[ofa-general] [PATCH RFC v4 1/2] RDMA/Core: MEM_MGT_EXTENSIONS support

[Roland Dreier]

 > Perhaps we should just pass in a u32 rkey for both WRs instead of the
 > mr pointer?  Then the code could put the old rkey in the invalidate
 > WR, and the newly updated rkey in the fast_reg WR and chain the two
 > together and do a single post.

Makes sense to me.  The only thing I would worry about would be if some
device needs the actual mr struct pointer to post the work request, but
mlx4 and I guess cxgb3 don't at least and I don't see a good reason why
another device would.  Let's go for it.

 - R.