Eli Cohen wrote:
The outcome of send with invalidate involves an implicit "sync_tpt" like
operation although it syncs the caches to invalidate only the specific
memory key (as opposed to sync tpt command which has a more global
nature).
But I think that the idea is not to save the overhead of sync tpt
commands but to provide security.
Yes, if send-with-invalidate causes a sync-tpt which applies only to the
specific rkey (I assume its documented in the PRM) this can be used to
make the mellanox fmrs --much-- more secure. Are you thinking on
enhancement to support that for consumers that use FMRs through the pool
at the core?
Or.
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
