Michael S. Tsirkin wrote: > Quoting r. Or Gerlitz <[EMAIL PROTECTED]>:
>> This porting relies on Sean's work to allow for joining/leaving "IP >> Multicast" mcast groups from user space. And anyway, when you expose >> something to user space, you might not "see" the usage as it can be not >> an open source one. > I don't know whether a single closd-source application shows something > is useful. Anyway, you don't need code upstream just to do development. > Pushing userspace-interfacing code upstream before there are actual users > leads to ABI instability and other such issues. The only current rdma cm user space code i am aware to is uDAPL. Intel MPI is a closed source package using uDAPL which drove much (almost 100%) of the uDAPL and uCMA testing and the uCMA is candidate to be submitted for 2.6.20 (ie ~3 months from today). I have seen some feedback provided on the uCMA UD Mcast code by Steve/Tom and as i said we are going to use it immediately so more feedback would be provided over this list. As of the NO synchronization (to remove doubt, i am not in favor for such sycn and in favor of not maintaining the ib kernel code in the svn!) between the openib SVN and Roland's GIT tree, to make the testing of uCMA consumers robust, applying this patch set on the for-2.6.20 branch of the IB git tree is essential. Or. _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
