**** BEGIN LOGGING AT Wed Jan 27 02:59:59 2016 Jan 27 03:32:33 leds continue flashing? (including the recognizable "heartbeat" led) Jan 27 03:33:00 if so, then you've just booted into a running system and not a flasher Jan 27 03:33:59 to force booting from SD card, be sure to hold the "S2" button (the one closest to the card slot) while applying power Jan 27 03:34:22 as soon as the power led is on you can release it Jan 27 03:38:14 filt3r: really, I want to strangle whoever made McASP's interrupt-enable bits mismatch its actual interrupt bits -.- Jan 27 03:38:35 :D Jan 27 03:41:16 but ok, on baremetal my little loopback test works fine, even with no dma/fifo (only 2 samples in circulation) at 192 kHz Jan 27 03:42:10 zmatt, thanks for the reply will try that now Jan 27 03:43:07 baremetal, without linux? Jan 27 03:43:45 "no OS" is I think the commonly accepted definition of baremetal :P Jan 27 03:44:03 ^^ Jan 27 03:44:41 and using McASP without dma is not going to work well on linux... something about interrupt latency :P Jan 27 03:44:57 Bare metal does not require the absence of an OS, just an OS that can GTFO of the way and let a privileged user mode app get at the hardware. Jan 27 03:44:58 ye true Jan 27 03:45:13 zmatt, no joy. and yes i have never seen the heartbeat one, think that is supposed to show up on led3 but nothing there Jan 27 03:45:41 Micro|MTW: what image did you flash onto the card? Jan 27 03:46:33 zmatt, Debian 7.9 (BeagleBone, BeagleBone Black, Seeed BeagleBone Green - 4GB SD) 2015-11-12 Jan 27 03:46:42 the one linked from the getting started page Jan 27 03:47:09 CathyBlue: hmm, I can get at the hardware just fine from linux userspace, but as long as linux is running it can still get in your way in numerous ways, even if merely by causing irq latency by disabling irqs often and for too long Jan 27 03:47:25 there is a note about removing a # on a command line but that line doesnt exist. i tried adding it as well but also no luck Jan 27 03:48:00 Micro|MTW: get an actual flasher image, I'd recommend the latest jessie snapshot from http://elinux.org/Beagleboard:BeagleBoneBlack_Debian Jan 27 03:48:18 will give it a go, thanks Jan 27 03:48:45 Micro|MTW: you can pick either the 4gb or 2gb lxqt flasher... Jan 27 03:49:30 CathyBlue: an OS that "can GTFO of the way" to such a degree isn't an OS but a library Jan 27 03:50:00 ;) Jan 27 03:51:29 * zmatt stares a bit mesmerized at the scope... Jan 27 03:51:51 (I increment the input sample every time I feed it back as output) Jan 27 04:19:25 zmatt, got a heartbeat with the 2g version. the 4g one is not online currently. seems to be flashing, thanks for the help Jan 27 04:20:11 ehh, you should *not* be getting a heartbeat while it's flashing Jan 27 04:21:33 ... the lights are acting different for the first time and led0 is blinking what seems to be a heartbeat and one of the others is on dim intermittently Jan 27 04:21:53 what file did you download exactly? Jan 27 04:22:35 since the led behaviour you're describing is again just a running system, not a flasher Jan 27 04:22:40 https://rcn-ee.com/rootfs/bb.org/testing/2016-01-24/lxqt-2gb/BBB-eMMC-flasher-debian-8.3-lxqt-2gb-armhf-2016-01-24-2gb.img.xz Jan 27 04:23:04 weird Jan 27 04:23:22 the leds are def not acting as they did Jan 27 04:24:09 there might be some difference (especially since the latest images for some reason use an rt kernel which I think doesn't support the cpu-activity led), but a heartbeat means a running system Jan 27 04:24:30 the 2g image doesnt seem to have the windows partition stuff is that correct? the card is not readable after using win32 disk to program it Jan 27 04:24:55 it's a single ext4 partition Jan 27 04:25:25 so it's correct it can't read the partition (though it should still understand the partition map) Jan 27 04:26:41 there's an easy way to make sure you did the boot-button thing right: try it *without* card inserted, it should then not boot at all (only power led on) Jan 27 04:27:11 you can then insert the card and it should start the flasher Jan 27 04:27:53 kk, will try that real quick. anything i need to do if it is really in a write cycle ? Jan 27 04:28:12 you're overwriting eMMC anyway, so no doesn't really matter Jan 27 04:28:45 kk Jan 27 04:31:05 that worked much better. got the blue led cylon pattern running like on my 3d printer Jan 27 04:32:05 ok, so the button probably wasn't pressed properly the previous attempt or something like that Jan 27 04:32:10 it can be a bit fiddly Jan 27 04:32:39 now it's flashing, when it's done all leds stay lit I think Jan 27 04:32:50 that is what i read Jan 27 04:33:06 again thanks for the help Jan 27 04:33:20 yeah for me it's just hearsay, I've never run a flasher myself, I just flash directly via USB Jan 27 04:33:47 but the utility for that is too brittle to seriously recommend for general use Jan 27 04:34:25 (also, only runs on linux, not windows) Jan 27 04:34:53 shame. fiddly is right about the button. eww, yeah i need to work on the fiddly :) i cant even spell linux correctly yet Jan 27 04:36:09 yeah, I have an advanced button-pressing-tool for that... you can see it near the bottomleft in this pic -> http://gerbil.xs4all.nl/bbb-pmic-testing.jpg Jan 27 04:37:23 :) to pricey for me Jan 27 04:38:14 (I've since then put a small bit of colored heat-shrink tubing on it to make it look like an object with a purpose and not a random piece of garbage people would keep throwing away) Jan 27 04:38:56 right now i cant even see the top of my desk been working on to dam many projects and don't seem to be able to finish any Jan 27 04:40:30 tip: bigger desk Jan 27 04:40:33 ;) Jan 27 04:43:23 started playing with mcu about 30 years ago with a motorola 68701 the learning curve on this one is steeper than others ive played with over the years Jan 27 04:44:46 well the programming seems about par but the layers on top of it to make it work has been a pain Jan 27 04:44:49 yeah the am335x is a funny thing, being sort of a cross between a microcontroller and a smartphone SoC Jan 27 04:45:08 layers-- Jan 27 04:45:26 * zmatt generally just removes layers whenever they're being annoying Jan 27 04:46:46 figures the 4g version of jessie is now working. oh well will give me practice fiddling Jan 27 04:47:39 Micro|MTW: btw, unless you really need an RT kernel I suggest installing a normal one Jan 27 04:48:03 (assuming rcn is still shipping images with RT kernels) Jan 27 04:50:31 iffy if i have to. right now the plan is to offload some critical timing for some sensors off to a dedicated atmel and kick the data back via i2c Jan 27 04:50:59 how critical is "critical timing" ? Jan 27 04:51:07 (btw, don't forget there's also PRU) Jan 27 04:52:06 the pru should be able to handle it but unless i read something wrong i will need to drop to asm to properly access it unless its some basic stuff, or have things changed since that article was written Jan 27 04:52:16 there's a C compiler Jan 27 04:53:14 k, so full access to the registers with C but not the other common high level languages like python then Jan 27 04:53:27 pru is often programmed in assembly because that lets you create extremely accurate timing simply by counting instructions Jan 27 04:53:41 are you running python on the atmel processor? :P Jan 27 04:54:41 nada, and if i can avoid learning another language i am pretty happy :) i am getting old and to the point that i fear learning new stuff i might forget something important like how to piss Jan 27 04:55:48 yeah so the C compiler is fine to use the PRU cores more-or-less as two on-chip microcontrollers Jan 27 04:56:17 if you need 5ns-accurate timing then you'll need to drop into assembly.... probably regardless of what processor you're using Jan 27 04:58:24 i might get lucky then Jan 27 05:35:14 i dont know to run my application on BBB boots up pls help me Jan 27 05:45:42 what is your application? Jan 27 05:58:13 Where can I read the description of the register C24? ( it's PRU0/1 Local Data) Jan 27 06:10:33 http://www.ti.com/product/am3358 under tech docs Jan 27 06:13:05 Sure! But I was interested in a more detailed description Jan 27 06:14:28 I reading pru_refering_guide, section 5.2.1 and using the table 9 Jan 27 06:16:49 but i can't find how this register (24) works. Exactly, how Data RAM0 is filled? Jan 27 06:18:34 i haven't JTAG emu, because i want to calculate it yourself Jan 27 06:21:47 zmatt, veremit: it works Jan 27 06:21:57 and filt3r Jan 27 06:22:41 \o/ Jan 27 06:22:51 | Jan 27 06:23:01 / \ Jan 27 06:23:19 \o/ Jan 27 06:23:25 \o/ Jan 27 06:23:28 | Jan 27 06:23:33 _/ \_ Jan 27 06:28:21 snowSTAFF: did you look at the ARM Cortex A8 TRM and/or the ARM Architecture Reference Manual? Jan 27 06:28:41 or the TI AM335x TRM? Jan 27 06:29:10 * yates buys zmatt, veremit, and filt3r a Guiness Stout Jan 27 06:30:54 i deleted few packages from my beaglebone black Jan 27 06:31:18 and now when i am connecting it througf usb it is not getting detected Jan 27 06:31:26 i am unable to reboot it Jan 27 06:31:59 tha packages which i deleted are as follows Jan 27 06:32:13 kanishka_: have you tried ssh through the wired network ? Jan 27 06:32:41 you can always just reflash, too Jan 27 06:33:50 the manners of folks... Jan 27 06:35:14 yates, yes, i did, but only am335x Sitara Processors TRM Jan 27 06:36:29 apache2.2-common apache2-mpm-worker dbus dbus-x11 hostapd openssh-server ppp screen sudo Jan 27 06:36:50 reflashing is not working Jan 27 06:37:06 my board is not at all getting detected by my laptop Jan 27 06:37:09 kanishka_: are you flashing the sdcard? Jan 27 06:37:25 or the internal mmc? Jan 27 06:37:26 yeah i am flashing the board through sd card Jan 27 06:37:42 are you holding the boot switch when you power on? Jan 27 06:37:47 yeah Jan 27 06:38:06 are your certain your flashing procedure is correct and your sdcard is good? Jan 27 06:38:33 yes the procedure is correct and sd card is also good Jan 27 06:38:42 then you have a hardware failure Jan 27 06:38:43 the boot button can be " fiddly" :) learned this just tonight Jan 27 06:38:44 i can say dat bcoz i have two boards Jan 27 06:38:55 and the otherone got booted up nicely Jan 27 06:39:10 oh really? the sdcard works on one board and not the other? Jan 27 06:39:39 this particular board was working fine before i deleted these package Jan 27 06:39:43 apache2.2-common apache2-mpm-worker dbus dbus-x11 hostapd openssh-server ppp screen sudo Jan 27 06:39:55 the packages have nothing to do with it if you've reflashed Jan 27 06:40:07 okey Jan 27 06:40:16 the sdcard works on one board and not the other? Jan 27 06:40:17 i am new to this thing Jan 27 06:40:34 yeah it worked fine for the othee new board Jan 27 06:40:46 it got flashed properly and is workin fine Jan 27 06:40:56 mb using TTL boot manualy? Jan 27 06:40:59 any leds lighting up? Jan 27 06:41:01 and yet it won't work on your first board? Jan 27 06:41:13 the exact same sdcard? Jan 27 06:41:18 yeah all the leds are lighting up Jan 27 06:41:25 yeah Jan 27 06:41:43 the leds are working all fine as if the board is booting Jan 27 06:42:04 how do you know it isn't booting then? Jan 27 06:42:05 but after 1 hour of led blinking when i m removing the sd card Jan 27 06:42:15 i m getting no results Jan 27 06:42:38 can you ssh in via 192.168.7.2? Jan 27 06:42:41 my laptop is not detecting the boardwhile the otherboard is working just fine Jan 27 06:42:46 nope Jan 27 06:42:52 i cannot do that also Jan 27 06:43:13 did you unplug (the usb) the second board while trying the first board? Jan 27 06:43:17 and to add a little information i have two board with similar problem Jan 27 06:43:31 yeah Jan 27 06:43:50 i left the first board thinking it has been damaged Jan 27 06:43:52 which program did you flash? Jan 27 06:44:01 debian Jan 27 06:44:07 image Jan 27 06:44:14 kanishka_, do the boot this way. worked for me. remove power. remove the sd card. hold down the boot button while you power up. then install the SD card and let go of the boot button Jan 27 06:44:34 okey i will try it Jan 27 06:44:39 thank you so much Jan 27 06:44:54 also how are you powering it Jan 27 06:44:57 I am new to this thing so have a little less knowledge Jan 27 06:45:02 thanks for helping Jan 27 06:45:18 are you powering via the usb? Jan 27 06:45:27 yes Jan 27 06:45:29 don't use USB power Jan 27 06:45:35 transformer Jan 27 06:45:46 okey!!! Jan 27 06:46:11 snowSTAFF: what is the name of the register? Jan 27 06:47:12 kanishka_: or you might try plugging in the usb on a different computer Jan 27 06:47:19 or try a different cable Jan 27 06:47:28 I dont have knowledge about registers Jan 27 06:47:44 but the same set up is working well for the other board Jan 27 06:47:46 kanishka_: i was tlking to snowSTAFF Jan 27 06:47:58 okey Jan 27 06:48:27 yates: PRU0/1 Local Data Jan 27 06:48:32 kanishka_: if the same cable/usb port/sdcard works on one board and not another, well, i'd say you have a bad board. Jan 27 06:48:57 I'd still get a uart cable and investigate it properly Jan 27 06:49:14 kanishka_: are the lights coming on solid but never blink? Jan 27 06:50:19 no they are blinking Jan 27 06:50:25 as if it is booting up Jan 27 06:50:50 so you're getting through u-boot. Jan 27 06:51:04 yeah Jan 27 06:51:30 and you've tried connecting both through 192.168.7.2 and the eth port (wired)? Jan 27 06:51:48 are you sure you got the right ip address for the wired port? (dhcp) Jan 27 06:52:12 yeah the ip is ri8 Jan 27 06:52:23 but i have not tried through eth port Jan 27 06:52:37 tbr: yeah, hooking up a serial cable would be a good idea at this point. you can login through it Jan 27 06:52:57 okey i will try that also Jan 27 06:53:16 ri8 now little short of resources but will do it soon Jan 27 06:53:26 http://www.ftdichip.com/Support/Documents/DataSheets/Cables/DS_TTL-232R_CABLES.pdf Jan 27 06:53:52 kanishka_: are you using linux on your laptop? Jan 27 06:54:01 doesn't matter, nm. Jan 27 06:54:23 nope i am using it through putty Jan 27 06:55:21 snowSTAFF: see section 4.5 of the ti trm. it summarizes all the pru registers. Jan 27 06:55:34 or heck, the whole of chapter 4 Jan 27 06:56:48 i don't like relying on the rndis / 192.168.7.2 interface. better to try it though the wired eth network Jan 27 06:57:04 okey Jan 27 06:57:34 can you find out on your router which dhcp addresses are assigned? Jan 27 06:57:56 also, are you connecting two boards at the same time to your pc? Jan 27 06:58:21 she said no to that earlier. Jan 27 06:58:26 ok Jan 27 06:58:36 would be good to confirm, though Jan 27 07:00:54 no i donno the dhcp either Jan 27 07:02:43 get a uart cable Jan 27 07:02:57 okey Jan 27 07:04:00 i have a uart cable Jan 27 07:05:27 yates: I think that i found, but it's not exactly what i want. Anyways, i got it a little for start. Jan 27 07:19:37 What makes command ST32? Jan 27 08:00:26 How to determine the processor version of ones BeagleBone Black (the x in AM335x) ? Jan 27 08:41:27 How to determine the processor version of ones BeagleBone Black (the x in AM335x) ? Jan 27 08:44:58 it's printed on the processor, and it's an AM3358 in all except the earliest batches Jan 27 08:45:34 Ahh, obviously.. Thanks a bunch! :) Jan 27 09:14:55 hi guys. i am back. well we were able to connect BBB with windows via SSH. but in ubuntu 14.04 it is not connecting. we tried giving dhcp and so on. but nothing happend. it just tries to look for connection but never actually create an network connection Jan 27 09:15:21 i thought it is because of faulty usb cable but it is working fine in windows Jan 27 09:16:14 also we try to run start.html it shows step 1 as green and step 2 and 3 are orange. step is driver installation which is not required with ubuntu Jan 27 09:16:33 instead we ran bash file provided but still same problem Jan 27 09:44:46 how working assembly code with C project? Each string C connected with full execution of the code assembler? Jan 27 14:04:24 hi, so everytime I start up my BBB the ip address in /etc/resolv.conf is back at 127.0.0.1 and I have to change it to 192.168.1.1 Jan 27 14:04:46 how can I keep it at my router's address? Jan 27 14:06:09 probably the image runs a local resolver that gets its configuration via dhcp Jan 27 14:42:39 Hi, using kernel 4.4.0-RC8, cpufreq-info is not working and cpufreq folder in /sys/devices/system/cpu/cpu0/cpufreq is not.. Any idea on which kernel config I missed during my update ? (note that CPU_FREQ=y and CPU_FREQ_GOV_PERFORMANCE) Jan 27 14:49:48 Hi, can someone help about restoring BBB to factory settings? Jan 27 15:12:07 Nenad: just get a debian flasher image Jan 27 15:16:14 aand gone Jan 27 15:45:10 Hello Jan 27 15:45:30 Good afternoon, I speak with? Jan 27 17:19:44 Hi Greetings of the day Respected Sir I am Kiran Jot Singh. I work as assistant professor at Chandigarh University,Mohali Punjab, India. I have been writing an article on IOT and embedded platforms.I would like to use images of Beagleboard boards in that article. So I seek your kind permission for the same. Kindly let me know. Watling for a swift reply from your side. Regards Kiran Jot Singh Jan 27 17:23:53 kiran_, I believe the people you wish to contact can be found at the bottom of this page. http://beagleboard.org/about Jan 27 17:27:14 thanks Jan 27 18:12:51 rcn-ee: hi there, help. new device tree is hard to grasp Jan 27 18:14:07 citylight2, where are you having issues? Jan 27 18:14:30 I wrote to the forum with the subject How to use current universal cape device tree and ADC at the same time Jan 27 18:14:37 http://beagleboard.org/Community/Forums Jan 27 18:14:47 http://beagleboard.org/Community/Forums?place=msg%2Fbeagleboard%2FroMzJwLgdF8%2FcAFCyzaCAQAJ Jan 27 18:15:42 please explain what you ment Jan 27 18:15:58 citylight2, upgrade your overlay repo: cd /opt/source/bb.org-overlays/ ; git pull ; ./install.sh Jan 27 18:16:34 P8_43 was a conflict... that i had fixed. ;) Jan 27 18:17:09 its making dtbo files Jan 27 18:17:38 please answer in the forum so others may know Jan 27 18:18:13 it'll fix P8_43, but the adc cape and universal will conflict.. Jan 27 18:18:28 you should be ale to just do config-pin <> adc Jan 27 18:18:58 after the universal cape is loaded after booting? Jan 27 18:19:03 yeah Jan 27 18:19:06 oa need it load before? Jan 27 18:19:09 o ok Jan 27 18:19:32 (didnt' know you were using cape universal, so i went the bb-adc overlay route. ;) ) Jan 27 18:19:48 right Jan 27 18:20:02 do you post release notes? Jan 27 18:21:17 can I use the pru to write the adc to a memory and read it? Jan 27 18:21:38 you'd have to ask the pru guys. ;) Jan 27 18:30:17 hi, I've seen that the beaglebone black has these realtime PRU processors integrated. TI claims these are suitable for industrial ethernet applications, so my question would be if the ethernet port on the bb black is accessible from the PRUs? Jan 27 18:44:14 caytchen, http://www.ti.com/tool/tidep0001 Jan 27 18:46:36 caytchen: it doesn't appear it is. see tables 4-7 and 4-8 in the AM335x TRM. Jan 27 18:50:06 mhm yeah I've looked over the schematics/references and it seems the only parts of the PRU exposed are a bunch of gpios Jan 27 18:50:59 they may do the ethercat stuff by passing data via shared memory. Jan 27 18:51:20 i don't know, though - just guessing. Jan 27 18:51:36 sure, but that voids all of the timing guarantees once it has to pass through linux Jan 27 18:52:25 rcn-ee: these look nice but they're lacking the HDMI and consumer stuff thats on the black Jan 27 18:53:23 the idea was to run android but still have the PRUs for the timing critical ethernet stuff Jan 27 18:54:00 caytchen: precisely which "industrial ethernet" did you want to run? Jan 27 18:55:14 I was looking at PROFINET but I think the precise variant doesn't matter too much, it's the low-depth access to the eth phy that matters Jan 27 18:58:18 caytchen: there is none, hence you are selecting the wrong processor for this application. Jan 27 18:58:53 none other than the standard linux device driver, that is. Jan 27 18:59:42 e.g., you mean you want to support the irt profile? Jan 27 19:02:31 IRT would be ideal, but thats of course precisely where you need the dedicated ethernet connection and processor. Jan 27 19:03:53 theres probably enough brunt in the ARM to get the lower RT levels working okay with just the normal ethernet, but the jitter/response requirements on IRT are insanse Jan 27 19:04:35 there are just not many companies out there that couple a consumer ARM with HDMI and all the gadgets _and_ something like the PRU :o Jan 27 19:06:50 perhaps it could be done on the am335x, but you'd probably have to write your own PHY driver.. do time-critical stuff in interrupt handlers.. Jan 27 19:07:13 probably no small task.. Jan 27 19:07:59 just guessing,though. Jan 27 19:08:27 zmatt may have a more informed opinion. Jan 27 19:09:21 rcn-ee: i got your custom linux kernel build running on our custom board. woo-hoo! Jan 27 19:09:36 thanks for that site - helped tremendously. Jan 27 19:09:42 yates, good job! ;) Jan 27 19:09:46 thank you. Jan 27 19:10:01 let me know if you want me to carry your custom *.dts ;) Jan 27 19:10:20 you mean integrate it into the tree? Jan 27 19:10:51 it doesn't really exist right now - i just hacked the am335x-boneblack.dts and family... Jan 27 19:10:55 yeah, i've got a few other custom board *.dts. (some are still not for sale) Jan 27 19:11:25 thank you for the offer, i will consider it when i get things more organized and if our project manager is agreeable. Jan 27 19:12:34 * zmatt is almost here and will read scrollback in a moment to see if he has an informed opinion Jan 27 19:18:45 caytchen: I don't know how low-level access you need to the PHY exactly... you can have PRUSS handle all ethernet traffic and pass non-real-time stuff to linux via some mechanism Jan 27 19:19:24 all four irqs from the ethernet switch subsystem are available in PRUSS Jan 27 19:21:47 it doesn't get you ethercat-like low level access... but for that you need a phy on different pins, some of which conflict with LCDC. moreover you need an AM3359 instead of an AM3358, which was only placed on some early versions of the BBB. Jan 27 19:22:57 and the ethercat functionality of PRUSS has no real public docs... it seems TI prefers to deliver ready-made solutions there rather than having people brew their own stuff (and then pesking tech support with their issues :P ) Jan 27 19:28:25 caytchen: how insane are these jitter/response requirements exactly? Jan 27 19:29:11 I don't have any experience with them, but I know there are various solutions for co-hosting linux and an RTOS on the cortex-a8 (e.g. xenomai) Jan 27 19:30:36 IRT has <1ms response and 1us jitter requirements Jan 27 19:31:33 1ms is plenty of time. max 1 us jitter is a bigger problem Jan 27 19:31:36 so basically a context switch ruins it Jan 27 19:34:32 and this is about recognizing certain packets and replying to them in some fashion? Jan 27 19:35:08 (I know virtually nothing about any of these industrial ethernet flavors) Jan 27 19:36:15 recognizing and replying yes, but also sending with correct timing Jan 27 19:36:55 which dts is yates and rcn-ee reffering to? Jan 27 19:37:01 would using the ieee 1588 timers on the phy work for the timing? Jan 27 19:37:18 ddrown: those only perform timestamping Jan 27 19:37:23 right Jan 27 19:37:30 you get your timestamp after the fact Jan 27 19:37:43 don't know if that works for this application or not Jan 27 19:37:44 citylight2, he has a custom board Jan 27 19:38:31 but sending with right timing is not hard in principle... e.g. have PRU submit the packets at the appropriate moment Jan 27 19:38:32 do you have a question in ee wiki to tell us how to load the ADC? Jan 27 19:39:35 rcn-ee: I rebooted after running ./install and /sys/devices/platform/bone_capemgr/slots is empty Jan 27 19:39:56 debian@boiler-bone:~$ sudo config-pin overlay cape-universala Jan 27 19:39:57 Loading cape-universala overlay Jan 27 19:39:57 bash: line 0: echo: write error: Invalid argument Jan 27 19:40:06 Error loading device tree overlay file: cape-universala Jan 27 19:41:39 citylight2, in /boot/uEnv.txt find, cmdline=.... and add "cape_universal=enable" let the boot script load the corrct "overlay" file:https://github.com/RobertCNelson/boot-scripts/blob/master/boot/am335x_evm.sh#L227-L303 Jan 27 19:41:39 having PRUSS inspect incoming packets is slightly trickier since you need to be careful you don't spend ages waiting for reads of distant memory. ideally ethernet DMA would be set up to deliver packets directly into PRUSS shared memory (assuming it is able to, need to verify), which allows PRU to quickly inspect it. if the packet data is needed but not urgently then quickly ship it out using EDMA to free u Jan 27 19:41:45 p the limited space for new packets Jan 27 19:41:48 citylight2, as there are a half dozen versions.. Jan 27 19:42:30 getting all this to work would be non-trivial Jan 27 19:42:50 rcn-ee I rather do it in my service , if I can Jan 27 19:44:14 citylight2, then pick the one that matches your configuration: https://github.com/RobertCNelson/boot-scripts/blob/master/boot/am335x_evm.sh#L258-L291 Jan 27 19:44:23 and load that "${overlay}" Jan 27 19:45:08 am335x-bonegreen-overlay for my beaglebone green Jan 27 19:45:19 zmatt, you mentioned the ethernet irqs can be mapped to the PRUs, is there some place to read up on that? Jan 27 19:45:27 citylight2, "univ-all" ;) Jan 27 19:45:54 so I need to run sudo config-pin overlay univ-all? Jan 27 19:45:56 well not so much "can be mapped" ... they are received there by default (in addition to the cortex-a8) Jan 27 19:46:03 citylight2, correcto.. Jan 27 19:46:16 SWEET Jan 27 19:46:28 and thanks again robert Jan 27 19:47:22 caytchen: https://goo.gl/O64hMr Jan 27 19:47:24 I can even name my wife who came out of a warm shower, heated using a beaglebone green and a calender Jan 27 19:47:45 citylight2, if i ever find time, i want to split it up, univ-most (aka all but hdmi/audio/emmc) univ-hdmi univ-audio, univ-emmc then we can progressivly load them safely.. Jan 27 19:48:19 caytchen: oh and according to the TRM ethernet DMA can't deliver directly into PRUSS Jan 27 19:50:57 so second best option would be OCMC... and if the cost of having PRU perform reads outside the subsystem (to inspect the packets) becomes too high, have EDMA copy the packet data (or whatever portion you're interested in) into PRUSS Jan 27 19:52:33 can anyone here help me use the ADC to measure a AC peak to peak and convert to RMS? Jan 27 19:52:58 I can indeed do a simple pooling and find the MAX value Jan 27 19:53:15 is there a bteer way? Jan 27 19:53:28 yes, calculate the RMS :P Jan 27 19:54:14 I want to output what current is running now, so if it drops to zero, I can sound an alarm Jan 27 19:57:09 in any case, thanks very much for your help zmatt! has given me much to think about Jan 27 20:00:22 caytchen: you're welcome! Jan 27 20:00:47 it would definitely not be an easy project Jan 27 20:01:40 especially since if you have timing-critical traffic going over the same interconnects as linux fluff you'll also have some SoC-level performance analysis/tuning to do, e.g. set initiator priorities Jan 27 21:54:30 rcn-ee: something seems very hosed on this 4.4.0-bone2 Jan 27 21:54:36 in the packages Jan 27 21:54:43 apt-get emacs returns nothing.. Jan 27 21:54:50 apt-get install emacs, that is Jan 27 21:55:05 yates-on-wifes-c, "sudo apt-get update ; sudo apt-get install emac" ? Jan 27 22:05:04 i just tried to upgrade my kernel with apt-get install linux-kernel-4.4.0-ti-r5 and then rebooted and got nothing. how can I revert to the previous kernel? I am booted off an SD card right now, but I don't know what files to delete/replace Jan 27 22:05:27 clockman, "uname_r" in /boot/uEnv.txt Jan 27 22:05:40 well, shoot, i tried that. no luck Jan 27 22:06:19 i'll try again to make sure it saved... Jan 27 22:06:44 clockman, do you have a usb-serial adapter to see what's going on during bootup? Jan 27 22:07:10 ooooo, good call i'll do that Jan 27 22:27:55 Sooo.... this is what I get: http://pastebin.com/WsnhKXxr Jan 27 22:27:59 not very helpful Jan 27 22:30:47 rcn-ee: doh! yeah, that did it. Jan 27 22:32:18 clockman, something else is going on, it's not even getting to the u-boot prompt... Jan 27 22:32:33 power supply? Jan 28 00:46:52 I can't for the life of me get Arch ARM to work on BBB **** ENDING LOGGING AT Thu Jan 28 02:59:58 2016