#linuxcnc | Logs for 2012-02-10

Back
[02:11:06] <Loetmichel> mornin'
[02:12:28] <Jymmm> Howdy
[02:13:47] <Loetmichel> *yawn* ... drove 15 minutts to work, working about an hour now, STILL not really awake :-(
[02:14:09] <Jymmm> temperature?
[02:15:08] <Loetmichel> -12°c
[02:15:34] <Jymmm> Ha! I'd have the water heater on high, and STILL be in the shower!!!
[02:17:40] <Loetmichel> hihi
[02:17:59] <mazafaka> it's also relatively warm here.
[02:25:00] <Jymmm> Finally! Fucking FF
[02:26:50] <Jymmm> In Firefox, you have the SEARCh box, but when you search for somethign you get a url that has all kinds of crap in it. You can edit the serach engine xml file (or actually copy and create a new one so you dont get that anymore, but you have to recreate it after you update/upgrade FF.
[02:27:09] <Jymmm> EDITED = https://www.google.com/search?q=qu
[02:27:26] <Jymmm> ORIGINAL = https://www.google.com/search?q=qu&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a
[02:30:17] <Jymmm> http://codepad.org/OtZ5QPUd
[02:31:01] <Jymmm> It also seems so speed up search results too.
[08:42:23] <Gromits> I put a new ssd on my linuxcnc machine and reinstalled a fresh copy of ubuntu and linuxcnc 2.4.7. I then used the buildbot to install 2.5.0pre. When I run 2.5 I get these errors: http://pastebin.com/VGbVFUD3
[08:42:31] <Gromits> Did I do something wrong in the install?
[08:45:40] <awallin> does 2.4.7 work OK?
[08:45:52] <awallin> "Realtime system did not load" is the problem in that pastebin
[08:45:53] <Gromits> I can run the simulators in 2.4.7
[08:46:19] <Gromits> it says that module-help is not suid root.
[08:46:31] <Gromits> What is/where is module-helper?
[08:46:49] <Gromits> and why wouldn't it have been setuid root from the distribution?
[08:47:02] <awallin> I don't know. 2.4.7 might be the one to use for now, untin linuxcnc stabilizes..
[08:47:21] <awallin> you could try building 2.5.0.pre from source
[08:47:51] <Gromits> I wanted to avoid that, which is why I was using the buildbot
[08:48:07] <jthornton> did you uninstall 2.4?
[08:48:12] <Gromits> I had previously built 2.5 from source and it was working, but I like the idea of using the buildbot....
[08:48:30] <Gromits> 2.4.7 was installed from the LiveCD (along with Ubuntu)
[08:49:09] <Gromits> I can run simulators in 2.4.7, it only fails if I try to run something in 2.5
[08:50:00] <jthornton> I went to yhe synaptic package manager and uninstalled 2.4 then installed the deb
[08:50:36] <jthornton> for 2.5
[08:51:04] <jthornton> iirc
[08:51:51] <Gromits> I installed 2.5 according to the instructions here: http://buildbot.linuxcnc.org/
[08:52:02] <Gromits> 2.4.7 is still on my system.
[08:52:33] <Gromits> I go into /emc2/emc2-dev and run ". scripts/rip-environment" and then run "linuxcnc"
[08:53:26] <jthornton> I'm pretty sure I uninstalled 2.4 from both computers that are running 2.5 deb but I'll be out in the shop in a 1/2 hr or so and can verify
[08:54:09] <jthornton> if you open up a terminal and type in emc or linuxcnc what version shows up
[08:54:24] * jthornton heads for the shower
[08:54:55] <Gromits> a2ms@a2msg:~$ linuxcnc
[08:54:57] <Gromits> LINUXCNC - 2.5.0-pre2-847-gf449347
[08:54:59] <Gromits> a2ms@a2msg:~$
[08:55:13] <Gromits> but 2.4.7 is still there...
[09:09:48] <Gromits> Here is the dmesg output, but I don't see anything obvious. Just all the seg faults.... http://pastebin.com/HgPL3vFY
[09:13:55] <JT-Shop> ok, I'm out in the shop singing "light my fire"
[09:15:17] <JT-Shop> pretty much the same thing I get
[09:15:27] <Gromits> what, the dmesg output?
[09:15:47] <JT-Shop> either try halrun -U or reboot and try again
[09:16:27] <Gromits> ok, will try reboot.... back in a few....
[09:18:55] <Gromits> no luck, same error
[09:24:11] <JT-Shop> are you trying to run one of the sample configs?
[09:24:27] <Gromits> It doesn't matter. Sample config or my config
[09:25:07] <Gromits> Here is what I get when I do halrun -U
[09:25:10] <Gromits> a2ms@a2msg:~/emc2/emc2-dev$ halrun -U
[09:25:12] <Gromits> RTAPI: ERROR: could not open shared memory (errno=2)
[09:25:14] <Gromits> ERROR: rtapi init failed
[09:25:15] <Gromits> HALCMD: Release Mutex failed!
[09:25:17] <Gromits> RTAPI: ERROR: could not open shared memory (errno=2)
[09:25:19] <Gromits> /home/a2ms/emc2/emc2-dev/scripts/halrun: line 10: 2087 Segmentation fault halcmd stop
[09:25:21] <Gromits> RTAPI: ERROR: could not open shared memory (errno=2)
[09:25:23] <Gromits> /home/a2ms/emc2/emc2-dev/scripts/halrun: line 10: 2088 Segmentation fault halcmd unload all
[09:25:24] <Gromits> module_helper is not setuid root
[09:25:26] <Gromits> module_helper is not setuid root
[09:25:27] <Gromits> module_helper is not setuid root
[09:25:29] <Gromits> module_helper is not setuid root
[09:25:30] <Gromits> module_helper is not setuid root
[09:25:32] <Gromits> module_helper is not setuid root
[09:25:34] <Gromits> module_helper is not setuid root
[09:27:26] <cradek> http://www.google.com/search?&q=module_helper+is+not+setuid+root
[09:27:28] <JT-Shop> I'll verify in a moment but I'm 99% sure I uninstalled 2.4 from both machines running the 2.5 deb
[09:28:20] <cradek> second hit has the correct answer
[09:28:49] <cradek> (sudo make setuid)
[09:29:14] <Gromits> cradek, installing from buildbot one needs to do a make?
[09:29:15] <JT-Shop> he is running a deb
[09:29:36] <cradek> no he's not: /home/a2ms/emc2/emc2-dev/scripts/halrun:
[09:30:01] <JT-Shop> oh, shortcut still pointing to the RIP
[09:30:25] <Gromits> I am confused...
[09:30:36] <JT-Shop> I'm glad someone is awake this morning
[09:31:28] <Gromits> I am opening a terminal, going to ~emc2/emc2-dev/, running . scripts/rip-environment, and then running linuxcnc. SHould I not be doing that?
[09:31:39] <JT-Shop> not for a deb
[09:32:19] <JT-Shop> Applications > CNC > EMC2
[09:32:33] <JT-Shop> or from a terminal linuxcnc
[09:33:21] <Gromits> Oh, duh.
[09:33:32] <Gromits> I need to move my configs from the dev directory...
[09:33:35] <JT-BP> john@D510MO:~$ emc
[09:33:37] <JT-BP> EMC2 - 2.5.0-pre2-626-g402b6cd
[09:33:58] <JT-BP> pick a sample config and it will set up the directories for you
[09:34:15] <JT-Shop> yep
[09:34:37] <Gromits> Yeah, I had done that already, but my configs from the other drive where in the emc2-dev directory -- because that one was built from source...
[09:41:15] <JT-Shop> I checked and I removed 2.4 from both computers via the synaptic package manager before installing the 2.5 deb to be sure I got the 2.5 manuals in the Applications > CNC menu
[09:59:22] <Gromits> Thanks for the help JT-Shop and cradek. Still working out some changes due to disk swap but at least I can run 2.5....
[10:08:49] <Gromits> we wrote a custom hal module (a variation on limit3 called limit3step) when we had source. I now need to incorporate that module into the deb build. I see limit3step.comp but also limit3step.ko and limit3step.o as well as limit3step.mod.o. Which of these do I actually need and how do I figure out where it(they) go?
[10:14:04] <JT-Shop> did you use comp?
[10:14:35] <Gromits> I see, for example, limit3.ko in /usr/realtime-2.6.32-122-rtai/modules/linuxcnc, so I will try copying our limit3step.ko there and see what happens....
[10:14:56] <Gromits> not sure what you mean by comp...?
[10:16:53] <JT-Shop> http://linuxcnc.org/docview/2.5/html/hal/comp.html
[10:17:25] <Gromits> Ah. Not sure, I didn't write, it a guy working with me did. I will have to ask him./
[10:17:36] <JT-Shop> well I see you did reading back
[10:17:47] <JT-Shop> http://linuxcnc.org/docview/2.5/html/hal/comp.html#sec:Compiling-realtime-components
[10:18:01] <Gromits> cool, thanks!
[10:18:05] <JT-Shop> eyes are fuzzy this morning
[10:18:15] <JT-Shop> so comp --install my.comp
[10:18:30] <JT-Shop> from the same directory as the .comp file is located in
[10:23:06] <Gromits> hmm, comp appears to be an MH command to compose a message...
[10:24:57] <Gromits> ok, found a comp in my old emc source directory...
[10:29:48] <Gromits> nope, that doesn't work - probably because comp was built in a different directory...
[10:31:08] <Gromits> make[1]: Leaving directory `/usr/src/linux-headers-2.6.32-122-rtai'
[10:31:09] <JT-Shop> you should just use the installed version of comp
[10:31:10] <Gromits> cp limit3step.ko /home/a2ms/emc2/emc2-dev/rtlib/
[10:31:42] <Gromits> where would that be?
[10:32:22] <Gromits> the installed version of comp is not linuxcnc "comp", but mail handler "comp"
[10:32:24] <JT-Shop> copy limit3step.comp to your home folder then open a terminal and say comp --install limit3step.comp
[10:32:47] <JT-Shop> well that would be a problem
[10:32:52] <Gromits> yes
[10:32:57] <JT-Shop> what does man comp show
[10:33:09] <Gromits> it shows me how to compose a message for mail :-)
[10:33:52] <JT-Shop> did you install some kind of mail program after you installed the liveCD?
[10:34:01] <Gromits> no
[10:34:40] <Gromits> well, I didn't. I suppose Peter (my partner in crime) might have...
[10:34:52] <Gromits> I could try to reinstall the deb build...
[10:34:59] <JT-Shop> yea that seems likely
[10:35:16] <JT-Shop> I'd uninstall 2.4 then install 2.5 to be sure
[10:35:30] <Gromits> ok
[10:35:40] <Gromits> thx again, will give it try
[10:38:33] <Gromits> In synaptic there is emc2 and emc2-dev, do I need both or just emc2?
[10:39:04] <jthornton> I have both for some unknown reason
[10:39:20] <Gromits> will try just emc2 and see if I get comp...
[10:40:03] <Connor> Hey guys, I'm in the market to get a new set of indexed drill bits.. Suggestions? I'm not looking to spend a arm and a leg... I'm thinking HSS will be fine..
[10:40:39] <jthornton> good drill bits cost an arm and a leg and a foot now a days
[10:41:40] <Gromits> hrumph, re-install didn't change comp, it's still email message compose..
[10:42:00] <jthornton> uninstall it
[10:46:40] <Gromits> well, that helped a little. man comp now shows the hal comparator and when I run comp --install limit3step.comp it tells me to apt-get install nmh (mail handler). Apparently Peter was working on this too last night ;-)
[10:47:51] <JT-Shop> ouch
[10:48:12] <Gromits> so need to find where linuxcnc hides it binaries i guess
[10:48:36] <cradek> you need to install the -dev package to get comp
[10:48:49] <JT-Shop> src somthing or other
[10:50:57] <JT-Shop> there you co Gromits someone that really knows :)
[10:51:09] <JT-Shop> co/go
[10:51:21] <Gromits> cool
[10:52:46] <Gromits> hey, man comp now gives me the right thing!
[10:53:39] <JT-Shop> YEA!
[10:53:48] <Gromits> ...and there was much rejoicing.... rah.
[12:11:00] <Gromits> I want to add our machine configs to the wiki for examples on THC, Glade, Gantry, etc. Should I just create a new page or do I start by adding a link from another page and if so, what other page?
[12:12:27] <mhaberler> sure
[12:12:42] <mhaberler> how did the gladevcp stuff work out after all?
[12:12:52] <Gromits> great, it is all working now
[12:12:53] <mhaberler> add a new page
[12:13:08] <mhaberler> whoa! I think other folks will like it
[12:13:26] <mhaberler> I think there was some examples collection.. let me see
[12:15:05] <TRW_Ags_Mx> Some idea about the current source/drain in the enable output of a mesa 7i33? thanks
[12:15:40] <mhaberler> Configuring LinuxCNC/User configurations looks like a good section
[12:15:51] <Gromits> Here was our final .py file (if you want to look at it): http://pastebin.com/jHXqJUni
[12:15:52] <mhaberler> the save_state issue is gone?
[12:16:05] <Gromits> yes
[12:16:58] <mhaberler> re blink rate of leds, see faq section in gladevcp manual - confusing but works; I dont think you need to set it in py file afterall
[12:17:35] <Gromits> I can hand edit the .ui file after running glade, but they are set to 0 after running glade. So, we just put them in the .py file...
[12:17:56] <Gromits> Not sure why glade seems to ignore the setting but it does...
[12:18:06] <mhaberler> ah, I see.. the debug states were saved too but you want to default them on startup
[12:18:30] <mhaberler> better than explicitly listing the desired widgets for restoring though
[12:18:43] <Gromits> we have checkboxes to enable some debugging states and we want to make sure those are off when Axis first comes up
[12:18:58] <Gromits> right
[12:19:03] <Gromits> less typing :-)
[12:19:06] <mhaberler> well widget_defaults(select_widgets(self.builder.get_objects(), hal_only=True,output_only = True))
[12:19:22] <mhaberler> selects ALL output widgets including the debug states
[12:19:33] <mhaberler> overall, looks very good!
[12:19:53] <mhaberler> gladevcp101: passed;)
[12:19:57] <Gromits> :-)
[12:20:02] <Gromits> when is the final?
[12:20:10] <Gromits> I will need to do more studying
[12:20:19] <mhaberler> what are them plans?
[12:20:59] <Gromits> the plans are to get the plasma torch back (from repair) and cut some metal and stop messing with software!
[12:21:09] <mhaberler> sounds reasonable
[12:21:16] <mhaberler> hope you're set
[12:21:51] <Gromits> for the time being, definitely. we may come up with more tweaks (that might require help) but will cross that bridge when we fall off it ;-)
[12:24:14] <mhaberler> http://despair.com/risks.html
[12:24:36] <Gromits> heh. very nice
[12:25:00] <mhaberler> the whole collection is stunning: http://despair.com/viewall.html
[12:25:31] <mhaberler> like 'Government' or 'Consulting';)
[12:25:37] <Gromits> I love the Tradition one...
[12:26:28] <mhaberler> or self-esteem.
[12:27:09] <mhaberler> ok, enjoy & good luck
[12:27:46] <Gromits> thanks for the help yesterday
[12:27:56] <mhaberler> sure
[12:44:11] <awallin> latest and greatest pocketing toolpath: http://www.anderswallin.net/wp-content/uploads/2012/02/ma_pocket.png
[12:53:10] <bill20r3> Pretty
[12:58:12] <alex4nde1-> awallin: nice,. I've been messing with much simpler implementations of something similar, doing spiral and trochoidal cuts.
[13:05:53] <Gromits> mhaberler, you there?
[13:05:59] <mhaberler> mmm
[13:06:12] <Gromits> We get this error when we quit axis...
[13:06:30] <mhaberler> pasteit
[13:06:34] <Gromits> "/usr/bin/gladevcp:263: GtkWarning: GdkWindow 0x4200003 unexpectedly destroyed
[13:06:36] <Gromits> gtk.main()"
[13:06:56] <Gromits> So, when we tried to use Window1 in this:
[13:07:00] <Gromits> def vbox1_destroy_cb(self,obj,data=None):
[13:07:02] <Gromits> self.ini.save_state(self)
[13:07:06] <Gromits> it doesn't work.
[13:07:14] <Gromits> vbox1 however, works fine
[13:07:17] <mhaberler> hm, that is a consequence of emc shutdown which is kinda brute
[13:07:22] <Gromits> just seems like a bug and thought I would mention it...
[13:07:57] <Gromits> i get that even if I use Exit from Axis menus...
[13:08:05] <mhaberler> one way to grind this cat is to use a signal handler. I'll look into it because it bugged me already, too
[13:08:30] <Gromits> ok
[13:08:44] <mhaberler> tacking it onto an event handler on a gtk window relies on, uh, controlled shutdown
[13:09:17] <Gromits> the examples you have show using window1, so either you should change those or...
[13:10:22] <Gromits> well controlled should be from the Axis menu (exit). We usually just close the window though, but...
[13:11:06] <mhaberler> I need to figure how the linuxcnc script and/or axis does this, but its a good point
[13:23:23] <Jymmm> Gromits: you gt the same issue with tk as well?
[13:23:58] <Jymmm> Gromits: or any other non-AXIS iface.
[13:24:46] <Jymmm> Might help to isolate if it's lcnc, axis, gtk, etc.
[14:38:13] <JT-Shop> or the back side of it
[14:39:20] <JT-Shop> my daughter would not clean her room ONCE, I put all her crap in a trash bag and put it out by the curb for the garbage pick up... never had another problem with a dirty room from her
[14:40:27] <Jymmm> heh, did you throw out any good stuff?
[14:43:44] <JT-Shop> I did not discriminate good from bad, but mostly clothing and shoes
[14:44:56] <Jymmm> Ah, a Girl's worse nightmare. gotcha.
[14:45:10] <Jymmm> how old was she at the time?
[14:45:14] <JT-Shop> yea, she is a hard working girl now
[14:45:19] <JT-Shop> 16
[14:45:57] <Jymmm> Sadly, I know someone's Mom that is worse (even to this day).
[14:46:09] <Jymmm> dirty clothes down the hallway etc
[14:48:21] <JT-Shop> http://www.youtube.com/watch?v=mvXGJ8eP1B0
[15:01:35] <Jymmm> EEEEEEEEEEEEEEEEEEEHHHHHHHHHHHHaaaaaaaaaaaaaaaaaaaaa
[15:14:24] <Jymmm> JT-Shop: How old is she now?
[15:14:51] <Jymmm> (please don't say 16.5 lol)
[15:15:21] <Jymmm> or even 15.5 NOW
[15:18:18] <JT-Shop> 40
[15:18:56] <JT-Shop> Jymmm: you like the gris gris?
[15:19:14] <Jymmm> JT-Shop: heh, you should remind her one day.
[15:20:07] <Jymmm> JT-Shop: that link you posted? Eh, not really my thing. I've relaized I like most things with a heavy base to them.
[15:20:09] <JT-Shop> of how well she turned out?
[15:20:22] <Jymmm> But I do listen to pretty much everything.
[15:20:40] <Jymmm> JT-Shop: well that too, but that you tossed her stuff out.
[15:20:43] <JT-Shop> I listen to all music... I don't consider rap crap music
[15:21:15] <JT-Shop> I somehow don't think she has forgot that day...
[15:21:16] <Jymmm> Some rap is okey. As is some non-nasal country.
[15:22:01] <Jymmm> JT-Shop: Well if you remind her, be sure to have a gift card to some clothing store handy.
[15:22:12] <JT-Shop> http://www.youtube.com/watch?v=kiVDtll8th0
[15:23:03] <Jymmm> Yes, THAT!
[16:04:16] <Gromits> very odd problem with glade. We are adding some bells and whistles, just buttons to touch off, and go home, etc. Touch off was working, I add the go home buttons and now touchoff doesn't work (but home buttons do).
[16:04:26] <Gromits> .ui file: http://pastebin.com/U7SDpQbr
[16:04:53] <Gromits> look at touch off X for example
[16:05:43] <Gromits> it should be issuing a G10 L2 P1 X0 since the button is assoc with the action, but nothing happens....
[16:15:30] <Gromits> no errors in the terminal when this happens....
[16:26:32] <Gromits> Nevermind. I am using G10 incorrectly.
[16:31:01] <Gromits> what g-code command is used (or is it a gcode) when I use the Touch-off button in axis?
[16:31:21] <Gromits> I want g10 l2 p1 X{My current position}
[16:36:18] <skunkworks_> Gromits: interesting...
[16:37:58] <skunkworks_> Gromits: maybe you could first to a g28.1 - that would put the current location in 5161-5166
[16:38:26] <skunkworks_> then do g10 l2 p1 X#5161 maybe
[16:38:35] <skunkworks_> I don't know if there is an easier way...
[16:39:56] <Gromits> g10 L20 P1 X0Y0X0 is the magic....!
[16:40:17] <skunkworks_> heh - that is easier...
[16:40:55] * skunkworks_ will have to remember that...
[16:43:19] <Gromits> here is our glade panel config: http://bgp.nu/~tom/pub/Glade-Panel.png
[16:46:54] <skunkworks_> Gromits: Cool!
[16:47:33] <Gromits> We have been polishing the turd (as it were) since our torch is broken...
[16:48:01] * skunkworks_ needs to play with glade
[16:48:03] <Gromits> I am loving Glade the more I use it.
[16:48:25] <skunkworks_> a lot more options than pyvcp?
[16:48:28] <Gromits> Silly things, like being able to have tooltips on every button when you hover over it is really cool!
[16:48:52] <Gromits> running python scripts, or gcode commands....
[16:49:11] <Gromits> blinking leds!!
[16:49:30] <Gromits> It's the little things in life :-)
[16:49:36] <Jymmm> BLINKING BUNNYS!!!!!!!!!!!!!!!!
[16:49:51] <Gromits> that too!
[16:49:59] <Jymmm> WOOHOO!!!
[16:54:48] <Gromits> I will put the configs up on the wiki this weekend or next week...
[18:07:47] <JT-Shop> dang Elvis left the building
[18:16:13] <pfred1> JT-Shop but he made a video before he did http://www.youtube.com/watch?v=fHPKaHLzXes
[18:17:08] <JT-Shop> clap, clap, clap they call him the "clapper"
[18:18:29] <Jymmm> *clap clap* CLAP ON....
[18:19:48] <JT-Shop> lol
[18:25:30] <JT-Shop> 1 1/2 rondel, large to go
[18:26:39] <JT-Shop> s/rondel/rondelle
[18:49:18] <JT-Shop> http://www.youtube.com/watch?v=9LjbMVXj0F8
[18:55:39] <pfred1> check out this genius running his BOB on conductive bubble wrap http://www.youtube.com/watch?v=F_AVZIhVi9U
[18:56:28] <pfred1> I have to hand it to him it does appear to work for them
[19:06:43] <pfred1> A Windows ad I'd like to see:
[19:06:45] <pfred1> Steve Ballmer holding up a copy of Windows 8 saying, So do you feel lucky punk? Well, do you? Did our coding team fix 6 bugs, or did we only fix 5? In all the excitement I kind of forgot myself ...
[19:10:14] <olson_> Hello, I'm a newb struggling with EMC2 and a PCIe Parallel port.
[19:10:57] <olson_> I've got a NetMos 9901, and the IO Ports go disabled on my constantly
[19:11:29] <olson_> I've gotten stepconfig to drive my X and Y access for a few moments once before they went offline.
[19:11:43] <olson_> It's hooked up to a Sherline driver box and XY table.
[19:13:29] <pfred1> olson_ what does this command say lsmod | grep parport
[19:18:41] <olson_> At initall boot it shows nothing
[19:19:14] <olson_> I can load parport_pc io=0x3010 irq=3 and it will pick up the card
[19:19:45] <pfred1> olson_ I don't think EMC uses the parport driver so nothing is good
[19:19:46] <olson_> I am fresh booting now so I can confirm that. :)
[19:20:17] <olson_> I've read that you can use it and specify your port using in the hal config by index number.
[19:20:39] <olson_> But doing that and firing up the mini interface geets me an instant lock up
[19:21:35] <pfred1> I don't use it here and it works for me
[19:22:07] <pfred1> I just put the io port into stepconf and let RTAI handle it
[19:22:59] <pfred1> olson_ sometimes if I run into a problem on my system I will examine /var/log/syslog for some clues as to what is going on and how i might fix it
[19:23:15] <olson_> I do see parport, but not parport_pc
[19:23:37] <pfred1> let me check on my cnc box it works
[19:25:10] <olson_> Huh, ports show enabled after that reboot.
[19:25:20] <olson_> This thing is so nondeterministicv
[19:25:30] <olson_> Or I so don't understand what I'm doing
[19:26:37] <pfred1> actually i get something called hal_parort when I am running
[19:26:48] <pfred1> hal_parport even
[19:27:48] <olson_> Well, I'm using the Sherline load, loadrt hal_parport cfg="0x3010" in my standrad_pinout.hal
[19:28:20] <olson_> My lspci -vv is at http://pastebin.com/CTJ5Bdx9
[19:28:29] <olson_> (Just the Parallel part)
[19:28:46] <olson_> Let me try step config and see what happens this time.
[19:28:49] <pfred1> your system does not have a built in port?
[19:29:02] <olson_> Alas no
[19:29:08] <olson_> And it's only PCIe
[19:29:44] <olson_> The machine is a Dell Optiplex 390, EMC2 booting off of a USB Drive
[19:31:42] <pfred1> this is what mine looks like before I run RTAI then after http://pastebin.com/a9S4fZrr
[19:32:21] <olson_> Okay I fired up stepconf, set the address to 0x3010, Sherline Driver type and pin out.
[19:32:24] <pfred1> mine works good
[19:32:59] <olson_> Select test axis for the X axis, and the stepper twitches for a sec, then both I/O ports show up as disabled immediately afterwards
[19:33:51] <olson_> Yeah, I don't see that hal_parport afterwards
[19:34:15] <olson_> Just the lone parport module.
[19:35:08] <pfred1> so you've 2 ports on your card one at 0x3010 and another at 0x3000
[19:35:40] <olson_> Before trying stepconf I can load the parport_pc module and it works, afterwards it appear to be locked out as well.
[19:35:45] <pfred1> and you say it has worked so I can only assume you've found the right one but you say you're not getting hal_parport to load
[19:35:47] <olson_> It's a single port card
[19:36:04] <pfred1> well it has 2 ports
[19:36:17] <olson_> I'm not sure what the extra IO Port range is for, but it appears to be common for PCI/PCIe Parallel ports to have multiple ranges with several unused
[19:36:20] <pfred1> says so right in your lspci
[19:36:43] <olson_> I understand the 9901 has a proprietary chip chaining system so it may be for that.
[19:38:15] <olson_> The chipset could support up to two serial ports as well, so it might also have been for those.
[19:38:27] <olson_> Havn't gotten far enough through the dataheet to tell
[19:47:00] <pfred1> this is what mine looks like http://pastebin.com/cvvdiUDa
[19:49:53] <pfred1> olson_ what does dmesg say when you try to run your RTAI?
[19:51:15] <pfred1> olson_ mine looks like this http://pastebin.com/Ztn3xfyq
[19:55:03] <olson_> It looks pretty similar, I don't see anything specificly about the parallel card in either.
[19:55:37] <olson_> Of course, most of the time mine crashes (complete hand or Keyboard/Mouse unresponsive) if the card isn't disabled.
[19:56:02] <pfred1> olson_ just X or the whole system?
[19:56:37] <pfred1> I guess if the keyboard locks up it is hard to tell
[19:56:46] <olson_> Indeed
[19:56:58] <pfred1> not impossible though that is what syslog if for
[19:57:04] <pfred1> is for even
[19:57:42] <olson_> True, but I have to hard reset it and thats somewhat difficult to distinguish from a hard hang.
[19:58:42] <pfred1> syslog is persistant between reboots
[19:59:06] <olson_> Yeah, but the logs just suddenly quit
[19:59:16] <olson_> Nothing about any last minute failures
[19:59:53] <olson_> Trying it from a remote X session
[20:00:35] <olson_> Hmm, I beleive I can confirm the system is unhappy
[20:01:03] <olson_> Although.. It isn't locked up
[20:01:10] <pfred1> olson_ but you say it has worked?
[20:01:18] <olson_> When it goes bad it seems like drive access if toast.
[20:01:26] <olson_> Not with the card enabled
[20:01:31] <olson_> Only stepconf
[20:01:36] <pfred1> OK lets forget the card for a moment and check something else out
[20:01:38] <olson_> Which I beleive doesn't go into RT mode
[20:01:59] <olson_> Ouch!
[20:02:05] <pfred1> you've checked your base period latency haven't you?
[20:02:06] <olson_> cd / ; ls
[20:02:23] <olson_> bash: /bin/ls: No such file or directory
[20:02:35] <pfred1> having it set wrong can lead to some unfortunate events
[20:02:42] <olson_> I did a latency test at once point early on.
[20:03:16] <olson_> Hmmm, well I never set it
[20:03:18] <pfred1> often while debugging it is standard practice to raise up your time to see if it helps you any
[20:03:37] <olson_> The Sherline guide I started from just told me to use their prebuild config
[20:03:50] <pfred1> well they didn't give you a PC
[20:03:53] <olson_> unfortunate - drive control broke
[20:04:00] <olson_> or unfortunate - hard drive vanishes?
[20:06:15] <olson_> I'm thinking that a USB boot drive is maybe bad mojo with emc2
[20:07:21] <pfred1> it may be I never tried it myself with EMC I have wiht other Linux distros though
[20:07:38] <olson_> It works right up until I run the emc command
[20:07:51] <olson_> Maybe the RT mode no likely USB disk....
[20:08:24] <pfred1> once the RTAI module loads things do happen
[20:08:34] <pfred1> that don't happen when it isn't loaded
[20:08:52] <pfred1> I had issues with GL here myself
[20:09:08] <olson_> I guess I'll give the Live CD a shot
[20:09:24] <olson_> I think that's on a eSata port
[20:09:27] <pfred1> why can't you use a regular HDD?
[20:09:53] <olson_> I'm working it someone elses equipment, and they didn't want to loose the main OS.
[20:10:51] <olson_> I suggested an extra drive and it showed up with a USB one. I figured that it probably wouldn't be terribly fast, but would work.
[20:11:21] <olson_> (I've certainly run Linux off of enough flash sticks in my time)
[20:13:29] <pfred1> then you know live isn't quite the same as installed
[20:13:44] <pfred1> almost sort of most of the time but not all of the time
[20:14:16] <pfred1> thing about EMC is it is pretty demanding some times
[20:15:35] <pfred1> olson_ but your latency testing went well?
[20:17:05] <olson_> It did as I recall
[20:17:13] <olson_> I didn';t know what it should be.
[20:17:21] <pfred1> it is what it is
[20:17:47] <pfred1> latency isn't something you can do very much about in most cases
[20:18:34] <olson_> I've been using the Sherline install as well, so switching to the LinuxCNC distributed should get me to a perhaps better supported version
[20:19:20] <pfred1> I am not familiar wiht what Sherline provides
[20:19:33] <pfred1> it sounds similar but there may be some differences
[20:19:44] <olson_> It copmes preconfigured
[20:19:54] <pfred1> in any event i cannot imagine yo ucan skip the latency testing
[20:19:55] <olson_> And is supposed to largely plug and play with their products
[20:20:03] <olson_> Since I'm having a Plug and Pray experience....
[20:20:08] <pfred1> there is no way they can guess your system's latency
[20:20:15] <olson_> It may have done something automatically
[20:20:24] <pfred1> nope not how it works
[20:20:28] <olson_> Okay
[20:20:38] <olson_> I'll need to learn how it works then
[20:20:51] <pfred1> yeah having your timing wrong is a show stopper
[20:21:10] <pfred1> your machine besically runs itself into hte ground not being able to keep up with itself
[20:21:23] <pfred1> it locks up
[20:21:59] <pfred1> so get the best latency figures you can from the systme you are on and use them properly
[20:22:34] <pfred1> there should be a button right in stepconf where you can do the testing
[20:22:44] <olson_> My early test got me 8190ns on the base thread
[20:23:15] <pfred1> did you try to read the USB drive while you were testing?
[20:23:34] <olson_> It was botted from it
[20:23:40] <olson_> booted even
[20:24:06] <pfred1> thats nice but what is more important is the results during the test not an event that happened beforehand
[20:24:45] <pfred1> the test is how your machine performs with RTAI
[20:24:47] <olson_> I ran the testing tool and a bunch of software at the same time
[20:24:52] <olson_> glxgears
[20:24:55] <olson_> firefox
[20:24:59] <olson_> wiggled windows
[20:25:00] <olson_> etc...
[20:25:24] <pfred1> so you used 8190ns as your base period?
[20:26:04] <pfred1> I think the default is 15000ns if you do not change anything
[20:26:10] <olson_> I think it waS 10000
[20:26:17] <olson_> yEAH, YOUR RIGHT
[20:26:22] <olson_> Argh capslock
[20:26:46] <pfred1> well if 8190ns is correct then 10 or 15K should be OK
[20:27:50] <pfred1> how does the sherline ini differ from the generic mill file?
[20:28:19] <olson_> Don't know
[20:28:40] <pfred1> might be a good idea to try to generate a generic ini then diff the two
[20:28:45] <olson_> Never got far enough through stepconf to generate one
[20:28:52] <olson_> port dies at axis test usually
[20:29:02] <olson_> I'm trying to get the live CD up
[20:29:04] <pfred1> then don't test the axises
[20:29:12] <olson_> After that I can compare
[20:29:24] <pfred1> but you're saying even the generic ini kills your system?
[20:29:27] <olson_> I saved the Sherline confs to a stick
[20:30:13] <olson_> I shall l;ikely be back
[20:30:19] <olson_> Thats for the assistance
[20:30:22] <olson_> Thanks
[20:30:24] <pfred1> I will say this when I set my machine up I didn't use any insmod or modprobe or anything I just put my port into stepconf and it worked for me
[20:30:41] <olson_> Hoping that will work with the officialish media
[20:31:01] <pfred1> yeah i have to hit the sack here early day tomorrow
[20:36:43] <olson_> Thanks again!