>Since there's no CM involved in creating UD QPs, and since there's no multicast
>rdma, rdma_cm seems like a weird place for multicast.
The CM is involved in using UD QPs through the use of SIDR, and it also requires
path record lookups, which is provided by the rdma_cm.
>Would it make more sense
Hi Amith,
On Wed, 2006-04-19 at 19:39, amith rajith mamidala wrote:
> Hi Sean,
>
> I have a few basic questions:
>
> 1. Does the API which waits for join to complete
> ensure that the multicast forwarding tables in the switches have been
> updated.
This is not an API issue. The IB spec (archite
Quoting r. Sean Hefty <[EMAIL PROTECTED]>:
> Subject: RFC userspace / MPI multicast support
>
> I'd like to get some feedback regarding the following approach to supporting
> multicast groups in userspace, and in particular for MPI. Based on side
> conversations, I need to know if this approach w
amith rajith mamidala wrote:
1. Does the API which waits for join to complete
ensure that the multicast forwarding tables in the switches have been
updated. This is one of the main problems that we had studied:
The join is asynchronous. Completion of the join would not be reported until
the s
Hi Sean,
I have a few basic questions:
1. Does the API which waits for join to complete
ensure that the multicast forwarding tables in the switches have been
updated. This is one of the main problems that we had studied:
(Please refer to the following EURO PVM/MPI paper for details)
http://www.cs