**** BEGIN LOGGING AT Wed Oct 30 02:59:58 2013 Oct 30 03:31:55 My beagleboneblack is not working. The Power light comes on but the USR0,1,2,3 show no change Oct 30 03:32:00 They're all silent Oct 30 04:53:26 Hi beagle. I successfully interfaced tca8418 keypad with beaglebone-black. It working with direct keys of QWERTY keymap.But i need multiple keys setup. Foe example Q key can be used for * after presssing some Fn key. Oct 30 04:53:59 I tried with Documentation/devicetree/bindings/input/matrix-keymap.txt. Oct 30 04:55:29 I cant figure hw it actually works? Anyone tried with linux,fn-keymap before? Oct 30 04:57:16 And also one another problem i faced with gpio as irq for this keyboard. The kernel hangs while i configure the gpio1_16 as my irq in my dtsi file.I think something problem with gpio-controller configuration in bbb.Please check it out. Oct 30 08:12:15 yay just got my first word to execute in my arm forth. it was "bye" but it executing proved the hashed dictionary searches work Oct 30 08:12:21 small step but an important one Oct 30 08:12:47 had a few stupid bugs in the kernel that i have uncovered so far and i know thers going to be more but im on the home stretch now :P Oct 30 08:13:31 arm forth? Oct 30 08:13:54 as the programming language? Oct 30 08:14:25 * KotH blinks and wonders whether there arent already a dozen arm forth kernels available Oct 30 08:14:26 forth the programming language Oct 30 08:14:32 for ar, Oct 30 08:14:33 arm Oct 30 08:14:53 99% of which will be ANS forth Oct 30 08:15:03 which is an abomination worse than the C programming language Oct 30 08:15:55 ah.. i see i see Oct 30 08:16:02 * KotH never got into forth Oct 30 08:16:29 probably because i used asm for low level stuff, and perl for everything else Oct 30 08:17:00 this will be released for both the beagle board and the raspberry pi but ill have to fix my pi before i release it for that Oct 30 08:17:17 its based on my linux x86 forth at www.isforth.com Oct 30 08:56:16 checking for XML::Parser... configure: error: XML::Parser perl module is required for intltool Oct 30 08:56:16 <--- err wth? Oct 30 08:56:24 erm misschan Oct 30 08:56:55 is it written in asm? Oct 30 08:57:03 or what is so ARM specific? Oct 30 08:58:06 its written in asm Oct 30 08:58:12 i dont write forth kernels in anything but asm Oct 30 08:58:13 ever Oct 30 08:58:27 any forth not written in forth is not forth Oct 30 08:59:08 my forth kernel is written in forth/asm but the gnu assembler assembles it Oct 30 08:59:15 any forth not written in forth/asm i mean Oct 30 09:01:59 can one write forth in forth? Oct 30 09:02:13 and so forth.... Oct 30 09:02:16 yes Oct 30 09:02:29 all of a forth system is written in either forth or in assembler Oct 30 09:02:45 you write your primitives in assembler and use those to create higher level forth definitions Oct 30 09:03:55 this way all you have to do is write an assembler extension so that your forht compiler can also assemble, a meta compiler (like a target compiler) and your forth system can now be entirely self replicating Oct 30 09:04:08 use the forth + assembler + metacompiler to compile the sources to the forth kernel Oct 30 09:04:33 use the forth kernel to compile the various extensions.. yu now have an updated version of your entire forth system Oct 30 09:16:22 mark4: do you have some code / documentation I can look at? Oct 30 09:16:38 http://www.isforth.com/ looks... weird Oct 30 09:28:08 forth runs past my house, it is very wet Oct 30 09:32:30 and speaking of forth the language I have still not got my hands on a jupiter ace :-) Oct 30 09:46:07 hi guys~ i have a question. if i install linux omap kernel again, my beagleboard become reset? Oct 30 09:46:17 keesj, i do for the x86 compiler Oct 30 09:46:21 at www.isforth.com Oct 30 09:46:29 sdfafaf: ??? Oct 30 09:46:32 im not releasing the arm one till its more "Stable" Oct 30 09:46:48 xora you WANT an ace? Oct 30 09:46:49 lol Oct 30 09:47:59 @ i'm downloading omap linux kernel for using gpio from git Oct 30 09:50:51 from what git? Oct 30 09:51:07 and you can use gpio with the stock angstrom image Oct 30 09:52:13 mark4: I have a zx80 and 8 zx81's, they pine for their lost sibling :-) Oct 30 09:52:25 http://devnull-as-a-service.com/ Oct 30 09:52:53 xora do you have the ram pack for the zx81? Oct 30 09:52:57 did you shave off its feet? Oct 30 09:53:33 mark4: I think I PVC taped it on Oct 30 09:53:54 mark4: oh, one of them I decased, it held much better without the cased Oct 30 09:53:56 mark4: just a few lines of code or anything would do. otherwise I don't care Oct 30 09:54:11 keesj a few lines of my arm code? Oct 30 09:58:37 mark4: and the memotech ram addons don't suffer wobble issues Oct 30 09:58:45 no, just anything , what does a forth "kernel" or driver look like Oct 30 09:58:57 www.isforth.com Oct 30 09:59:08 that has the entire sources to my linux x86 forth compiler Oct 30 09:59:12 32 bit Oct 30 09:59:43 xora a 3rd party ram module? Oct 30 10:00:39 mark4: yes, its case is shaped to fit snugly on back of a zx81, these days I have a ZX-Pand in a memtech case Oct 30 10:01:53 xora i had an x81 with the original ram pack, i lost a 16k program i spent all day typing in. i put the zx81 away and bought a c64 Oct 30 10:02:07 well. actually i bought that a few weeks later after i learned 6502 asm Oct 30 10:02:50 * XorA laughs Oct 30 10:03:03 I was a Sinclair guy until the Amiga Oct 30 10:03:24 yea i moved to the amiga late. i got a 1200 but never did any real coding with it Oct 30 10:03:43 my A500 with HD got me through programming courses at UNI Oct 30 10:03:48 i lived in cambridge england at the time of the zx81 tho i never met sir clive Oct 30 10:03:55 saved me hours in their shoddy labs Oct 30 10:04:26 i put a desktop hard drive in my 1200, wish i could have afforded a laptop drive :) Oct 30 10:05:59 I have a 16G compact flash in my A1200 these days Oct 30 10:06:20 oh how such tech became cheap too late Oct 30 10:07:15 amigaos is the best multi tasking os i ever saw Oct 30 10:07:28 copy first external floppy to second external floppy while formatting internal floppy. Oct 30 10:07:41 meanwhile have two instances of the same chess program plahying chess against each other Oct 30 10:07:44 * XorA waits for the RiscOS fans in the channel to erupt :_D Oct 30 10:07:58 with music playing in the background and sprites boucning randomly all over the place and not one stutter Oct 30 10:08:06 just isnt a safe os Oct 30 10:08:11 any app can crap all over any app Oct 30 10:08:24 mark4: thanks I found the download Oct 30 10:08:52 i call that the fastest compiler of ANY non trivial language btw keesj Oct 30 10:08:55 Amiga + 1 Oct 30 10:09:15 amiga <3 Oct 30 10:09:16 nobody in over 10 years has shown me any compiler of any non trivial lang that can compile as fast as that compiler does Oct 30 10:09:22 but, it's dead jim Oct 30 10:09:31 stop kicking that dead horse Oct 30 10:09:41 dead ? o_O Oct 30 10:09:42 dice was a good c compiler but im not very impressed with dragonfly Oct 30 10:10:01 idiot removed forth as the boot loader for dragonfly lol Oct 30 10:10:06 corpsicle, depends on your definition of alive I guess Oct 30 10:10:06 I remember DICE, I could never afford a license so I could not use floats Oct 30 10:10:15 if not completely decomposed OK Oct 30 10:10:25 brb lunch Oct 30 10:10:35 make me lunch! Oct 30 10:10:39 panto: still new Amiga hardware manufactured, so not stopped breathing yet :-D Oct 30 10:10:41 sudo make me lunch Oct 30 10:11:01 xora the 68k processor is no longer manufactured Oct 30 10:11:13 you cannot buy it other if you find someone with it stockpiled Oct 30 10:11:23 mark4: there is large stocks though, I just bought a new 68030 accellerator card Oct 30 10:12:16 and there is already an FPGA clone Oct 30 10:12:36 is the clone as fast as stock? Oct 30 10:12:41 faster Oct 30 10:12:42 12mhz? Oct 30 10:12:57 the 020 was 12mhz not sure about the 030 Oct 30 10:13:30 40 or 50Mhz Oct 30 10:13:46 bitbang 68k on the PRU Oct 30 10:13:51 i did like 68k asm i just never did it much Oct 30 10:14:28 I think there is a propeller bitbanging 68k as well Oct 30 10:14:39 yea not really interested in the prop Oct 30 10:14:54 or arduino Oct 30 10:15:02 yet i like avr Oct 30 10:15:12 i have an unreleased avr forth too Oct 30 10:15:29 * av500 wants 74xx forth Oct 30 10:15:43 74xx - that an NEC device? Oct 30 10:15:56 forth would make arduino much more fun Oct 30 10:16:26 http://en.wikipedia.org/wiki/7400_series Oct 30 10:16:38 i developed my forth for the atmega32u4 but the assembler will work with ANY avr device Oct 30 10:16:48 and wont allow opcodes not supported by the selected device Oct 30 10:17:00 oh TI Oct 30 10:18:01 7400 where you can use quad 2 input nand gates to build anything :-D Oct 30 10:18:28 forth2rtl Oct 30 10:18:44 chinese forth ? Oct 30 10:18:57 ogra_: RRPN? Oct 30 10:19:02 I heard forth is considered very unlucky in china Oct 30 10:19:37 a lot of buildings don't even have a forth floor Oct 30 10:19:54 like 13 in this country Oct 30 10:19:59 how evil ... do you have to jump from the third to the 5th ? Oct 30 10:20:04 1 2 3 4 5 6 7 8 9 10 11 12 14 Oct 30 10:20:06 mark4, or java Oct 30 10:20:20 java isnt as evil as c Oct 30 10:20:22 or vb Oct 30 10:20:36 developing embedded applications in C is like opening a can Oct 30 10:20:38 with a rock Oct 30 10:20:57 yes, clearly, since the majority of the world's infrastructure runs on c Oct 30 10:21:35 thats because every moron and his autistic brother can learn to code C in 24 hours and be "Experts" Oct 30 10:21:46 yes, we are all moron here Oct 30 10:21:48 +s Oct 30 10:21:59 I'll invite my autistic brother Oct 30 10:22:02 no, I like, "we are all moron" better Oct 30 10:22:15 down here we all moron Oct 30 10:22:17 develop your non trivial application in c. ill do the same application in forth - ill be finished in hours what takes you months Oct 30 10:22:27 and what are you trying to say about autistic people here? Oct 30 10:22:28 your binary will be hundreds of times larger than mine Oct 30 10:22:30 really? Oct 30 10:22:34 are you 12? Oct 30 10:22:50 12 and know what a zx81 is? lol Oct 30 10:23:02 oh, ok, so you are just bigoted Oct 30 10:23:31 or have more experience with both C development and forth development than you? maybe? Oct 30 10:23:40 mark4, wut? Oct 30 10:23:49 * av500 checksout forth projects on github Oct 30 10:24:05 mark4, so forth developers enjoy making fun of autistic folks? Oct 30 10:24:29 Russ: only a moron would think so Oct 30 10:24:41 we all moron here av500 Oct 30 10:24:50 I moron more! Oct 30 10:25:02 don't moron on my moron! Oct 30 10:25:33 now, if there was forth in .js Oct 30 10:30:09 I love airport wifi "Ping reply from NickServ: 40.75 second(s)" Oct 30 10:30:32 it's because their routeurs are not written in forth Oct 30 10:30:52 +1 Oct 30 10:31:01 I think they are sending packets back and forth Oct 30 10:31:06 I don't know, I do feel like I'm trying to open a can with a rock here Oct 30 10:31:08 * ogra_ prefers back to forth anyway Oct 30 10:31:33 I tried to open the can but I banged the rock on my fingers Oct 30 10:31:37 now I type with a limp Oct 30 10:31:40 Russ: maybe you CAN rigROCK it? Oct 30 10:31:48 sig* Oct 30 10:31:52 panto: you're ready for java Oct 30 10:32:17 das: java is for people that dont even have an autistic brother Oct 30 10:32:21 I figured opening a can with a rock is actually very efficient, so long as you don't care about the contents Oct 30 10:32:48 in forth the can opens you Oct 30 10:33:07 ouch Oct 30 10:33:13 who cares about contents of rocks anyway Oct 30 10:33:21 gold miners? Oct 30 10:34:00 .js rock stars? Oct 30 10:34:36 mark4, let us know when to stop making fun of you Oct 30 10:34:51 ah, here we go http://i.imgur.com/8jnvgsG.gif Oct 30 10:34:57 I'll stop now, I have a can to open Oct 30 10:35:11 y u no scm isforth.com? Oct 30 10:35:29 I'll write a filesystem in forth Oct 30 10:35:42 av500, as long as it's a database Oct 30 10:36:03 hi! i would like to change the default access rights and owner of the user led files of my beaglebone black (e.g. /sys/class/leds/beaglebone:green:usr0/brightness). doing it manually works, but after reboot, the defaults (root:root) are restored again. how can i get Debian to automatically set 'debian Oct 30 10:36:14 as owner of the devices? Oct 30 10:36:17 udev script Oct 30 10:36:20 rule Oct 30 10:36:23 suonds like modifying udev rules. Oct 30 10:36:25 Russ: ++ Oct 30 10:36:26 i tried to, didn't work Oct 30 10:36:46 i had SUBSYSTEMS=="leds", GROUP="debian" Oct 30 10:37:19 didn't change anything Oct 30 10:38:24 mark4, why are there executables and object files in the isforth release? Oct 30 10:39:10 i also tried other combinations with SUBSYSTEMS=="gpio.leds" or KERNELS=="gpio" etc. but nothing worked Oct 30 10:39:40 http://www.reactivated.net/writing_udev_rules.html#basic Oct 30 10:39:46 and the udev rules file was /etc/udev/rules.d/50-leds.rules, that should work, right? Oct 30 10:39:50 heh, the isforth source is in assembly Oct 30 10:39:52 there are no object modules, the executable you create Oct 30 10:40:00 you need the nasm assembler installed Oct 30 10:40:15 mark4, why is there a kernel.o in the source? Oct 30 10:40:17 make will create the kernel called kernel.com <-- a joke from my dos days Oct 30 10:40:42 Russ: thanks, that was my yesterday lecture, didn't bring me further Oct 30 10:40:50 that file wont always exist so im not bothering with worrying about it Oct 30 10:41:01 after doing make and creating kernel.com you run ./extend Oct 30 10:41:05 that creates ./isforth Oct 30 10:41:07 hanzzz, did you read the "testing and debugging" section? Oct 30 10:41:10 which is the fully extended forth Oct 30 10:41:28 don't put object files in a source release Oct 30 10:41:43 that shouldnt have been in there lol Oct 30 10:41:48 but im not worried about it Oct 30 10:42:02 just run make. it will assemble the kernel sources and create a new one Oct 30 10:42:09 Russ: yes, running the test never showed any changes Oct 30 10:42:19 and also, it's much more community friendly to put this sort of thing on github rather than dump tarballs over the wall Oct 30 10:42:30 i dont use github Oct 30 10:42:39 with monitor i rechecked that i'm using the correct subsystem (leds) Oct 30 10:42:52 mark4, you don't get much of a choice in what other people know and use though Oct 30 10:42:55 Russ: github is moron Oct 30 10:43:08 does it have an autistic brother ? Oct 30 10:43:10 av500, there is a talk somewhere on why you should stop using git Oct 30 10:43:15 gitorious Oct 30 10:43:24 heh Oct 30 10:43:25 * av500 uses that Oct 30 10:44:00 does forth implement argument parsing on the stack? Oct 30 10:44:17 I wish I could see more gitoris Oct 30 10:44:28 looks like for x86, the first argument is passed in register (ebx) and the second on the stack Oct 30 10:45:22 thats implementation specific Oct 30 10:45:30 the ebx is a cached top of stack. not all forths do that Oct 30 10:45:39 this code would really be a lot simplier in c Oct 30 10:45:48 and then you'd be able to compile for other arches easily too Oct 30 10:46:11 and then it would not be forth :) Oct 30 10:46:19 I mean the kernel Oct 30 10:46:24 and it would not be simpler in C Oct 30 10:46:26 your wrote the kernel in x86 assembly Oct 30 10:46:56 but the assembler produces exactly the same object code that a forth+assembler extension+meta compiler would produce Oct 30 10:46:58 the kernel is forth Oct 30 10:47:02 not c. Oct 30 10:47:32 so if the kernel is forth, why do I see no forth under src/kernel/ Oct 30 10:47:56 you do. but you dont recognize it as such Oct 30 10:47:59 look at macros.s Oct 30 10:48:11 'mov esi, [dp+5]' is forth? Oct 30 10:48:35 that would be mov esi, [ ' dp >body ] Oct 30 10:48:40 in an assembler extension Oct 30 10:48:42 which i dont have yet Oct 30 10:49:09 ' shr ebx, byte 1' is forth? Oct 30 10:49:21 it being asm does not make it not forth Oct 30 10:49:29 the individual opcodes are "Coded" definitions Oct 30 10:49:40 in forth terminology that means that that definition was coded in asm Oct 30 10:49:44 so I can compile this just fine for arm then? Oct 30 10:49:46 a "colon" definition is high levle forth Oct 30 10:49:58 no. the kernel is not portable Oct 30 10:50:02 nor should it be Oct 30 10:50:04 so its x86 assembly Oct 30 10:50:13 Russ: yes, its forth Oct 30 10:50:26 you can keep splitting hairs if you want Oct 30 10:50:42 it compiles with nasm, it isn't forth Oct 30 10:51:03 I may not be a very smart man, but I know that nasm isn't a forth compiler Oct 30 10:51:10 when i write the assembler extension and metacompiler the fully extended forth will be used to COMPILE the kernels soruces Oct 30 10:51:20 the result will be a kernel.com that is identical to the one produced by nasm Oct 30 10:51:43 this is a BETA pre-release Oct 30 10:51:46 so you are writing an assembler for 16-bit x86 code Oct 30 10:51:48 tho its been in beta for a long time Oct 30 10:51:59 in forth Oct 30 10:52:11 i wont condsider it a finished work until it has its assembler extension and meta compiler Oct 30 10:52:13 32 bit Oct 30 10:52:21 Russ: think of the cans! Oct 30 10:52:23 and maybe 64 bit but not 16 Oct 30 10:52:24 ... and so forth ... Oct 30 10:52:44 why nasm and not as? Oct 30 10:53:15 because AS was never intended to be used as a general purpose assembler. i was forced to use gnu as for the arm verison Oct 30 10:53:31 it took me 6 months to get the macros.s stuff done with the gnu assembler Oct 30 10:53:36 it took 10 minutes with nasm Oct 30 10:53:45 clearly we are the morons Oct 30 10:53:57 russ give it a break i did not call you a moron Oct 30 10:54:21 i simply stated that EVERYONE codes c and you dont need to be a competent ENGINEER to do so Oct 30 10:54:34 'ill be finished in hours what takes you months' Oct 30 10:54:42 go spend 20 years as a realtime embedded control applications CONSULTANT and tell me that 99% of all C developers are not morons Oct 30 10:55:02 yes. not because you are a moron, you made taht mental leap Oct 30 10:55:02 I don't care what other people do, I really don't Oct 30 10:55:10 mark4, do you think that around here we're some kind of teenagers? Oct 30 10:55:11 but because you are developing it in C Oct 30 10:55:23 I'm sure 99% of people that use knives are morons Oct 30 10:55:32 and no, we're not dicking around in forth in trivial problems Oct 30 10:55:35 I'm sure 99% of people that use quite a few things are morons, like cars Oct 30 10:55:56 everybody except me in a car is a moron! Oct 30 10:56:05 av500, you are the 1% Oct 30 10:56:06 everyone but a car is a moron! Oct 30 10:56:27 no. there are 3 kinds of drivers. thers you, theres assholes and theres maniacs. the people driving slower than you are morons, the people driving faster than you are maniacs Oct 30 10:56:41 slower are assholes i mean Oct 30 10:56:47 bah missquote the joke :/ Oct 30 10:56:47 which ones do have the knives now ? Oct 30 10:56:49 I didn't see the arm version on isforth.com my first go around Oct 30 10:56:56 because its not released Oct 30 10:57:04 so its on github then? Oct 30 10:57:04 its not yet fully working Oct 30 10:57:04 blame as Oct 30 10:57:07 no Oct 30 10:57:12 i dont use git Oct 30 10:57:18 cvshub.com? Oct 30 10:57:21 av500, thats part of it Oct 30 10:57:48 svnorious.org? Oct 30 10:57:55 so using git is like opening a car with a knife ? Oct 30 10:57:59 cvsheaven.net Oct 30 10:58:01 no why would i use some external version control i can use hg and keep my entire repository LOCAL. i do not distribute via version control Oct 30 10:58:10 das, like cutting through butter with a plasma torch Oct 30 10:58:12 das, with a canned rock Oct 30 10:58:39 you are certainly behaving like a bunch of morons Oct 30 10:58:42 mark4, yes, certainly, you totally lose control when you push like that Oct 30 10:58:48 I blame you guys for forgetting what it was I wanted to do right now Oct 30 10:58:53 cats and dogs living together Oct 30 10:58:54 but ok go ahead and keep twisting my words to suit your tiny minds Oct 30 10:59:01 autistic brothers branching your code Oct 30 10:59:27 russ no. i just dont want anything to do with github or sourceforge or the like Oct 30 10:59:39 go ahead and branch. Oct 30 10:59:45 mark4, too many trolls? Oct 30 11:00:02 did i say that? Oct 30 11:00:10 I'm just wondering why Oct 30 11:00:14 you have the sources Oct 30 11:00:22 it's very trendy these days to share what you are working on Oct 30 11:00:31 i have no interest in taking the time to "manage" an account on some external site Oct 30 11:00:39 so when you talk about this arm forth stuff, I can take a look Oct 30 11:00:50 if i was interested in trendy why would i develop forth? lol Oct 30 11:01:12 russ you want to see some of my not yet fully debugged arm code? Oct 30 11:01:54 mark4, otherwise, you are just wasting people's time talking about it Oct 30 11:02:09 +1 Oct 30 11:02:15 no im currently out of work so debugging this is my full time job right now Oct 30 11:02:22 mark4: he might even want to help you with it Oct 30 11:02:25 it will be released Oct 30 11:02:36 av500, not me, please don't make me Oct 30 11:02:38 av500 i doubt it Oct 30 11:02:44 baby steps with his tiny mind though Oct 30 11:02:52 av500, but yes, other people interested in forth could share and share alike Oct 30 11:03:07 russ when i have it "working" it will be released Oct 30 11:03:25 anyway, time to board Oct 30 11:03:41 good think the linux kernel isn't waiting until it's "working" to release :) Oct 30 11:03:48 or working to a degree where i would not mind others seeing it Oct 30 11:04:12 you think linus had his entire sources released as he was developing the kernel from day 1? Oct 30 11:04:14 i doubt it Oct 30 11:04:28 i guess its possible it was totally open from day 1 but i doubt it Oct 30 11:04:32 these days he might Oct 30 11:04:44 back then, that was not convenient Oct 30 11:05:12 since git didn't exist yet Oct 30 11:05:14 * XorA wonders if the Sinclair QL version of linux is still about Oct 30 11:07:35 btw russ, your comment asking if "specific x86 opcode" is forth the answer CAN be "yes it is" Oct 30 11:08:04 if you look at swiftx by forth inc their high level definitions lay down very heavilly optimized assembler Oct 30 11:08:40 also iforth by marcel hendrix - isforth however is not optimized because it was intended that anyone could read the sources Oct 30 11:08:53 ive been told by non asm coders that they can read those sources Oct 30 11:09:40 but if you want to see some non trivial forth code look in the extensions directory, the ext/term and ext/memory would be good places to look Oct 30 11:10:00 thats a terminfo parser and text windowing system much like ncurses yet more powerful and easier to use Oct 30 11:10:21 and my someone ok ish memory manager (more like a memory sub allocator than a full manager of course) Oct 30 11:10:38 do you have an H264 decoder too? Oct 30 11:10:41 definately not as "Good" as the one in glib or wherever it is Oct 30 11:10:45 i could have Oct 30 11:10:52 if i had need i would develop it Oct 30 11:11:18 a competent forth coder could create it, why would he not be able to? Oct 30 11:12:39 http://bpaste.net/show/145148/ Oct 30 11:12:52 thats one of the souce files from my arm forth Oct 30 11:13:11 that i believe is mostly working now, maybe fully working but im not calling that yet Oct 30 11:16:37 im sure someone more competent in arm asm could improve that code somewhat Oct 30 11:23:40 anyone knows where I can start whit a BLDC motos ? Oct 30 11:25:13 sorry, I don't know. Oct 30 11:27:58 .. Oct 30 11:28:43 tried google? Oct 30 11:33:10 I´m tried Oct 30 11:33:50 I will try again... Oct 30 12:30:41 https://plus.google.com/111930368766985978471/posts/9QPuicSLWUt Oct 30 12:43:25 hey Oct 30 13:30:45 Can anyone recommend a decent synthetic benchmark tool that cross builds easily? I'm messing around with the cgroup memory resource controller and I want to run some tests in a few different scenarios. Oct 30 13:34:06 hello Oct 30 13:34:25 i have stange problem regarding touchscreen calibration Oct 30 13:34:45 I'm updating the software, it has been 50minutes since i start. but all 4 USRx LEDs don't lit solid. what should I do ? Oct 30 13:35:37 im using 5.7 inch lcd display with Robert nelosns ubuntu 13.04 demo image. on top of which i have installed lxde. Oct 30 13:36:42 whenever i do touch screen calibration using ubuntu xinput_calibrator command it works fine for 3 or 4 reboots Oct 30 13:37:23 then i need to recalibrate again ... thats the problem!! Oct 30 13:42:32 I think I'll try lmbench Oct 30 13:44:34 I noted down values from calibrator utility and tried to give xinput set-int-prop "ti-tsc" "Evdev Axis Calibration" 32 3733 367 79 3992 Oct 30 13:44:42 but that didnt help Oct 30 13:44:58 any idea how can i fix this? Oct 30 14:49:02 Hi, where can I find some info about the naming conventions of BB overlay files? For instance in BB-FOO-01-00A0.dts what does 01 mean? Is it optional? Can it be an arbitrary string? Oct 30 14:50:03 panto: ^^^ Oct 30 14:50:33 ao2, anything up to -00A0 is just a name Oct 30 14:50:45 the 00A0 is the revision Oct 30 14:51:07 note, don't go crazy on punctuation Oct 30 14:57:06 panto, from one of my tests it looks like there is a length limitation in the name: file named BB-GPIO-KEYS-WHEEL-00A0.dtbo, I run "echo BB-GPIO-KEYS-WHEEL:00A0 > /sys/devices/bone_capemgr.8/slots" and in dmesg I get "failed to load firmware 'BB-GPIO-KEYS-WHE-00A0.dtbo'" see the truncated part name? Oct 30 14:57:13 yes Oct 30 14:57:22 it comes from the EEPROM format Oct 30 14:57:45 legacy bogging us down Oct 30 14:57:47 16 bytes should be enough for everybody Oct 30 14:57:49 32 char limits Oct 30 14:57:55 or that Oct 30 14:57:59 I bet its not unicode either Oct 30 14:58:06 no ideograms Oct 30 14:58:40 it is 16 bytes then, OK, I just didn't see this limitation documented in the tutorial I read Oct 30 14:58:46 thanks for the info Oct 30 14:59:26 it is buried somewhere in the TRM of the bone Oct 30 14:59:33 where the eeprom format is documented Oct 30 15:04:47 panto, I am looging at it, and the doc says Board Name is supposed to be 32 bytes, bit in any case what I was missing was the link between the overlay file names and the eeprom format, it's not immediate from a newbe point of view. Now I have some more background, thanks. Oct 30 15:04:56 s/bit/but/ Oct 30 15:05:01 part-number Oct 30 15:05:04 not board name Oct 30 15:06:36 ah, right, like in the same field in the dts file, the "number" in the name confused me, thanks again Oct 30 15:08:34 np Oct 30 15:31:21 hi, i'm having trouble trying to run an executable on my BBB. it just says no such file or directory, though i can cat it just fine and i already gave it the chmod 777 treatment Oct 30 15:31:53 is there some additional file flag or permission bit needed to execute it? Oct 30 15:31:53 run file on it Oct 30 15:32:16 I guess its x86, not arm Oct 30 15:32:22 ./ ? Oct 30 15:32:23 file Oct 30 15:32:24 ELF 32-bit LSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 3.1.1, BuildID[sha1]=0x688fad6cd0db8b779abcc0002d95060d1e219cb8, not stripped Oct 30 15:32:29 palesius, is it in a filsystem with noexec? maybe under /tmp? Oct 30 15:32:36 yeah i did the ./ Oct 30 15:32:44 arm, ok Oct 30 15:33:19 no it is in /home/root and i copied /bin/true there and it ran just fine from the same directory Oct 30 15:33:23 <`Ishq> palesius: ldd file and ensure that all dynamic libraries are present. Oct 30 15:33:53 ldd or ld? Oct 30 15:33:56 <`Ishq> ldd Oct 30 15:34:14 ldd gives command not found. Oct 30 15:34:27 <`Ishq> Ok, did you download this file from the Internet or compile yourself? Oct 30 15:34:54 it seems odd that it would say file not found though. Oct 30 15:35:04 compiled it on a fedora box Oct 30 15:35:07 <`Ishq> Linux does that when it cannot find all the dynamic libs it is linked to Oct 30 15:35:13 <`Ishq> Ok, so not on your own bbb Oct 30 15:35:19 <`Ishq> On the fedora box Oct 30 15:35:23 <`Ishq> Run ldd file Oct 30 15:35:39 <`Ishq> And copy all the dynamic libs that show up and that are not present on your bbb to your bbb Oct 30 15:35:51 <`Ishq> That means that you are missing a library Oct 30 15:36:17 says not a dynamic executable Oct 30 15:36:42 <`Ishq> It says it's dynamically linked on your bbb Oct 30 15:36:57 <`Ishq> Are you sure it's the same file? Oct 30 15:37:11 it uses libusb-1.0 and libpthread Oct 30 15:37:50 let me make sure i'm running the arm ldd, i was probably running x64 Oct 30 15:37:50 <`Ishq> It might also have other libs linked to it based on stuff that libusb needs or other stuff that the compiler may have linked to it for whatever reason Oct 30 15:41:35 it reurns the following on fedora: [root@endymion usb-mitm-master]# /opt/gcc-linaro-arm-linux-gnueabihf/bin/arm-linux-gnueabihf-ldd --root /opt/gcc-linaro-arm-linux-gnueabihf/lib usb-mitm libusb-1.0.so.0 not found libpthread.so.0 not found libc.so.6 not found Oct 30 15:41:59 and i see all 3 in /lib on the BBB Oct 30 15:42:49 a version mismatch maybe? Oct 30 15:43:06 <`Ishq> The same exact filenames are present on the bbb? Oct 30 15:43:17 <`Ishq> If that's the case, it might just be looking in the wrong directory Oct 30 15:43:37 <`Ishq> for instance, looking in /opt/gcc-linaro-arm-etc on your bbb when it should be looking in /lib Oct 30 15:44:15 any way to check that? Oct 30 15:44:24 <`Ishq> ya Oct 30 15:44:25 <`Ishq> run Oct 30 15:44:29 <`Ishq> on your bbb Oct 30 15:44:32 FWIW it did work under debian on the BBB Oct 30 15:44:47 <`Ishq> export LD_LIBRARY_PATH="/lib" Oct 30 15:44:49 <`Ishq> ./file Oct 30 15:45:43 same thing Oct 30 15:46:45 try running it with LD_DEBUG=all ./file Oct 30 15:47:19 expot ld_debug? Oct 30 15:47:26 export i mean? Oct 30 15:48:03 just put it before ./file Oct 30 15:49:02 no change Oct 30 15:49:21 no output at all? Oct 30 15:49:46 just no such file or directory Oct 30 15:50:13 i tried LD_DEBUG=all /bin/true and got many lines Oct 30 15:50:30 that's what i expected Oct 30 15:51:52 is there anything else (like noexec on the fs) that could be stopping it from even attempting to execute it? Oct 30 15:52:25 i scped it over from fedora to bbb, then did a chmod 777 on it Oct 30 15:57:05 no Oct 30 15:57:13 most likely some lib does not match up Oct 30 15:57:27 building on fedora and running on angstrom might have issues Oct 30 15:58:43 but shouldn't i get something from LD_DEBUG if that is the case? Oct 30 16:00:03 ld cannot check if two library versions are binary compatible Oct 30 16:00:48 but shouldn't there be SOME output from LD_DEBUG, rather than just file not found? Oct 30 16:03:17 palesius: well, that would mean that it was looking for a specific lib file and that's not around Oct 30 16:03:28 palesius: make sure the name matches _exactly_ Oct 30 16:03:43 palesius: oh.. and run ldconfig as well Oct 30 16:05:31 the names match exactly including the numbers after .so Oct 30 16:06:40 <`Ishq> Just install ldd and be sure Oct 30 16:08:03 what packageing system does angstrom use? Oct 30 16:08:22 opkg? Oct 30 16:09:55 Use it just like apt-get. Oct 30 16:10:25 It's based on apt-get but modified enough. The same commands for update, install, upgrade, etc... work. Oct 30 16:11:04 what is the package name for ldd? Oct 30 16:11:16 gcc is already installed, and i assumed it would be part of that Oct 30 16:13:38 <`Ishq> palesius: ldd? Oct 30 16:13:42 <`Ishq> http://www.angstrom-distribution.org/repo/?pkgname=ldd Oct 30 16:14:17 i tried that, it said unknown package Oct 30 16:14:47 I am sure stock angstrom ships with ldd, no? Oct 30 16:15:09 i would have though so, but find / | grep ldd give me absolutely nothing Oct 30 16:15:47 <`Ishq> Just download it from the site then Oct 30 16:16:07 ldd is some shell script Oct 30 16:16:20 on my fedora box, it comes from something called glibc-common Oct 30 16:16:23 not sure how it gets packaged by OE Oct 30 16:17:17 i think i'm going to go back to debian. I was having trouble with gadgetfs mounting there, but i had the same issue initially on angstrom, and hopefully the same fix will apply Oct 30 16:52:16 <`Ishq> Is it possible to upgrade the bbb without an SD card? Oct 30 16:58:20 not really Oct 30 16:58:27 anything is possible with the power of wishful thinking Oct 30 16:58:28 doable yes, simple no Oct 30 16:58:32 get a card Oct 30 16:58:34 its cheap Oct 30 16:59:02 <`Ishq> Yeah. I guess I should have foreseen it. I have one in my phone, but I'm reluctant to reformat that. Oct 30 17:04:15 Hi ! Anyone knows if Android runs on the original beagleboard (rev C4) ? If yes, any video hardware acceleration possible ? Oct 30 17:09:48 HW runs SW Oct 30 17:09:56 its all possible Oct 30 17:10:13 now, can you download an image and be happy Oct 30 17:10:15 no Oct 30 17:20:07 <`Ishq> Is ssh enabled on the bbb by default? Oct 30 17:20:16 <`Ishq> I keep getting remote connection closed Oct 30 17:20:27 update it Oct 30 17:23:07 Hi, is there a good doc out there that describes what exactly is happening when the beagelbone black boots? Oct 30 17:23:37 parco: are you connected via ssh or a serial cable? Oct 30 17:23:42 yes sir Oct 30 17:23:49 parco, last chapter in the AM3359 TRM :) Oct 30 17:24:06 well actually, i just ran though the tutorial and im pulling the page over 192.168.7.2 Oct 30 17:24:10 I would just say dmesg and google away. Oct 30 17:24:40 shoragan_: i have that pulled, am i right by saying the AM3359 has firmware preloaded on it? Oct 30 17:24:57 yes Oct 30 17:24:57 parco: login to the system and drop dmesg into a file and go through that line by line. You will see EVERYTHING going on. Oct 30 17:24:59 parco, yes, it has an on-chip rom Oct 30 17:25:00 ROM code Oct 30 17:25:25 whats dmesg? Oct 30 17:25:30 i am not a linux dude Oct 30 17:25:38 that one loads a small u-boot into sram, which configures clocks, pinmux and ddr, and then loads a bigger bootloader Oct 30 17:25:41 it's all the messages at system level in linux Oct 30 17:25:53 parco: there is google for that Oct 30 17:25:56 too Oct 30 17:25:56 that one then goes on to setup the device tree and loads the kernel Oct 30 17:27:22 so the chips ROM code / firmware, configures all the on board pins and interfaces with the main circuitry, but then somehow hands off code execution to a .img? Oct 30 17:28:42 i'm an EE and haven't taken too many OS courses, I'm getting confused after the ROM code executes Oct 30 17:30:00 shoragan_: by "that one", you are referring to the on-chip rom, correct? Oct 30 17:30:56 the first time yes :) Oct 30 17:31:21 rom -> small bootloader in SRAM -> full bootloader in RAM -> kernel Oct 30 17:32:09 ok, and we are capable of changing GPIO configuration at any one of those stages? Oct 30 17:32:46 I'm assuming yes from the cloud9 tutorial Oct 30 17:32:56 more or less yes Oct 30 17:34:10 is it possible to modify the preloaded on-chip rom code? Oct 30 17:35:32 no Oct 30 17:35:36 why would you? Oct 30 17:35:40 it does little Oct 30 17:35:53 besides loading your code and passing control to it Oct 30 17:36:03 then your code can go wild Oct 30 17:37:14 ok, so when i see tutorials saying "flash your program", it is not referring to any modification of the AM3359's ROM Oct 30 17:37:37 indeed Oct 30 17:37:41 that's a little more clear now, thanks! Oct 30 17:37:45 RO = read only Oct 30 17:44:05 Is it written in firmware that when the Boot button is held down during power-up, to flash the contents of sd-card to the eMMC? Oct 30 17:44:49 parco, that behavior is just software running on the SD (which checks the status of that button) Oct 30 17:45:15 so in general, when the boot button is held it will boot from SD? Oct 30 17:45:24 yes Oct 30 17:45:36 that particular program just so happens to then write to the eMMC Oct 30 18:25:18 I'm trying to use the PRU's internal cycle counter so I can get accurate timings in my programs. Here is what I'm doing right now to try to clear and disable the PRU cycle counter: http://pastiebin.com/52714f022b7c4 Oct 30 18:25:35 *clear and enable Oct 30 18:32:40 e Oct 30 18:57:19 Anyone have a camera cape for sale? Oct 30 21:03:46 Hi anybody in here? Oct 30 21:05:47 * KotH watches the tumbleweeds pass by Oct 30 21:05:58 * rocky|arch chases it Oct 30 21:06:26 yeah, it's a quiet day Oct 30 21:06:32 seems so Oct 30 21:07:03 seen a few people ask questions, but no answers. *summary of past 5 hours* Oct 30 21:08:06 but maybe somebody knows if it is possible to invert a shiftout or SPI signal Oct 30 21:08:31 using the hardware itself? Oct 30 21:08:39 MrFaul: invert the data? Oct 30 21:08:51 MrFaul: why not send inverted data directly? Oct 30 21:08:56 ^ Oct 30 21:09:01 nope invert the data don't works Oct 30 21:09:06 why not? Oct 30 21:09:18 what exactly are you trying to do? does the device just have inverted logic levels? Oct 30 21:09:36 because the 1 bit and the 0 bit are timed HIGH levels Oct 30 21:10:07 MrFaul: you can set the clock polarity as well Oct 30 21:10:15 MrFaul: see spi.txt in Documentation somewhere Oct 30 21:10:44 na my hardware is 5V so i soldert a board with a optocoupler but the damn thing inverts the signal Oct 30 21:11:00 I'd just stick a transistor on the other side to invert it. Oct 30 21:11:02 *sigh* Oct 30 21:11:17 MrFaul: standard problem Oct 30 21:11:30 MrFaul: two solutions: invert everything on the software side (easy) Oct 30 21:11:52 MrFaul: or just wire up the optocopler such that it does not invert your signal (needs soldering iron) Oct 30 21:13:23 dude if that were possible i'd have done it by now.. Oct 30 21:13:46 the transistor or inverting software side? Oct 30 21:14:04 MrFaul: all three solutions are possible. trust me on this Oct 30 21:14:08 MrFaul: i'm a professional troll Oct 30 21:14:41 with that said, i'm off to bed Oct 30 21:14:49 n8t Oct 30 21:14:50 fairwell! Oct 30 21:15:04 MrFaul: oh.. ein deutscher! Oct 30 21:15:11 jo Oct 30 21:15:34 *hutlupf* Oct 30 21:15:35 bis morgen Oct 30 21:15:48 ^^ Oct 30 21:20:11 hm maybe writing a own driver would be a soulution does but i don't know whats the min switching time for a GPIO Oct 30 21:20:25 it's pretty fast Oct 30 21:20:43 ns please... Oct 30 21:20:52 well Oct 30 21:21:15 let me measure it. I'm actually working on that kind of stuff right now :p software PWM on the PRU Oct 30 21:21:34 let me guess it depends on the CPU load Oct 30 21:21:45 Not on PRUs Oct 30 21:22:06 I'll test it on not PRU then :/ Oct 30 21:22:12 happy? Oct 30 21:22:40 thx ^^ but both would be informative Oct 30 21:23:01 I'll do pru first then because it's more accessible to me right now. Oct 30 21:23:27 also, I don't think the number will change much because they both go through the same subsystem Oct 30 21:24:19 having a bit of trouble giving my bone an ip address. I'm just gonna restart and brb Oct 30 21:24:58 hm writing a own driver would also allow me to run the stuff on max fps Oct 30 21:26:24 back Oct 30 21:26:40 ^^ Oct 30 21:27:13 I've written some bitbanging i2c code for the pru Oct 30 21:27:14 already Oct 30 21:27:33 and darn it dhcpcd! GIVE MY BEAGLE AN IP! ~:| Oct 30 21:27:52 xD Oct 30 21:28:27 ok, had to replug the bone. now things are happy ^^ Oct 30 21:29:50 well i'm trying to drive my new lamp with addressable LED pixel https://www.dropbox.com/s/m4s707awfonn8k5/20131029_201153.jpg Oct 30 21:30:05 is this one of those led strips? Oct 30 21:30:10 jop Oct 30 21:30:21 is that a yes? Oct 30 21:30:28 yes^^ Oct 30 21:30:52 actually it's around 6m Oct 30 21:31:07 i was bored... Oct 30 21:31:13 one of my profs has a reel of it Oct 30 21:31:38 blah Oct 30 21:31:47 stupid prus Oct 30 21:32:39 They seam to stop responding if you don't let the program terminate. Oct 30 21:32:48 Haven't found a fix for it yet. Oct 30 21:33:06 -> plug Oct 30 21:33:15 besides that >.< Oct 30 21:33:23 ^^ Oct 30 21:34:39 I'm gonna see if I can send you the scope picture. Oct 30 21:34:48 so you can get an idea of the rise/fall time going on here Oct 30 21:35:27 yeah, I don't have a usb drive on me. I'll just send a picture Oct 30 21:36:28 ah someone has the right equipment to measure the time^^ Oct 30 21:36:45 yep, at my college's signals lab Oct 30 21:37:44 ok where are you? since that sounds like daytime Oct 30 21:39:15 Rose Hulman Oct 30 21:39:23 It's 5:40 pm here Oct 30 21:40:21 nope that doesn't ring anything i'm in Wuppertal btw. and it's 22:40 Oct 30 21:40:48 where's that? Oct 30 21:41:00 https://www.dropbox.com/s/8spqog4peui3lmp/IMG_20131030_173819.jpg Oct 30 21:41:17 it's 50ns/div, 1v/div Oct 30 21:41:52 germany 51.284682, 7.201511 Oct 30 21:42:15 USA, Indiana Oct 30 21:43:40 oh, the picture is missing the frequency! Oct 30 21:44:13 you can deduce it though. Oct 30 21:44:30 well it says 50ns meas it one square? Oct 30 21:44:43 around 4.4 MHz Oct 30 21:44:47 yeah Oct 30 21:45:56 ... ah damn i'm an idiot looking into the wrong sheet xD hang on a sec Oct 30 21:47:10 not sure if the rise/fall time could be better, can't really factor out the impedance of my probe. Oct 30 21:47:48 it says it's around 15 pf for 10 MHz Oct 30 21:48:50 do you still care about how fast it can go on the ARM? Oct 30 21:49:01 it'll be basically the same limit if you do mmap'd gpio Oct 30 21:51:21 nope that fast enough Oct 30 21:51:58 I'm not sure if you could get up to those speeds with the device tree. Then again, I wouldn't think it'd matter beyond a certain point. Oct 30 21:52:24 I've gotta go, bbl Oct 30 21:52:26 i think i write a direct driver for the WS2811 maybe i release it Oct 30 21:52:39 that'd be cool Oct 30 21:52:42 thx was helpful Oct 30 21:52:47 np Oct 30 21:58:46 anyone have experience with 4 user led's blinking at the sametime well trying to flash the emmc? Oct 30 22:22:34 anyone interested in yet another parallel port for beaglebone? Oct 30 22:26:27 eh? what sort of parallel port? Oct 30 22:31:48 memory mapped x86 compatible 5v Oct 30 22:36:00 http://uploads.oshpark.com/uploads/project/top_image/qfaEYUJo/i.png Oct 30 22:36:11 2 ports via IDC26 ribbon Oct 30 23:06:28 Hello People. Oct 30 23:07:51 What are the common problems preventing people from connecting via the debug cable? Oct 30 23:08:48 I have the right COM port, speed at 115200, data and stop at 8 and 1, no parity and no flow control. Oct 30 23:09:08 The console opens, putty thinks it is connected but there is no traffic and not response. Oct 30 23:10:35 I have problems with putty connecting via the USB serial port, I had it running but then it would no longer establish a connection. Oct 30 23:10:56 I have never had it work. Oct 30 23:11:10 I can connected to the web server and all that, just no serial debug. Oct 30 23:11:28 s/connected/connect Oct 30 23:14:16 what OS is on the workstation? Oct 30 23:19:51 N2TOH: Win8 Oct 30 23:19:59 It fails from OSX as well Oct 30 23:20:10 I have not tried it with the Linux box. Oct 30 23:21:46 I just tried mine again and putty claims there is a fatal error when trying to open the serial port Oct 30 23:21:59 the gimic drive also no longer works Oct 30 23:28:17 VirG: is this on a beaglebone black? Oct 30 23:28:40 VirG: i almost exclusively connect to both the -xm and black via serial (from linux) Oct 30 23:29:39 common problems on windows, in my experience anyways, was choosing the wrong com port (windows) or not configuring putty correctly Oct 30 23:30:29 what are the correct settings for putty? Oct 30 23:31:01 115200 8n1 Oct 30 23:31:26 unless they changed it up on the bbb Oct 30 23:31:47 thanks. just trying to unbrick my bbb, what a headache! heh Oct 30 23:32:10 anyone know the limit of re-flashes one can do to the emmc? Oct 30 23:32:14 I confirm 115200 8-n-1, i'm using it right meow Oct 30 23:33:12 any flow control? Oct 30 23:33:23 flow control is the n Oct 30 23:33:27 no Oct 30 23:33:46 k, thought that was for parity Oct 30 23:33:51 that's none to right? Oct 30 23:34:07 hmm, maybe, but yes they're both off Oct 30 23:34:12 k, thanks Oct 30 23:55:19 Hmmm Oct 30 23:57:04 grrr... almost ready to throw this pos in the can and buy something else! Oct 30 23:57:19 packet over sonet? Oct 30 23:57:22 Sorry, dealing with several things Oct 30 23:57:37 flow control is currently off as is parity Oct 30 23:57:49 Changing (again) to see what happens. Oct 30 23:58:14 what specifically it the problem... have been working and focusing on other things Oct 30 23:58:38 I am unable to establish a serial connection to the BBB. Oct 30 23:58:58 COM4, 115200, 8, 1, No Parity, No Flow Control. Oct 30 23:59:07 Using Putty on Win8 Oct 30 23:59:55 what serial cable are you using? is win8 recognizing it properly? Oct 30 23:59:56 poor win 8 is cracking at the seems already? Oct 31 00:00:00 :D Oct 31 00:00:07 indeed. Oct 31 00:00:16 although I think it is likely to be Operator Error Oct 31 00:00:19 anyone know the default user/pass for an emmc angstrom flasher image? I have 4 blinking leds and want to get in a see what dmesg says Oct 31 00:02:33 have you tried root/blank Oct 31 00:03:04 yeh Oct 31 00:04:53 I thought the LEDs blink during the flashing Oct 31 00:05:11 oh they do, but not like this... they blinking in sync Oct 31 00:05:37 do you have a display hooked up? Oct 31 00:05:44 yeah Oct 31 00:05:47 or just serial console? Oct 31 00:05:51 blank with the emmc Oct 31 00:06:05 blank with the emmc flasher image that is. Oct 31 00:06:21 got ubuntu to somewhat boot, but just goes blank after a few seconds. Oct 31 00:07:11 someone claims I have tried to reflash the emmc to many times... but I have re-flashed under 10 times, so it can't be that. Oct 31 00:08:28 gonna just give the 4gb non-emmc flasher image a try and strickly boot off the sd card. Oct 31 00:08:59 have a feeling the partition table is messed on the emmc and the flasher's are having dificulty figuring it out. just my thought's tho. Oct 31 00:10:47 anyone have any thoughts on my serial connection fail? Oct 31 00:11:26 jkridner: ping Oct 31 00:11:44 VirG does the serial port you're using disappear when you unplug the cable from the host machine? Oct 31 00:11:51 Yes Oct 31 00:12:29 I was a little surprised to see the name as "Gadget Serial", though. Oct 31 00:12:49 do you have a virtual machine you could test on? Oct 31 00:13:01 I can SSH to the BBBs IP so I know it is operational. Oct 31 00:13:15 I can make a VM, but what would I be checking? Oct 31 00:13:57 just interested if the device functions in linux Oct 31 00:14:08 that does occur sometimes Oct 31 00:14:09 VirG: what distro you running on the bbb? Oct 31 00:14:29 sheldon: stock angstrom. Having issues with FreeBSD. Oct 31 00:15:27 Thurgood: you're thinking that I use a linux vm to try to connect, is that right? Oct 31 00:15:45 yeah Oct 31 00:16:14 OK. I might try it from the linux box in a bit. However, I am told that this should work from Win8. It has failed from OSX too. Oct 31 00:16:16 in case it's a driver issue with win8 Oct 31 00:16:18 you have tty04 set correctly? dmesg | grep tty04 - anything show up? Oct 31 00:16:46 or check /dev for tty04 Oct 31 00:17:30 sheldon: I'm on the Win8 box right now. Oct 31 00:17:36 on the OSX box, no. Oct 31 00:17:44 err, nothing is displayed Oct 31 00:18:06 no no, on your BBB itself Oct 31 00:18:11 No tty04 on the OSX machine Oct 31 00:18:12 you said you can ssh to it. Oct 31 00:18:12 oh Oct 31 00:18:13 heh Oct 31 00:18:17 lemme check Oct 31 00:18:19 ;) Oct 31 00:19:01 nothing is displayed on the BBB Oct 31 00:19:27 and it is not listed in tty04 Oct 31 00:19:30 http://hipstercircuits.com/enable-serialuarttty-on-beaglebone-black/ Oct 31 00:19:34 give that a go. Oct 31 00:19:36 what do I have to do? I will check online. Oct 31 00:19:40 ahh. ok. Thank you. Oct 31 00:20:09 np Oct 31 00:20:46 stty -F /dev/ttyO4 raw Oct 31 00:20:47 | stty -F /dev/ttyO4 9600 Oct 31 00:20:59 should create a basic stty for yeah Oct 31 00:21:09 can change the baudrate I think to 115200 tho Oct 31 00:21:22 no such file or directory. Oct 31 00:21:26 Reading the rest of the article. Oct 31 00:22:04 your sudo'ing, or in root? Oct 31 00:22:09 root Oct 31 00:22:14 hrm, odd. Oct 31 00:38:02 the process at the enable serial / uart/ etc page fails on echo enable-uart5 > /sys/devices/bone_capmgr.*/slots I tried with the dts extension as well. Oct 31 00:38:10 any suggestions? Oct 31 00:39:32 I am going to update the firmware to the most recent. Oct 31 00:41:17 hmm. that's from Jun 20. I think I already have it. Oct 31 00:51:31 VirG: when using a serial uart, the device is like /dev/ttyO4, where that's a capital O, not a zero Oct 31 00:52:45 VirG: also, if you're trying to make a serial connection from your pc to the beaglebone black, you have to use the little row of pins, not gpio pins Oct 31 00:53:11 I use this for watching the boot process, for instance Oct 31 00:53:20 he's trying directly over usb cable, no? Oct 31 00:53:44 so won't see boot, just get a stty afterwards. Oct 31 00:53:47 sheldonw: beats me, usb cable only gets you a console after the system boots Oct 31 00:53:53 aye Oct 31 00:54:12 i dunno what the usb connection shows up like for him, on my system it presents as a network interface Oct 31 00:54:25 the ubuntu distro i'm using stands up a busybox dhcp server and everything Oct 31 00:56:22 wish mine would just even boot! heh. could be more of help here. Oct 31 00:57:08 well what is he ultimately trying to do? communicate w/another device via serial (like an arduino) or just get a console connection for debugging? Oct 31 00:57:22 sheldonw: have you tried booting from sd? i don't push crap down to my emmc yet, not smrt enough Oct 31 01:00:01 yeah, tried just about everything, just comes down to 4 sync'd flashing led's. With angstrom, can't login, root account isn't accepting blank passwd. Ubuntu let's me login with the ubuntu account, but I'm stuck with read-only bunked system. dmesg gives me ifile size errors. Oct 31 01:00:53 sheldonw: i'm using ubuntu 13.04 from an sdcard with no issues Oct 31 01:01:13 sheldonw: so when you boot ubuntu from the sdcard, you can log in and look around, but nothing works? Oct 31 01:01:50 yeah, something is crapping out during boot and it's not fully booting, going into single user mode with a read-only rootfs. Oct 31 01:02:20 i'd reimage a different card and see if that improves life Oct 31 01:02:35 been there, done that. :P Oct 31 01:02:57 sorry friend, i'm not sure what else to recommend. have you tried the card on another beagle? this could rule out hardware. Oct 31 01:03:34 no, don't feel like buying another one, just tempted to scrap beagle and go pi at this point. Oct 31 01:04:03 i have a few pis as well and they're pretty sweet Oct 31 01:04:17 i prefer the 2gig emmc even though i'm not currently using it, but know its there makes me happy Oct 31 01:04:35 yeah, might hold out and get a cubietruck. the emmc is what origianlly sold me on the beagle. Oct 31 01:04:50 cubietruck's are going to sport 2gb ram, would be much nicer! Oct 31 01:04:52 what are you guys setting the swappiness setting to? Oct 31 01:05:19 N2TOH: i haven't changed my swappiness Oct 31 01:06:23 sheldonw: 2gb would be nice but the cubietruck isn't cheap Oct 31 01:06:43 i can buy 2 beagle blacks for the price of a single cubietruck Oct 31 01:06:56 sheldonw: imagine! you could have two devices not booting! Oct 31 01:08:02 haha... yeah. I imagine a werkin bbb! have a big project that this was intended for. Oct 31 01:09:01 sheldonw: did you modify to partition structure on the emmc? Oct 31 01:10:11 sheldonw: i'm out, best of luck! Oct 31 01:10:17 peace s5fs Oct 31 01:10:29 thurgood: I must have. Oct 31 01:11:41 you reading my post's on beagleboard.org? :P Oct 31 01:11:53 no Oct 31 01:12:00 smart man than. heh Oct 31 01:12:15 just trying to think what could have done this Oct 31 01:12:25 your pretty much right I think. Oct 31 01:12:46 just can't figure out how to fix the partition table when I'm stuck in a read-only system. Oct 31 01:13:14 boot from an SD? Oct 31 01:13:28 s5fs: hahahahahaha. That's it. OK. thanx. Oct 31 01:13:51 The 0 vs O Oct 31 01:14:01 that's the only way I can get to this read-only fs... with ubuntu as it's the only thing I can atcually log in with. Oct 31 01:14:09 from a sd card that is. Oct 31 01:14:19 shit don't boot at all from emmc. Oct 31 01:14:45 hoping robert c. n. can figure it out. I'm stumped. Oct 31 01:15:45 sheldonw, is there a card in the SD slot? Oct 31 01:15:51 hmm, what does it show for the contents? have you tryed unmounting it and running fsck? Oct 31 01:15:57 N2TOH, yes Oct 31 01:16:12 unmounting the emmc? Oct 31 01:16:18 that is why it will not boot from the emmc Oct 31 01:16:18 it's not mounted at all. Oct 31 01:16:29 trust me remoce the card Oct 31 01:16:32 yes, you wsiad ti was read only Oct 31 01:16:40 N2TOH, with or without a sd card it won't work. Oct 31 01:16:42 said even Oct 31 01:16:49 oh Oct 31 01:16:59 or not. Oct 31 01:17:01 :\ Oct 31 01:17:06 :/ Oct 31 01:17:08 JTAG? Oct 31 01:17:11 or is it the SD that is mounting read only? Oct 31 01:17:32 the rootfs of the sd filesystem is read-only... Oct 31 01:18:02 ah, then try running fsck on the SD card itself in a host computer Oct 31 01:18:32 and make sure the write protect switch is not enabled Oct 31 01:18:40 I've got multiple sd card's all running diff. distros, all giving me the same headache. Oct 31 01:19:32 is there a physical write protect switch on the BBB? Oct 31 01:19:47 on the SD Oct 31 01:20:07 s5fs: I'm just trying for a serial connection so I can see the boot process. Oct 31 01:20:36 on SD cards there is a write protect switch, but not on micro's. if so, it is so damn small I can't find them! Oct 31 01:20:37 VirG: you're using the 5pin ftdi cable right? Oct 31 01:20:51 I can't write fine to them from other os's on pc. Oct 31 01:21:04 so you have 2 problems? sd is read only and emmc just won't boot? Oct 31 01:21:08 ah... perhaps it's the SD socket Oct 31 01:21:14 thurgood: no, the 4 pin adafruit cable Oct 31 01:21:20 do you have any capes ? Oct 31 01:21:22 VirG, you will need an actual serial jtag based setup, usb can't do that with BBB. Oct 31 01:21:44 VirG, oh, didn't know you had one of those. Oct 31 01:22:08 no capes, vanilla setup. Oct 31 01:22:18 2A supply of power Oct 31 01:22:46 this one: http://circuitco.com/support/index.php?title=BeagleBone_Black_Accessories#Adafruit_4_Pin_Cable_.28PL2303.29 Oct 31 01:22:58 thurgood: I do, but they're in the box. Oct 31 01:23:07 sheldon: unF. OK. Oct 31 01:23:07 thurgood, think you could private msg me your partition table of your emmc? Oct 31 01:23:39 just so I know what it "should" be. Oct 31 01:25:35 VirG: ah, and you've got the pins wired correctly? Oct 31 01:26:04 jonpry, precisely. Oct 31 01:27:07 I've tried to remount the rootfs rw, but no go either. Have done that many times on a pc in past when issues like this arise. Oct 31 01:27:27 I believe so. Black to 1, Green on Receive, White on Transmit: https://lh4.googleusercontent.com/-3oca40vrH-g/Ubd5TjD-0_I/AAAAAAAAALU/-GibZBaTvKk/s1024/beaglebone-black-serial.jpg Oct 31 01:27:53 I've tried reversing the receive and transmit thinking maybe translation was occurring. Oct 31 01:28:07 why I can't boot off of an sd card is beyond me, even if my partition talbe of the emmc is screwed. Oct 31 01:31:42 VirG: and you've installed the drivers, or are you using auto-installed drivers? Oct 31 01:36:48 I installed the drivers from the getting started page, BONE_D64.exe. This was to network-over-usb. I did *not* install the FTDI USB to Serial drivers from that page. I did install PL2303 USB to Serial Driver which did not help Oct 31 01:37:46 I did not think I needed the FTDI driver as I get a COM port when I plug it in. Oct 31 01:39:11 http://www.adafruit.com/products/954#Downloads just because it shows up doesn't necessarily mean it's functonal Oct 31 01:39:21 yeah. :) Oct 31 01:39:48 I have three so maybe I should try one of the other cables. Oct 31 01:39:59 I'll try after I install the FTDI drivers. Oct 31 01:40:30 but if I don't have a /dev/ttyO4, then I don't see this happening. Oct 31 01:40:51 unless I misunderstand this Oct 31 01:41:55 I'm pretty sure you'll need th FTDI drivers to work properly. Otherwise your just trying to use typical serial communications. Oct 31 01:42:20 there's a difference, speically if you have a cable with a FTDI chip embedded in it. Oct 31 01:43:14 OK. No effect. Oct 31 01:43:22 Putty still shows a pretty green cursor. Oct 31 01:43:22 hrm Oct 31 01:43:35 heh Oct 31 01:43:51 Interestingly enough, a new unknown device shows up "USB-Serial Controller" Oct 31 01:44:08 ikes Oct 31 01:44:20 I'm going to do the windows thing and restart. Oct 31 01:44:23 sudo dmesg | grep tty Oct 31 01:44:36 OK. Oct 31 01:45:06 on my Ubuntu box it comes up as, /dev/ttyACM0 Oct 31 01:45:15 well it did Oct 31 01:45:24 I'm not sure what happened Oct 31 01:45:47 maybe I shoule check the solder joints of the USB connector Oct 31 01:45:50 dmesg returns nothing Oct 31 01:45:56 noting? Oct 31 01:46:03 nothing* Oct 31 01:46:31 nada Oct 31 01:46:36 hmm Oct 31 01:46:40 can someone send me there partition table output please? Oct 31 01:46:44 ls /dev/tty gives me a lot. /dev/tty4 being one Oct 31 01:46:46 of the emmc that is. Oct 31 01:47:09 sheldon: is there a way for me to get that if I ssh to the BBB? Oct 31 01:47:32 ttyO0, Oct 31 01:47:55 yeah, fdisk /dev/ then hit "p" and copy and paste what comes up. Oct 31 01:48:28 just don't fuck up and hit "d" or "w" or any weirdness. don't want you to end up bunked like mine! :P Oct 31 01:49:12 not sure what your mmcblk for emmc will be, mine was random based on wether I had a sd card in or not. Oct 31 01:49:29 fdisk? heh. Oct 31 01:49:40 that gives me an unexpected token 'newline' Oct 31 01:49:50 ahh. ok. Oct 31 01:50:30 either will be mmcblk0 or mmcblk1 Oct 31 01:51:03 using fdisk /dev/mmcblk0 I get: Oct 31 01:51:15 if your unfamiliar with fdisk, don't worry about helping me out mang. Oct 31 01:52:03 VirG: that cable specifically says it is not ftdi Oct 31 01:52:16 I have results, but I forgot that putty and copy are not friends. Oct 31 01:52:23 shedldon... don't have access to bbb at the moment, sorry Oct 31 01:52:32 thurgood, no worries Oct 31 01:52:53 VirG, just select the text it copy's it to clipboard automatically. Oct 31 01:53:07 right click pastes in putty. Oct 31 01:53:09 Check prives Oct 31 01:53:13 -e Oct 31 01:53:40 you can you pastebin rather than private msg if you prefer Oct 31 01:54:03 that was mmcblk0, I also have mmcblk0boot0 and mmcblk0boot1 Oct 31 01:54:17 thurgood: yeah. I was just being lazy Oct 31 01:54:22 yeah thanks, I'll start with that. Oct 31 01:54:48 there is also mmcblk0p1 and p2. huh. I guess one is swap? Oct 31 01:55:06 no, unless you made a swap, there isn't one automatically. Oct 31 01:55:17 p1 is the fat, p2 is the ext. Oct 31 01:55:23 one should be a small boot area Oct 31 01:55:33 ahh yeah, I see that now Oct 31 01:56:16 well jesus... not sure what the deal here is no, but about 200 attempts later it's FLASHING my emmc!!!!!! heh Oct 31 01:56:33 I'm going to try a restart and see if that has an effect with the FTDI drivers Oct 31 01:56:35 sorry for the spam ppl! :P Oct 31 01:56:43 you hit the majik number Oct 31 01:57:06 420 woulda been a better number! :P Oct 31 02:03:52 Jeez, that gerald guy on beagleboard.org is a moron... Hi if your reading this Mr. G. Oct 31 02:04:14 and thanks to all on here that attempted to help. My bbb just flashed successfully. Oct 31 02:10:02 gerald is pretty sharp, generally Oct 31 02:11:01 moron isn't the word... rude is more appropiate, the last msg he sent me was insulting. Oct 31 02:11:41 something that one in his position should never be, just alittle to arogant for my liking. Oct 31 02:14:37 hehe Oct 31 02:14:41 that's pretty funny Oct 31 02:20:20 yeah, he's funny... power trippin mr. moderator. Oct 31 02:21:20 no matter how smart he may be, I woulda fired his ass long ago. heh. Oct 31 02:21:24 okay, I'm done. Oct 31 02:21:37 happy my bbb is booting properly now. Oct 31 02:26:19 VirG, any luck? **** ENDING LOGGING AT Thu Oct 31 02:59:58 2013