Home
release-team@conference.openafs.org
Wednesday, March 8, 2017< ^ >
Room Configuration
Room Occupants

GMT+0
[14:25:40] meffie joins the room
[14:56:59] wiesand joins the room
[15:01:04] <wiesand> Hello
[15:01:13] mvita joins the room
[15:01:20] <mvita> present
[15:02:04] <wiesand> I may have found a nit in 12524
[15:03:38] <mvita> I saw your comment, but haven't investigated yet
[15:04:07] <mvita> I think in general LINUX24 is left to bitrot
[15:04:33] <mvita> since it doesn't exist in master
[15:04:39] <wiesand> I'm afraid that's true
[15:05:11] <wiesand> But axing it during the 1.6 series would be bad style
[15:05:38] <mvita> no, I'm not suggesting that
[15:06:31] <wiesand> Well, letting it bitrot conciously  is not much better
[15:06:52] <mvita> you do have a point
[15:07:20] <mvita> Mike or I will check it out
[15:07:29] <wiesand> Thanks.
[15:07:44] <mvita> and good catch, thank you
[15:08:01] <meffie> oh. nice catch.
[15:08:03] <wiesand> I prepared draft release notes in 12562
[15:08:45] <meffie> yes, i can do a 1.6.x only patch to fix the linux24 define in 12524.
[15:09:31] <wiesand> That would be good. How likely is it that more changes are missing for Linux 2.4?
[15:10:08] <wiesand> (from the 12524..12527 stack)
[15:10:37] <wiesand> Except for those, I think we're only waiting for 12558
[15:10:59] jhg joins the room
[15:11:13] <wiesand> Everything else foreseen to go into 1.6.20.2 was merged, unless I'm missing something.
[15:11:18] <jhg> hello
[15:11:18] <wiesand> Am I?
[15:11:26] <wiesand> Hello Joe
[15:11:34] <jhg> Hi Stephan
[15:12:28] <wiesand> So, I guess the question is whether we further delay the point release for the outstanding solaris changes.
[15:12:51] <mvita> Stephan, I see a few things in that stack that might have linux24 implications
[15:13:25] <wiesand> NB we seem to be having buildbot woes
[15:13:38] <wiesand> Mark: Sigh… thanks for checking.
[15:14:25] <mvita> re: solaris, I will have ps3 today for 12558
[15:15:19] <mvita> I would like to see that merged for inclusion in 1.6.20.2
[15:16:27] <wiesand> Fine, I personally don't need that release anytime soon…
[15:16:46] <meffie> > NB we seem to be having buildbot woes
sorry, we had a hardware outage. it should be back online
[15:16:46] kadukoafs@gmail.com/barnowlD058797B leaves the room
[15:17:19] <meffie> back online, as the new parts have been installed yesterday.
[15:17:44] <mvita> 12558 will fix a problem reported in October by mpg.de
[15:18:14] <wiesand> So, one more week I guess…
[15:18:21] kadukoafs@gmail.com/barnowl538BDB78 joins the room
[15:18:32] <wiesand> Which of course trigger the L question… any news on Linux 4.11?
[15:18:41] <mvita> yes
[15:18:50] <jhg> Yes
[15:18:51] <meffie> yes
[15:18:59] <wiesand> Good ones, surely?
[15:19:00] <mvita> I defer to Joe
[15:19:13] <jhg> I am working through some Autoconf breaks right now
[15:19:17] <mvita> not horrible
[15:19:50] <jhg> simple thing. I am mostly learning the build system.
[15:19:58] <wiesand> ok, sounds like it could be worse… thanks for caring
[15:20:24] <jhg> I should have a patch up today or tomorrow.
[15:20:59] <wiesand> If we wait for 12558 anyway, it would be nice to ship that with 1.6.20.2 too.
[15:21:12] <mvita> indeed
[15:21:16] <jhg> bleeding edge linux compatibility =)
[15:21:36] <wiesand> Not having to act upon Linux 4.11 would increase the chance to eventually get a 1.6.21 out.
[15:22:14] <wiesand> And I think there are some changes in the pipeline calling for a 1.6.21 release.
[15:22:40] <mvita> but… the linux commit that broke us also broke a few minor things in linux itself - so there might be a few more changes in subsequent -rcs
[15:22:57] <wiesand> oh well
[15:24:25] <wiesand> I think this concludes the 1.6 topic for today?
[15:24:55] <mvita> the Good News is that the linux builders worked as intended - they both gave us early headsup on Linux changes that affect us.
[15:25:01] <wiesand> There's a new ticket in RT, but…
[15:25:16] <kadukoafs@gmail.com/barnowl538BDB78> Whoops, I missed that.
[15:25:18] <mvita> in a completely automated fashion
[15:25:40] <wiesand> ben: the ticket? don't worry too much
[15:25:46] <kadukoafs@gmail.com/barnowl538BDB78> Oh, you mean 134189?
[15:25:50] <kadukoafs@gmail.com/barnowl538BDB78> I see.
[15:26:19] <wiesand> yes :-)
[15:27:19] <wiesand> deja vu, insufficient data
[15:27:46] <mvita> well, there is a log attached
[15:27:59] <kadukoafs@gmail.com/barnowl538BDB78> Not quite the log I would want, from the file name.
[15:29:07] <meffie> i need to follow up with ted to see what he is doing.
[15:29:17] <kadukoafs@gmail.com/barnowl538BDB78> I don't know that a whole lot happened on the 1.8 front this week.
[15:29:43] <kadukoafs@gmail.com/barnowl538BDB78> The other bug report we got seems to be related to some
system:authuser permissions, and somehow the cache manager is caching
bogus stuff, or something like that.
[15:32:55] <meffie> hmm.
[15:34:10] <wiesand> isn't system:authuser a bogus concept per se?
[15:34:56] <meffie> i've never found it very useful.
[15:35:29] <meffie> normally, it's better to use groups.
[15:40:12] <wiesand> Is there more to discuss today?
[15:40:45] <kadukoafs@gmail.com/barnowl538BDB78> I don't think I have anything.
[15:41:04] <mvita> wait, I have something
[15:41:05] <meffie> i dont have anything else. sorry about the build slaves.
[15:41:28] <wiesand> mike: np
[15:41:35] <wiesand> mark: hmm?
[15:41:41] <mvita> for 12558
[15:42:31] <mvita> someone pointed out an existing ARM commit for a similar issue - no floats allowed on ARM either?
[15:43:11] <mvita> anyway, is there any value in generalizing a config parm for KERNEL_NOFLOAT or something like it?
[15:43:11] <meffie> yes
[15:43:35] <mvita> rather than one-offs for each platform
[15:43:39] <meffie> correct.
[15:44:06] <meffie> that's why we abandoned 12560
[15:44:35] <mvita> found it 11939
[15:46:21] <meffie> yes, it is mentioned in 12560 and 12558
[15:47:12] <mvita> should I work up a generalized way to do this, based on 11939 and 12558?
[15:47:13] <meffie> yes, we need it for aarch64 (arm64)
[15:47:22] <mvita> if so, where should it go?   a new commit?
[15:47:32] <meffie> yes
[15:47:36] <mvita> okay.
[15:47:43] <mvita> tx, that's all I have.
[15:47:46] <meffie> 12558 is solaris specific.
[15:47:54] <mvita> I realize that.
[15:48:11] <meffie> i'd leave it as is.
[15:48:18] <mvita> yes.
[15:48:27] <meffie> thanks Mark
[15:48:49] <kadukoafs@gmail.com/barnowl538BDB78> I think we can probably defer putting in a generic solution until the
second arch that needs the define comes in.
[15:49:08] <mvita> ARM & SOLARIS x86
[15:49:11] <mvita> that's two
[15:49:38] <mvita> that's my point
[15:50:01] <kadukoafs@gmail.com/barnowl538BDB78> The relevant ARM (aarch64) bits are sitting in gerrit waiting for
review; they're not in the tree yet.
[15:50:02] <meffie> i've got an aarch64 builder now, i'm going to be adding it to the buildbot as a nightly builder.
[15:50:45] <meffie> but, yes, more review is needed too.
[15:50:55] <mvita> Let's keep 12558 and 11838 as they are.   If I have time, I'll work up a new commit to generalize them
[15:51:08] <kadukoafs@gmail.com/barnowl538BDB78> Seems reasonable to me.
[15:51:16] <mvita> sorry 11939
[15:51:28] <meffie> sounds good.
[15:51:35] <mvita> ugg, this meeting would go so much smoother for me if I didn't have to type
[15:52:04] <mvita> <obligatory whine about text meetings>
[15:54:00] <wiesand> <obligatory remark about the usefulness of the transcripts>
[15:54:16] <mvita> IMPASSE
[15:54:23] <mvita> ;-)
[15:54:32] <meffie> should i post minutes again? to -devel ?
[15:54:36] <wiesand> i'm not getting that one…
[15:54:55] <wiesand> Mike: if you can, please go ahead.
[15:55:40] <meffie> ok, will do.
[15:56:15] <wiesand> looked up impasse, got it. new vocabulary :)
[15:56:38] <wiesand> Ok, let's adjourn. Thanks a lot everyone!
[15:56:45] <mvita> English, borrowed from French.
[15:56:56] <jhg> Ciao!
[15:57:06] <mvita> Adieu!
[15:57:38] <wiesand> Merci beaucoup, au revoir.
[15:57:52] wiesand leaves the room
[15:58:03] <mvita> a bientot
[17:53:07] meffie leaves the room
[20:50:29] meffie joins the room
[20:53:04] meffie leaves the room
[21:01:59] meffie joins the room
[21:02:20] <meffie> the build slaves running at sna are all back on line. sorry for the outage.
[21:04:35] meffie leaves the room