Is _NTO_CHF_SENDER_LEN currently set by the resource manager package?
That is, can ctp->info.srcmsglen be assumed to be reliable without
having to use the undocumented _dispatch_create()? If not, what are the
conditions that make it “unreliable”? (I haven’t been able to come up
with a test case where it fails.) Also, if it’s still really necessary
to call ChannelCreate(), then _dispatch_create(), and
_NTO_CHF_COID_DISCONNECT is included in the ChannelCreate flags,
shouldn’t _DISPATCH_CHANNEL_COIDDEATH be included in the
_dispatch_create flags?