- Thursday, October 7, 2021

release-team@conference.openafs.org

Thursday, October 7, 2021< ^ >

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

Room Configuration

Room Occupants

GMT+0
[03:06:40] Cheyenne leaves the room
[13:10:16] mbarbosa joins the room
[13:11:22] meffie joins the room
[13:43:27] Cheyenne joins the room
[15:49:47] wiesand joins the room
[15:59:51] mvita2 joins the room
[16:00:01] <mvita2> howdy
[16:00:08] <wiesand> Hello
[16:00:25] <Cheyenne> Hello all.
[16:00:33] <meffie> hi
[16:01:46] <wiesand> Not much to report from my side today, except that I dug out 14825 after re-reading Andrew's comment in 14503
[16:02:36] <wiesand> Not too serious it seems, but I guess this should be part of the next maintenance release.
[16:03:20] <wiesand> And the hot topic today is probably Linux 5.14?
[16:03:22] <meffie> thank you
[16:03:42] <Cheyenne> There is a required fix for linux-5.14 for an issue that was reported in the openafs-info list.  Simple fix, just checking to see what "extra" stuff might be needed (should be maybe 2 lines to fix)
[16:04:12] <wiesand> Good news, thanks.
[16:05:16] <Cheyenne> just need to see if the function declaration location changed (e.g. changed .h )
[16:05:42] <wiesand> Almost everything we merged on 1.8.x since the 1.8.8 release is FBSD support.
[16:06:37] <wiesand> Exceptions are 14793 14788 14787 14740
[16:07:30] <wiesand> I believe those are supposed to cause no behaviour change at all, or only very minor ones.
[16:07:49] <meffie> those were for the freebsd changes, right?
[16:08:03] <mvita2> the first one is
[16:08:39] <mvita2> 14788 14787 are tangential (probably found during fbsd work)
[16:08:39] <meffie> all except 14740
[16:08:59] <wiesand> Which means we could issue a 1.8.8.1 with just FBSD + Linux 5.14 (& 5.15?) without branching, and maybe even without reverting anything.
[16:09:40] <meffie> sounds good!
[16:09:47] <mvita2> aye +1
[16:11:21] <wiesand> Right, I think all except 14740 were part of the FBSD stacks by Andrew. And 14740 is supposed to be fully backward compatible IIUC.
[16:12:08] <wiesand> Looks like we have a near term plan then?
[16:12:59] <meffie> +1
[16:13:04] <Cheyenne> I should haver the 5.14 fix ready in just a bit.
[16:14:01] <wiesand> Great, thanks! Remind me of the Linux 5.15 status?
[16:15:13] <wiesand> 14791 is all we need for 5.15, right?
[16:15:20] <Cheyenne> I haven't compiled today, but there hasn't been a lot of activity either, so I think it's still "good"
[16:15:48] <Cheyenne> I will check that once I have the 5.14 stuff fixed
[16:16:11] <wiesand> Very good, thanks.
[16:17:19] <Cheyenne> q: the function I need was originally located in mm.h but was moved around 2006 to buffer_head.h (which has been around since original git commit).  Any problem with just doing an include for buffer_head.h w/out a autoconf test? -- makes it a 2 line fix.
[16:17:39] <wiesand> Meanwhile I'll try to continue going through the wish list and maybe pull up more changes, but won't merge for the time being.
[16:18:29] <kaduk@jabber.openafs.org/barnowl> I'm here now; sorry to be so late.
[16:18:35] <meffie> welcome!
[16:18:36] <mvita2> 2006 is what, 2.x series?
[16:19:14] <Cheyenne> let me check..
[16:19:25] <Cheyenne> (it's after 2.6.18 that's for sure)
[16:19:50] <mvita2> if it's in the Linux git history, we usually try to support it… so I personally would like to see an autoconf test for the header include
[16:20:10] <mvita2> (buffer_head.h or mm.h)
[16:20:42] <kaduk@jabber.openafs.org/barnowl> It's probably worth asking the question of when we will next bump our
minimum supported linux version.  The cutover to git was a nice
milestone to pass, but if something is 15 years old, it might be time
to let it go...
[16:20:49] <Cheyenne> both are present since first commit in linux repo.  it's just that the function I need was moved from mm.h into buffer_head.h in 2006
[16:20:58] <mvita2> but for now, you can just submit with buffer_head.h and mention what release that goes back to in the commit msg
[16:21:49] <mvita2> again, imo
[16:22:01] <Cheyenne> okay
[16:23:57] <wiesand> EL5 has 2.6.18 kernels. I know it's EOL since long, but breaking EL5 deliberately feels… icky…
[16:24:27] <kaduk@jabber.openafs.org/barnowl> Yeah, I'm not feeling in a hurry to specifically break EL5.
[16:24:54] <mvita2> or you could cheat and do both includes
[16:24:56] <Cheyenne> having the include won't break anything.. just that the include is not needed until the commit that moves __set_page_dirty_buffers into buffer_head.h
[16:25:07] <mvita2> actually I think I prefer that
[16:25:15] <Cheyenne> mm.h is already there.
[16:25:21] <mvita2> yeah
[16:25:27] <wiesand> OTOH, folks still using such an old distro should be happy with older openafs releases.
[16:25:55] <wiesand> Oh, that sounds good to me
[16:25:58] <mvita2> yup, you probably still need mm.h for other stuff
[16:26:09] <kaduk@jabber.openafs.org/barnowl> Oh, yeah, that seems reasonable
[16:26:10] <mvita2> just include both and be done
[16:26:23] <mvita2> no autoconf required
[16:26:36] <wiesand> :)
[16:26:39] <kaduk@jabber.openafs.org/barnowl> :)
[16:26:42] <Cheyenne> including both.. will work the whole way back to original linux git commit
[16:27:13] <mvita2> so let it be written, so let it be done
[16:27:29] <wiesand> nice
[16:27:36] <Cheyenne> (doing test build right now w/5.15 + the 5.14 fix) on against linux-5.15
[16:29:12] <wiesand> Ben, what's your stance on adding that 5.14 fix + 5.15 support + maybe a few more FBSD ones recently merged on master on top of the current HEAD of 1.8.x and call that 1.8.8.1?
[16:29:42] <kaduk@jabber.openafs.org/barnowl> That seems like a good idea
[16:34:30] <kaduk@jabber.openafs.org/barnowl> Should I be jumping on on master/1.9 now?
[16:34:39] <meffie> 1.8.8.1 is a palindrome
[16:34:50] <meffie> sounds good.
[16:34:52] <wiesand> unless anyone has other 1.8.x topics?
[16:35:50] <kaduk@jabber.openafs.org/barnowl> I don't have much specific to report, but I did at least make some
progress over the weekend, merging a handful of things and reviewing a
few more.  No real structured plan to it, though, mostly just trying
to catch up on stuff I already had open/had been meaning to look at.
[16:36:12] <kaduk@jabber.openafs.org/barnowl> In particular, no progress on the recvmmsg front that had gotten email
on -info a couple weeks ago
[16:37:25] <meffie> thank you for the merges!
[16:39:12] <kaduk@jabber.openafs.org/barnowl> The things currently sitting open are the "remove kdump" stack, safe
string functions/bosserver, and some of the test suite work Andrew did
(IIRC as precursor to rxgk-phase2)
[16:39:43] <meffie> thank you
[16:39:47] <Cheyenne> ben, if you want I can go through and rebase the rx.h cleanup and kdump removal
[16:40:35] <kaduk@jabber.openafs.org/barnowl> gerrit does not seem to be complaining about needing a rebase,
interestingly enough
[16:40:47] <kaduk@jabber.openafs.org/barnowl> at least for the rx.h change
[16:41:37] <kaduk@jabber.openafs.org/barnowl> I don't think I had any other specific topics for today, just a todo
list of things I need to do...
[16:42:27] <meffie> ok, thanks. let us know if there is anything we can help with. maybe we can do gerrit admin for you ?
[16:43:12] <kaduk@jabber.openafs.org/barnowl> Thanks for the offer to help, I will try to mention things where help
is useful.
[16:43:36] <meffie> +1
[16:43:37] <kaduk@jabber.openafs.org/barnowl> This past week I was mostly kept busy by preparing for the IESG
telechat that was just a couple hours ago.
[16:44:25] <kaduk@jabber.openafs.org/barnowl> Were there other topics for today?
[16:44:36] <wiesand> Not from my side
[16:45:00] <meffie> i have nothing
[16:45:24] <mvita2> nada
[16:45:45] <kaduk@jabber.openafs.org/barnowl> Okay.  Sounds like time to adjourn, then
[16:45:54] <mvita2> thank you all
[16:46:03] <meffie> thank you, have a safe day
[16:46:27] <kaduk@jabber.openafs.org/barnowl> thanks everyone, and have a good week
[16:46:30] <wiesand> Thanks a lot everybody!
[16:46:31] <Cheyenne> Will have the 5.14 patch uploaded once I finish writing the commit message and finish the test build
[16:46:40] <kaduk@jabber.openafs.org/barnowl> Thanks Cheyenne!
[16:49:24] meffie leaves the room
[16:52:32] wiesand leaves the room
[16:57:35] <Cheyenne> clean 5.15 build (w/ the 5.14 fix as well)
[21:33:50] mbarbosa leaves the room
[22:15:25] Cheyenne leaves the room