#emc-devel | Logs for 2008-08-28

Back
[12:10:42] <CIA-37> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gcode/main.lyx: add missing item to quick reference table
[15:57:43] <CIA-36> EMC: 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py:
[15:57:43] <CIA-36> EMC: restore soft limits box for lathes
[15:57:43] <CIA-36> EMC: the old code got limits like [x, 0, z] while the new code gets limits like [x, +-1e99, z].
[15:57:43] <CIA-36> EMC: This is apparently an implementation detail in task when initializing limits
[15:57:43] <CIA-36> EMC: for axes that are not used. So let's treat any number bigger in absolute value than 1e30
[15:57:45] <CIA-36> EMC: as 0, so that lathe limits get drawn in the plane y=0 which is the plane of the
[15:57:47] <CIA-36> EMC: screen, instead of outside the view volume.
[15:58:03] <cradek> cool
[16:01:38] <SWPadnos> jepler_, I know I've asked this before, but I can't find the answer again: is the .config you used for the 8.04/AMD64 kernel online somewhere?
[16:03:06] <jepler_> you mean besides in the source and binary packages?
[16:03:33] <skunkworks_> wait - was that a slam? ;)
[16:03:36] <SWPadnos> right - nevermind. I have the RT kernel installed here, so I actually have the config
[16:03:38] <SWPadnos> yes
[16:04:09] <SWPadnos> I was thinking it had to come from /proc/config.gz, but I can't boot that kernel
[16:04:25] <SWPadnos> but it can easily come from /boot/ :)
[16:13:07] <alex_joni> SWPadnos: I stuck the i386 config into CVS
[16:13:14] <alex_joni> maybe you can do the same with the amd64
[16:13:36] <SWPadnos> ok - where did you put it? (/debian ?)
[16:14:05] <alex_joni> no, in infrastructure
[16:14:08] <SWPadnos> ok
[16:14:24] <alex_joni> it's not emc2 version specific
[16:14:29] <SWPadnos> once I get the new HD set up for CVS checkins, I'll do that
[16:14:35] <alex_joni> http://cvs.linuxcnc.org/cvs/infrastructure/livecd/kernels/hardy/
[16:15:26] <SWPadnos> hmmm. that's the name I have
[16:16:07] <SWPadnos> I wonder if it isn't the A64 kernel that got installed
[16:16:22] <SWPadnos> it should be, since that's what I installed with originally
[16:16:54] <SWPadnos> hmmm. it is A64
[16:58:57] <alex_joni> bbl
[18:23:11] <alex_joni> SWPadnos: do you know what FPGA on the m5i20 is ?
[18:23:30] <SWPadnos> XC2S200 I think - lemme checl
[18:23:33] <SWPadnos> check
[18:24:34] <skunkworks_> sparton xc2s200
[18:24:37] <SWPadnos> yep - XC2S200 PQG208AMS0553
[18:24:53] <SWPadnos> I think the G means "RoHS compliant"
[18:25:02] <SWPadnos> and the speed rating is 5C
[18:25:06] <alex_joni> ah, ok.. thanks
[18:25:17] <SWPadnos> sure
[18:25:35] <alex_joni> my board is at work..
[18:25:40] <alex_joni> so I can't check :)
[19:55:21] <alex_joni> I remember seeing some site with "free" fpga designs
[19:55:28] <alex_joni> does that ring a bell?
[19:57:16] <SWPadnos> www.opencores.org
[19:59:27] <alex_joni> looking for something to drive a colour LCD
[19:59:32] <alex_joni> the type without controller
[19:59:50] <SWPadnos> an LVDS-connected LCD
[20:00:47] <SWPadnos> you should need external hardware for that, unless you're just interfacing to COG (chip-on-glass) refresh controllers
[20:30:08] <alex_joni> I wonder why emc2 is so big
[20:30:20] <alex_joni> others seem to achieve with way less :)
[20:30:29] <SWPadnos> yeah, look at TurboCNC
[20:30:33] <alex_joni> "Includes Verilog source code (about 200 lines) and C++ source code (about 400 lines).
[20:30:36] <alex_joni> Implements linear motion on up to 4 axis simultaneously.
[20:30:39] <alex_joni> "
[20:31:39] <jepler_> I once wrote cnc in about that many lines too
[20:31:51] <jepler_> it was fine for controlling my etch-a-sketch with G1 moves
[20:32:03] <jepler_> only I think I wrote it in all python or somehthing :-P
[20:32:13] <alex_joni> heh
[20:32:38] <cradek_> I'm stumped on this bug, can you guys help?
[20:33:00] <cradek_> in mdi, s100 m3, then hit escape. spindle stops. mdi s200, spindle incorrectly starts again
[20:33:01] <alex_joni> we can try
[20:33:19] <cradek_> it's due to spindleOn in emccanon.cc becoming out of date after the abort
[20:33:41] <cradek_> so I rewrote it to check GET_EXTERNAL_SPINDLE_* instead, but those fields in teh stat buffer are always zero
[20:33:44] <jepler_> alex_joni: (is that the knjn.com motion control software?)
[20:33:55] <alex_joni> jepler_: yeah
[20:34:07] <jepler_> personally I wouldn't buy it *cough*
[20:34:07] <alex_joni> cradek_: hang on, rebooting into hardy
[20:34:11] <cradek_> in emctop, I can see that those fields in stat are being updated
[20:34:23] <cradek_> cradek_ is now known as cradek
[20:35:58] <cradek> hm, in mdi mode, those fields are getting updated, and my fix works right
[20:36:41] <cradek> but not when running a program
[20:36:47] <jepler_> readahead
[20:36:51] <jepler_> ?
[20:36:57] <cradek> like s300m3 / g1x1 / s400 / g1x0
[20:37:15] <cradek> yes maybe, but I see EMC_SPINDLE_ON_TYPE is EMC_TASK_EXEC_WAITING_FOR_MOTION_AND_IO
[20:37:16] <jepler_> when s400 is executed, s300m3 hasn't been executed yetr
[20:37:26] <jepler_> interp_list
[20:37:41] <cradek> here is where I'm lost, I think
[20:39:26] <cradek> I'm suspicious any time any layer saves some state to refer to later - it seems like every one of those is a bug
[20:39:34] <alex_joni> heh
[20:39:42] <cradek> could just be that I'm suspicious because I don't understand all the layers
[20:39:52] <SWPadnos> if there's no synchronization, then they certainly can be bugs
[20:40:09] <alex_joni> even with sync they can easily turn into bugs
[20:40:17] <SWPadnos> sure
[20:40:20] <alex_joni> cradek: so what's the latest problem?
[20:40:34] <cradek> in mdi, s100 m3, then hit escape. spindle stops. mdi s200, spindle incorrectly starts again
[20:40:40] <cradek> it's due to spindleOn in emccanon.cc becoming out of date after the abort
[20:41:05] <SWPadnos> does abort issue S0 or M5?
[20:41:12] <cradek> alex_joni: read back 20 lines?
[20:41:15] <jepler_> SWPadnos: it does neither
[20:41:18] <jepler_> it doesn't send any mdi commands
[20:41:25] <alex_joni> cradek: I already read that
[20:41:30] <SWPadnos> well, the canon / NML equivalent
[20:41:35] <cradek> ok
[20:41:45] <cradek> sorry I did not understand your question then
[20:41:49] <alex_joni> is STOP_SPINDLE_TURNING() beeing called?
[20:42:14] <cradek> no, just emcAxisAbort(), emcTrajAbort(), etc
[20:42:52] <alex_joni> right, and one of them stops the spindle motion (in motmod)
[20:42:59] <cradek> yes
[20:43:13] <cradek> but interp and canon's state is not updated
[20:43:54] <alex_joni> do you know what stops the spindle?
[20:44:07] <alex_joni> not EMCMOT_ABORT ..
[20:44:09] <cradek> I thought using the canon call GET_EXTERNAL_SPINDLE_* (which reads stat) would fix it
[20:44:42] <cradek> EMC_TASK_ABORT
[20:45:12] <alex_joni> emcMotionAbort() calls emcSpindleAbort()
[20:45:21] <alex_joni> but only if localSpindleMode == 0
[20:45:38] <alex_joni> (that was something I added to let it keep running for people who don't want it to stop on Esc)
[20:45:42] <alex_joni> or modechange
[20:45:58] <cradek> I want it to stop on ESC but not mode change :-/
[20:46:51] <alex_joni> check taskintf.cc / 1295
[20:47:13] <alex_joni> //FIXME: maybe add other modes / or replace completely with run-levels..
[20:47:14] <alex_joni> :D
[20:47:21] <cradek> yes I was there too...
[20:47:34] <alex_joni> ok, that doesn't solve your problem..
[20:47:44] <alex_joni> I think the issue is somewhere in emctaskmain
[20:47:56] <alex_joni> can you put a patch online so I can try it out?
[20:53:01] <alex_joni> on my checkout only an ESTOP causes the spindle to stop
[20:53:14] <alex_joni> but if I program Sxx it starts again, so I agree there's a bug there
[20:57:21] <cradek> oops, I'm back now
[20:57:33] <cradek> what patch do you want from me? my patch to canon?
[20:57:57] <alex_joni> yeah
[20:59:49] <cradek> http://pastebin.ca/1187848
[21:01:05] <alex_joni> missing some parts?
[21:01:55] <cradek> ?
[21:02:04] <alex_joni> doesn't seem complete to me
[21:02:17] <cradek> that's what I thought would fix it, and it does fix it for mdi mode, but not running a program, I assume because of some queue
[21:02:44] <alex_joni> I don't see GET_EXTERNAL_SPINDLE ..
[21:02:50] <cradek> that was already there!
[21:02:52] <SWPadnos> line 46
[21:02:53] <cradek> it just doesn't work
[21:03:04] <cradek> also GET_EXTERNAL_SPEED
[21:03:24] <alex_joni> hmm.. why don't I have it then?
[21:03:43] <cradek> ??
[21:03:44] <alex_joni> found it now :(
[21:03:44] <cradek> trunk?
[21:03:49] <alex_joni> yeah.. sorry
[21:06:32] <alex_joni> cradek: remember the thing to put in the config to make it stop on Esc?
[21:07:01] <cradek> sure
[21:07:10] <alex_joni> I meant what is it ..
[21:07:12] <cradek> well that's the default
[21:07:18] <cradek> oh I don't know that :-)
[21:07:19] <alex_joni> it doesn't stop here
[21:07:20] <cradek> I'll look
[21:07:25] <alex_joni> when I push Esc
[21:07:26] <SWPadnos> SPINDLE_MODE or something like that
[21:07:57] <alex_joni> SWPadnos: right
[21:08:02] <alex_joni> SPINDLE_MODE in EMCIO
[21:08:08] <SWPadnos> heh - that wasn't the question I guess :)
[21:08:16] <cradek> if (NULL != (inistring = inifile.Find("SPINDLE_MODE", "EMCIO"))) {
[21:08:17] <alex_joni> SWPadnos: it was..
[21:08:21] <SWPadnos> oh
[21:08:32] <cradek> that's an obscure name...
[21:09:18] <cradek> driving home now...
[21:09:23] <cradek> thanks for looking at this alex
[21:09:24] <alex_joni> ok
[21:09:30] <alex_joni> np
[21:13:53] <alex_joni> I don't understand what is wrong with running a program
[21:14:09] <alex_joni> hitting Esc stops the program and the spindle..
[21:14:21] <alex_joni> going to MDI and typing Sxxx doesn't start it again
[21:19:29] <alex_joni> jepler_: ever seen this: http://pastebin.ca/1187860 ?
[21:43:59] <alex_joni> I get that when I try to select a line to start from..
[21:45:58] <alex_joni> wonder if it's related to the ini changes regarding how rotation gets treated
[21:47:45] <alex_joni> jepler_: running sim/axis.ini and selecting lines further down in the program gets this error
[21:49:40] <jepler_> alex_joni: not familiar to me
[21:50:32] <alex_joni> might be display driver specific?
[21:51:21] <alex_joni> I see that it doesn't highlight the line in the preview (it used to colour it cyan iirc)
[21:52:41] <jepler_> 2.2? trunk?
[21:52:45] <alex_joni> trunk
[21:53:12] <alex_joni> freshly built as sim
[22:02:48] <alex_joni> jepler_: can I offer any other details?
[22:03:18] <alex_joni> it's an intel 945GM
[22:04:45] <alex_joni> off to bed for me, travelling tomorrow
[22:05:30] <alex_joni> cradek: your change looks ok for me.. I'm not sure what you mean that's different with running a program
[22:16:53] <cradek> run the program s300m3 / g1x2f10 / s400 / g1x0 / m2
[22:16:59] <cradek> at the s400 line the spindle speed will not change
[22:44:56] <jepler_> jepler_ is now known as jepler
[23:56:34] <skunkworks> fenn: you take crap well ;)