14:00:12 <licquia> #startmeeting LSB Bug Triage 2014 Apr 4
14:00:12 <lsbbot> Meeting started Fri Apr  4 14:00:12 2014 UTC.  The chair is licquia. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:12 <lsbbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:00:25 * mwichmann arrives
14:01:29 <licquia> #info lsb: 6 new, 3 pleasetest
14:02:11 <licquia> #info fhs: 27 new
14:02:21 * licquia waves to mwichmann
14:02:50 * orc_fedo peers in
14:03:08 <licquia> ok, let's roll
14:03:14 <licquia> !lsbbug 3722
14:03:17 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3722 normal, P2, 4.1-updates, licquia, PLEASETEST , app checker doesn't like some RPMs
14:03:41 <orc_fedo> there are some I dont like either
14:04:00 <licquia> heh
14:04:42 <mwichmann> <comment=no>
14:04:54 <mwichmann> none, I meant (botch my own joke)
14:05:09 <licquia> looks fairly simple to test; a bare debian vm checking a rpm with app checker
14:05:20 <licquia> i can probably give this a look later
14:05:43 <mwichmann> closed bugs will be rewarded with one jelly bean each
14:05:54 <licquia> #action licquia test 3722 on a debian vm
14:05:56 <mwichmann> licquia could get a rather big bag...
14:06:02 <orc_fedo> seems like an alien failure to check for a userid set before trying to use it in the chown
14:06:08 <licquia> none of which i could eat :-)
14:06:22 <mwichmann> oh, well
14:06:49 <licquia> !lsbbug 3946
14:06:53 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3946 normal, P2, 4.1-updates, licquia, PLEASETEST , Add scandir and alphasort interface with tests
14:07:46 <mwichmann> possibly tweak the summary to include olver so it's a little clearer (like next one does)
14:07:50 <licquia> same deal; we just need to make sure olver builds and doesn't fail with this change
14:08:16 <licquia> done
14:08:44 * licquia checks buildbot status
14:09:36 <orc_fedo> licquia: any chance of putting a 'what are you doing?' and a 'poke it with a stick script on a box mats has access to?
14:09:38 <licquia> yup, currently building (modulo build issues on s390 archs)
14:10:40 <licquia> orc_fedo: being slow getting what those scripts would do
14:11:06 <orc_fedo> licquia: permits mats, when he notices a seemingly dead BB, and you are our of pocket, to remidiate it
14:11:13 <orc_fedo> out*
14:11:24 <licquia> ah
14:11:36 * licquia thinks he has access to that vm; checking
14:12:18 <orc_fedo> perhaps then just a usage note he can consult if his memory fails
14:12:35 <licquia> yes and no; access to vm possible, not sudo though
14:12:43 <licquia> that should be easy to fix
14:13:06 <orc_fedo> perhaps a restricted ssh command to do limited keyed access or such
14:13:14 <licquia> eh; it's mats
14:13:25 <licquia> not worried about his malicious hacking adventures :-)
14:13:29 <orc_fedo> and machines don't get compromised?
14:13:35 <orc_fedo> I dont even trust myself
14:14:04 <orc_fedo> we use such heavily in pmman
14:14:14 <licquia> that's true
14:14:30 <licquia> might think about that
14:15:26 <licquia> anyway, on this bug, the patch built, so there's nothing left but to test it
14:16:18 <licquia> could probably do that at the same time as the app checker test
14:17:11 <licquia> #action test 3946 using olver-core snapshot
14:17:35 <licquia> !lsbbug 3947
14:17:38 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3947 normal, P2, 4.1-updates, licquia, PLEASETEST , Extend mutex interface in OLVER to support prioceiling and protocol
14:18:22 <licquia> same deal, as observed; the same test should verify both bugs
14:18:43 <licquia> #action test 3947 using olver-core snapshot
14:19:06 <licquia> !lsbbug 3953
14:19:09 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3953 normal, P2, ---, mats, NEW , add link checker job to puppet to run periodically
14:19:49 <mwichmann> sorry for distraction, just finished up a list of 17 enums that need attention
14:20:03 <licquia> excellent
14:21:42 * licquia looks at url_checker.sh
14:24:16 <licquia> looks fairly straightforward; just need something like "url_checker.sh | mail -s 'report' some@mail 2>&1"
14:24:37 <licquia> any volunteers?
14:24:42 <mwichmann> not too sophisticated, and I'm not sure we want to start by dropping everything at the start
14:24:52 <orc_fedo> licquia: hand it to me please
14:24:58 <licquia> ok
14:25:02 <mwichmann> in light of bug 3954, which asks for a cache
14:25:34 <orc_fedo> that was what I was thinking .. a unified descriptive table of cacheing and checking and deviations is prolly needed
14:25:43 <licquia> #action orc_fedo set up auto runs of url_checker per bug 3953
14:25:45 <orc_fedo> hand me 3954 as well
14:25:54 <licquia> !lsbbug 3954
14:25:56 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3954 normal, P2, ---, mats, NEW , link checker script should build cache ... more
14:26:29 <licquia> should 3953 depend on 3954?
14:26:34 <orc_fedo> sure
14:26:40 <licquia> ok
14:27:09 <licquia> #action orc_fedo set up cache for url_checker per 3954
14:27:27 <licquia> #agreed set 3953 to depend on 3954
14:31:08 <licquia> alright
14:31:15 <licquia> !lsbbug 3955
14:31:18 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3955 normal, P2, ---, mats, NEW , should there be a TypeStd table?
14:32:28 <licquia> along with that question, thought comes to mind: do we need to fill in TypeStd entries for old types?
14:33:16 <mwichmann> this one was more a "make sure the question is remembered"
14:33:42 <mwichmann> I'm not overly upset if we don't act on it
14:34:04 <licquia> ok, so mark assigned and deal with later?
14:34:36 <mwichmann> unless someone here feels strongly it should be done for gtk/gdk 3.0 for LSB 5
14:34:58 <mwichmann> I'm not even sure if it makes sense
14:35:25 <orc_fedo> yes as a future item we may never address
14:35:40 <orc_fedo> we may turn out to need it for an update
14:35:47 <licquia> ok
14:35:57 <licquia> #agreed mark 3955 assigned
14:36:25 <licquia> !lsbbug 3956
14:36:27 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3956 normal, P2, ---, licquia, NEW , stop spam through old SF mailing lists, mark as abandoned
14:36:54 <licquia> this pretty much done with the list changes?
14:37:17 <orc_fedo> not til we get the credentials from teh next layer
14:37:26 <orc_fedo> updating bug -- you were gonna ping the aussies
14:37:59 <mwichmann> it must have kicked me off because I've not seen any of this spam
14:38:00 <orc_fedo> detail added cyoeh and rusty r
14:38:18 <orc_fedo> mwichmann: I have some tasty spam from it earlier today
14:38:52 <mwichmann> I absolutely don't care what happens here, so whatever solution is fine
14:39:25 <licquia> actually, as admins, don't we have the power to do things with all the lists?
14:39:44 <orc_fedo> licquia: no
14:39:54 <orc_fedo> they compartmentalize how they implemented it
14:40:04 <licquia> that's silly
14:40:18 <orc_fedo> it was state of the art when they started
14:40:51 <orc_fedo> why does unit not use a ldap auth and a DB backend everywhere?
14:41:03 <orc_fedo> s.unit/Linux/
14:41:13 <orc_fedo> same observation
14:41:44 * licquia supposes
14:41:59 <licquia> ok, will take this bug, then, and see if i can make progress on it
14:42:47 <licquia> #action licquia on 3956, still need to get access to fhs list on sourceforge somehow
14:44:55 <licquia> still find it exceedingly odd that a project admin can't manage the lists
14:45:08 <licquia> anyway...
14:46:54 <licquia> #action licquia try to gain admin access to the sourceforge lists per 3956
14:47:06 <licquia> !lsbbug 3957
14:47:08 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3957 major, P2, ---, licquia, NEW , /tset/LSB.os/network/accept/T.accept  /tset/LSB.os/network/connect/T.connect tests failed due to fcntl
14:48:03 <mwichmann> is it by accident we have not had problems for this?  or have we?
14:48:31 <licquia> checking problem db
14:49:43 <licquia> how odd; a commented-out prdb entry
14:50:03 <licquia> anyway, yes, we've had problems here before
14:50:18 <licquia> !lsbbug 2529
14:50:20 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=2529 normal, P2, 4.1, licquia, ASSIGNED , /tset/LSB.os/network/accept/T.accept 7 FAIL on ppc32
14:50:48 <licquia> !lsbbug 2332
14:50:50 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=2332 normal, P3, 4.1, stewb, RESOLVED FIXED, /tset/LSB.os/network/connect/T.connect 2 test can fail due to a hardcoded IP address
14:51:52 <mwichmann> not necessarily the same thing
14:53:00 <mwichmann> I see on fcntl doc that for F_GETFL the third arg is optional (linux says "ignored"), slightly better patch might be to omit it
14:53:30 * mwichmann reminds himself we are triaging, not fixing....
14:54:46 <licquia> patch does appear to contain at least one fix to something odd
14:55:02 <licquia> flags = (sockfd, F_GETFL);
14:55:04 <mwichmann> if((fcntl(sockfd, F_GETFL, flags)) == -1)
14:55:04 <mwichmann> {
14:55:04 <mwichmann> tet_printf("fcntl syscall returned error");
14:55:04 <mwichmann> return;
14:55:04 <mwichmann> }
14:55:05 <mwichmann> flags=flags | O_NONBLOCK;
14:55:07 <mwichmann> if((fcntl(sockfd, F_SETFL, flags)) == -1)
14:55:09 <mwichmann> {
14:55:11 <mwichmann> tet_printf("fcntl syscall returned error");
14:55:13 <mwichmann> return;
14:55:15 <mwichmann> }
14:55:29 <mwichmann> yeah, that does look wrong.... we're not changing the gotten flags
14:55:57 <mwichmann> because we don't save what we got
14:56:32 <licquia> not sure how we've escaped this previously; looks like a good patch
14:56:40 <licquia> 4.1 updates?
14:56:49 <mwichmann> yes
14:56:54 <licquia> (and 5.0, of course, but assuming this isn't 5.0 specific)
14:59:36 <licquia> #agreed assign 3957 to 4.1 updates, problemreport
14:59:58 <licquia> !lsbbug 3958
15:00:00 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3958 normal, P2, ---, dsilakov, NEW , can/should mkheader emit funcptrs separately?
15:01:01 <licquia> just assign, to deal with later?
15:01:07 <licquia> i.e. not a 5.0 issue
15:01:17 <orc_fedo> sure
15:01:55 <mwichmann> well...
15:01:56 <licquia> wondering though; mkheader changes could be a 5.0 impactor
15:02:26 <mwichmann> if we don't fix mkheader, we have to fix some other way, since the current state is broken header
15:02:54 <mwichmann> a thought is at the very end of this - using HeaderGroups to separate out funcptrs
15:03:03 <mwichmann> at this stage I think I don't really care how it's fixed
15:03:22 <mwichmann> but as things sit, this situation does block 5.0
15:03:42 <licquia> is there a dependent bug for the specific gtk situation?
15:03:54 <mwichmann> I don't recall atm
15:04:03 <mwichmann> there may be
15:04:14 <licquia> ok; we should probably tie directly to 5.0, then, to make sure it gets solved
15:04:21 <licquia> one way or another
15:04:24 <mwichmann> earlier on, when this problem was less prevalent, because there were more missing,
15:04:40 <mwichmann> I just made a small patch using the usual header mechanism, which moved the types up
15:04:53 <mwichmann> and I think that may have been bugged
15:05:18 <mwichmann> now there are 38 type complaints (probably not 38 types), I'd rather not try to manually order in a patch
15:05:25 <licquia> right
15:06:18 <licquia> ok, marking as a 5.0 blocker
15:06:32 <licquia> #agreed target 3958 at 5.0
15:07:16 <licquia> ok; anything else we need to talk about here?
15:07:26 <orc_fedo> licquia: another question: have you any ideas off the top of your head where in the VCS the lsbbot configs and customization/extensions are?
15:07:40 <licquia> lemme check...
15:07:57 <orc_fedo> I tried find and egrep to no success
15:08:56 <licquia> such as we have is in puppet-lsb, in modules/supybot
15:09:28 <orc_fedo> thanks -- will look and see if I can get that 'poke' buildbot into the otehr bot
15:10:14 <licquia> i note, however, that certain config files for that aren't in puppet, which should probably get fixed too
15:10:29 <orc_fedo> /opt/supybot/config/lsbbot.conf
15:10:35 <orc_fedo> for one example
15:10:42 <licquia> yup
15:11:20 <licquia> i would not be at all sore if that problem got fixed :-)
15:11:34 <licquia> ok, calling this meeting; can continue conversation in #lsb
15:11:39 <licquia> #endmeeting