#linuxcnc-devel Logs

Feb 25 2017

07:55 AM stustev: guys - thanks - I had put the path in my .bashrc file - removing it allows everything to work as expected
08:27 AM Tom_itx: he shoots.. he scores!
08:51 AM stustev: nutin but net
10:48 AM KGB-linuxcnc: 03andypugh 05master cb9dc75 06linuxcnc 10scripts/update_ini update_ini: Fix issues #230 and #236 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cb9dc75
11:12 AM pcw_home: rtapi_shmem_new failed due to shmget(key=0x130cf406): Permission denied
11:12 AM pcw_home: SCOPE: ERROR: failed to get shared memory (key=0x130cf406, size=264)
11:12 AM pcw_home: I get this error on master and Mint18/Ubuntu 16.04 (not on Mint17/Ubuntu 14.04)
11:12 AM pcw_home: any ideas?
11:13 AM cradek: the udev rule for /dev/rtai-whatever?
11:27 AM pcw_home: 50-LINUXCNC-shmdrv.rules = KERNEL=="shmdrv", MODE="0666"
01:09 PM seb_kuzm1nsky: pcw_home: what does `ipcs` say? do you have an old shm segment with the wrong perms lying around?
01:10 PM seb_kuzm1nsky: i'm repeating a project i did a while ago, and this time i'm refactoring/rewriting some of the g-code i used last time
01:11 PM seb_kuzm1nsky: i found it handy to run rs274 to produce the canon commands from before and after the rewrite, to see if i introduced any inuntended changes
01:14 PM seb_kuzm1nsky is now known as seb_kuzminsky
02:05 PM pcw_home: ipcs shows many shared memory segments not sure what to make of them
02:24 PM seb_kuzminsky: pcw_home: i think, look for a shared memory segment with key 0x48484c34
02:24 PM seb_kuzminsky: or just try "halrun -U" to force-unload whatever linuxcnc has loaded
02:26 PM pcw_home: there is no such key unless linuxcnc is running
02:27 PM seb_kuzminsky: right
02:27 PM seb_kuzminsky: hmm
02:27 PM seb_kuzminsky: how are you starting linuxcnc and halscope?
02:27 PM pcw_home: when running:
02:27 PM pcw_home: 0x48484c34 21692475 pcw 600 8 4 locked
02:28 PM pcw_home: the chooser/ then HALScope from axis
02:31 PM seb_kuzminsky: pcw_home: the key it can't access is the SCOPE_SHM_KEY frmo halscope
02:32 PM seb_kuzminsky: it's the non-realtime part of halscope that's failing, not the realtime part
02:32 PM seb_kuzminsky: are you on rtai or uspace?
02:38 PM seb_kuzminsky: huh, halscope's scope_rt.c calls rtapi_shmem_new(), but rtapi_shmem_new's manpage says it's not allowed from realtime tasks
02:38 PM seb_kuzminsky: i guess the initialization part is ok, just not a realtime thread
02:41 PM seb_kuzminsky: pcw_home: i see that you are on uspace, ignore that question
02:41 PM KGB-linuxcnc: 03Dewey Garrett 05dgarr/external_offsets c8745c0 06linuxcnc 10(43 files in 13 dirs) External Axis Offset hal pins * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c8745c0
02:41 PM pcw_home: uspace (works with 14.04, not 16.04)
02:42 PM seb_kuzminsky: is 16.04 xenial?
02:42 PM seb_kuzminsky: yes it is
02:43 PM seb_kuzminsky: i've got a xenial vm here, i'll try on it
03:23 PM seb_kuzminsky: axis won't run because of a missing dependency on pango
03:25 PM seb_kuzminsky: it's part of python-gtk2 on stretch
03:25 PM seb_kuzminsky: oh, which is a Dependency of linuxcnc.deb, but not checked for by our configure script
03:25 PM seb_kuzminsky: that's fine.jpg
03:26 PM seb_kuzminsky: pcw_home: it works for me
03:26 PM seb_kuzminsky: i built master for uspace on 16.04, ran sim/axis/axis, and started halscope via the Machine menu
03:29 PM seb_kuzminsky: pcw_home: http://imgur.com/a/zbKZs
03:29 PM seb_kuzminsky: i wonder if there's a race between scope.c and scope_rt.c both creating the scope shm?
03:30 PM seb_kuzminsky: in scope.c, after it shells out to halcmd to loadrt scope, put in a 2-second sleep, and see if that changes anything
03:30 PM seb_kuzminsky: on this VM i don't have a special udev rule for shm, by the way
03:55 PM seb_kuzminsky: who knows of a good value insert facemill?
04:36 PM pcw_home: Yeah the udev rule is old
04:36 PM pcw_home: the delay didn't make a difference (other that delaying the error message)
04:45 PM cradek: seb_kuzminsky: not many of them work for aluminum. be sure it takes high rake aluminum inserts
05:36 PM seb_kuzminsky: cradek: thanks for the warning!
05:38 PM seb_kuzminsky: pcw_home: that's so weird
05:39 PM seb_kuzminsky: after scope loadrts the realtime half, can you see the scope shm key in ipcs? what are its permissions? Can you see the scope rt component loaded and ready?
05:40 PM seb_kuzminsky: (make the delay really long, like a minute, and you can poke around in a separate shell during the interim)
05:40 PM seb_kuzminsky: cradek: my naive mind wants a facemill that shares inserts with my lathe tooling
05:42 PM seb_kuzminsky: the lathe tools use TCxx or TPxx inserts (triangular, 7 or 11 degree relief angle)
05:44 PM seb_kuzminsky: i'm imagnining something like this, but with a cylindrical shank: http://www.shars.com/products/indexable-cutting/indexable-milling/r8-1-1-4-1-1-2-2-carbide-indexable-end-mill
05:45 PM seb_kuzminsky: shars sells them with a 1" head on a cylindrical shank
05:46 PM seb_kuzminsky: ooh, this might work, with modification: http://www.shars.com/products/indexable-cutting/indexable-milling/30nmtb-shank-2-carbide-indexable-face-mill
05:46 PM seb_kuzminsky: cradek: didn't you pull some caper with a bench grinder on the table of your mill? ;-)
05:49 PM JT-Shop: my face mill uses a semm insert and my larger end mills use an apet insert both polished with a sharp edge for alum
05:50 PM seb_kuzminsky: do they work well?
05:51 PM JT-Shop: yea
05:52 PM JT-Shop: I have one box of tmng inserts for alum on the lathe but the face mills don't seem to use them
05:53 PM JT-Shop: yea all my face mills use the 4 edge square insert
07:06 PM seb_kuzminsky: JT-Shop: make & model?