- Thursday, August 18, 2022

release-team@conference.openafs.org

Thursday, August 18, 2022< ^ >

kaduk@jabber.openafs.org/barnowl has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[15:52:52] meffie joins the room
[15:59:09] wiesand joins the room
[16:00:05] <kaduk@jabber.openafs.org/barnowl> Greetings
[16:00:16] <wiesand> Hello
[16:00:46] <meffie> hello
[16:02:15] <wiesand> No news from my side today, but work on 1.8.9 will commence this evening.
[16:02:28] Cheyenne joins the room
[16:02:33] <Cheyenne> Hello all
[16:02:42] <kaduk@jabber.openafs.org/barnowl> Looking forward to 1.8.9!
[16:04:30] <kaduk@jabber.openafs.org/barnowl> I got up a bit early this morning and made progress clearing out "to
be reviewed" tabs in my browser.  Merged a few things, even...
[16:04:31] <Cheyenne> I pushed the 5.19 commits (15093-15095).  I also pushed a 1.8.x version the "linux: Call put_page if errors in aops->readpages" commit
[16:04:59] <wiesand> Thanks!
[16:04:59] <Cheyenne> (5.19 for 1.8.x that is)
[16:05:11] <meffie> thanks all!
[16:05:40] <wiesand> I have to set up a new test system. The old one was decommissioned.
[16:05:44] <kaduk@jabber.openafs.org/barnowl> I should probably do something with my local branch adding 5.19
support for the debian packaging ... like, maybe upload to debian.
[16:08:26] <meffie> wiesand: is there anything we can do to help setup a new test system?
[16:08:45] <wiesand> No, thanks.
[16:09:08] <meffie> ok, thanks.
[16:09:20] <wiesand> It's just that I don't merge code before I have seen it run.
[16:09:29] <kaduk@jabber.openafs.org/barnowl> Good policy!
[16:09:38] <meffie> yes, thanks!
[16:11:52] <kaduk@jabber.openafs.org/barnowl> Did anyone have anything to say about the flurry of AIX-related
activity courtesy of Ben Huntsman?
I'm still hoping we can get a buildbot worker again...
[16:12:48] <meffie> yes, cheyenne and i have been helping with git and gerrit questions.  and ben h is looking into making an buildbot worker.
[16:12:56] <kaduk@jabber.openafs.org/barnowl> excellent!
[16:13:09] <Cheyenne> meffie and I have been helping him out with some git related questions, as well as some afs stuff
[16:13:12] <meffie> i've given him a name and password and some info.
[16:13:28] <kaduk@jabber.openafs.org/barnowl> Thank you both!
[16:13:57] <meffie> he says the aix6.1 box is only used by him, so he should be able to run a worker on it. i dont know how hard it will be to install python there.
[16:14:53] <meffie> it looks like there is only one aix commit left to sort out. the one where aix does not have strlcat and friends
[16:16:12] <Cheyenne> I took a quick look at that.  strlcat, strlcpy and snprintf.  Looks like the strlcat and strlcpy shouldn't be a problem, but snprintf might be.
[16:16:28] <meffie> also, i dont know if he actually got /afs to mount yet.
[16:17:28] <kaduk@jabber.openafs.org/barnowl> I thought he ran into some weird in-kernel API behavior shift that
prevented getting /afs mounted
[16:17:50] <meffie> yes, that's the last i saw as well.
[16:17:51] <kaduk@jabber.openafs.org/barnowl> But maybe that was only on the newer AIX version and he also has an
older version available
[16:19:00] <kaduk@jabber.openafs.org/barnowl> I'll also thank Mark and Marcio and the crew for working on the viced
deadlock in failed callback recovery; a gnarly situation to untangle!
[16:19:13] <meffie> yes.
[16:19:34] <meffie> it took a while to get a reproduction in the lab.
[16:19:40] <kaduk@jabber.openafs.org/barnowl> I can imagine
[16:20:20] <meffie> it would be nice to have the fix in 1.8.9, but it seems to be a rare bug so i guess we should not block on it?
[16:21:01] <kaduk@jabber.openafs.org/barnowl> I don't think I can commit to reviewing the whole situation in a fast
enough timescale to avoid holding up 1.8.9, and I really don't want to
hold up 1.8.9
[16:21:10] <meffie> agreed
[16:21:39] <wiesand> ok
[16:22:15] <kaduk@jabber.openafs.org/barnowl> Anyway, I currently have "use make to build the man pages" open again.
Not sure if there are other things I should be prioritizing (other
than the aforementioned viced deadlocks)
[16:22:36] <meffie> i think the situation basically comes down to "blocking inside rx_Multi is asking for deadlocks"
[16:22:53] <kaduk@jabber.openafs.org/barnowl> When you put it like that...
[16:23:13] <meffie> :)
[16:23:28] <meffie> woot, make man pages :)
[16:23:43] <meffie> sorry it's such a big patch.
[16:24:33] <kaduk@jabber.openafs.org/barnowl> It's the price that has to be paid :)
[16:25:09] <kaduk@jabber.openafs.org/barnowl> Did we have other topics for today?
[16:25:33] <Cheyenne> I'm cleaning up a linux 6.0 patch.. still fairly simple
[16:25:57] <meffie> i have nothing else. i have been continuing to make progress on my gerrit review backlog.  getting to some really old changes now.
[16:27:02] <kaduk@jabber.openafs.org/barnowl> Happy to see progress on the review backlog, but sad to think about
how many things have been in the queue for years
[16:27:11] <Cheyenne> Oh.. I did dig a little on the LTO stuff (thanks meffie) it appears to be only 2 problems.. but I haven't "fixed" anything yet.
[16:27:22] <meffie> LTO?
[16:27:29] <kaduk@jabber.openafs.org/barnowl> link-time optimization
[16:27:31] <Cheyenne> link-time-optimization
[16:27:39] <kaduk@jabber.openafs.org/barnowl> It produced some new build errors, IIRC
[16:27:41] <meffie> ah, thanks.
[16:27:51] <Cheyenne> yeah.. turned out it's just 2 errors.
[16:28:29] <meffie> the case you mentioned last week look like a trival bug. we dont use the one that is a char*, it's just to appease the linker.
[16:29:10] <Cheyenne> correct.. I didn't dig into the other problem (other than seeing that there were just the 2 warnings/errors).
[16:30:22] <meffie> ok, thanks
[16:31:20] <meffie> motion to adjourn?
[16:31:38] <Cheyenne> nothing more here... thanks everyone :)
[16:32:00] <meffie> please stay safe and stay healthy all!
[16:32:06] <wiesand> motion sustained - thanks a lot!
[16:32:13] <kaduk@jabber.openafs.org/barnowl> Yes, thanks everyone and have a good week!
[16:32:39] meffie leaves the room
[16:33:19] wiesand leaves the room