#linuxcnc-devel | Logs for 2016-07-11

Back
[06:53:58] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #99: Has been fixed already in April 2016 (11.04.2016),... 02https://github.com/LinuxCNC/linuxcnc/issues/99#issuecomment-231708794
[06:55:47] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #99: Has been fixed already in March 2016 (30.03.2016),... 02https://github.com/LinuxCNC/linuxcnc/issues/99#issuecomment-231708794
[06:55:50] <jepler> looks like some newer boost headers will suffice to make the lock free queue available on wheezy. hopefully other old OSes are similar... I'll take the cowards way and just copy them into our source tree...
[06:58:04] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #82: I fixed that with commit f102b9f118b52346c17a4c3f966327dcf0fc5127... 02https://github.com/LinuxCNC/linuxcnc/issues/82#issuecomment-231709519
[06:59:54] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler commented on issue #99: @gmoccapy just to confirm, the fix is in 2.7 branch? If so we'll close this issue when the next 2.7 release is made. 02https://github.com/LinuxCNC/linuxcnc/issues/99#issuecomment-231709811
[07:09:46] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #99: Hallo Jeff,
[07:11:36] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler commented on issue #82: We'll have to manually close this issue once the next 2.7 release is made. 02https://github.com/LinuxCNC/linuxcnc/issues/82#issuecomment-231711735
[07:12:37] <jepler> seb_kuzminsky: how can we help with making a release of 2.7 happen?
[09:21:58] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 0dfb884 06linuxcnc 10(40 files in 7 dirs) build: include a copy of boost lockfree for heritage platforms * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0dfb884
[09:21:58] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 164ed30 06linuxcnc 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: Introduce lockfree queue for rtapi_print_msg * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=164ed30
[09:21:58] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d1a7a61 06linuxcnc 10src/rtapi/rtapi_uspace.hh 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: rtapi_delay will need a different implementation for rtai * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d1a7a61
[09:22:02] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 9a9029c 06linuxcnc 10src/configure.in configure: fall back to uspace realtime if rtai not found * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9a9029c
[09:22:05] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d6f9282 06linuxcnc 10src/module_helper/module_helper.c module_helper: Allow loading modules associated with running kernel * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d6f9282
[09:22:10] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d5e22d0 06linuxcnc 10src/module_helper/module_helper.c module_helper: always allow .ko objects * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d5e22d0
[09:22:13] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 47b2481 06linuxcnc 10scripts/realtime.in realtime: Load modules even for uspace, if requested by rtapi.conf * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=47b2481
[09:22:17] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d2003ed 06linuxcnc 10src/Makefile build: even on uspace, make linuxcnc_module_helper setuid * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d2003ed
[09:22:22] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 07e9507 06linuxcnc 10src/Makefile build: install additional programs as setuid in uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=07e9507
[09:22:25] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 96a2044 06linuxcnc 10debian/configure packaging: drop special treatment of linuxcnc_module_helper * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=96a2044
[09:22:30] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus e7a4a0b 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: ensure a more orderly shutdown * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e7a4a0b
[09:22:33] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 409d3b3 06linuxcnc 10(7 files in 3 dirs) uspace: add uspace+rtai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=409d3b3
[09:22:37] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 2c174c0 06linuxcnc 10(6 files in 2 dirs) uspace: add uspace+xenomai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2c174c0
[09:22:41] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 1508a3b 06linuxcnc 10(8 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=1508a3b
[09:22:45] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 02d5b56 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=02d5b56
[09:25:33] <linuxcnc-build> build #4330 of 1306.rip-precise-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/4330 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[09:25:46] <linuxcnc-build> build #4328 of 1300.rip-precise-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/4328 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[09:26:19] <jepler> hmph.
[09:30:20] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 4dc488c 06linuxcnc 10(40 files in 7 dirs) build: include a copy of boost lockfree for heritage platforms * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4dc488c
[09:30:20] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 952e4fc 06linuxcnc 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: Introduce lockfree queue for rtapi_print_msg * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=952e4fc
[09:30:20] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 823ca27 06linuxcnc 10src/rtapi/rtapi_uspace.hh 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: rtapi_delay will need a different implementation for rtai * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=823ca27
[09:30:24] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 00c45a6 06linuxcnc 10src/configure.in configure: fall back to uspace realtime if rtai not found * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=00c45a6
[09:30:28] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 286ac31 06linuxcnc 10src/module_helper/module_helper.c module_helper: Allow loading modules associated with running kernel * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=286ac31
[09:30:31] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 5e07e5e 06linuxcnc 10src/module_helper/module_helper.c module_helper: always allow .ko objects * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5e07e5e
[09:30:36] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 4f83e4d 06linuxcnc 10scripts/realtime.in realtime: Load modules even for uspace, if requested by rtapi.conf * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4f83e4d
[09:30:40] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 9d02872 06linuxcnc 10src/Makefile build: even on uspace, make linuxcnc_module_helper setuid * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9d02872
[09:30:43] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 2a8dfee 06linuxcnc 10src/Makefile build: install additional programs as setuid in uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2a8dfee
[09:30:48] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 904f169 06linuxcnc 10debian/configure packaging: drop special treatment of linuxcnc_module_helper * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=904f169
[09:30:51] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 0807f20 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: ensure a more orderly shutdown * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0807f20
[09:30:56] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 1ca86d6 06linuxcnc 10(7 files in 3 dirs) uspace: add uspace+rtai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=1ca86d6
[09:30:59] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus f8b9388 06linuxcnc 10(6 files in 2 dirs) uspace: add uspace+xenomai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f8b9388
[09:31:03] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 2927cbf 06linuxcnc 10(8 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2927cbf
[09:31:07] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 0b5feec 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0b5feec
[09:56:30] <linuxcnc-build> build #4343 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4343 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:00:43] <jepler> > King of Swamp Castle: When I first came here, this was all swamp. Everyone said I was daft to build a castle on a swamp, but I built in all the same, just to show them. It sank into the swamp. So I built a second one. That sank into the swamp. So I built a third. That burned down, fell over, then sank into the swamp. But the fourth one stayed up. And that's what you're going to get, Lad, the stro
[10:00:49] <jepler> ngest castle in all of England.
[10:01:09] <jepler> buildbot is my swap and fire
[10:01:38] <jepler> swamp
[10:19:24] <seb_kuzminsky> the buildbot is our designated driver: http://assets.amuniversal.com/eb729c3025890132a011005056a9545d
[10:24:00] <linuxcnc-build> build #2001 of 1402.rip-wheezy-rtpreempt-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1402.rip-wheezy-rtpreempt-i386/builds/2001 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:24:10] <cradek> dummy, Jeff Epler
[10:24:25] <cradek> I wish it wasn't so mean
[10:27:40] <jepler> hmmm that's the same failure I was chasing saturday
[10:27:46] <jepler> I thought I'd fixed
[10:27:47] <jepler> it
[10:28:06] <jepler> it could be a sign the lockfree queue "backport" is not right
[10:32:24] <jepler> hmph, I have a jessie with rt-preempt (doesn't use the backport) and that test succeeds 256/256 times. and a wheezy (but with vanilla kernel) and the test succeeds 256/256 times.
[10:33:39] <cradek> I have a machine up that's running wheezy on the linuxcnc 3.4.55-4linuxcnc rtai kernel. can that help?
[10:36:11] <jepler> cradek: maybe. I don't know if the uspace+rtai build works there at all, so it would be an interesting data point
[10:36:30] <jepler> after building I'm just running that specific failed test a bunch of times: runtests tests/module-loading/rtapi-app-main-fails/{,,,}{,,,}{,,,}{,,,}
[10:37:45] <jepler> oh wait, now I have 4/512 fails so I can reproduce it
[10:38:50] <jepler> .. yay ?
[10:39:20] <cradek> yay!?
[10:42:00] <seb_kuzminsky> jepler: 2.7.5 is mostly done in a branch here
[10:42:02] <linuxcnc-build> build #2689 of 1404.rip-wheezy-rtpreempt-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2689 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:42:03] <linuxcnc-build> build #4344 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4344 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:42:13] <seb_kuzminsky> i wanted to wait a little bit after that hairy Task fix/breaking i did, to see what else needs fixing
[10:43:49] <seb_kuzminsky> the original commits were in early June, the one bugfix since then was July 1
[10:44:14] <seb_kuzminsky> so maybe i should muster my courage and just release it
[10:44:45] <seb_kuzminsky> i'm super grateful to Gene Heskett for finding and reporting the bug
[10:46:13] <seb_kuzminsky> uhm, so maybe tuesday (tomorrow) at the hackspace i'll do it
[10:53:55] <cradek> Running test: tests/module-loading/rtapi-app-main-fails
[10:53:55] <cradek> loadrt did not find the test component
[10:53:55] <cradek> *** tests/module-loading/rtapi-app-main-fails: FAIL: checkresult exited with 1
[10:54:29] <cradek> jepler: what am I missing? this is branch jepler/master/uspace-plus built from clean
[10:56:12] <jepler> cradek: isn't that the error from buildbot? so it seems generally to affect wheezy, :(
[10:56:50] <jepler> what's happening is, some rtapi_print messages are going missing...
[10:57:22] <jepler> they go through this new queue thing so they can get out of real-time without making syscalls...
[11:06:53] <cradek> ok, it's failing every time on wheezy+rtai
[11:08:23] <cradek> wait, http://paste.ubuntu.com/19086935/
[11:09:27] <seb_kuzminsky> heh
[11:09:35] <seb_kuzminsky> the module failed to load, but not like it should
[11:09:49] <cradek> first (real?) failure: http://paste.ubuntu.com/19087032/
[11:09:55] <seb_kuzminsky> is that a negative false positive?
[11:10:01] <cradek> so confused
[11:10:41] <cradek> jepler: if I'm not helping, just tell me to stop
[11:14:53] <jepler> I don't even understand that second failure...
[11:15:20] <cradek> that's after I ran /etc/init.d/realtime stop manually to unload the stuck rtai modules
[11:15:47] <jepler> either of the failures you pasted in fact
[11:15:50] <cradek> the failure leaves 3 rtai modules loaded
[11:17:27] <jepler> there are at least two problems then. maybe stop helping for now...
[11:17:51] <cradek> ok
[11:18:32] <jepler> thanks though!
[11:18:37] <cradek> welcome
[12:13:40] <jepler> ISTM that the original failure (not seeing the specific load-failed error message) indicates that the backported boost stuff is not right.
[12:22:49] <jepler> no, it's not just the backport. I get eventual failures on my jessie system too, just took more than 4096 tries
[12:23:37] <seb_kuzminsky> i'm amused at your {,,,}
[12:24:00] <seb_kuzminsky> so something good has come out of all your hard work
[12:24:03] <jepler> it's not an elegant way to say it, but it works and it has become a habit
[12:24:22] <seb_kuzminsky> it's a lot less typing that writing a for loop in bash
[12:24:38] <jepler> yeah but it destroys the evidence
[12:24:50] <jepler> need to add an option: runtests --ttf (test-to-failure)
[12:25:05] <jepler> or something
[12:25:15] <jepler> or just stop-at-first-failure
[12:25:49] <jepler> sigh I hate 1-in-100-or-less bugs
[12:36:21] <jepler> "pthreads is stranger than we can imagine"
[12:37:26] <jepler> .. I have found an additional problem with my use of pthreads that would explain this test failure
[12:37:46] <jepler> I was unaware of the fact that a thread could get canceled in the middle of a printf(3) / write(2)
[12:38:12] <jepler> so the behavior is: message gets pulled off queue, write(2) syscall starts, thread gets canceled -> message is lost
[12:38:55] <jepler> I'll let this run to 10,000 iterations if it doesn't fail first
[12:39:01] <jepler> or one lunch of iterations
[13:27:51] <jepler> yay at 25k+ iterations no failures
[13:29:01] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 53f2237 06linuxcnc 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: Introduce lockfree queue for rtapi_print_msg * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=53f2237
[13:29:01] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d911278 06linuxcnc 10src/rtapi/rtapi_uspace.hh 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: rtapi_delay will need a different implementation for rtai * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d911278
[13:29:01] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 814a067 06linuxcnc 10src/configure.in configure: fall back to uspace realtime if rtai not found * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=814a067
[13:29:04] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 78312cd 06linuxcnc 10src/module_helper/module_helper.c module_helper: Allow loading modules associated with running kernel * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=78312cd
[13:29:09] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 22d77d7 06linuxcnc 10src/module_helper/module_helper.c module_helper: always allow .ko objects * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=22d77d7
[13:29:12] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus b26ef10 06linuxcnc 10scripts/realtime.in realtime: Load modules even for uspace, if requested by rtapi.conf * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b26ef10
[13:29:17] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus a75dacb 06linuxcnc 10src/Makefile build: even on uspace, make linuxcnc_module_helper setuid * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a75dacb
[13:29:21] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 36bc7c9 06linuxcnc 10src/Makefile build: install additional programs as setuid in uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=36bc7c9
[13:29:25] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 9586e57 06linuxcnc 10debian/configure packaging: drop special treatment of linuxcnc_module_helper * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9586e57
[13:29:28] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus a6a81a0 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: ensure a more orderly shutdown * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a6a81a0
[13:29:33] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 4cb4d92 06linuxcnc 10(7 files in 3 dirs) uspace: add uspace+rtai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4cb4d92
[13:29:36] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus cfa5617 06linuxcnc 10(6 files in 2 dirs) uspace: add uspace+xenomai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cfa5617
[13:29:40] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus c0261db 06linuxcnc 10(8 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c0261db
[13:29:44] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus e35b7a7 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e35b7a7
[13:31:18] <jepler> at some point, we'll have the strongest castle in all these lands
[13:34:29] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler closed issue #105: compiling against libmodbus version 3.1.0 or greater fails due to API changes 02https://github.com/LinuxCNC/linuxcnc/issues/105
[13:37:37] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler closed issue #95: Axis GUI: spurious Teleop commands when idle 02https://github.com/LinuxCNC/linuxcnc/issues/95
[13:38:01] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler commented on issue #94: ("calibrate" is inherited from tklinuxcnc and is a standlone Tk program) 02https://github.com/LinuxCNC/linuxcnc/issues/94#issuecomment-229324292
[13:39:21] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler closed pull request #102: rs274: work around boost::python bug (06master...06jepler/2.7/boost-workaround) 02https://github.com/LinuxCNC/linuxcnc/pull/102
[13:40:55] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler closed pull request #92: halshow.tcl: use no special grouping for signals (06master...062.6-halshow_signals) 02https://github.com/LinuxCNC/linuxcnc/pull/92
[14:00:19] <jepler> http://imgur.com/DrEinPB
[14:00:29] <jepler> (discussion at https://news.ycombinator.com/item?id=12071240)
[14:10:03] <jepler> seb_kuzminsky: how does buildbot decide what args to invoke debian/configure with?
[14:10:41] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 34872bd 06linuxcnc 10(9 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=34872bd
[14:10:42] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 895affa 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=895affa
[14:21:02] <jepler> seb_kuzminsky: hm I'm worried I wedged a builder :-/ http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2690/steps/runtests/logs/stdio
[14:31:41] <linuxcnc-build> build #2690 of 1404.rip-wheezy-rtpreempt-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2690 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[14:31:42] <linuxcnc-build> build #4345 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4345 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[14:32:02] <jepler> maybe it's sorted itself out, wonder what'll happe next build
[14:32:16] <jepler> needless to see I haven't seen a hang like that in 20,000+ halruns with my branch locally
[14:32:57] <jepler> [267742.880012] INFO: rcu_preempt detected stall on CPU 1 (t=3122070 jiffies)
[14:33:25] <cradek> does that mean kernel bug?
[14:33:54] <jepler> mmmmaybe
[14:34:53] <jepler> http://stackoverflow.com/questions/35401317/rcu-preempt-self-detected-stall-on-cpu-0
[14:35:07] <jepler> or it could maybe happen if a RT task is not yielding the CPU
[14:35:16] <jepler> > ... A CPU-bound real-time task in a CONFIG_PREEMPT_RT kernel that is running at a higher priority than the RCU softirq threads. This will prevent RCU callbacks from ever being invoked, and in a CONFIG_PREEMPT_RCU kernel will further prevent RCU grace periods from ever completing. Either way, the system will eventually run out of memory and hang.
[14:37:06] <jepler> https://www.kernel.org/doc/Documentation/RCU/stallwarn.txt
[14:58:41] <cradek> I expect stallwarn.txt to be about software patents or cell phone surveillance or something
[15:08:20] <jepler> indeed
[15:24:14] <jepler> now that same builder seems to be stalled during the main build
[15:24:26] <jepler> last line of output has been the same for several minutes
[16:08:37] <seb_kuzminsky> i'm back
[16:08:39] <seb_kuzminsky> err
[16:09:16] <seb_kuzminsky> i'll bounce it
[16:11:15] <seb_kuzminsky> it's back
[16:12:15] <seb_kuzminsky> jepler: the debian/configure args are specified by the buildmaster config, on a per-builder basis
[16:12:25] <seb_kuzminsky> http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2690
[16:13:04] <seb_kuzminsky> in the Build Properties table it shows the property name ("debian/configure argument"), the value ("uspace") and the Source of the property ("Builder")
[16:14:28] <seb_kuzminsky> this is the best pic i can find in 60 seconds showing how "Builder" fits into the buildbot architecture: http://docs.buildbot.net/0.8.3/Buildmaster-Architecture.html#Buildmaster-Architecture
[16:18:12] <jepler> seb_kuzminsky: OK, so no convenient way to test my branch with different debian/configure arguments except I guess to hard-code it
[16:18:49] <seb_kuzminsky> yeah not currently
[16:19:45] <seb_kuzminsky> i can imagine adding something like scripts/platform-is-supported, a script the buildbot could look for and run if it's there to override the default in the config
[16:42:08] <jepler> seb_kuzminsky: is buildbot still confused? the 0000.checkin build seems to still be spinning, but no other build is taking place.
[16:42:46] <jepler> trying to interrupt it..
[16:43:27] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 837a49f 06linuxcnc 10(9 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=837a49f
[16:43:27] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus f951891 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f951891
[16:45:56] <jepler> buildbot "successfully configured for 'uspace-Debian-8.2'-'uspace+rtai+xenomai'.. "
[16:46:30] <jepler> hmm but without including a rtai-modules package in the build-depends, whoops
[16:51:00] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus d2f82b6 06linuxcnc 10(9 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d2f82b6
[16:51:00] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus fda24b3 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=fda24b3
[16:51:34] <jepler> nope still not right
[16:51:39] <jepler> $ apt-cache policy ""
[16:51:39] <jepler> N: Unable to locate package
[16:51:43] <jepler> $ echo $?
[16:51:43] <jepler> 0
[16:54:08] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus c3d2078 06linuxcnc 10(9 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c3d2078
[16:54:08] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 49f3f16 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=49f3f16
[16:57:52] <seb_kuzminsky> jepler: looks like 0000.checkin didnt like it when the wheezy-rtpreempt-amd64 buildslave stalled and was powercycled, thanks for resetting it
[17:22:21] <linuxcnc-build> build #2527 of 1405.rip-wheezy-armhf is complete: Failure [4failed install-missing-build-dependencies compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/2527 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[17:22:21] <linuxcnc-build> build #4347 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4347 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[17:39:28] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 50f41e5 06linuxcnc 10(9 files in 2 dirs) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=50f41e5
[17:39:28] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus a83a620 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a83a620
[17:55:13] <linuxcnc-build> build #2528 of 1405.rip-wheezy-armhf is complete: Failure [4failed install-missing-build-dependencies compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/2528 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[17:55:13] <linuxcnc-build> build #4348 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4348 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[18:22:04] <jepler> seb_kuzminsky: I've been trying to get the uspace builds from my branch to automatically pick up whether to build rtai or xenomai support based on info from apt-cache, but now I see two things: first, <distro>/base isn't in source.list (based on searching in http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/2495/steps/apt-get-update/logs/stdio) and anyway the metadata might
[18:22:10] <jepler> not be there yet since apt-get update is after configure debian/control
[18:32:19] <seb_kuzminsky> jepler: you want to see if a xenomai kernel is available in the apt sources, but you don't care what's currently running on the build host?
[18:34:15] <seb_kuzminsky> i think it's likely i left the 'linuxcnc.org $DIST base' out of sources.list, since it was not needed at the time i set the wheezy machines up
[18:34:19] <seb_kuzminsky> that's probably a bug on my part
[18:38:05] <jepler> seb_kuzminsky: rtai, actually. xenomai development files are available in the base system on x86 even if no kernels are
[18:38:14] <jepler> (xenomai have a stable ABI while rtai is :shrug: about ABI)
[18:38:43] <jepler> don't change anything right now, we can work it out later
[18:55:24] <skunkworks> jepler, is it the userspace or kernel space version of xenomi?
[19:01:19] <linuxcnc-build> build #482 of 4017.5.deb-wheezy-armhf is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.5.deb-wheezy-armhf/builds/482 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[19:06:11] <linuxcnc-build> build #3492 of 4007.deb-precise-i386 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4007.deb-precise-i386/builds/3492 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[19:13:33] <jepler> hm
[19:14:24] <jepler> I guess it uses the same .dscs for i386, amd64, and armhf builds (how can it not) so it ends up requiring libxenomai-dev on armhf
[19:14:41] <jepler> I guess I can statically qualify that as only being on i386 and amd64 platforms, until someone says otherwise
[19:15:12] <linuxcnc-build> build #2320 of 4009.deb-precise-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4009.deb-precise-rtai-i386/builds/2320 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[19:35:28] <linuxcnc-build> build #3492 of 4008.deb-precise-amd64 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4008.deb-precise-amd64/builds/3492 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[19:50:04] <skunkworks> jepler, great work!
[20:21:46] <linuxcnc-build> build #2060 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/2060 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[20:22:29] <linuxcnc-build> build #2064 of 4017.deb-wheezy-amd64 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-amd64/builds/2064 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[20:35:57] <linuxcnc-build> build #1345 of 4014.deb-wheezy-rtpreempt-i386 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4014.deb-wheezy-rtpreempt-i386/builds/1345 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[20:47:48] <jepler> skunkworks: oops I missed that you asked me a question earlier. it is the userspace API of xenomai, specifically their "posix skin".
[20:48:39] <skunkworks> is the LXRT better supported for more kernels than rtai?
[20:48:45] <skunkworks> (userspace)
[20:48:52] <skunkworks> jepler, thanks
[20:49:18] <skunkworks> (I may be asking the wrong question..)
[20:50:39] <linuxcnc-build> build #1380 of 4015.deb-wheezy-rtpreempt-amd64 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4015.deb-wheezy-rtpreempt-amd64/builds/1380 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[20:52:45] <jepler> skunkworks: lxrt is the name of the userspace API of rtai. I think it is optional, so not all rtai kernels may have it.
[20:53:44] <skunkworks> ah.. so it is on top of rtai?
[20:56:38] <jepler> right. with this branch, the same build of uspace can work in 4 different ways: no specific kernel (simulator mode), any preempt-rt kernel, any properly-configured xenomai kernel, and any properly-configured rtai kernel
[20:57:05] <jepler> in the case of xenomai and rtai it is not tied to a single kernel version number, it's a matter of that kernel providing an ABI to userspace programs (posix-skin or lxrt respectively)
[20:57:53] <skunkworks> neat
[20:58:52] <Tom_itx> latency vary between versions?
[21:00:03] <jepler> Tom_itx: yes, each different kernel has different latency performance.
[21:00:39] <Tom_itx> so as before any certain hardware may like one over another
[21:01:10] <jepler> right
[21:02:11] <Tom_itx> sounds pretty neat
[21:02:11] <jepler> but the ability to support any of those with the same build of linuxcnc should be a big improvement
[21:02:17] <Tom_itx> yeah
[21:02:27] <linuxcnc-build> build #729 of 4022.deb-jessie-amd64 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4022.deb-jessie-amd64/builds/729 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[21:02:40] <jepler> to be fair, the machinekit people had the same idea first, what with wanting to support different ARM computers all with different kernels
[21:02:42] <linuxcnc-build> build #1752 of 4018.deb-wheezy-rtai-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4018.deb-wheezy-rtai-i386/builds/1752 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[21:03:43] <linuxcnc-build> build #645 of 4019.deb-jessie-rtpreempt-i386 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4019.deb-jessie-rtpreempt-i386/builds/645 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[21:06:10] <linuxcnc-build> build #731 of 4021.deb-jessie-i386 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4021.deb-jessie-i386/builds/731 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[21:07:29] <linuxcnc-build> build #644 of 4020.deb-jessie-rtpreempt-amd64 is complete: Failure [4failed shell_3 shell_4] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4020.deb-jessie-rtpreempt-amd64/builds/644 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[21:25:22] <jepler> it's nice that the congratulations can begin at the point where I've broken *ALL* the package builds (on a branch, but still)
[21:30:58] <skunkworks> but still ;)
[21:31:10] <skunkworks> You have always figured it out
[21:33:09] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus a792bba 06linuxcnc 10(7 files in 3 dirs) uspace: add uspace+rtai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a792bba
[21:33:09] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus e14fc74 06linuxcnc 10(6 files in 2 dirs) uspace: add uspace+xenomai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e14fc74
[21:33:09] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus ebac00b 06linuxcnc 10(8 files) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ebac00b
[21:33:11] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus fa8d52c 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=fa8d52c
[21:42:09] <jepler> this is what, castle #18? the last one got blown to bits by sappers from the kingdom of lintian...
[22:08:32] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15jepler commented on issue #108: getting close to building on all platforms in our buildbot, though because of technical limitations that doesn't include rtai support now. 02https://github.com/LinuxCNC/linuxcnc/pull/108#issuecomment-231922744
[22:10:21] <linuxcnc-build> build #2695 of 1404.rip-wheezy-rtpreempt-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2695 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[22:10:21] <linuxcnc-build> build #4350 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4350 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[22:22:56] <jepler> > command timed out: 1200 seconds without output, attempting to kill
[22:22:57] <jepler> argh
[22:23:07] <jepler> [20427.404021] INFO: rcu_preempt detected stall on CPU 1 (t=15001 jiffies)
[22:23:10] <jepler> more of this
[22:24:48] <jepler> seb_kuzminsky: I fear you'll have to bounce that machine again. but before you do, can you see if there are spinning processes (rtapi_apps maybe)?
[22:25:10] <seb_kuzminsky> checking...
[22:27:00] <seb_kuzminsky> "top" just sits at the terminal, doesn't redraw the screen
[22:27:12] <seb_kuzminsky> actually it locked the console up
[22:27:17] <jepler> too bad
[22:27:25] <seb_kuzminsky> Alt-F2 switched consoles though...
[22:27:40] <jepler> this a new failure mode right?
[22:27:44] <seb_kuzminsky> new to me, yeah
[22:27:54] <seb_kuzminsky> i've seen similar things while debugging rtai
[22:27:59] <seb_kuzminsky> including the rcu stall
[22:28:14] <seb_kuzminsky> i can log in on the other console and ps gives some output before hanging
[22:28:21] <seb_kuzminsky> does not get back to the shell prompt
[22:29:34] * seb_kuzminsky pulls the plug
[22:29:58] <seb_kuzminsky> you know...
[22:30:14] <seb_kuzminsky> both failures today were on wheezy-rtpreempt-amd64
[22:30:22] <seb_kuzminsky> none on wheezy-rtpreempt-i386
[22:30:55] <seb_kuzminsky> well no, they both have just 1 cpu
[22:31:24] <seb_kuzminsky> i got different baffling results on rtai when running with more CPUs, i'm going to try bumping wheezy-rtpreempt-amd64 to 2 vcpus
[22:31:42] <jepler> the main difference that affects preempt-rt (and vanilla for that matter) is this new system of sending rtapi_print messages out through this lockfree queue
[22:32:05] <jepler> hm my wheezy test system has 12 threads :-P
[22:32:44] <jepler> but that new thread doesn't run with elevated priority of any kind
[22:33:18] <seb_kuzminsky> the 19th one got infested with alligators
[22:34:43] <seb_kuzminsky> i just rebuild the rtai kernel with all lock debugging options turned on
[22:34:54] <jepler> linuxcnc-build: force build --branch=jepler/master/uspace-plus 0000-checkin
[22:34:54] <linuxcnc-build> no such builder '0000-checkin'
[22:35:09] <jepler> linuxcnc-build: force build --branch=jepler/master/uspace-plus 0000.checkin
[22:35:10] <linuxcnc-build> build forced [ETA 1h36m29s]
[22:35:10] <linuxcnc-build> I'll give a shout when the build finishes
[22:36:20] <jepler> that builder looks like it had two (v)CPUs already
[22:36:25] <jepler> [21328.024081] Pid: 7497, comm: rtapi_app Not tainted 3.2.0-4-rt-amd64 #1 Debian 3.2.81-1 Bochs Bochs
[22:36:46] <jepler> [20967.776304] Pid: 3, comm: ksoftirqd/0 Not tainted 3.2.0-4-rt-amd64 #1 Debian 3.2.81-1 Bochs Bochs
[22:36:49] <jepler> [21147.900379] Pid: 0, comm: swapper/0 Not tainted 3.2.0-4-rt-amd64 #1 Debian 3.2.81-1 Bochs Bochs
[22:36:58] <jepler> mostly, comm is rtapi_app
[22:37:26] <seb_kuzminsky> hmm, yes
[22:37:31] <seb_kuzminsky> http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/2695/steps/environment-report/logs/stdio
[22:38:01] <seb_kuzminsky> it disturbs me that nproc in the guest disagrees with libvirt's theory of the guest's cpu count
[22:38:26] <jepler> I wonder if the message-printing task shold simply wait long between polling
[22:38:34] <jepler> it is going every 10us which is actually frightfully often
[22:38:43] <jepler> is that right?
[22:38:56] <jepler> + struct timespec ts = {0, 10000};
[22:38:57] <jepler> + rtapi_clock_nanosleep(CLOCK_MONOTONIC, 0, &ts, NULL, NULL);
[22:39:17] <seb_kuzminsky> 10,000 nanoseconds, yep
[22:41:20] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 36fe278 06linuxcnc 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: Introduce lockfree queue for rtapi_print_msg * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=36fe278
[22:41:21] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus ffd7a41 06linuxcnc 10src/rtapi/rtapi_uspace.hh 10src/rtapi/uspace_common.h 10src/rtapi/uspace_rtapi_app.cc 10src/rtapi/uspace_ulapi.c uspace: rtapi_delay will need a different implementation for rtai * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ffd7a41
[22:41:21] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 98ee7f1 06linuxcnc 10src/configure.in configure: fall back to uspace realtime if rtai not found * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=98ee7f1
[22:41:24] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus ec63b08 06linuxcnc 10src/module_helper/module_helper.c module_helper: Allow loading modules associated with running kernel * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ec63b08
[22:41:28] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus c4b87c7 06linuxcnc 10src/module_helper/module_helper.c module_helper: always allow .ko objects * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c4b87c7
[22:41:32] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus be528b7 06linuxcnc 10scripts/realtime.in realtime: Load modules even for uspace, if requested by rtapi.conf * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=be528b7
[22:41:36] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus ff3fbb3 06linuxcnc 10src/Makefile build: even on uspace, make linuxcnc_module_helper setuid * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ff3fbb3
[22:41:40] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 30df045 06linuxcnc 10src/Makefile build: install additional programs as setuid in uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=30df045
[22:41:44] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 05a0efa 06linuxcnc 10debian/configure packaging: drop special treatment of linuxcnc_module_helper * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=05a0efa
[22:41:48] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 25a055c 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: ensure a more orderly shutdown * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=25a055c
[22:41:52] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus f66f0ee 06linuxcnc 10(7 files in 3 dirs) uspace: add uspace+rtai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f66f0ee
[22:41:56] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus c3ce4ec 06linuxcnc 10(6 files in 2 dirs) uspace: add uspace+xenomai realtime * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c3ce4ec
[22:42:00] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus ccc4512 06linuxcnc 10(8 files) packaging: rtai, xenomai are sub-packages of uspace * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ccc4512
[22:42:03] <KGB-linuxcnc> 03Jeff Epler 05jepler/master/uspace-plus 5b0e05e 06linuxcnc 10docs/src/code/building-linuxcnc.txt docs: document new RTOS support * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5b0e05e
[22:42:38] <jepler> it disturbs me if simply trying to run a *non-rt* task even at 100kHz even in a VM could make something go wonky
[22:43:06] <jepler> but it's not yet clear whether that's the right theory anyway
[22:43:33] <seb_kuzminsky> it's extra disturbing if a single thread on a multi-cpu VM can make it crash
[22:45:53] <jepler> what's the virtualization software, qemu with kvm support?
[22:46:12] <jepler> I see those messages say "bochs" but that's seriously old stuff
[22:48:31] <jepler> 'night, its late here
[22:50:12] <seb_kuzminsky> yeah, qemu/kvm
[22:50:21] <seb_kuzminsky> the wheezy VMs are running on a jessie hypervisor
[23:11:39] <seb_kuzminsky> i guess it's running the bochs bios
[23:41:41] <KGB-linuxcnc> 03Dewey Garrett 05master d95374e 06linuxcnc 10src/emc/motion/control.c control.c teleop wheel jog: no feed override JA * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d95374e
[23:41:41] <KGB-linuxcnc> 03Dewey Garrett 05master 263c646 06linuxcnc 10src/emc/usr_intf/axis/scripts/axis.py axis.py: allow feedoverride adjust always JA * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=263c646
[23:47:14] <KimK> From the "Uh-Oh" department? (via Distrowatch): "Lately we have been seeing more Linux distributions dropping support for 32-bit x86 architectures. ..."--Distrowatch news item "...making a case for phasing [out] our 32-bit support slowly over the next two and a half years. ..."--Dimtri John Ledkov, Ubuntu developer Distrowatch news item: http://distrowatch.com/weekly.php?issue=20160711#news Ubuntu dev thread:
[23:47:15] <KimK> https://lists.ubuntu.com/archives/ubuntu-devel/2016-June/039420.html
[23:47:19] <KimK> Oops
[23:52:34] <KimK> Flood, sorry. I thought HexChat would have had a character-counter by now. (I'm in the process of upgrading from 10.04 to Mint 18). I had a character-counter script running on Xchat/10.04 but probably won't bother on HexChat/Mint 18.
[23:53:22] <seb_kuzminsky> KimK: come over to debian stable, the water is fine!