**** BEGIN LOGGING AT Wed Mar 04 02:59:58 2015 Mar 04 06:56:50 IN SOCKET COMMUNICATION IS ADK is needed for bbb Mar 04 07:12:56 Hi, I am trying to use the Serial UARTS on the BeagleBone Black for my application. I have enabled them through the uEnv.txt file through "capemgr.enable_partno". However, I noticed that I can't access the RTS pin. Mar 04 07:13:17 Running the command "sudo cat /proc/tty/driver/OMAP-SERIAL" gives me: this: uart:OMAP UART4 mmio:0x481A8000 irq:45 tx:0 rx:0 CTS|DSR Mar 04 07:14:01 How can I access the RTS pin on the board? I am using Debian by the way Mar 04 07:18:06 echo BB-UART1 > /sys/devices/bone_capemgr.9/slots Mar 04 07:18:13 to turn on UART1 ? Mar 04 07:18:18 then etc etc for the rest Mar 04 07:21:39 I tried that earlier, but I get "bash: /sys/devices/bone_capemgr.9/slots: Permission denied" even as root Mar 04 07:24:14 The UARTS are already enabled, but it seems like I am not getting the pins like RTS. I read about the Device Tree and tried things like "sudo echo ttyO4_armhf.com > /sys/devices/bone_capemgr*/slots" (using a file I downloaded while trying to follow a guide), but I keep getting that error Mar 04 07:32:39 is this the stock debian or a custom image ? Mar 04 07:33:28 I am running a stock debian image Mar 04 07:36:26 This is the dmesg log from the system: http://pastebin.com/cVze5t2Q Mar 04 07:42:52 illtry and compare with mine later. Mar 04 07:43:13 it just worked for me once i didthe slots echo as above Mar 04 07:43:21 fresh image, off it goes. Mar 04 07:49:10 I'll re-image the thing and see if it works Mar 04 08:02:34 how to run video using vnc server ? Mar 04 08:03:16 do i need to download anything extra Mar 04 08:04:58 what do you mean by 'running video'? Mar 04 08:10:16 i have a video on my sdcard , i want play it Mar 04 08:10:30 its in mp4 format Mar 04 08:11:21 you'll need a video player for that. Like VLC or gstreamer or avplay Mar 04 08:32:37 Hi, I am new here. I am interested to participate in Beagleboard development. Mar 04 09:19:43 Hi friends.. Mar 04 09:22:05 I'm trying to update Angstrom linux by using opkg update. but it tries to connect to angstrom-distribution.org/feeds Mar 04 09:22:58 angstrom-distribution.org/feeds doesn't exist and it shod connect to feeds.angstrom-distribution.org/feeds Mar 04 09:23:28 help me pleasee.. Mar 04 09:30:48 anybody's there? Mar 04 09:32:36 Agol: I'd say just reinstall with latest image Mar 04 09:40:45 thanks.. where can i find latest image of linux angstrom? Mar 04 09:41:21 not sure, but you could always build one with narcissus, custom to your needs Mar 04 09:44:15 id move to the new debian image unless you had reasons not to Mar 04 09:44:58 yeah, that's the other option Mar 04 09:50:10 <_av500_> Agol: do you really need angstrom? Mar 04 09:50:16 <_av500_> if not, install the latest debian Mar 04 09:56:04 i don't really know Mar 04 09:56:42 angstrom is awful..is it possible to install ubuntu or debian on beagleboard? Mar 04 09:56:50 angstrom is a bit of an orphan, debian is a far better thing to code against. Mar 04 09:57:10 beagleboard or beaglebone Mar 04 09:57:20 there are official debian images Mar 04 09:57:40 afaik they have abandoned angstrom for the bone, debian is the new default Mar 04 09:57:41 <_av500_> there is also debain Mar 04 09:57:47 <_av500_> and debian Mar 04 09:57:52 <_av500_> did I mention debian? Mar 04 09:58:44 _av500_: what about that debian ? Mar 04 09:58:57 <_av500_> jo Mar 04 09:59:15 in fact i'm using BeagleBoard-xM Board based on TI DM3730 Mar 04 10:02:02 _av500_: do we still keep android under angstrom projec ? seen on the ML a post, somebody asked about it Mar 04 10:02:07 if we still ditch angstrom Mar 04 10:12:31 <_av500_> vvu: I dont push for that Mar 04 10:20:58 agol: the beagleboard-xm is a bit of an orpahn itself Mar 04 10:21:11 i think its only got the old angstrom image Mar 04 10:21:29 i got one of those from work, however havent dont much with it due to lack of io Mar 04 10:24:07 you mean , that it isn't possible to install ubuntu on beagleboard-xm? :( Mar 04 10:24:13 or debian Mar 04 10:24:15 ? Mar 04 10:24:43 it says it has ubunutu, my mistake, i have the basic beagleboard Mar 04 10:24:56 i mostly use beaglebone so i get confused on the other ones sometimes Mar 04 10:25:22 oh...thanks Mar 04 10:25:43 i found this blog useful: http://www.brianhensley.net/2013/01/beagleboard-xm-how-to-install-ubuntu.html Mar 04 10:26:56 http://elinux.org/BeagleBoardUbuntu Mar 04 10:27:36 thanks alot.. Mar 04 10:29:12 hi Mar 04 10:29:14 xm looks ok, it actually has io, the basic one i have only has single usb Mar 04 10:38:04 Is there anyone with DTO and LCD driver experience for the beaglebone black on Debian? Mar 04 10:38:05 I'll pay for your assistance Mar 04 11:43:27 is the socket communication is possible in Beaglebone black in android Mar 04 11:43:54 and android device with static ip address Mar 04 11:44:05 sure Mar 04 11:44:21 sockets are the same for all OS Mar 04 11:44:29 without using ADK is it possible Mar 04 11:45:25 don't know what ADK is , but as long as yu can ping , you should be able to send/receive from cokets Mar 04 11:47:40 is it possible for static ip address Mar 04 11:48:09 yes Mar 04 11:48:53 any link for reference Mar 04 11:49:28 any resource or examples for reference Mar 04 11:50:04 man nc Mar 04 11:50:23 ? Mar 04 11:51:12 any references available Mar 04 11:51:46 every programming language has a socket library Mar 04 12:02:24 ok Mar 04 12:43:22 Hi Mar 04 12:44:11 I'm in problems to enable GPIO2(P9_22) in Beaglebone Black Mar 04 12:44:29 IN AM3359AZCA ok worked.... Mar 04 12:44:40 IN AM3358BZCZ problem, not enable Mar 04 12:45:11 did anyone tried to run miracast from the bbb yet? Mar 04 12:45:12 # ls -1 /sys/class/gpio/gpio2/ Mar 04 12:45:13 active_low direction edge power subsystem uevent value Mar 04 12:45:36 # cat /sys/kernel/debug/gpio GPIOs 0-31, gpio: gpio-2 (sysfs ) out hi Mar 04 12:46:09 Distribution Linux generate with Yocto Project 1.7 , using kernel 3.14 Mar 04 12:46:52 apparently with kernel 3.8 function ok Mar 04 12:49:05 Generate device tree source with PinMux to Texas to GPIO (0x150), no to uart2 or spi0_clk Mar 04 13:10:48 has anybody built an openpandora-like case for the pi2? Mar 04 13:10:56 oops wrong channel Mar 04 13:24:55 hello everyone Mar 04 13:25:04 i have a question Mar 04 13:25:08 any help? Mar 04 13:25:21 was that your question? Mar 04 13:25:29 answer is depends Mar 04 13:25:53 which processor is used in beagleboard xm? Mar 04 13:26:13 am3715 or dm3730? Mar 04 13:26:43 ? Mar 04 13:26:56 that is my question! Mar 04 13:27:19 <_av500_> dm3730 Mar 04 13:28:25 below webpage shows that the processor is am37X Mar 04 13:28:29 http://beagleboard.org/beagleboard-xm Mar 04 13:29:01 i'm confused Mar 04 13:29:20 finally, DM3730 or AM3715? Mar 04 13:36:43 <_av500_> its the same Mar 04 13:37:02 <_av500_> TI made similar CPUs under different names Mar 04 13:37:27 <_av500_> in order to confuse people Mar 04 13:37:45 <_av500_> AM37xx == DM37xx Mar 04 13:37:57 <_av500_> == OMAP36xx Mar 04 13:38:54 <_av500_> xx might not always be the same Mar 04 13:39:09 <_av500_> looks like AM3715 = DM3730 Mar 04 13:49:35 0/ Mar 04 13:49:46 they probably renamed it to get rid of the omap name Mar 04 13:50:20 can someone help to newbie BBB user? :) Mar 04 13:50:31 I've got a custom board based on beaglebone and I'm trying to figure out what kernel version to use Mar 04 13:50:54 should I get the one from github github.com/beagleboard/linux ? Mar 04 13:51:08 or from kernel.org? Mar 04 13:51:18 codehero1: that will also depend on what you want to do with your board Mar 04 13:51:52 USB, ethernet, UART ports, ADC Mar 04 13:51:56 and gpio hogging Mar 04 13:52:19 and SPI and i2c of course Mar 04 13:52:44 Just got my BBB right out of Amazon box, but Win 7 x64 refusues to recognize it at all. I\ve got the x64 drivers from official biut no good with it. Mar 04 13:54:25 codehero1: sounds like going with kernel.org should work for you. Mar 04 13:55:04 laborant: how long did you wait? did the pc recognize anything? any leds on? Mar 04 13:56:16 tbr: thanks; I guess my question would be why have the separate github beagleboard linux branch then? Mar 04 13:56:36 hello everyone! i want to use my pru in beaglebone but no idea from where to start. can any one help? Mar 04 13:56:39 up to 10 mins, double restarted, USR0 & USR 2 are flashing Mar 04 13:56:54 codehero1: because it wasn't always so nice Mar 04 13:58:38 codehero1: for quite long many things worked best on the TI frankenkernel, also it had the convenient overlay mechanics with capemanager. now overlays are upstream at least Mar 04 13:58:59 I've got HDMI to VGA cord, when I try to start BBB in direct, it shows me Linux logo for couole of secs and then just go off the black screen Mar 04 13:59:33 tbr: OK makes sense. Mar 04 14:00:16 laborant: sounds like the board is working then. Mar 04 14:00:49 laborant: did you reboot after installing the drivers? Mar 04 14:01:01 sure double time Mar 04 14:01:44 If BBB is Ok where I can find it in Windows Device Manager? Mar 04 14:02:19 laborant: does the device manager show any unknown devices? I think that was a yellow question mark. Mar 04 14:02:59 No It doesn't Mar 04 14:04:24 Oh, I ca see Beagle_Bone in Portables now, but it still not connected as USB drive in WE Mar 04 14:07:44 laborant: it should also show up as a removable USB drive Mar 04 14:08:22 Let me check disk manager could be logical drive issue Mar 04 14:31:15 hi all Mar 04 14:31:46 help me Mar 04 14:32:05 <_av500_> if only we knew how Mar 04 14:33:10 i need flash zImage for beagleboardblack Mar 04 14:33:25 but i don't know Mar 04 14:33:48 Can i help me? Mar 04 14:34:37 http://lmgtfy.com/?q=flash+beagleboneblack+kernel Mar 04 14:34:56 <_av500_> in most cases its not flash, but "copy" Mar 04 14:41:09 rcn-ee: I posted patch set for pruss on ti 3.14 on my page http://www.aceware.ca/2015/03/enable-pruss-on-beagle-bone-black-with-ti-kernel-3-14/ Mar 04 14:43:40 Catslab: what's new about the patch? Wasn't there already some PRU support in that kernel? Mar 04 14:45:42 jkridner: pru support using remoteproc yes, but not anyone is willing to go that way Mar 04 14:46:15 Catslab: beaglelogic.net and pruspeak both did remoteproc. Mar 04 14:47:17 jkridner: I now, but it's another learning curve to port to remote proc from an already exixting project using the old method Mar 04 14:47:36 Catslab: does your patch break remoteproc? Mar 04 14:48:18 <_av500_> what is the issue with remoteproc? Mar 04 14:51:55 jkridner: for pru utilisation yes, I replaced something in dts, for the rest I don't know. Rcn-ee wanted to know what I did to get pru working on ti 3.14 Mar 04 14:52:47 that's just a patch for people wanting to get pru the old way Mar 04 14:54:31 Catslab: thanks Mar 04 14:55:29 jkridner: if you know a good how-to remoteproc, I,ll take it :) Mar 04 14:56:45 Catslab: The code examples of beaglelogic and pruspeak are the main how-to's I know about right now. http://beagleboard.org/pru shows some steps to reproduce pruspeak. Mar 04 14:59:42 ok thanks Mar 04 15:01:07 I bookmarked your github repo and I'll have a look at that later. I first need to get my project runing without remoteproc Mar 04 15:16:14 l;.................................................................+ Mar 04 15:16:19 +--++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Mar 04 15:16:53 *meow* ? Mar 04 15:21:30 catlike typing detected Mar 04 15:24:27 ahahah Mar 04 15:34:32 can anyone tell me how GPIO_INSTANCE_ADDRESS like (SOC_GPIO_1_REGS) and GPIO_INSTANCE_PIN_NUMBER like (23) are inter related ?? writing libraries in starterware.. Mar 04 15:49:58 anyone somewhat familiar with starterware programming ?? Mar 04 15:51:25 geekswine_: I never bothered with starterware for my client's custom boards Mar 04 15:51:55 :( Mar 04 17:20:00 need to learn how to write device tree files for bbb custom cape. where to start? Mar 04 17:20:31 i copied an existing one and made the necessary modifications Mar 04 17:32:51 Rotti: that's the approach i'm following. but feel like i'd need a basic overview about dt syntax and low level hardware management Mar 04 17:33:41 samael: it's nearly two years ago i did it, but i was googling a lot... there have been some blog entries and so on with information... and a lot trial&error Mar 04 17:37:55 there's a device tree compiler thingy rcn-ee has put together .. but there are web tutorials on device-tree Mar 04 17:38:42 indeed, there are a lot of informations around, but for a beginner the difficult part is to choose the right ones. anyway, i guess i need to rtfm a lot Mar 04 17:39:03 fraid so Mar 04 17:39:05 samael, are you targetting v3.8.x, v3.14.x or mainline? Mar 04 17:41:53 rcn-ee: 3.8 for the moment. i'm using capemgr and still too immature about the platform to be productive with 3.14 Mar 04 17:43:16 samael, start with looking thru: https://github.com/beagleboard/linux/tree/3.8/firmware/capes as pretty much every interface has been used at some point.. Once you get that working, we can help you port it to v3.14.x/mainline.. Mar 04 17:43:55 ping panto Mar 04 17:45:20 rcn-ee: thanks a lot Mar 04 17:48:10 i am getting some error implementing opencv on BBB Mar 04 17:48:31 especially when using a webcam Mar 04 17:48:45 can anyone help Mar 04 17:49:44 i am getting famous 'VIDIOC_QUERYMENU: Invalid Argument' error on runtime Mar 04 17:50:43 maybe.. if... you ... asked... in ... an...opencv... channel? Mar 04 17:51:18 program has no compilation error Mar 04 17:51:50 i was following derek molloy's tutorial Mar 04 17:52:38 dikshant, http://linuxtv.org/downloads/v4l-dvb-apis/vidioc-queryctrl.html it's an ioctrl, compare your kernel version with the v4l tools... Mar 04 17:52:49 somethign is probally too old.. Mar 04 17:59:45 Hi Mar 04 18:00:16 Somebody go problem with Kernel 3.14 and GPIO2(P9_22) in AM3358BZCZ? Mar 04 18:00:28 With AM3359ZA is ok with Kernel 3.14 Mar 04 18:01:05 rcn-ee ichecked the link u posted and couldn't understand Mar 04 18:01:31 your point Mar 04 18:07:05 dikshant, "uname -r" mi Mar 04 18:07:08 might he Mar 04 18:07:13 help* on the running system Mar 04 18:07:17 GRR enter key Mar 04 18:09:33 michael can u tell me how coz that command just echoes the version of angstrom i am using Mar 04 18:12:06 perhaps it was 'uname -a' Mar 04 18:12:33 Hello, are all Beagle Bone models currently using this: https://en.wikipedia.org/wiki/Das_U-Boot , as the boot system? Mar 04 18:12:37 eg. mine says "Linux ds-miche2 3.12.21-gentoo-r1 #1 SMP Wed Jul 2 14:25:23 BST 2014 x86_64 AMD A8-6500 APU with Radeon(tm) HD Graphics AuthenticAMD GNU/Linux Mar 04 18:12:37 " Mar 04 18:13:14 kernel version is "3.12.21-gentoo-r1" in my case .. rcn was suggesting you check that vs the instructions you are following Mar 04 18:16:40 Abhishek_, pong Mar 04 18:16:57 Anyone? Mar 04 18:16:58 Hi panto Mar 04 18:17:20 hi Mar 04 18:17:39 This is about establishing a PRU framework Mar 04 18:18:47 If you are aware, there's are projects on upstreaming PRU support this year in GSoC and we wanted you to join the discussion Mar 04 18:19:44 Abhishek_, website, forum links? Mar 04 18:21:16 panto: I mean building upon the work you did on the pru remoteproc driver in 3.8.13 and extending it to 3.14 and then to mainline Mar 04 18:21:36 ok Mar 04 18:21:44 have you figured out what to do with the TI PRU work? Mar 04 18:22:16 Not yet. Mar 04 18:22:57 ok Mar 04 18:23:01 I think they are moving twds rpmsg Mar 04 18:23:24 yeah, but rpmsg is not high performance enough IMO Mar 04 18:23:35 rpmsg sounds very heavy weight Mar 04 18:24:18 panto: here: http://elinux.org/BeagleBoard/GSoC/Ideas#Heterogeneous_co-processor_support_in_open_source_operating_systems_and_libraries Mar 04 18:24:20 yeah, I looked at it when I started Mar 04 18:24:56 the PRUs have such small code space that you got to tread carefully Mar 04 18:24:58 Abhishek_: 1000KHz capture rates! Mar 04 18:25:15 panto: I agree Mar 04 18:25:32 not only that... unrolling loops in the PRU is one way to meet timing for tight stuff Mar 04 18:26:09 I mean the whole framework has to be lightweight, less boilerplate Mar 04 18:26:12 I think the use of rpmsg comes from it's use at TI DSPs Mar 04 18:27:07 s-anna has written the new driver for 3.14 Mar 04 18:27:18 have people thought of NOT treating the PRUs like a processor but rather a hardware blob? Mar 04 18:28:48 I mean, that's the way I looked at the PRUs when I did BeagleLogic, and it did a fantastic job Mar 04 18:29:06 more as in "Software defined peripherals" side Mar 04 18:29:12 yep Mar 04 18:29:16 that was the idea Mar 04 18:29:31 but? Mar 04 18:29:33 but software people have an itch to over-architect everything Mar 04 18:29:58 I see the PRU as an FPGA that's programmed in C and assembly Mar 04 18:30:05 pretty much. Mar 04 18:30:32 in fact, I would very much like to see the PRU use the new FPGA framework Mar 04 18:30:34 you give it some data and let it clock it out independent of the main CPU for example Mar 04 18:30:41 it might be a better fit that the remoteproc one :) Mar 04 18:30:44 docs? Mar 04 18:30:53 google is your friend Mar 04 18:31:42 panto: YES!!!! Mar 04 18:31:47 I really liked the concept of upcalls and downcalls Mar 04 18:31:57 yeah, that is key Mar 04 18:32:20 upcall/downcalls? Mar 04 18:32:38 ds2, we use it for comms between the ARM & PRU Mar 04 18:33:01 instead of bundling a bunch of data in a message you issue RPC calls back and forth Mar 04 18:33:11 oh that blah Mar 04 18:33:13 the trick is that there is no 'message' Mar 04 18:33:29 why not interrupts + "registers" (emulated via the shared mem) Mar 04 18:33:43 There could be both approaches, I thought about this Mar 04 18:33:44 you can't emulate a register with memory Mar 04 18:33:52 there are interrupts ofcourse Mar 04 18:33:59 as in, halting the PRU before writing the RAM? Mar 04 18:34:02 why not? Mar 04 18:34:08 it seems to work for me Mar 04 18:34:17 no halting Mar 04 18:34:19 cause you can't emulate the write to clear behavior Mar 04 18:34:27 Oh that Mar 04 18:34:28 with guaranteed results Mar 04 18:35:00 design around that Mar 04 18:35:16 then it's not a real-register interface Mar 04 18:35:22 i just put interrupt status as a 32bit counter, new interrupt from PRU bumps it. ARM only does reads Mar 04 18:35:34 we do that too Mar 04 18:35:35 in that sense, sure Mar 04 18:35:47 anyway, it's s/w Mar 04 18:35:53 there's more than one way to skin that cat Mar 04 18:35:59 but it isn't much worse then the i2c slave stuff on a tiny MCU Mar 04 18:36:07 STM32? Mar 04 18:36:16 *nod* Mar 04 18:36:25 even i2c master is a pain Mar 04 18:36:27 Abhishek_: NXP but same idea Mar 04 18:37:32 wish the traditional processor guys would toss out more PLD/FPGA stuff in their silicon to get more people to think differently Mar 04 18:38:47 well, standard cores are much much faster than low cost FPGAs Mar 04 18:39:07 and you can integrate them on your die since they use the same process as your big ass cpu core Mar 04 18:39:11 what I really wanted with proposing the PRU framework idea was for someone not to duplicate the effort while developing BeagleLogic Mar 04 18:39:18 *something like Mar 04 18:39:22 or even a small app Mar 04 18:39:36 ADI had some ARM7's w/a PLD on there Mar 04 18:39:56 think some MSPs had them too. Mar 04 18:39:58 That's what I personally liked more about the libprussdrv thingy Mar 04 18:40:12 I could rapidly prototype Mar 04 18:40:40 Abhishek_: the main drawback is libprussdrv didnt expose a kernel version of it Mar 04 18:40:59 is gpio-of-helper usage deprecated? Mar 04 18:41:08 yup, that was the API I wanted to (help) develop Mar 04 18:41:28 so what's the alternative? Mar 04 18:41:36 Abhishek_: it doesn't sound like upstream is too receptive :( Mar 04 18:41:48 yup, I understand that Mar 04 18:41:53 Abhishek_: but if you can get momentum, I'd support that and help Mar 04 18:42:12 codehere1: hmm? Mar 04 18:42:23 That's why the "PRUSS Support for newer kernels" proposition Mar 04 18:43:07 ds2: I wanted to reserve some GPIO in the devicetree but it seems mainline linux does not include gpio-of-helper Mar 04 18:43:34 ds2: I downloaded a patch for gpio-of-helper a while back but I wasn't sure if it was maintained sitll Mar 04 18:44:07 codehere1: I cannot see a proper use case where that is needed. Mar 04 18:44:40 if you want gpio's, request them. Mar 04 18:44:51 reserve makes no sense Mar 04 18:45:19 perhaps I'm missing something Mar 04 18:45:27 you can set up LEDs and gpio-keys in the devicetree Mar 04 18:45:42 I thought the same happened for generic GPIO Mar 04 18:45:57 if you can't see the case for generic GPIO why are gpio-keys special? Mar 04 19:04:18 \ Mar 04 19:04:36 yes but your driver can just request those GPIOs Mar 04 19:04:46 no reason to "reserve" it Mar 04 19:07:16 rcn-ee: got CONFIG_BACKLIGHT_GPIO=m into the debian kernel officially, so LCD7 at least should end up working with Jessie's kernel + custom .dtb Mar 04 19:08:00 still haven't figured out the minimal config diff to get USB working ok Mar 04 19:08:07 (or needed patches) Mar 04 19:08:27 vagrantc, i saw that!! very nice... Mar 04 19:08:55 btw, hopefully the 3.16-clk guys pick up: https://git.kernel.org/cgit/linux/kernel/git/stable/stable-queue.git/tree/queue-3.18/arm-dts-am335x-bone-usb0-is-hardwired-for-peripheral.patch Mar 04 19:09:06 fixes some random reboots on the bones. ;) Mar 04 19:09:48 yeah, the musb is tied behind 3-4 different config labeles... Mar 04 19:12:53 nope missed clk8... Mar 04 19:13:33 nm, it's in the repo, never made it to my gmail list.. Mar 04 19:21:46 rcn-ee: what to look forward to in kernel 4.0 ? Mar 04 19:22:07 kernel patching without reboot! ;) Mar 04 19:22:30 now we know why skynet chose, 4.1.x. ;) Mar 04 19:25:43 is the machines ready to rise? Mar 04 19:26:34 i think they have, look at all the drones... that last 8 minutes... Mar 04 19:27:29 evil cpu_idle.... the am335x pm firmware doesn't like that option at all.. Mar 04 19:27:48 the am335x pm firmware has never worked completely right Mar 04 19:28:47 half tempted to use that M3 for something else Mar 04 19:30:49 it sorta works on v3.14.x-ti... but with cpu_idle + firmware, network transfer speed maxes out at 125Kb/s... Mar 04 19:31:02 with "either" removed, 10Mb/s.. Mar 04 19:33:38 wonder if it is doing something dumb like shutting things down and bring it up on cpuidle Mar 04 19:34:22 that's my guess too, probally fighting either other tooth and nail for control.. Mar 04 19:34:44 they can't fight on the am335x Mar 04 19:35:03 and idel cpu usage is down too.. ;) Mar 04 19:35:07 A8 does a WFI, M3 gets control and does the take down Mar 04 19:35:34 A8 here can't do certain things like on the AM37x/AM35x Mar 04 19:35:53 cpuidle doing a WFI seems like a reasonable thing Mar 04 19:36:25 it's going to have to be eventually fixed in mainline.. i know in v3.14.x they'll just add it to the notes.. Mar 04 19:39:15 is there fully working PM for the 33x in ANY version? Mar 04 19:39:51 it goes into suspend and comes out.. does that count? Mar 04 19:40:44 multiple times? Mar 04 19:42:13 i know better.. i only tested once. ;) Mar 04 19:47:30 :D Mar 04 19:54:09 hey ! whenever i use apt-get command, like in this example, to download v4l-utils, i get this huge list of errors. need help with this please. Mar 04 19:54:12 http://pastebin.com/JWPVCCLE Mar 04 19:56:07 ZeekHuge, element14 screwed up when they "modified" bb.org's images delete: sudo rm /etc/init.d/led_aging.sh and reboot.. Mar 04 19:56:38 or jsut reflash with a non-modified image: http://elinux.org/Beagleboard:BeagleBoneBlack_Debian#Debian_Image_Testing_Snapshots Mar 04 19:56:47 rcn-ee: The PM firmware sources are available? Mar 04 19:57:11 yeap: http://git.ti.com/gitweb/?p=ti-cm3-pm-firmware/amx3-cm3.git;a=commit;h=d85e6a17da1025adac9ab7f97dd7aa2a58c517ce Mar 04 19:57:32 better link: http://git.ti.com/gitweb/?p=ti-cm3-pm-firmware/amx3-cm3.git;a=tree;h=refs/heads/ti-v3.14.y;hb=refs/heads/ti-v3.14.y Mar 04 19:58:30 but as i remember, it didt show up, earlier, for the same command . just one day, it suddenly came up . Mar 04 19:59:03 it's always been there, bunch of threads on beagleboard google group... Mar 04 19:59:19 ok, i'll delete that file. Mar 04 19:59:23 thank you ! Mar 04 20:33:59 Hi there, I have a BBB and I just replaced Angstrom with the latest Debian image and now I need to mount the micro sd card into my file system so I can install programs. How do I do that? Mar 04 20:35:34 sudo mount /dev/mmcblk** ./ Mar 04 20:38:03 thanks for that - what's ? Mar 04 20:38:22 the you chose to mount it too... Mar 04 20:39:07 okay, should I erase my micro sd first? and will apt-get use the micro sd to install new software? Mar 04 20:40:10 that's not trivial for your average user, either use the eMMC as is, or boot with a bigger microSD card. Mar 04 20:43:21 rcn-ee: in the old ubuntu image based on 3.8 we had /sys/class/i2c-adapter/i2c-1/new_device Mar 04 20:43:34 in jessie (8.0) its missing Mar 04 20:43:43 what to do to get it back? Mar 04 20:44:00 it should be there. (not sure if i2c-1 is enabled by default) Mar 04 20:44:32 I'm booting off the 2 gigs on the bbb itself - maybe mount my micro sd as /opt? Mar 04 20:45:10 I only see i2c-0 Mar 04 20:45:21 I only see i2c-1 back? Mar 04 20:45:24 cityLights, uname -r? Mar 04 20:45:28 I need it for RTC Mar 04 20:45:43 3.19.0-bone3 Mar 04 20:47:32 cityLights, yeah... https://github.com/RobertCNelson/linux-stable-rcn-ee/blob/3.19-bone3/arch/arm/boot/dts/am335x-boneblack.dts i didn't add the cape eeprom bus by default yet.. Mar 04 20:48:05 right, saw it in rc3 , but didn't want to bother you then Mar 04 20:48:12 now that 3.19 is out... Mar 04 20:48:17 cityLights, if you add this include: https://github.com/RobertCNelson/linux-stable-rcn-ee/blob/3.19-bone3/arch/arm/boot/dts/am335x-cape-rtc-ds1307.dtsi Mar 04 20:52:14 is the included dtc on the image good, or need I run your script to build the compiler from source? Mar 04 20:52:46 sorry, I have to build the dtc Mar 04 20:53:18 i don't remember. ;) it's probally best to grab the 3.19 branch of dtb-rebuilder and just add that *.dtsi to the main beaglebone black dts. .. :) Mar 04 20:54:05 if you need the i21 bits: grab it from here: https://github.com/RobertCNelson/linux-stable-rcn-ee/blob/3.19-bone3/arch/arm/boot/dts/am335x-cape-bbb-exp-c.dtsi#L48 Mar 04 20:56:45 What if I create /home, /usr and /opt on my microsd and mount them with fstab on every boot? would that work? Mar 04 20:56:47 I just need to patch the am335x-boneblack.dts to include am335x-cape-rtc-ds1307.dtsi Mar 04 20:56:55 then build it, right Mar 04 20:57:30 cityLights, correct... and grab the i2c1 bits if you want that bus too enabled.. Mar 04 20:57:45 twoten, and /var/ & /lib Mar 04 20:58:08 by then, you might as well run off the microSD and add eMMC as opt ;) Mar 04 20:58:34 I am so grateful your are here rcn Mar 04 21:00:54 I foresee a lot of resizing if I put 5 system partitions on the microsd, what is your advice since my ideas suck? Mar 04 21:02:04 use the "standalone" microSD image... Mar 04 21:03:12 okay, and I can wipe debian off the emmc and use it for /opt? Mar 04 21:04:36 sure, or whatever else you want to use for it Mar 04 21:05:00 alright that's a plan, thank you very much! Mar 04 21:06:34 oh and one more thing, what's the biggest microsd card the bbb can take? Mar 04 21:08:45 32gb microSD Mar 04 21:09:05 (sdhc) Mar 04 21:09:18 thanks very much! Mar 04 21:17:21 $ /usr/local/bin/dtc -O dtb -o am335x-boneblack.dtbo -b 0 -@ am335x-boneblack.dts Mar 04 21:17:22 Error: am335x-boneblack.dts:10.1-9 syntax error Mar 04 21:18:29 if i completely zero out the eMMC, i haven't had good luck loading u-boot off of the SD card... Mar 04 21:26:07 vagrantc, even holding down the uSD switch? that sounds like an issue .. Mar 04 21:26:47 cityLights, use the make scripts in the dtb-rebuilder: http://elinux.org/Beagleboard:Capes_3.8_to_3.14#Custom_dtb (theres a 3.19 branch) the #includes make dtc have a fun time.. Mar 04 21:27:20 right, sorry , forgot Mar 04 21:27:34 stt_michael: ah, maybe i didn't try that. Mar 04 21:28:09 i do recall it working when u-boot was files on a partition, but had trouble when it was dd'ed to the uSD Mar 04 21:28:28 i'll have to run through it agian Mar 04 21:28:30 again Mar 04 21:28:59 * vagrantc forsees many BBB reboots in the near future tweaking the kernel... Mar 04 21:29:35 vagrantc .. my understanding from rcn-ee was the config on the eMMC u-boot tried uSD before eMMC anyway .. hence why it 'appears' to magically work without the boot select switch .. but if you zero out the emmc .. it's not going to ever 'see' that .. and relies on the actual hardware boot jumpers Mar 04 21:29:54 d'oh lol Mar 04 21:30:00 I found connman is far better then network-manager in keeping the connection up even when the wifi gateway powers off Mar 04 21:30:30 stt_michael, it's complex. ;) bootrom looks 1: eMMC 2: microsd Mar 04 21:30:45 u-boot: (even when in eMMC) looks to microsd first for "uenv.txt".. Mar 04 21:32:15 d'oh Mar 04 21:32:43 rcn-ee, where does the boot switch affect, then? Mar 04 21:33:39 the boot switch changes the bootrom order... (omap's usually have around 8 io pins, depending high/low, they tweak the 'order') Mar 04 21:34:04