**** BEGIN LOGGING AT Sat Dec 01 02:59:58 2012 Dec 01 03:04:46 mru, what do you use your beagles for? you mentioned that you have several. Dec 01 03:05:25 only about 10 Dec 01 03:07:10 is they for career work, or a hobby? Dec 01 03:07:14 *are Dec 01 03:07:25 both Dec 01 03:07:57 some are running automated tests, some are for hacking, some just lying around, a couple dead Dec 01 03:08:19 make a beaglewulf cluster Dec 01 03:08:25 are you a test engineer? Dec 01 03:08:27 lol Dec 01 03:08:44 strata: I made the beaglewall Dec 01 03:10:05 I'm trying to synch my beagleboard to GPS time through the serial port Dec 01 03:11:42 nobdoor: from my experience, changing boot.scr/uEnv.txt to simply get rid of the console=/dev/ttyXn sometimes doesn't work. your best bet is to change it from ttyS2 to ttyS1, then disable the getty. Dec 01 03:12:23 one of the kernels i ran had a default commandline with console=/dev/ttyO0 in it. Dec 01 03:12:53 default command lines only kick in if you don't provide one at all Dec 01 03:13:06 (or they override whatever you provide completely) Dec 01 03:13:11 intersting. are these commands a part of a script, or do I need to build a new kernel? Dec 01 03:13:16 mru: right. and in my case, after i got rid of console=/dev/ttyO0, there was no commandline. Dec 01 03:13:34 so if the kernel is using _something_ from the bootloader-provided command line, it's using exactly that command line Dec 01 03:13:50 strata: oh, no root= even? Dec 01 03:13:55 nope. Dec 01 03:14:05 root=/dev/mmcblk0p2 was already in there Dec 01 03:14:17 in where? Dec 01 03:14:29 I thought the only compiled in default root was fd0 Dec 01 03:14:50 i was playing around with some whack distro, can't recall what it was. was when i first got my beaglebone Dec 01 03:15:18 note to self: don't blindly install distros from random blog posts Dec 01 03:15:33 heh Dec 01 03:16:20 mru: since you have so many beagles, perhaps you might know an answer to this: i was doing some serial comm between a zipit Z2 and my beaglebone a few days ago. Dec 01 03:16:45 i had to configure the OMAP mux option to configure the pins to be a uart. Dec 01 03:17:03 strangely after that, my usb hub stopped working (until i rebooted) Dec 01 03:17:09 err. usb host. Dec 01 03:17:18 no voltage on the pins at all Dec 01 03:17:29 does setting certain mux options turn it off? Dec 01 03:28:51 well since all else is silent I'll throw another question into the great abyss Dec 01 03:29:46 where can I find the uenv/uboot files on the XM? seems this question is too simple for google but too complicated for yahoo Dec 01 03:29:55 strata: thank you Dec 01 03:30:01 mru: thank you Dec 01 03:30:13 i didn't expect to find help here like i did Dec 01 04:08:50 * Russ googles 'how do I google something' Dec 01 04:21:21 http://lmgtfy.com/?q=lmgtfy Dec 01 04:25:11 emeb_mac, love it... Dec 01 04:25:31 :) Dec 01 04:25:44 how's things ka6sox-away ? Dec 01 04:29:36 emeb_mac, better...bbiaf...taking kidlet to LAN Party Dec 01 04:29:50 whee! Dec 01 04:47:48 so my beaglebone was going REALLY slow, even though I have nearly all the stock services disabled and removed (cloud9, node, gateone). I tried rebooting it just now, and once it rebooted, it stayed up for about 10 seconds, then i got "System HALTED". This happens every time now. Dec 01 04:48:20 I had to fix this once before by just formatting the sd card, but i really would like to figure out wtf is going on Dec 01 04:49:10 anyone run into this? Dec 01 05:04:01 well, I suppose it's time for my usual peak season for a furnace of blazing Dec 01 05:04:42 isn't that called stepping outside in PHX? Dec 01 05:06:02 it's from a machine translated programming manual Dec 01 05:06:33 it also has a 'counterfeit a furnace of blazing' Dec 01 05:27:23 I see Dec 01 09:04:21 <_av500_> sakoman: check your G+ :) Dec 01 09:25:42 _av500_: fors? Dec 01 09:27:50 christ the pru repo is full of ^M's Dec 01 09:30:11 ahem jkridner_ :) Dec 01 10:35:26 mranostay: check https://github.com/modmaker/am335x_pru_package Dec 01 10:35:37 mranostay: that has the s3kr1t instructions Dec 01 13:34:05 new to beaglebone, was wondering why most docs seem to prefix python script execution with 'python', i.e., python bbpwm.py Dec 01 13:43:10 some people run python scripts by explicitly calling the interpreter.. just a style thing Dec 01 13:44:42 line 1 #!/usr/bin/python a dark secret? Dec 01 13:44:56 yes Dec 01 13:45:32 windows ppl :) Dec 01 13:45:56 did you have a question or did you come here just to feel smug? Dec 01 13:46:03 you may have to do it that way on a fat partition as well Dec 01 13:47:15 there's always a couple of exceptions, explicitly calling the interpretter should always work if it's installed Dec 01 13:47:37 don't understand the fat partition thing, can you expain Dec 01 13:47:46 permissions Dec 01 13:47:55 ah Dec 01 13:48:04 can't change them? Dec 01 13:48:37 you can change the ones that exist Dec 01 13:51:51 hope i didn't offend the attitude police Dec 01 13:53:02 Oat: when you're new somewhere, it's usually a good idea not to start by telling everybody they're wrong Dec 01 13:53:52 actually I may have been wrong about even being able to change the flags on a fat partition under linux :P Dec 01 13:53:54 was simple speculation on why unnecessary keystrokes commonly appear in docs Dec 01 13:54:44 clarity maybe? Dec 01 13:55:40 probably so, for those who may not be familiar with the line one feature of unix scripts Dec 01 13:56:08 but probably the more likely to work also could be the reason Dec 01 13:56:21 Oat: we're quite well aware of that feature Dec 01 13:56:29 please take your attitude elsewhere Dec 01 13:56:37 no guarantee everyone is Dec 01 13:57:50 mru it's not necessarily attitude... at least someone's chattting Dec 01 14:11:21 Oat: for python they usually even do #!/usr/bin/env python (http://www.velocityreviews.com/forums/t324727-what-does-usr-bin-env-python-do.html) Dec 01 14:11:57 as if env is always in /usr/bin Dec 01 14:13:48 not even /bin/sh is required to exist Dec 01 14:16:44 I guess les often :P http://en.wikipedia.org/wiki/Env also mentions that problem (but I have never encountered it) Dec 01 14:17:21 but with the little python I know it's not weird. (setup.py and such do some more hardcoding) Dec 01 14:18:57 I like that they state "more portable" in the wikipedia entry. Dec 01 14:20:31 I wouldn't expect env to be anywhere else on a linux system Dec 01 14:20:34 wasn't aware of env. Similar function to 'which" i guess? Dec 01 14:20:48 http://pubs.opengroup.org/onlinepubs/9699919799/utilities/env.html Dec 01 14:22:33 does busybox put env in /usr/bin by default Dec 01 14:22:45 good to know if you transport your scripts to various unix platforms Dec 01 14:24:07 I can't name a system that has it elsewhere, but that doesn't mean they don't exist Dec 01 14:25:41 always have to change line 1 of perl scripts when i take them to Sun Dec 01 14:26:54 mru: btw I used your clock.c code to do some testing. Dec 01 14:27:48 I was very supprised (and I still don't really understand) that I needed to lower the amount of instructions run to prevent the CCNT clock from wrapping Dec 01 14:28:44 at one point do you a spin = 1<<30; and run the loop 1<<30 time (that takes 2 asm instructions) Dec 01 14:29:55 so it should take roughly 1<<31 cycles Dec 01 14:30:08 or less Dec 01 14:30:09 the amount of instructions should not be > 1<<32 , so why would ccnt wrap? Dec 01 14:30:31 what makes you think it wraps? Dec 01 14:31:14 when I lower that value (bellow 25) I get stable results of 600 MHz Dec 01 14:31:33 afther that is's bingo time Dec 01 14:31:35 it is Dec 01 14:32:34 running many other things? Dec 01 14:32:36 well .there is of course also the interrupt handling and other stuff happening at the same time Dec 01 14:33:17 just that code and and interrupt handler(1000 Hz) Dec 01 14:33:34 that's a high tick rate Dec 01 14:34:05 I'm getting consistently 3.2G cycles on A8 Dec 01 14:34:09 should I move more to the 60/100 Hz range? Dec 01 14:34:47 I was thinking the 1ms timer was designed just for that Dec 01 14:35:09 of course you can use a 1ms timer if you want Dec 01 14:35:21 there's just usually no reson to do so Dec 01 14:36:10 doesn't linux run a that rate (and keeps track of jiffies like that?) Dec 01 14:36:59 or is the port "tickless" now :P Dec 01 14:37:08 linux is usually run tickless Dec 01 14:37:14 which isn't quite as tickless as it seems Dec 01 14:37:25 there are times when a periodic interrupt is still running Dec 01 14:37:34 the rate of that is configurable on most systems Dec 01 14:38:03 100, 250, 300, and 1000 Hz are the usually available choices Dec 01 14:39:24 you're using a counter at full rate? no division? Dec 01 14:39:31 yes Dec 01 14:39:41 really need that kind of resolution for something? Dec 01 14:40:04 so the problems I was having with clock are probably related to the amout of interrupts being handled Dec 01 14:40:17 Oat: it runs at that rate Dec 01 14:40:32 have played with another arm sbc in C (been a while) Dec 01 14:40:43 and seems i remember you can divide it down Dec 01 14:41:00 i usually divided it down to microsecond Dec 01 14:41:03 you can choose full rate or /64 iirc Dec 01 14:41:16 this is the cpu clock we're talking about Dec 01 14:41:24 oic Dec 01 14:41:38 if you're counting cycles, you usually want them all Dec 01 14:42:02 i was playing with servos, microseconds were handy for setting pulse widths Dec 01 14:42:23 we were talking about measuring the cpu clock rate Dec 01 14:42:34 sorry, misunderstood Dec 01 14:43:14 keesj: btw, a9 runs that loop in ~1G cycles Dec 01 14:44:05 keesj: btw, offtopic minix question Dec 01 14:44:25 I'm running minix3 under vmware and it's not responding to arp requests for some reason Dec 01 14:45:07 am I missing some trick? Dec 01 14:45:26 does ifconfig show correct configuration Dec 01 14:45:42 I said it's not responding to arp Dec 01 14:45:48 everything else is exactly as it should Dec 01 14:45:55 oh arp only Dec 01 14:45:57 interesting Dec 01 14:46:32 how can others talk to it if it's not responding to arp Dec 01 14:46:33 mru: I don't know I will ask for you. Dec 01 14:46:48 Oat: if it talks first, it works Dec 01 14:47:10 so others cache its mac from traffic Dec 01 14:47:15 keesj: vmware esxi with a virtual e1000 nic fwiw Dec 01 14:49:01 it's no big deal, just mildly annoying Dec 01 14:49:28 no idea about that vm stuff Dec 01 14:49:51 i haven't used minix since linux first came out on floppies Dec 01 14:49:54 how useful is it Dec 01 14:50:17 for me, not particularly Dec 01 14:50:25 how complete is it Dec 01 14:50:33 in what regard? Dec 01 14:50:49 it has most of the netbsd userland Dec 01 14:51:00 any conspicuous omissions from typical distributions? Dec 01 14:51:10 mru: you did some porting of libav to minix right? Dec 01 14:51:20 keesj: not much porting required Dec 01 14:51:51 seems to me it used to be a one floppy distro Dec 01 14:52:09 Oat: yes, in 1989 Dec 01 14:52:20 that would be about what i remember Dec 01 14:52:33 minix3 is quite different Dec 01 14:52:54 was eager to have something unix like on my pc Dec 01 14:53:05 xenix! Dec 01 14:53:28 that wasn't a freebie was it? Dec 01 14:53:39 I doubt it Dec 01 14:54:57 played around with Coherent something Dec 01 14:55:02 not free but reasonable Dec 01 16:01:00 where can one find the original SD image shipped with the BBxm? Dec 01 17:32:01 <_av500_> ha, I almost bought coherent Dec 01 17:32:09 <_av500_> then I discovered that linux Dec 01 17:32:15 <_av500_> big mistake :) Dec 01 17:46:18 jkridner_: is there some beagleboard contest going on? Dec 01 17:59:29 wmat, beaglebone cape design contest is going on Dec 01 17:59:58 when it's over, we'll know who is more capeable Dec 01 18:00:14 *splash* Dec 01 18:01:11 I think it was planned to coincide with the layoff..knowing we needed people to design boards for us Dec 01 18:01:27 next there'll be an SoC design contest Dec 01 18:01:44 mdp: yeah, I saw that one Dec 01 18:02:16 but I received an email via elinux.org asking how to enter the beagleboard contest Dec 01 18:02:23 perhaps they're mistaken Dec 01 18:02:32 there's also the weekly trivia contest Dec 01 18:02:43 http://beagleboard.org/CapeContest/weekly/ Dec 01 18:03:12 I like my cape: https://lh3.googleusercontent.com/-jEn066fvMw8/TutiDHo_IvI/AAAAAAAAAPU/Mqi0pzWAMCo/s848/ Dec 01 18:03:14 gotta enter by tomorrow Dec 01 18:03:43 mru, when they let me on the panel, that's a sure winner Dec 01 18:04:26 "Notice the patterns left by the fabric scissors! This one has everything!" Dec 01 18:08:20 ugh, personal assistant cleaned my office while I was gone...can't find any boards now Dec 01 20:28:26 hi, I am trying to use the DSP on the BeagleBoard using C6Run. I ran into a module "cmemk.ko" being incompatible with my kernel 3.0.17+. This is a TI module (part of the ti-cmem-module package). Opkg downloads this module into the 2.6.32/.... directory in my /lib/, not into the 3.0.17+/..... How can I find the right module? Dec 01 20:30:27 ik try opkg update Dec 01 20:30:50 i did but it downloads the same version Dec 01 20:31:02 does that mean that they have not released it for the newer kernel? Dec 01 20:31:30 it should compile Dec 01 20:39:08 right so the problem is that it's for some reason compiling it against the 2.6.32 kernel and not 3.0.17+ which I am running Dec 01 20:40:24 no the problem is why koen did not compile it for 3.0.17 kernel Dec 01 20:40:36 which beagleboard you have? Dec 01 20:40:50 standard beagleboard c5 Dec 01 20:42:57 so should i downgrade to the previous kernel version to make it work? Dec 01 20:43:19 ik can you run opkg again and tell me the full version of cmem Dec 01 20:44:07 the latest kernel in the feeds seems to be 3.0.28 Dec 01 20:44:08 http://feeds.angstrom-distribution.org/feeds/core/ipk/eglibc/armv7a/machine/beagleboard/ Dec 01 20:44:19 Package ti-cmem-module Version: 1:2_25_05_11-r100d.6 Dec 01 20:45:26 hm try to install 3.0.28 and cmem ti-cmem-module_2_26_01_02-r119e_beagleboard.ipk Dec 01 20:46:25 ok i'll try, thanks Dec 01 20:47:44 ik be aware that the kernel ipkg install the new kernel to /boot Dec 01 20:47:56 you may have to copy it to some other partition Dec 01 20:48:01 depending on your setup Dec 01 20:51:02 if I opkg install the new kernel, will it just give me a uImage that I will have to mount on the SD card? Dec 01 20:52:27 err nevermind Dec 01 21:32:02 <_av500_> hmm Dec 01 21:33:49 mmh Dec 01 21:52:12 has anyone encountered "pcm_read:1801: read error: Input/output error" while trying to use arecord to read usb audio from a webcam? I've tested on an ubuntu 12.04 system and it works OK. Any ideas on the first things I should be checking? (I've googled the error and haven't found any solutions) Dec 01 22:52:46 ds2: you spoke of arks earlier: http://freethinker.co.uk/images/uploads/2012/12/ark.jpg Dec 01 23:00:16 hello gentlemen Dec 01 23:00:25 hello troll Dec 01 23:00:46 i feel so honored Dec 01 23:05:17 i'm having a boot problem with my beaglebone. it boots up, then waits about 1 minute, then shuts down. I THINK it's happening RIGHT after "memtester 1M 1" finishes Dec 01 23:05:25 any clue how to disable this or stop it from crashing? Dec 01 23:06:06 memtester? Dec 01 23:07:03 http://pastebin.com/y3n6sHY8 Dec 01 23:07:49 immediately after the last grep, the thing give me an "Unmounting file systems..." etc etc then system halted Dec 01 23:25:35 Here's the complete output from the boot process: http://pastebin.com/tHtTP06a It just boots up, then shuts down immediately Dec 01 23:28:30 hi, I am trying to upgrade the kernel on my Beagleboard (C5) to 3.0.28. I used opkg and copied the uImage to the SD card. I rebooted and got the error "Verifying Checksum ... Bad Data CRC ERROR: can't get kernel image!" Dec 01 23:29:27 do I need to go through the process of repartitioning the SD card? Dec 01 23:35:12 no? Dec 01 23:35:23 but it to the place where uboot is expecting it Dec 01 23:35:27 ups Dec 01 23:35:29 put Dec 01 23:37:52 ok, so i knoe uboot is looking at 0x80300000 but how do I place it at that location? Dec 01 23:37:55 *know Dec 02 00:17:48 hi again Dec 02 00:19:30 ik__: well clearly the image isn't complete or corrupted Dec 02 00:19:44 or you have a bad SD card Dec 02 00:19:57 * mru suspects failure to unmount/sync properly Dec 02 00:20:09 hmm Dec 02 00:21:42 i just scp'd from /boot to /media/mmcblk0p1/uImage and synced Dec 02 00:22:05 did you try unmouting Dec 02 00:22:20 before copying it over? Dec 02 00:23:57 er no Dec 02 00:25:13 After syncing I just reset the BB... Dec 02 00:32:06 so just to make sure I'm not missing something obvious, is updating the kernel as simple as replacing the old image with the new one on the SD card? Dec 02 00:33:05 in the theory if it is the correct one Dec 02 00:33:29 if they are compatible, yes Dec 02 00:33:48 if the new kernel needs a different command line that will have to be adjusted Dec 02 00:34:09 yeah but you would get farther than CRC error :) Dec 02 00:34:30 oh yes Dec 02 00:37:27 heh the PRU has a nice amount of GP registers Dec 02 00:37:48 just shy of 30 iirc Dec 02 00:37:58 awesome Dec 02 00:38:13 note how the multiplier is wired up Dec 02 00:38:24 ? Dec 02 00:38:34 there isn't a multiply instruction as such Dec 02 00:39:04 you just put the operands in two registers and read the result back from a third Dec 02 00:39:31 something like r26-r28 Dec 02 00:40:08 ah Dec 02 00:40:28 the price you pay for everying being one clock cycle? Dec 02 00:40:36 thanks for your help guys. Out of curiosity (and inexperience), is there a reason why opkg update/upgrade did not automatically download the new 3.0.28 kernel? Could it be b/c it is incompatible or something else? Dec 02 00:40:44 every cycle it multiplies whatever is in the operand registers and puts the product it in the destination register Dec 02 00:40:56 ik__: it should download whatever is in the feed Dec 02 00:41:06 there's also a multiply-accumulate mode Dec 02 00:41:32 in that mode, a mac is triggered by writing to a control register Dec 02 00:41:36 it's all in the manual Dec 02 00:41:45 yes i'm reading it now Dec 02 00:42:45 what assembly syntax is pasm using? Dec 02 00:42:57 its own I'd imagine Dec 02 00:43:09 it should be documented somewhere Dec 02 00:44:48 mranostay: i looked at the feed [http://feeds.angstrom-distribution.org/feeds/core/ipk/eglibc/armv7a/machine/beagleboard/] and that image is there... that's where I manually downloaded it from. I just didn't know if it can somehow be incompatible **** ENDING LOGGING AT Sun Dec 02 02:59:59 2012