[openib-general] tavor quirks etc (opensm compliance etc)

[Eitan Zahavi]

Or Gerlitz wrote:
> Eitan Zahavi wrote:
>   
>> Or Gerlitz wrote:
>>     
>>>> Assuming the value M represents the lowest MTU on the path
>>>> We denote by M-1 the MTU value one level below M (e.g. 1K if M=2K)
>>>> R represents the MTU value in the request. Similarly R-1 is one below 
>>>> R and R+1 is one above R.
>>>>
>>>> Query-MTU | Query-Sel | Resp by Spec     | OpenSM Should  | OpenSM 
>>>> Quirk w. Tavor End Port
>>>> ----------------------------------------------------------------------------------------- 
>>>>
>>>> UNDEFINED | UNDEFINED | <= M             | M              | min(M,1K)
>>>> R         | <         | <= min(R-1, M)   | min(R-1, M)    | min(R-1, 
>>>> M, 1K)
>>>> R         | =         | R if M>=R /ERR   | R if M>=R /ERR | R if M>=R /ERR
>>>> R         | >         | R < <= M         | R+1 if M>R /ERR| R+1 if M>R /ERR
>>>>         
>
>   
>>> 1st maybe its clear to everyone expect me, but what do you mean by 
>>> /ERR in the table above, is it what opensm would return before the 
>>> patch you suggested?
>>>       
>
>   
>> By ERR I mean that the path being evaluated is rejected from being 
>> included in the paths group of the response to the provided query.
>>     
>
> so when you say
>
> "X if some relation holds on (Y,Z) /ERR"
>
> you mean that it "should return X but if r(Y,Z) holds return no record" 
> and this how the code is written with the patch?
>
>   
No:
R if M>=R /ERR mean:
Return R if M is bigger or equal to R or else this path does not match 
the request.

R+1 if M>R /ERR meas:
Return R+1 if M is bigger then R or else this path does not match the request.

If no paths math the request you the response depends on the query method:
For Get(PathRecord) you will get an error.
For GetTable(PathRecord) you will get zero number of returned records
For GetMulti(MultiPathRecord) you should get zero number of returned records

EZ
>>> 2nd can you post the open sm tavor quirk patch?
>>>   
>>>       
>> What do you mean? The old patch introducing the "opensm quirk" mode?
>> It is GIT versions: 86077144ed956ddb32a0f8d067d5bb00fd564ac6 followed by 
>> 03e3b3a6fa934202c0f4270a2c69d64ac486b1ca
>> or SVN: 9497 followed by 9518
>>     
>
> OK, thanks, i guess you mean to the svn trunk or its the ofed 1.1 
> branch? can be cool if you send a pointer to the SVN...
>   
This is trunk
> Or.
>
>
> _______________________________________________
> openib-general mailing list
> openib-general at openib.org
> http://openib.org/mailman/listinfo/openib-general
>
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
>