17:17:42 <gitomat> [illumos-gate] 14944 xhci silently drops ugen transfers sometimes but not always -- Joshua M. Clulow <josh⊙so>
20:47:15 <bahamat> Is there something for gerrit that will automatically update reviewed by? Or do I need to manually edit the commit message and push it?
20:47:28 <richlowe> you need to manually do it
20:47:46 <richlowe> I think a handful of us have scripts to give you the list from the gerrit API
20:47:59 <richlowe> mine is bad, but I think andyf's isn't
20:48:18 <bahamat> OK. And I'm assuming this won't invalidate the reviews if I'm only modifying the commit message?
20:48:23 <andyf> I can share what I've got if you like
20:48:32 <jclulow> bahamat: Yes as long as you don't also rebase at the same time
20:48:38 <richlowe> I just usually don't update the message in gerrit, I just do it for the RTI
20:48:50 <rmustacc> It's fine to make it the advocate's problem.
20:49:04 <jclulow> core team member
20:49:07 <jclulow> but yes
20:49:09 <jclulow> :D
20:49:11 <richlowe> oh yeah, that too, though them being in the email helps.
20:50:54 <bahamat> Ok...and I pushed the wrong thing.
20:51:14 <richlowe> to where?
20:51:18 <jclulow> not master!
20:51:24 <rmustacc> It doesn't appear so.
20:51:33 <rmustacc> At least not on 14967.
20:51:35 <bahamat> I did HEAD:refs/for/master, and only afterward remembered that I'm supposed to push to this change ID or something
20:51:42 <richlowe> that's fine
20:51:51 <richlowe> gerrit is clever and updates because the change-id
20:51:52 <jclulow> that's totally fine -- it looks at the Change-Id: in there to figure out what to do with it
20:51:53 <jclulow> yeah
20:51:53 <richlowe> that's what the change-id is for
20:51:58 <rmustacc> Exactly.
20:52:07 <bahamat> Oh, no the output from git push says it's the same cr number
20:52:19 <richlowe> it'll say like "updated: 1"
20:52:34 <jclulow> https://illumos.org/docs/contributing/gerrit/#3-prepare-your-commit-and-push-to-gerrit
20:52:45 <jclulow> (and also "4. Update your change with new contents")
20:52:53 <bahamat> OK, that looks like I did it right.
20:54:25 <richlowe> if you can't push to master, it's almost foolproof
20:54:31 <richlowe> or rather to branches other than refs/for/...
20:55:14 <bahamat> Ok, RTI is away. I think I got it all.
21:00:28 <danmcd> And richlowe --> I'll push it for bahamat upon approval.
21:01:22 <richlowe> I'm trying to land my python thing you just approved, for marcel
21:01:35 <richlowe> so I haven't looked at anything beyond hitting the button in topicbox
21:01:47 <danmcd> SOrry, I shouldn't have picked on you per se.  That should've been more generic.
21:02:10 <danmcd> Take your time on the python one.  (I'm just glad today's merge trash-fire was all my fault.)
21:17:19 <andyf> To be fair, I made exactly the same mistakes with the omnios merge... 
21:17:40 <andyf> and now s10 zones are broken in omnios bloody, which is also probably my fault
21:17:41 <gitomat> [illumos-gate] 14847 illumos should default to python 3.9 -- Richard Lowe <richlowe⊙rn>
21:39:56 <andyf> danmcd - s10 brand has been broken for us since OS-4151 (fenix) - panic in DEBUG, memory leak otherwise
21:39:57 <fenix> OS-4151: setbrand hooks should be sane during fork (Resolved)
21:39:57 <fenix> ↳ https://smartos.org/bugview/OS-4151
21:40:09 <andyf> so it isn't new, and not AVX512 related, thankfully :)