**** BEGIN LOGGING AT Tue Oct 23 03:00:01 2012 Oct 23 03:01:38 If data is properly aligned for the sampling operation (by using the processor's own clock and tweaking a delay) would it be possible to read data on the GPIOs at 100 MHz? Oct 23 04:57:29 ok demo really done now :) Oct 23 04:59:25 when do you present? Oct 23 05:01:18 emeb_mac: at E-ELC? Oct 23 05:04:43 * mranostay dd's his SD card just in case both get nuked in transit Oct 23 05:15:55 mranostay: are you keep it next to your 'samples'? Oct 23 05:15:59 or test sources? Oct 23 05:16:21 keeping Oct 23 05:16:28 heh i am not taking anything nuclear with me Oct 23 05:19:20 probably a good idea Oct 23 05:24:24 <_av500_> nucular Oct 23 05:26:02 although i learned techically it is illegal to open smoke detectors :-) Oct 23 05:26:07 http://www.youtube.com/watch?v=YZMsNpyaHcA Oct 23 05:26:25 i want to know how to enable internal sound card on BeegalBoard-XM Oct 23 05:26:52 i want a beegalboard Oct 23 05:27:01 <_av500_> mranostay: you should build an earthquake detector cape next Oct 23 05:27:02 i want a Bengalboard! Oct 23 05:27:09 <_av500_> i want a steak sandwich Oct 23 05:27:19 i want to know how to enable internal sound card on BeegalBoard-XM Oct 23 05:27:39 any one knows how to enable internal sound card on BeegalBoard-XM Oct 23 05:28:54 _av500_: funny you should say that Oct 23 05:29:14 * emeb_mac is laughing Oct 23 05:29:47 but doesn't know why... Oct 23 05:33:20 emeb_mac: we are weird is why? Oct 23 05:33:44 revel in the weird. Oct 23 05:35:47 <_av500_> somesh: its part of the kernel config Oct 23 05:36:00 <_av500_> use angstrom and the sound works Oct 23 05:50:28 I need a bagelboard Oct 23 05:50:50 but no one has it in stock Oct 23 05:51:47 fucking magic smoke... Oct 23 05:52:20 bagelboard.... hmmm.... bagels... Oct 23 05:56:50 wonder what the term is for that Oct 23 05:57:01 if necrophiliac is with dead.... Oct 23 06:23:17 ds2, hold on, I think I found one Oct 23 06:23:18 ! Oct 23 06:23:21 http://www.cooksdirect.com/product/win-holt-bb1826-bagel-board?utm_source=googlepepla&utm_medium=adwords&gclid=CKP8h4bAlrMCFQXhQgodsXEAiA Oct 23 07:24:48 _av500_: pickle me elmo? Oct 23 07:25:59 mranostay: winter is coming Oct 23 07:28:13 sigh, it's the ant and the grasshopper all over again Oct 23 07:28:25 grasshopperboard? Oct 23 07:28:42 if you haven't stored up enough muppet for the winter, don't come crying to me Oct 23 07:30:28 * av500 opens a jar of seasoned Kermit Oct 23 07:36:21 weirdos Oct 23 07:58:41 koen: when running with FLIP in /etc/powervr.ini I get this on boot: http://pastebin.com/zeFFd6C4. with FRONT I do not. it seems to be trying a blanking of /dev/fb1 using the sgx. seen/know this issue? Oct 23 08:06:14 haven't seen it yet Oct 23 09:47:35 koen, does usb work for you on the bone with your latest stack? Oct 23 09:51:54 for small values of "works", yes Oct 23 09:52:07 keyboard, mouse, audio Oct 23 09:52:17 the usual musb problems are still there Oct 23 09:59:12 is this with daniel mack's musb series? Oct 23 09:59:12 i didn't see them in your stack Oct 23 10:10:52 it is Oct 23 10:11:18 shoragan: https://github.com/beagleboard/kernel/tree/3.7/patches/usb Oct 23 10:11:55 shoragan: http://pastebin.com/W6CBNHFY Oct 23 10:12:13 ah, git fail on my side :/ Oct 23 10:17:40 thanks for the pointer Oct 23 11:48:20 mdp, what kind of seating issues with RS232 cape cause Ethernet issues? How does that happen? Oct 23 11:48:30 SilicaGel, did you ever get your SDIO card working? Oct 23 11:50:36 Hello everyone, I want to enable UART2 in my Beagleboard xM rev. C2. So i did pin muxing in linux kernel. How can i test it? If you can help me i'll be very glad. Oct 23 12:02:27 bradfa, I have nfc...I had the cape on..it worked..but then I had those link issues... Oct 23 12:02:50 pulled it off, put it back on and things are rock solid Oct 23 12:02:54 magic Oct 23 12:03:02 magic indeed Oct 23 12:03:18 do you observe my LED inversion issue with u-boot 2012.10? Oct 23 12:03:39 that's on the bone rj45 leds? Oct 23 12:04:05 I never seem to have these boards oriented to have those visible...so no. I can look Oct 23 12:07:23 mdp, yes, on the Ethernet jack Oct 23 12:07:59 Bone A6 is my test platform but A5 does it too, just that the A5 has the orange LED wrong due to resistor settings so it's less easy to observe than the A6 Oct 23 12:08:09 Haven't tested A3 boards Oct 23 12:08:43 Don't really care about A3 boards any more... Oct 23 12:10:08 I only have an A1 and two A5s Oct 23 12:11:07 mdp, you can still see it on an A5 Oct 23 12:11:23 so remind me the case where I see this issue? Oct 23 12:11:48 With 100 Mb full dup connection in u-boot, if you stop the count down, the network gets init'ed. On first power up the Green LED will be ON. Oct 23 12:12:20 if you then issue a 'reset' command from u-boot command line, and interrupt u-boot count down again, now when the Ethernet gets init'ed the Green LED will be OFF Oct 23 12:12:33 if you 'reset' again, Green LED comes back on when you interrupt count down Oct 23 12:12:37 this continues Oct 23 12:12:50 it's like the PHY never gets the LED bits set, they just get inverted Oct 23 12:13:01 and the PHY doesn't get reset when am335x gets reset Oct 23 12:13:15 I need to check the PHY data sheet, u-boot code, and bone schematics to understand this better Oct 23 12:13:31 Linux just seems to invert the LEDs when the interface goes up/down Oct 23 12:13:44 u-boot does the same, except it shouldn't Oct 23 12:14:00 end result is after reboots, Ethernet LEDs might be in reversed mode from expected Oct 23 12:14:20 which is fine, if your CTO doesn't _ABSOLUTELY_LOVE_ LEDs Oct 23 12:14:25 :) Oct 23 12:19:14 lol Oct 23 12:19:41 functional but not blingy enough? Oct 23 12:20:40 thurbad, functional but blinky the wrong way Oct 23 12:21:08 well, blinky one of two ways, one is wrong and the wrong way happens after every other reboot Oct 23 12:21:44 which probably isn't a horrible thing as my bones shouldn't ever reboot, but if I can find excuses to send patches apparently I will try... Oct 23 12:27:15 LEDs on an A6, as I understand it should be: Orange -> 100 ON, 10 or no link OFF. Green -> link ON, no-link OFF, activity BLINK OFF BRIEFLY Oct 23 13:20:36 Hello, Can anyone assist me about UART2 enabling in BeagleBoard xM (rev. C2) ? Oct 23 14:17:31 koen: nightly build of github/beagleboard/kernel failed Oct 23 14:17:33 http://pastebin.com/YaTaiGd2 Oct 23 14:21:16 cehh: it works over here Oct 23 14:22:32 hmm. OK, I will try running manually instead of using jenkins Oct 23 14:25:28 koen: fyi. I'm using 3.7 branch, commit ef967104 Oct 23 14:26:48 cehh: that's a few commits behind already :) Oct 23 14:27:34 ok. I will pull/retry Oct 23 14:29:41 koen: btw, your 'staging' tree is now part of the nightly build/test Oct 23 14:29:58 the latest results can always be found at http://arago-project.org/testresults/linux/beagle-staging/latest/ Oct 23 14:33:44 bradfa: no. :( Oct 23 14:35:25 koen: still no love. http://pastebin.com/ByWmM2tT Oct 23 14:36:40 SilicaGel, did you test with a SD card? Oct 23 14:36:55 not yet, but that's still on my list of things to try Oct 23 14:37:21 SilicaGel, your SDIO slot is hand wired up? Oct 23 14:37:32 and you hav 1.8 to 3.3 level shifters, rigth? Oct 23 14:37:38 I do have level shifters yeah Oct 23 14:37:53 well not exactly hand wired, we did make a circuit board Oct 23 14:38:02 to mount the TiWi-R2 module on, level shifters, and various other things Oct 23 14:39:11 is your TiWi-R2 hard mounted to the bus or is it on a SDIO card? Oct 23 14:40:21 it's mounted to an actual circuit board. Oct 23 14:40:26 Here, let me go take a photograph of it, i'll show Oct 23 14:43:30 bradfa: http://goo.gl/te8Qs Oct 23 14:43:48 the TiWi-R2 is the silvery thing right in the middle Oct 23 14:44:33 SilicaGel, pretty cape :) Oct 23 14:44:43 It's getting there Oct 23 14:44:47 how would you wire up an SD card to that? Oct 23 14:44:51 it has a 3-port USB hub on it too Oct 23 14:45:02 SilicaGel, USB is evil :) Oct 23 14:45:06 and on the bottom is a Mini PCI Express slot for cell card Oct 23 14:45:43 well i was planning on using a blank pcb and dead bugging it Oct 23 14:45:45 PCI-e on a bone? Oct 23 14:45:51 cool Oct 23 14:45:59 bradfa: no for usb to the cell card Oct 23 14:46:01 well it's Mini PCI Express Oct 23 14:46:03 yeah Oct 23 14:46:16 mini pcie carries also usb and a few other things Oct 23 14:46:23 Mini PCI Express = USB + PCIx1 I think on the same slot; cell cards only use the USB piece of it so that's the only part we hooked up Oct 23 14:46:28 the cell card works great Oct 23 14:46:30 it was stupidly easy Oct 23 14:46:33 opkg install pppd Oct 23 14:46:36 basically that was it Oct 23 14:46:38 dm8tbr, ExpressCard carries USB, mini PCI-e does to? Oct 23 14:46:38 :) Oct 23 14:46:48 yes Oct 23 14:46:51 yep Oct 23 14:47:05 thats how all the modems in laptops work so far Oct 23 14:47:13 cehh: ah, I forgot to push, try it now Oct 23 14:47:14 yeah this is basically a laptop card we're using Oct 23 14:48:01 It still strikes me as funny how the SDIO clock is a sine wave Oct 23 14:48:10 but it's too perfect a sine wave to be a coincidence Oct 23 14:48:47 but the thing is, I think I'm dead in the water until somebody here agrees to buy me an SD-aware logic analyzer Oct 23 14:49:26 Right now it just fails to load the firmware, and it thinks that the device id it sees on sd is some somewhat-randomish number like 0x66666666 or 0xEEEEEEEE or sometimes 0xFFFFFFFF Oct 23 14:49:28 beware, the plague is upon us Oct 23 14:49:31 it seems to change, haven't figured out why. Oct 23 14:50:15 on a scope, the level converters seem to be doing the right thing, waveforms are relatively crisp and clean Oct 23 14:50:49 koen: it works now, thanks! Oct 23 14:51:22 I need to update my recipe now to use mainline v3.7-rc2 Oct 23 14:52:11 On the upside, though, I replaced the firmware in my sparkfun USB-Weather-v2 board with something that emits ZCL messages that get sent over zigbee (xbee) to my beaglebone Oct 23 14:52:12 one more question... when you build for beaglebone are you using the defconfig under configs/beaglebone Oct 23 14:52:14 and taht part seems to work Oct 23 14:53:03 and now I dig zigbee. I wish they didn't have a moron license. Oct 23 14:59:24 SilicaGel, indeed Oct 23 15:01:29 SilicaGel, clock really shouldn't be a sine wave Oct 23 15:01:47 but neat to see zigbee working Oct 23 15:01:48 i thought that too but Oct 23 15:02:08 one day I flipped the beaglebone upside down and I probed the clock right on the MicroSD card and it was also very sinusoidal Oct 23 15:02:10 not squarewaveish Oct 23 15:02:25 I love exact technical terms Oct 23 15:02:37 SilicaGel, my scope plot I sent you was fairly square wave Oct 23 15:02:40 and I figured that the BB folks would have taken care about lineimpedances and stuff Oct 23 15:02:46 that was direct on pull-up for bone clock line Oct 23 15:04:26 yeah, waht you sent did at least have actual edges. Oct 23 15:04:28 SilicaGel, what's your zigbee going to do? report the weather? Oct 23 15:04:55 That's one thing, yeah Oct 23 15:05:22 Right now though i'm racing toward something else :D This TiWi thing, taht's for work. my home project is muchless useful, but more interesting Oct 23 15:05:32 SilicaGel, now till May -> snow. May till June -> rain. July to August -> hot. September to now -> overcast and rain Oct 23 15:05:55 I did a layout for a 2 sided circuit board that's a 120VAC to 3.3VDC power supply, an Xbee Programmable module (with the second microcontroller on it), and four solid state relays Oct 23 15:05:57 SilicaGel, Rochester has great weather :) Oct 23 15:06:11 SilicaGel, cool! Oct 23 15:06:19 I'm trying to hurry up before Christmas to make synchronized chrismast lights I can put on my lawn, and my neighbors lawns Oct 23 15:06:32 and try to do a kind of low-speed synchronized christmas display across multiple yards Oct 23 15:06:52 SilicaGel, that's way cooler than any work project :) Oct 23 15:07:02 especially if you can make car crashes Oct 23 15:07:07 from distraction Oct 23 15:07:14 oh boy, I hope not, but I have one of those umbrella policies Oct 23 15:07:53 If I were a more powerful villain, then at some point it would be revealed that all these Christmas trees were, in fact, Daleks, and they would exterminate all the traffic on state route 31 Oct 23 15:07:56 maybe next year Oct 23 15:08:50 What's the tiwi thing for? Oct 23 15:09:42 it's for asset health monitoring for equipment. Vehicles, generators, pumps, that sort of thing. Oct 23 15:09:56 That board is to allow it to communicate out. Data and alarms. Oct 23 15:10:21 So each asset that gets monitored gets a bone with cape? Oct 23 15:11:24 yeah, and if it's a vehicle, it also gets the other cape, which is an OBD-II interface, GPS receiver, and 6-axis accelerometer Oct 23 15:11:49 well, and also J1939/J1708 Oct 23 15:11:57 but only when I actually finish my J1708 implementation in the PRUSS Oct 23 15:12:37 SilicaGel, any relation to Vnomics? Oct 23 15:12:47 ................ Oct 23 15:12:50 why do you ask that I wonder Oct 23 15:13:10 now i'm probably in trouble Oct 23 15:13:12 haha Oct 23 15:13:13 SilicaGel, heard through the grape vine :) Oct 23 15:19:35 je suis Oct 23 15:20:36 hello, i am developping a proyect with the beagle board xm using the distribution system "angstrom". Oct 23 15:21:08 USB does stink, but with each kernel release it seems to get a little bit better Oct 23 15:25:48 av500: should I have said something about plugging the do-hickey into the thing-a-ma-bob too? :) Oct 23 15:28:03 yep Oct 23 15:30:42 huh, Bus Speed Mode 7 DDR50: 1.8V signaling, clock up to 50 MHz, sampled on BOTH CLOCK EDGES Oct 23 15:30:44 that's nice. Oct 23 15:31:02 Too bad I don't have any clock 'edges' haha Oct 23 15:31:15 <_Unsolo> mru: so i got my mmu configured i think mapping everything from 0x000xxxxx to 0xFFFxxxxx into sram (just seemed like a safe place to stick the tables) but when i try to enable D cache or I cache or L2 cache it all blows up on me.. Oct 23 15:31:21 SilicaGel, I believe you have to handle 3.3V signalling first though to get to that mode Oct 23 15:31:43 So it seems unlikely the tiwi module is using said mode Oct 23 15:34:03 SilicaGel, probably not, but am335x can't do that mode anyway :) Oct 23 15:34:13 oh, interesting Oct 23 15:34:33 any of the DDR modes I believe are SD v3.00 things, am335x only supports SD 2.00 Oct 23 15:34:48 Oh, good. I don't need that complication right now. Oct 23 15:35:32 * bradfa needs lunch! Oct 23 15:35:50 * _Unsolo needs a MMU dm37x guide Oct 23 15:58:33 <_Unsolo> If i enabled the mmu and program seams to execute i have done something right right ? Oct 23 16:04:33 maybe Oct 23 16:08:40 <_Unsolo> well its executing but waay slower than the pace it had when I cache was enabled.. Oct 23 16:13:34 <_Unsolo> av500: with mmu on and i cache on it outperforms mmu off i cache on Oct 23 16:13:44 <_Unsolo> now what i need is to get the d cache on i belive Oct 23 16:16:44 <_Unsolo> but from what i can read some think that is just a matter of fliping a bit in c1,c0 Oct 23 16:26:34 yes Oct 23 16:26:49 im just checking my ancient mmu code from arm9 times :) Oct 23 16:28:04 <_Unsolo> a bigger question is if my dma routines will fail epically with the D cache on and if L2 needs to be enabled for L1 to work Oct 23 16:30:32 <_Unsolo> at least that would explain my crash when i enable D cache Oct 23 16:32:36 you don't need L2 for L1 Oct 23 16:33:11 I-cache can always be enabled Oct 23 16:33:25 since you're not allowed to execute from device/s-o memory anyway Oct 23 16:35:44 <_Unsolo> so if I cache + MMU outperforms only I cache i should be able to also enable D-cache (L1)right? Oct 23 16:36:21 <_Unsolo> simply by flipping the C bit in c1,c0 Oct 23 16:39:14 <_Unsolo> this is almost interesting.. Oct 23 16:41:25 almost Oct 23 16:41:59 <_Unsolo> so im going to try with A and V bit disabled Oct 23 16:42:12 <_Unsolo> but I and D enabled Oct 23 16:42:18 <_Unsolo> and MMU enabled Oct 23 16:42:31 enable all the bits! Oct 23 16:42:37 <_Unsolo> hmm Oct 23 16:42:39 <_Unsolo> sure ? Oct 23 16:42:50 <_Unsolo> im quite confident it wil epically fail on me Oct 23 16:43:11 bit-bang it! Oct 23 16:43:24 bit-bangbus Oct 23 16:43:39 <_Unsolo> only bit banging i will do will be on r0 Oct 23 16:44:18 <_Unsolo> its either ORR or BIC Oct 23 16:45:17 we need a cpu with a BNG instruction Oct 23 16:45:43 <_Unsolo> so you can really go out with a Bang ? Oct 23 16:46:46 <_Unsolo> mru should it matter if i dma or not with 1:1 mapping ? Oct 23 16:47:06 <_Unsolo> again enable D-cache results in failure Oct 23 16:47:18 <_Unsolo> trying all the bits Oct 23 16:47:27 BBNG r0,[r3,#0]? Oct 23 16:47:28 if you use dma you must do proper cache maintenance Oct 23 16:47:44 <_Unsolo> mru perhaps thats whats failing Oct 23 16:47:54 <_Unsolo> i will try to disable dma Oct 23 16:48:01 clean before dma read, invalidate after dma write Oct 23 16:48:49 <_Unsolo> hmm ok i guess i can fix that after i get D and L2 online Oct 23 16:49:25 then L3 Oct 23 16:49:26 and L4 Oct 23 16:49:35 skip L2 for now Oct 23 16:49:45 <_Unsolo> L3 and L4 are working Oct 23 16:49:46 cache all the levels Oct 23 16:49:50 <_Unsolo> oO Oct 23 16:49:56 av500, need to get the bitbang cape on kickstarter Oct 23 16:50:04 <_Unsolo> its just a slightly to smart booter Oct 23 16:50:36 hello, i have a beagleboard xm with serial DB9 connector, I tried to connect a device on this connector, but when I try read cat /dev/ttyO0 cat /dev/ttyO1 cat /dev/ttyO2 cat /dev/ttyO3 cat /dev/ttyO4 cat /dev/ttyO5 not happens i'm using a debian squeeze. i had disabled the console boot on uEnv.txt Oct 23 16:50:36 <_Unsolo> mru: would still be fun to cache all the levels. Oct 23 16:50:52 get it working with only L1 first Oct 23 16:51:12 rgimenes: there could still be a getty on that uart Oct 23 16:51:16 console is only for kernel Oct 23 16:51:39 also you need to setup baud rate Oct 23 16:52:36 i tried to use a setserial of debian, but this script only run with ttyS devices Oct 23 16:52:38 mdp: bitbang cape, awesome! Just a bunch of gpio pins, but have arbitrary labels for SPI, I2C, CAN, USB.... I wonder who much you could raise. Oct 23 16:53:19 the DB9 connector means what file output? /dev/ttyO? what number Oct 23 16:53:43 alan_o, yep, and the s/w will preconfigure all the pinmuxes to mode7 for you. I'll have to say, it's quite a stunning idea Oct 23 16:54:17 alan_o, I was thinking it would come with some blank stickers that you could relabel what the pins are for Oct 23 16:54:30 mdp: nooo Oct 23 16:54:38 a little OLED next to every screw terminal Oct 23 16:54:42 at the $200 backer level I'll throw in a yoctopus Oct 23 16:54:53 like that russian keyboard Oct 23 16:54:56 lol Oct 23 16:55:07 of course the OLEDs are bitbanged Oct 23 16:55:14 of course! Oct 23 16:55:36 I'm not just a member of bitbang club...I'm the president Oct 23 16:55:45 hah Oct 23 16:56:42 <_Unsolo> mru enabling the D-cache still fails on me.. Oct 23 16:57:06 check how u-boot does it Oct 23 16:57:12 make some example programs for it that do things as slowly as possible, then make comments about how Linux is slow compared to a $30 Arduino Oct 23 16:57:42 alan_o: it will all be done high level in python Oct 23 16:57:59 "compared to The Greatest Microcontroller Ever(tm)" Oct 23 16:58:00 using RPC to talk to sysfs Oct 23 16:58:20 yeah, sysfsd Oct 23 16:58:53 alan_o, http://www.kickstarter.com/projects/18182218/freesoc-and-freesoc-mini, if you didn't see that one before. Oct 23 16:58:58 systemsysfsd Oct 23 16:59:18 python, even better... I was actually thinking last night about how device tree is just another interpreted language, and how we might as well just embed a python interpreter in the kernel to make our board files in. Oct 23 16:59:43 "It's just data" Oct 23 17:00:05 err, how about no... Oct 23 17:00:05 riiiight... just like a jar file :) Oct 23 17:00:14 panto, quiet! Oct 23 17:00:27 panto: luddite! Oct 23 17:00:32 yeah, panto, you just hate new technology. Oct 23 17:00:40 shame! Oct 23 17:00:52 <_Unsolo> mru: any tips as to where in the u-boot source i should look for it lowlevel_init.S doesn't cover it Oct 23 17:00:56 DT is a nice description of the relationship of modules Oct 23 17:01:08 You C guys are the past. Python in kernel is the future. In fact, I saw some benchmarks that _proved_ that python is faster than C in all cases. Oct 23 17:01:14 _Unsolo: there are some files with 'cache' in their names iirc Oct 23 17:01:19 panto, I hear it's a tree of devices Oct 23 17:01:33 in fact it's a bit-banged tree of bit-banged devices Oct 23 17:01:41 * mdp staggers backward Oct 23 17:01:44 whoa Oct 23 17:01:50 alan_o: its enough to prove that is is never slower Oct 23 17:01:53 BDBT Oct 23 17:02:01 panto, you just blew mdp's mind... Oct 23 17:02:18 * av500 has a bit bang theory Oct 23 17:02:24 I'm going on sabbatical to ponder bit bang theory Oct 23 17:02:25 he'll get over it Oct 23 17:02:53 we should ask for a tv-show Oct 23 17:02:54 panto, I shall go to a zombie-proof meditation location in Meteora Oct 23 17:03:05 god, I wish I'd go too Oct 23 17:03:05 <_Unsolo> next thing ill see here is that python has very nice RT properties Oct 23 17:03:33 pythonRT solves that Oct 23 17:03:44 calling dibs on the nicer WC spots Oct 23 17:03:45 _Unsolo, that's crazy talk, but JS is actually pretty good for RT stuff. Oct 23 17:03:59 panto, LOL Oct 23 17:04:03 <_Unsolo> specially on predictable delay RT execution Oct 23 17:04:20 well, RT calls about predictability, not speed Oct 23 17:04:32 with python you could say that you are going to be predictably slow :) Oct 23 17:04:56 I predict failure Oct 23 17:04:58 a jitter of a few msec is nothing when everything takes a few tens of seconds Oct 23 17:05:30 RT implies that time actually passes Oct 23 17:05:45 in a geological scale it does Oct 23 17:05:46 av500, time stands still when you're writing JS? Oct 23 17:06:16 this time business is all just a theory Oct 23 17:06:17 <_Unsolo> at least when you try to read what you have written Oct 23 17:06:22 since it flies when you are having fun, that is indeed correct Oct 23 17:06:30 :) Oct 23 17:06:35 bit bang theory.. oh god Oct 23 17:06:49 mranostay, you're in charge of writing the parody song. Oct 23 17:06:53 and its not even friday here... Oct 23 17:07:06 <_Unsolo> prefers the bitch bang theory Oct 23 17:07:17 that's x-rated Oct 23 17:07:24 mranostay, can you make the song be in the spirit of the team america soundtrack? Oct 23 17:07:45 is someone singing an nda? Oct 23 17:07:54 in the rain? Oct 23 17:08:27 <_Unsolo> that would be crying not bit Oct 23 17:08:28 hrm, aptina nda terms set to music..I like this Oct 23 17:08:59 team bit bang? Oct 23 17:09:24 <_Unsolo> there is an x-rated alternative to that as well. Oct 23 17:09:34 perhaps Psy can choreograph a dance for us? Oct 23 17:09:46 <_Unsolo> whoppa bit bang style Oct 23 17:10:15 the dancers can be holding RPi's Oct 23 17:10:50 I hate to visualize the dancers being in that music video Oct 23 17:11:00 not going to be very pleasing to the eye Oct 23 17:11:25 well, let's try to keep this mass appeal, let's not recruit dancers at ELC-E, ok? Oct 23 17:11:29 * _Unsolo wonders why this cannot be some asian chicks dancing. Oct 23 17:12:20 * mranostay thinks _Unsolo is missing the point Oct 23 17:12:25 _Unsolo: on the iphone assembly line? Oct 23 17:12:37 that would be an underage video :) Oct 23 17:13:06 panto: in your jurisdiction! Oct 23 17:14:02 * panto checks age of consent here... 15... Oct 23 17:14:16 which is a little creepy to know Oct 23 17:14:45 of course it is 16 min for any US citizen even overseas Oct 23 17:14:53 creepy i know that as well :) Oct 23 17:15:06 mranostay, wha??? is that true? Oct 23 17:15:15 That can't be true Oct 23 17:15:20 The law doesn't go with you. Oct 23 17:15:28 oh yes it does Oct 23 17:15:37 alan_o: god's law goes everywhere :) Oct 23 17:15:40 (neither does the bill of rights, to the surprise of many moronic tourists) Oct 23 17:15:44 * av500 fought the law, the law won Oct 23 17:16:56 * _Unsolo doesnt want to know the age Oct 23 17:17:09 mranostay, I think I see what you're saying. If the overage person goes overseas, yes I've heard of that. Oct 23 17:17:29 overweight most likely too Oct 23 17:17:44 hah, yeah. I saw something on 20/20 about that one time. Oct 23 17:17:49 alan_o: which really even if it is legal it is sooooo creepy :) Oct 23 17:18:36 lets get back to banging bits Oct 23 17:18:43 oh yes! Oct 23 17:19:14 yeah, it got dark in here in a hurry..... that's a lot mranostay ...... Oct 23 17:19:26 it is what i do Oct 23 17:20:36 _av500_: bring back the light Oct 23 17:20:42 have a sunrise photo? Oct 23 17:21:08 <_Unsolo> mru: still lookinf for the code that enables d-cache.. Oct 23 17:21:16 <_Unsolo> -f+g Oct 23 17:22:56 _Unsolo, I haven't seen the code, but in the uboot for beagle bone, there's a dcache off and dcache on command. Oct 23 17:23:10 should be straight-forward enough to find Oct 23 17:23:32 geez, I leave for 5 minutes... Oct 23 17:23:36 * mranostay cranks up tunes to drown out the nearby nodejs talk Oct 23 17:24:10 skrillex? Oct 23 17:24:47 Call Me Maybe? Oct 23 17:25:14 AZIZ, LIGHT! Oct 23 17:25:32 mranostay, crank that up! Oct 23 17:26:36 mranostay, but everyone knows threads don't scale. Oct 23 17:27:03 mranostay, and non-blocking I/O. Oct 23 17:27:24 nodejs is a bad ass rock star Oct 23 17:28:28 and it's webscale Oct 23 17:28:32 ewwwwwwwwwwwwwwwwwwwwwwwwwwwwww javascript Oct 23 17:28:42 blah...ick ick ick Oct 23 17:29:02 I read javascript is the future Oct 23 17:29:12 in a trade magazine Oct 23 17:29:12 <_Unsolo> cache_v7.c seems to take it quite far Oct 23 17:29:19 yeah, where's my copy of Information Week Oct 23 17:29:30 <_Unsolo> i was hoping to get away with waay less Oct 23 17:30:24 mranostay, thanks a lot for the "call me maybe.." Oct 23 17:30:54 mranostay, I'm gonna have to play the PSY video to get it out of my head now.... Oct 23 17:31:06 <_Unsolo> call me maybe or the psy video Oct 23 17:31:26 _Unsolo, yeah, I'll be out of the frying pan into the fire. Oct 23 17:31:28 have they remixed those togather yet? Oct 23 17:31:54 I'm sure they're working on it. Oct 23 17:33:26 it's friday? Oct 23 17:33:52 so I think I finally got my embedded board to stop locking up under heavy ethernet traffic. Found a race condition in the driver. Oct 23 17:34:40 http://www.youtube.com/watch?v=Co4Rsne-KD4 Oct 23 17:35:06 no need to listen to two songs Oct 23 17:36:48 hehe Oct 23 17:44:06 av500: i sent my full question on beagleboard group (google). Oct 23 17:44:55 aw the weekly digest i never read :) Oct 23 17:45:35 hic.. Oct 23 17:46:26 too early to start drinking Oct 23 17:46:35 mranostay: it's almost 7pm Oct 23 17:47:09 oh i keep forgetting Europe lives in a time warp Oct 23 17:48:12 cant say u-boot source code made me a lot wiser.. Oct 23 17:48:38 Unsolo_: made you hate life more right? :) Oct 23 17:49:24 well no , more hate generalization of not so general stuff Oct 23 17:52:10 interesting Oct 23 17:54:36 mru whats most interesting is probably the cp_delay in the u-boot Oct 23 17:55:04 it seems they run a delay between coprocessor (which isnt a co processor) read and write Oct 23 17:55:16 where? Oct 23 17:55:22 :q Oct 23 17:55:36 arch/arm/lib/cache-cp15 Oct 23 17:55:40 .c Oct 23 17:58:41 Hello all! I want to enable uart2 in my bb xm rev. c2. What path do i need to follow? Oct 23 17:58:44 something tells me it might be easier to write this in c.. Oct 23 18:00:45 Bolkar, might have to do some pin muxing Oct 23 18:01:22 look at /arch/arm/mach-omap2/omap3-beagle.c, specifically under the section where it sets up the zippy2 board. Oct 23 18:01:36 zippy2 has another serial port on it Oct 23 18:01:38 not sure if it's uart2 Oct 23 18:01:42 might be uart1 Oct 23 18:01:50 should get you pointed in the right direction Oct 23 18:01:55 omap3evm has uart1 up wich is somewhat different Oct 23 18:02:06 see the second last post in this thread: https://groups.google.com/forum/?fromgroups=#!topic/beagleboard/-msQlYWlw_Q Oct 23 18:03:44 alan_o , Unsolo_ and wmat , thank you for your answers. I'll immediately look and feedback here. Oct 23 18:04:49 guys, if you have a beaglebone and have interest in capes give this a whirl: git@github.com:pantoniou/linux-bbxm.git branch capebus-v1 Oct 23 18:05:17 docs are pretty scarce ATM, but if you take a look at the bone dts you'll get the idea Oct 23 18:06:00 mru: is it also neccesary to setup the dram_bank_mmu ? Oct 23 18:07:05 uh? Oct 23 18:07:11 that function is what builds the page tables Oct 23 18:07:31 gee, thanks much, Russ :P Oct 23 18:08:06 mmu_setup() first builds a flat uncached mapping covering everything Oct 23 18:08:22 then it enables cache for the pages covering the dram Oct 23 18:08:31 i see Oct 23 18:09:02 Russ, checkout the capebus branch; it has the generic cape support thing we talked about Oct 23 18:09:41 so dram bank mmu_setup is over writing the page_tables with data such as WriteThrough etc. Oct 23 18:09:56 that's what it looks like to me Oct 23 18:10:02 wmat, i look to the post. I need to recompile the u-boot , not kernel right? Oct 23 18:12:13 Bolkar: first test that it's not already setup. That thread has the tests in it. Oct 23 18:12:28 mru: so if i know my "ram" is from 800 to 9ff i could simply add 0x1A to all those ? Oct 23 18:12:53 and perhaps that is what is failing ? Oct 23 18:13:01 you should set up mapping for all the memory Oct 23 18:13:09 otherwise your mmio registers will be inaccessible Oct 23 18:13:21 s/all memory/full address space/ Oct 23 18:13:33 ofc 0x000xxxx to 0xfffxxxxx Oct 23 18:14:16 panto, in who's repo? Oct 23 18:14:50 mru: my question is if the d cache needs 0x1A or 0x1e to be set to work Oct 23 18:15:00 eg on the actual memory Oct 23 18:15:40 look up those bits in the manual Oct 23 18:16:09 Russ: https://github.com/pantoniou/linux-bbxm/commits/capebus-v1 Oct 23 18:16:49 wmat, i tested uart2_tx and mcbsp3_fsx and they are exactly same with the post. Oct 23 18:17:54 the bit 0 means inner cachable Oct 23 18:18:04 L1 is inner cache Oct 23 18:18:20 L2 is configurable inner or outer Oct 23 18:18:26 if you don't enable it you don't need to care Oct 23 18:19:31 Anyone experienced with rootfs images of angstrom on beaglebone mind helping me out? Oct 23 18:19:40