Home
release-team@conference.openafs.org
Wednesday, November 22, 2017< ^ >
Room Configuration
Room Occupants

GMT+0
[13:45:36] meffie joins the room
[14:05:32] kadukoafs@gmail.com/barnowlEBF84E62 joins the room
[14:56:24] mvita joins the room
[16:01:12] <meffie> good day.
[16:01:37] <mvita> indeed.
[16:05:24] <mvita> <crickets>
[16:09:08] wiesand joins the room
[16:09:27] <wiesand> sooorryyyy
[16:10:57] <wiesand> If you're still here: any news on the EL7.4 kerne; issues?
[16:11:04] <mvita> yes
[16:11:21] <mvita> as I wrote to the list, I know the root cause and I'm working on a patch
[16:11:32] <mvita> I should submit a provisional patch to gerrit today
[16:11:42] <wiesand> Saw that. It's great that you found it.
[16:11:45] <mvita> and perhaps a better (autoconf) test next week.
[16:12:06] <mvita> the autoconf thing is difficult for this case
[16:12:11] <wiesand> Sounds like a good plan.
[16:13:00] <wiesand> I'd like to delay the next release to include this fix.
[16:13:05] <mvita> okay
[16:14:12] <wiesand> If Fedora pushes the 4.14 update very soon, or Ubuntu backports 4.14 to the 4.0.4  name, we can still scramble
[16:15:03] <wiesand> The other Linux stuff is still being worked on too anyway.
[16:15:24] <mvita> yes
[16:15:26] <wiesand> But I admit to have lost track of what happened during the last week :-(
[16:16:07] <wiesand> So, anything else to discuss today?
[16:16:09] <mvita> for the other problem (kernel_read), someone found that Damien's code is okay for 4.13 - it's only broken for 4.14
[16:16:20] <mvita> (signature change in Linux, sigh)
[16:16:26] <wiesand> Ah
[16:16:51] <wiesand> IIRC Mike remarked that it breaks on EL7.4 too
[16:16:52] <mvita> so Marcio is adding more autoconf to make ti work on both
[16:17:13] <meffie> Ben beat marcio to it.
[16:17:32] <mvita> oh, is that what you were talking about ??
[16:17:36] <meffie> there's an update in gerrit now waiting for review
[16:17:41] <mvita> okay, cool
[16:18:11] <meffie> gerrit 12769
[16:18:35] <mvita> bringing up the rear in Linux land, our dailies are currently not building for 4.15 pre-rc1
[16:18:35] <meffie> it looks good, but i want to look at the linux versions closer.
[16:18:50] <mvita> I don't think anyone's had time to look at why yet
[16:19:23] <meffie> damien's not able to look yet, maybe after the getcwd() stuff we can look.
[16:19:31] <wiesand> yes, saw that too
[16:19:31] <mvita> yes.
[16:19:55] <wiesand> "checking whether kernel module builds work… no"
[16:20:05] <mvita> <sadness>
[16:20:54] <wiesand> there's a several thousand lines pull request for kafs, but I'm not sure it has been merged yet
[16:21:32] <meffie> linus is too busy yelling at security people
[16:21:37] <mvita> heh
[16:21:43] <meffie> (i dont blame him)
[16:21:57] <mvita> Linus is too busy yelling.  Period.
[16:22:00] <mvita> ftfy
[16:22:04] <meffie> heh
[16:22:06] <wiesand> I missed something here
[16:22:21] <kadukoafs@gmail.com/barnowlEBF84E62> Can I send now?
[16:22:28] <mvita> ack
[16:22:33] <wiesand> "yes you can"
[16:22:39] <kadukoafs@gmail.com/barnowlEBF84E62> Oh yay.  gmail was having trouble earlier, and I had another meeting
cross-scheduled with this
[16:22:58] <kadukoafs@gmail.com/barnowlEBF84E62> So, uh, sorry I'm late.
[16:23:06] <meffie> no worries
[16:23:16] <wiesand> only slightly more than me…
[16:23:20] <kadukoafs@gmail.com/barnowlEBF84E62> w.r.t. kernel_read, IIRC Damien's patch is good for 4.14 but breaks
for 4.13 (i.e., the other way around)
[16:23:37] <mvita> oh, sorry
[16:23:55] <kadukoafs@gmail.com/barnowlEBF84E62> Well, I want to double-check to make sure that I'm not
misunderstanding :)
[16:24:29] <mvita> I feel like I'm going through the Linux car-wash.  Sans car.
[16:24:36] <kadukoafs@gmail.com/barnowlEBF84E62> On the EL7.4 thing, with d_invalidate() (not) returning -EBUSY, didn't
we already have some code to cope with this in a previous patch?  (So
that we just need to update what configure checks we do, now?0
[16:24:48] <mvita> correct
[16:25:01] <mvita> but the configure check is gonna be difficult
[16:25:18] <kadukoafs@gmail.com/barnowlEBF84E62> Oh, I don't doubt it.  I just wanted to check my memory.
[16:25:20] <mvita> very "red-hat-ish"
[16:25:39] <kadukoafs@gmail.com/barnowlEBF84E62> Even getting the configure check for the kernel_read() thing was a
little exciting
[16:25:53] <mvita> but I have a stopgap I will submit today while I continue to work on autoconf
[16:26:02] <kadukoafs@gmail.com/barnowlEBF84E62> But I think I did enough testing to verify that it's doing what I
want.
[16:26:08] <kadukoafs@gmail.com/barnowlEBF84E62> *nods*, thanks
[16:26:28] <kadukoafs@gmail.com/barnowlEBF84E62> Mike, do we know what we need to do to get buildbot commenting on
changes again?
[16:26:59] <meffie> i need to fix the opensuse stuff.
[16:27:13] <meffie> let me make that a priority today.
[16:27:18] <kadukoafs@gmail.com/barnowlEBF84E62> Okay.  I thought that was what it looked like, but it's always hard to
be sure until it's done :)
[16:28:09] <kadukoafs@gmail.com/barnowlEBF84E62> Hopefully I am not stomping over Stephan's agenda when I ask if you
wanted to re-review the rx event patches after I made changes for
Mark's review.
[16:28:25] <kadukoafs@gmail.com/barnowlEBF84E62> But I think the buildbot should probably take precedence over that.
[16:28:39] <wiesand> no, I think we were finished with 1.6
[16:28:57] <meffie> ok
[16:29:46] <kadukoafs@gmail.com/barnowlEBF84E62> > I think we were finished with 1.6
Okay.  I think waiting for the EL7.4 fix makes sense, at least for
now, FWIW.  Debian is only on a 4.13 kernel at this point, so no huge
rush for 4.14 support from that end.
[16:30:21] <kadukoafs@gmail.com/barnowlEBF84E62> meffie: and I think given your previous round of review on the rx
event stuff, I'd be okay taking just mark's re-review of the latest
iteration, if you don't want to put more time in it.
[16:30:35] <wiesand> Thanks for the data point
[16:30:39] <meffie> ok, thank you.
[16:31:01] <kadukoafs@gmail.com/barnowlEBF84E62> so it sounds like I should not wait for you on the rx event stuff,
then?
[16:31:34] <kadukoafs@gmail.com/barnowlEBF84E62> I mostly just don't want to merge it and have you feel like you had
still wanted to look, but if you don't care, let's go ahead and get it
in.
[16:33:24] <meffie> i dont recall the changes let me take a quick look.
[16:33:36] <kadukoafs@gmail.com/barnowlEBF84E62> Okay, thanks.
[16:34:12] <kadukoafs@gmail.com/barnowlEBF84E62> In a related vein, is anyone planning to review the kernel_read()
change (where I stomped on Marcio's change)?
[16:34:32] <mvita> yes
[16:35:00] <mvita> I can do that after I push my provisional getcwd fix
[16:35:00] <kadukoafs@gmail.com/barnowlEBF84E62> Just checking :)
[16:35:38] <kadukoafs@gmail.com/barnowlEBF84E62> My recollection is that after we get that and the rx-event stuff in,
we're good for a 1.8.0pre3.
[16:36:24] <meffie> ben, thanks for the fix to 12769. marcio was dreading the autoconfig goo.
[16:36:27] <mvita> yay!
[16:36:54] <wiesand> x-mas time :)
[16:37:02] <kadukoafs@gmail.com/barnowlEBF84E62> Oh, that's good to know -- I was worried that marcio would be annoyed
that I caused him to have wasted time on what he had already gotten
started
[16:37:17] <kadukoafs@gmail.com/barnowlEBF84E62> because yes, that goo is never much fun
[16:38:24] <kadukoafs@gmail.com/barnowlEBF84E62> At least twice I had gone and added the AC_DEFUN the function for
this specific check and then failed to actually *call* the new
function, which is frustrating when configure takes 10 minutes or
whatever.
[16:39:07] <mvita> yes, I was trying to find a way to test just a single autoconf but didn't get very far
[16:39:28] <kadukoafs@gmail.com/barnowlEBF84E62> Anyway, are there other 1.8 (or 1.6) topics?
[16:40:10] <wiesand> A parallel configure would be a blessing…
[16:40:26] <wiesand> But I have no further topics.
[16:40:30] <kadukoafs@gmail.com/barnowlEBF84E62> As far as master, Andrew made a lot more useful comments on the rxgk
stuff, but I haven't really had a chance to do much with them.  And
there's also the aklog debugging output that I need to look at.
So my summary of master stuff is "blocking on me"
[16:41:55] <kadukoafs@gmail.com/barnowlEBF84E62> So if you see Andrew, please tell him "thank you" again for the rxgk
review.
[16:42:16] <meffie> also, i need to redo the aklog akimperonate patches based on andrews feedback. those are not blockers for 1.8 tho.
[16:42:27] <meffie> i will tell him. thanks!
[16:43:08] <kadukoafs@gmail.com/barnowlEBF84E62> Oh right, that also happened.
[16:43:53] <meffie> those changes are just to make testing easier for us, so should not block 1.8.0
[16:43:54] <kadukoafs@gmail.com/barnowlEBF84E62> I wish I had a better understanding of what was going on with
akimpersonate, as I used to understand it pretty well.
[16:44:13] <kadukoafs@gmail.com/barnowlEBF84E62> *nods*
[16:44:46] <meffie> it would be nice to have a separate command called 'akimpersonate', but i guess that would confict the the existing hiemdal command.
[16:45:47] <kadukoafs@gmail.com/barnowlEBF84E62> We could presumably find a different name, but yeah.
[16:49:09] <kadukoafs@gmail.com/barnowlEBF84E62> Okay.  So, Mark has a temporary workaround to push to gerrit while
working on better autoconf for the EL7.4 thing, and reviewing 12769
(kernel_read); Mike has fixing the buildbot; I have merging bits to
make 1.8.0pre3 and dealing with the rxgk/aklog/akimpersonate stuff on
master; we're waiting for the EL7.4 stuff to do anything with 1.6.  Is
that about it?
[16:49:38] <wiesand> Nice summary
[16:49:49] <mvita> and the current linux daily failures, bringing up the rear
[16:50:19] <wiesand> there are weeks to fix that ;-)
[16:50:32] <wiesand> (luxury)
[16:51:13] <mvita> aye
[16:51:19] <wiesand> Looks like we're done for today. Adjourn?
[16:51:30] <mvita> d'accord
[16:51:54] <kadukoafs@gmail.com/barnowlEBF84E62> Sounds like a plan.
Thanks everyone!
[16:52:33] <wiesand> Thanks a lot everyone! Sorry again for being late today. Unfortunately, there's some chance that could happen next week too though…
[16:52:48] <wiesand> Bye
[16:52:52] wiesand leaves the room
[16:52:56] <mvita> np Ben, thanks
[16:53:07] <mvita> Happy Thanksgiving to those who celebrate it
[16:58:34] meffie leaves the room
[17:38:12] mvita leaves the room
[17:48:54] mvita joins the room
[18:35:51] meffie joins the room
[22:00:37] meffie leaves the room