[ewg] Re: summary on OFED 1.4 plans - re adding new kernel features/verbs through ofed

[Or Gerlitz]

Steve Wise wrote:
> Tziporet Koren wrote:
>>
>> * OFED 1.4: *
>> 1. Kernel base: since we target 1.4 release to Sep we target the 
>> kernel base to be 2.6.27
>>     This is a good target, but we may need to stay with 2.6.26 if the 
>> kernel progress will not be aligned.
>> 2. Suggestions for new features:
>>
>>     * Verbs: Reliable Multicast (to be presented at Sonoma)
>>     * IPoIB - continue with performance enhancements
>>
> Sorry I missed these meetings.  For iWARP,  here is my plan:
> New iWARP Verbs:
> - stag_alloc/dealloc
> - nsmr_fastreg
> - read-with-inv-local-stag
> - inv-local-stag
> Note the above verbs might be transport-independent.  I believe the 
> IBTA has defined a fastreg verb too?
> - peer-2-peer support in IWCM/Drivers
Steve, Tziporet,

So you are talking about adding new verbs/features to the Linux RDMA 
stack. Are you intending to do this through the mainline kernel cycles, 
eg the general list, the maintainer (Roland), etc? and if not, why?

Same for for the ipoib performance enhancements, with all the troubles 
we experienced with the ipoib features merged to ofed 1.3 between rc3 
and rc4 (selective TX signaling, 4K MTU), I think such patch must be 
first reviewed in the general list. For example, the merge of  the 
selective TX signaling patch was suspended by Eli since he doesn't see 
any performance improvement on top of the CQ separation etc patch, so if 
it would have been reviewed for mainline in the first place and the 
question "how much does this improves" would have been asked, etc we 
could avoid much of the troubles that were associated with this patch.

Or.



Or.

Or.