**** BEGIN LOGGING AT Wed Mar 27 02:59:58 2013 Mar 27 03:04:26 jkridner: ping Mar 27 03:06:08 don't you guys *ever* sleep Mar 27 03:06:09 ? Mar 27 03:07:10 ka6sox-faraway: no Mar 27 03:07:12 hehe Mar 27 03:08:50 pong Mar 27 03:09:52 prpplague: pong Mar 27 03:11:28 well the bus feature is useless for this part.. Mar 27 03:13:11 koen: did you know the latest GateOne is back to giving an infinite redirect loop? Mar 27 03:24:48 * ka6sox puts on his flameproof suit before sending PRs to the kernel repos :P Mar 27 03:42:08 mranostay: did you get your uart board ordered? Mar 27 03:43:37 can I buy a BBT NXP HDMI cape today? Mar 27 03:48:12 sigh... pwm is a long string of patches on the 335x Mar 27 03:57:35 . Mar 27 04:03:53 .. Mar 27 04:33:31 ... Mar 27 04:49:14 woof. Mar 27 05:26:36 <_av500_> miaow Mar 27 06:17:46 Russ: new thing for you to mock :P Mar 27 06:21:20 and _av500_ Mar 27 06:23:29 prpplague: yeah got those UART capes Mar 27 06:23:43 had to rework them of course Mar 27 06:24:27 * Russ checks Mar 27 06:25:07 so is that some sort of voltage ladder? Mar 27 06:25:23 where? Mar 27 06:25:35 that IC at the bottom? Mar 27 06:25:52 sot23, chip component, sot23 chip component, sot23... Mar 27 06:26:25 Russ: boost converter Mar 27 06:26:33 where does the via with the plus sign go? Mar 27 06:26:54 Russ: test point Mar 27 06:27:03 make it bigger? Mar 27 06:27:20 TP to GND so i can write to the EEPROM Mar 27 06:27:34 ah, just for you Mar 27 06:27:53 typical for most capes now Mar 27 06:28:18 move the vias around underneath the eeprom so that there is a short path between the decoupling cap ground and eeprom ground Mar 27 06:28:42 mranostay: pls pm me your home address, I need to ship something to you Mar 27 06:28:50 this in fact: http://www.history.didaktik.mathematik.uni-wuerzburg.de/ausstell/winkelmesser/geodreieck_gedreht.gif Mar 27 06:28:59 * mranostay is scared Mar 27 06:29:06 I will highlight a certain angle Mar 27 06:29:23 * Russ tries to locate the non-45 trace Mar 27 06:29:40 power rails don't count right? Mar 27 06:29:46 op, there they are Mar 27 06:29:51 sda/scl Mar 27 06:30:06 on the pullups? come on :) Mar 27 06:30:10 the component on the lower left Mar 27 06:30:17 the orientation marker is likely wrong Mar 27 06:30:22 unless its an led Mar 27 06:30:25 is it an led? Mar 27 06:30:29 yes Mar 27 06:30:32 ok Mar 27 06:31:10 mranostay: the 2 traces on the bottom Mar 27 06:31:30 to the EEPROM? Mar 27 06:31:34 ya Mar 27 06:31:57 I do wonder why you don't ground pour the bottom, power pour the top Mar 27 06:32:08 should i? Mar 27 06:32:13 it doesn't matter Mar 27 06:32:22 it makes it easier to route stuff for me at least Mar 27 06:32:38 mranostay: also, add a via in that GND-less hole about the left IC Mar 27 06:32:47 saves the environment Mar 27 06:32:49 also, if you oshpark, laen would prefer copper pours Mar 27 06:33:17 there isn't any unconnected copper Mar 27 06:33:23 av500: via to gnd? Mar 27 06:33:30 and there is no gnd on the bottom :) Mar 27 06:33:52 mranostay: I assume the bottom was filled too Mar 27 06:33:55 assumed Mar 27 06:33:56 nope Mar 27 06:33:57 non? Mar 27 06:34:01 ew Mar 27 06:34:02 non Mar 27 06:34:13 all that copper going into the creek Mar 27 06:34:19 hehe Mar 27 06:34:29 if its on a mixed panel, it makes process control difficult Mar 27 06:34:31 hey i'm a sw guy mostly.. i'm learning :P Mar 27 06:34:36 sure Mar 27 06:35:06 also, there is no bacon Mar 27 06:41:13 need to find a CCCP eagle logo Mar 27 06:41:18 Russ: better? Mar 27 06:42:48 I don't see a logo yet Mar 27 06:42:56 and I need to go to sleep Mar 27 06:53:56 I need to go to work Mar 27 06:53:58 till later Mar 27 07:01:14 i need to stock take cape connectors Mar 27 07:01:14 :-) Mar 27 07:01:16 lol. Mar 27 07:01:29 stowaht? Mar 27 07:07:30 okay I am done playing with Irish Snakes for the night... Mar 27 07:48:19 mranostay: ping Mar 27 07:48:43 about to sleep but pong Mar 27 07:50:28 g+ Mar 27 08:16:18 morn Mar 27 08:24:32 ing Mar 27 08:25:36 av500, are you sure you want to appear to collude with a greek person? Mar 27 08:25:49 and being serbian to boot... tsch tsch.. Mar 27 08:47:35 lo Mar 27 08:48:37 gin: Mar 27 08:49:11 opless channels can be a pain Mar 27 08:50:27 aholler: what is the problem? Mar 27 08:51:55 panto: aint we friends? Mar 27 08:52:04 united against the turks Mar 27 08:52:11 keesj: you don't see join/quit-messages, do you? Mar 27 08:52:22 sure, you go first Mar 27 08:52:42 after coffee Mar 27 08:52:43 I am and looking at http://arun.wordpress.com/2009/09/27/ignore-channel-joinpart-messages-in-irssi/ as we speak Mar 27 08:53:07 I don't want to ignore them Mar 27 08:53:24 there is the problem again ;) Mar 27 08:53:43 jkridner has op Mar 27 08:53:50 I dont Mar 27 08:55:03 I've recently seen a macosx-user who wasn't aware that his mac awakes every hour, logged in to irc, gone to sleep for an hour all the time Mar 27 08:56:15 but he didn't understood irc at all. Mar 27 08:56:28 what is irc? Mar 27 08:56:35 aholler: did you look into ti-rtos? Mar 27 08:56:44 I guess you love xdc already Mar 27 08:56:55 very ritsy Mar 27 08:57:01 no, you've said only cc*, so I don't need to look at it ;) Mar 27 08:57:15 aholler: I didnt say that Mar 27 08:57:22 you dont need ccs Mar 27 08:58:06 [13:35] < av500> but xdc and ti cc Mar 27 08:58:16 ok,I'm rwrong Mar 27 08:58:28 but ti cc isn't an option too ;) Mar 27 08:58:47 aholler: hmm Mar 27 08:58:50 not sure it needs ti cc Mar 27 08:59:10 gcc should work too Mar 27 09:00:56 ok, I will start to download it. we'll see what happens with the rgb-led of the stelaris launchpad ;) Mar 27 09:01:06 rt-blink Mar 27 09:02:19 that needs hard RT Mar 27 09:03:23 aholler: I guess it will need some tweaking to use another toolchain Mar 27 09:03:28 like linker scripts Mar 27 09:03:39 dotn know how gcc compatible TI is these days Mar 27 09:14:57 I hopethey have a rgb-led-driver included ;) Mar 27 09:23:03 I wonder how "zero-latency interrupts do work with broken FIQ's ;) Mar 27 09:24:06 :) Mar 27 09:24:16 I think IRQ has the same latency Mar 27 09:25:39 you just dont have the shadow registers Mar 27 09:26:30 there's no such thing as zero-latency interrupts :) Mar 27 09:26:46 FIQ is a holdover from the old days of ARM Mar 27 09:27:03 don't think of it an IRQ, think of it as a way to do fake DMA Mar 27 09:27:51 well, its a 2nd interrupt vector Mar 27 09:27:52 it was just a joke. I don't even know what's broken with the FIQ. I've just played "polly wants a cookie" ;) Mar 27 09:28:06 aholler: on am335x FIQ is broken Mar 27 09:28:16 something about the "secure" variant Mar 27 09:28:33 I wish we had trusted boot on the bone Mar 27 09:29:01 that I remember having heard, therefor the parrot-joke ;) Mar 27 09:29:08 I guess they use the FIQ for the secure side Mar 27 09:31:08 yeah Mar 27 09:31:23 FIQ is very nice (if you can handle it's craziness) Mar 27 09:33:34 speaking of which, what is the point of 'secure domain' in a community board? Mar 27 09:33:54 breaking gpl Mar 27 09:38:06 panto: there is non Mar 27 09:38:08 none Mar 27 09:38:13 bone has the GP part Mar 27 09:38:17 not the secure one Mar 27 09:38:33 but as usual, secure stuff breaks features on GP parts Mar 27 09:38:36 yeah, but FIQ is still b0rken Mar 27 09:38:45 since, nobody would ever use the GP parts Mar 27 09:38:55 are you sure? Mar 27 09:39:14 I wouldn't trust anything to be 'secure' that run on s/w on the general purpose processor Mar 27 09:39:18 with omap parts, the usual reaction from ti was "oh, you are on GP" Mar 27 09:39:28 if the GP parts are cheaper, that what I'd use Mar 27 09:39:41 of course Mar 27 09:40:05 but the mindset seems to be "make secure work" Mar 27 09:40:09 so all this gnashing about 'secure' is bogus, since customer would always go for the cheaper part Mar 27 09:40:17 panto: but no Mar 27 09:40:22 handset people need secure Mar 27 09:40:30 otherwise you are crashing the AT&T network.... Mar 27 09:40:47 horrible things can happen Mar 27 09:40:58 cats and dogs living together? mass hysteria? Mar 27 09:41:08 yes Mar 27 09:41:29 so the secure people get to mess with the silicon, fucking up GP in the process Mar 27 09:42:03 if you base your wireless infrastructure on client devices 'behaving' your are clueless IMO Mar 27 09:42:28 maybe that worked when h/w design was super expensive, but now? Mar 27 09:42:39 ti-rtos includes telnet. that reminds me at http://internetcensus2012.bitbucket.org/ ;) Mar 27 09:43:40 aholler, Allegro lol Mar 27 09:44:05 used that crap almost what? 13 years ago Mar 27 09:44:13 don't know if they had a different name then Mar 27 09:44:24 ? Mar 27 09:44:33 Allegro ROM pager Mar 27 09:44:36 in the paper Mar 27 09:45:00 ah Mar 27 10:05:23 ah Mar 27 10:05:30 the red equal sign is not bacon Mar 27 10:09:32 now I have a 360mb executable ... Mar 27 10:14:26 jkridner: I did some cleanup in the flash scripts to make them safe to use on the host as well Mar 27 10:14:51 k Mar 27 10:15:13 tightly integrating irc in multi-protocol chat- or messaging-apps isn't really what should be done. there are are to much users which can't handle it ;) Mar 27 10:41:32 Hi guys! I was wondering if somebody has experience with cross-compiling code for the PRUSS and in particular the am335x_pru_package Mar 27 10:42:49 not native compiling? :) Mar 27 10:44:24 lol Mar 27 10:51:37 technically it's impossible since there is no compiler for PRU, only an assembler Mar 27 10:54:23 that's an obstacle at most Mar 27 10:54:57 any CS student who is worth half the books he owns can write a compiler in a weekend or two ;) Mar 27 10:57:47 Lol Mar 27 10:58:20 Well, let me make myself more clear Mar 27 10:58:49 the am335x_pru package has a driver that allows you to load the assembled code on the PRU and run it Mar 27 10:58:59 yes Mar 27 10:59:07 I don't necessarily want to do that, but in order to follow that way of execution I need to cross-compile said driver Mar 27 10:59:23 ah Mar 27 10:59:27 I am open also to the pathway of using the assembler and manually running the resulting code Mar 27 10:59:38 but I have no idea how to manually run the code Mar 27 10:59:41 you want to compile the PRU loader Mar 27 10:59:46 yes Mar 27 10:59:47 not the code running in the PRU Mar 27 10:59:53 or learn how to manually run PRU code without it Mar 27 10:59:54 yes Mar 27 11:00:07 I would imagine it makes no difference where you run the assembler Mar 27 11:00:12 as it would give the same result either way Mar 27 11:00:17 KotH: a c compiler is only a glorified macro assembler anyway Mar 27 11:00:39 apropos.. does anyone have the url to a good cross compilation tutorial at hand? Mar 27 11:00:47 av500: wasn't that forth? Mar 27 11:01:40 So, any suggestions :) Mar 27 11:02:09 * KotH had recently someone tell him, that forth is so much better for uC's than c, because with c you never know what the compiler does while with forth it's crystal clear Mar 27 11:02:15 KotH: what do you want to cross-compile? Generally you just need a good toolchain and off you go Mar 27 11:02:22 Stefan__: google for linux cross compilation Mar 27 11:02:38 Stefan__: oh.. i wanted an url to give to guys like you :) Mar 27 11:03:09 Stefan__: it's an question that pops up every once in a while so a good tutorial would come handy Mar 27 11:03:24 KotH: I have cross compiled a kernel for the BB, so I am not an absolute noob to cross-compilation, I was mostly wondering what environment variables I need to specify for correct cross compilation Mar 27 11:03:42 Stefan__: oh.. then you are half way there Mar 27 11:03:42 apart from CROSS_COMPILE, what else is there? Mar 27 11:03:50 Stefan__: now you need to cross compile a user space prog Mar 27 11:03:59 well the driver I believe is a kernel space app Mar 27 11:04:04 Stefan__: how you do that, depends on the system you used for building your userspace Mar 27 11:04:07 as you need to insmod it to use it Mar 27 11:04:34 Stefan__: wait for mdp to wake up Mar 27 11:04:39 http://blog.boxysean.com/2012/08/12/first-steps-with-the-beaglebone-pru/ Mar 27 11:04:39 he is master of all things PRU Mar 27 11:04:50 I would like to follow the above link but by cross-compiling instead Mar 27 11:04:57 Stefan__: ah.. Mar 27 11:04:59 Should I offer some tribute to wake him up :) Mar 27 11:05:05 Stefan__: isn't "KERNEL_DIR=blah" enough? Mar 27 11:05:23 Stefan__: make some strong coffee and leave the window open Mar 27 11:05:25 well that's what I don't know is it KERNEL_DIR or KSRC or what :) Mar 27 11:05:28 Stefan__: iirc the module make files take everything they need from there Mar 27 11:05:33 but then, he's from the US, make any coffee Mar 27 11:05:53 Stefan__: you need the kernel source code somewhere Mar 27 11:05:54 Stefan__: it's KERNEL_DIR Mar 27 11:06:06 KotH: Good, let me try that :) Mar 27 11:06:20 Stefan__: iirc (i've never done any PRUSS stuff) Mar 27 11:06:32 Stefan__: the trolls here refuse to give me a bb :) Mar 27 11:06:49 KotH: and the swiss franc is worthless? Mar 27 11:07:00 KotH: Did you try bying one :P Mar 27 11:07:09 av500: i'm a poor student, i dont have any money Mar 27 11:07:37 av500: you know, humanities doesnt pay well ;) Mar 27 11:07:56 cough Mar 27 11:10:20 Hmm... compiled with a bunch of warnings Mar 27 11:10:25 just eat less chocolate. will save the environment too Mar 27 11:10:39 but it only produced a .o file, shold it not be a .ko? Mar 27 11:10:46 as in kernel? Mar 27 11:10:47 should Mar 27 11:11:24 or does it place the ko somewhere else? Mar 27 11:11:36 let me double check but I think not] Mar 27 11:12:19 Nope, it does not Mar 27 11:12:28 Would you like to see the output of the compile? Mar 27 11:12:37 It's about 5 lines Mar 27 11:13:07 Stefan__: did you do a "make modules" ? Mar 27 11:13:21 facepalm! good point Mar 27 11:14:48 double facepalm -- no such target Mar 27 11:15:32 try make foo.ko Mar 27 11:15:52 if it doesn't work, just read the Makefile Mar 27 11:16:00 and maybe the README Mar 27 11:16:15 or whatever documentation is provided Mar 27 11:16:49 I read the Makefile, it only confused me Mar 27 11:16:56 there are several targets and I've made all of them Mar 27 11:18:15 .o0(Genesis: Land of Confusion) Mar 27 11:21:51 Apparently there's an SDK... however, what worries me is my kernel is homebrewed (sort of), as it includes Xenomai support Mar 27 11:22:50 av500: I don't think I will look further into ti-rtos. I've just read the getting started guide and it only looks like a path into vendor locked foo Mar 27 11:23:19 bu Mar 27 11:23:20 but Mar 27 11:23:22 but Mar 27 11:23:23 its Mar 27 11:23:25 free Mar 27 11:24:58 i prefer some rtos which consist of a few files/drivers and some makefiles, not a highly complicated framework ;) Mar 27 11:25:38 aholler: here: https://chromium.googlesource.com/chromiumos/platform/ec/ Mar 27 11:25:45 it even runs on LM4 Mar 27 11:25:48 well my experience with Xenomai (once the kernel was working) has been positive, it's relatively easy to write code for it, and it works quite well Mar 27 11:26:46 av500: there are several free rtos around. Mar 27 11:26:53 i know Mar 27 11:26:57 like freertos Mar 27 11:28:17 So any suggestions on my PRU problem :( ? Mar 27 11:29:21 aholler: this TI rtos is using sysbios and thus brings all the xdc package legacy Mar 27 11:29:50 yes, that's what I've just seen. Mar 27 11:31:01 it's just something renamed to rtos to catch more people Mar 27 11:55:25 might be reasonable for the clicking windows crowd, but not for me Mar 27 11:58:45 * KotH double clicks on aholler Mar 27 11:59:59 btw: is there any rtos out there that has a good power handling system? ie one that allows to run the system in the lowest possible power stage when nothing is going on? Mar 27 12:00:26 * KotH has so far not found anyone that does more than the basic power handling Mar 27 12:04:51 KotH: that would be nice and complex ... Mar 27 12:05:53 yeah.. unfortunately Mar 27 12:06:05 KotH: free? Mar 27 12:06:34 for a modest fee Mar 27 12:07:17 alan_o: $company wouldnt mind to pay something, for my private: has to be free as in beer Mar 27 12:07:30 s/my/me/ Mar 27 12:07:37 but you don't drink Mar 27 12:07:39 try linux Mar 27 12:08:11 he said rtos... Mar 27 12:08:15 mru: ok, free as in lunch Mar 27 12:08:18 with pm Mar 27 12:08:26 there are the rt-patches Mar 27 12:08:27 linux doesn't really do either of those Mar 27 12:08:30 hmm.. well the power management thing seems like it would suffer from multi-arch OS's, so I wonder what Q-Kernel's handling of power is, since it's PIC24/32 only Mar 27 12:09:53 for the stuff someone uses normally a rtos, just disabling the clocks and modifying the idle-call might be easy. Mar 27 12:09:59 KotH: Q-Kernel uses a lot of pic-specific stuff to do what it does on the RT side, so I assume they do the same with power. Mar 27 12:10:25 alan_o: ok.. i'll have a look then Mar 27 12:10:28 alan_o: thanks Mar 27 12:10:33 alan_o: and good morning :) Mar 27 12:10:53 np, and good afternoon :) Mar 27 12:10:57 Hi. yesterday I got some advice to use the ubuntu 12-10 from elinux.org. Mar 27 12:11:03 Hey guys, so it seems that I figured out the way the PRU code works, however, since I am using my own kernel, which does not have PRU support, I can't quite use it, does anybody know if it's possible to load PRU programs otherwise? Mar 27 12:11:04 aholler: just disabling clocks and modifying idle calls doesnt cut it Mar 27 12:11:48 builda general solution will just end up in a very complicated way to do things Mar 27 12:11:52 aholler: depending on what subsystems you run, you want to change clock frequency and switch between various clock sources (crystal, pll, rc-osci...) Mar 27 12:12:13 KotH: thats what I said, you know what you are running Mar 27 12:12:29 therfor you know what you can disable. Mar 27 12:12:32 aholler: subsystems as in uC subsystems Mar 27 12:12:40 Stefan__: so add pru to your kernel Mar 27 12:12:42 *sigh* Mar 27 12:12:54 aholler: modern uC systems are a tad bit more complex than you think Mar 27 12:13:02 you know what you use from your µC Mar 27 12:13:26 KotH: you don't know what I think Mar 27 12:14:03 its getting philosophical Mar 27 12:14:37 aholler: you probably think i'm an asshole... most people do that :) Mar 27 12:14:48 av500: yeah, he should avoid his stupid assumptions about other people Mar 27 12:14:58 I know what you all think..it's in the borislog Mar 27 12:15:03 KotH: exactly Mar 27 12:16:52 av500: Easier said than done -- I just had a look at what menuconfig is giving me as options on what I've got as kernel source -- the PRU driver is not one of them Mar 27 12:17:05 av500: I have no clue on how to add just that to my sources Mar 27 12:17:12 av500: Suggestions? Mar 27 12:17:13 define your sources Mar 27 12:17:17 not angstrom? Mar 27 12:17:20 if not, why not Mar 27 12:17:23 definitely not Mar 27 12:17:31 utterly not? Mar 27 12:17:40 Stefan__: it's uio_pruss Mar 27 12:17:50 I started with Arago, then merged to mainstream 3.21 I believe, then merged the Xenomai stuff Mar 27 12:18:07 there was no other way that I could see to get the Xenomai to run at the time Mar 27 12:18:54 mdp: I know it's uio_pruss, but how do I get it merged to my kernel sources? Mar 27 12:19:02 git merge :) Mar 27 12:19:13 adding the directory and modifying a Makefile Mar 27 12:19:37 av500: Lol -- you overestimate my git magic :) Mar 27 12:19:48 aholler: you think it would work? Mar 27 12:20:01 why not add xenomai to angstrom kernel? Mar 27 12:20:27 CONFIG_UIO_PRUSS=y Mar 27 12:20:48 mdp: I did a menuconfig, and it was not listed Mar 27 12:21:14 you're using an unsupported kernel Mar 27 12:21:19 yes Mar 27 12:21:19 I am :) Mar 27 12:21:28 [13:17:50] I started with Arago, then merged to mainstream 3.21 I believe, then merged the Xenomai stuff Mar 27 12:21:31 I predict misfortune and woe to come your way Mar 27 12:21:39 an evil hobbyist kernel Mar 27 12:21:52 mdp: lol Mar 27 12:21:54 so, which kernel would you guys suggest I start with? Mar 27 12:21:58 mdp: good morning btw Mar 27 12:22:04 Stefan__: angstrom Mar 27 12:22:05 mdp: Stefan__ has a coffee ready for you Mar 27 12:22:06 mdp: Misfortune and woe are here already :) Mar 27 12:22:12 KotH: good afternoon ;) Mar 27 12:22:32 Stefan__: and apparently pain theory does not work on you...ok then. let's proceed Mar 27 12:22:38 KotH: I've got coffee for everyone who's willing to help, cookies too Mar 27 12:23:02 Stefan__: sorry, i'm picky about my coffee... turkish heritage and all ;) Mar 27 12:23:09 mdp: So you recommend I get a hold of the Angstrom kernel? Mar 27 12:23:14 KotH: In a djezve? Mar 27 12:23:25 +10 Mar 27 12:23:28 Stefan__: cezve :) Mar 27 12:24:06 KotH: Excuse me, my Turkish is rusty, you guys left @140 yrs ago, so not much practice since ;) Mar 27 12:24:39 Џезва Mar 27 12:24:42 av500: I don't want to hear about this Angstrom kernel -- it means I have to merge Xenomai in yet again Mar 27 12:24:50 so? Mar 27 12:24:58 av500: I am kidding :) Mar 27 12:24:59 Stefan__: I would recommend using the beagleboard/kernel 3.8 Mar 27 12:25:12 mdp: Do you have a link? Mar 27 12:25:16 mdp: I heard 3.9 will be all the rage Mar 27 12:25:22 however, you seem to have a constraint with wanting to patch xenomai Mar 27 12:25:39 av500, rage on Mar 27 12:25:50 mdp: In an ideal world, I'd like to have Xenomai support AND PRU support, but I suppose I could live with either/or Mar 27 12:26:06 you could maybe time-share Mar 27 12:26:06 Stefan__: so you're not from edinburgh? Mar 27 12:26:09 https://github.com/beagleboard/kernel/tree/3.8 Mar 27 12:26:22 KotH: I am not from, but in; I'm Bulgarian Mar 27 12:26:37 KotH: Also, way to stalk me :D Mar 27 12:26:39 * KotH gets out the aussault rifle Mar 27 12:26:53 stalk? Mar 27 12:26:54 stalk? you're new to the intarwebz! Mar 27 12:27:24 I guess :) Mar 27 12:27:45 an evil student kernel then Mar 27 12:28:19 av500: Yes with some help from nice peoplez from the interwebz Mar 27 12:28:27 Stefan__: and just for the record, the turkish are still in bulgaria... even though 90% left in '89 Mar 27 12:28:44 KotH: Yes I know, but officially I mean ... and all that jazz Mar 27 12:29:03 Stefan__: yeah..yeah.. Mar 27 12:29:04 KotH: My grandma's village has some very nice turkish people -- friendly and helpful Mar 27 12:29:30 * KotH cannot say the same about his grandmother and the bulgarians there.... Mar 27 12:29:32 but i disgress Mar 27 12:29:33 mdp: Would I need a special rootfs? Mar 27 12:29:45 mdp: I am currently using a barebones one from the Yocto project Mar 27 12:30:20 KotH: Sorry, unfortunately not all people are good :( Mar 27 12:30:36 doesn't matter...armv7 is what's special Mar 27 12:30:55 Stefan__: yeah.. and lots of bad blood have been running between bulgarians and turkish Mar 27 12:31:31 so just leave your national feelings out of this channel Mar 27 12:31:38 they are just annoying Mar 27 12:31:53 aholler: I don't see anyone fighting Mar 27 12:32:01 so what's your problem? Mar 27 12:32:20 Stefan__: the other thing you can do if you want to be on 3.2 is go grab https://github.com/koenkooi/linux and the most recent 3.2.24 branch at the top of the branch list Mar 27 12:32:33 mru: calls for jihad, war to infidels, ... Mar 27 12:32:36 extract out the uio_pruss am33xx enablement patches if you like Mar 27 12:32:57 aholler: for fuck's sake, grow a sense of humour Mar 27 12:33:06 impossible for some things Mar 27 12:33:13 then get the fuck of the internet Mar 27 12:33:14 Stefan__: given that everybody here pretty much is synced with the latest community driven kernel (3.8), I doubt you'll find much assistance though Mar 27 12:33:24 mru: good suggestion Mar 27 12:33:32 jeez Mar 27 12:33:53 certainly not friday Mar 27 12:34:10 ^^' Mar 27 12:34:12 sorry boys Mar 27 12:34:18 mdp: It's wednesday today :D Mar 27 12:34:19 KotH: you did nothing wrong Mar 27 12:36:52 mdp: I don't necessarily want to stick with 3.2.24 or 21, if I however want Xenomai -- I think they're up to 3.4 or 3.5 right now Mar 27 12:36:52 Stefan__: "friday" has a special meaning in this channel Mar 27 12:37:05 KotH: Care to enlighten me :) Mar 27 12:37:12 Stefan__: there's no 3.4/3.5 kernel for am335x/bone Mar 27 12:37:21 3.2 then jump to 3.6/7/8 Mar 27 12:37:24 your choice Mar 27 12:37:51 mdp: I feel lost :) Mar 27 12:38:00 Okay a general question then Mar 27 12:38:12 Stefan__: the day when the channel is particullarly crazy Mar 27 12:38:52 I am developing a communications prototype and would ideally like to drive some pins really fast with as little jitter as possible Mar 27 12:38:53 Stefan__: you've to experience it yourself.. there are no words to describe it Mar 27 12:38:57 n7segment: heyo! Mar 27 12:38:59 what would you recommend Xenomai or PRU? Mar 27 12:39:32 do you mean you're bitbanging some protocol? Mar 27 12:39:44 Erm -- depends :D Mar 27 12:39:48 Kind of I suppose Mar 27 12:39:51 out of tree software solution or a vendor lock-in real time core found on some TI parts Mar 27 12:39:56 hrm Mar 27 12:40:16 depends on my use case Mar 27 12:40:30 It's not a protocol -- it really is a prototype -- it's on a very low level Mar 27 12:40:42 I am using the BB to simulate purpose build hardware I suppose Mar 27 12:41:01 in that case the ti lock-in is of no concern Mar 27 12:41:08 If this were to be actually manufactured it would probably all go into a chip, and be done by hardware rather than software Mar 27 12:41:20 It's all just proof of concept Mar 27 12:41:29 the pru certainly gives the most accurate timing Mar 27 12:41:45 I am aware -- but would it be able to handle large data transfers Mar 27 12:41:51 i.e. 100 Mbps? Mar 27 12:42:02 What worries me is the transfer between DDR and the PRU memory Mar 27 12:42:36 The pipe width is not a problem i.e. I could use 16 pins for example Mar 27 12:43:00 pru has limited i/o pins Mar 27 12:43:18 also loses determinism going out on the crossbar (e.g. to DDR) Mar 27 12:43:31 also, where do you get 100mbps from? Mar 27 12:43:43 av500: Just an example Mar 27 12:44:01 dma data from ddr to ping-pong buffers in pru-internal memory Mar 27 12:44:04 well, you need to know your constraints first Mar 27 12:44:06 mdp: Yeah I get the determinism loss Mar 27 12:44:30 brb Mar 27 12:44:35 Stefan__: tell us more about your application Mar 27 12:44:47 Stefan__: what i've seen sofar sounds interesting :) Mar 27 12:45:57 Hi. Yesterday I got some advice to use the 12.10 Ubuntu Version from eLinux.org. Unfortunately, I have problems when installing. Is there a .img File that I can use? At the moment I use preinstalled OMAP3 Precise (12.04) but I have some problems with the I2C. I cant see my devices. Mar 27 12:46:33 Bernd_: _what_ is your problem? Mar 27 12:46:51 mdp: Okay I'm back Mar 27 12:46:58 KotH: I am about to :) Mar 27 12:48:25 mdp, KotH: So let's say I want to do the following with a BB and an FPGA --> on the transmitter side I'd like to be able to move data from DDR to the PRU memory at speeds of at least 10 Mbps and then be able to toggle some pins via the PRU at something like 4-5MHz Mar 27 12:48:47 M bits or bytes Mar 27 12:48:55 bits Mar 27 12:49:08 :) Mar 27 12:49:36 and on the receiver side I'd like to capture the signal via FPGA, process it and then do the reverse -- feed the decoded data to the BB Mar 27 12:50:02 I don't think I'll get away without FPGA as there's a bit of processing to be done and it has to be real time Mar 27 12:50:46 the former can be done on one PRU with some reserved edma channels...then shoved shared quickly with the other PRU via the scratchpad registers. the latter would be done on the other PRU...bitbanging a 4-5MHz isn't really an issue as the core runs 200MHz and can toggle PRU GPOs on each clock. Mar 27 12:52:07 mdp: Can you point me in the direction of some material on edma reserved channels? Mar 27 12:52:18 and also on the memory sharing between PRUs Mar 27 12:52:31 KotH: I try to get some data form a I2C device connected to my BB but I dont get anything. I tried these tutorial http://dominion.thruhere.net/koen/cms/using-the-beaglebone-as-a-weatherstation. I'm not sure I there is something "wrong" at my ubuntu. I also have to enable my I2C with modprobe i2c-dev to see my i2c bus. The theme is new for me so I'm not sure where could the failure be. Mar 27 12:52:32 er Mar 27 12:52:35 spruh73 Mar 27 12:52:46 I've got that one :) Mar 27 12:52:50 Stefan, spruh73c.pdf has all the programming information Mar 27 12:52:55 you must have Rev C Mar 27 12:52:59 google has it cached Mar 27 12:53:09 ok, I'll have a look, I've got a local copy ;) Mar 27 12:53:09 the golden rev Mar 27 12:53:20 mru: golden rev? Mar 27 12:53:28 rev c Mar 27 12:53:34 kk Mar 27 12:53:46 the later ones have a lot of pru stuff removed Mar 27 12:53:48 stefan: did I see you are doing PRU work? Mar 27 12:53:56 cmicali: About to :) Mar 27 12:54:04 stefan: I've been doing a lot of PRU stuff for a data aq project, let me know if you have any specific questions Mar 27 12:54:05 mdp: did they remove said information in later revisions? Mar 27 12:54:21 cmicali: I DO :) mdp has been very helpful so far though :) Mar 27 12:54:35 KotH: all of it. that's why I suggest that revision to anybody trying to do real work on the PRU Mar 27 12:54:44 stefan: i'm around, happy to help. Mar 27 12:55:02 the documents in the am335x pru package are missing key info too Mar 27 12:55:16 mdp: Oh, I see, I mgiht have 76 Mar 27 12:55:39 SPRUH73C–October 2011–Revised December 2011 Mar 27 12:56:09 caxapa.ru/thumbs/302179/spruh73c_.pdf Mar 27 12:56:12 there Mar 27 12:56:16 its in the channel logs now Mar 27 12:56:19 forever Mar 27 12:56:34 av500, some of us can't do that ;) Mar 27 12:56:57 what, type a random url and hope it links to something? Mar 27 12:57:02 that's all I did Mar 27 12:57:09 heh Mar 27 12:57:19 mdp: hmm? do you know the reason why they removed it? Mar 27 12:57:26 mdp: it somehow doesnt make sense to me Mar 27 12:57:27 yes Mar 27 12:57:30 TI does not need reasons Mar 27 12:57:39 KotH: one word: effort Mar 27 12:58:00 av500: ti is not elenril Mar 27 12:58:06 it is Mar 27 12:58:15 ^^; Mar 27 12:58:32 Hello team! Mar 27 12:58:41 grezi single player Mar 27 13:00:03 btw, did we reach consensus on where to source stackable headers for the bone capes? Mar 27 13:00:35 mrpackethead wanted to buy a 1000 or so from a chinese vendor Mar 27 13:00:50 er.. taiwanese Mar 27 13:00:53 mdp: Thanks for the info -- I think I'll try with the PRU approach Mar 27 13:01:20 Can anyone give me a hint or a hand over a Debian installation if it is not to much of an issue? Mar 27 13:01:22 np Mar 27 13:01:24 mdp: Anyway, I have a somewhat working solution with Xenomai which I can just put on ice, and see if the PRU can perform better Mar 27 13:01:53 mdp: You're here often? So I know where to come back to when the questions start raining :) Mar 27 13:02:31 sure...mranostay takes the late shift for PRU stuff ;) Mar 27 13:02:59 mdp: GREAT :) I guess I'd better get on getting that kernel you suggested going Mar 27 13:03:07 with PRU support of course :) Mar 27 13:03:28 Fico: http://www.debian.org/CD/live/ Mar 27 13:03:30 Stefan__: pls dont forgot to fill out the survey Mar 27 13:03:34 and mention mdp Mar 27 13:03:38 forget* Mar 27 13:03:51 alternatively, poke him on FB Mar 27 13:03:57 lol :D Mar 27 13:03:59 you just need to set up your pin muxes and PRU is usable. Mar 27 13:04:03 just give me a link :) Mar 27 13:04:12 av500, good luck finding me on FB :P Mar 27 13:04:15 KotH: I did follow it line by line! but got stuck in "starting up partinioner" during the instalation :S Mar 27 13:04:25 did you read between the lines? Mar 27 13:04:29 Fico: uh... you might want to try #debian Mar 27 13:04:32 mdp: you mean so that the PRU accessible pins are muxed to the the connectors? Mar 27 13:04:37 Fico: this is not a general linux help channel Mar 27 13:04:37 yes Mar 27 13:04:45 Stefan__: everything is muxed Mar 27 13:05:03 KotH: I will TY! Mar 27 13:05:05 stefan: i have a device tree change that mixes all of the PRU pins you need for the 3.8 kernel Mar 27 13:05:17 there you go Mar 27 13:05:23 its a PRU love fest here today Mar 27 13:05:27 stefan: so once you get the 3.8 kernel building, you can use that to get all the pins Mar 27 13:05:33 cmicali: as long as he doesn't need a different set of pins for his project ;) Mar 27 13:05:43 ... all of the PRU ... Mar 27 13:05:56 #pruallthethings Mar 27 13:06:07 mdp: eh :D everyone should only need 10 outputs and 2 inputs Mar 27 13:06:09 I LOVE that hashtag Mar 27 13:06:33 cmicali: right! except for my 6502 cape Mar 27 13:06:34 mdp: I need like 4-8 outputs and 16 inputs :) Mar 27 13:06:36 and .... Mar 27 13:06:45 and everything other than my specific project Mar 27 13:06:56 stefan: you probably can't get that many inputs :( Mar 27 13:07:28 stefan: you can use the ti pinmux tool to figure out what you can bring out, the bone is limited though as not all of the am335x pins are exposed in the 2 headers Mar 27 13:07:33 cmicali: It's OK, I probably can get away with moving the data through general GPIO Mar 27 13:07:57 memory is hazy, but i think you can get a total of about 8-10 PRU pins out, almost any mix of in/out Mar 27 13:08:15 stefan: http://www.ti.com/tool/pinmuxtool this thing ended up being pretty helpful Mar 27 13:08:22 If that goes at over 5MSps then it's all good :) Mar 27 13:09:27 https://docs.google.com/spreadsheet/ccc?key=0As0aJokrBccAdGkxeHkyYW1qRHNQdm5yZDhPQlRNR2c#gid=0 Mar 27 13:10:38 you can get your 16 inputs but need to drive them from both prus via the DEBUG register Mar 27 13:12:09 you'd use PRU1 GPI 0..13 then PRU0 14..15 to be able to use r30/r31 naturally with the lower halfword Mar 27 13:13:18 that then leaves 7 output pins left on PRU0 Mar 27 13:13:55 prufect Mar 27 13:15:19 Stefan__: I did the debug register scheme in https://github.com/ohporter/b6502/blob/master/pruss/b6502_pruss.p that you'll need to do copy something similar Mar 27 13:15:49 in order to get an aligned 16-bit read Mar 27 13:16:07 ahh, the i2c numbering patch again Mar 27 13:16:13 in my case it's was handling 8-bits in and out plus control stuff Mar 27 13:16:33 Jacmet: yeah, when benoit went MIA after OMAPageddon I just gave up for a while Mar 27 13:17:06 Jacmet: the fortran numbering one for dts? Mar 27 13:17:14 mdp: yes Mar 27 13:17:31 yeah, doublechecking to give an ack Mar 27 13:17:44 mdp: I agree it should be the same as the trm, but by now there's probably quite some dts written with the old numbers :/ Mar 27 13:18:03 back when I did it...it was a smaller patch...since it got dropped things have snowballed Mar 27 13:18:27 Jacmet: there's never a bad time to do things right Mar 27 13:18:46 plus...all the out-of-tree stuff is irrelevant to upstream Mar 27 13:18:51 mdp: true, but the cost of changes increase of time Mar 27 13:19:06 mdp: Thanks -- I don't need to do duplex at this point -- just simplex Mar 27 13:19:14 that's the vendor kernel maintainer problem ;) Mar 27 13:19:21 mdp: So that probably means I am safe in terms of number of pins Mar 27 13:19:40 speaking of omapagedon... what would be a good replacement part for the 3530? will dm3730 be shut down too... is am335x a better solution Mar 27 13:20:41 thurbad, it depends if it has what you need...they all have different features Mar 27 13:21:00 3730 is the obvious immediate replacement for 3530 Mar 27 13:21:29 similar feature set to 3530, similar pinout, long term availability Mar 27 13:21:51 3730 is mostly just newer versions of all the blocks Mar 27 13:21:53 mdp: did you use DMA on your b6502 project? Mar 27 13:21:55 and higher clock rates Mar 27 13:22:24 yeah, we've already got a board designed with dm3730, my only worry is that it may suffer the same fate as the omap Mar 27 13:22:25 dunno anything about _DM_ but from an AM37xx POV, it's the highest performing, lowest power part that the Sitara folks have Mar 27 13:22:37 so it would be difficult to go away Mar 27 13:22:50 am335x can't replace it in many applications Mar 27 13:23:39 there's no reason to stop making a perfectly good chip as long as there are buyers Mar 27 13:24:13 might be discouraged for new designs of course Mar 27 13:24:26 'k I was just wondering if I was being worried for no reason... sounds like I probably am Mar 27 13:25:07 I might have seen a mention of long term support for dm3730 Mar 27 13:25:14 for every chip, they should be able to tell you how long they're committed to making it Mar 27 13:25:21 ask your ti rep Mar 27 13:25:37 thurbad: you should be able to get a written commitment from ti on dm3730 Mar 27 13:26:01 heh, assuming we can get a response...they haven't been really helpful in the past Mar 27 13:26:20 thurbad: it also took some months here Mar 27 13:26:59 they've been to our offices once and sold my boss at the time on the omap4... :( Mar 27 13:27:47 the omap4 looks very nice on paper Mar 27 13:28:31 as a paperweight Mar 27 13:28:46 thurbad: there should be a line of happy sales reps from distributors who want to sell you catalog parts like dm3730. try designing something with omap4460 instead... :) Mar 27 13:28:49 this is the same boss that decided to go with the 3530 over the 3730 board we had in the works :P Mar 27 13:29:16 that sounds stupid Mar 27 13:29:27 3730 is cheaper than 3530 Mar 27 13:29:31 and better Mar 27 13:29:34 and 3530 is completely unsupported Mar 27 13:29:38 he didn't think it was mature enough at the time Mar 27 13:29:52 it's the matured version of 3530 Mar 27 13:30:03 I know Mar 27 13:30:29 but it was a new chip as far as he was concerned Mar 27 13:30:55 sounds like wrong person making decisions Mar 27 13:31:17 this was about 2 and a half years ago Mar 27 13:31:57 the world was fucked up beyond repair much earlier than that Mar 27 13:32:08 lol Mar 27 13:32:37 the xM had only recently been released Mar 27 13:34:10 and I wouldn't say he was the wrong person to be making decisions.. just that that one was a bad one Mar 27 13:34:45 he made the decision based on insufficient information then Mar 27 13:34:57 that's probably fair Mar 27 13:35:50 hi Mar 27 13:36:01 lo Mar 27 13:36:05 Z Mar 27 13:37:35 Jacmet: I'm still missing 3/5 that you replied to ;) odd..got your reply. Mar 27 13:37:49 mdp: strange Mar 27 13:38:00 mdp: spam filter? Mar 27 13:38:40 Jacmet: btw, I do feel bad that my original patch was not considered important then since now we have more churn...but tough luck I guess Mar 27 13:39:01 mru: the 3530 is a manured chip by now Mar 27 13:39:02 I even mentioned that it needed to be fixed before more garbage went in with the wrong labels Mar 27 13:39:18 mdp: yes, I remember Mar 27 13:39:25 mdp: I even think I ack'ed it Mar 27 13:40:22 so who's going to send the patch to make it stop doing sequential numbering? Mar 27 13:40:25 not even in spam..very odd Mar 27 13:40:26 (not that it matters much) Mar 27 13:40:31 koen, different issue Mar 27 13:40:32 I want /dev/i2c0,/dev/i2c2 Mar 27 13:41:00 koen, technically, my sub-team colleague and I own that driver now. :) Mar 27 13:41:08 it sure isn't going to be me jumping on that Mar 27 13:41:21 then your sub colleague Mar 27 13:41:41 he's in .in so I don't have contact Mar 27 13:41:48 koen: probably want you really want is /dev/i2c-foo and /dev/i2c-bar instead if you really wanted consistent naming Mar 27 13:42:10 * mdp purchases koen a udev rule Mar 27 13:42:16 can a UUID help here? Mar 27 13:45:50 koen: just gave your recent image a try and still see the xmodem CCCCC Mar 27 13:46:48 hmmm Mar 27 13:47:01 Jacmet: can you login and do: 'systemctl status emmc' ? Mar 27 13:47:15 (and pastebin the output) Mar 27 13:47:29 koen: you mean boot from the sd and run systemctl while the emmc script runs? Mar 27 13:48:53 koen: http://pastebin.com/1SZw3ubx Mar 27 13:50:36 ok, that seems to be doing a lot better than the previous incarnation Mar 27 13:50:42 the kpartx thing looks odd Mar 27 13:51:03 I reused the omap3 script Mar 27 13:51:09 that has support for loop files Mar 27 13:52:25 Does anyone have any advice for me, what it could be that I do not see any I2C devices? I plugged my BMP085 from Sparkfunk directly to the BB. Mar 27 13:52:53 Bernd_, i2c busses do not probe for devices Mar 27 13:53:23 you have to explicitly address the device, either from user space or from device tree Mar 27 13:53:46 panto, Jacmet: can you check the contents of /dev/mmcblk1p1 and see if MLO is present? Mar 27 13:54:14 koen: root@beaglebone:~# ls /media/1 Mar 27 13:54:18 MLO u-boot.img uEnv.txt Mar 27 13:54:20 but I should see something with i2cdetect -r -y 3 ?? Mar 27 13:54:32 Bernd_, probably Mar 27 13:54:45 are you sure about the bus numbering though? Mar 27 13:54:48 Jacmet: OK, so that worked, now I need to figure out why the ROM doesn't like it Mar 27 13:54:54 and the pinmuxing Mar 27 13:55:04 pin muxing ftw Mar 27 13:55:11 Jacmet: and did you powercycle? warm reboot won't work, another ROM issue Mar 27 13:55:38 koen: ahh, I'm fairly certain I only pressed the reset button Mar 27 13:55:48 koen: I'll let if finish once more and pull the power Mar 27 13:56:07 mdp: you mentioned reserved channels on the EDMA controller - would you mind explaining that? PRU could use those instead of having to have the edma linux kernel driver allocate channels? Mar 27 13:56:21 koen: is that rom issue documented anywhere? I don't remember anything like that in the errata Mar 27 13:56:34 cmicali, the linux edma driver only allocates channels for linux use Mar 27 13:57:09 mdp: yeah, although one of the PRU examples exposes a user-land driver to allocate/deallocate channels that their PRU example code then uses Mar 27 13:57:19 mdp: so I wasn't sure if that is the correct approach, or if there is a way to bypass having to do that Mar 27 13:58:01 hey emeb_mac Mar 27 13:58:04 cmicali: in the 3.8 vendor kernel there's a DT property to reserve channels/slots Mar 27 13:58:14 hi panto Mar 27 13:58:31 cmicali: ti,edma-reserved-channels and ti,edma-reserved-slots which you can set a list of channels that the linux driver will ignore. Mar 27 13:58:34 mdp: I saw that, https://lkml.org/lkml/2013/1/11/23 right? Mar 27 13:58:56 mdp: ok, so you pick a couple channels you want to use, add them there, linux ignores them, and then they are available for use by PRU directly Mar 27 13:59:02 then you can use those channels on the PRU as they are reserved for "other processor" use Mar 27 13:59:06 right Mar 27 13:59:11 choose wisely, though Mar 27 13:59:16 koen, emmc boot seems to work Mar 27 13:59:17 mdp: that was the next question Mar 27 13:59:27 mdp: what are the reserved ones already in the DT being used for :) Mar 27 13:59:28 now stuck at: systemd-fsck[82]: Angstrom: clean, 68935/112672 files, 330160/449820 blocks Mar 27 13:59:34 fyi, those properties are going away upstream...will probably be replaced by configfs options Mar 27 13:59:38 how much time does it take? Mar 27 13:59:41 panto: Yes Im sure about the bus number. There is a tutorial at Koen's Blog that I tried. But there is no new address neither on 1 /2 or 3 Mar 27 13:59:57 cmicali: you could just use those...they are just useless ones I shouldn't have had in there Mar 27 14:00:00 mdp: thanks Mar 27 14:00:01 pinmuxing check time Mar 27 14:00:10 it's an outdated patchset in the 3.8 beagle kernel Mar 27 14:00:33 mdp: ok i will start trying the ones already there Mar 27 14:00:46 mdp thanks Mar 27 14:00:58 koen: with a power cycle it boots Mar 27 14:01:19 hey _chase_1 Mar 27 14:01:57 koen: but hanging at fsck like panto said Mar 27 14:10:45 Jacmet: dunno what happened..had to recover 3/5 from gmane. :( Mar 27 14:11:04 mdp: funky. You were in To: even Mar 27 14:12:26 the internets hate me today Mar 27 14:14:56 did you do something to clog up the tubes? Mar 27 14:15:49 Do I have to do something before, that I can see the the address of the connected device with the i2cdetect command? Mar 27 14:19:03 check pinmux Mar 27 14:19:09 do you see the I2C pins change? Mar 27 14:21:58 I dont know. Have to have a look how this works. Mar 27 14:25:45 both are in Mode 0 (i2c2_scl/sda) Mar 27 14:26:49 Bernd_: what av500 meant is, that you should connect an osciloscope or a logic analyzer to the pins and see whether something happens Mar 27 14:27:19 ah ok. Mar 27 14:27:25 Bernd_: prefered way of doing is first connecting an oscilloscope to see pins toggeling, then a logic analyzer to see whether the data is the one you expect Mar 27 14:27:45 k Mar 27 14:27:58 if your sequence is short enough, the osciloscope will do for the second step as well Mar 27 14:28:16 i will connect it to the oszi Mar 27 14:32:49 koen, why is there a i2c1 pinmux patch in audio? Mar 27 14:33:18 1.13.1&0a;BME R:80 AFE <0:e1>&0a;IFE Mar 27 14:33:27 worst version string EVER Mar 27 14:33:46 panto: apparently fsck takes 30+ mins Mar 27 14:34:21 * panto facepalms Mar 27 14:35:14 wtf Mar 27 14:35:44 hehe Mar 27 14:36:07 Jacmet: fsck'ing a 4TB ext2 formated disk? :) Mar 27 14:37:10 ka6sox-away, there's a codec on the i2c bus Mar 27 14:37:12 biot: ping Mar 27 14:37:20 KotH: no, 2GB of slow emmc ;) Mar 27 14:37:53 panto, which codec (so I can annotated it as to what requires it) Mar 27 14:38:04 jackmitchell: pong Mar 27 14:38:14 tlv320aic3x on the dvi cape Mar 27 14:38:25 it might not be at the hdmi case Mar 27 14:38:38 biot: I'm getting an error building sigrok-gtk-git on ArchLinux 64bit Mar 27 14:38:58 I only have the 1,8V but no signal. Perhaps the 1,8V are not enught for the BMP085. If I power up the bmp085 with 3,3V (without conection to the board) will I see some signals on the sda line? Mar 27 14:39:05 panto, thanks, I'll rename it so it makes sense. Mar 27 14:39:09 biot: http://pastebin.com/YCnucFYF Mar 27 14:39:32 jackmitchell: yeah, don't worry about sigrok-gtk and sigrok-qt, they're unmaintained/obsolete Mar 27 14:39:38 ah, ok Mar 27 14:39:48 should I be looking at pulseview then? Mar 27 14:39:52 yup Mar 27 14:40:03 biot: gotcha, thanks :) Mar 27 14:42:22 Jacmet: ah.. the embedded equivalent of an 2TB disk :) Mar 27 14:49:31 biot: haven't tried recently but is pulseview getting decoder support? Mar 27 14:53:53 morning Mar 27 14:53:59 mourning even Mar 27 14:54:13 mranostay: free coffee at Stefan__'s Mar 27 14:55:12 mranostay: looks almost good now Mar 27 14:55:25 I would clean up that scattered parts in top right Mar 27 14:55:31 * mdp sees mranostay and punches out. Mar 27 14:55:39 transistors and resistors in all orientations Mar 27 14:55:53 electrons will be confused Mar 27 14:57:47 mdp: eventually, but it'll take a while yet Mar 27 14:58:06 only one guy is working on pulseview, and he doesn't have all that much time Mar 27 14:58:16 ok Mar 27 14:58:24 mdp: patches welcome Mar 27 14:58:28 yeah yeah Mar 27 14:58:33 as in... "ok I'll do it?" Mar 27 14:58:42 not really an area I can help Mar 27 15:01:50 uhm... Mar 27 15:02:00 pulseview does not find my ols here Mar 27 15:02:09 my OLS arrived today Mar 27 15:02:23 no errors.. the ols-client does work.. any ideas what i should look for? Mar 27 15:02:33 12 days from payment to recieve in the UK which I thought was acceptable Mar 27 15:03:07 it's useful, I would have loved it to be able to do continuous sampling but that's a bit much to ask from a $50 device :D Mar 27 15:04:17 KotH: it's on a serial port, typically /dev/ttyACM0 Mar 27 15:04:17 jackmitchell: the range of use cases you can handle is quite large for a $50 board, yes. Mar 27 15:04:29 we don't probe serial ports (anymore) Mar 27 15:04:30 KotH: i'm trying to use pulseview here, I can get it recognised (sometimes it seems a bit flaky) but I can get it sampling Mar 27 15:04:44 s/can/can't Mar 27 15:04:58 so can recognised, can't sample Mar 27 15:05:19 there's a known bug where it'll sample the first time, but it'll hang the second time Mar 27 15:05:39 biot: I can't seem to set the sample rate Mar 27 15:07:30 biot: yes, the udev rule is there. i have a /dev/ttyACM0 and the symlink OpenLogicSniffer that the ols software wants Mar 27 15:07:51 biot: giving these by hand doesnt help either Mar 27 15:08:00 er Mar 27 15:08:18 you know I tried out pulseview for the first time in months last night Mar 27 15:08:29 can't really help you much with it :) Mar 27 15:08:42 jhol on #sigrok is the man for it Mar 27 15:08:47 I'll see where good old cli can get me ;) Mar 27 15:08:59 and then _maybe_ I can see if I can hlep with pulseview Mar 27 15:09:09 biot: oki.. thanks anyways :) Mar 27 15:09:14 jackmitchell: I'll take that as a firm vow Mar 27 15:09:18 but working with gtk (and as such glib) sends shivers down my spine Mar 27 15:09:32 jackmitchel: i got an OLS a couple weeks ago - i had lots of problems until i updated the firmware on it Mar 27 15:09:40 because gint so much better than a normal int Mar 27 15:09:48 oh yeah, update the firmware Mar 27 15:09:51 sigrok-cli --driver ols:conn=/dev/ttyACM0 --device samplerate=100k --samples 64 Mar 27 15:09:52 cmicali: ah, I had assumed it was at latest Mar 27 15:09:58 ^^ that'll do it Mar 27 15:10:00 jackmitchell: so had I, but it was way behind Mar 27 15:10:08 jackmitchell: and that fixed 90% of my issues Mar 27 15:10:23 cmicali: good shout, I'll look into that, I got your email earlier too, which I'll reply to soon ;) Mar 27 15:10:29 jackmitchel: no rush :) Mar 27 15:11:09 the OLS is nice because it can go up to 200mhz, but the salea logic i have is amazing, it really is worth the extra $50 Mar 27 15:11:15 well, $100 Mar 27 15:11:41 cmicali: can the salea do continuous sampling? Mar 27 15:12:07 no unfortunately, but it doesn't have an onboard buffer so you can use extremely large buffers client side Mar 27 15:12:24 it can actually, but IIRC the saleae software can't Mar 27 15:12:32 but sigrok can :) Mar 27 15:12:36 really! Mar 27 15:12:40 yes Mar 27 15:12:45 ah ok, so it can do straight to ram? or... if it doesn't have a buffer, where does it store the data.. Mar 27 15:12:49 * cmicali goes to download sigrok Mar 27 15:13:06 directly to ram, the client software has a buffer it writes into Mar 27 15:13:20 works well, but can be a bit dicey at 24mhz on slower or loaded machines Mar 27 15:13:24 everything is streaming in sigrok... even stuff like the OLS, which is literally a store + buffer dump -- the driver still streams it Mar 27 15:13:39 4 Mar 27 15:13:57 i think the logic 16 has an onboard buffer Mar 27 15:14:04 memory is hazy though Mar 27 15:14:16 if only the salea was a little bit faster, 50MHz say Mar 27 15:14:21 that is the problem for me Mar 27 15:14:23 panto, I see i2c2 for dvi+audio but not i2c1 Mar 27 15:14:37 the logic 16 can sample faster, but it's limited to a 2 channels at it's max rate Mar 27 15:14:40 jackmitchell: then it wouldn't be able to stream live Mar 27 15:14:51 ka6sox, you need i2c1? Mar 27 15:14:55 biot: where is the bottleneck? Mar 27 15:15:04 as cmicali said, at its max 24MHz rate, it'll saturate (more or less) the USB bus Mar 27 15:15:17 panto, no, there is a patch in audio that sets up i2c1 and its not used that I can tell. Mar 27 15:15:19 i.e. don't wiggle the mouse Mar 27 15:15:44 ah, so with USB 3 here it would be possible to bump it up a notch Mar 27 15:15:59 ka6sox: the nxp i2c moved busses in between revs Mar 27 15:16:26 jackmitchell: well yeah, for example the bladerf has USB 3 for that purpose Mar 27 15:16:35 koen, so both are needed still? Mar 27 15:17:14 ka6sox, point me to that patch plz? Mar 27 15:17:55 https://github.com/beagleboard/kernel/blob/3.8/patches/audio/0002-arm-dts-add-DT-pinmux-details-for-I2C1.patch Mar 27 15:19:26 av500: thunderbird works well as a crude proof of concept rss-to-imap gateway. I've found *piles* of preexisting scripts though and am going to evaluate them. Mar 27 15:21:46 dm8tbr: hmm Mar 27 15:21:58 dm8tbr: but no nice mobile app Mar 27 15:22:56 ka6sox, that's just the pinctrl definitions Mar 27 15:23:18 you have to enable i2c1 bus device for it to take effect Mar 27 15:25:11 ka6sox: no, only the most recent one Mar 27 15:25:51 ka6sox: iirc the rev with the old bus had the connector wired wrong, so there was never a working unit with that config Mar 27 15:26:08 koen, then I'll just drop that and free up the resource. Mar 27 15:26:09 av500: K9? Mar 27 15:26:22 thanks! Mar 27 15:27:26 dm8tbr: fast Mar 27 15:28:04 dm8tbr: but maybe Mar 27 15:39:58 cmicali: bingo, just updated the firmware (in a rather crude fashion, the utilities weren't very good) and it seems to have sprung into life, thanks Mar 27 15:40:16 jackmitchel: no problem, i beat my head against the wall for 4 hours before trying that Mar 27 15:53:11 Eleventy billion windows updates for my XP virtual machine. 5 or 7 at a time, then a reboot Mar 27 15:55:02 use a real OS :) Mar 27 15:55:07 or at least Ubuntu :P Mar 27 16:08:36 I hate this bone on my desk, I believe it's the silicon revision with the ethernet IRQ storm Mar 27 16:09:01 all the others are tied up in prototypes, I should really get another one Mar 27 16:11:07 jackmitchell: what rev? Mar 27 16:11:19 mranostay: A5 Mar 27 16:11:19 dumbledore board? Mar 27 16:12:01 mranostay: gandalf the white Mar 27 16:12:29 hmmm i thought that was fixed before A5 Mar 27 16:13:25 all I know is the board works fine, but the ethernet is dog slow Mar 27 16:14:19 and when I look at /proc/interrupts the the ethernet ones are climbing fast Mar 27 16:14:48 both int:57 and int:58 climb at about 30 a second Mar 27 16:45:50 mranostay: i need to re-think the bacon board some Mar 27 16:46:14 not enough fat? Mar 27 16:46:38 ka6sox-farfarawa: need it to be chocolate covered Mar 27 16:46:40 <_av500_> prpplague: add cheese Mar 27 16:46:54 _av500_: indeed Mar 27 16:47:10 <_av500_> but I am sure mranostay is happy to rip up an reroute Mar 27 16:47:15 <_av500_> he found the knack for it Mar 27 16:49:42 _av500_: i forgot about the new uart config Mar 27 16:50:09 <_av500_> you mean the debug uart? Mar 27 16:50:56 <_av500_> yesterday I soldered a 6pin FTDI header to an XM :) Mar 27 16:51:04 <_av500_> so I can use one cable for both Mar 27 16:51:11 _av500_: hehe Mar 27 16:51:23 _av500_: i've started using the ftdi cable with the 3.5mm jack Mar 27 16:52:27 <_av500_> I hate these Mar 27 16:52:36 <_av500_> forces are too high Mar 27 16:52:44 <_av500_> on prototypes etc Mar 27 16:53:27 _av500_: i have an adapter i built up to convert it to the 6pin Mar 27 16:53:41 _av500_: main thing is i cna use it easily with my toy hacing Mar 27 16:53:43 hacking Mar 27 16:53:52 <_av500_> yeah Mar 27 16:54:05 _av500_: http://www.elinux.org/InnoTab_3.5mm_Uart Mar 27 16:54:09 <_av500_> yes Mar 27 16:54:10 <_av500_> i know Mar 27 16:54:25 <_av500_> i have the same adapter :) Mar 27 16:54:32 <_av500_> our frenchies used 3.5mm too Mar 27 16:54:43 <_av500_> on tiny wee wires that broke all the time Mar 27 16:59:56 _av500_: hehe Mar 27 17:00:11 _av500_: not the same adapter... i made mine, hehe Mar 27 17:12:09 prpplague: rethink how? :) Mar 27 17:12:33 mranostay: i forgot about the new uart config Mar 27 17:12:40 grrr you want to pop a FTDI on it? :) Mar 27 17:13:04 mranostay: no, just that we had originally planned on it being a "wing" which would not cover up the uart connector Mar 27 17:13:11 prpplague: just put a stackable header for those 6 pins done Mar 27 17:13:20 ah that was reasoning for the wing Mar 27 17:13:28 and i thought jkridner was crazier than normal Mar 27 17:13:31 mranostay: hehe Mar 27 17:13:51 mranostay: so yea we need to maybe at a pass through or route the uart somewhere Mar 27 17:14:14 <_av500_> just pass it to the top Mar 27 17:14:18 mranostay: plus i need to get you the correct 7-seg symbol/package Mar 27 17:14:23 prpplague: do we have traces going through there? Mar 27 17:14:41 mranostay: i'll have to go back and look at your original config Mar 27 17:14:49 mranostay: i played around with it some last night Mar 27 17:15:03 always something Mar 27 17:15:23 mranostay: i have to finish the minnow x3 reviews and make sure the fab gets started on the trainer lure Mar 27 17:18:56 <_av500_> mranostay: we can always fix it in SW Mar 27 17:20:14 _av500_: what there is a magic via creater kernel module? Mar 27 17:26:18 mranostay: modprobe drill Mar 27 17:34:52 http://www.reddit.com/r/talesfromtechsupport/comments/1b2hp0/the_b_manager_from_hell_pt22_fallout/ Mar 27 18:04:26 _av500_: how badly would you mock me if i used that pcb art plugin Mar 27 18:04:47 what is a pcb art plugin? Mar 27 18:06:49 Russ: see Brooks comment on my latest post Mar 27 18:09:05 I usually use inkscape for my pcb art, but the image posted just seems like overkill Mar 27 18:09:30 a pcb should be aesthetically pleasing, but still functional Mar 27 18:40:08 <_av500_> mranostay: awesome Mar 27 18:40:47 av500: i'm sure the curved traces wouldn't get a mocking Mar 27 18:41:34 <_av500_> if you can make it as artsy as that Mar 27 18:42:15 * mranostay spies a wild jkridner Mar 27 18:42:15 many of the hand-drawn boards of old were works of arts Mar 27 18:42:53 mru: now it is all autorouter crap? Mar 27 18:43:12 mranostay: wild and crazzzzy? Mar 27 18:43:44 mranostay: not many people use autorouter Mar 27 18:43:57 auto router boards are a fast way to a poorly working board Mar 27 18:44:05 yeah well of course Mar 27 18:44:28 mranostay: either they cannot afford a good autorouter (which is damn expensive and costs a lot of time to annotate all constraints) or they simply dont know what they are doing Mar 27 18:46:23 even manually routed boards look mechanical Mar 27 18:46:35 compared to hand-drawn traces Mar 27 18:46:57 you mean hand taped ? Mar 27 18:47:06 drawn, taped, whatever Mar 27 18:47:09 I hand drew a board once Mar 27 18:47:30 a wild Crofton|work appears! Mar 27 18:47:59 i can tell your too young to remember a light box and rolls of tape and pads Mar 27 18:48:20 you're right, I've never done it myself Mar 27 18:48:27 but I've seen such boards Mar 27 18:48:35 <_av500_> Wulfman: I have such tape somewhere Mar 27 18:48:41 Crofton|work: how it turn out? Mar 27 18:48:43 it was fun when you needed to rip and retry lol Mar 27 18:48:49 and what year was this? Mar 27 18:48:56 earily 80s Mar 27 18:49:08 I forget Mar 27 18:49:14 i went so fast to pads cad Mar 27 18:49:17 the board was ok, not sure the circuit worked Mar 27 18:49:23 it was so much easer Mar 27 18:50:02 <_av500_> here is your "board" http://www.luedeke-elektronic.de/images/product_images/popup_images/3899_0.jpg Mar 27 18:53:43 Wulfman: i used to watch my dad with the tape and dots Mar 27 18:53:44 :p) Mar 27 18:53:52 and i made my first pcb that way Mar 27 18:53:53 :-) Mar 27 18:55:24 its the old ways Mar 27 18:55:37 now is much better and faster Mar 27 19:02:24 Crofton|work: wtf is Forest, VA? Mar 27 19:03:05 a place Mar 27 19:04:18 hmm, pub dinner or not, that is the question... Mar 27 19:05:09 nomnomnom Mar 27 19:06:07 is that a yes? Mar 27 19:06:33 indeed it is Mar 27 19:07:49 why not a low end frozen dinner roulette? Mar 27 19:07:59 you might even land on horse protien! Mar 27 19:08:14 mru: drink your dinner Mar 27 19:15:30 * KotH recomends kfte Mar 27 19:15:31 ;) Mar 27 19:15:54 not on the menu today Mar 27 19:16:05 ask your friendly turkish restaurant Mar 27 19:16:13 they'll be happy to help you with that Mar 27 19:17:07 oh I'm sure it can be had somewhere in the city Mar 27 19:18:07 kebab Mar 27 19:19:46 hi everybody Mar 27 19:20:14 I'm struggling with an issue and i thought somebody could give me a hand Mar 27 19:20:32 i'm trying to debug the kernel using CCS following http://processors.wiki.ti.com/index.php/Linux_Debug_in_CCSv5#Stop_Mode_Debug Mar 27 19:21:30 everything seems to work fine until the point where i have to launch the debug session Mar 27 19:21:34 .c Mar 27 19:21:47 i obtain an error saying: An internal error occurred during: "Launching am335x_debug.ccxml". java.lang.NullPointerException Mar 27 19:22:14 i found that there are some other folks that have gone through this http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/t/185703.aspx?pi73866=1 Mar 27 19:22:20 but they don't clarify how Mar 27 19:22:43 has anybody debug the kernel on a beaglebone? Mar 27 19:23:34 <_av500_> yes Mar 27 19:23:38 <_av500_> but not with ccs Mar 27 19:24:05 i'm open to other posibilities Mar 27 19:24:13 would you mind sharing how you did it? Mar 27 19:25:15 erle: you probably have the wrong .ccxml for your ccs version Mar 27 19:28:07 mru i might be doing something wrong, however i create a new .ccxml containing just the Texas Instruments XDS100v2 USB Emulator and selecting AM335X as the board Mar 27 19:28:20 mru and the connection works Mar 27 19:28:56 oh, then I don't know Mar 27 19:29:14 jkridner: ping Mar 27 19:29:28 mru too bad :(, still, thanks a lot for the help Mar 27 19:34:39 <_av500_> erle: not many people here using ccs Mar 27 19:34:47 <_av500_> most use printk :) Mar 27 19:36:28 I sometimes use printf on u-boot Mar 27 19:36:50 when I want to let down my hair Mar 27 19:39:29 * ogra_ sometimes uses CSS ... can make printf and prink output shiny Mar 27 19:52:44 all right :) thanks for your tips Mar 27 19:53:21 ccs is anything but right Mar 27 19:54:18 mdp: growing that mullet finally? Mar 27 19:54:28 you know it Mar 27 19:54:48 to match the neckbeard? Mar 27 19:55:03 * mru hides in terror of the mental image Mar 27 20:03:46 <_av500_> finally it wont bring down the AT&T network: http://linuxgizmos.com/first-yocto-compatible-carrier-grade-linux/ Mar 27 20:05:16 i found a bargin Mar 27 20:05:20 an energy creator Mar 27 20:05:27 Input power 0.5-8W Mar 27 20:05:27 Input power SSB 5-15W Mar 27 20:05:27 Output Power 50W Mar 27 20:05:45 lol.. gotta love ebay adverts Mar 27 20:06:38 <_av500_> buy 2 Mar 27 20:06:51 they don't accept bitcoin Mar 27 20:07:06 <_av500_> try snake oil Mar 27 20:08:31 heh magic Mar 27 20:09:19 im considering buyin one of these Mar 27 20:09:20 http://www.ebay.com/itm/T962A-Infrared-SMD-BGA-IC-Heater-Reflow-Oven-30X32CM-/171009769536?_trksid=p2045573.m2042&_trkparms=aid%3D111000%26algo%3DREC.CURRENT%26ao%3D1%26asc%3D27%26meid%3D6545823992251792085%26pid%3D100033%26prg%3D1011%26rk%3D1%26sd%3D171009769536%26 Mar 27 20:10:04 tinyurl mofo! Mar 27 20:11:34 i got a grumpy email that i was attempting to profiteer on beagle cape connectors Mar 27 20:11:37 geepers. Mar 27 20:12:15 mranostay, dangerousprototypes has one, you might watch their video Mar 27 20:12:21 er, damn autocomplete Mar 27 20:12:22 mrpackethead, Mar 27 20:12:35 I went with a convection toaster oven Mar 27 20:12:41 i have one Mar 27 20:12:43 I don't think the T962A has convection Mar 27 20:12:45 it works ok. Mar 27 20:13:08 mrpackethead: profiteer on what now? Mar 27 20:13:10 i have an oven that Mar 27 20:13:24 well, apparently, at $89/100 i'm making a massive profit Mar 27 20:14:16 <_av500_> 2nd ferrari? Mar 27 20:14:50 mrpackethead: for what headers? Mar 27 20:14:59 <_av500_> bone stcking ones Mar 27 20:15:01 stacking connectors for capes Mar 27 20:15:03 <_av500_> 2x2x23 Mar 27 20:15:07 i've bought 1000 Mar 27 20:15:11 have 600 surplus Mar 27 20:15:12 from who? adafruit? Mar 27 20:15:16 <_av500_> china Mar 27 20:15:21 taiwan Mar 27 20:15:22 you didn't get the ones with the really long pins, right? Mar 27 20:15:38 they are the same as the ones that are on the Audio Cape Mar 27 20:18:12 the MOQ was 1000 pces Mar 27 20:18:27 but 1000 was cheaper than 400 x $2.10 Mar 27 20:21:56 http://www.youtube.com/watch?v=yRxcYO0nuD8 Mar 27 20:22:01 wow, pnp for $3500! Mar 27 20:22:59 mrpackethead: angry email from who? Mar 27 20:23:16 some plonker on teh beagle-list Mar 27 20:23:26 300 of the 600 spares are sold now Mar 27 20:32:15 <_av500_> any went to germany? Mar 27 20:34:24 _av500_: planning on robbing the shipment? Mar 27 20:34:30 no.. Mar 27 20:34:31 :-) Mar 27 20:34:43 _av500_: are you in .de and need some Mar 27 20:34:43 mrpackethead: how much to ship to the states? Mar 27 20:34:52 $89 including freight Mar 27 20:34:54 for 100 Mar 27 20:35:24 which is? Mar 27 20:35:48 $89 for 100 connectors, delivered Mar 27 20:36:26 or you can pay $59 for the connectors, and $30 for the freight Mar 27 20:36:30 you choose Mar 27 20:36:32 i dont' mind Mar 27 20:36:33 <_av500_> speaking of PCBs: http://www.adafruit.com/adablog/wp-content/uploads/2013/03/P3140798_600.jpg Mar 27 20:36:49 <_av500_> :) Mar 27 20:37:51 mrpackethead: ok i was just wondering if was extra :P Mar 27 20:38:16 no. Mar 27 20:38:39 _av500_: one of your early early early preOMAP1 protos? :D Mar 27 20:38:49 i'll share any profits around Mar 27 20:38:56 as long as you take bitcoin. Mar 27 20:39:13 <_av500_> ds2: http://kuzyatech.com/keithley-197-microvolt-dmm-in-pictures Mar 27 20:39:23 <_av500_> guard traces Mar 27 20:39:25 <_av500_> nice Mar 27 20:40:09 <_av500_> http://kuzyatech.com/wp-content/uploads/2013/03/P3140786.jpg Mar 27 20:41:59 mrpackethead: take paypal? Mar 27 20:42:49 yes. Mar 27 20:42:52 that is easy Mar 27 20:43:39 wont' have them for a week or two Mar 27 20:43:45 so don't stress on it Mar 27 20:44:22 do i look stressed? :) Mar 27 20:45:05 <_av500_> did you reroute bacon already? Mar 27 20:45:12 <_av500_> checked all angles? Mar 27 20:50:20 heh Mar 27 20:50:34 won't be a big reroute Mar 27 21:24:08 i will be selling a reflow controller Board on ebay for about 90 bux that you put in a convection oven for reflowing boards Mar 27 21:24:39 the ovens you see on ebay are not that good if you search for reviews on them Mar 27 21:45:42 is #beagle now a swap meet? Mar 27 21:45:51 <_av500_> yes Mar 27 21:46:10 <_av500_> anybody got hub caps for a 72 pinto? Mar 27 21:47:52 i have some slighlty used low milage omap4 and omap5 products.... i only used them on sunday for church projects Mar 27 21:48:51 lo Mar 27 21:48:58 lol Mar 27 21:54:21 * KotH wonders why people think that buffer will magically speed up any data transfer Mar 27 21:54:52 <_av500_> add more buffers Mar 27 21:54:58 <_av500_> to be even faster Mar 27 22:03:48 _av500_: nice EE-porn :) Mar 27 22:04:40 <_av500_> :) Mar 27 22:09:20 hey not at work now! Mar 27 22:09:32 do that at home :P Mar 27 22:13:22 http://www.youtube.com/watch?v=xzIv5DDVLKk Mar 27 22:13:55 mranostay: i am at home :) Mar 27 22:14:45 http://kuzyatech.com/part-of-the-week-tag-connect <- nice stuff Mar 27 22:26:29 tag connect looks strangely like a beacon board Mar 27 22:27:02 well.. they are basically probe pins in a connector Mar 27 22:27:09 prpplague: will they ressurect themselves this Sunday? Mar 27 22:27:16 nothing fancy.. stufy any EE has done hundreds of times in his life Mar 27 22:27:21 I mean the layout style Mar 27 22:27:22 but these are ready made sold :) Mar 27 22:27:25 ah.. Mar 27 22:27:34 dont know the beacon board Mar 27 22:27:58 hey, something to program cape eeproms Mar 27 22:29:24 * KotH has no cape Mar 27 22:32:47 _av500_: hehe i should delete your comment, hehe Mar 27 22:33:13 <_av500_> yeah, that's fool people Mar 27 22:33:16 <_av500_> that'll Mar 27 22:33:22 hehe Mar 27 22:33:27 <_av500_> the pcb looks kinda darkish Mar 27 22:34:00 odd lighting and it hasn't been washed yet Mar 27 22:34:11 <_av500_> right Mar 27 22:34:18 <_av500_> it'll come out snow white Mar 27 22:35:04 indeed, hehe Mar 27 22:44:58 * Russ falls out of his chair Mar 27 22:45:29 * KotH puts a carpet over Russ Mar 27 22:49:20 Russ: something interesting? Mar 27 22:56:37 * Russ is glad to be safely under the carpet at this point Mar 27 22:59:21 * ccssnet fills the room with packing peanuts while russ is busy under the carpet Mar 27 23:10:49 pink ones i hope Mar 27 23:30:40 jkridner: ping Mar 27 23:37:10 * mranostay hurls NAND chips at ds2 Mar 27 23:38:02 * ds2 catches them and turns them into a RAID Mar 27 23:38:16 make a raid of NOR chips :P Mar 27 23:40:27 I trust my NOR chips Mar 27 23:50:43 ds2: for how many writes? :) Mar 27 23:51:12 more those those pesky MLC NANDs Mar 27 23:51:18 heh Mar 27 23:51:26 * ds2 hugs his 29F010's Mar 27 23:51:26 i thought they are all SLC now Mar 27 23:51:40 eh? Mar 27 23:52:48 * mranostay pokes mru Mar 27 23:52:57 * mru ducks Mar 27 23:57:33 mmmm duck Mar 28 00:01:43 goose Mar 28 00:02:07 mongoose Mar 28 00:05:02 mongoose is so misleading Mar 28 00:11:19 mongooses are cool Mar 28 00:11:27 we used to have them living in our backyard Mar 28 00:11:42 they are way cooler than squirels Mar 28 00:13:42 is the plural not mongeese? Mar 28 00:16:47 mru: hmmm i wonder Mar 28 00:17:11 The plural form is mongooses,[3] or, rarely, mongeese.[4] It has also been spelled "mungoose".[ Mar 28 00:17:38 hah, trolled you Mar 28 00:17:55 http://oxforddictionaries.com/definition/english/mongoose Mar 28 00:18:17 webster does mention mongeese though Mar 28 00:22:00 * mru trolls apple dude on llvm list Mar 28 00:22:05 this could get fun Mar 28 00:22:30 linky-poo Mar 28 00:22:36 hey! leave LLVM alone ;) Mar 28 00:22:59 ds2: go play with your NAND chips : Mar 28 00:23:01 :P Mar 28 00:23:41 ds2: just doing my job Mar 28 00:24:48 mru: which one? linaro-gcc, gcc, armcc, or? Mar 28 00:25:38 which what? Mar 28 00:25:46 which job Mar 28 00:26:00 linaro toolchain Mar 28 00:29:04 mru: trolling ain't eay Mar 28 00:29:06 *easy Mar 28 00:31:33 I didn't choose the troll life? Mar 28 00:33:04 you are born a troll Mar 28 00:33:10 or rather spawned one Mar 28 00:35:56 Russ: since you use UML... what do you think about getting it going on ARM? Mar 28 00:41:04 it'd be fun Mar 28 00:41:23 think it would be useful though? Mar 28 00:41:42 probably not, especially with all the container stuff Mar 28 00:42:43 but containers don't allow for new kernels Mar 28 00:43:21 uml kernels are specially built Mar 28 00:43:53 I know Mar 28 00:44:06 but for some features like binder, it can be done in the UML side only Mar 28 00:44:24 you think you could implement binder in a uml kernel without any host kernel support? Mar 28 00:44:45 maybe Mar 28 00:44:56 it has a better chance then a device driver talking to HW Mar 28 00:45:26 if its really binder you want, you might be able to do it with a LD_PRELOAD, it'd be slower than the real binder Mar 28 00:46:04 who knows, maybe someone will step up with a binder wrapper for af_dbus Mar 28 00:46:15 *shrug* Mar 28 00:47:02 just thinking of GSoC ideas Mar 28 00:47:52 ds2: wash my car app Mar 28 00:48:05 mow my lawn app is another idea Mar 28 00:48:18 much rather have toss mronostay's car into the bay app Mar 28 00:48:39 hey i have insurance go ahead Mar 28 00:48:57 ds2: btw tab/ typing fail :) Mar 28 00:49:53 there was a good article somewhere, maybe lwn, on current state of android mainline efforts recently Mar 28 01:50:34 trouble is back Mar 28 01:50:35 :p) Mar 28 01:57:42 i see that Mar 28 02:18:35 back back back Mar 28 02:39:33 front front front Mar 28 02:50:16 mranostay, you here? Mar 28 02:50:35 ka6sox: think he's headed out Mar 28 02:50:49 ka6sox: yeah Mar 28 02:50:55 wasn't there some comments about emptying barrels of brew? Mar 28 02:51:02 * mranostay opens a beer Mar 28 02:51:22 where is the groundplane? Mar 28 02:51:37 it is there Mar 28 02:51:59 I *don't* see. Mar 28 02:55:22 dotted lines Mar 28 02:55:28 that is just an outline from Eagle Mar 28 02:56:11 ka6sox: eagle is quirky about the flood fills Mar 28 02:56:22 ka6sox: if you hit the ratsnest Mar 28 02:56:26 eh? Mar 28 02:56:27 prpplague: have to run RAT Mar 28 02:56:28 ka6sox: it will do the fill Mar 28 02:56:39 oh you are calling THAT quirky Mar 28 02:57:07 ds2: eagle is quirky in general, that is just one of the oddities Mar 28 02:58:57 prpplague: compared to the older classic stuff, it isn't that quirky Mar 28 02:59:27 class == non PC based stuff Mar 28 02:59:52 ds2: yea well, i can old compare to what i know **** ENDING LOGGING AT Thu Mar 28 02:59:58 2013