#linuxcnc-devel Logs

Dec 07 2017

#linuxcnc-devel Calendar

08:23 AM cradek: ooh zultron is back on state-tags!
08:55 AM -!- #linuxcnc-devel mode set to +v by tolkien.freenode.net
09:41 AM skunkworks: yaY!
09:44 AM skunkworks: pcw_home, you're arguing again... ;)
10:05 AM hazzy: Hurray for zultron!
10:07 AM seb_kuzminsky: yeah, that dude's amazing :-)
10:08 AM skunkworks: hmm - what did he do? Is it in master?
10:08 AM seb_kuzminsky: does anyone remember why we reverted the drain_interp_queue stuff mentioned in the issue #49 that he just reopened?
10:08 AM skunkworks: :)
10:30 AM skunkworks: did anyone make a wheezy livecd with uspace?
10:32 AM seb_kuzminsky: i think the only uspace livecd we have is the stretch one jepler made
10:32 AM seb_kuzminsky: we should probably release it for real, i think
10:33 AM seb_kuzminsky: http://linuxcnc.org/testing-stretch-rtpreempt/
10:34 AM seb_kuzminsky: we've reached -r9 on that iso, so if we make one release it'll be -r10 and it'll sort wrong, so now's the time
10:34 AM skunkworks: I have used it often - no issues
10:34 AM seb_kuzminsky: same here
10:34 AM seb_kuzminsky: i've been running my bridgeport on it for a while now (it's a servo-thread only machine, with hm2)
10:35 AM seb_kuzminsky: software stepping might be a problem on some hardware :-(
10:35 AM skunkworks: The only thing that doesn't work is gmoccopy - which I understand is fixed in master?
10:35 AM seb_kuzminsky: wait, that's news to me
10:35 AM seb_kuzminsky: fixed in master but not in 2.7? boo
10:35 AM skunkworks: I run my lathe with uspace - 50us base thread
10:36 AM skunkworks: IIRC
10:36 AM seb_kuzminsky: -50 us! wow!
10:36 AM skunkworks: no issues
10:36 AM seb_kuzminsky: except the tachyon emissions
10:37 AM skunkworks: heh
10:37 AM seb_kuzminsky: what's wrong with gmoccapy on the stretch iso?
10:37 AM skunkworks: hold on
10:37 AM skunkworks: Found an error!
10:37 AM skunkworks: The following information may be useful in troubleshooting:
10:37 AM skunkworks: Traceback (most recent call last):
10:37 AM skunkworks: File "/usr/bin/gmoccapy", line 102, in <module>
10:37 AM skunkworks: from gmoccapy import player # a class to handle sounds
10:37 AM skunkworks: File "/usr/lib/python2.7/dist-packages/gmoccapy/player.py", line 27, in <module>
10:37 AM skunkworks: import gst
10:37 AM skunkworks: ImportError: No module named gst
10:38 AM skunkworks: it was mentioned on the list recently
10:38 AM skunkworks: something about the video player or something
10:38 AM seb_kuzminsky: ooh, yeah i think i remember seeing emails about that
10:38 AM * seb_kuzminsky blinks
10:39 AM seb_kuzminsky: commit 7c400794e3ce0b2ef994d0a3c24f3a6987536841 in 2.7 purports to fix gst in gmoccapy
10:40 AM skunkworks: hmm - This is 2.7.11
10:40 AM seb_kuzminsky: https://github.com/LinuxCNC/linuxcnc/commit/7c400794e3ce0b2ef994d0a3c24f3a6987536841
10:40 AM seb_kuzminsky: which is in 2.7 but not in 2.7.11
10:40 AM seb_kuzminsky: so maybe we need 2.7.12 and a new iso before the official release
10:41 AM skunkworks: Maybe!
10:41 AM seb_kuzminsky: can you repro that bug? can you test is on the tip of 2.7?
10:41 AM seb_kuzminsky: or does it just happen anytime you start gmoccapy?
10:42 AM skunkworks: anytime you start gmoccopy
10:42 AM seb_kuzminsky: great! i'll test it right now
10:42 AM skunkworks: I can build 2.7 and test too
10:42 AM skunkworks: (running updates right now)
10:44 AM seb_kuzminsky: cool
10:47 AM skunkworks: yes - swiching to 2.7 branch - gmoccopy launches
10:47 AM seb_kuzminsky: great!
10:48 AM skunkworks: v2.7.11-52-gde97fdd10
10:48 AM seb_kuzminsky: great
11:01 AM skunkworks: well crap - now I am arguing
11:27 AM hazzy1 is now known as hazzy
11:30 AM * JT-Shop never argues
11:32 AM seb_kuzminsky: JT-Shop: yeah you do
11:57 AM JT-Shop: lol
01:30 PM Tom_itx is now known as Tom_L
01:39 PM linuxcnc-build_: build #1298 of 4019.deb-jessie-rtpreempt-i386 is complete: Failure [4failed shell_1] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4019.deb-jessie-rtpreempt-i386/builds/1298 blamelist: Jeff Epler <jepler@unpythonic.net>
02:42 PM linuxcnc-build_: build #3390 of 1405.rip-wheezy-armhf is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/3390 blamelist: Dewey Garrett <dgarrett@panix.com>
03:26 PM linuxcnc-build_: build #5239 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/5239 blamelist: Dewey Garrett <dgarrett@panix.com>
03:32 PM seb_kuzminsky: hi dgarr
03:33 PM dgarr: linuxcnc-build: force build --branch=dgarr/limit3_simple_tp 0000.checkin
03:33 PM dgarr: i suppose the buildbot fail is an arm prob, runtests worked on several machines i tested
03:34 PM seb_kuzminsky: maybe...
03:35 PM seb_kuzminsky: the wheezy failure looks different from the usual flaky arm buildslave problems
03:35 PM seb_kuzminsky: it failed in the jogwheel-joint test, because it jogged to the wrong place:
03:35 PM seb_kuzminsky: joint jogged from 0.000000 to 0.001000 (1 counts at scale 0.001000)
03:35 PM seb_kuzminsky: joint jogged from 0.000000 to -0.250000 (10 counts at scale -0.025000)
03:35 PM seb_kuzminsky: joint jogged from 0.000000 to -10.001000 (-100 counts at scale 0.100000)
03:35 PM seb_kuzminsky: joint 2 didn't get to target (start=0.000000, target=-10.000000, got to -10.001000)
03:36 PM seb_kuzminsky: not sure what to make of that
03:36 PM pcw_mesa: close enough for Jazz
03:36 PM seb_kuzminsky: the jessie failure before that looks like the network glitched
03:37 PM seb_kuzminsky: pcw_mesa: is Jazz the codename for Mach4?
03:37 PM dgarr: close enough for armhf maybe
03:38 PM seb_kuzminsky: if it only does it once in 100 tests, on average it's within 0.00001 of the commanded position and surely that's fine
03:38 PM dgarr: linuxcnc-build_: force build --branch=dgarr/limit3_simple_tp 0000.checkin
03:38 PM linuxcnc-build_: The build has been queued, I'll give a shout when it starts
03:40 PM dgarr: seb_kuzminsky: is this relevant: task: 2991 cycles, min=0.000537, max=0.026711, avg=0.001233, 3 latency excursions
03:40 PM seb_kuzminsky: probably not
03:40 PM seb_kuzminsky: jogwheels dont go through task
03:41 PM seb_kuzminsky: some ui/nml tasks can get thrown off by Task sluggishness, but i think this test should be immune
03:42 PM dgarr: i guess i have a hard time caring about an armhf failure that doesnt occur in x86 machines
03:42 PM seb_kuzminsky: i'd care if it happened regularly, on any buildslave known to be generally stable and trustworthy
03:43 PM seb_kuzminsky: but in this case i'm happy to make a mental note of it and move on with life
07:29 PM linuxcnc-build_: build forced [ETA 2h02m51s]
07:29 PM linuxcnc-build_: I'll give a shout when the build finishes
09:50 PM linuxcnc-build_: build #427 of 1540.rip-jessie-armhf is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1540.rip-jessie-armhf/builds/427
09:55 PM dgarr: motion-logger/mountaindew: USRMOT: ERROR: command timeout
09:56 PM dgarr: motion/jogwheel-joint: exit 0
09:58 PM dgarr: using simple_tp planner refactored to current limit3
10:04 PM linuxcnc-build_: build #5242 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/5242