- Wednesday, November 30, 2016

release-team@conference.openafs.org

Wednesday, November 30, 2016< ^ >

Room Configuration

Room Occupants

GMT+0
[03:57:53] Jeffrey Altman leaves the room
[03:58:06] Jeffrey Altman joins the room
[04:56:23] Jeffrey Altman leaves the room
[04:56:24] Jeffrey Altman joins the room
[06:09:00] Jeffrey Altman leaves the room
[06:18:02] Jeffrey Altman joins the room
[14:40:20] meffie joins the room
[14:49:04] <meffie> oh dear. get well soon.
[15:00:34] wiesand joins the room
[15:00:52] <wiesand> Hello
[15:00:57] <kadukoafs@gmail.com/barnowlE45450E9> I'm getting better, but am still not 100%
[15:01:35] <wiesand> Glad to hear the part before the comma
[15:01:42] <meffie> good to hear you are feeling better.
[15:02:37] <wiesand> I've been in non-afs mode for other reasons :-(
[15:03:44] <wiesand> I think our plan is still to issue 1.6.20 a.s.a.p. followed swiftly by 1.6.20.1 ?
[15:04:29] mvita joins the room
[15:04:35] <mvita> hello
[15:04:43] <kadukoafs@gmail.com/barnowlE45450E9> Yes ... though I did not get to test proto-1.6.20 las week :(
[15:05:28] <wiesand> It's still looking good on my test systems (typos aside)
[15:06:10] <wiesand> Alternatively, we can postpone 1.6.20 once more and go for 1.6.19.1 next week
[15:06:44] <kadukoafs@gmail.com/barnowlE45450E9> Well, I'm willing to accept your testing :)
[15:06:44] <mvita> hmm
[15:07:01] <kadukoafs@gmail.com/barnowlE45450E9> So, shall we start that process right now (TM)?
[15:07:12] <wiesand> But then Linux 4.9 may have an RC8…
[15:07:17] <wiesand> Ben: don't ;-)
[15:07:27] <mvita> rc8 still looks likely
[15:07:31] <mvita> but not guaranteed
[15:07:45] <wiesand> Ben, I'd rather trust Mark's testing
[15:08:11] <kadukoafs@gmail.com/barnowlE45450E9> *nods*
[15:08:15] <mvita> I just built openafs master on Linux v4.9-rc7 without error
[15:08:44] <mvita> but as usual, this is just a "header" build - I did not test on Linux 4.9 system
[15:09:25] <mvita> as far at 1.56.20 goes, we've been running it in production for a while here.
[15:09:32] <mvita> on our public cell
[15:09:42] <mvita> sorry 1.6.20 equiv
[15:10:07] <wiesand> So Mark, thumbs up for starting the process (TM)?
[15:10:26] <mvita> thumbs up from me, fwiw
[15:10:51] <wiesand> Objections anyone?
[15:11:00] <meffie> none
[15:11:49] <wiesand> So Ben, if you feel up to it, please go ahead whichever way you prefer to actually do it
[15:12:01] <wiesand> I'll prepare a volume now…
[15:12:03] <kadukoafs@gmail.com/barnowlE45450E9> We'll try it your way, I think.
[15:12:14] <wiesand> Great, let's try
[15:12:27] <kadukoafs@gmail.com/barnowlE45450E9> 12456-12460
[15:19:20] <wiesand> Mark, Mike: do those look ok?
[15:19:29] <kadukoafs@gmail.com/barnowlE45450E9> stephan, do you want me to do the "make 1.6.20" commit?
[15:19:32] <mvita> correcting a commit msg now
[15:21:15] <wiesand> Ben, if you're prepared to do it, fine
[15:21:35] <wiesand> The volume is in place now
[15:22:08] <mvita> 12458 needs a commit msg tweak
[15:24:40] <mvita> everything else is good
[15:24:51] <kadukoafs@gmail.com/barnowlE45450E9> > 12458 needs a commit msg tweak
done
[15:26:21] <mvita> green light
[15:26:38] <kadukoafs@gmail.com/barnowlE45450E9> [waiting for buildbot]
[15:32:18] <mvita> will 12460 be merged first?  since it has comments summarizing the other commits?
[15:32:39] <mvita> or last?
[15:32:53] <mvita> so it's "on top"?
[15:34:44] <kadukoafs@gmail.com/barnowlE45450E9> last, I think
[15:34:54] <mvita> okay
[15:34:55] <meffie> ha "memset is not as expensive as it was in the 1980s."
[15:35:14] <mvita> I have no preference, so either way is good
[15:38:04] <wiesand> I'm preparing a web change. The boilerplate can probably stay the same as for 1.6.17?
[15:40:13] <wiesand> Buildbot will check the full stack first, right?
[15:41:29] <kadukoafs@gmail.com/barnowlE45450E9> buildbot will chck the whole stack, yes
[15:41:47] <kadukoafs@gmail.com/barnowlE45450E9> > boilerplate can probably stay the same
probably
[15:45:08] <mvita> do we need to write an OpenAFS Security Advisory for this?
[15:52:02] <kadukoafs@gmail.com/barnowlE45450E9> Arguably
[15:54:32] <mvita> I can do it - most of the content is already in the commit msgs
[15:54:45] <kadukoafs@gmail.com/barnowlE45450E9> thanks
[15:58:12] <mvita> how do you want it?  as commits against openafs web?
[15:59:17] <kadukoafs@gmail.com/barnowlE45450E9> that works, sure
[15:59:18] <mvita> oh, yes, I see that's how it was done for previous ones
[16:03:10] <mvita> we did not obtain CVE numbers for any of these, correct?
[16:05:52] <wiesand> Do we wait for all build slaves to complete before proceeding with the 1.6 commits?
[16:25:03] <wiesand> I guess we are. I think in this case I'll drive home and rejoin the party from there (until I fall asleep).
[16:25:22] <wiesand> Later
[16:25:28] wiesand leaves the room
[16:46:54] <kadukoafs@gmail.com/barnowlE45450E9> Okay, my other meeting ended (well, I dropped off the VPN and they
were basically done), so starting to merge things and fixup commit
messages for "cherry-picks".
[16:51:24] <kadukoafs@gmail.com/barnowlE45450E9> 1246[123] for stable (so far)
[17:04:24] stephan.wiesand joins the room
[17:05:10] <stephan.wiesand> Back
[17:06:51] <kadukoafs@gmail.com/barnowlE45450E9> That solaris11_x86-builder is really holding us back ;)
[17:07:21] <stephan.wiesand> Surprise ;-)
[17:07:40] <kadukoafs@gmail.com/barnowlE45450E9> This means I should yell at meffie, right? ;)
[17:08:03] <stephan.wiesand> When I proposed this procedure, I assumed a security release would comprise a change or two...
[17:08:41] <stephan.wiesand> I still like it better, even if we end up having to finish this tomorrow!
[17:08:57] <meffie> gah.
[17:09:18] <stephan.wiesand> Well the solaris10 builder isn't that much faster ;-)
[17:09:21] <meffie> ok, i have some faster hardware available now. i'll setup a new builder on it.
[17:09:46] <stephan.wiesand> Anyway, good news.
[17:10:03] <meffie> the new machines have fast local disks.
[17:10:26] <stephan.wiesand> What do the old ones have?
[17:12:30] <meffie> SAN
[17:12:52] <meffie> (a suboptimal SAN)
[17:14:14] <kadukoafs@gmail.com/barnowlE45450E9> We'll have to keep the number of commits in mind for next time.
[17:15:42] <stephan.wiesand> Or rule not to wait for the slowest build slaves, which is still much better than what we did in the past.
[17:16:13] <stephan.wiesand> Fedora, Ubuntu, FBSD were finished a while ago.
[17:18:42] <kadukoafs@gmail.com/barnowlE45450E9> Yeah, we have a verified+1 from mvitale we could use
[17:19:41] <stephan.wiesand> But I'm fine with waiting. Though that *will* mean I'll drop out before we're done.
[17:20:46] <kadukoafs@gmail.com/barnowlE45450E9> I'm fine with not waiting for the last two on master.
[17:24:42] <kadukoafs@gmail.com/barnowlE45450E9> 12464-12467 also on stable, including NEWS and version numbers
[17:28:16] <stephan.wiesand> Thanks. Mark, Mike, any chance you take a look at the 1.6 changes?
[17:29:43] <meffie> ok
[17:30:19] <mvita> I will look after I finish my advisory draft
[17:30:36] <stephan.wiesand> Thanks. I'm working on cherry-picking these into a local tree and doing/running EL5/6/7 test builds
[17:31:08] <mvita> however, right now I need food
[17:31:32] <stephan.wiesand> Enjoy!
[17:38:09] <kadukoafs@gmail.com/barnowlE45450E9> buffer.c: In function 'DNew':
buffer.c:452:5: error: implicit declaration of function 'memset'
[-Werror=implicit-function-declaration]
buffer.c:452:5: error: incompatible implicit declaration of built-in
function 'memset' [-Werror]
cc1: all warnings being treated as errors
fascinating
[17:38:59] <stephan.wiesand> memset?!
[17:39:05] <stephan.wiesand> which one?
[17:39:13] <kadukoafs@gmail.com/barnowlE45450E9> https://buildbot.openafs.org/builders/opensuse12-x86_64-builder/builds/1878/steps/compile/logs/stdio
[17:40:31] <kadukoafs@gmail.com/barnowlE45450E9> (I have a tab open to buildbot.openafs.org/grid)
[17:40:51] <stephan.wiesand> Do you know this "I hate it when I'm right" experience?
[17:41:34] <kadukoafs@gmail.com/barnowlE45450E9> All too well...
[17:41:52] <kadukoafs@gmail.com/barnowlE45450E9> (looks like master has roken.h which presumably will make sure there's
a memset prototype)
[17:43:22] <stephan.wiesand> I swear I built and am running 1.6.19 + the stable diff for 133060 on EL5/6/7...
[17:46:17] <kadukoafs@gmail.com/barnowlE45450E9> Lots of compilers will just give you a memset, and the opensuse
machine may be running with particularly pedantic compiler flags.
[18:02:14] <meffie> buffer.c:452:5: error: implicit declaration of function 'memset' [-Werror=implicit-function-declaration]
[18:02:56] <meffie> that took some searching to find ;)
[18:20:12] <mvita> back
[18:20:20] <mvita> oh  the dumb memset thing
[18:21:14] <mvita> I've hit that one already as well when backporting from master
[18:30:06] <kadukoafs@gmail.com/barnowlE45450E9> Whoops, missed a message.
The procedure for getting CVE numbers has changed since the last time
we did this, and the obvious way to get one requires disclosing the
nature of the vulnerability.
(The non-obvious way is to become a CNA, which is supposedly supported
for us.)
[19:15:15] <stephan.wiesand> buildbot verdicts are trickling in for the 1.6 changes, but no CR+1s, and it's getting late here...
[19:16:36] <stephan.wiesand> NB EL5/6/7 smoke tests look good here (including ps2 for 12463)
[19:19:44] <stephan.wiesand> Thanks Mike
[19:47:09] <stephan.wiesand> And MArk
[19:47:21] <stephan.wiesand> We're getting close...
[20:08:09] <stephan.wiesand> Looks like we're ready?
[20:10:10] <kadukoafs@gmail.com/barnowlE45450E9> Indeed it does.
[20:10:32] <mvita> just wrapping up the advisory
[20:10:42] <kadukoafs@gmail.com/barnowlE45450E9> Thanks for staying up for it (how late is it, for you?).
[20:10:59] <mvita> gotta configure to push to gerrit
[20:11:26] <stephan.wiesand> rolling tarballs... Ben could you tag 2161e741da94b3ceb60cbd57a78643dd4169dfb7 as openafs-stable-1_6_20 ?
[20:11:59] <stephan.wiesand> (it's 9:11pm here)
[20:15:33] <mvita> hmm, I goofed something…
[20:15:46] <stephan.wiesand> relax
[20:15:48] <kadukoafs@gmail.com/barnowlE45450E9> tag pushed
[20:15:54] <stephan.wiesand> thanks
[20:15:56] <mvita> host11:~/src/openafs-web mvitale$ git push gerrit HEAD:refs/for/master
Counting objects: 8374, done.
Delta compression using up to 8 threads.
Compressing objects: 100% (1849/1849), done.
Writing objects: 100% (8374/8374), 3.40 MiB | 288.00 KiB/s, done.
Total 8374 (delta 6466), reused 8365 (delta 6462)
remote: Resolving deltas: 100% (6466/6466)
remote: Processing changes: refs: 1, done
To ssh://gerrit.openafs.org:29418/openafs
! [remote rejected] HEAD -> refs/for/master (no common ancestry)
error: failed to push some refs to 'ssh://gerrit.openafs.org:29418/openafs'
host11:~/src/openafs-web mvitale$ git l10
d805c32 (HEAD -> refs/heads/v1.6.20) Release 1.6.20 security advisory
f129dac (refs/remotes/origin/master, refs/remotes/origin/HEAD, refs/heads/master) Release 1.6.19
8d6be56 Rototill roadmap.html
40b6ca9 Release 1.6.18.3
[20:16:23] <kadukoafs@gmail.com/barnowlE45450E9> git remote -v |grep gerrit
[20:16:44] <mvita> gerrit    ssh://gerrit.openafs.org:29418/openafs (fetch)
gerrit    ssh://gerrit.openafs.org:29418/openafs (push)
[20:16:45] <stephan.wiesand> it's refs/publish/master now
[20:16:53] <mvita> oh.
[20:16:55] <kadukoafs@gmail.com/barnowlE45450E9> You need the gerrit url to be openafs-web
[20:17:04] <mvita> oh oh
[20:17:10] <kadukoafs@gmail.com/barnowlE45450E9> refs/for/master has been working fine for me still...
[20:18:17] <mvita> host11:~/src/openafs-web/.git mvitale$ git remote -v | grep gerrit
gerrit    ssh://gerrit.openafs.org:29418/openafs-web (fetch)
gerrit    ssh://gerrit.openafs.org:29418/openafs-web (push)
[20:18:21] <mvita> like dat?
[20:18:37] <kadukoafs@gmail.com/barnowlE45450E9> looks like
[20:19:32] <mvita> 12468
[20:19:45] <mvita> html untested
[20:22:28] <mvita> I just tested it, it looks passable to me
[20:24:54] <stephan.wiesand> uploading
[20:25:48] <mvita> btw  refs/for/master worked for me too
[20:29:23] <stephan.wiesand> okok
[20:30:26] <kadukoafs@gmail.com/barnowlE45450E9> pushed new advisory rev
[20:31:27] <mvita> tx
[20:31:46] <mvita> any recommendations/procedures for making the patch files?
[20:32:37] <stephan.wiesand> git show ;-)
[20:33:01] <mvita> edgy!
[20:33:45] <kadukoafs@gmail.com/barnowlE45450E9> Nah, git diff is better here
[20:33:52] <kadukoafs@gmail.com/barnowlE45450E9> (rev 3 includes the patchfiles, FWIW)
[20:33:56] <meffie> git format-patch
[20:33:59] <meffie> ;)
[20:34:31] <mvita> oh, you already did them, okay
[20:34:32] <kadukoafs@gmail.com/barnowlE45450E9> Though stephan merged the NEWS and version number changes in the
middle, so that makes more work ... except for me, since I had the
changes staged locally in the right order.
[20:34:50] <mvita> I would have gone w/ format-patch, since that's what my sensei advises
[20:35:26] <kadukoafs@gmail.com/barnowlE45450E9> format-patch will make a chain of diffs and ~requires git to reapply;
git diff's output should be usable by any patch(1).
[20:35:54] <mvita> makes sense
[20:36:13] <mvita> except to sensei?
[20:36:56] <mvita> thank you Ben
[20:38:44] <kadukoafs@gmail.com/barnowlE45450E9> thank you for doing the hard part (writing the text) :)
[20:39:10] <mvita> I did it in haste, so proofread carefully
[20:39:31] <mvita> I still have to write something similar for SNA
[20:40:08] <stephan.wiesand> sorry for merging things out of order - i was kind of eager to get things done
[20:40:47] <stephan.wiesand> tarballs uploaded, volumes released (I hope)
[20:43:07] <kadukoafs@gmail.com/barnowlE45450E9> don't worry about merging out of order; it's easy enough to cope
[20:47:32] <stephan.wiesand> proposed web change is 12469
[20:50:02] <kadukoafs@gmail.com/barnowlE45450E9> merged both web changes; going off to find the command to trigger a
rebuild...
[20:52:36] <stephan.wiesand> remctl www-01.central.org <http://www-01.central.org> export_htdocs openafs
[20:54:46] <stephan.wiesand> I just executed that
[20:55:25] <kadukoafs@gmail.com/barnowlE45450E9> So did I ;)
[20:55:50] <stephan.wiesand> Whatever did it, the openafs.org <http://openafs.org> web content looks ok now.
[20:56:02] <stephan.wiesand> Now all we need is an announcement
[20:56:16] <stephan.wiesand> Ben, would you?
[20:57:05] <stephan.wiesand> It's 10pm now. I shouldn't draft announcements before I had a good night's rest ;-)
[20:58:12] <mvita> which parts of this should be signed?
[20:58:21] <mvita> just the announcement?
[20:59:21] <stephan.wiesand> er... we actually sign stuff?
[20:59:58] <mvita> well, the OPENAFS-SA-2016-001.txt was signed
[21:00:27] <mvita> I noticed when I was looking for examples and then forgot to mention it when I pushed to gerrit
[21:01:58] <stephan.wiesand> That's finetuning IMHO.
[21:02:18] <stephan.wiesand> Though I'm a fan of singed content, especially nowadays
[21:04:30] <kadukoafs@gmail.com/barnowlE45450E9> I can send an announcement.
[21:04:42] <stephan.wiesand> Thanks.
[21:04:42] <mvita> thank you Ben
[21:04:48] <stephan.wiesand> !
[21:06:01] <mvita> It's best that Ben does it - Stephan is so burnt-out tired that he wants to singe the content.
[21:06:15] <stephan.wiesand> just remember it's jhutz and cg2v you have to pester to get it approved ;-)
[21:06:16] <mvita> ;-)
[21:07:58] <stephan.wiesand> Yes I am tired... I'll sign off for today now...
[21:08:33] <mvita> have a good rest
[21:08:37] <mvita> well earned
[21:08:42] <stephan.wiesand> Mark: s/singed/signed/... [growls]
[21:08:50] <mvita> heh
[21:11:20] <kadukoafs@gmail.com/barnowlE45450E9> Thanks for all the help, everyone
[21:12:01] <stephan.wiesand> Ben, Mark, Mike: Thanks a lot! Pulling off this stunt was a nice example of teamwork indeed.
[21:12:21] <mvita> it was a very good experience
[21:12:59] stephan.wiesand leaves the room
[22:30:44] meffie leaves the room
[22:30:46] meffie joins the room
[22:39:15] meffie leaves the room