Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-02-20

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:11 bala joined #gluster-dev
01:30 shyam joined #gluster-dev
01:52 kdhananjay joined #gluster-dev
02:40 Frankl joined #gluster-dev
02:51 bharata-rao joined #gluster-dev
03:17 inodb joined #gluster-dev
03:20 shubhendu joined #gluster-dev
03:39 itisravi joined #gluster-dev
03:42 kdhananjay joined #gluster-dev
03:46 semiosis joined #gluster-dev
04:20 hagarth joined #gluster-dev
04:32 semiosis joined #gluster-dev
04:34 semiosis_ joined #gluster-dev
04:36 semiosis joined #gluster-dev
04:50 hagarth joined #gluster-dev
04:53 ndarshan joined #gluster-dev
04:56 shyam joined #gluster-dev
04:57 bala joined #gluster-dev
04:57 ajha joined #gluster-dev
05:02 ppai joined #gluster-dev
05:05 ndarshan joined #gluster-dev
05:08 kanagaraj_ joined #gluster-dev
05:31 aravindavk joined #gluster-dev
05:41 lalatenduM joined #gluster-dev
05:47 semiosis joined #gluster-dev
05:50 semiosis_ joined #gluster-dev
05:51 semiosis joined #gluster-dev
05:59 mohankumar joined #gluster-dev
06:05 raghu joined #gluster-dev
06:12 hagarth joined #gluster-dev
06:57 spandit joined #gluster-dev
08:10 kdhananjay joined #gluster-dev
08:55 ppai joined #gluster-dev
09:19 aravindavk joined #gluster-dev
09:22 ajha joined #gluster-dev
09:28 bharata_ joined #gluster-dev
09:32 shubhendu joined #gluster-dev
09:43 hagarth joined #gluster-dev
09:46 ndarshan joined #gluster-dev
10:00 bharata__ joined #gluster-dev
10:01 mohankumar__ joined #gluster-dev
10:29 Frankl_ joined #gluster-dev
10:43 ndarshan joined #gluster-dev
10:47 ajha joined #gluster-dev
10:50 aravindavk joined #gluster-dev
11:19 hagarth joined #gluster-dev
11:49 ira joined #gluster-dev
12:01 itisravi_ joined #gluster-dev
12:09 edward2 joined #gluster-dev
12:12 pk1 joined #gluster-dev
12:13 hagarth joined #gluster-dev
12:36 kdhananjay joined #gluster-dev
12:37 pk1 joined #gluster-dev
12:58 kkeithley1 joined #gluster-dev
13:19 kkeithley1 ndevos,jclift_,hagarth,a2: regression/rpm.t failures seem to be because configure is trying to do "gcc   /d/build/install/lib/ conftest.c  >&5".  Note space before conftest.c.
13:21 kkeithley1 But it looks like that cc cmd ought to be $CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5
13:22 kkeithley1 AFAICT /d/build/install/lib is only set in /opt/qa/regression.sh
13:23 portante joined #gluster-dev
13:23 kkeithley1 Some recent commit to configure.ac has borked the CFLAGS or CPPFLAGS?
13:24 kkeithley1 e.g. http://review.gluster.org/6957  ??
13:27 kkeithley_ still looking....
13:40 pk1 left #gluster-dev
13:52 shyam1 joined #gluster-dev
13:53 spandit_ joined #gluster-dev
13:55 hagarth1 joined #gluster-dev
13:58 JoeJulian_ joined #gluster-dev
14:02 [o__o] joined #gluster-dev
14:11 semiosis joined #gluster-dev
14:11 mohankumar__ joined #gluster-dev
14:16 semiosis joined #gluster-dev
14:17 badone joined #gluster-dev
14:39 ilbot3 joined #gluster-dev
14:39 Topic for #gluster-dev is now Gluster Development Channel - http://gluster.org | For general chat go to #gluster | Patches - http://review.gluster.org/ | Channel Logs - https://botbot.me/freenode/gluster-dev/ & http://irclog.perlgeek.de/gluster-dev/
14:42 ndevos kkeithley_: CFLAGS is indeed a little broken atm, http://review.gluster.org/7029
14:43 ndevos and I'm not very happy with that change either, we should not really set CFLAGS if there is no need, and keep whatever is set in the environment
15:06 lalatenduM joined #gluster-dev
16:02 jclift_ kkeithley_: The patch that breaks CFLAGS is the reason I was asking in gluster-meeting the other day about build breakers.
16:03 wushudoin joined #gluster-dev
16:03 * jclift_ kindof things we should push through 7029 now to fix the breakage just on principle
16:03 jclift_ s/things/thinks/
16:04 jobewan joined #gluster-dev
16:05 kkeithley_ I think I'd prefer to revert what broke it first rather than keep piling fixes on top of fixes.
16:05 kkeithley_ It's already brittle enough
16:06 kkeithley_ Let's see what happens with 7141 first
16:06 kkeithley_ I hate debugging through 27 layers of sh*t
16:06 kkeithley_ autoconf, automake, libtool, etc., etc.
16:08 kkeithley_ if I wanted to dig through layers of things I've have been a paleontologist or an archeologist
16:08 kkeithley_ why doesn't pidgin spell check like archeologist?
16:20 jclift_ kkeithley_: Yeah, I'd be ok with reverting it.
16:21 kkeithley_ make the person who broke it do it over — correctly
16:21 jclift_ I'm all for it.
16:21 jclift_ The problem in this instance is that the patch works as intended on EL6.
16:21 jclift_ It's just on F19/F20/EL7 where it breaks
16:22 jclift_ But we don't yet automatically test non-EL6 stuff
16:22 jclift_ So it wasn't caught until after it was committed
16:22 kkeithley_ well, it mostly works correctly on el6, except in our jenkins regression test environment
16:22 jclift_ Good point
16:23 kkeithley_ I wonder how it worked in the patch regression but somehow broke everything after that.
16:23 kkeithley_ in the patch's own regression test
16:24 kkeithley_ and /me wonders how it became my job to fix it.
16:25 jclift_ kkeithley_: You volunteered. :D
16:26 jclift_ hagarth: The build breaker is still in master.  Can we please either revert the commit that broke it, or merge the fix?
16:26 jclift_ hagarth: (like, in the next hour or something?)
16:26 jclift_ hagarth: Breaker: http://review.gluster.org/7029
16:27 jclift_ Dammit, bad cut-n-paste
16:27 jclift_ Breaker: http://review.gluster.org/6957
16:27 jclift_ 7029 is the fix
16:30 jclift_ Hmmm, I'm definitely sounding shittier in my communications than I should be
16:30 jclift_ Sorry. :(
16:33 hagarth jclift_: no worries,
16:33 hagarth I will merge this and ask lpabon to update us if this breaks anything
16:34 jclift_ Thanks. :)
16:35 jclift_ kkeithley_: You might want to chime in here
16:35 jclift_ kkeithley_: Just in case reverting is better than merging
16:36 jclift_ hagarth: It sounds like the original broken patch is also the cause of the rpm.t regression failures we're seeing in automated testing online
16:36 jclift_ ... and I've just realised that 7029 might not fix that bit
16:38 hagarth jclift_: done
16:38 kkeithley_ http://review.gluster.org/7141 is running the regression/rpm.t right now. reverting versus a fix on top makes very little difference to me.
16:39 jclift_ kkeithley_: k.  The fix is in master now.  Thanks hagarth. :)
16:40 kkeithley_ sorry for the delay, I was moving cars around to shovel some snow around on my driveway.
16:41 kkeithley_ now can we solve whatever it is that's filling the regression test logs with "loop: can't get info on device /dev/loop2: Transport endpoint is not connected"
16:42 kkeithley_ and I volunteered because I thought we were talking about a different problem with rpm.t
16:42 * kkeithley_ is grumpy today
16:43 jclift_ "Transport endpoint is not connected" <-- this indicates a glusterfsd crash doesn't it?
16:43 jclift_ kkeithley_: Yeah, I seem to be grumpy too. :(
16:43 jclift_ Phase of moon thing maybe?
16:43 * jclift_ shurgs
16:43 kkeithley_ full moon is past
16:43 jclift_ shrugs even
16:44 jclift_ Maybe that's is?  We get grumpy just after the full moon has passed?
16:44 jclift_ :D
16:44 kkeithley_ I don't think glusterfsd crash is the case here. There's be core files
16:44 jclift_ Good point
16:44 * jclift_ gets back to work.  Stops being distracted.
16:44 ndevos jclift_: sounds like the glusterfs-fuse client exited without unmounting
16:45 jclift_ ndevos: Interesting.
16:45 * jclift_ learns something new :D
16:45 ndevos jclift_: or, glusterfs-fuse is hanging, and access to the mountpoint is timing out
16:45 kkeithley_ well, reverting http://review.gluster.org/6957 wasn't the fix anyway. Not the droid I was looking for.
16:46 ndevos jclift_: when you mount a volume, attach gdb to the process and try to access the mountpoint from an other terminal
16:46 jclift_ kkeithley_: Just to point out, the 6957 breaking commit, is very different to the fix that was applied.
16:46 kkeithley_ okay
16:46 jclift_ The fix isn't a full reversion of the breaking commit, it just adds $CFLAGS back in or something
16:47 kkeithley_ as long as it works again
16:47 kkeithley_ I guess
16:47 jclift_ So if 6957 added in an extra argument, or space (etc) in the gcc line where it shouldn't be, then the 7029 fix that was applied probably didn't help the rpm.t case.
16:48 ndk joined #gluster-dev
16:48 jclift_ Saying this because it sounds like 6957 might be the droid you're looking for, and 7029 was a first go at a jedi hand wave trick to distract you. ;)
16:49 hagarth will take a look in my morning - have had a very long day today.
16:49 jclift_ hagarth: No worries.
16:49 jclift_ hagarth_afk: Thanks for your help. :)
17:16 mohankumar__ joined #gluster-dev
17:52 jobewan joined #gluster-dev
17:53 jobewan joined #gluster-dev
18:10 jobewan joined #gluster-dev
19:19 ira joined #gluster-dev
21:06 glusterbot joined #gluster-dev
21:07 semiosis joined #gluster-dev
21:23 glusterbot` joined #gluster-dev
21:34 lkoranda joined #gluster-dev
21:55 purpleid1a joined #gluster-dev
22:01 JoeJulian joined #gluster-dev
22:02 jobewan joined #gluster-dev
22:02 hchiramm__ joined #gluster-dev
22:06 semiosis joined #gluster-dev
22:23 tg2 joined #gluster-dev
22:36 semiosis joined #gluster-dev
22:46 semiosis joined #gluster-dev
23:09 jobewan joined #gluster-dev
23:22 jclift_ kkeithley1: Am stepping through the code for rpm.t atm, to better understand it.
23:23 jclift_ kkeithley1: Something seems weird in the lines under the "# checkout the sources to a new directory to execute ./configure and all"
23:23 jclift_ It has "COMMIT=$(git describe)", then a few lines later "git checkout -q -b rpm-test ${COMMIT}"
23:25 jclift_ It _seems_ like the git describe is trying to retrieve the closest git tag, and then the git checkout is trying to create a new branch (rpm-test) using that.
23:26 jclift_ However, from reading the git-describe man page that can't work (and doesn't seem to), since git describe doesn't return a straightforward tag name nor even a usable commit.
23:27 jclift_ Anyway, just mentioning, as I'm not yet sure what the intended purpose of this code is. (still stepping through and watching)

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary