| 00:03.32 | *** join/#edev ico2__ (n=ico2@modem-1956.cougar.dialup.pol.co.uk) |
| 00:09.40 | *** join/#edev nhoman_ (n=homann@79.97.106.107) |
| 01:18.53 | *** join/#edev linac (n=lin@222.76.81.123) |
| 01:47.11 | *** join/#edev CP|home (n=nobody@www.crouse-house.com) |
| 02:16.08 | *** join/#edev monmotha_ (n=monmotha@66.208.186.250) |
| 02:34.37 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 02:49.38 | *** join/#edev dijenerate (n=dijenera@66.54.118.16) |
| 03:08.18 | *** join/#edev cajaka (n=none@cpe-74-67-237-9.twcny.res.rr.com) |
| 03:19.41 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 03:57.56 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 04:04.02 | *** join/#edev Appleman1234 (n=Appleman@131.181.98.192) |
| 04:07.09 | *** part/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 04:07.15 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 04:17.00 | RobotGuy-Gentoo | Lively bunch tonight |
| 04:27.49 | RobotGuy-Gentoo | sends W.A.L.T.E.R. around the channel to see if everyone is awake.. |
| 04:29.50 | Appleman1234 | Hi |
| 04:30.05 | RobotGuy-Gentoo | Hello there. |
| 04:30.16 | Appleman1234 | I am awake, wondering why the sata hdd won't spin up on my thecus n1200 |
| 04:30.24 | RobotGuy-Gentoo | What are you doing with embedded systems? |
| 04:31.02 | RobotGuy-Gentoo | Is that an SBC? |
| 04:32.13 | Appleman1234 | it is a nas |
| 04:32.35 | Appleman1234 | has an onboard flash on the motherboard that starts up fine |
| 04:32.50 | RobotGuy-Gentoo | Oh, OK. How much RAM and Flash? |
| 04:33.15 | Appleman1234 | but the hdd (which has the rest of operating system) doesn't spin up, even though the sata pins get power and the sata drive lines up with the pins |
| 04:33.54 | Appleman1234 | 128MB ram, 8mb flash |
| 04:34.03 | RobotGuy-Gentoo | Have you checked the HD with another board or a PC? Have you got the SATA drivers enabled in your kernel? |
| 04:35.48 | Appleman1234 | http://onbeat.dk/thecus/index.php/N1200_Hardware are the details, I haven't checked the HD with another PC but the supplier did and said that the hdd was fine (and that they won't fix the problem via warranty as they reckon I caused), the kernel is the default kernel which supported the hdd perfectly fine before the drive stopped getting power |
| 04:37.30 | RobotGuy-Gentoo | Don't depend on what the supplier says. Check the voltages going into the SATA drive and make sure they are right. |
| 04:38.19 | RobotGuy-Gentoo | You need to verify that the drive is good by connecting it to a different system. |
| 04:39.18 | RobotGuy-Gentoo | Why does the supplier think you caused the problem? |
| 04:39.56 | Appleman1234 | I have checked the voltages into the pins, I am currently waiting on cables so I can test the drive on a different system |
| 04:40.01 | RobotGuy-Gentoo | Check the voltages coming out of the board's sata connectopr with and without the HD connected. |
| 04:41.25 | RobotGuy-Gentoo | Is the sata cable to the HD a standard one or custom? |
| 04:41.31 | Appleman1234 | The supplier provided the device in a working state, but in installing the hdd they created a crack in a plastic in the connector, after some use of the device, the plastic crack caused a small claw to break off |
| 04:42.04 | Appleman1234 | they reckon I caused the plastic to break, when it broke due the crack mark they caused |
| 04:42.21 | Appleman1234 | RobotGuy-Gentoo: there is no cable |
| 04:42.33 | RobotGuy-Gentoo | I'd say the supplier causes the problem, not you, if that crack was there when you got the drive. |
| 04:42.36 | RobotGuy-Gentoo | No cable?? |
| 04:42.40 | Appleman1234 | RobotGuy-Gentoo: it plugs directly into the connector |
| 04:43.04 | Appleman1234 | RobotGuy-Gentoo: I say that too, but they still won't fix under warranty |
| 04:43.39 | RobotGuy-Gentoo | OK. Well, it won't be possible to help further until you ahave verified the drive is good or discover it is bad. You can still check the voltages from the board's sata connector with and without the HD connected. |
| 04:43.57 | RobotGuy-Gentoo | Who is the supplier for this HD? |
| 04:44.11 | Appleman1234 | the supplier of the device, Altech ... |
| 04:44.32 | RobotGuy-Gentoo | I get all my HDs from Newegg.Com. |
| 04:44.41 | Appleman1234 | I checked the voltages in the pins on the connector and they were fine (without hdd) |
| 04:45.00 | RobotGuy-Gentoo | How are the voltages with the HD connected?? |
| 04:45.29 | Appleman1234 | the voltages on the back of pins (they are exposed in the connector) are fine when the drive is connected, but I haven't been able to check the front as they plugs directly into the drive .... |
| 04:46.06 | RobotGuy-Gentoo | You should be able to check the voltages to the connector by looking just before the connector on the board. |
| 04:47.19 | Appleman1234 | the voltages to the connector are fine then |
| 04:47.33 | Appleman1234 | it is getting from the connector the drive that seems to be the issue |
| 04:47.58 | RobotGuy-Gentoo | OK, well now you will have to wait until you can try the HD on another system to make sure it is really OK. |
| 04:48.18 | RobotGuy-Gentoo | WHich could mean the problem is on the HD. Check the voltages just after the connector on the HD. |
| 04:48.32 | RobotGuy-Gentoo | Check for broken traces around the SATA connector on the HD. |
| 04:49.08 | Appleman1234 | Thanks |
| 04:49.24 | Appleman1234 | I will do that, when I have the cables and access to the device |
| 04:49.30 | RobotGuy-Gentoo | How are the voltages there? You're very welcome. |
| 04:51.19 | RobotGuy-Gentoo | Check both connectors for damage. |
| 04:52.00 | RobotGuy-Gentoo | I've helped as much as I can for now. I hope you find the problem and can get it fixed. |
| 04:52.08 | Appleman1234 | thanks , so do I :) |
| 05:10.49 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 05:11.02 | RobotGuy-Gentoo | returns |
| 05:15.00 | *** part/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 05:47.05 | *** join/#edev lyakh (n=lyakh@p57BD21AA.dip0.t-ipconnect.de) |
| 05:51.05 | *** join/#edev steliosk (n=Stelios@85.75.210.59) |
| 06:05.26 | *** join/#edev drath_ (i=vmaster@p5B07D7A7.dip.t-dialin.net) |
| 06:44.49 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 06:46.16 | RobotGuy-Gentoo | Dr{Who}: ping |
| 07:16.13 | *** join/#edev [AD]Turbo (n=turbo@host24-241-dynamic.22-79-r.retail.telecomitalia.it) |
| 07:18.28 | [AD]Turbo | yo |
| 07:19.30 | RobotGuy-Gentoo | oy |
| 07:30.49 | *** join/#edev monmotha__ (n=monmotha@66.208.186.250) |
| 08:09.08 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 08:34.18 | *** join/#edev Mr_Bonga (n=Rui_Silv@194.65.138.120) |
| 08:59.21 | *** join/#edev Appleman1234 (n=Appleman@131.181.98.192) |
| 09:17.35 | *** join/#edev Sascha_ (n=sisawess@cssun.rrze.uni-erlangen.de) |
| 09:18.33 | *** join/#edev ChanServ (ChanServ@services.) |
| 09:18.34 | *** join/#edev agaffney (n=agaffney@gentoo/developer/pdpc.active.agaffney) [NETSPLIT VICTIM] |
| 09:18.34 | *** join/#edev ohsix (n=ohsix@66.220.111.222) [NETSPLIT VICTIM] |
| 09:18.34 | *** mode/#edev [+o ChanServ] by irc.freenode.net |
| 09:20.09 | *** join/#edev ChanServ (ChanServ@services.) |
| 09:20.09 | *** mode/#edev [+o ChanServ] by irc.freenode.net |
| 10:00.47 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 10:02.47 | RobotGuy-Gentoo | I hate configuring X. |
| 10:05.56 | *** part/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 10:22.10 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 10:33.34 | *** join/#edev RobotGuy-Gentoo (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 11:02.03 | *** join/#edev monmotha_ (n=monmotha@66.208.186.250) |
| 11:14.26 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 11:16.24 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 11:39.14 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 11:39.14 | *** mode/#edev [+o sjhill] by ChanServ |
| 11:46.15 | *** join/#edev Appleman1234 (n=Appleman@121.222.170.32) |
| 12:24.01 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 12:24.01 | *** mode/#edev [+o GPSFan] by ChanServ |
| 12:35.09 | *** join/#edev bbradley (n=bbradley@92-235-133-146.cable.ubr17.azte.blueyonder.co.uk) |
| 13:08.39 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 13:14.34 | *** join/#edev robotguy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 13:16.56 | RobotGuy | g1 |
| 13:18.18 | RobotGuy | G'Mornin |
| 13:21.01 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:24.46 | RobotGuy | prpplague: Greetings, Salutations, and Halcinations to you. :) |
| 13:25.03 | prpplague | ho ho ho, merry freakin friday |
| 13:25.17 | RobotGuy | Rough week for you? |
| 13:35.45 | prpplague | RobotGuy: lately every week at aml is rough |
| 13:43.21 | prpplague | RobotGuy: if you speak to rusty, no need to mention my whinning |
| 13:44.02 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 13:44.46 | *** join/#edev mbuf (n=shakthim@61.16.248.242) |
| 13:54.27 | *** join/#edev sweetdream (i=sweetdre@ip70-187-210-104.dc.dc.cox.net) |
| 13:55.06 | sweetdream | hi |
| 13:55.16 | sweetdream | info? |
| 13:56.17 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 13:57.45 | chouimat|work | morning |
| 13:58.01 | sweetdream | selling cc? |
| 14:01.17 | RobotGuy | prplague: I never have mentionend anything like that. |
| 14:02.58 | RobotGuy | X is being cranky for me. It won't even run with the exact config file I was using under Kubuntu. |
| 14:07.04 | *** join/#edev dijenerate (n=dijenera@72.22.143.246) |
| 14:12.02 | RobotGuy | sweetdream: We only sell hammers and nails here. |
| 14:16.21 | prpplague | RobotGuy: actualyl we don;t sell them here, just discuss them, hehe |
| 14:16.56 | RobotGuy | prpplague: :P You are so picky. |
| 14:17.18 | prpplague | RobotGuy: hehe, wording is everything |
| 14:18.14 | RobotGuy | I managed to wipe out my /home partition when I was installing Gentoo. |
| 14:18.59 | prpplague | doh |
| 14:19.32 | RobotGuy | I was one off on the partition number. :( All e-mail gone, all scripts gone. :( |
| 14:22.53 | prpplague | RobotGuy: no backups? |
| 14:23.29 | RobotGuy | prpplague: I could never find reverse.. :P |
| 14:24.16 | RobotGuy | I wouldn't normally have formatted /home. |
| 14:27.10 | prpplague | RobotGuy: " I could never find reverse.." ? |
| 14:29.13 | RobotGuy | prpplague: Backup, reverse.. As in none. |
| 14:31.46 | g1powermac | alrighty, my plans are slowly coming together. . . |
| 14:31.55 | g1powermac | world domination is in my grasp! bwahahaha |
| 14:54.39 | g1powermac | hmm, neat little box: http://www.revogear.com/ProductDetails.asp?ProductCode=KURO-BOX%2FPRO |
| 14:57.25 | g1powermac | and only about $20 more than the retail cost of the zipit |
| 14:58.08 | GPSFan | g1powermac: the kuroboxen used to be PPC based, nice to see an ARm varient. |
| 14:58.20 | g1powermac | yea, they still have the ppc ones too |
| 15:01.15 | g1powermac | http://buffalo.nas-central.org/index.php/KuroBoxPro |
| 15:01.18 | g1powermac | full specs of the ARM one |
| 15:02.07 | GPSFan | g1powermac: thanks haven't had time to look closer |
| 15:02.29 | g1powermac | its a pretty neat little box |
| 15:03.05 | g1powermac | and since its got pci express slot in it, I'd imagine you could do quite a few different things with it |
| 15:04.36 | GPSFan | a decent amount of ram and gig-e too |
| 15:04.55 | g1powermac | http://buffalo.nas-central.org/index.php/Category:Kurobox |
| 15:05.04 | g1powermac | full comparison of the different boxen |
| 15:06.27 | g1powermac | its kinda cool that they're sold by buffalo, they made the openwrt running routers I have |
| 15:06.51 | g1powermac | I really like the routers, especially with the modified massive wifi antenna I stuck on them :-) |
| 15:13.13 | GPSFan | the original kurobox architecture was similar to the A3000 (not amiga) that several of us got a few years ago, no data ever made it onto the wiki. |
| 15:18.01 | *** join/#edev Mr_Bonga (n=Rui_Silv@194.65.138.120) |
| 16:11.40 | RobotGuy | <PROTECTED> |
| 16:13.33 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 16:23.52 | *** join/#edev monmotha_ (n=monmotha@66.208.186.250) |
| 16:31.36 | *** join/#edev bbradley (n=bbradley@92-235-133-146.cable.ubr17.azte.blueyonder.co.uk) |
| 16:33.14 | g1powermac | heya RobotGuy |
| 16:44.49 | RobotGuy | g1powermac: Hi |
| 16:45.18 | *** join/#edev Dr{Who} (n=mathewss@dev.null.nutech.com) [NETSPLIT VICTIM] |
| 16:46.00 | RobotGuy | Dr{Who}| Greets |
| 16:47.14 | g1powermac | is prepping for his new college adventure |
| 16:47.24 | g1powermac | trying to get everything together from the old college |
| 16:47.43 | RobotGuy | can't get X to run. :( |
| 16:49.21 | g1powermac | *sigh* I really need to review the math stuff, I did calculus 1 and 2 over two years ago, and I don't remember it ;-) |
| 16:49.56 | RobotGuy | If you don't use it, you lose it. |
| 16:50.08 | g1powermac | heck yea |
| 16:50.53 | g1powermac | even simple algebra is, well, interesting :-) |
| 16:51.19 | RobotGuy | I used to love doing algebra, manipulating equations, etc. |
| 16:51.44 | g1powermac | ya, I was great at the stuff, got A's in all my math courses |
| 16:51.59 | g1powermac | but man, you really do forget stuff if you don't use it in a long time |
| 16:52.12 | RobotGuy | Yep, that's for sure. |
| 16:53.12 | RobotGuy | Dr{Who}| ping |
| 16:55.18 | g1powermac | ponders about this. . . |
| 16:55.30 | g1powermac | there's 58 questions on the GRE, plus two essay questions |
| 16:55.41 | g1powermac | but on the ACT, there's over 200 questions |
| 16:56.13 | RobotGuy | I bet those two essay questions are a big part of the score of the test. |
| 16:56.38 | g1powermac | not really, the essay questions get their own grade by themselves |
| 16:57.14 | g1powermac | its upto the graduate school to want to use the essay question scores or not |
| 16:57.24 | RobotGuy | Ahhhh |
| 16:57.52 | RobotGuy | It's all in how well you can BS and cover it up so nobody knows. :) |
| 16:57.53 | g1powermac | heck, actually, its upto the graduate school to want to use the verbal or quantitative sections too |
| 16:58.28 | g1powermac | its interesting that one of the top biomedical engineering schools in the country only want to see your quantitative scores |
| 17:01.36 | *** join/#edev Dr{Who} (n=mathewss@dev.null.nutech.com) [NETSPLIT VICTIM] |
| 17:02.06 | RobotGuy | Dr{Who}: WB |
| 17:02.11 | *** join/#edev mathewss_ (n=mathewss@dev.null.nutech.com) |
| 17:04.24 | RobotGuy | Right now, I just need to get X to run so I can get on with using my PC again. |
| 17:11.19 | prpplague | g1powermac: http://imagebin.ca/view/tH3RGx.html |
| 17:12.00 | g1powermac | prpplague, oh cool |
| 17:12.17 | g1powermac | besides the windows mobile thing :-) |
| 17:12.59 | prpplague | hehe, actually thats just a default lcd image the mechanical engineer uses when doing renderings |
| 17:13.21 | g1powermac | ahhh |
| 17:14.40 | prpplague | solid works comes with that as the default image for a 32x240 lcd |
| 17:18.43 | prpplague | ~quote scoxq.pk |
| 17:27.41 | MothaWork | anybody know if it's OK to call printk from an IRQ context? |
| 17:27.45 | MothaWork | I don't know if it sleeps or not |
| 17:35.54 | *** join/#edev ce_geek (n=ce_geek@roger-blanchette-4.umelst.maine.edu) |
| 17:37.52 | prpplague | MonMotha: printk inside of irq call back is ok, as long as it doesn't create a massive delay |
| 17:38.17 | prpplague | MonMotha: i usually use two or three printks inside a irq handler |
| 17:38.34 | prpplague | MonMotha: i.e. one on entry, one on status of irq, and one on exit |
| 17:38.46 | prpplague | ce_geek: hey hey |
| 17:38.50 | prpplague | ce_geek: heard from rick yet? |
| 17:39.46 | ce_geek | prpplague: Sorta. I did get an A in the class, so apparently he liked it :) |
| 17:40.06 | prpplague | ce_geek: good deal |
| 17:40.08 | ce_geek | prpplague: Gonna go over some stuff with him this afternoon. Basically walk him throug the steps so he can experience it. |
| 17:40.17 | prpplague | ce_geek: ahh |
| 17:40.30 | prpplague | ce_geek: well give me a update when you can |
| 17:40.32 | chouimat|work | hi * |
| 17:40.45 | MothaWork | prpplague: ok, that's what I thought |
| 17:40.50 | MothaWork | at this point, I'm only calling it once |
| 17:41.13 | MothaWork | the instant the IRQ fires, the system locks up, despite the only thing my IRQ doing being disabling the IRQ and returning "IRQ_HANDLED" |
| 17:41.43 | prpplague | hmm |
| 17:41.54 | ce_geek | prpplague: Sure thing. |
| 17:42.10 | prpplague | MonMotha: hmm, sure something else isn't registering the irq? |
| 17:42.20 | MothaWork | I hope not |
| 17:42.26 | MothaWork | it's on the FIQ line of the processor |
| 17:42.32 | *** mode/#edev [+o prpplague] by ChanServ |
| 18:01.50 | MothaWork | also, if I don't register the IRQ, nothing happens when the IRQ fired |
| 18:07.50 | roxfan2 | can you check what's going on with the jtag? |
| 18:08.35 | MothaWork | I'm not set up to do that with this board |
| 18:08.43 | MothaWork | I could theoretically do so, yes |
| 18:08.54 | MothaWork | but I don't have any of the software necessary to do so set up |
| 18:09.00 | MothaWork | I don't normally write software at all |
| 18:09.25 | MothaWork | for a hint as to what I normally do, I have the ECAD software used to design this board up on one of my monitors to look up testpoints |
| 18:13.08 | prpplague | ce_geek: what do you think? http://imagebin.ca/view/tH3RGx.html |
| 18:14.13 | MothaWork | prpplague: any idea why the system would hard lock when an IRQ fires with a handler that does nothing but disable the irq (irq_disable_nosync) and return IRQ_HANDLED? |
| 18:14.44 | MothaWork | it is the FIQ line, but I don't see why that would make any difference |
| 18:15.57 | prpplague | MonMotha: sorry i can't think of anything, are you able to fire the IRQ manually? i.e. with a push button |
| 18:16.22 | MothaWork | I can fire it at will, if that's what you're asking |
| 18:16.25 | prpplague | yea |
| 18:16.27 | MothaWork | it lock sup the instant the IRQ fires |
| 18:16.34 | ce_geek | prpplague: Sliiiiiiiiiiick. |
| 18:16.43 | ce_geek | prpplague : What is it? :) |
| 18:16.49 | prpplague | MonMotha: what i would do is a small hello world kernel module, with nothing but the irq request and control |
| 18:17.01 | prpplague | ce_geek: new industrial pda soon to be released |
| 18:17.15 | MothaWork | that's just about what this module is down to now |
| 18:17.22 | prpplague | ce_geek: so far we can 24 hours of continuous use with 802.11b/g |
| 18:17.28 | prpplague | MonMotha: ahh |
| 18:17.30 | ce_geek | prpplague: Very cool. |
| 18:17.34 | MothaWork | it's loading in more code, but none of it gets called unless I do something like write to the device |
| 18:17.38 | ce_geek | prpplague: Price point set yet? |
| 18:17.44 | ce_geek | prpplague: Hackable? |
| 18:18.04 | prpplague | <ce_geek> prpplague: Price point set yet? very very expensive |
| 18:18.12 | prpplague | <ce_geek> prpplague: Hackable? very |
| 18:18.50 | ce_geek | Windows?! *gasp* |
| 18:18.51 | prpplague | ce_geek: put a barcode scanner in anything and it instantly jumps in cost by a factor of 10 |
| 18:19.09 | prpplague | <prpplague> g1powermac: http://imagebin.ca/view/tH3RGx.html |
| 18:19.09 | prpplague | <g1powermac> prpplague, oh cool |
| 18:19.09 | prpplague | <g1powermac> besides the windows mobile thing :-) |
| 18:19.09 | prpplague | <prpplague> hehe, actually thats just a default lcd image the mechanical engineer uses when doing renderings |
| 18:19.09 | prpplague | <g1powermac> ahhh |
| 18:19.09 | prpplague | <prpplague> solid works comes with that as the default image for a 32x240 lcd |
| 18:19.17 | ce_geek | prpplague: Yikes. Parts or just value? |
| 18:19.37 | prpplague | just overall retail price |
| 18:19.46 | g1powermac | ce_geek, heh, we think alike |
| 18:20.05 | prpplague | the m5900 sells for a little over $1000 and it doesn't even have wifi |
| 18:21.21 | *** join/#edev steliosk (n=Stelios@athedsl-284816.home.otenet.gr) |
| 18:30.34 | MothaWork | prpplague: is this even the correct way to ask for an interrupt on the rising edge of FIQ: "request_irq(AT91_ID_FIQ, irq_handler_func, IRQF_TRIGGER_RISING, myname, mycookie)"? |
| 18:34.59 | prpplague | MonMotha: that looks correct, assuming all of the arguments are defined properly |
| 18:35.19 | prpplague | it is possible that "myname" and "mycookie" are not defined properly |
| 18:35.38 | prpplague | or that the irq function define isn't correct |
| 18:36.07 | prpplague | short paste |
| 18:36.10 | prpplague | +static int gpiomatrix_irq(int irq, void *id) |
| 18:36.10 | prpplague | +{ |
| 18:36.10 | prpplague | + struct device *dev = (struct device *)id; |
| 18:36.10 | prpplague | + genmatrix_report_irq(dev); |
| 18:36.10 | prpplague | + return IRQ_HANDLED; |
| 18:36.10 | prpplague | +} |
| 18:36.12 | prpplague | + |
| 18:36.33 | prpplague | MonMotha: that basically how you have your irq handler defined? |
| 18:44.35 | *** join/#edev sweetdream (i=sweetdre@ip70-187-210-104.dc.dc.cox.net) |
| 18:46.35 | *** join/#edev dijenerate (n=dijenera@72.22.143.246) |
| 18:49.56 | MothaWork | prpplague: static irqreturn_t irq_handler_func(int irq, void *handle) { disable_irq_nosync(irq); return IRQ_HANDLED; } |
| 18:50.40 | MothaWork | "myname" is "spi->dev.driver->name" from the spi "probe" function, and "mycookie" is a pointer to a static struct I use to keep track of internal state (so it should be unique) |
| 18:51.33 | MothaWork | I've also tried putting the literal AT91_ID_FIQ into disable_irq_nosync, and it does the same thing |
| 18:52.39 | ds2 | maybe FIQ support is broq? |
| 18:55.26 | MothaWork | that would be very, very bad |
| 18:55.45 | ds2 | are there anything else that uses FIQ's? |
| 18:55.49 | MothaWork | because that means I have to change the hardware to get the line re-routed to a different IRQ, none of which are available |
| 18:55.59 | prpplague | MonMotha: hmm |
| 18:56.02 | MothaWork | I have no idea. Nothing else on this board is on FIQ |
| 18:56.08 | ds2 | no, I mean FIQ support broq in the kernel |
| 18:56.26 | MothaWork | the rationale for using FIQ is that the interrupt handler itself is actually very, very small. All it is going to do is touch off an async SPI transfer and return |
| 18:56.31 | prpplague | MonMotha: might try changing your request to this |
| 18:56.49 | MothaWork | it's also somewhat time-critical, since I have to get the data off the chip before more arrives |
| 18:56.53 | ds2 | right, but don't see many drivers using FIQ's |
| 18:57.01 | prpplague | "request_irq(AT91_ID_FIQ, irq_handler_func, IRQF_TRIGGER_RISING, "mytestdev", irq_handler_func)" |
| 18:57.07 | prpplague | just for testing |
| 18:57.40 | prpplague | MothaWork: just to rule out some null pointer or other issue in the your structs |
| 18:57.56 | MothaWork | I guess |
| 19:01.17 | MothaWork | gah, now it's hard locking the instant I load the driver |
| 19:01.26 | MothaWork | and this system takes FOREVER to boot |
| 19:01.34 | MothaWork | mostly because it's got tons of crap on it we don't need |
| 19:02.26 | ds2 | could the interrupt be firing before things are fully registered? i.e.enabling hardware to send FIQ prior to installing the handler? |
| 19:03.05 | MothaWork | well, the hardware sends interrupts whenever it darned well feels like it, but it's an edge triggered interrupt, and the kernel masks the interrupt until it's requested (that I've determined, things don't lock up if I never request the interrupt) |
| 19:03.31 | MothaWork | and still locks when I fire the interrupt |
| 19:03.55 | ds2 | are there erratas with the eddge mode? |
| 19:03.57 | ds2 | edge |
| 19:04.25 | MothaWork | I've tried level mode, too |
| 19:04.35 | MothaWork | let me see if I can route this to a standard IRQ |
| 19:22.24 | ce_geek | prpplague: Grumble. Tried as an exercise to disable the kernel debug messages in the kernel and compile a kernel and use OpenOCD, but we didn't get the kernel stuff to go away. |
| 19:23.05 | ce_geek | prpplague: Is this the line that needs to be changed? CONFIG_DEBUG_S3C2410_UART=0 |
| 19:24.40 | prpplague | ce_geek: actually the debug stuff you are seeing is part of the inittab file |
| 19:24.45 | prpplague | ce_geek: in the rootfs |
| 19:26.55 | MothaWork | ds2: and the answer is...don't use FIQ |
| 19:26.59 | prpplague | ce_geek: but you can turn that off as well in the kernel as well |
| 19:27.02 | prpplague | MonMotha: hehe |
| 19:27.16 | MothaWork | unfortunately, that requires a board mod to move |
| 19:27.36 | MothaWork | fortunately, we don't have thousands of these sitting around, so we may be able to just rev it. There's one other known issue (floating ground pin) |
| 19:27.47 | MothaWork | package has two ground pins, so it works, but I hate to leave ground floating like that |
| 19:28.17 | ce_geek | Hrm. Which option controls this in the kernel config? |
| 19:29.38 | RobotGuy | MothaWork: Is this with a Samsung processor? |
| 19:31.56 | ds2 | MothaWork: have you determined it is a hardware errata or a kernel problem? |
| 19:33.46 | MothaWork | RobotGuy: AT91 |
| 19:33.49 | MothaWork | ds2: no idea |
| 19:34.06 | MothaWork | I don't believe there is any documented errata on the processor relating to this |
| 19:34.09 | MothaWork | let me pull up the docs |
| 19:35.21 | ds2 | which AT91? |
| 19:38.44 | cbrake | prpplague: do you have any pointers on how to configure a 8686 wifi device for a WPA network? I'm looking through http://wiki.debian.org/WiFi, but iwlist is returning "no private ioctls" |
| 19:39.17 | cbrake | prpplague: "iwlist eth1 scanning" works fine |
| 19:39.34 | ds2 | wpa_supplicant? |
| 19:42.49 | prpplague | cbrake: yea you need wpa_supplicant t do wpa |
| 19:43.04 | cbrake | ds2: prpplague: ok, thanks |
| 19:43.13 | prpplague | RobotGuy: ce_geek is using the hammer kit |
| 19:45.17 | RobotGuy | prpplague: I know. |
| 19:52.49 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 19:55.24 | *** join/#edev MothaWork (n=monmotha@66.208.186.250) |
| 20:00.37 | cbrake | prpplague: http://pastebin.ca/1020429, I'm now getting a ioctl[SIOCSIWAUTH]: Operation not supported |
| 20:28.00 | prpplague | cbrake: sounds like your either your wpa_supplicant doesn't have the marvell support or the marvel driver you are using doesn't have wpa support |
| 20:28.18 | prpplague | cbrake: did you patch your wpa_supplicant? |
| 20:28.29 | prpplague | cbrake: i assume you are using the marvell reference driver right? |
| 20:28.37 | prpplague | cbrake: not the libertas driver |
| 20:34.16 | g1powermac | prpplague & GPSFan, hows the zipit hacking? |
| 20:34.44 | prpplague | has been too busy |
| 20:34.57 | cbrake | prpplague: no, I'm using the libertas driver |
| 20:35.03 | prpplague | cbrake: ahh |
| 20:35.04 | g1powermac | k |
| 20:35.17 | prpplague | cbrake: then the standard wpa_supplicant should work |
| 20:35.34 | cbrake | prpplague: I have the source for the marvell driver, but would like to avoid then unless there is a good reason to use the Marvell driver |
| 20:35.52 | prpplague | cbrake: indeed |
| 20:36.08 | prpplague | cbrake: can you pastebin.ca your conf file for wpa? |
| 20:36.10 | cbrake | prpplague: have you observed any benefits to the libertas driver over the marvell one? |
| 20:36.14 | cbrake | prpplague: sure |
| 20:36.25 | g1powermac | ponders how the libertas driver is shaping up for the wireless chip in the zipit. . . |
| 20:36.30 | prpplague | cbrake: its open source and much more rebust |
| 20:36.46 | prpplague | g1powermac: we are talking about the same chip |
| 20:36.51 | g1powermac | oh, cool |
| 20:37.11 | g1powermac | so its coming along well then :-) |
| 20:37.31 | prpplague | g1powermac: libertas support for the 8686 in sdio mode is fully functional |
| 20:37.41 | g1powermac | awesome |
| 20:37.44 | cbrake | prpplague: http://pastebin.ca/1020489 |
| 20:38.07 | g1powermac | so now we have nothing stopping us from upgrading the kernel in the zipit now |
| 20:38.49 | prpplague | g1powermac: no, the zipit2 need GSPI support |
| 20:38.54 | prpplague | cbrake: interesting |
| 20:39.11 | g1powermac | prpplague, doh, yea |
| 20:39.25 | prpplague | cbrake: its like either the wpa_supplicant is using the wrong driver interface, or the libertas driver doesn't have wpa support |
| 20:39.41 | cbrake | prpplague: so this should all work? |
| 20:40.22 | prpplague | cbrake: yea |
| 20:40.37 | cbrake | prpplague: ok, I'll keep poking around then |
| 20:41.35 | g1powermac | ponders how the gspi support is coming for the zipit |
| 20:42.07 | prpplague | cbrake: what kernel version are you using? |
| 20:43.12 | prpplague | g1powermac: someone posted a bunch of the wince driver code for the GSPI support |
| 20:43.16 | cbrake | prpplague: 2.6.25 |
| 20:44.02 | g1powermac | prpplague, yea, I heard that |
| 20:44.56 | prpplague | cbrake: http://osdir.com/ml/linux.laptop.olpc.devel/2007/msg01288.html |
| 20:45.31 | prpplague | cbrake: you are getting a similar problem which leads me to believe that either your wpa_supplicant is old, or you have something not enabled for your kernel driver |
| 20:47.24 | cbrake | prpplague: ok, thanks for the pointers. I'll keep digging into this -- has to be something simple ... |
| 20:50.54 | prpplague | np |
| 21:16.55 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 21:20.24 | MothaWork | prpplague: hey, are there any good generic queue structures in the kernel? I have a bunch of incoming data, and I need to have my "read" FileOp wait on the queue to have data, then consume it from the queue when the user wants it |
| 21:25.16 | prpplague | there are some generic work queues , not sure if thats what you are looking for |
| 21:25.35 | MothaWork | no, I'm looking to actually queue up data for later consumption as requested by userspace |
| 21:25.52 | MothaWork | I'm actually using a workqueue |
| 21:25.57 | MothaWork | for something else, though |
| 21:29.20 | prpplague | ahh |
| 21:29.34 | prpplague | sorry not something i've had usage for |
| 21:36.35 | MothaWork | gotcha |
| 21:36.45 | MothaWork | well, my solution for now is that userspace just has to consume the data as fast as it comes in |
| 21:36.49 | MothaWork | otherwise you lose a frame |
| 21:36.51 | MothaWork | that's ok |
| 21:36.53 | MothaWork | for now |
| 21:37.04 | ds2 | MothaWork: RT patches ;) |
| 21:37.26 | ds2 | MothaWork: which AT91 is this? |
| 21:37.30 | MothaWork | this is a character driver for an 802.15.4 radio (though we're using it as a fancy wireless UART, there's no MAC) |
| 21:37.34 | MothaWork | AT91SAM9263 |
| 21:38.46 | ds2 | is this one of those things you just hook up to a ChipCon chip? |
| 21:41.20 | MothaWork | it's actually an Atmel radio |
| 21:41.44 | MothaWork | it's meant to run IEEE 802.15.4 (the underlying PHY and MAC of ZigBee), but we only have the PHY stuff at this point, no MAC |
| 21:41.59 | MothaWork | seems to have pretty good RF performance, though |
| 21:42.13 | ds2 | *nod* TI was talking/showing those at ESC |
| 21:42.21 | ds2 | (ti bought chipcon) |
| 21:43.42 | MothaWork | yeah, I looked at the chipcon stuff, too |
| 21:44.05 | MothaWork | the Atmel radio has better RF specs, though I haven't tested either chip to be sure that it meets the published specs (I lack the equipment required) |
| 21:44.13 | g1powermac | zigbee is interesting. . . |
| 21:44.45 | ds2 | heh... know if Atmel made a raw radio chip like Chipcon CH2500? there is a neat project using one of those a 2.4GHz spec an |
| 21:45.20 | MothaWork | this chip is pretty raw |
| 21:45.23 | MothaWork | the MAC is software |
| 21:45.38 | ds2 | does it give you an RSSI value? |
| 21:45.45 | MothaWork | so about the only part of 802.15.4 that is enforced by the chip is the modulation and the preamble |
| 21:45.47 | MothaWork | I believe so, yes |
| 21:46.03 | ds2 | for receive that won't matter |
| 21:46.29 | ds2 | unless it wants to demod before providing an RSSI |
| 21:46.33 | g1powermac | could use to find an ultra low power really small wireless transmitter that only needs to transmit for about 6-8 ft |
| 21:46.52 | ds2 | g1powermac: the ez430-RF kit! |
| 21:47.00 | MothaWork | I don't know if you can request an RSSI measurement independent of a frame reception |
| 21:47.12 | g1powermac | ds2, hmm, I'll look into that |
| 21:47.15 | ds2 | MSP430 + CH2500 on a premade board... you get 2 of those for $50 |
| 21:48.03 | g1powermac | hmm, still too big |
| 21:48.15 | ds2 | okay, what about the linx stuff? |
| 21:48.17 | g1powermac | I need something that can run off a coin cell type battery |
| 21:48.32 | g1powermac | and be about the size of one too |
| 21:48.41 | ds2 | the ez430-RF can run off a coin battery...it is just prewired for 2 AA's |
| 21:48.42 | SpeedEvil | g1powermac: premade, or are you willing to cook it? |
| 21:49.03 | g1powermac | SpeedEvil, depends, obviously premade is always nice :-) |
| 21:49.34 | g1powermac | my actual goal is to make wireless eeg/ecg sensors that can be received by my zipit |
| 21:49.41 | SpeedEvil | freescales stuff otherwise |
| 21:49.48 | SpeedEvil | look under zigbee |
| 21:49.49 | g1powermac | so the smaller and less cumbersome the better |
| 21:50.04 | SpeedEvil | they've got $3 or so chips with integrated MCU and zigbee phy |
| 21:50.16 | g1powermac | hmm |
| 21:50.36 | g1powermac | that would be interesting to have MCUs in them as well |
| 21:51.05 | g1powermac | they could theoretically do some filtering right on the sensor instead of shifting it off to the zipit side |
| 21:51.27 | SpeedEvil | yeah |
| 21:51.38 | SpeedEvil | few tens of K of RAM and ROM |
| 21:51.51 | g1powermac | checks out freescale |
| 21:53.25 | g1powermac | hmm, which ones have MCUs in them? |
| 21:53.52 | ds2 | g1powermac: have you looked at the linx chips then? |
| 21:54.01 | g1powermac | ds2, no. . . |
| 21:54.07 | g1powermac | got part numbers? |
| 21:54.10 | ds2 | look up linx technology |
| 21:54.18 | ds2 | Digikey carries them... all ISM chips |
| 21:54.32 | g1powermac | k |
| 21:57.12 | g1powermac | ahh, finally found the mcu+zigbee freescale chips |
| 22:00.07 | g1powermac | sadly it looks like they're all bgas |
| 22:01.31 | g1powermac | looks like linx's site is down atm |
| 22:18.56 | g1powermac | hmm, the linx stuff is interesting |
| 22:29.44 | g1powermac | hmm, you know what would be interesting is to have something like wireless I2C |
| 22:35.25 | g1powermac | I am really liking the linx hp3 series, it can even transmit pure analog signals. . . |
| 22:35.35 | g1powermac | would making the sensors super simple |
| 22:35.40 | g1powermac | *make |
| 22:36.13 | ds2 | is that a chip or a module? |
| 22:37.00 | g1powermac | the module |
| 22:39.14 | g1powermac | though its prolly not the best to use it to transmit the analog signal from the sensors directly |
| 22:39.39 | g1powermac | if I combine the module up with a simple 8051, I'd have a nice pair |
| 22:39.48 | g1powermac | though i wonder how power efficient that would be |
| 22:40.33 | ds2 | MSP430F2xxx series |
| 22:40.40 | ds2 | lower power then most 8051's |
| 22:40.57 | g1powermac | its gotta have a really good adc on it though |
| 22:41.14 | ds2 | the 2013 I think has a very nice ADC |
| 22:41.20 | g1powermac | k |
| 22:41.49 | ds2 | those 430day free seminar/boards are earning their keep ;) |
| 22:42.51 | g1powermac | hmm, so this module takes about 14ma. . . |
| 22:43.10 | ds2 | constant or peak? 14mA is huge |
| 22:43.19 | g1powermac | typical |
| 22:43.51 | ds2 | yikes the smaller stuff like a CR2016 would at best power it for an hour or so |
| 22:44.02 | g1powermac | yea |
| 22:44.30 | g1powermac | though I wonder if the power amount changes depending on how much power you let it use to transmit |
| 22:47.08 | g1powermac | ok, if I go with the LC series module which is really basic, the power goes down to 3ma |
| 22:47.29 | g1powermac | of course then i need a decorder/encoder to use the thing |
| 22:49.35 | ds2 | how quickly does it go from cold no power to ready to transmit/receive? |
| 22:50.05 | g1powermac | which one? |
| 22:50.17 | ds2 | either one |
| 22:50.27 | ds2 | cuz if they are fast, maybe you can switch power to it from the MCU |
| 22:51.00 | g1powermac | looks like 7ms |
| 22:52.59 | g1powermac | actually, the max time is 10ms |
| 22:53.46 | g1powermac | ponders about the feasibility of that |
| 22:55.41 | g1powermac | euh, just noticed these suckers are really expensive |
| 22:55.46 | g1powermac | about $43 a piece |
| 22:56.11 | g1powermac | oh wait, thats the receiver |
| 22:56.19 | ds2 | what frequency are these? |
| 22:56.19 | g1powermac | the transmitter is cheaper at $30 a piece |
| 22:56.26 | g1powermac | 900mhz |
| 22:56.35 | ds2 | I seem to recall the ones i played with were $15ish from digikey |
| 22:56.53 | g1powermac | well, they do have those, problem is that they're single channel |
| 22:56.58 | ds2 | Hmmm the ones I played with ere the 433MHz ones |
| 22:57.01 | ds2 | ah I see |
| 22:57.01 | g1powermac | I need multiple channels |
| 22:57.14 | g1powermac | lots of transmitters to one receiver |
| 22:59.47 | g1powermac | takes a look at atmel's zigbee offerings |
| 23:00.32 | g1powermac | they do make nice avr + zigbee chips, hmm |
| 23:01.35 | g1powermac | even though I'm not a huge fan of avrs. . .this might still work nicely |
| 23:04.59 | ds2 | g1powermac: what about the Nordic semi stuff? |
| 23:08.10 | g1powermac | hmm, what makes the nordic stuff better than zigbee? |
| 23:08.28 | ds2 | they are generic radios |
| 23:08.46 | ds2 | and I do not believe in zigbee? :) |
| 23:08.48 | SpeedEvil | seperate radio - i2c or spi |
| 23:08.56 | g1powermac | ahh |
| 23:09.14 | ds2 | if you want a precanned solution can try your luck with ANT |
| 23:09.41 | ds2 | they make little sensor transceivers out of the Nordic's but I have had mixed experience with their software interface |
| 23:10.07 | g1powermac | was looking at the nordic stuff on sfe |
| 23:17.42 | g1powermac | these avr+zigbee chips though are really interesting |
| 23:19.13 | g1powermac | hmmmmmmm |
| 23:19.23 | g1powermac | ti makes an 8051+zigbee chip. . . . |
| 23:19.33 | g1powermac | and it has a nice adc in it too |
| 23:19.43 | g1powermac | http://focus.ti.com/docs/prod/folders/print/cc2430.html |
| 23:24.40 | g1powermac | http://focus.ti.com/docs/prod/folders/print/cc2431.html |
| 23:24.44 | g1powermac | now thats interesting |
| 23:28.06 | g1powermac | its a bit more logic than I wanted to use though, however, I could use these chips to gather multiple signals in a general region on the body and then transmit to the zipit |
| 23:28.25 | g1powermac | I could then use larger batteries |
| 23:33.27 | *** join/#edev omero (n=omero@host30-30-dynamic.182-80-r.retail.telecomitalia.it) |
| 23:35.29 | omero | sigh... ARM? anything else is OT... uh? |
| 23:35.46 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 23:35.55 | g1powermac | hey RobotGuy |
| 23:36.05 | RobotGuy | Hi |
| 23:36.26 | g1powermac | found very interesting 8051s. . . |
| 23:36.32 | g1powermac | http://focus.ti.com/docs/prod/folders/print/cc2431.html |
| 23:36.35 | RobotGuy | Really? |
| 23:36.44 | g1powermac | 8051s with zigbee |
| 23:37.16 | RobotGuy | Ooooooh, interesting |
| 23:37.27 | g1powermac | and a very nice adc |
| 23:38.09 | g1powermac | sees multi-sensor wireless modules that'll process and send eeg/ecg/emg data from the body to the zipit for display and further processing |
| 23:38.47 | g1powermac | and with the location sensing system on it, I see some further interesting features of it tracking extremity movement |
| 23:38.52 | ds2 | g1powermac: what kind of dev tools do those use? |
| 23:38.56 | ds2 | dev tool as in programmer |
| 23:39.11 | g1powermac | for coding, sdcc should work |
| 23:39.14 | RobotGuy | ACK, they are QFN. |
| 23:39.20 | g1powermac | to load code, I'll have to check |
| 23:39.34 | ds2 | ack ... I mean downloading/burning/flashing/etc |
| 23:39.52 | g1powermac | usually its simple serial port stuff |
| 23:39.57 | RobotGuy | Would be good for any wireless sensor type application, including robotics. |
| 23:40.04 | ds2 | 'usually'? He HE HE HE HE |
| 23:40.17 | ds2 | I am spoiled by the Silabs toolstick |
| 23:40.23 | g1powermac | it depends on the manufacturer |
| 23:42.36 | omero | gentlemen, a plase where to ask buildroot (generic) questions? |
| 23:42.59 | omero | s/plase/place/g |
| 23:43.05 | ds2 | omero: ask.. if someone knows... |
| 23:43.07 | omero | gh |
| 23:43.19 | omero | i'm having a problem |
| 23:46.11 | omero | gcc is not working in the target |
| 23:46.28 | omero | ld complains it cannot find libgcc |
| 23:46.50 | ds2 | buildroot on the target? never seen that before |
| 23:48.39 | omero | mm... I'm dazed and confused too |
| 23:51.11 | omero | this has surely nothing to do with busybox |
| 23:51.55 | omero | I just need a target system uClibc-based with binutuls and a compiler |
| 23:54.36 | RobotGuy | omero: Are you saying you have a native toolchain on the target? |
| 23:55.19 | omero | I doubt it |
| 23:56.00 | omero | although that would be the goal ;) |
| 23:56.06 | RobotGuy | You mentionded gcc not working on the target. |
| 23:57.01 | RobotGuy | Do you have the c++ options enabled in your toolchain config? |
| 23:57.32 | g1powermac | ok, this chip is awesome |
| 23:57.34 | omero | nope... |
| 23:57.35 | RobotGuy | What is the target processor you are working with? |
| 23:58.01 | omero | RobotGuy: that's the fun part... plain i486 |