#linuxcnc-devel | Logs for 2013-12-08

Back
[12:20:10] <KGB-linuxcnc> 03Norbert Schechner 05master 68b13e1 06linuxcnc 10(9 files in 7 dirs) gmoccapy_0_9_9_5 - introduced color chooser for homed color * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=68b13e1
[13:33:45] <KGB-linuxcnc> 03Sebastian Kuzminsky 05dgarr/config-cleanup d23258d 06linuxcnc 10docs/src/Master_Getting_Started_es.txt 10docs/src/index_es.tmpl docs: fix Spanish master doc to use translated docs * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d23258d
[13:33:45] <KGB-linuxcnc> 03Sebastian Kuzminsky 05dgarr/config-cleanup fb1af58 06linuxcnc 10(7 files in 2 dirs) docs: rebrand some page names * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=fb1af58
[13:33:45] <KGB-linuxcnc> 03Sebastian Kuzminsky 05dgarr/config-cleanup 4966a21 06linuxcnc 10docs/src/Submakefile docs build: check all english and french docs for broken links * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4966a21
[13:33:48] <KGB-linuxcnc> 03Sebastian Kuzminsky 05dgarr/config-cleanup 8e24e72 06linuxcnc 10docs/src/Submakefile docs: check links in spanish docs, but only warn on failure * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=8e24e72
[13:33:52] <KGB-linuxcnc> 03Norbert Schechner 05dgarr/config-cleanup 21ebdbd 06linuxcnc 10(9 files in 7 dirs) gmoccapy_0_9_9_5 - introduced color chooser for homed color * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=21ebdbd
[13:33:56] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/config-cleanup f94595c 06linuxcnc 10tcl/bin/pickconfig.tcl pickconfig: support flat or hierachical user dir * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f94595c
[13:48:15] <andypugh> Is there a matrix anywhere of what IO hardware works with what RT system? My impression, possibly wrong, is that the 7i80 ethernet card only works with Xenomai and RT-PREEMPT user-space realtime, and the PCI cards only work with RTAI or Xenomai Kernel-space realtime?
[17:41:52] <memleak> if the 7i80 works with RT net it will work with RTAI, however depending on the programming of linuxcnc, it might not work with that.
[17:44:18] <memleak> ("that" meaning linuxcnc) on the RT side everything will function properly with RTAI, especially if it works with xenomai
[17:44:33] <memleak> i believe xenomai shares the same RT net code with RTAI but not 100% certain
[17:53:07] <micges> rtnet is based on rtdm interface which is in both RTAI and xenomai
[17:53:09] <andypugh> memleak: The new branch just created doesn't use rtnet
[17:53:52] <andypugh> http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=commit;h=60e91dca489d6360d18ab4c887c5f2482b437e69
[18:02:44] <memleak> mesanet is an internal library of linuxcnc?
[18:02:56] <memleak> if so then the RT subsystem should work completely independant from it
[18:03:24] <memleak> havent looked at the source though to verify this.
[18:03:29] <memleak> (of mesanet i mean)
[18:04:06] <micges> mesanet is manfacturer of 7i80
[18:06:39] <memleak> ah. well this comment here in the commit message "this driver don't use rtnet, but standard system sockets" implies that regardless of RT system it should work fine, considering there is no changes of the core networking infrastructure within RTAI, preempt-rt and xenomai
[18:07:27] <memleak> all the system socket code stays the same between RT systems so i'd imagine it would all work ok.
[18:08:13] <micges> that's right
[18:10:09] <memleak> :)
[18:10:23] <andypugh> I guess that kernel threads can't access the standard sockets, though?
[18:10:32] <cradek> yarrrrghhhghghhh
[18:11:06] <cradek> I just spent the day massaging hardware and installing xubuntu12 to try memleak's kernel and seb's linuxcnc master build
[18:11:15] <cradek> but I installed the amd64 version
[18:11:38] <andypugh> That's bad?
[18:11:51] <memleak> i never tested the debs, only 64-bit code compiled myself
[18:12:02] <cradek> the rtai deb is i386
[18:12:08] <memleak> ah..
[18:12:21] <memleak> RTAI needs to be compiled for 64-bit if 64-bit distro
[18:12:40] <memleak> its not like firefox ;)
[18:12:41] <cradek> no kidding
[18:13:23] <cradek> the linuxcnc debs are also i386
[18:13:32] <andypugh> Still baffled by the lcd component in rt-preempt, insmod returns -1, despite the rtapi_app_main() code in lcd running through with no error and returning a 0.
[18:13:51] <memleak> linuxcnc you might be able to get by running 32 on 64 but RTAI most definitely not.
[18:14:13] <memleak> RTAI hard codes a lot of things specific to whether if 32 or 64-bit
[18:14:18] <cradek> yeah that is certainly not a goal of mine
[18:14:37] <andypugh> No compiler errors (now), no error messages anywhere, in fact. Odd.
[18:15:37] <andypugh> It even gets added to a thread with no real bother, and appears to take a variable thread time.
[18:16:31] <andypugh> And the output pin changes as expected too.
[18:23:44] <andypugh> And, in fact, other than the fact that the insmod error abouts the hal file load, it all works perfectly.
[18:25:18] <andypugh> micges: So, we have ethernet to 8i20 to parallel cable to 7i76 to smart-serial to 7i73 to a ribbon cable to an LCD display, all seeming to work.
[18:25:32] <andypugh> Sorry, not 8i20. 7i*0
[18:25:40] <andypugh> 7i80 eve
[18:25:42] <andypugh> n
[18:25:48] <andypugh> (Bed time!)
[18:26:29] <micges> andypugh: that's great, thnaks for test
[18:26:53] <micges> if you can leave it running with sserial for few hours
[18:27:12] <andypugh> An 8i20 would be a good test, it would indicate if the timing was OK, with hall signals heading one way, and motor control the other.
[18:27:39] <andypugh> I just need to plug it all back together and we can try that with the washing machine motor.
[18:29:24] <micges> first Peter needs to backport 7i80 eth interface to 8i20
[18:29:51] <andypugh> No, 8i20 is smart-serial
[18:30:09] <andypugh> I was meaning 7i80 to 7i76 to 8i20 to motor
[18:30:17] <micges> ah
[18:30:29] <micges> I messed up, here it's also bed time :)
[18:30:58] <andypugh> 7i80 to 7i44 to 8 x 8i20 would be a fun setup :-)
[18:31:16] <andypugh> (7i74, I suppose, with the 7i80db)
[18:32:44] <micges> seems like utimate test
[19:12:16] <andypugh> Nothing like a 1kw motor and 100V to check your confidence in a system :-)
[19:13:13] <micges> :)
[19:13:56] <micges> I'm not so confident, my first test will be 75W dc motor
[19:14:31] <micges> 7i80 + 7i40
[19:17:27] <andypugh> Right, I am going to set the system running overnight to see how it goes.
[19:17:49] <micges> thanks
[19:17:52] <micges> good night
[19:31:23] <cradek> http://pastebin.com/raw.php?i=gFj18F1Q
[19:31:24] <cradek> wheeee
[19:34:28] <cradek> although ... http://pastebin.com/raw.php?i=2HpNQz4S
[19:36:05] <cradek> Generating grub.cfg ...
[19:36:05] <cradek> Found linux image: /boot/vmlinuz-3.4.55-rtai-1
[19:36:06] <cradek> Found linux image: /boot/vmlinuz-3.2.0-57-generic
[19:36:06] <cradek> Found initrd image: /boot/initrd.img-3.2.0-57-generic
[19:36:19] <cradek> and I'm not getting an initrd image for this new kernel. I wonder will it boot?
[19:37:09] <seb_kuzminsky> hi cradek
[19:38:39] <seb_kuzminsky> i also do not have an initrd for the rtai kernel, and it boots
[19:38:46] <seb_kuzminsky> that's kind of broken though :-(
[19:46:18] <seb_kuzminsky> the deb is made with 'make-kpkg', maybe that does it wrong somehow