**** BEGIN LOGGING AT Mon Jul 07 02:59:59 2014
Jul 07 04:49:02 <zmatt> hey, they re-added the PRU chapter to the TRM
Jul 07 04:49:46 <av500> \o/
Jul 07 04:53:18 <zmatt> I spot differences with the separately released (but "unsupported") am335xPruReferenceGuide though
Jul 07 04:55:31 <zmatt> both more info and less info, in different areas
Jul 07 06:43:07 <mmilano> i got the panda wifi card working sorta. i can ping my router and google from the beaglebone, but i cannot ping the beaglebone or ssh to the wlan0 ip from another system on my network.  when i’m wired up on eth0, i can access the bbb just fine from other systems.
Jul 07 06:43:41 <mmilano> any ideas on why i can’t hit the bbb from another system via the wifi interface?
Jul 07 06:52:45 <dey> is the only way to access gpio pins really via writing High/low into the pin related file in /sys/...? Isnt that awefully slow?
Jul 07 06:53:26 <dey> awfully*
Jul 07 06:55:59 <av500> no
Jul 07 06:56:26 <av500> if you need high speed with somewhat reliable timing, consider a kernel driver
Jul 07 06:57:36 <dey> av500: a kernel driver allows me to directly access the registers?
Jul 07 06:58:29 <av500> yes
Jul 07 06:58:37 <av500> you can also access the registers from user space
Jul 07 06:58:49 <av500> but thats kind of hacky and you will get user space timing
Jul 07 07:00:13 <dey> i dont really need the speed. But it sounds interesting. I think i look into it after i finish the driver
Jul 07 07:01:04 <dey> does writing into the gpio file create an interrupt? Or is the kernel constantly checking for changes in the folder?
Jul 07 07:01:14 <av500> neither
Jul 07 07:01:21 <av500> its a virtual filesystem
Jul 07 07:01:30 <av500> it ends directly inside the kernel
Jul 07 07:01:57 <dey> that means i basically give a value directly to a kernel function?
Jul 07 07:03:00 <av500> yes
Jul 07 07:04:04 <dey> i thought i write something into a file on my disk :P which gets parsed, whenever the kernel feels like :D
Jul 07 07:04:30 <dey> needless to say, i was amazed that it actually works
Jul 07 07:07:20 <woglinde> gm
Jul 07 07:10:38 <KotH> hoi woglinde
Jul 07 07:10:50 <KotH> alles klar im grossen kanton im norden?
Jul 07 07:10:53 <woglinde> hi koth
Jul 07 07:11:12 <woglinde> koth we will see it tomorrow
Jul 07 07:11:45 <KotH> what is going to happen tomorrow?
Jul 07 07:11:50 <dey> ein schweizer :)
Jul 07 07:12:16 <KotH> dey: natürli! was häsch denn du dänkt? das i en schoob seg?
Jul 07 07:46:43 <mmilano> i have wlan0 configured, but cannot ping the ip assigned to it. is there something else to configure so i can ping it like i can eth0 when i’m wired up?
Jul 07 07:50:07 <faber> hi everybody!
Jul 07 07:50:31 <faber> is there a I2C cape for beaglebone?
Jul 07 07:56:23 <maquefel> 2faber for black bleaglebone?
Jul 07 07:57:06 <maquefel> anyway for what you have 2 i2c ports on both beagles
Jul 07 08:01:45 <maquefel> 2faber https://learn.adafruit.com/adding-a-real-time-clock-to-beaglebone-black/overview
Jul 07 08:14:05 <faber> 2maquefel thank you! Not black one, only beaglebone. i know BB has a i2c bus, but do i need a special CAPE to exit with a standard i2C connector or do i have it directly on the BB master board ?
Jul 07 08:14:44 <faber> i dont understand this reading around on the net...
Jul 07 08:14:58 <maquefel> 2faber you can make one
Jul 07 08:15:03 <maquefel> buy protocape
Jul 07 08:15:17 <maquefel> and use this drawing to connect i2c device
Jul 07 08:15:21 <maquefel> as an example
Jul 07 08:15:28 <maquefel> file:///home/maquefel/Downloads/RTC_Cape_A1_Schematic.pdf
Jul 07 08:15:38 <maquefel> opss
Jul 07 08:15:39 <maquefel> soz
Jul 07 08:15:47 <maquefel> https://github.com/CircuitCo/RTC-Cape/blob/rev_a1/RTC_Cape_A1_Schematic.pdf?raw=true
Jul 07 08:15:49 <maquefel> whis one
Jul 07 08:16:20 <maquefel> and you don't need a special cape
Jul 07 08:20:04 <faber> 2maquefel : thank you again ! bye
Jul 07 09:50:55 <luastoned> hey, does anyone know what happened to this project http://beagleboard.org/project/BBB/ ?
Jul 07 10:02:41 <av500> luastoned: like many of these projects it went nowhere
Jul 07 10:05:12 <mkad> what projects?
Jul 07 10:05:22 <sumit_> i have been assigned the task of minimising the rootfs can some1 lead me to from where should i start ....?
Jul 07 10:06:16 <Letothe2nd> sumit_: use apt-get remove
Jul 07 10:06:33 <mkad> sumit_, debian bootstraping probably and creating minimal bootable installation of Debian ?
Jul 07 10:06:47 <mkad> sumit_, also recompiling kernel to include only essential modules
Jul 07 10:07:05 <Letothe2nd> or if you need it way smaller, rebuild something minimal, be it either debootstrap, or yocto for example
Jul 07 10:07:13 <mkad> sumit_, I would start with that
Jul 07 10:08:12 <mkad> sumit_, but bare in mind, that for someone asking "where should I start" that might mean a lot of reading, learning etc... and experimenting
Jul 07 10:09:41 <mkad> sumit_, probably you will need to know how to set up flat device tree, bootloader, recompiler kernel, bootstrap minimal system, glue that all parts together
Jul 07 10:09:55 <Letothe2nd> well both ahve documentation, don't they? ;)
Jul 07 10:10:46 <sumit_> i am newbie to this field and this task assigned but i can learn no problem....
Jul 07 10:11:43 <sumit_> thnks guys by ur replies it seems this would be quite intresting....
Jul 07 10:12:02 <Letothe2nd> though, i suspect the task being incomplete anyways - why do you want to "minimize" and by what amount
Jul 07 10:14:15 <sumit_> currently i am using arago filesystem provided by ti (55M) and aim is to bring it down to 18M or so....
Jul 07 10:14:31 <Letothe2nd> hrhrhr, good luck.
Jul 07 10:14:36 <mkad> oh
Jul 07 10:14:37 <Letothe2nd> read up on yocto, then.
Jul 07 10:14:38 <mkad> good luck
Jul 07 10:15:27 <mkad> Letothe2nd, I think deboostrap alone is like 100mb
Jul 07 10:15:34 <Letothe2nd> mkad: yeah, probably.
Jul 07 10:16:07 <Letothe2nd> you'll need something along the busybox + udev lines, maybe including a standard bash. and yes, no loads of modules.
Jul 07 10:16:18 <mkad> exactly
Jul 07 10:16:41 <mkad> well with busybox there is not need for bash
Jul 07 10:16:45 * Letothe2nd wonders why one would need a 18mb rootfs for a BBB though
Jul 07 10:17:05 <Letothe2nd> or is this maybe more like some commercial thing and not beagle related at all? ;)
Jul 07 10:17:07 <mkad> Letothe2nd, maybe a company making router devices ^^
Jul 07 10:19:17 <dey> i cant see japanese/german characters in my putty window. I guess its an unicode problem? the question is, is it a linux problem or a putty problem :/
Jul 07 10:19:39 <Letothe2nd> dey: probably putty, set it to UTF8
Jul 07 10:19:56 <dey> how do i do that?
Jul 07 10:20:15 <Letothe2nd> dey: check puttys options ;)
Jul 07 10:21:37 <dey> Letothe2nd: hm its already set to utf8. Why would someone even not use utf8? I expected it to be unchangable.
Jul 07 10:22:00 <av500> everything is changeable
Jul 07 10:22:16 <dey> the characters are all shown as '?'. If that helps
Jul 07 10:22:43 <dey> av500: just not via easily accessable options :p
Jul 07 10:23:19 <av500> there is always a focussed ion beam....
Jul 07 10:24:05 * dey shrieks in terror
Jul 07 10:24:54 <dey> ööääü
Jul 07 10:25:02 <dey> wait they work
Jul 07 10:25:04 <dey> wtf
Jul 07 10:25:06 <dey> ß
Jul 07 10:25:38 <dey> or not
Jul 07 10:25:52 <dey> some seem to work some dont
Jul 07 10:29:01 <dey> japanese doesnt work at all
Jul 07 10:29:18 <maquefel> sumit_: start and end here http://www.linuxfromscratch.org/
Jul 07 10:30:03 <sumit_> thanks maquefel....
Jul 07 10:30:07 <Letothe2nd> maquefel: which is basically just the very verbose, manual and geeky version of just using yocto
Jul 07 10:30:25 <Letothe2nd> plus, without heavy modificaion, will get you nowhere near 18M
Jul 07 10:30:29 <maquefel> well it a linux
Jul 07 10:30:35 <maquefel> not some kind of distrib
Jul 07 10:31:10 * Letothe2nd still is waiting for some pointers why a <20M rootfs is needed for something beagle
Jul 07 10:31:18 <maquefel> and i think yocto started reading LFS
Jul 07 10:31:50 <Letothe2nd> maquefel: yes of course, its just an automated, reproductible way of archieving the same goals (heavily simplified)
Jul 07 10:31:58 <maquefel> special device on cape for example
Jul 07 10:32:05 <maquefel> boot over network
Jul 07 10:32:08 <maquefel> etc etc
Jul 07 10:32:22 <maquefel> crypted rootfs
Jul 07 10:32:35 <maquefel> saving space for video
Jul 07 10:32:40 <maquefel> on nand he he he
Jul 07 10:32:44 <Letothe2nd> maquefel: why don't we let himself/herself answer?
Jul 07 10:32:47 <maquefel> and die in a half year
Jul 07 10:33:12 <Letothe2nd> maquefel: because i still smell something fishy.
Jul 07 10:33:14 <maquefel> and the main reason for LFS - wipe out unneeded crap
Jul 07 10:33:28 <luastoned> well other question, is there any library that does SPI via mmap?
Jul 07 10:33:35 <maquefel> lol
Jul 07 10:33:46 <maquefel> linux question ?
Jul 07 10:34:01 <Letothe2nd> maquefel: sounds like you haven't really compared lfs/yocto... lfs comes with lots of bloat, at least in the stadard documentation.
Jul 07 10:34:03 <maquefel> if not linux then give me back my lol
Jul 07 10:34:11 <Letothe2nd> luastoned: hopefully not.
Jul 07 10:34:33 <maquefel> Letothe2nd: When it comes to embedded devices and Nand I DON'T USE ANY DISTRIB
Jul 07 10:34:37 <Letothe2nd> maquefel: in sharp contrast to yoctos core-image-minimal, for example
Jul 07 10:34:38 <maquefel> ok ;) ?
Jul 07 10:34:45 <luastoned> Letothe2nd: why is that? It should be way faster than normal access via /dev/spi?
Jul 07 10:34:53 <Letothe2nd> luastoned: nope.
Jul 07 10:35:10 <luastoned> Letothe2nd: okay, that's interesting :0
Jul 07 10:35:11 <Letothe2nd> luastoned: which operation do you expect to be faster?
Jul 07 10:35:13 <sumit_> Letothe2nd the final product uses very less disk space bbb is just the evaluation board for the same processor ...
Jul 07 10:35:13 <maquefel> spi driver is responsible mmap
Jul 07 10:35:26 <maquefel> not some kind of library
Jul 07 10:35:49 <maquefel> luastoned: let me guess eeprom ? :-D
Jul 07 10:35:49 <luastoned> ill report back after i've tried some things
Jul 07 10:36:25 <luastoned> maquefel: I haven't started yet.. but I want to get ~30k writes out over spi
Jul 07 10:36:35 <maquefel> so?
Jul 07 10:36:41 <maquefel> how it comes to mmap?
Jul 07 10:36:51 <Letothe2nd> luastoned: without naming a timeframe, that means nothing. 30k writes in one year is about nothing
Jul 07 10:37:01 <luastoned> yeah, in a second Letothe2nd
Jul 07 10:37:07 <Letothe2nd> luastoned: still not very much.
Jul 07 10:37:16 <maquefel> mmap won't give you anything actually
Jul 07 10:37:24 <maquefel> in this case
Jul 07 10:37:29 <Letothe2nd> except of course if "1 write" means like a megabyte.
Jul 07 10:37:50 <dey> still not very much
Jul 07 10:38:08 <Letothe2nd> dey: 30M/sec over standard spi is at least not too slow.
Jul 07 10:38:41 <luastoned> Well, I want to write 4 bytes to 3 chips each 30k times per second
Jul 07 10:39:05 <dey> i just kept your ignorance spree warm :3
Jul 07 10:39:17 <Letothe2nd> sounds like some dac thing, in which case you should read up on the IIO subsystem.
Jul 07 10:39:19 <maquefel> Letothe2nd: and this guy is writing a driver for ADC with 30 kSps
Jul 07 10:39:52 <luastoned> what exactly is the IIO subsystem?
Jul 07 10:40:03 <maquefel> forget about mmap the bottle neck is master/slave spi
Jul 07 10:40:17 <Letothe2nd> luastoned: people wanting to do mmap should be able to read up on a linux realted buzzword themselves... ;)
Jul 07 10:40:28 <maquefel> Letothe2nd: back to holy war!!!! why do you think youcto is better?
Jul 07 10:40:28 <dey> what does he do with 30kSps? Its to much for most things and far to less for the rest :/
Jul 07 10:40:40 <elinuxer> Hai , am working on beaglebone white to interface tca6416 io expander . I connected io expander pins to  i2c lines 19,20 in the beagle board. when i use this "i2cdetect -y -r 2" command am getting "[ 1858.615328] omap_i2c 4819c000.i2c: controller timed out" error. Am newbie to i2c and beaglebone. Any idea to solve this?
Jul 07 10:40:46 <Letothe2nd> maquefel: its reproductible.
Jul 07 10:40:48 <luastoned> Letothe2nd: sure, I was just referring to http://beagleboard.org/project/BBB/ .. although it being dead
Jul 07 10:41:05 <maquefel> Letothe2nd: like anything else...
Jul 07 10:41:12 <Letothe2nd> maquefel: and by reproductible i do not mean copypasting/typing lists of commands, but having it automated.
Jul 07 10:41:29 <cityoflights2> is this the place to ask about any one used a valve?
Jul 07 10:41:42 <maquefel> Letothe2nd: lol, do you differ bash script from python script or c program in this case?
Jul 07 10:41:50 <cityoflights2> I am seeking to use a small 1/4" electric valve
Jul 07 10:41:56 <cityoflights2> 12v
Jul 07 10:41:56 <Letothe2nd> maquefel: well, depends on the definition of reproductible. for mine, LFS is not. nor is debootstrap.
Jul 07 10:41:59 <maquefel> Letothe2nd: i can use gentoo =) and still build LFS
Jul 07 10:42:06 <Letothe2nd> maquefel: nor is gentoo.
Jul 07 10:42:15 <maquefel> Letothe2nd: it is
Jul 07 10:42:16 <dey> cityoflights2: doesnt sound very bbb related to me.
Jul 07 10:42:23 <cityoflights2> is there a place to download a BBB gentoo image?
Jul 07 10:42:30 <Letothe2nd> maquefel: use whatever floats your boat :)
Jul 07 10:42:39 <Letothe2nd> maquefel: and so will i.
Jul 07 10:42:48 <maquefel> Letothe2nd: so it is another distrib hole war
Jul 07 10:42:57 <Letothe2nd> maquefel: not at all.
Jul 07 10:43:12 <maquefel> Letothe2nd: and yocto is no better then anything else
Jul 07 10:43:19 <Letothe2nd> anyways, i'm not in the mood for that.
Jul 07 10:45:57 <maquefel> elinuxer: i2cdetect -y -r 1
Jul 07 10:47:14 <elinuxer> yes , i2cdetect -y -r 1 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
Jul 07 10:47:21 <maquefel> yes - what?
Jul 07 10:47:37 <maquefel> 19, 20 is first i2c bus not second
Jul 07 10:50:16 <elinuxer> maquefel :  i2cdetect -y -r 1, i cant set or dump using i2ctools. it is throwing error.
Jul 07 10:50:30 <maquefel> ahhhh
Jul 07 10:50:35 <maquefel> 5 sec
Jul 07 10:52:02 <maquefel> ls -l /sys/bus/i2c/devices/i2c-* ?
Jul 07 10:53:19 <elinuxer> maquefel : error "Error: Write start address failed, return code -1
Jul 07 10:53:20 <elinuxer> " while dumping
Jul 07 10:54:20 <elinuxer> maquefel :  ls -l /sys/bus/i2c/devices/i2c-* , getting i2c-0. i2c-1,i2c-2.
Jul 07 10:57:43 <maquefel> have you wired correctly?
Jul 07 10:58:00 <maquefel> 20 is SDA
Jul 07 10:58:07 <maquefel> 19 is SCL
Jul 07 10:59:08 <maquefel> what exactly distrib you are using?
Jul 07 11:00:42 <mkad> elinuxer, i2cdetect -y -r 1  or -r 0
Jul 07 11:01:05 <maquefel> 0 is closed
Jul 07 11:01:13 <maquefel> only eeprom there and no wires
Jul 07 11:01:18 <mkad> what ?
Jul 07 11:01:24 <mkad> I have two buses 0 or 1
Jul 07 11:01:28 <mkad> and both work fine
Jul 07 11:01:30 <maquefel> there 3
Jul 07 11:01:43 <mkad> that depends how your OS enumerates them
Jul 07 11:01:52 <maquefel> 0 is where eeprom hangs and it's wires are not exposed
Jul 07 11:02:04 <maquefel> 1 and 2 are avaiable
Jul 07 11:02:12 <mkad> ok but 0 can be 1 bus depends how OS enumerates
Jul 07 11:02:12 <maquefel> 2 should be mux enabled
Jul 07 11:02:34 <mkad> its not the same as hardware
Jul 07 11:03:21 <elinuxer> <maquefel> : i wired correctly. i checked again.
Jul 07 11:03:28 <maquefel> distrib?
Jul 07 11:03:44 <elinuxer> <maquefel>: how can i read device id
Jul 07 11:03:58 <mkad> elinuxer, by scanning with i2cdetect
Jul 07 11:04:40 <elinuxer> <mkad>i2cdetect -y -r 1, i using 19, 20 pins
Jul 07 11:04:53 <maquefel> i2cdetect doesn't work for him
Jul 07 11:05:09 <elinuxer> but am not getting id.
Jul 07 11:05:21 <mkad> well I use the same pins
Jul 07 11:05:24 <mkad> 10-th from the top
Jul 07 11:05:30 <mkad> and for me its bus 1
Jul 07 11:05:38 <mkad> so I use the same command
Jul 07 11:05:45 <mkad> but answer to maquefel what distro
Jul 07 11:05:50 <maquefel> cat /sys/devices/bone_capemgr.8/slots
Jul 07 11:06:05 <maquefel> it worked for me also
Jul 07 11:06:14 <maquefel> do you have common ground with bbb?
Jul 07 11:06:44 <elinuxer> i build custom kernel using yocto toolchain...
Jul 07 11:06:59 <maquefel> loooooooooool
Jul 07 11:07:04 <mkad> lmao
Jul 07 11:07:13 <mkad> fix your kernel :)
Jul 07 11:07:19 <elinuxer> cat /sys/devices/bone_capemgr.8/slots
Jul 07 11:07:19 <elinuxer>  0: 54:PF---
Jul 07 11:07:20 <elinuxer>  1: 55:PF---
Jul 07 11:07:20 <elinuxer>  2: 56:PF---
Jul 07 11:07:20 <elinuxer>  3: 57:PF---
Jul 07 11:07:35 <maquefel> lsmod
Jul 07 11:09:35 * maquefel kicked Letothe2nd go help with yocto :-D
Jul 07 11:09:53 <maquefel> last question
Jul 07 11:09:59 <maquefel> do you have common GND ?
Jul 07 11:10:08 <maquefel> bbb and your i2c device
Jul 07 11:10:49 <elinuxer> lsmod
Jul 07 11:10:49 <elinuxer>     Not tainted
Jul 07 11:10:49 <elinuxer> is there any entries for tca6416 in dsti file
Jul 07 11:12:52 <elinuxer> Am using beagleboard white , yes i have common ground.
Jul 07 11:14:05 <maquefel> i2cdetect -l
Jul 07 11:14:58 <elinuxer> i2cdetect -l
Jul 07 11:15:08 <elinuxer> i2c-0   i2c             OMAP I2C adapter                        I2C adapter
Jul 07 11:15:08 <elinuxer> i2c-1   i2c             OMAP I2C adapter                        I2C adapter
Jul 07 11:15:08 <elinuxer> i2c-2   i2c             OMAP I2C adapter                        I2C adapter
Jul 07 11:15:51 <maquefel> and i2cdetect -V
Jul 07 11:16:17 <elinuxer> i2cdetect version 3.1.0
Jul 07 11:16:23 <maquefel> and what are trying to wire
Jul 07 11:16:32 <maquefel> chip and mnifacturer
Jul 07 11:16:50 <maquefel> and uname -a
Jul 07 11:18:29 <elinuxer> i couldn't read device id , read or write using i2c tools, is this wiring issue.
Jul 07 11:18:32 <elinuxer> Linux beaglebone 3.8.13-00720-g465c7df-dirty #9 SMP Mon Jul 7 15:24:53 IST 2014 armv7l GNU/Linux
Jul 07 11:19:46 <elinuxer> i cant access tca6416, i doubt that is there any registration for tca6416 in dtsi file?
Jul 07 11:22:51 <maquefel> it is not a matter of dtsi
Jul 07 11:26:02 <maquefel> zcat /proc/config.gz | grep I2C_CHARDEV
Jul 07 11:29:09 <maquefel> and last idea try doing the same with device not connected
Jul 07 11:33:06 <elinuxer> zcat /proc/config.gz | grep I2C_CHARDEV
Jul 07 11:33:06 <elinuxer> CONFIG_I2C_CHARDEV=y
Jul 07 11:34:31 <elinuxer> getting the same result, device is not in connection
Jul 07 11:35:27 <maquefel> well run out ideas - sorry
Jul 07 11:35:36 <maquefel> you can try the same with Angstrom
Jul 07 11:35:39 <maquefel> if fails
Jul 07 11:35:58 <maquefel> then i2c circuit is malfunctioning
Jul 07 11:36:03 <maquefel> if succeds
Jul 07 11:36:08 <mkad> maquefel, he should try booting from sd card a standard distro
Jul 07 11:36:14 <maquefel> then yocto project if spoiled
Jul 07 11:36:30 <mkad> maquefel, its like 5min of work to flash image on it
Jul 07 11:36:35 <maquefel> well mkad is right - it is an easy way
Jul 07 11:37:22 <maquefel> and i am very suspicious about all i2c enabled by default
Jul 07 11:37:33 <maquefel> showing nothing in cat /sys/devices/bone_capemgr.8/slots
Jul 07 11:37:55 <maquefel> somebody messed with dbt
Jul 07 11:38:13 <elinuxer> oka, let me try Angstrom
Jul 07 11:38:32 <maquefel> and yocto guy dissapeared
Jul 07 11:39:01 <elinuxer> and then to yocto
Jul 07 11:42:27 <maquefel> ah
Jul 07 11:54:40 <Undertasker> Hi. Does anybody know if the PRUs can access the hardware SPI?
Jul 07 12:06:47 <maquefel> nope
Jul 07 12:08:14 <zmatt> sure, why not?
Jul 07 12:08:29 <zmatt> PRU has access to nearly the whole L3
Jul 07 12:13:33 <zmatt> any access to an address >= 0x80000 is routed to the L3, and the table in chapter 10 shows it should have access to EMIF, OCMC, EDMA, GPMC, ADC/TSC, L4_Per, L4_Fast, L4_Wkup, DebugSS, L3 config registers, and the ever-mysterious "expansion slot"
Jul 07 12:15:49 <zmatt> of course making any access to the L3 means sacrificing the rigid predictable timing PRU normally has
Jul 07 12:18:09 <Undertasker> It's more a matter of getting the SPI receive register read before the next byte is received.
Jul 07 12:19:39 <zmatt> PRU interrupt controller receives an irq from McSPI0
Jul 07 12:19:56 <Undertasker> I want to sniff and decode the SPI communication between a micro and a sensor. So I want to receive MOSI and MISO data simultaneously, log the configuration data comming from the micro, decode the sensor data in realtime and send it out as a PWM value.
Jul 07 12:21:22 <Undertasker> Or output it as a 16bit parallel value, maybe.
Jul 07 12:22:43 <zmatt> I'm not really familiar with McSPI, dunno if you can configure both data pins as input
Jul 07 12:23:39 <Undertasker> The buspirate can sniff SPI communication, but it is ways too slow for a 10MHz SPI bus.
Jul 07 12:24:05 <zmatt> 10 MHz ?  sounds like PRU's direct GPIO could handle that easily
Jul 07 12:26:18 <maquefel> http://hackaday.com/2013/12/07/speeding-up-beaglebone-black-gpio-a-thousand-times/
Jul 07 12:27:06 <zmatt> maquefel: that's still using the normal GPIO from the ARM
Jul 07 12:27:14 <zmatt> which is definitely too slow
Jul 07 12:30:57 <maquefel> just curios
Jul 07 12:30:59 <zmatt> (especially he's just doing back-to-back writes, which get buffered so you're not seeing the effects of interconnect latency... reads however definitely will)
Jul 07 12:31:06 <maquefel> why PRU ?
Jul 07 12:31:06 <maquefel> Up to Two Master and Slave McSPI Serial Interfaces
Jul 07 12:31:06 <maquefel> Up to Two Chip Selects
Jul 07 12:31:07 <maquefel> Up to 48 MHz
Jul 07 12:32:07 <maquefel> can't we connect the bbb to the end of spi bus and get all traffic?
Jul 07 12:32:18 <zmatt> yeah, using both McSPIs would work too
Jul 07 12:33:13 <zmatt> only one of them gives an interrupt to PRU which is apparently where Undertasker wants to process the data, though if both McSPIs are running from the same clock and CS you only need an irq from one of them I guess
Jul 07 12:33:36 <av500> you can also use the PRU to sample the raw bus data
Jul 07 12:33:44 <av500> and extract SPI from thatz
Jul 07 12:33:45 <zmatt> that's what I said earlier
Jul 07 12:33:46 <av500> that
Jul 07 12:34:10 <av500> PRU is 200MHz, your spi is 10MHz?
Jul 07 12:35:05 <zmatt> hence the "sounds like PRU's direct GPIO could handle that easily"
Jul 07 12:35:23 <maquefel> well...
Jul 07 12:35:45 <maquefel> a logging oscilloscope?
Jul 07 12:40:48 <BennyB_> Hi who can help me with OneWire - GPIO? I found this posting which tells from some changes in (i think kernel module): http://e2e.ti.com/support/arm/sitara_arm/f/791/p/264704/995958.aspx#995958
Jul 07 12:40:55 <BennyB_> how can I do this?
Jul 07 12:42:32 <BennyB_> I found some source of this file here: https://code.google.com/p/beagle-borg/source/browse/kernel/arch/arm/mach-omap2/board-am335xevm.c
Jul 07 12:42:55 <BennyB_> but what do I have to compile? Do I need to recompile kernel?
Jul 07 12:46:25 <KotH> BennyB_: step #1: write code
Jul 07 12:46:26 <av500> BennyB_: board files are a thing of the past
Jul 07 12:46:32 <KotH> BennyB_: step #2: debug code
Jul 07 12:46:38 <KotH> BennyB_: step #3: ...
Jul 07 12:46:42 <KotH> BennyB_: step #4: profit
Jul 07 12:47:13 <KotH> BennyB_: beside: doesnt the kernel already have a generic 1-wire driver?
Jul 07 12:47:30 <av500> BennyB_: the angstrom and debian kernels dont use board files any more
Jul 07 12:47:33 <av500> but device tree
Jul 07 12:47:36 <BennyB_> sure I can code somehow in C++.
Jul 07 12:47:48 <av500> for that e2e post to apply, you need to use some TI frankenkernel
Jul 07 12:47:52 <BennyB_> I'm using debian 3.8.13
Jul 07 12:48:25 <mkad> av500, is programing pru difficult?
Jul 07 12:48:29 <BennyB_> the devbicetree shows me no slaves when using cat /sys/devices/w1_bus_master1/w1_master_slave_count
Jul 07 12:48:51 <av500> mkad: no idea, never did it
Jul 07 12:48:58 <av500> but there are many example already available
Jul 07 12:49:05 <BennyB_> If this posting mentioned in TI forum is obsolete I need to search anywhere else
Jul 07 12:49:06 <av500> so find the closest one and modify
Jul 07 12:49:34 <av500> https://www.google.com/search?q=beaglebone+1+wire
Jul 07 12:49:47 <BennyB_> yea, I did search alot and find one matching solution but whatever I do no DS18S20 slave is detected
Jul 07 12:49:48 <mkad> av500, I am working on drone with beagle board
Jul 07 12:50:06 <mkad> av500, but I get freezes when reading from i2c so maybe I should use pru..
Jul 07 12:50:24 <av500> freezes?
Jul 07 12:50:41 <mkad> yeah reading and printing to terminal coordinates stops
Jul 07 12:50:47 <mkad> for like 1.5s
Jul 07 12:50:53 <mkad> and than again goes on
Jul 07 12:50:59 <av500> find out why
Jul 07 12:51:00 <BennyB_> av500 people on the TI-forum post did also complain that there is no connection to DS18S20 slaves, so I thought this must be the solution
Jul 07 12:51:09 <av500> BennyB_: no idea
Jul 07 12:51:44 <BennyB_> av500 but the board files are out to date with debian, am I right?
Jul 07 12:51:51 <av500> yes
Jul 07 12:51:54 <BennyB_> ok
Jul 07 12:52:02 <av500> http://www.bonebrews.com/temperature-monitoring-with-the-ds18b20-on-a-beaglebone-black/
Jul 07 12:52:23 <BennyB_> I know this one. but my slaves are not detected
Jul 07 12:52:33 <av500> sorry, no idea
Jul 07 12:52:39 <av500> does the pin toggle?
Jul 07 12:52:44 <av500> looked at it with a scope?
Jul 07 12:52:48 <BennyB_> I evern installed OWFS (one wire file system), but it's the same, no slaves are detected
Jul 07 12:53:15 <BennyB_> no I didn't look at it with the scope
Jul 07 12:53:40 <mkad> av500, thats a nice link you posted
Jul 07 12:54:37 <BennyB_> the bonebrews link is very good, this one is also not bad: http://hipstercircuits.com/dallas-one-wire-temperature-reading-on-beaglebone-black-with-dto/
Jul 07 12:55:04 <av500> dont thank me, thank google:  https://www.google.com/search?q=beaglebone+1+wire
Jul 07 12:56:49 <BennyB_> but you must be careful for me the pin P8.11 didn't work. it was already occupied by emmc. I musing now P9.15
Jul 07 12:57:57 <BennyB_> just to vfferify: is my dts file correct for my purpose? http://pastebin.com/BH0sZL85
Jul 07 12:58:41 <BennyB_> line 30 confuses me
Jul 07 12:58:58 <BennyB_> but there is no error while loading/booting
Jul 07 13:01:47 <BennyB_> here is dmesg output. something is strange at the end: http://pastebin.com/hRXFaA6F
Jul 07 13:03:27 <av500> [  358.737560] of_get_named_gpio_flags: can't parse gpios property
Jul 07 13:03:38 <av500> so its not good
Jul 07 13:03:40 <BennyB_> what does it mean?
Jul 07 13:04:40 <BennyB_> dts file says: gpios = <&gpio1 16 0>; /*grrr I think this means gpio1_16 (using 1 to 4 instread of 0-3)*/
Jul 07 13:05:00 <BennyB_> should it be ... &gpio0 16 0>;  ??
Jul 07 13:10:50 <BennyB_> i ust looked GPIO_1_16 is P9.16, so maybe my dts file is wrong, it claims exclusive use for P9.15
Jul 07 13:11:29 <BennyB_> owch, totally confused now: https://github.com/jadonk/bonescript/blob/master/src/bone.js
Jul 07 13:11:56 <BennyB_> line 1045 says GPIO1_16 is P9_15
Jul 07 13:13:40 <BennyB_> where are the device tree freaks? ;-)
Jul 07 13:14:04 <BennyB_> mux reg offset 0x40 is correct?
Jul 07 13:14:59 <BennyB_> how can I be sure in my dts file this is correct for 1wire? " 0x040 0x37 /*pin P9_15 input with pullup mode 7 - w1-gpio */"
Jul 07 13:15:25 <BennyB_> the second number "0x37" how can I change is to NOT use pullup?
Jul 07 13:41:34 <Jarrett__> yo
Jul 07 14:34:56 <mmilano> still stumped. when i enable my wifi interface on the BBB, it acts like there’s a firewall blocking incoming traffic. when i’m wired up, incoming works fine. any ideas on this? i’m using a panda wireless usb card + the latest debian.
Jul 07 15:29:22 <maquefel_> a really idiotic question - tssop and ssop has any difference apart of height?
Jul 07 18:30:07 <gwilson> Anyone have any experience cross compiling from ubuntu to BBB debian with QT4 installed?
Jul 07 22:38:13 <MrYaah> Is the ethernet port on the beaglebone black a dedicated ethernet port or is it shared with the usb like the raspberry pi
Jul 07 22:39:20 <mkad> MrYaah, what do you mean ? I can see two devices usb0 and eth0 on my BBB
Jul 07 22:43:10 <MrYaah> mkad: I'm just wondering if they share a bus or something like that. Basically if im using ethernet and usb to full capacity for different purposes would the total capacity be the 100mbps for the ethernet + whatever the max for usb is or would it be just the max for usb
Jul 07 22:45:42 <MrYaah> mkad: tbh im not really sure what the difference is all i know is that the hummingboard video comparison against the raspberry pi https://www.youtube.com/watch?v=xhFWedNbHHI claims a dedicated ethernet port as one of its advantages over the pi
Jul 07 22:48:33 <mkad> MrYaah, us hummingboard open-hardware? unlike pi ?
Jul 07 22:48:35 <mkad> is?
Jul 07 22:49:21 <MrYaah> mkad: I am not sure, I think it might be
Jul 07 22:49:42 <MrYaah> mkad: actually i dont know at all, some stuff about it is open, not sure about hardware
Jul 07 22:49:42 <mkad> rpi is not afaik
Jul 07 23:03:36 <mkad> MrYaah, do you know if this hummingboard has the same format as old rpi ? does it fit into the same case?
Jul 07 23:08:06 <mkad> by the way
Jul 07 23:08:20 <mkad> if there are caps for beagle bone white, do they also  work on bbb ?
Jul 08 00:09:23 <chris37879> Anyone know of a way to connect the beaglebone black to an android device via OTG and not have it disable the device's network connection?
Jul 08 00:09:43 <chris37879> Android sees the device as a network connection and routes all my traffic over that instead of my Wifi
Jul 08 00:28:02 <ddaysword> So, I have a BBB rev B and an atrix lapdock (pictures of the back and front at http://chinetti.me/bbb). The lapdocks native resolution is 1366x768, but the bbb is rejecting those modes.
Jul 08 00:28:15 <ddaysword> Any idea why?
Jul 08 00:29:10 <ddaysword> It _does_ work at 1360x768, but that causes slight aliasing, which is the /worst!/
Jul 08 00:35:23 <ddaysword> Connection to chinetti.me closed.
Jul 08 00:35:23 <ddaysword> alarm P F CHINETTI ~:
Jul 08 00:35:24 <ddaysword> 450 >  cat /proc/cmdline
Jul 08 00:35:35 <ddaysword> well.
Jul 08 00:35:46 <ddaysword> meant to paste my /proc/cmdline
Jul 08 00:36:00 <ddaysword> but... I plugged in my mouse in between
Jul 08 00:36:07 <ddaysword> console=ttyO0,115200n8 quiet drm.debug=7 capemgr.disable_partno=BB-BONELT-HDMI video=HDMI-A-1:1360x768MR@60 root=/dev/mmcblk0p2 rw rootfstype=ext4 rootwait fixrtc
Jul 08 00:52:56 <palesius> anyone have any idea how to disable/reenable usb host port (such that it will stop providing power to device) on the BBB through code/command line?
Jul 08 01:05:07 <rcn-ee> "sudo halt" will shutdown the port. ;)  it's wired to the 5volt.. not going to happen..
Jul 08 01:16:44 <ds2> you could disconnect it and put in a PFET
Jul 08 01:36:11 <mmilano> anyone using the panda wireless on the BBB?
Jul 08 01:49:13 <router> When I use media-ctl, the pad fmt doesn't show up. anyone can help
**** ENDING LOGGING AT Tue Jul 08 02:59:58 2014