01:06:27 <richlowe> _impl is usually where things are hidden behind an opaque struct
01:06:33 <richlowe> to contain the real struct
05:46:02 <gitomat> [illumos-gate] 16547 savecore should report progress when saving compressed dump -- Andy Fiddaman <illumos⊙fn>
16:42:00 <gitomat> [illumos-gate] 16532 smbd: conflicting types for 'smbd_logon_final' -- Toomas Soome <tsoome⊙mc>
18:42:34 <jbk> hrm.. sometimes it'd be nice when dtrace is displaying an array of all 0s if it'd abbreviate it with say { 0 }.. instead of writing out 256 of them...
19:46:13 <richlowe> yeah mdb has a limit, you could do that with like an -x option
19:46:48 <richlowe> actually, I dont' know if mdb will condense to {0}, but it save itself from displaying unnecessarily massive buffers etc.
22:26:15 <jbk> ok, i'm pretty sure i'm completely confused about the relationship between flow_entry_ts and mac_soft_ring_set_ts
22:27:11 <jbk> flow_entry_t seems to suggest that each one could have multiple rx SRSs and 1 tx SRS
22:27:50 <jbk> and other places suggests that a SRS could have multiple flow_entry_ts associated with it
22:28:27 <jbk> but then looking at how the bw control is managed, it makes me think it's the other way around
22:29:15 <jbk> unless I completely misread it, the various state bits for BW management appear to be copied to each SRS from the flow_entry_t
22:30:24 <jbk> which i don't see how that can work unless the SRS only has 1 flow
23:09:09 <sommerfeld> jbk: have you looked at the "OVERVIEW NOTES FOR DATAPATH" in mac_datapath_setup.c ?
23:12:40 <jbk> yes, but it just leaves me more confused
23:23:08 <jbk> just because i can't reconsile the comments with the other bits of code i've looked at.. which makes me think there's some place i've missed looking at but don't know what that is