- Friday, March 20, 2020

release-team@conference.openafs.org

Friday, March 20, 2020< ^ >

meffie has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[12:14:15] meffie joins the room
[12:25:26] mbarbosa joins the room
[15:33:42] cwills joins the room
[15:52:13] kaduk@jabber.openafs.org/barnowl joins the room
[15:59:45] <kaduk@jabber.openafs.org/barnowl> Greetings all
[16:00:55] <cwills> Good morning / afternoon
[16:01:16] <meffie> good day
[16:03:21] yadayada joins the room
[16:03:28] <yadayada> Hi All
[16:03:53] <kaduk@jabber.openafs.org/barnowl> Hi Yadav!
[16:04:21] <yadayada> Hello Ben
[16:05:08] wiesand joins the room
[16:05:09] wiesand leaves the room
[16:05:11] wiesand joins the room
[16:05:45] <wiesand> Hello. Sorry, connecting took some time.
[16:06:14] <kaduk@jabber.openafs.org/barnowl> It is looking like 14082 and 14099 will make it into pre2?
[16:08:02] <wiesand> er, no, I don't have those on my "must make 1.8.6" list yet- should I?
[16:08:23] <kaduk@jabber.openafs.org/barnowl> Well, hmm.  I was just guessing.
[16:09:14] <wiesand> They haven't been discussed here unless I missed it. I'm open of course.
[16:09:15] <kaduk@jabber.openafs.org/barnowl> It would only be critical if we had the one but not the other; we
could probably wait to take both if needed.
[16:09:46] <kaduk@jabber.openafs.org/barnowl> I have no doubt that the AppArmor issue they try to fix is real,
though.
[16:11:03] <wiesand> Merged the Linux-5.6 ones yesterday and hoped for a positive robotest. No successful buildbot result though. But it looks like an issue due to external problems?
[16:12:18] <cwills> That should have been resolved.. upstream kernel-ppa failed yesterday, but is clean today.  Mike kicked the buildbot for a redo.  It finished cleanly
[16:12:24] <meffie> yes, it was external, i retried and they built.
[16:13:37] <wiesand> Frankly, I'd rather get 1.8.6(pre2) out with as few additional changes as possible (besides those required for Linux-5.6), and postpone anything else to 1.8.7.
[16:13:51] <kaduk@jabber.openafs.org/barnowl> Okay, that seems reasonable.
[16:13:57] <meffie> we use the ubuntu kernel team's ppa for the nightly linux images. they are pretty good at keeping it working, but sadly it was out for just a bit.
[16:14:41] <cwills> I run a weekly build against the latest rc from the Linux repo ("vanilla linux")
[16:14:44] <wiesand> Noticed the successful build for master, but obviously missed the on for 1_8_6
[16:15:07] <meffie> it just ran
[16:15:38] <wiesand> ok, thanks a lot for caring!
[16:16:18] <wiesand> Ben: really sorry for breaking my promise to get stuff done last weekend!
[16:17:12] <wiesand> Turns out corona has a pretty paralyzing effect on me.
[16:18:18] <kaduk@jabber.openafs.org/barnowl> Just the entire situation, or are you perseonally affected??
[16:18:32] <kaduk@jabber.openafs.org/barnowl> I hope you are okay!
[16:19:47] <wiesand> In addition, the "everyone please work from home now" order took it's toll on us -  for days we were overwhelmed with requests of folks who are now expected to do "home office" but have no clue how to set up and use ssh tunnels or socks5 proxies.
[16:21:05] <meffie> yes, i've heard the same for many, many other people
[16:21:23] <kaduk@jabber.openafs.org/barnowl> I can imagine, yes.
[16:21:49] <meffie> most organizations were not setup for distributed work, and the vpns are failing.
[16:21:53] <kaduk@jabber.openafs.org/barnowl> Lots of my friends have been noting that the wirecutter-recommended
webcam is out of stock ~everywhere.  (Dell has/had some, apparently.)
[16:23:09] <wiesand> I'm not sure whether I'm personally affected. I'm showing symptoms compatible with Corona. But you won't get tested unless you stayed in a "high risk area" or had a proven contact to someone already tested positively.
[16:23:50] <kaduk@jabber.openafs.org/barnowl> Seems wise; what you do isn't really going to change one way or the
other
[16:24:12] <kaduk@jabber.openafs.org/barnowl> I have been a little sick this week, but the symptoms don't really
match up with the COVID-19
[16:24:32] <wiesand> But then those symptoms are compatible with a flu or all the other stuff you typically contract this time of year.
[16:24:42] <kaduk@jabber.openafs.org/barnowl> Exactly!
[16:26:40] <wiesand> Except that Corona is much more likely to kill me than the other stuff. (I qualify for "high risk" in several ways).
[16:29:00] <wiesand> I'm not the one to panic. But I'll admit I'm scared.
[16:29:38] <meffie> it is a scary situation.
[16:29:55] <kaduk@jabber.openafs.org/barnowl> It is.
[16:30:41] <wiesand> Back on topic: how crucial is the "Linux creds" stuff?
[16:31:20] <meffie> i believe it came from jhutz?
[16:31:21] <kaduk@jabber.openafs.org/barnowl> If you're running AppArmor in an affected configuration , users will
seemingly randomly get EPERM for normal operations that should succeed
[16:31:38] <kaduk@jabber.openafs.org/barnowl> CMU is running with local patches for it, IIUC
[16:31:39] <cwills> Well.. I ran robotests on a couple different platforms to test the 14099 patch.
[16:31:48] <kaduk@jabber.openafs.org/barnowl> MIT has seen the behavior in question as well
[16:32:15] <wiesand> More bluntly: should I merge those before we tag 1.8.6pre2?
[16:32:33] <kaduk@jabber.openafs.org/barnowl> I think your proposal to wait until 1.8.7 for them is fine.
[16:32:48] <wiesand> Or would it be ok to postpone this to 1.8.7?
[16:33:34] <meffie> maybe a good idea. we did just fix a bug  on master for it.
[16:33:52] <meffie> maybe more time on master is a good idea?
[16:34:06] <kaduk@jabber.openafs.org/barnowl> Sure
[16:34:43] <wiesand> And BTW, it may be more of a formality, but *some* review on 14096 and 14097 would be nice indeed…
[16:36:20] <yadayada> 14082 will it have impact for SELinux enforcing mode also ? from commit msg it looks like
[16:36:34] <kaduk@jabber.openafs.org/barnowl> I expect so, yes (re SELinux)
[16:36:41] <yadayada> ok
[16:37:56] <cwills> I've ran robotests in centos7 / 8 with selinux when I was doing my tests
[16:41:21] <cwills> On a different topic .. a heads up.  gcc-10 changes from the default setting -fcommon to -fno-common.  -- I just went through and "patched" the failures.. but haven't tested anything (I just added the -fno-common to the cflags to see what breaks).
[16:42:47] <kaduk@jabber.openafs.org/barnowl> The manual implies that issues are likely to arise at link-time, not
runtime.
[16:42:58] <cwills> yes
[16:43:11] <meffie> thanks for staying on top of the gcc changes
[16:43:19] <wiesand> So: the remaining changes set for 1.8.6pre2 (14096/7) were reviewed this evening, and it looks like we'll postpone anything else regarding 1.8.6, right?
[16:43:20] <cwills> just want to make sure it doesn't break anything.. most of the changes are adding extern
[16:43:45] <meffie> yes
[16:45:51] <wiesand> Ok, I'll get those merged, build tarballs, ask Ben for the tag etc. a.s.a.p. If you have objections, raise them now please.
[16:46:13] <kaduk@jabber.openafs.org/barnowl> *raises anti-objection*
[16:46:18] <meffie> lol
[16:46:26] <wiesand> grin
[16:46:35] <wiesand> thanks!
[16:47:24] <wiesand> So much for the stable branch then.
[16:47:38] <wiesand> On to master?
[16:48:26] <kaduk@jabber.openafs.org/barnowl> I don't know how much there is to say, for master.
I did some minor housekeeping in terms of merging things that had +2
and buildbot verification, and reviewed a couple smaller changes.  But
I didn't put much time in.
[16:48:47] <kaduk@jabber.openafs.org/barnowl> Just this morning I did pick up the 1.9.0 NEWS again, so hopefully
there will be something there today.
[16:49:10] <cwills> just stumbled across a real minor item associated with the opr_cache.. I'll push a fix (cache_t needs to be added to tests/.gitignore)
[16:50:22] <kaduk@jabber.openafs.org/barnowl> That should be a quick fix; I like those :)
[16:50:37] <meffie> we need a "core" entry there too.
[16:50:49] <cwills> okay
[16:50:51] <kaduk@jabber.openafs.org/barnowl> Heh
[16:51:13] <kaduk@jabber.openafs.org/barnowl> "maybe we could just fix all the bugs instead"
<ducks>
[16:51:27] <meffie> if i remember correctly. one of the softsig test leaves a core behind
[16:51:36] <meffie> it crashes on purpose! :)
[16:52:25] <kaduk@jabber.openafs.org/barnowl> I left discussion-y questions on 14025 and 14026 (don't need to
discuss now per se, but just in case the relevant people didn't see it
already)
[16:52:36] <meffie> ack
[16:53:15] <kaduk@jabber.openafs.org/barnowl> (I guess on 12578 as well, but that's a FreeBSD-specific one so I may
be on my own.)
[16:53:40] <meffie> i just glanced at 14025, need to look closer.
[16:56:21] <yadayada> 14026 is interesting as it can avoid log flooding. will try it
[16:56:38] <meffie> thank you yadayada
[16:59:35] <kaduk@jabber.openafs.org/barnowl> We're back at the top of the hour; anything else for today?
[17:00:16] <yadayada> nothing from my side
[17:00:23] <meffie> i pushed the changes to the wiki i mentioned last week to the rpm howto: https://wiki.openafs.org/devel/HowToBuildOpenAfsRpmPackages/
[17:00:34] <kaduk@jabber.openafs.org/barnowl> Thanks Mike!
[17:00:35] <meffie> tested on centos 6, 7, 8, and fedora 31
[17:01:15] <wiesand> Let's adjourn. Thanks a lot everyone!
[17:01:23] <cwills> Thanks.
[17:01:25] <meffie> thanks! stay safe!
[17:01:35] <cwills> take care everyone.
[17:01:39] <kaduk@jabber.openafs.org/barnowl> Indeed, stay safe everyone!
[17:01:51] <yadayada> thanks
[17:01:58] cwills leaves the room
[17:02:03] wiesand leaves the room
[17:03:47] meffie leaves the room
[18:56:45] cwills joins the room
[18:56:46] kaduk@jabber.openafs.org/barnowl leaves the room
[19:00:00] cwills leaves the room
[19:05:19] mbarbosa leaves the room
[19:05:32] mbarbosa joins the room
[21:26:01] yadayada leaves the room
[21:43:52] mbarbosa leaves the room