14:01:50 <licquia> #startmeeting LSB Bug Triage 2013 Mar 21 14:01:50 <lsbbot> Meeting started Thu Mar 21 14:01:50 2013 UTC. The chair is licquia. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:50 <lsbbot> Useful Commands: #action #agreed #help #info #idea #link #topic. 14:02:08 <licquia> good morning 14:02:15 <mwichmann> mornin' 14:02:44 <licquia> looks like 2 new bugs 14:03:06 <mwichmann> indeed 14:03:33 <mwichmann> and they're test suite bugs, so I can nap for a bit? 14:03:36 * licquia has a list of bugs to reply to that aren't new 14:03:46 <licquia> heh 14:04:15 <mwichmann> okay. I have limited time today, but can we do a separate session a bit later to look at new libc interfaces? 14:04:47 <licquia> if you're time-limited, we can start with those 14:04:56 * licquia can probably take those two new bugs on himself 14:05:10 <mwichmann> we can do the libc stuff later 14:05:16 <mwichmann> you said you had other bugs to reply to? 14:05:34 <licquia> just have a list of test suite bugs from ksrot 14:06:03 * mwichmann is starting to get really irritated with db naming inconsistency 14:06:47 <mwichmann> I know "don't do things manually" is a useful rule, but there's no tool for what I'm fiddling with right now 14:07:01 <mwichmann> it's Interface.Ilibrary but Header.Hlib 14:07:28 <licquia> yeah, those can be a pain 14:07:45 <licquia> so, quickly: 14:07:52 <mwichmann> Interface.Iid but ArchInt.AIint 14:07:52 <licquia> !lsbug 3768 14:07:53 <lsbbot> licquia: Error: "lsbug" is not a valid command. 14:07:54 <mwichmann> ok 14:07:58 <licquia> !lsbbug 3768 14:08:00 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3768 normal, P2, ---, licquia, NEW , tset/POSIX.os/files/ioctl/T.ioctl 3 failed 14:08:32 <licquia> this is a kernel change; ioctl now returns ENOTTY instead of EINVAL in some cases 14:09:13 <mwichmann> goody goody 14:09:21 <licquia> done very deliberately, too 14:09:28 <licquia> not sure if this has a spec impact 14:09:29 <mwichmann> that makes it partly my business too... sounds like errata needed to allow new behavior 14:10:00 <licquia> maybe; that, or the test was too picky 14:11:32 <mwichmann> okay... where does this EINVAL->ENOTTY thing happen? 14:12:28 <licquia> so, there's an internal kernel error code that gets translated 14:12:59 <mwichmann> our own manpage is already different from POSIX 14:13:13 <licquia> hmm 14:13:16 * licquia checks that 14:13:27 <licquia> it looks like the POSIX man page is all about STREAMS 14:13:48 <mwichmann> http://refspecs.linux-foundation.org/LSB_4.1.0/LSB-Core-generic/LSB-Core-generic/baselib-ioctl-2.html 14:13:58 <mwichmann> yes, we redefine that 14:16:46 <licquia> so it appears that the test is wrong even according to our spec 14:16:54 <mwichmann> aha 14:17:14 <licquia> so, no spec errata 14:17:30 <mwichmann> but if it passed before, apparently matched former behavior 14:17:38 <mwichmann> but nice if spec not affected 14:17:56 <licquia> correct; the kernel folks are calling this a bug in the kernel 14:18:22 <licquia> supposedly, there's workaround code in other parts of the kernel for this bug, and they're unwinding all that 14:18:46 <licquia> so yes, the test follows the kernel's old (wrong) behavior 14:20:11 <mwichmann> so looks like you have data to update bug 14:20:43 <mwichmann> I guess this was an original bit of test code, maybe never modified to reflect LSB rules 14:20:46 <licquia> yup; plan to tack on to next stable update, add to problem_db2 14:21:23 <mwichmann> tset/POSIX.os means we didn't originally write it, though might have modified at some point 14:23:53 <licquia> looks like most of this test actually traces back to stewb 14:24:21 <licquia> yup; here's the copyright: 14:24:45 <licquia> Copyright 2008 IBM Corp. 14:25:09 <mwichmann> oh, that sounds like some of the contributed stuff from LTC 14:25:17 <licquia> yup 14:25:18 <mwichmann> anyway, let's move on 14:25:34 <mwichmann> (that's not stewb, except importing into suite) 14:25:39 <licquia> right 14:26:10 <licquia> #agreed 3768 is a test bug hidden by a kernel bug until recently 14:26:26 <licquia> #action fix 3768 in next stable update, add to problem_db2 now 14:26:47 <licquia> and my browser crashes, how fun 14:28:23 <licquia> !lsbbug 3769 14:28:26 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3769 normal, P2, ---, licquia, NEW , OLVER complains about glob64_spec: "Mediator failed" (no further details) 14:29:04 <licquia> our old friend 14:29:06 <mwichmann> "mystery stuff you don't get to understand has had a problem" 14:30:10 <licquia> so i'm thinking this gets a waiver and attached to the next stable update 14:30:24 <licquia> (of course, assuming we can figure out what's going on well enough to fix it) 14:31:10 <mwichmann> right 14:31:41 <licquia> #agreed 3769 is probably a bug in olver 14:31:52 <mwichmann> (side question: at what point does snapshotspecs update, is that time-based or at some point on commit) 14:32:03 <licquia> #action attach 3769 to next stable update, add to problem_db2 14:32:50 <licquia> looking... 14:34:53 <licquia> cron job, once a day at 3 am pacific 14:35:13 <mwichmann> okay, missed it... I'll post in #lsb in a moment 14:35:19 <mwichmann> more here? 14:35:29 <licquia> that's it for "interesting" bugs 14:35:48 * licquia hasn't updated bugs themselves yet; will do that later 14:35:55 <mwichmann> no prob 14:36:09 <licquia> so... libc? 14:36:15 <mwichmann> let's do later 14:36:25 <licquia> ok 14:36:35 <licquia> i have a few on my plate to look at 14:37:11 <mwichmann> I have to run an errand today, and with both phone lines dead (believe damaged by electric co. folks digging on road), I have to get done early 14:37:32 <licquia> oh, ok; my bugs are all test bugs, so might not be as interesting 14:37:40 <licquia> so if you need to take off, we can end early 14:40:26 <licquia> alright 14:40:29 <licquia> #endmeeting