**** BEGIN LOGGING AT Tue Apr 21 02:59:58 2015 Apr 21 04:59:39 Hello to all Apr 21 05:00:07 does any one using Kivy framwork on the BBB? Apr 21 08:43:17 morning guys. Just a quick question: is there a command to display the current pinmuxing mode of the pins for BBB? Apr 21 08:43:48 I mena if the pin is used as input/output and for which mode it is configred Apr 21 08:43:56 I mean... Apr 21 08:44:07 there are pinmux files in /sys Apr 21 08:45:05 ok, where in sys? I donno see a pinmux folder Apr 21 08:45:32 I have debian running Apr 21 08:45:32 bennyb which kernel? Apr 21 08:45:41 3.8.13 Apr 21 08:46:47 BennyB_: don't remember the path, but did you try a find? Apr 21 08:47:08 no, but what should i search for? Apr 21 08:47:12 pinmux? Apr 21 08:48:57 look at /sys/kernel/debug/pinctrl or something like that Apr 21 08:49:51 ah, ok ther are these directories: pinctrl-devices pinctrl-handles pinctrl-maps Apr 21 08:49:53 there should be a device folder with a bunch of files that give you the current state. Apr 21 08:51:24 ther is a 44e10100.pinmux folder Apr 21 08:51:37 yes, inside there. Apr 21 08:51:43 some files Apr 21 08:51:51 i know, you should read them :) Apr 21 08:52:04 gpio-ranges pinconf-groups pinconf-pins pingroups pinmux-functions pinmux-pins pins Apr 21 08:52:08 which one? Apr 21 08:52:17 whichever you are interested in Apr 21 08:52:27 i'll try Apr 21 08:52:28 less them and se. Apr 21 08:52:30 *see Apr 21 08:56:38 hm, pinmux-pins seems to be the list but I don't understand the output Apr 21 08:57:03 what was the site name to post bigger files again... don't remember it right now Apr 21 08:57:11 pastebin? Apr 21 08:57:17 yepp Apr 21 08:57:19 thanks Apr 21 08:57:34 hm wasnt there graphical tool for 3.8? Apr 21 08:58:13 that is the output of pinmux-pins : http://pastebin.com/Mh1yvxBr Apr 21 08:58:47 I just found this: http://roshan.info/blog/2014/03/03/beaglebone-black-pinmux-and-dts-helper/ Apr 21 08:58:55 could that be helpful for mew? Apr 21 08:59:01 ..me ? Apr 21 09:01:27 jepp Apr 21 09:01:37 did you read the entry? Apr 21 09:01:50 To display the current pin configuration, just run the utility without any parameters: Apr 21 09:02:19 yes I'll try Apr 21 09:06:06 OK, that helped me. Apr 21 09:06:42 I have the problem that two pins seem to be high all the pime when I confgure them as input Apr 21 09:07:05 I now found out that they are defines as i2c_scl an sda Apr 21 09:07:23 how can I disable the i2c function on this pins Apr 21 09:08:17 the python tool says : Apr 21 09:08:22 Header P9_18 Name: I2C1_SDA Number: 86 GPIO: 4 0x958/158 Mode: I2C1_SDA (2) Direction: in PUD: pull-down Apr 21 09:08:23 Header P9_17 Name: I2C1_SCL Number: 87 GPIO: 5 0x95c/15c Mode: I2C1_SCL (2) Direction: in PUD: pull-down Apr 21 09:12:07 exacter question: how can I disable i2c1? Apr 21 09:21:55 anybody? come on guys! Apr 21 09:23:26 cant you use some other pins? Apr 21 09:23:28 there are so many Apr 21 09:23:46 no I'm afraid I cannot use other pins Apr 21 09:23:46 if you want to disable I2C on them you need to edit the DTS file and remove i2c Apr 21 09:24:18 ok, which dts file? the boneblackdts? Apr 21 09:26:25 the one that gets loaded by the kernel Apr 21 09:27:40 so the am335x-bone*.dts ... but where is it located in the sources ? Apr 21 09:28:15 ah, sorry for asking, found it... Apr 21 09:41:33 hm, it seems to bbe in am33xx.dtsi, but it say status="disabled" so how is it automatically enabled at boot? I'm not so experienced with dts files... Apr 21 09:49:17 check the files included Apr 21 09:49:30 there might be an override somewhere Apr 21 09:56:25 hm, in bone-common.dtsi it s an paragraph for i2c2_pins. Can I simply delete it and rebiuld the dtb file and then copy it to the boot/uboot/dtbs folder? Or is the procedure different? Apr 21 09:57:38 BennyB_: i2c2 != i2c1 Apr 21 09:58:58 oh, thanks. Apr 21 09:59:32 but then there is nothing for i2c1 only for 0 and 2... Apr 21 10:00:13 the python tool sayd the pins are on mode 2 Apr 21 10:02:17 dtsi says for i2c0 mode 0 and i2c2 mode 3.... for i2c1 I can't find anything. In am3xx file the is this: http://pastebin.com/fRupBvZs Apr 21 10:02:43 they are alld disabled there Apr 21 10:03:26 i2c0: ..... hwmods="i2c1" .... so now WHO'S WHO ? Apr 21 10:04:52 bone-common says this: http://pastebin.com/PbuizVfh Apr 21 10:06:14 dts files are lik a book with 7 locks.... Apr 21 10:21:00 hello Apr 21 10:21:43 BennyB_: are you sure those are even muxed then? Apr 21 10:22:03 need help on Beagle Apr 21 10:22:23 BoneBlack.. I want to install Angstrom Apr 21 10:22:28 BennyB_: did you check if those pins are shared with anything else and muxed to some other function? Apr 21 10:24:03 hello need help on beaglebone black with angstorm Apr 21 10:25:22 sidh: well, what do you need help with? Apr 21 10:25:49 tbr, I want to port angstorm on BBB Apr 21 10:28:06 sidh: you're a couple of years late. It has been done already. Apr 21 10:29:26 tbr, yeah that I know. basically I want to learn . I want to boot BBB from SD card with angstorm Apr 21 10:30:18 sidh: it's as easy as downloading an image, writing it to a card, sticking it into the BBB and powering it up Apr 21 10:30:45 tbr, I tried same but board not booting.. please point me to some doc Apr 21 10:37:14 sidh: are you on linux or windows? Apr 21 10:37:25 tbr, linux Apr 21 10:41:09 sidh: and which angstrom image did you download? Apr 21 11:06:07 tbr: sorry I was away for lunch. I'm not sure how the state of the pins is, that's why I asked to a tool to show the pinmux state of the pins. Apr 21 11:07:17 tbr: my posting 90 mins ago showed that the pins are configureed as mode 2 for i2c, but I'm totally sure I exported them be my c++-program to be inputs Apr 21 11:07:39 tbr: so I thought the pins are in mode 7. Apr 21 11:08:23 tbr: but my pins are always high independent of which signal I apply to them. that is what made me search for the problem. Apr 21 11:09:18 tbr: how can I search if the pins are double configured/muxed ? Apr 21 11:12:27 BennyB_: one way is to look up the pins and their potential mux functionality and then check if any of that is enabled/used Apr 21 11:13:05 tbr: and how to do this? - sorry for asking, but I do not know... Apr 21 11:15:26 i can cat /sys/class/gpio/gpio5/direction or gpio4 . this gives me always a "0" Apr 21 11:15:45 ah sorry, give me "in" Apr 21 11:16:00 value give me always 0 Apr 21 11:17:39 BennyB_: the TRM should have a listing, but you'll also find pages "on the internet" listing potential functions of pins Apr 21 11:19:52 tbr: this is the output the python tool: http://pastebin.com/n9XKWnue for p9_17 + 18 it says i2c mode 2 Apr 21 11:22:46 tbr: did you mean the SRM? it states for the two pins P9_17+18 mode 7 should be gpio, but they run in mode 2 which is i2c. Apr 21 11:25:03 BennyB_: did you try adding definitions of those pins to a dtsi? Apr 21 11:25:57 and obviously then rebuilding the dtb and using it Apr 21 11:26:45 no I didn't add anything to a dtsi for these pins I simply exported them from my c++ prog to be inputs Apr 21 11:27:14 how? by f*cking around with /dev/mem? Apr 21 11:28:58 no open the file /sys/class/gpio/export and write e.g. "4" to it Apr 21 11:29:19 like on consolse echo 4 > /sys/class/... Apr 21 11:29:53 for all the other pins I use this works, because theey are not used by the system Apr 21 11:31:27 how about doing it properly and defining all your pins in DT? Apr 21 11:34:42 but I need to reconfigure my pins during the run of the program Apr 21 11:38:48 my problem right now is simply that the two pins are not in the mode I need them. Do you know where I need to thange it in the boneblack respectively dtsi files to disable the two pins beeing in i2c mode? Apr 21 11:40:42 I posted some pastebins with the content of the dts files, but I don't know what to delete/modyfy ther to achieve my desired result. Apr 21 11:43:29 I do not find the place where i2c1 is activated in the dts files... only for i2c0 and i2c2 Apr 21 11:47:17 try just defining the muxes for those pins Apr 21 11:47:46 which kernel are you on anyway? Apr 21 11:50:58 as far as I see the "address" of one pin is 0x95c I do not find anything with this number in the dts files Apr 21 11:51:08 i use kernel 3.8.13 Apr 21 11:52:37 then you could even make an overlay Apr 21 11:57:24 I tried thy python tool. is the output correct, can you please check? Because I am confused about the address 0x158 the tool outputs: http://pastebin.com/jur48rxP Apr 21 11:59:18 in no parameters mode the python tool outputs address 0x95c/15c for pin p9_17 Apr 21 13:26:55 hey guys, how can i stop the BBB from powering up when USB is connectd? Apr 21 14:47:54 gta Apr 21 16:38:55 attempting to connect bbb to mac for the first time - doesn't show up at all on my network interfaces. any suggestions? Apr 21 16:41:04 ( have downloaded and installed serial & network drivers for mac ) Apr 21 16:46:04 tom___: does it show up on other computers? which distro are you running on the BBB? Apr 21 16:46:59 yep. shows up on network interfaces for windows and other ubuntu 14.10 / 14.04 computers. running 14.04 on the BBB Apr 21 16:47:54 hmm, the 14.04 image comes with that enabled? we've had lots of people complaining, because it didn't enable the network and serial thing Apr 21 16:48:16 hi Apr 21 16:48:27 I guess you should check if the mac even sees the device Apr 21 16:49:20 I found interesting issue. I wanted to supply beagle with cable DC jack -> USB A which I attach to charger (5V 2A) Apr 21 16:49:29 however board keeps resetting Apr 21 16:50:00 any ideas ? Apr 21 16:51:37 tbr: yea, we had to use the g_ether module, but it does show on other OS's network interfaces as usb0 -- which i believe is different from the initial angstrom install which is on eth0 Apr 21 16:52:08 tomkbrady: then it _should_ Apr 21 16:52:48 syrioosh: maybe your charger isn't that good after all Apr 21 16:53:19 syrioosh: does it boot if you power the BBB through the mini-usb port instead? Apr 21 16:53:37 yes it does Apr 21 16:54:02 but I need to use some usb device so I must have better power supply Apr 21 16:54:28 AFAIK micro usb connector will limit max current Apr 21 16:55:29 I dont know whether charger does some handshaking about max current Apr 21 16:55:34 it is charger from phone Apr 21 17:00:59 syrioosh: which distro? Apr 21 17:01:22 ubuntu 14 headless Apr 21 17:01:41 but it even doesnt reach kernel loading Apr 21 17:01:56 on my DC cable Apr 21 17:03:40 i think i need some resistors on D+ D- line to simulate device Apr 21 17:04:02 but not sure Apr 21 17:04:09 tbr: http://pastebin.com/BZ1PDncL. That's what I'm getting with dmesg. Apr 21 17:07:36 you're trying to power the bbb over the standard USB, not otg? Apr 21 17:08:57 thurgood: I am trying to use mobilephone charger to power bb via DC connector Apr 21 17:09:20 i have cable DC <-> USB A Apr 21 17:11:13 tomkbrady: looks like it is setting up things. so you should have a network Apr 21 17:12:22 syrioosh: mobilephone chargers are notoriously crap, especially if they are not from a known brand. Apr 21 17:15:26 I found something for those who are intrested Apr 21 17:15:29 http://www.maximintegrated.com/en/app-notes/index.mvp/id/4803 Apr 21 17:17:25 seems shorting d+ d- may help Apr 21 17:20:44 Algum brasileiro? Apr 21 18:40:13 Anyone know how to turn on logging for the ssh daemon in bbb angstrom? Apr 21 19:39:55 can anyone tell me if it's possible to get SDL 2 working on the black with open gl es/2 Apr 21 19:55:43 squishface should be possible Apr 21 19:55:50 with the sgx drivers Apr 21 20:04:19 I’m getting a compile error when trying to compile a kernel module on beaglebone debian make[2]: *** No rule to make target `arch/arm/tools/gen-mach-types', needed by `include/generated/mach-types.h'. Stop. Apr 21 20:04:25 am I missing a compiler flag? Apr 21 20:08:57 i have the SGX drivers installed, and the OpenGL ES demos work Apr 21 20:09:09 but SDL is reporting that i don't have an available video device Apr 21 20:11:21 squishface hm okay strace your application Apr 21 20:11:31 to see what SDL is doing Apr 21 20:11:48 Lightsword no Apr 21 20:12:28 Lightsword there is proably no kernel-headers package Apr 21 20:13:02 woglinde, I thought thats what it was at first but those are installed Apr 21 20:13:25 apt-get install linux-headers-$(uname -r) Apr 21 20:14:13 squishface which sdl version? Apr 21 20:18:35 thurgood sdl2 Apr 21 20:22:32 nothing shocking in strace Apr 21 20:22:50 just opening the libs Apr 21 20:26:05 sdl 2 doesn't work with fb if I recall Apr 21 20:27:37 previosus version does though Apr 21 20:28:32 i'm trying to use sdl 2 for the opengl backend Apr 21 20:28:46 i had 1.2 working already, just trying to get a performance boost Apr 21 20:36:51 squishface opengl? Apr 21 20:36:56 you mean egl/gles Apr 21 20:37:49 but I am off now Apr 21 20:47:03 woglinde, yes i do Apr 21 22:36:22 any idea what this error message is? https://gist.github.com/jameshilliard/9b5633101f7b97696be1 Apr 22 00:09:49 hi all Apr 22 00:12:50 i ahev got a question Apr 22 00:13:03 i updated kernel from 3.8 to 3.19 Apr 22 00:13:27 i can't make my outputs working Apr 22 00:13:39 here is my .dtsi file Apr 22 00:13:41 http://pastebin.com/iFysknH5 Apr 22 00:13:54 dtb-rebuilder compiles fine Apr 22 00:14:05 but after reboot Apr 22 00:14:18 there is no gpios in /sys/class/gpio/ Apr 22 00:14:25 what do I do wrong Apr 22 00:14:29 ? Apr 22 00:14:53 anybody? Apr 22 00:25:13 everybody in bed? Apr 22 01:18:04 kinkazu_: you should have four gpiochips in sys/class/gpio Apr 22 01:18:16 and more importantly, 'export' and 'unexport' Apr 22 01:20:05 echo the gpio number you want to access from userspace to /sys/class/gpio/export and voila, it should appear as a subdir (annoyingly they are numbered contiguously across the four banks, so gpio 1.21 would be called gpio53 ) Apr 22 01:32:41 * GenTooMan is faking sleeping mostly. Apr 22 01:33:32 Isn't that method of accessing GPIO rather slow? Apr 22 01:33:50 of course, sysfs is meant for config Apr 22 01:34:51 apart from the SPI their isn't any faster IO port apart from USB2 with 1 connector right? Apr 22 01:34:58 I memory-map the relevant GPIO controller, but of course eternal damnation awaits those who travel down the road of /dev/mem ... or so some people seem to think Apr 22 01:36:01 direct gpio is pretty fast... you can set or clear any subset of the 32 pins of a single GPIO bank using a single 32-bit write (the GPIO controller has "atomic set" and "atomic clear" registers) Apr 22 01:36:40 zmatt: i wanted to have them automatically loaded into /sys/class and activated high low Apr 22 01:36:51 same as I had in 3.8 kernel and cape manager Apr 22 01:37:30 kinkazu_: I have no experience with 3.8 and cape manager really... dunno if you can "pre-export" them using device-tree Apr 22 01:38:00 it is possible in 3.8 Apr 22 01:38:09 in any case, exporting them (if not already exported) and echoing "high" or "low" to their direction (that sets direction to output and also sets the level) Apr 22 01:38:17 so... use the same thing? Apr 22 01:38:23 :D Apr 22 01:38:26 yes yes Apr 22 01:38:40 but I wante dto have this done during bootup Apr 22 01:38:44 as soon as possible Apr 22 01:38:48 (how btw? I saw something about "gpio hogs" but I haven't looked into it yet) Apr 22 01:38:52 and not through init scripts Apr 22 01:38:52 yes, hence device tree Apr 22 01:39:15 note that bone capemgr stuff was also loaded by an init script Apr 22 01:39:25 if you want it early, put it in your main device tree Apr 22 01:39:58 and that is how it is in 3.19? Apr 22 01:40:11 all in one Apr 22 01:40:14 dtb Apr 22 01:40:42 unless the 3.19 you're using has been patched to support dynamic loading, yes Apr 22 01:41:08 I think the old capemgr thing is only maintained in 3.8, and the new dynamic overlays require a 4.0 kernel Apr 22 01:41:37 but to be honest, I've never had much need for overlays myself Apr 22 01:41:44 I installed 4.0 kernel, but was scared Apr 22 01:41:53 dtb-rebuuilder is for 3.19 Apr 22 01:41:58 I am so lame Apr 22 01:42:03 what's "dtb-rebuilder" ? Apr 22 01:42:12 you see Apr 22 01:42:26 you can do what you want with device tree Apr 22 01:42:29 * zmatt just changed and recompiled the mainline dts Apr 22 01:42:30 i need scripts Apr 22 01:42:32 :D Apr 22 01:42:44 git clone https://github.com/RobertCNelson/dtb-rebuilder Apr 22 01:42:50 well I have a Makefile for it... I'm lazy too ;P Apr 22 01:42:55 I feel so pity for myself Apr 22 01:43:02 I have spen 3 hours and I am nowhere Apr 22 01:43:09 and it is 3am Apr 22 01:43:17 almost Apr 22 01:43:31 only 3 hours? why are you complaining? Apr 22 01:43:40 moaybe 4 Apr 22 01:43:53 more to come Apr 22 01:44:00 this suppouse to be the easiest part Apr 22 01:44:04 I have other problem Apr 22 01:44:15 in both 3.8 and 3.19 Apr 22 01:44:19 bbb crashes Apr 22 01:44:30 when i am scanning can0 Apr 22 01:44:40 building my own device tree was something I also deferred for a long time because it seemed complicated at first Apr 22 01:44:46 and doing something with uart4 or uart2 Apr 22 01:44:55 until I just did it and it turned out a piece of cake Apr 22 01:45:01 eh Apr 22 01:45:04 maybe someday Apr 22 01:45:08 as for now Apr 22 01:45:11 black magic Apr 22 01:45:20 was easy in 3.8 Apr 22 01:45:36 it is probably easy in 3.19 Apr 22 01:45:42 to be honest, those overlays used in 3.8 ... *those* I consider black magic Apr 22 01:45:43 but not for me Apr 22 01:45:52 horrible to read Apr 22 01:46:00 you are just kind Apr 22 01:46:10 and you are trying to stop my tears Apr 22 01:46:13 :) Apr 22 01:46:13 often not preprocessed, so people used raw hex constants for things like pinmux Apr 22 01:46:35 hmm Apr 22 01:46:42 maybe someday Apr 22 01:47:28 I will try to load prepared dtsis Apr 22 01:47:42 and will chek which one does export Apr 22 01:47:46 automatically Apr 22 01:47:51 maybe tomorrow Apr 22 01:48:09 and will move to 4.0 Apr 22 01:48:43 will have to learn more about device tree anyway Apr 22 01:49:07 thanks zmatt , its time fo rme to go Apr 22 01:49:43 * kinkazu_ is sleeping Apr 22 01:52:29 You know KotH is right their are lots of things you would want to use an FPGA for but can't without using some proprietary tool. Apr 22 02:08:43 kinkazu_: hold on... (if still here) Apr 22 02:09:02 I installed the Xilinx IDE once.... Apr 22 02:10:46 * GenTooMan congratulates zmatt. "that the WEB tool one of the 'windows' based one?" Apr 22 02:11:28 hm, kinkazu_ really gone it seems... oh well, I wanted to show him a snippet of the DTS I'm using currently (along with my syntax highlighting rules) -> http://gerbil.xs4all.nl/snippet.dts.html Apr 22 02:12:02 coworker mentioned it was the first time he could actually sort of read DTS (compared with the cape/overlay things which tend to look like voodoo) Apr 22 02:12:15 ehh, ISE or something? Apr 22 02:13:01 they also installed their new Java-based thing (Vivante?), tried it briefly but couldn't get it to even start Apr 22 02:13:36 ultimately during the time I was playing with it I ended up mostly using the commandline tools :P Apr 22 02:13:55 I know what you mean Apr 22 02:14:44 especially since the target was a 32-macrocell Coolrunner II.... like, I'd actually prefer to fill that one in _manually_ Apr 22 02:15:27 Heh yeah cause you are talking what a 44 pin TQFP or something? Anyhow I need sleep have headache and my eyes are twitching general bad sign for not enough sleep as ussual. Stayed up late reading my book on SDR. Apr 22 02:15:33 (I actually managed to reveng a large part of the programming bitstream for it Apr 22 02:15:34 ) Apr 22 02:15:48 it's what's on the XDS100v2 JTAG debugger Apr 22 02:16:51 with a little tweaking, it probably significantly work around the deficiencies of the FTDI FT2232H chip Apr 22 02:17:54 xc2c32_vq44 I think Apr 22 02:18:16 do you happen to know if I can find the BBB component layout in PDF form somewhere? Apr 22 02:18:44 all I can find are those stupid allegro files, and the low-rew images in the System Reference Manual (which actually seem to be screenshots) Apr 22 02:18:47 they have the pick and place files don't they? Apr 22 02:19:06 don't they have the gerber plots etc? Apr 22 02:20:02 I will look tomorrow morning having a hard time seeing too much late night reading on SDR :) Apr 22 02:21:50 I mean something like http://gerbil.xs4all.nl/bbb-top.png Apr 22 02:21:52 but PDF Apr 22 02:22:00 (sorry for the red color, that's my fault) Apr 22 02:22:28 pl Apr 22 02:22:30 *(ok Apr 22 02:22:39 I still got stuff to do too Apr 22 02:23:05 just seems strange it'd so hard to find a pdf instead of some weirdass format :/ **** ENDING LOGGING AT Wed Apr 22 02:59:58 2015