**** BEGIN LOGGING AT Tue May 12 02:59:59 2015 May 12 03:39:07 Hey guys, to connect sensors to my beaglebone, how can I search for the wire that I need? So I can get a roll. May 12 03:45:36 22/24/26/28 gauge? What have you guys used? May 12 04:12:00 you mean like jumper cables tundraka? May 12 04:14:05 :) May 12 04:16:04 I think 22 would work, there are spec sheets for known working products that you could look at May 12 04:16:20 here is one: https://www.adafruit.com/datasheets/20218.pdf May 12 05:12:34 sigh... pru timing is too tight :( May 12 05:21:03 temperature sensitive behavior in sw is no fun May 12 05:23:05 pru timings changes with temp? May 12 05:23:48 no May 12 05:24:06 pru timing is fixed but the signal I am acquiring is right on the edge of the PRU timing May 12 05:24:17 so it crosses instruction timing boundaries May 12 05:25:53 ah May 12 05:25:58 you need a faster pru May 12 05:26:07 superpru May 12 05:26:09 got a P/N to order? May 12 05:26:24 08/15 May 12 05:26:53 input signal is a 27MHz BT656 stream May 12 05:27:15 @ 200MHz, that gives me around 5-6 instructions per clock to save things out May 12 05:27:31 but that 27MHz clock is not synchronous to the 200MHz clock May 12 05:28:58 ah May 12 05:29:01 good ole 27mhz May 12 05:29:19 ds2: you need to pll to the 27mhz May 12 05:29:25 yes May 12 05:29:37 and every few hundred ticks, I slip May 12 05:29:37 or run the pru at 270mhz :) May 12 05:30:00 where is the source clocked from? May 12 05:30:09 since that is a YUYV stream, there is very noticeable artifacts May 12 05:30:16 tvp5150AM1 May 12 05:30:23 ah May 12 05:30:26 so standalone May 12 05:30:34 no way to use the 27mhz as input on the AM335? May 12 05:31:05 won't work as the tvp5150am1 generates the clock from the view source May 12 05:31:07 video May 12 05:31:13 ah right May 12 05:31:15 so if video goes away, am335x dies May 12 05:31:28 too bad :) May 12 05:31:43 then you need a local clock to takeover May 12 05:32:03 or I find a chip that outputs a 25MHz signal May 12 05:33:36 there is a lot of dead space between the SAV/EAV signals May 12 05:34:22 hmm May 12 05:34:34 so you need to be in clock sync only for one line at a time? May 12 05:34:55 that not enough? May 12 05:35:08 that should be fine May 12 05:35:34 so detect SAV and then run a fixed loop for one line May 12 05:35:43 (kinda) May 12 05:35:46 * av500 hand waves May 12 05:37:32 yes May 12 05:41:03 morning! May 12 05:41:36 vvu: please, people are trying to sleep here May 12 05:42:06 do not mind me May 12 05:42:11 * vvu just passing by May 12 05:42:25 in ca? May 12 05:42:52 moaning May 12 05:42:54 nop, still .de until end of June May 12 05:43:30 so no makerfaire for you? May 12 05:44:01 if there are some from july1 until end of sept i'll be going, but otherwise no May 12 05:44:34 * vvu had some adventures finding housing in SF May 12 05:44:45 hehe May 12 05:48:40 hmmm different camera and it works better May 12 06:05:08 face detect at 3fps on the Beagle May 12 06:06:12 openCV ftw? May 12 06:06:15 no May 12 06:06:18 did you use -O7? May 12 06:06:40 donno... need to look at the makefile for that lib May 12 06:07:07 need to verify if neon and the good stuff is enabled May 12 06:21:41 keeping stats sucks... numbers are worse then 3fps...closer to 1.2fps :/ May 12 06:22:05 what lib? May 12 06:22:34 some simple lib called ccv May 12 06:22:41 trivial to use C interface May 12 09:32:51 can anyone here help me package a deb from a simple shell script? May 12 09:35:26 ? May 12 09:39:43 hi hi woglinde May 12 09:40:25 noticed I changed roles yesterday from a guy that asks here to one that answered the guy , about connecting to his wifi? May 12 09:41:18 so it seems hard to use debian helper tools to create a package May 12 09:41:27 and #debian offer no help May 12 09:43:16 hm? May 12 09:43:19 read the guide May 12 09:43:32 you only have a script to package? May 12 09:45:14 kind of May 12 09:45:19 I got this error: May 12 09:45:22 http://nopaste.linux-dev.org/?475364 May 12 09:45:40 Add one D in the beginning May 12 09:45:47 I used debuild May 12 09:47:10 http://dep.debian.net/deps/dep3/ May 12 09:47:15 as the comment says May 12 09:47:18 I want to package a few /etc/ config files a shell script and a python script May 12 09:47:40 reading , thanks May 12 11:44:48 ho hum... just checking, do all debian bone kernels have the problem of a 5-second delay between pressing the power button and the system turning on? May 12 11:45:28 (I think I recall they do, but I'm running a custom-built kernel atm so that's why I'm asking) May 12 11:46:15 because I just realized that means the kernel puts the PMIC in SLEEP-state instead of OFF-state... which is very very bad May 12 11:47:02 (in the sense of violating absolute maximum ratings and such) May 12 11:55:21 *at shutdown May 12 11:55:30 anyhow, afk, back in a bit May 12 12:14:22 woglinde: I copied the first example to my changelog and ran debuild again , more errors: http://nopaste.linux-dev.org/?476572. May 12 12:15:02 here is my change log: http://nopaste.linux-dev.org/?476573. May 12 12:25:11 well seems woglinde is away, can anyone else help with packaging? May 12 12:28:37 Hi any links for beginner May 12 12:29:18 I am new to beagleboard.. could any one help me May 12 15:01:54 hi guys how would i go about using device tree overlays on the latest kernels from the rcnee repos May 12 15:42:20 woglinde: hi May 12 15:42:41 here is my change log: http://nopaste.linux-dev.org/?476573 May 12 15:42:49 more errors: http://nopaste.linux-dev.org/?476572 May 12 15:43:15 I followed your advice, but it failed... May 12 15:45:09 cityLights write a sane debian/changelog file May 12 15:45:29 best ist copy one and edit it with your settings May 12 16:09:47 hi guys what kernel has preempt_rt in the rcnee repos May 12 16:10:48 mistawright, http://elinux.org/Beagleboard:BeagleBoneBlack_Debian#3.14.x-ti_BeagleBone.2FBeagleBone_Black.2FBeagleBoard-X15_.2B_RT May 12 16:12:20 rcn-ee, unfortunately I have to use the 3.8.13 kernel as I need device tree overlay support. I am not sure if the overlays are working in the latest 4.0 kernel as I have not been able to find an answer May 12 16:12:39 im also running ubuntu on this board May 12 16:13:09 my repo's are enabled for all ubuntu/debian releases.. May 12 16:13:34 last i tried 3.8.13 + preempt_rt = crashed a horrible death.. May 12 16:13:53 mistawright, which 'overlay' do you need, we need testers for v4.1-rc3 + overlay... May 12 16:15:05 I am trying to run ardupilot on my beaglebone black. https://github.com/mirkix/BBBMINI, I need a realtime kernel and there is device tree overlays provided for the beaglebone black including the pru. Would everything work in 4.0 such as pwm etc May 12 16:16:43 mistawright, where's there overlay? May 12 16:18:13 rcn-ee, these are the overlays I am using: https://github.com/diydrones/ardupilot/tree/master/Tools/Linux_HAL_Essentials/devicetree/bbbmini it also has firmware for the pru but this dts doesnt appear to be using it. will rectify with creator May 12 16:21:09 mistawright, we should be able to get most of that working this week in v4.1.0-rc3 ( i need to seperate out the hdmi/eMMC and make a generic empty bbb base.dtb) I'll import that *.dts into my bb.org-overlay right now.. May 12 16:29:29 Im not sure if I need a kernel with preempt_rt or just preempt. I know that on this page they recommend building the rt kernel. http://dev.ardupilot.com/wiki/building-the-code/building-for-beaglebone-black-on-linux/ I dont mind having to build a kernel. May 12 16:30:05 I also wouldnt mind testing the device tree overlays on the 4.1.0-rc3 kernel. May 12 16:37:52 mistawright, i've merged the dts with some generic changes: https://github.com/RobertCNelson/bb.org-overlays/blob/master/src/arm/BB-BBBMINI-00A0.dts May 12 16:38:09 do they have a cape i can order, it would be nice to have that actually uses the spi port.. May 12 16:38:58 hmmm? May 12 16:39:58 you can use the spi port with simply a piece of wire May 12 16:40:11 or if you are playing with pinmuxes, use something like 1K or 470ohm resistor May 12 16:42:01 ds2, something hardwired, something to make sure the dt bindings are correct.. May 12 16:42:10 that is hard wire May 12 16:42:14 jumper MISO to MOSI May 12 16:42:25 a "loopback" device May 12 16:45:27 if you want something that will check the clock lines too, throw in a shift register May 12 17:08:57 I finally got my beaglebone black to boot May 12 17:09:03 my 128GB microSD was too large May 12 17:09:13 I'm at 16GB and 32GB SDHC now May 12 17:09:15 and it boots May 12 17:29:26 rcn-ee: as far as you know, do debian bone kernels in general exhibit the phenomenon that after shutdown, there's a 5-second delay between pressing the power button and actual boot (power led turning on etc) May 12 17:29:30 ? May 12 17:32:09 there's usually a couple of seconds in uboot, waiting for user interaction May 12 17:33:54 that's after the power led is already on May 12 17:39:20 zmatt: any responses? May 12 17:42:06 anyone know if the differences between 3.8 Device Tree and 4.1 Device Tree are documented anywhere? May 12 17:42:54 e.g. there seems to be some new macros for pinmux, etc. May 12 17:46:46 rcn-ee, the pixhawk firecape actually uses multiple devices through spi. the bbbmini is a diy version of this. I have a mpu9250 hooked up through spi and a ms5611 as well. May 12 17:49:13 rcn-ee, here is a link to it: https://erlerobotics.com/blog/product/pixhawk-fire-cape/ it failed an indiegogo campaign. mainly due to cost. Theres a bit of redundancy built into the cape May 12 17:50:29 may i also ask why you commented out the areas you did? the only thing I can say for sure that did not work with this dts was the pru side of things and I am trying to get that worked out May 12 17:53:20 rcn-ee: I see you made some changes to the CBB-Serial overlay for 4.1. You know if that's working, or should I test it out? May 12 18:10:51 mistawright, i was just looking at that cape too.. need to find where to order it and not get stuck behind backorder. ;) May 12 18:11:25 alexanderhiam, it looks okay from kernel land, not 100% on the rts/cts lines so those need to be added.. can 'should' work.. ;) May 12 18:15:43 rcn-ee: cool. Does it make sense to have those separate flow control overlays again? May 12 18:16:14 rcn-ee, this is the kernel i would need right: linux-image-4.1.0-rc3-bone0 and does that come precompiled with rt? May 12 18:17:38 alexanderhiam, let's combine them in the pinmux for the cape.. ;) Then when another cape comes out that uses the same usartX rx/tx/cts/rts/etc, i'll just use meld to copy it over as is. ;) May 12 18:18:09 sounds good May 12 18:18:43 mistawright, sorry: CONFIG_PREEMPT_VOLUNTARY=y May 12 18:19:02 rcn-ee: and are the 4.1 DT changes going to be able to be pushed to 3.8, or are overlay incompatibilities unavoidable? May 12 18:19:39 I'm thinking of continuing to use overlays in PyBBIO on 4.1, and whether or not I would need different versions May 12 18:19:54 how would i go about building the 4.1 kernel with preempt. Wondering if I can build this on an x86 gentoo machine or archlinux as well. not really wanting to wait to build it on the beaglebone black May 12 18:20:00 "some" things will work with only a valueX = value[X-1], change... May 12 18:20:55 alexanderhiam, and i'm not sure how pwm/ecap works in 4.1 dt... most should port over easily.. May 12 18:21:22 or perhaps that's a bad idea and I should just bite the bullet and convert to using universal-io on all versions May 12 18:21:28 alexanderhiam, panto, also pushed new dtc branch for dtc, i didn't look to see if it's 100% compatible with v3.8.x's dtc.. ;) May 12 18:21:39 hi rcn-ee May 12 18:21:55 hi panto May 12 18:21:56 I'm pushing updates to my tree May 12 18:22:04 some cleanups May 12 18:22:30 panto, cool! i'll add them asap.. btw, the new omap 8250 serial, doesn't like stuff added, (so we went back to omap_serial) May 12 18:22:38 yeah May 12 18:22:48 expect lots of broken drivers May 12 18:23:40 well, omap_serial, can, tilcd, led's, work. .;) May 12 18:23:46 wow May 12 18:23:49 awesome! :) May 12 18:24:05 working on the lcd4 right now, they should be working fine.. May 12 18:24:17 hopefully find a spi cape so i can test that too.. May 12 18:24:53 * alexanderhiam heads afk, but will do more capemgr testing soon May 12 18:25:15 and thanks panto and rcn-ee! May 12 18:25:28 hopes alexanderhiam, on the rate things are doing, i'm nuking v3.8.x in my tree for defualt images. ;) May 12 18:25:56 awesome May 12 18:26:30 still need to tackle the new "default" bbb base *.dtb (with eMMC/hdmi disabled) i've been cheating by using the bbw... May 12 18:30:06 rcn-ee, any reason for commenting out the dts file i supplied where you did? May 12 18:31:35 ds2: at least on the kernel I built if I shut down and then press the power button to turn it on again there's a 5-second delay before the power led turns on... I think I recall the same behavior on stock kernels but I'm wondering if I maybe remember wrong May 12 18:32:03 mistawright, haven't had a chance to test spi.. and we need a blank bbb.dtb, otherwise the pru (aka lcd) pins aren't going to work.. ;) May 12 18:33:18 in any case, any kernel exhibiting that behaviour is doing something Very Bad™ on shutdown and on BBB rev A6A or later risks hardware damage May 12 18:35:00 rcn-ee, so at this point the pru is not working right with this kernel? Or is this something done after boot? May 12 18:36:02 ds2: I thought it'd be something obvious enough to just ask people in general whether this happens, but given the lack of response (does nobody ever shut down their BBB?) I'm now installing a standard kernel again to check it :/ May 12 18:37:18 mistawright, pru: just dt changes... May 12 18:38:56 hmmm i happen to have the cape i made with me at work. I could try to see if spi works that way. as long as the spi devices are named the same i can even run arducopter to see if it picksup gyro and barometer correctly May 12 18:39:06 zmatt, for the bbb, on later mainline kernel's, it's a combination rtc/pmic that does the final shutdown.. and if you battery back that rtc, you can actually get it to "wake" up.. ;) May 12 18:40:59 Hi, I just programm the beagle bone to perform a simple task (Blink an external LED) I am woundering if I can run this program after I disconeccted from the PC and supply it with power scource? May 12 18:41:04 mistawright, that would be awesome! here's what i used in v3.14.x: https://github.com/RobertCNelson/dtb-rebuilder/blob/3.14-ti/src/arm/am335x-bone-spi0-spidev.dtsi might give you a hint on what to chagne from the 3.8.x version.. May 12 18:43:27 Hi, I just programm the beagle bone to perform a simple task (Blink an external LED) I am woundering if I can run this program after I disconeccted from the PC and supply it with power scource? May 12 18:43:49 okay, now i'm just lazy... come on build bot.. hurry up with panto's changes. ;) http://builds.beagleboard.org/builders/build-kernel May 12 18:48:11 rcn-ee: it's normally always an rtc/pmic combination that does the shutdown, but "shutdown" into pmic sleep-state (aka "RTC only sleep") is absolutely forbidden on a beaglebone unless you do multiple hardware patches May 12 18:48:58 rcn-ee, should i just be able to uncomment it and I try? I noticed the file you linked to didnt have any fragments and seemed to be slightly different May 12 18:49:29 unless it does shutdown into off-mode and it's a pmic bug, but I don't think so (but I can check) May 12 18:50:13 zmatt, i don't remember which, but it was mentioned on l-o when the rtc shutdown was fixed this winter.. May 12 18:50:24 rcn-ee: do you happen to know if SYS_5V stays powered in shutdown? (that's a clear sign of pmic sleep-mode when running on DC power) May 12 18:51:02 i know d1 goes off, how is that strapped to Sys_5v... May 12 18:51:13 d1 ? May 12 18:51:20 the led you mean? May 12 18:51:30 yeah the pwr led, sorry.. May 12 18:51:32 power led is on ldo2, you can actually turn it on and off in software May 12 18:52:16 mistawright, give it a shot, if spi works, send me a git merge and we will enable it. ;) May 12 18:52:41 i2c mw 24 0b 6b ; i2c mw 24 16 7e May 12 18:52:54 that turns off the power led :) May 12 18:53:00 (in u-boot ) May 12 18:53:30 requires BBB rev A6 or later May 12 18:53:42 before that there'd be hell to pay if you execute those commands May 12 18:54:12 (it would turn off the 3v3b regulator) May 12 18:55:54 zmatt i guess SLEEP mode: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/arch/arm/boot/dts/am335x-boneblack.dts?id=672e2b14741368134c8640e3851c841b422f6c45 May 12 18:57:41 yep, just measured sys_5v, it stays on May 12 18:59:28 that needs to be fixed FAST... pmic sleep mode on an unpatched beaglebone is absolutely not permitted May 12 18:59:54 zmatt, black or white? (the white it's bad..) May 12 19:00:41 on mainline, it's setup for only the black.. May 12 19:00:59 dunno the white, black is very very bad May 12 19:01:23 except versions older than A6 May 12 19:01:28 what's a few dead beagles among friends:D May 12 19:01:30 A5A is safe May 12 19:02:11 but A6 introduced the 3v3b regulator bug, which applies in off-mode on battery power, and sleep-mode on any kind of power May 12 19:02:44 zmatt: considered writing a summary of versions/battery power/issues? May 12 19:02:45 for the cpu, in the patch review, Felipe metioned for the processor it has to be ES1.0 or later.. May 12 19:03:01 rcn-ee: 2.0 you mean? May 12 19:03:17 well, yeah.. es1.0 it's broken.. May 12 19:03:23 rtc-only mode was iirc indeed buggy in am335x May 12 19:03:38 but only in the sense that the rtc doesn't actually work May 12 19:03:43 which is a bummer in rtc-only mode ;P May 12 19:03:54 it's not bad in a hardware-damaging way afaik May 12 19:04:49 ds2: for the BBB there are just 3 cases: before A6, A6, A6A and later May 12 19:05:03 zmatt: a chart summarizing it would be good May 12 19:05:33 I suppose I could... what would be the most suitable place to put it? May 12 19:05:53 elinux? May 12 19:06:26 rcn-ee: btw, in such situations (shutdown on battery power, or sleep mode) on BBB A6 or later... in any case, do _not_ leave a serial cable connected May 12 19:06:55 rcn-ee, is it no longer necessary to echo capes into slots? May 12 19:07:13 preferably avoid such states altogether, but external hardware actually makes a significant difference, and the serial driver IC is one of those May 12 19:07:14 zmatt, by battery power what pins are you connecting? May 12 19:07:20 i'm just using 5vdc... May 12 19:07:42 mistawright, actually haven't figured out how to "echo" . ;) so patched the initrd.img instead. .;) May 12 19:08:14 rcn-ee: TP5-TP8, so these problems used to be limited to a very small group of people. But the kernel change to use pmic sleep mode has now introduced those problems to the masses :P May 12 19:11:21 since sleep mode doesn't merely keep sys_5v and the rtc powered, (since rev A6) vdd_3v3b also remains up and (since rev A6A) the cpu 'vdds' supply also remains up May 12 19:11:25 but everything else goes down May 12 19:11:44 so external hardware will now be happily driving into unpowered processor I/Os May 12 19:11:54 rcn-ee, so i need to edit my uEnv.txt? it looks as thought I need to disable hdmi as well. I dont see it in the list of items when cat'ting slots May 12 19:12:52 even with no external hardware connected at all I measured 35 mA May 12 19:12:53 mistawright, it's only a couple days old.. ;) i need to write a blank beaglebone dtb with the hdmi/eMMC disabled.... May 12 19:12:59 leaving a serial cable connected... 80 mA May 12 19:13:09 zmatt, ouch! May 12 19:13:30 which I fear might mean those 45 mA are driven into a single cpu pin (uart0_rxd) May 12 19:16:48 rcn-ee: see here for scope pics of what happens to the supplies => https://goo.gl/RO5W2W May 12 19:17:04 rcn-ee, that would explain the conflicts when I tried echo'ing it into slots. It did load somewhat however or shows when catting slots that it is there May 12 19:17:32 mistawright, the conflicts worked! ;) May 12 19:17:40 i haven't tested that yet. ;) May 12 19:18:14 rcn-ee, http://pastebin.com/C1BH279y May 12 19:18:39 rebooting now to see how it looks after cloning your overlay repo updating dtc and installing the overlays May 12 19:19:08 rcn-ee: these show off-mode on DC supply vs battery, but sleep-mode (on any power source) will produce basically the same results as off-mode on battery power May 12 19:19:15 slightly worse May 12 19:20:51 (since vdds does in fact also leak heavily to 3v3a) May 12 19:21:23 although I suppose the 3v3b leakage already keeps 3v3a high enough to not notice that May 12 19:22:56 rcn-ee, http://pastebin.com/XYmpPRvD how should gpio4 be referenced as it appears it doesnt like the old way that worked May 12 19:23:18 ah no you actually do notice that, provided the BBB has no external hardware connected May 12 19:23:58 mistawright, move it to gpio3 May 12 19:24:45 mistawright, just pushed a blank dtb, watch this address for the build: http://builds.beagleboard.org/builders/build-kernel/builds/27 (should start building in 5 minutes) May 12 19:24:57 use: dtb=am335x-boneblack-overlay.dtb May 12 19:26:11 should i just pull in the latest changes and try again? from what you have told me the uEnv.txt wouldnt affect what is loaded May 12 19:27:04 rcn-ee: the stupid thing is that they also fucked up the implementation of sleep mode evidently considering the RTC doesn't respond to a wakeup request (power button), causing the pmic to power cycle the system after a 5-second timeout May 12 19:27:21 mistawright, i don't have a disable=eMMC,HDMI setup.. May 12 19:27:44 so you need to force the overly dtb in /boot/uEnv.txt that has eMMC/HDMI disabled... so you can use the pru (lcd) pins.. May 12 19:28:28 zmatt, the power button wake up actually works on my Rev C.. where my older none c, never shtudown.. May 12 19:29:09 rcn-ee: hmm, no 5-second delay? May 12 19:29:17 (A5B).. atleast it's not as crappy as my (A5A) with broken musb... May 12 19:30:29 I have the delay on my A5A with a kernel based on 4.0.0-rc6-bone-something May 12 19:31:18 if I understand the commit message right, adding the ti,pmic-shutdown-controller will fix the problem? May 12 19:31:27 gonna test that.. May 12 19:33:32 hmm, no "ti," in my dts... I guess that changed later May 12 19:34:55 it did... now it's generic May 12 19:35:30 the shutdown property would probably be on the pmic... let's see if I can find it in the driver May 12 19:36:34 https://github.com/beagleboard/meta-beagleboard/blob/master/common-bsp/recipes-kernel/linux/linux-mainline-3.8/rtc/0003-ARM-dts-AM33XX-Set-pmic-shutdown-controller-for-Beag.patch May 12 19:36:39 bingo May 12 19:37:59 mistawright, the linux-image-*.deb is up with the blank overlay: http://builds.beagleboard.org/linux/4.1/6c5e97c356d91c77257f9d41d40aec587f6f24f7/ May 12 19:38:12 rcn-ee: I'd urgently apply the same patch for the beaglebone black May 12 19:39:55 zmatt, so that's part of v3.14.x: https://github.com/RobertCNelson/ti-linux-kernel-dev/blob/ti-linux-3.14.y/patches/beaglebone/dts/0004-tps65217-Enable-KEY_POWER-press-on-AC-loss-PWR_BUT.patch#L22 May 12 19:40:12 I don't know how easily the am335x gets fried by powering it through the protection diodes of its processor IOs, but I'm in favor on not finding out :P May 12 19:40:44 never got mainline: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/am335x-bone-common.dtsi#n225 May 12 19:41:17 so we are doing the right thing on v3.8.x & v3.14.x-ti, but need to fix mainline.. May 12 19:41:25 nono May 12 19:41:34 actually that patch is wholly unrelated May 12 19:41:44 that's for pmic events when powered May 12 19:41:46 via irq May 12 19:42:07 it's unrelated, but the "ti,pmic-shutdown-controller;" is enabled in the tps node.. May 12 19:42:21 eh, I see no + there May 12 19:42:38 oh right May 12 19:42:40 that's what you meant May 12 19:42:46 (which means someone at ti patched it in their tree) but didn't push it! May 12 19:43:15 naughty May 12 19:43:56 i'll add it right now to v4.1.x-bone as i'm hacking on it.. May 12 19:43:56 the commit message should have included something along the line of "URGENT - avoids potential hardware damage" :P May 12 19:44:53 it should be in all dtbs where the rtc is the shutdown controller May 12 19:45:20 rcn-ee, installing debs now May 12 19:45:23 I'm wondering how it (tried to) shut down before then? May 12 19:45:24 as long as it's in am335x-bone-common.dtsi May 12 19:45:41 before, it wasn't... shutdown shut the core down... May 12 19:45:46 how? May 12 19:45:50 (or: idle...) May 12 19:46:05 you mean it just halted but stayed on? May 12 19:46:20 hmm, I'm pretty sure I did a shutdown before 4.0 May 12 19:46:33 exactly... before v4.0.x it didn't truly shut down.. .it just halted... May 12 19:46:51 were those patches maybe backported to 3.8-bone ? May 12 19:46:54 the rtc patches atleast moved it forward to a shutdown, looks like no one tested the tps to make sure.. May 12 19:47:30 the patch was for A5 May 12 19:48:26 Johan Hovold, rewrote the rtc driver last fall, before enabling the shtudown on mainline for the beaglebone's.. May 12 19:48:28 which does support rtc-only mode... although I'm wondering if the rtc misconfiguration is maybe another thing that ti fixed but failed to mainline May 12 19:48:44 (rtc misconfiguration = the 5-second delay) May 12 19:49:10 which I'm pretty sure also resets the rtc, negating the usefulness of rtc-only mode May 12 19:49:47 I actually have a rev C with working rtc-only mode, but it required two hardware patches :/ May 12 19:50:23 (and moving R8 to R9 is not a fun patch) May 12 19:50:51 located between the DC power socket and the ethernet socket May 12 19:52:21 yeah no way... May 12 19:55:23 colleague of mine patched it for me :) May 12 19:55:43 I probably would have turned the whole area into a scorched solder blob or something May 12 20:00:26 the rtc misconfig is fortunate though, otherwise I wouldn't have easily noticed the 4.0 kernel puts the pmic into sleep-mode May 12 20:04:29 the patch seems to work, it now responds immediately to the power button May 12 20:05:27 zmatt, sweet! so all you added was ti,pmic-shutdown-controller; May 12 20:07:00 yep. I got confused for a moment since sys_5v still seemed up, but then I realized that without the 3v3b regulator bug (I'm doing this test on my A5A) sys_5v will have almost no leakage in off-mode and it's backed by like 150 uF of capacitors May 12 20:07:20 so I'll leave it shutdown for a bit and measure it again :P May 12 20:10:09 (or possibly I measured the wrong pin) May 12 20:10:25 in any case it's 0.78 v now -- yay May 12 20:14:07 I'm hunting for TI's original patch that introduced that line in bone-common May 12 20:14:32 hay, can you guys help me package a deb file May 12 20:14:53 No c files to compile, so it should be easy - right? May 12 20:15:27 well i used debuild, and got an error from the change log file. so I used dch, still error May 12 20:15:38 how do you package rcn? May 12 20:18:26 rcn-ee: found it, c85d82b7cedad29af6ee54707acdadc61904fc49 May 12 20:19:25 cityLights, take a look at bb-customizations: https://github.com/rcn-ee/repos/tree/master/bb-customizations/suite/jessie/debian May 12 20:19:32 it just copies 4 scripts to the rootfs... May 12 20:19:42 so no *.orig, just a debian dir.. May 12 20:19:48 you are a great helper May 12 20:20:05 even better then deb helper May 12 20:20:09 hehe May 12 20:20:29 well, it uses... debhelper. ;) https://github.com/rcn-ee/repos/blob/master/bb-customizations/suite/jessie/debian/control#L5 May 12 20:21:08 yep, i just want to release a proper deb with the wifi solution May 12 20:21:34 rcn-ee: seems google doesn't index TI's commits, but manually constructing a link worked => https://git.ti.com/ti-linux-kernel/ti-linux-kernel/commit/c85d82b7cedad29af6ee54707acdadc61904fc49 May 12 20:21:34 zmatt, suprised j-keerthy didn't already push it mainline.. May 12 20:22:08 yeah i tried mainline, before moving to ti's git tree. ;) May 12 20:22:11 rcn-ee: again the commit message unfortunately lacks any explanation of why it's needed May 12 20:22:27 I see 4.0 is out, can you sum-up the best features in it? May 12 20:23:30 cityLights, 4.0 isn't important.. :) panto gave use capemgr for v4.1.0-rcX ;) May 12 20:23:34 right now I'd go for "urgent inclusion into mainline" but that's probably not a pleasant route May 12 20:24:04 thanks May 12 20:24:25 into stable mainline that is, since the shutdown patch is there already right? May 12 20:24:56 zmatt, step 1: add a bug to http://bugs.elinux.org/, step 2, post to l-o, and it'll hit every release.. May 12 20:25:13 greg, backported this change to every stable back to v3.12.x: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/arch/arm/boot/dts/am335x-bone-common.dtsi?id=67fd14b3eca63b14429350e9eadc5fab709a8821 May 12 20:26:01 as long as the commit has Fixes: bug link, it'll hit everywhere.. May 12 20:27:02 rcn-ee: it's clear you're a lot more familiar with the relevant processes than I am... mind if I leave this to you? May 12 20:27:20 eh, you sure you pasted the link you intended to? May 12 20:27:40 as long as you fill out the bug report, on bugs.elinux.org, otherwise they might see me trying to cheat. ;) May 12 20:27:46 oh right now I understand what you meant May 12 20:28:14 I am in a better position to explain the problem I guess :) let's see... May 12 20:28:48 make sure to add your beagleboard google group link, for all the background.. May 12 20:28:59 I'm guessing I need to register to file a bug? May 12 20:29:02 (and your signed off.) May 12 20:29:15 yeap, i can approve you after you click.. May 12 20:30:25 click May 12 20:30:53 and clicked.. May 12 20:44:46 rcn-ee: I got this error when trying to package, did your encounter it? http://paste.ubuntu.com/11101721/ May 12 20:46:07 rcn-ee: hmm, the commit: syntax shown in the text formatting help doesn't work? (for the mainlined patch) May 12 20:48:40 cityLights, might be missiing debian/control: https://github.com/rcn-ee/repos/blob/master/bb-customizations/suite/jessie/debian/control May 12 20:57:09 here is the control: http://paste.ubuntu.com/11101931/ May 12 21:06:30 ok I copied your changelog and I got a diffrent error May 12 21:06:45 dpkg-source: error: can't build with source format '3.0 (quilt)': no upstream tarball found at ../cwvw_1.20150507.orig.tar.{bz2,gz,lzma,xz} May 12 21:12:18 rcn-ee: what Priority should I select? May 12 21:12:37 cityLights, you have files outside ./debian/ so you need to tar those up as is and create an archive, i cheated and just have them copied from debian.. May 12 21:12:47 zmatt, doesn't matter, i'll do it tonight. ;) May 12 21:12:52 zmatt: I definitely think a written summary of your finding would be extremely valueable May 12 21:13:07 agree ^ May 12 21:13:34 well you've got one... http://bugs.elinux.org/issues/143 May 12 21:13:36 :) May 12 21:14:08 crap gotto run to catch last train, bbl May 12 21:14:23 why must I use version XX~YY ? when using 0.1 I get an error when sing 0.1~ I dont May 12 21:16:04 cityLights, just use the "1.20150507-1" syntax. ;) the extra ~'s are for me with reprepro limitations, and making sure things work when people dist upgrade to the next version.. May 12 21:16:22 1.(today's date)-1 ;) May 12 21:17:22 thnaks May 12 21:29:43 is the name cwvw too short? May 12 21:33:44 for a packege? May 12 21:38:24 I almost have Minix3 built for my BBB May 12 21:42:38 <_av500_> how is the usb support? May 12 21:42:58 USB support is experimental for Minix3 on the beaglebone black May 12 21:43:08 lol May 12 21:43:09 but contributions and patches are welcome May 12 21:43:38 um... let me post a link on minix3 support on the beagle boards May 12 21:44:00 http://wiki.minix3.org/doku.php?id=developersguide:minixonarm May 12 21:44:07 ^ see the Feature Matrix May 12 21:44:33 Minix3 supports Beaglebone Weather May 12 21:44:39 which is kind of neat I guess May 12 21:44:52 http://elinux.org/Beagleboard:BeagleBone_Weather May 12 21:48:57 rcn-ee: well I am lost. can you please see this: https://www.dropbox.com/s/dz0piv25pt2oy68/cwvw_1.20150512.orig.tar.xz?dl=0 May 12 21:49:14 its all the project along with the four debian files May 12 21:50:53 my changelog is so simple now: May 12 21:51:18 http://paste.ubuntu.com/11102464/ May 12 21:52:40 so hard to get a proper deb file May 12 21:56:52 panto: also working overlays: tsadc, tilcd-panel, gpio-keys, gpio-backlight.... ;) May 12 22:06:03 ok, now I get the error of a missing sig. this changelog was tricky May 12 22:56:54 back May 12 22:58:04 _av500_: who needs usb anyway... May 12 23:00:38 ds2: what I think would be really helpful would be simplified schematics/diagrams of the various generations of the beaglebone power supply scheme, but I'm not sure what to use... I still haven't found any decent diagramming program, and my experience with schematics programs is that they tend to be user-interface horrors May 12 23:04:51 zmatt: eagle should be fine May 12 23:04:54 print to PDF May 12 23:05:18 zmatt: or if you just want to purely draw stuff - LTspice May 12 23:05:33 the schematic editor there is not a total abomination for casual use May 12 23:06:37 LTspice has no linux version afaict May 12 23:09:59 it runs perfectly in wine May 12 23:10:20 or you can always use OrCAD May 12 23:10:28 then you can cut and paste from the official schematics ;) May 12 23:10:47 which is generally not useful when you want to make a simplified diagram May 12 23:11:02 if they did a proper hierchy.... May 12 23:11:26 then probably I would have trouble getting the relevant details at the top-level May 12 23:11:58 and getting everything else gone May 12 23:16:28 and based on reading a short tutorial on eagle I doubt that it's gonna make things pleasant May 12 23:16:47 considering it seems to be very focussed on also getting a pcb May 12 23:21:29 I really just want a flexible diagramming program, and I've wanted one for a long time already :/ May 12 23:31:33 zmatt: Dia. Fritzing. Visio. gEDA. May 12 23:33:41 yes I know about those May 12 23:34:08 (although I've never really used Visio) May 12 23:35:45 zmatt: YOu have exhausted the list of diagramming softwares I know of. Write your own. May 12 23:35:54 Or pay someone to write one per your spec. May 12 23:37:39 I may perhaps some day do the former :P not that I really have time for such a project, but the annoyance is one I run into sufficiently often that I might still attempt it May 12 23:38:06 (if you want to make your list longer btw, try yEd) May 12 23:38:34 * agmlego|skynet is sated with the current list, but thanks. May 12 23:50:38 :) May 13 00:59:02 sigh eve May 13 02:14:55 Hello everyone May 13 02:17:05 I have a one problem what my beaglebone is halted May 13 02:18:05 I think RAM is full May 13 02:19:21 It works 4 client code, and communicate to server program every 5 minutes May 13 02:19:43 What do you say about it? May 13 02:22:21 SpeedEvil are you there? **** ENDING LOGGING AT Wed May 13 02:59:58 2015