**** BEGIN LOGGING AT Mon Mar 14 02:59:58 2016 Mar 14 03:01:13 even dynamic filtering would already be very useful... like, I said, to have the option to only show meta-messages (join/part/quit) of a person if there's also a chatline within the last screenful or so Mar 14 03:09:27 many clients you can turn messages on/off but you'd have to write a clever module to adjust content .. I suspect it probably wouldn't work with most text 'blocks' Mar 14 03:11:30 irssi has pretty fancy scriptability support and event hooks all over the place Mar 14 03:12:02 just... of all the places, just not the right ones for this purpose :/ Mar 14 03:25:09 o,O Mar 14 03:28:12 ping zmatt jkridner :P Mar 14 08:54:43 Hi, everyone. I have some trouble with booting beaglebone black via UART. I've installe Processor SDK and followed instructions. When I see "CCC" in my minicom terminal firstly I send u-boot-spl.bin and its OK, secondly I send u-boot.img through y-modem and when transaction is complete I see "Retry 0: nak on sector" and nothing happens. Can somebody explain, what should I do? Mar 14 08:55:18 installed* Mar 14 09:18:20 try asking the u-boot people. They might know better what it wants Mar 14 11:00:40 It seems, I've solved my problem:) Mar 14 11:18:22 I am a beginner, i have to implement human detection techniques on beagle board, can u please tell me where can i find some basic help Mar 14 11:18:47 i am a noob right now. thanks Mar 14 11:23:43 im tryiing to debug the beagle bone black using the emulator xds200 Mar 14 11:24:19 my board powers on and draws very little current but fails to boot Mar 14 11:25:00 the power led powers on Mar 14 11:26:18 when i debug using the xds200 an error occurs saying the test for the jtag ir instruction path length failed Mar 14 11:26:24 what does this mean Mar 14 11:26:45 does it mean that the processor is faulty or not working?? Mar 14 12:07:05 Hi all. A question about power supply: I got a 5V DC / 2A one, but it doesn't work. Mar 14 12:07:18 The barrel is longer, than the board's input Mar 14 12:07:29 is that the case? Mar 14 12:18:44 hello. im not able to ssh to beaglebone via 192.168.7.2 Mar 14 12:18:49 and is it possible to shorten the barrel or somehow fix this... Mar 14 12:19:12 what can i do to ssh to beaglebone Mar 14 12:20:38 I had played with network/interfaces file to grant static address Mar 14 12:25:43 zmatt: a little birdy told me you have an AM57xx pinout spreadsheet like your AM335x one? (I really liked your AM335x one!) Mar 14 12:25:53 zmatt: are you able to share the AM57xx sheet? Mar 14 12:30:10 Hi. How do I write the debian image to the SD without using the windows tool. I want to do it in GNU/Linux directly. Mar 14 12:30:29 I'm using Beaglebone black rev. A5C. Mar 14 12:43:45 michaelis: cp? Mar 14 12:43:48 er, dd Mar 14 12:44:33 dd if= of=/dev/sd Mar 14 13:13:04 av500: Thank you. I will try that. Mar 14 13:46:56 Nice, my new BBB USB 4 port hub fits like i hoped it would. Now i need to buy parts and build one to see if it works like i hope it will. http://imgur.com/3ieHyVG http://imgur.com/Im7eYs1 Mar 14 13:52:02 veremit: ? Mar 14 13:52:50 Strontium: that's pretty cool. Mar 14 13:57:18 jkridner: Thanks Mar 14 13:57:41 Its a 4 port self or BBB powered USB2 High speed hub. Mar 14 14:06:33 if it works, i will try and fund a batch build so bbb users can get them. Mar 14 14:57:30 <= veremit: jkridner, I think zmatt was trying to ask you somethign about the beagle-blue .. he'd just msg'd after you left .. Mar 14 14:59:29 zmatt: you are wanting to look at the blue power management? Mar 14 15:07:26 jkridner, are we any closer to basic hardware details yet?! :) Mar 14 15:07:49 I'll probably release schematics this month. Mar 14 15:08:01 jkridner, I think zmatt was hoping there may be an opportunity to feed back some of the power supply issues he observed Mar 14 15:08:18 yeah, I'd love the feedback. Mar 14 15:08:19 03:27:42 < zmatt> jkridner: btw, since I spent a lot of time examining the BBB and especially its power scheme I discovered a number of issues that, while not show-stoppers, could often easily have been fixed and saved someone some misery (and perhaps saved a few uart0_rxd ports from dying)... if you want me to go with a comb over the Blue to see if I spot anything iffy, let me know Mar 14 15:08:25 I was suggesting he might need to have some form of conversation with gerald probably .. Mar 14 15:08:39 Defiant, thanks :) Mar 14 15:09:25 jkridner, I know he's written a lot up on the elinux page .. let me find the link .. sec Mar 14 15:12:18 jkridner, et al : http://elinux.org/BeagleBone_Power_Management Mar 14 15:26:58 still plenty missing on that page though... like I said, I'm better at acquiring knowledge/insight then converting that into text or code unfortunately Mar 14 15:28:33 bradfa: it's nowhere near the quality of the one for the beaglebone (obviously, since I don't even have an x15 yet), but here -> https://docs.google.com/spreadsheets/d/1mSqEpV_BAUHfeNApytxHcGhgTZwypy564GyOr66Nphs/edit Mar 14 15:31:20 need... tea... Mar 14 15:32:43 morning zmatt :P Mar 14 15:32:48 -.- Mar 14 15:33:13 tea is good :] Mar 14 15:38:34 jkridner: this post was Mar 14 15:39:04 jkridner: this post was my first one on BBB power issues and a nice example of the digging I did -> https://groups.google.com/forum/#!msg/beagleboard/7sxPePT7wkM/3vFMPydR20IJ Mar 14 15:39:22 (actually the issue of that topic dated back to the BBW even) Mar 14 15:41:03 we actually patch our BBBs to remove the 3v3b regulator and tie 3v3b to 3v3a instead Mar 14 15:43:42 I also found vdds-related leakage... I posted on E2E about it but never got any useful response from TI -> https://e2e.ti.com/support/arm/sitara_arm/f/791/p/421756/1515700 Mar 14 15:44:09 interestingly, the am572x turns out to have the same (afaict) issue, and there it's actually documented in the errata Mar 14 15:58:35 my last post in that thread contains a more clear exposition of the issue than the original post btw Mar 14 16:00:12 after that I kinda stopped caring :P Mar 14 16:20:45 zmatt: thanks for the link :) Mar 14 16:21:01 zmatt: hey I know that bbw power issue :) I started that thread :) Mar 14 16:23:23 oh hi :) Mar 14 16:23:57 I resolved it for you ;) (at least, fully identified the issue and enumerated various possible workarounds) Mar 14 16:24:28 zmatt: yes, thank you :) Mar 14 16:25:05 zmatt: i no longer work at the company where I was doing the bbw work, but as far as I know, they still haven't had any failures due to the power-off issues Mar 14 16:25:20 zmatt: but your fix seems much more comprehensive Mar 14 16:25:50 what did you tie the regulator control to again? Mar 14 16:26:30 zmatt: U8 enable got rewired to the power LED Mar 14 16:26:44 ah same as early BBB designs Mar 14 16:26:55 until A6 iirc Mar 14 16:26:55 yes, but then it got changed in later revs of bbb iirc Mar 14 16:27:16 gerald had recommended that change to us Mar 14 16:27:40 we found it worked, so implemented it and moved on Mar 14 16:27:57 you went a bit more in-depth to actually find the root cause Mar 14 16:27:58 yeah it creates a 1 ms window for trouble at power-on and power-off Mar 14 16:28:14 how much trouble depends on the external logic Mar 14 16:29:12 zmatt: our custom capes had an external reset holding circuit to ensure that reset pulses were always at least 1 ms long, not sure if this would have helped prevent the issue or not Mar 14 16:29:27 zmatt: this is all from memory for me now, it's been a few years... Mar 14 16:30:34 no the question is whether any processor pins are driven high (powered from 3v3exp) during these windows in the power sequencing... reset (POR in fact) is asserted at these times Mar 14 16:33:05 on the bbb this happens easily since the console "isolation" buffer / level-shifter is powered by 3v3b (kinda equivalent to 3v3exp on BBW), so if you have a debug cable connected during power sequencing you end up ramming more than 40 mA down that cpu pin Mar 14 16:33:56 (powering that buffer from 3v3a would have fixed that issue) Mar 14 16:36:08 in general, having two 3v3 power domains which are interconnected by whatever IO the end user tinkers onto the board was a _bad_ idea. power domain crossings require a lot of attention, while in this case it's almost certain to receive no attention at all Mar 14 17:15:20 Folks, need help. Got my Black, Flashed to eMCC, tested...software works but the HDMI or USB will not respond. Ethernet and mini usb respond as expected. is this a defective board, or could the software change these settings...can they be change back, and how. Mar 14 18:26:01 Folks, need help. Got my Black, Flashed to eMCC, tested...software works but the HDMI or USB will not respond. Ethernet and mini usb respond as expected. is this a defective board, or could the software change these settings...can they be change back, and how. Mar 14 18:28:16 Does anybody talk, or just looking Mar 14 18:29:12 Wow....bye Mar 14 19:38:28 jkridner: hmm, has it ever been considered to replace the tps65217 (whose unique features are unused on the bbb) by the tps65910a31 ? it's cheaper, has more useful features, and since it was originally designed for the omap3 I'd suspect it to be better designed than the tps65217 (which is noticably buggy) Mar 14 19:40:52 I'd probably use a different connection diagram than the suggested one though... I'd be more inclined to use the 1A "VIO" smps for user I/O ... avoids the need for a separate regulator for that, saves power by using an smps instead of ldo, and with a bit of care you can actually make a strapping option to select between 1.8V and 3.3V user I/O Mar 14 19:43:23 (afaict using it for RAM is complete overkill when you have a single x16 RAM chip operated as DDR3-800) Mar 14 19:46:31 the only problem with that is that it needs a separate regulator for the RAM, or a change to the eeprom configuration of the pmic (and, very annoyingly, TI doesn't let people (re)program their PMICs, they require factory programming) Mar 14 20:06:01 (of course there's little motivation to tinker with a design that works well enough most of the time, I'm just thinking out loud here...) Mar 14 20:22:04 With a BBB isn't plugged into the LAN, I'm hear the LAN8710A's 25MHz crystal come on about once a second. Mar 14 20:22:21 Is there anyway from debian Jessie to turn this off? Mar 14 20:22:57 The crystal is causing some RFI in another process. Mar 14 20:24:47 there's probably some management register in the PHY to shut it down entirely Mar 14 20:28:41 ecdhe: you hear 25MHz? Mar 14 20:29:10 KotH, and harmonics. With a spectrum analyzer. Mar 14 20:29:23 * KotH blinks Mar 14 20:29:31 Have isolated the source to the PHY xtal. Mar 14 20:29:45 if that causes detectable EMI than something is wrong with the borad design Mar 14 20:29:51 zmatt, the LAN8710A data sheet says it has a couple of low power modes. Mar 14 20:30:20 I'm pretty sure that when the XTAL is off, it's in a low power mode. Mar 14 20:30:34 But I don't know how to keep the kernel from waking it up? Mar 14 20:31:11 it seems unlikely to me the kernel is doing so, but have you tried setting the link down? Mar 14 20:32:08 just `ifconfig eth0 down` ? I haven't done. Mar 14 20:32:46 zmatt, I don't know what else besides the kernel would be driving the MDIO lines. Mar 14 20:32:47 ip link set eth0 down ... or presumably what you said (I don't know much of the syntax of the old ifconfig utility) Mar 14 20:33:04 why are you assuming they are? Mar 14 20:33:47 Because the LAN8710A datasheet suggests that the way for it to enter a low power mode in the first place is only over the MDIO lines. Mar 14 20:34:28 Since it's sleeping, then temporarily waking up, then sleeping again, it seems that it's probably the kernel initiating the action. Mar 14 20:34:41 But I haven't probed the lines, yet. Mar 14 20:34:50 I see no reason for the kernel to poll it though, the phy has an interrupt line to let the kernel know about link up/down Mar 14 20:36:22 in any case, you can definitely disable ethernet entirely using device tree Mar 14 20:36:37 if you think the kernel is responsible Mar 14 20:37:26 it's also worth mentioning that if the mdio controller of the ethernet subsystem is enabled, it continuously polls the phy Mar 14 20:38:02 (I don't know whether it is left enabled by the kernel when not actively being used) Mar 14 20:41:46 note however that the PHY is actually fully enabled by default Mar 14 20:44:14 so while you can easily prevent the kernel from using ethernet (by omitting it from device tree or just recompiling the kernel without support for it) you would probably have to manually configure the phy into powerdown mode Mar 14 20:50:00 but if the kernel indeed puts the phy in low-power mode by default but for some reason it wakes up now and then, I would definitely first try setting the link down Mar 14 20:50:23 also beware of any network manager that may be installed, like 'connman' Mar 14 20:51:17 (I don't use it, but I know it's installed by default in standard images) Mar 14 21:24:08 of course you can always get rid of ethernet the same way ds2 did... with a sharp knife Mar 14 21:25:47 though that was more about power saving I think (even in 'powerdown' the phy still consumes a non-negligible amount) Mar 14 23:08:07 Does the heartbeat (BBB Rev B) run regardless if an OS is loaded? Mar 14 23:11:30 no Mar 14 23:14:04 ok good to know Mar 14 23:14:32 Ferris: only the power led is automatically turned on by hw Mar 14 23:14:58 then the four leds going on in sequence is done in u-boot Mar 14 23:15:33 I guess it was hanging on a bad fstab entry which I was able to fix by booting from a uSD card Mar 14 23:15:34 everything after that (heartbeat, cpu/mmc0/mmc1 activity leds) are software-controlled by linux Mar 14 23:15:57 you can actually reconfigure what they indicate via sysfs Mar 14 23:17:01 yeah i have a program that usually runs at startup which turns them all off and only blinks once a minute Mar 14 23:33:36 thanks y'all Mar 15 00:23:15 zmatt: what triggers the heartbeat .. I may have use on another arm board .. Mar 15 00:23:26 or even any of them Mar 15 00:23:54 I should try some things with a tri-colour LED Mar 15 00:26:13 veremit: drivers/leds/trigger/ledtrig-heartbeat.c Mar 15 00:26:51 (combined with drivers/leds/leds-gpio.c obviously) Mar 15 00:26:54 zmatt: is that a beagle driver? Mar 15 00:27:02 or generic Mar 15 00:27:23 omg, is every linux source repository on the internet down, simultaneously ??!?? Mar 15 00:27:24 :P Mar 15 00:27:40 http://lxr.free-electrons.com/source/drivers/leds/trigger/ Mar 15 00:27:40 git.kernel.org ? Mar 15 00:27:52 YES it is Mar 15 00:28:51 cool Mar 15 00:28:59 will have to try that Mar 15 00:29:08 but for now .. I think sleep is due .. Mar 15 00:29:18 this early? Mar 15 00:29:30 27 hours conscious is enough ... Mar 15 00:29:55 since I gotta be compos mentis in another 9 :/ Mar 15 00:30:22 27 hours is nothing! though I don't necessarily recommend doing what I did last friday/saturday Mar 15 00:30:31 agsinst my better judgement :p Mar 15 00:30:37 did it involve beer and pizza? Mar 15 00:30:46 or hookers .. Mar 15 00:31:36 or all three :D Mar 15 00:31:42 and/or wine Mar 15 00:31:44 no, but it did involve going to work thursday evening, leaving work saturday afternoon, and going to a party that evening Mar 15 00:31:45 hehe Mar 15 00:32:00 oh dear .. I've come close Mar 15 00:32:11 the party turns quickly into a blur Mar 15 00:32:24 didn't have that problem really Mar 15 00:32:26 assuming you don't immediately pass out in a chair .. which I've also done Mar 15 00:32:57 note: I get dexamphetamine on prescription (ADD) :P Mar 15 00:33:03 and so speaking of .. Mar 15 00:33:14 hehe Mar 15 00:33:35 (and just maaaay have slightly exceeded my prescribed dose that saturday) Mar 15 00:33:38 catchya on the flip side :P Mar 15 00:33:44 nite nite Mar 15 00:34:41 ds2: do you happen to know where I can get a curiosectomy? I really need to be less curious Mar 15 00:35:28 I can't stop pondering things like what device contained PRU/PDSP v0 and its weird LFC/STC instructions Mar 15 00:38:11 or what the envisioned purpose of sxin/sxout/sxchg is... they don't seem to be used by anything Mar 15 02:30:42 zmatt: Lol, you can get that, but it comes with a free labototmy. Im not sure you want the freebie. **** ENDING LOGGING AT Tue Mar 15 02:59:58 2016