Re: [PATCH RFC] IB/mad: remove obsolete snoop interface

[Hal Rosenstock <hal@dev.mellanox.co.il>]

On 9/30/2015 12:32 PM, Weiny, Ira wrote:
>>
>> On 9/30/2015 2:01 AM, ira.weiny@intel.com wrote:
>>> From: Ira Weiny <ira.weiny@intel.com>
>>>
>>> This interface has no current users and is obsolete.
>>
>> There are no in tree users of this but there is Sean's madeye tool (which is out
>> of tree). This is still a useful debug tool for MADs.
> 
> Where is that?  

It's been out of tree for some time now. For one place, it can be found
on OpenFabrics site in some old and moldy OFED.

> I did not think it was supported any longer?

It still works AFAIK.

> I have a series of about 5 patches which implement tracing in the MAD stack which I was working through and was going to submit 

Does your MAD stack tracing include the dumping and decode of sent and
received MADs ?

-- Hal

> along with this patch (those are still being tested in my spare time).  I just wanted put this out here because of the patch Nicholas posted yesterday.
>>
>>> Signed-off-by: Ira Weiny <ira.weiny@intel.com>
>>> ---
>>>
>>> This has undergone a medium level of testing.  I have run it against
>>> mlx4, qib, and OPA hardware and it does not seem to cause any regressions.
>>>
>>>  drivers/infiniband/core/mad.c      | 226 +------------------------------------
>>>  drivers/infiniband/core/mad_priv.h |  13 ---
>>>  2 files changed, 5 insertions(+), 234 deletions(-)
>>
>> There are also snoop changes needed in include/rdma/ib_mad.h.
> 
> I'll look into it if there are no other objections to removing the snoop interface.
>
>>
>> Is it correct to assume that OPA_CAP_MASK3_IsSnoopSupported in
>> opa_port_info.h is not related to this ?
> 
> Nope, not related.
> 
> Ira
> 
>