 
> Minor updates:
> 	Clean up commit messages
> 	Fix/add const and bool usage
> 	Remove inline qualifiers (let complier handle inline)
> 	Add additional Immutable data checks
> 	Change WARN_ON to BUG_ON in drivers
> 	Add out_mad_pkey_index to process_mad in order to maintain the
> 	"constness" of the struct ib_wc parameter.
> 
> 
> Ira Weiny (14):
>   IB/mad cleanup: Clean up function params -- find_mad_agent
>   IB/mad cleanup: Generalize processing of MAD data
>   IB/mad: Split IB SMI handling from MAD Recv handler
>   IB/mad: Create a generic helper for DR SMP Send processing
>   IB/mad: Create a generic helper for DR SMP Recv processing
>   IB/mad: Create a generic helper for DR forwarding checks
>   IB/mad: Support alternate Base Versions when creating MADs
>   IB/core: Add ability for drivers to report an alternate MAD size.
>   IB/mad: Convert allocations from kmem_cache to kzalloc
>   IB/mad: Add support for additional MAD info to/from drivers
>   IB/core: Add OPA MAD core capability flag
>   IB/mad: Add partial Intel OPA MAD support
>   IB/mad: Add partial Intel OPA MAD support
>   IB/mad: Add final OPA MAD processing

There haven't been any further technical issues with this patchset.
There is the ongoing argument about usage of node_type, but I think I've
made my opinion on that matter clear.  Changes to the core code to make
CA versus SWITCH operation not rely on node_type can be done as follow
ons to this patch set.  As such, I've picked this up for 4.2.



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]