**** BEGIN LOGGING AT Wed Nov 11 02:59:58 2015
Nov 11 03:52:58 <brycecater> hi all
Nov 11 03:55:06 <brycedcarter> is anyone here using debian with the 4.1 kernel?
Nov 11 04:02:14 <brycecater> I am running into a weird issue when I try to use sysfs to control the GPIOs in bank 2
Nov 11 04:03:10 <brycecater> it lets me change the value between 1 and 0, but the actual pin on the header does not change state
Nov 11 04:36:47 <zmatt> brycecater: you're sure the pins are muxed correctly?
Nov 11 04:41:52 <brycecater> zmatt: hmmm... i thought so, but how can i double check. I have a custom cape that I am loading
Nov 11 04:41:53 <zmatt> (to check you can use http://gerbil.xs4all.nl/show-pins.pl.gz )
Nov 11 04:43:41 <brycecater> brycecater: ok, thanks, let me give that a try
Nov 11 04:44:49 <zmatt> it's just a little script that makes the pinctrl data from debugfs a bit more human-readable
Nov 11 04:45:47 <brycecater> dude! that script is AWESOME!
Nov 11 04:47:25 <brycecater> brycecater: i think i understand what is happening now... i think you were right, the pins are not muxed correctly. I am loading a custom cape, but it looks like the lcd cape is already controlling those pins
Nov 11 04:47:41 <zmatt> (you can also pass -v to show more pins or -v -v to show... even more pins)
Nov 11 04:47:55 <zmatt> you realize you're addressing yourself ? ;)
Nov 11 04:48:08 <zmatt> and such a conflict *should* give an error message
Nov 11 04:48:31 <brycecater> yeah... oops ;)
Nov 11 04:48:44 <zmatt> (upon loading of the custom cape)
Nov 11 04:49:07 <zmatt> (or the lcd cape, depending on who's first)
Nov 11 04:53:20 <Set_> Hello...
Nov 11 04:53:54 <brycecater> yeah, i just changed from angstrom to debian. on angstrom I had to unload the HDMI cape that was loaded by default. but now on debian the cape manager does not show that any are loaded by default. so i cannot unload the HDMI cape
Nov 11 04:54:35 <zmatt> if you have a recent filesystem image there should be comments in /boot/uEnv.txt
Nov 11 04:54:48 <zmatt> there are a few different base dts files now for a couple of cases
Nov 11 04:55:26 <zmatt> dtb=am335x-boneblack-emmc-overlay.dtb
Nov 11 04:55:46 <zmatt> that disables HDMI
Nov 11 04:56:48 <Set_> Does anyone deal with Xively?
Nov 11 04:59:21 <Set_> I am waiting for Xively people to get back to me.
Nov 11 04:59:31 <Set_> I just wanted to know if anyone knew anyone there...
Nov 11 05:00:29 <Set_> Hold that thought...I have to go and rip off one of those Window Sensor.
Nov 11 05:00:51 * zmatt isn't really putting any thought on hold
Nov 11 05:06:45 <Set_> Yes...
Nov 11 05:06:53 <Set_> Those dang sensors are hard to get off.
Nov 11 05:07:09 <Set_> I am just going to purchase one.
Nov 11 05:07:21 <Set_> I do not want to break the window.
Nov 11 05:08:57 <brycecater> zmatt: yeah, I see the comments in that file, do i just need to uncomment the emmc one? that is what I did and it did not seem to change the output of that script that you sent. should the one that I uncommented appear in the cape manager? i does not seem to be there
Nov 11 05:09:48 <zmatt> the "dtb=am335x-boneblack-emmc-overlay.dtb" line should appear in the uEnv.txt file
Nov 11 05:10:14 <zmatt> capemgr isn't involved anymore for the on-board logic (as you noticed, since there wasn't a cape to unload)
Nov 11 05:12:16 <brycecater> ok, gotcha... so I uncommented the "dtb=am335x-boneblack-emmc-overlay.dtb" line the the uEnv.txt file... should that change the output of the your show-pins script?
Nov 11 05:12:55 <zmatt> after reboot, yes
Nov 11 05:15:16 <Set_> I got it dude. I got me a window sensor.
Nov 11 05:16:52 <Set_> Now...I get to use my BBB as a alert mechanism.
Nov 11 05:17:09 <Set_> I am going to play Bob Marley when someone opens the door.
Nov 11 05:17:32 <brycecater> zmatt: something was funky after the first reboot :P now it is working great! THANK YOU SO MUCH!! this was causing such a headache
Nov 11 05:17:34 <Set_> Maybe someone else.
Nov 11 05:18:08 <Set_> I smashed my thumb in the process but it was worth it.
Nov 11 05:18:19 <brycecater> also, that script is totally great, I will treasure it always ;)
Nov 11 05:20:25 <zmatt> :)  hmm, maybe I should add columns for the current state of the pin.  and possible some checks and warnings
Nov 11 05:20:30 <zmatt> *possibly
Nov 11 05:21:02 <Set_> Hey Zmatt...
Nov 11 05:21:21 <zmatt> it's still a lowercase z
Nov 11 05:21:41 <Set_> I noticed that holding down the boot button on my BBB does nothing when loading a new image.
Nov 11 05:21:46 <Set_> sorry. zmatt.
Nov 11 05:22:06 <Set_> It just makes the SD card the system.
Nov 11 05:22:53 <Set_> When I do not hold down the boot button when applying a new image, I get it downloaded on my board quickly and correctly.
Nov 11 05:23:14 <Set_> I done "learn-a-did."
Nov 11 05:24:23 <zmatt> you're not making much sense... holding down the boot button during power-on changes the boot device order from { eMMC, μSD, uart, usb } to { spi, μSD, usb, uart }
Nov 11 05:24:46 <zmatt> i.e. it lets you bypass booting from the eMMC
Nov 11 05:24:59 <zmatt> it has no other effect
Nov 11 05:25:13 <Set_> Oh.
Nov 11 05:36:53 <Set_> Okay...I have my sensor connected.
Nov 11 05:37:15 <Set_> I will be back.
Nov 11 05:37:21 <Set_> I have to write a thing or two.
Nov 11 06:28:55 <brycecater> zmatt: hmmmm... i am still having issues. did you mean that the cape manager is not used at all anymore?
Nov 11 06:29:03 <zmatt> it's used for capes
Nov 11 06:29:07 <zmatt> actual ones
Nov 11 06:29:55 <zmatt> I don't know why it's not used anymore for the BBB's "virtual capes", or how that worked to begin with anyway
Nov 11 06:30:14 <brycecater> so i have a dts that i made and compiled into a dtbo... i have been trying to load it using the cape manager
Nov 11 06:30:50 <brycecater> I dont get any errors but loading it into slots does not seem to have any effect
Nov 11 06:31:07 <zmatt> you checked dmesg for error messages?
Nov 11 06:32:05 <zmatt> (note btw that 4.1 also has a shiny new mechanism to load overlays via configfs; using capemgr for that is technically slight abuse if the overlay isn't actually for a cape)
Nov 11 06:32:43 <zmatt> (unlike capemgr the configfs mechanism doesn't require you to make up bogus cape-metadata)
Nov 11 06:36:46 <brycecater> is there any documentation on how to use configfs?
Nov 11 06:37:13 <zmatt> yeah, in the kernel docs as usual (somewhere in Documentation/devicetree/ )
Nov 11 06:37:18 <zmatt> it's easy enough though
Nov 11 06:37:59 <zmatt> go to /sys/kernel/config/device-tree/overlays
Nov 11 06:38:30 <zmatt> make a new directory (pick any name you like)
Nov 11 06:39:33 <zmatt> inside it you find a 'status' property (still saying "unapplied" if you read it at this point) and two writeable ones: dtbo and path
Nov 11 06:39:45 <zmatt> you can either write the path to your dtbo to 'path'
Nov 11 06:39:52 <zmatt> or the content of your dtbo to 'dtbo'
Nov 11 06:40:04 <zmatt> your choice
Nov 11 06:40:46 <zmatt> if all is well, status will change to "applied"
Nov 11 06:41:33 <zmatt> dtbo and path become immutable, but you can read back what you've written.  to remove the overlay, rmdir the directory
Nov 11 06:42:18 <zmatt> dtbo is still the same as for capemgr, except no metadata is required (and any metadata present is ignored)
Nov 11 06:45:24 <brycecater> zmatt: sorry, i am still building up my sysfs knowledge... can i do "echo /lib/firmware/BB-PWM2A-00A0.dtbo > path"
Nov 11 06:45:44 <zmatt> yep
Nov 11 06:45:53 <zmatt> if that's where the dtbo is located
Nov 11 06:47:53 <zmatt> (random nitpick: although residing in /sys you're not actually in sysfs but in configfs at the moment... /sys/kernel/config is a mountpoint.  same goes for debugfs which is mounted at /sys/kernel/debug.  the different is not really important, they behave similar to sysfs)
Nov 11 06:48:03 <zmatt> *difference
Nov 11 06:49:44 <brycecater> oh, i see
Nov 11 06:50:29 <brycecater> when i try that command, it tells me "no such file or directory" even though that file does exist
Nov 11 06:50:46 <zmatt> check dmesg
Nov 11 06:51:28 <zmatt> the unfortunate side of using the filesystem as user interface is that it's basically impossible to give useful error messages that way, so those appear in the kernel log instead
Nov 11 06:52:06 <brycecater> oh, i understand..
Nov 11 06:52:13 <brycecater> thank you
Nov 11 06:52:29 <brycecater> dmesg says "Direct firmware load for /lib/firmware/BB-PWM1A-00A0.dtbo failed with error -2"
Nov 11 06:53:16 <zmatt> ... which is -ENOENT which is still "no such file or directory"
Nov 11 06:53:49 <zmatt> ohh, try using "echo -n" instead of "echo"
Nov 11 06:54:43 <brycecater> still the same error
Nov 11 06:56:44 <zmatt> interesting, I get the same error if I try to load a bullshit file
Nov 11 06:56:47 <zmatt> nice job guys
Nov 11 06:57:37 <brycecater> do you think my dtbo is bad?
Nov 11 06:57:45 <zmatt> try instead:  cat /lib/firmware/BB-PWM1A-00A0.dtbo >dtbo
Nov 11 06:57:56 <zmatt> that seems to give more useful errors here
Nov 11 06:58:28 <brycecater> oh! it worked... status is applied
Nov 11 06:58:49 <zmatt> path must be broken or works differently than I thought it does
Nov 11 07:08:43 <zmatt> ahh
Nov 11 07:08:48 <brycecater> zmatt: wow, thank you so much!! you have no idea how much time and suffering you have saved me
Nov 11 07:08:51 <zmatt> the path is passed along to request_firmware()
Nov 11 07:09:07 <zmatt> so most likely you were supposed to write just 'BB-PWM1A-00A0.dtbo' to path
Nov 11 07:09:48 <brycecater> ahhh
Nov 11 07:10:10 <zmatt> though using the dtbo property makes more sense to me
Nov 11 07:10:34 <brycecater> yeah, that is what i am doing
Nov 11 07:11:33 <zmatt> sorry, I don't really use overlays myself, otherwise I would have known about the path thing ;)
Nov 11 07:12:15 <brycecater> out of curiosity, what do you use instead of overlays?
Nov 11 07:12:21 <zmatt> customize the main dts
Nov 11 07:12:46 <zmatt> overlays can be.. fickle.  you never quite know how the kernel is going to react to runtime changes to the device tree
Nov 11 07:14:45 <brycecater> yeah, makes sense... I am pretty new to beagle bone and linux kernel in general for that matter, so working directly with the main dts is a bit daunting for me... trying to take baby steps
Nov 11 07:14:46 <zmatt> (removing overlays seems even more hazardous, when I briefly experimented with that I quickly got a kernel panic.  note btw you can only remove overlays in the reverse order they were applied)
Nov 11 07:15:22 <zmatt> yeah, it would be nice if the entry barrier to that would be made lower
Nov 11 07:15:57 <zmatt> or better yet, generate dts/dtb files from some kind of _sane_ format to describe your hardware
Nov 11 07:16:07 <zmatt> *hardware config
Nov 11 07:16:22 <brycecater> hahaha.. yeah
Nov 11 07:16:39 <brycecater> lucky for me i don't need to do any unloading of overlays :)
Nov 11 07:16:40 <ds2> I wonder how well does the clock management code react to overlay removal
Nov 11 07:17:23 <zmatt> ds2: well, as long as the device goes away there shouldn't be any problem
Nov 11 07:17:47 <zmatt> though the whole hwmod thing is still magic to me anyway (and apparently going to go away in the medium-term)
Nov 11 07:35:46 <brycecater> zmatt: should i only be able to load one overlay at a time?
Nov 11 07:36:12 <brycecater> any that i load after the first have their status change to applied but do not seem to have any effect
Nov 11 07:36:36 <zmatt> did you create a new dir for them?
Nov 11 07:36:44 <brycecater> yeah
Nov 11 07:36:50 <zmatt> what does the kernel log say?
Nov 11 07:37:02 <zmatt> oh
Nov 11 07:37:15 <zmatt> if they changed to applied they should have taken effect
Nov 11 07:38:26 <zmatt> you can inspect the current state of your device tree in /proc/device-tree btw
Nov 11 07:39:41 <brycecater> dmesg does not change after applying the overlays
Nov 11 07:39:51 <zmatt> yeah so that means success
Nov 11 07:40:26 <zmatt> for some value of "success"
Nov 11 07:44:07 <brycecater> how do i interpret the contents of /proc/device-tree?
Nov 11 07:44:25 <Super_guest> device tree
Nov 11 07:44:27 <Super_guest> ;D?
Nov 11 07:44:34 <zmatt> the directory structure is exactly the device tree
Nov 11 07:45:24 <zmatt> and the files are exactly the properties, though since they're binary they're less useful to read directly
Nov 11 07:45:58 <Super_guest> i am currently also working to achieve some device tree related functions
Nov 11 07:46:21 <Super_guest> i found the uEnv.txt finaly but not where people said it would be -.-
Nov 11 07:46:40 <Super_guest> some say armhf dtso wont work with uEnv.txt is this correct?
Nov 11 07:46:42 <zmatt> tree -Ud /proc/device-tree     is a nice way to just list the nodes
Nov 11 07:47:07 <Super_guest> cd /boot/uboot/dtbs/
Nov 11 07:47:12 <Super_guest> my uboot folder does not contain dtbs
Nov 11 07:47:37 <zmatt> I don't have anything in /boot/uboot whatsoever
Nov 11 07:48:02 <Super_guest> oh i found it
Nov 11 07:48:04 <Super_guest> its just in the /boot
Nov 11 07:48:05 <zmatt> dtbs (note: not the same as dtbos) are normally in /boot/dtbs/
Nov 11 07:48:37 <Super_guest> yes the dts stuff is confusing
Nov 11 07:48:44 <Super_guest> but i found multiple sources to help me out
Nov 11 07:48:46 <Super_guest> i can do this
Nov 11 07:48:54 <Super_guest> comple with dtc
Nov 11 07:49:02 <Super_guest> echo it somewhere
Nov 11 07:49:10 <Super_guest> and boot/dtbs for startup
Nov 11 07:49:12 <zmatt> brycecater: dtc -I fs -O dts /proc/device-tree    converts it back to dts format
Nov 11 07:49:39 <Super_guest> zmatt
Nov 11 07:49:43 <zmatt> brycecater: though not very readable since all phandle refs will have become numerical
Nov 11 07:49:54 <Super_guest> u recommned doing the boot configuration with boot/dtbs or with the uEnv.txt?
Nov 11 07:50:17 <zmatt> Super_guest: the question makes no sense to me... they configure very different things
Nov 11 07:50:42 <Super_guest> O:
Nov 11 07:50:53 <zmatt> uEnv.txt contains variables for the bootloader, such as which kernel, initrd, and dtb to load
Nov 11 07:50:57 <Super_guest> i cant find the am335x-boneblack.dtb
Nov 11 07:51:18 <zmatt> what filesystem image are you using?  (cat /etc/dogtag )
Nov 11 07:51:18 <Super_guest> which i had the locate function
Nov 11 07:51:22 <Super_guest> latest
Nov 11 07:51:38 <Super_guest> 2015-03-01
Nov 11 07:51:42 <zmatt> ancient, ok
Nov 11 07:51:53 <Super_guest> its still 2015
Nov 11 07:51:59 <Super_guest> how fast do u make new distros
Nov 11 07:53:06 <zmatt> last three official releases were 2015-03-01, 2015-07-28, and 2015-11-03
Nov 11 07:53:19 <Super_guest> o.o
Nov 11 07:53:20 <zmatt> note that for some reason the latest release still uses Wheezy, which is old crap
Nov 11 07:53:33 <Super_guest> the getting started says the latest is 1 not 11
Nov 11 07:53:35 <brycecater> zmatt: so the overlays say that they are applied (in the status file) but the pin muxing display by your script only shows the change that the first one makes... the pins from the subsequent overlays do not seem to change
Nov 11 07:53:35 <Super_guest> or did i misread
Nov 11 07:53:42 <zmatt> most likely it's just outdated
Nov 11 07:53:47 <Super_guest> yes i found it
Nov 11 07:53:53 <Super_guest> sir molloy had given it to me
Nov 11 07:53:54 <Super_guest> ;D
Nov 11 07:53:56 <zmatt> a ever-looming risk of info on the interwebs
Nov 11 07:54:03 <Super_guest> i knew i had seen that name somewhere
Nov 11 07:54:07 <Super_guest> he has it in his github
Nov 11 07:54:21 <Super_guest> the arm335x dts
Nov 11 07:54:39 <zmatt> Super_guest: less outdated "testing" images are also available (debian jessie)
Nov 11 07:54:40 <zmatt> ehm
Nov 11 07:54:49 <zmatt> do NOT grab an dts from a random place
Nov 11 07:54:55 <zmatt> or dtb
Nov 11 07:55:18 <zmatt> dtb files are always included with your kernel
Nov 11 07:55:38 <Super_guest> sir molloy is the greatest ;D
Nov 11 07:55:40 <zmatt> corresponding dts files can be found in the corresponding kernel sources
Nov 11 07:55:52 <Super_guest> u dont know him ;x?
Nov 11 07:55:59 <zmatt> grabbing a random version may very well lead to an unbootable system
Nov 11 07:56:02 <Super_guest> http://derekmolloy.ie/gpios-on-the-beaglebone-black-using-device-tree-overlays/
Nov 11 07:56:15 <Super_guest> zmatt, so u suggest flashing 2015-11-03?
Nov 11 07:56:18 <Super_guest> on the bbb?
Nov 11 07:57:09 <zmatt> personally, I strongly recommend a 'testing' image
Nov 11 07:57:24 <zmatt> given that they're _still_ using wheezy for the official releases, god knows why
Nov 11 07:57:48 <Super_guest> uuuugggh
Nov 11 07:58:12 <zmatt> note that the 'testing' images actually use debian stable (jessie), not debian testing (stretch)
Nov 11 07:58:47 <Super_guest> i don thave my barrel jack power supply with me
Nov 11 07:58:53 <Super_guest> and not recommended flashing over usb
Nov 11 07:58:55 <zmatt> we don't use anything older than stretch ourselves tbh, with debian that's old enough thanks
Nov 11 07:59:02 <Super_guest> o.o
Nov 11 07:59:11 <Super_guest> ill just make my system unbootable then ;(
Nov 11 07:59:16 <zmatt> but there are no ready-made images for stretch
Nov 11 07:59:29 <Super_guest> http://elinux.org/Beagleboard:BeagleBoneBlack_Debian
Nov 11 07:59:34 <zmatt> so unless you have debian experience, stick to jessie I guess
Nov 11 08:00:10 <Super_guest> i have wheezy
Nov 11 08:00:12 <Super_guest> or whatever
Nov 11 08:00:16 <Super_guest> these names tell me nothing
Nov 11 08:00:29 <Super_guest> ill just ftp that arm335
Nov 11 08:00:31 <Super_guest> and hope for the best
Nov 11 08:00:47 <zmatt> what on earth are you doing?
Nov 11 08:00:57 <zmatt> also, it's am335x
Nov 11 08:01:24 <Super_guest> i want to enable uart2 4 5
Nov 11 08:01:26 <Super_guest> http://www.armhf.com/beaglebone-black-serial-uart-device-tree-overlays-for-ubuntu-and-debian-wheezy-tty01-tty02-tty04-tty05-dtbo-files/
Nov 11 08:01:36 <Super_guest> i want them avaiable at start up
Nov 11 08:02:02 <Super_guest> i don thave the arm335x but i can download it from molloy
Nov 11 08:02:08 <veremit> uboot config :P
Nov 11 08:02:17 <zmatt> AM335x
Nov 11 08:02:18 <zmatt> ^^
Nov 11 08:02:22 <zmatt> ^^
Nov 11 08:02:29 <Super_guest> oh rite
Nov 11 08:02:37 <veremit> tbh you can't do anything until kernel loads anyway ..so it doesnm't matter
Nov 11 08:03:01 <Super_guest> https://billwaa.wordpress.com/2014/10/13/beaglebone-black-enable-all-uart-ports-at-boot/\
Nov 11 08:03:15 <Super_guest> but this guy says the armhf method wont work with uEnv.txt
Nov 11 08:03:21 <Super_guest> ill try it maybe it does work
Nov 11 08:04:10 <veremit> oh .. -that- late yea .. just automatically enable tge uart configs
Nov 11 08:04:41 <veremit> think capemgr has been fixed now, right?
Nov 11 08:05:09 <zmatt> was it broken?
Nov 11 08:05:17 * zmatt has no idea, never uses it
Nov 11 08:05:33 <zmatt> also he's still using an ancient release
Nov 11 08:05:38 <Super_guest> its not ancient ;(
Nov 11 08:05:42 <zmatt> Super_guest: what kernel version are you on
Nov 11 08:05:45 <Super_guest> its recommneed by beaglebone .org
Nov 11 08:05:59 <Super_guest> 3.8.13
Nov 11 08:06:04 <zmatt> veremit: that.
Nov 11 08:06:20 <veremit> ah thats probably 'original'
Nov 11 08:06:30 <Super_guest> had to install glibc >2.14 manualy
Nov 11 08:06:32 <zmatt> the march image
Nov 11 08:06:43 <Super_guest> does the nwe one come with glibc higher than 2.15?
Nov 11 08:06:46 <veremit> its the bits in the meantime .. I think 3-10-4 it didn't work
Nov 11 08:07:17 <Super_guest> ok guyz thank u for the info <(^_^<)
Nov 11 08:07:28 <Super_guest> if it crashes while i am doing this then i can still flash it with sd right?
Nov 11 08:07:31 <veremit> Super_guest .. check whats on the packages.debian.net site
Nov 11 08:07:31 <Super_guest> to the new one?
Nov 11 08:07:47 <Super_guest> i am new in the linux world sorry ;x
Nov 11 08:07:48 <zmatt> Super_guest: I have 2.19
Nov 11 08:08:07 <veremit> I don't appear to have glibc installed
Nov 11 08:08:08 <zmatt> Super_guest: you can always reflash using SD, yes
Nov 11 08:08:15 <zmatt> veremit: package name libc6
Nov 11 08:08:16 <Super_guest> thank u then
Nov 11 08:08:20 <veremit> ah
Nov 11 08:08:24 <veremit> apt-what-is it?
Nov 11 08:08:29 <veremit> show?
Nov 11 08:08:33 <zmatt> show, or policy
Nov 11 08:08:42 <Super_guest> try ldd --version
Nov 11 08:08:46 <zmatt> or that
Nov 11 08:08:58 <veremit> 2.13 here on an ancient build
Nov 11 08:09:06 <zmatt> apt-cache policy  is my usual way of checking package versions
Nov 11 08:09:07 <Super_guest> same build as mine probally
Nov 11 08:09:12 <veremit> but install what you want via the debian repos
Nov 11 08:09:37 <veremit> https://packages.debian.org/search?keywords=libc6
Nov 11 08:09:47 <Super_guest> thank u guyz tho i am first going to try what the tutorial says (>^_^)>
Nov 11 08:09:57 <veremit> yea upgrade to a decent jessie image
Nov 11 08:10:25 <veremit> if you're about through the day, will check with the kernel dudes here what the state of capemgr is
Nov 11 08:10:51 <veremit> otherwise you gotta do your own device-trees stuff
Nov 11 08:10:57 <veremit> or use zmatt's uio :P
Nov 11 08:10:59 <zmatt> veremit: you have any idea what's in the lxqt-4g image that's not in lxqt-2g ?
Nov 11 08:11:14 <veremit> zmatt .. alas no .. would be useful if RN did a package listing
Nov 11 08:11:26 <veremit> we'll have to nudge him about that - should be trivial
Nov 11 08:11:36 <veremit> libreoffice?! lol
Nov 11 08:11:37 <zmatt> veremit: uio is orthogonal to whether you use overlays or a custom main dts
Nov 11 08:11:45 <veremit> zmatt: cool :d
Nov 11 08:11:48 <Super_guest> what about setting up the uarts with crontab ;x
Nov 11 08:11:58 <Super_guest> then i can even stty them to raw and right baudrate
Nov 11 08:12:01 <veremit> ewwww
Nov 11 08:12:03 <Super_guest> ;(
Nov 11 08:12:09 <zmatt> generating custom dtb files should be made more accessible
Nov 11 08:12:13 <Super_guest> i am beginner
Nov 11 08:12:14 <veremit> just put the cape_mgr.enable blah in uenv .. its easy really
Nov 11 08:12:27 <veremit> that's what I think I did years ago
Nov 11 08:12:43 <veremit> but yea .. gotta be careful which versions if you need specific glib
Nov 11 08:12:53 <zmatt> once you're used to it you wonder why on earth you ever used those messy overlays, but right now the entry barrier is still to high
Nov 11 08:13:06 <zmatt> *too
Nov 11 08:13:16 <Super_guest> i have nothing to put on enable
Nov 11 08:13:21 <Super_guest> i dont have the am335x files
Nov 11 08:13:24 <Super_guest> o.O
Nov 11 08:13:32 <zmatt> the dtbos you mean
Nov 11 08:13:38 <zmatt> did you check /lib/firmware ?
Nov 11 08:13:47 <Super_guest> yes i did
Nov 11 08:13:53 <Super_guest> only python and adafruit
Nov 11 08:14:03 <zmatt> in /lib/firmware ??
Nov 11 08:14:06 <Super_guest> yep
Nov 11 08:14:08 <Super_guest> no uarts
Nov 11 08:14:15 <Super_guest> download or write tose myself
Nov 11 08:14:20 <Super_guest> according to the tutorials
Nov 11 08:14:22 <Super_guest> no BB-UART
Nov 11 08:14:43 <veremit> shuold be in the cape_mgr stuff
Nov 11 08:14:54 <zmatt> I have 'em in /lib/firmware
Nov 11 08:15:01 <zmatt> BB-UART1-00A0.dtbo
Nov 11 08:15:03 <Super_guest> yes but u use latest
Nov 11 08:15:05 <Super_guest> i use ancient
Nov 11 08:15:08 <veremit> yah.. associated with I'd imagine
Nov 11 08:15:17 <zmatt> they should be there on ancient too... those files *are* ancient
Nov 11 08:15:18 <zmatt> afaik
Nov 11 08:15:25 <veremit> yup should be there
Nov 11 08:15:32 <Super_guest> i can compile the armhf onces
Nov 11 08:15:34 <Super_guest> with dtc
Nov 11 08:15:45 <Super_guest> but i dont have them in my firmware folder
Nov 11 08:15:59 <zmatt> (also this system I'm looking at isn't latest but 2015-09-20 testing)
Nov 11 08:16:03 <veremit> hmm if you dmesg|grep cape_mgr does it show up??
Nov 11 08:16:18 <veremit> zmatt .. I think they put it 'back in' what kernel are you on?
Nov 11 08:16:44 <Super_guest> no it doesnt ;x
Nov 11 08:16:44 <zmatt> veremit: this image shipped with some 4.1 kernel
Nov 11 08:16:51 <veremit> and are you 'testing-jessie' or testing debian-sid?
Nov 11 08:16:56 <zmatt> veremit: (it's an lxqt image)
Nov 11 08:16:56 <Super_guest> but i do have  cape_mgr hidden somewhere
Nov 11 08:17:01 <Super_guest> i did get uart 4 running yesterday
Nov 11 08:17:06 <Super_guest> with cape_mgr echo or something
Nov 11 08:17:23 <zmatt> veremit: kernel or debian version doesn't matter since these files live outside package management
Nov 11 08:17:37 <zmatt> so all that matters is the dogtag
Nov 11 08:17:59 <veremit> really .. lib/firmware should be governed by the kernel
Nov 11 08:18:01 <zmatt> this was an lxqt testing image
Nov 11 08:18:41 <zmatt> (which is our only BBB where I could look since the rest is based on console images which don't have them at all)
Nov 11 08:18:52 <veremit> zmatt is that "debian testing" or "beagle testing"
Nov 11 08:18:57 <zmatt> rcn-ee testing
Nov 11 08:19:04 <veremit> ah .. probably jessie?
Nov 11 08:19:07 <veremit> debian STable
Nov 11 08:19:10 <zmatt> yes
Nov 11 08:19:15 <veremit> right
Nov 11 08:19:25 <zmatt> though that's not really an important factor in this
Nov 11 08:19:26 <veremit> I'd go with that then .. flash the console version and go from there ...
Nov 11 08:19:42 <zmatt> I wouldn't recommend a console image for Super_guest
Nov 11 08:19:45 <veremit> if you got all the firmware files ..
Nov 11 08:19:47 <veremit> oh ok
Nov 11 08:19:52 <veremit> well .. lxqt then :D
Nov 11 08:20:09 <Super_guest> maybe i can get it working
Nov 11 08:20:15 <Super_guest> lots of the tutorials are dated also u know
Nov 11 08:20:20 <Super_guest> 2013 mostly
Nov 11 08:20:22 <veremit> I'll try to boot my 'play' beagle later on
Nov 11 08:20:29 <zmatt> veremit: as I mentioned above, console images do not have the dtbos _at all_
Nov 11 08:20:32 <veremit> Super_guest.. most documentation is pitifully out of date
Nov 11 08:20:40 <veremit> zmatt .. really?! wtf.
Nov 11 08:20:43 <zmatt> rcn-ee should rename 'console' to 'minimal'
Nov 11 08:20:44 <veremit> need to push RN on that
Nov 11 08:20:57 <veremit> unless there's a package for it ..
Nov 11 08:21:02 <veremit> that's another useful option
Nov 11 08:21:41 <zmatt> dpkg-query: no path found matching pattern /lib/firmware/BB-UART1-00A0.dtbo
Nov 11 08:22:02 <zmatt> as I said, outside package management
Nov 11 08:22:10 <veremit> yaeh
Nov 11 08:22:17 <veremit> think we'll enquire about aht
Nov 11 08:23:12 <zmatt> all the crap living outside package management is exactly why I personally _avoid_ those images... much easier to build up from minimal than to hunt down and delete crap
Nov 11 08:23:26 <brycecater> i am getting an issue with applying an overlay. the dmesg reads "[  485.613940] Invalid device tree blob header [  485.619595] create_overlay: failed to unflatten tree"
Nov 11 08:23:54 <zmatt> brycecater: did you accidently cat the dts instead of the dtbo maybe? ;)
Nov 11 08:24:13 <Super_guest> <(^_^<) this community ;D
Nov 11 08:24:36 <zmatt> brycecater: I mean, invalid header really sounds like you grabbed the wrong stuff
Nov 11 08:24:50 <brycecater> zmatt: no, but i am worried that i may have made an error in my dts
Nov 11 08:24:59 <zmatt> not one that results in a wrong header
Nov 11 08:25:09 <zmatt> afaik
Nov 11 08:25:47 <brycecater> zmatt: https://gist.github.com/brycedcarter/94d020a2b2aafc3b4de2
Nov 11 08:26:23 <brycecater> that is the dts that i made... but i am just learning so i do not really know what I am doing
Nov 11 08:26:38 <zmatt> brycecater: you can remove all the properties of /, they're ignored by the configfs mechanism
Nov 11 08:26:50 <brycecater> it compiles to a .dtbo without errors though
Nov 11 08:27:09 <zmatt> only the nodes (traditionally named fragment@0 etc)
Nov 11 08:27:20 <zmatt> are used
Nov 11 08:27:58 <zmatt> ew
Nov 11 08:29:00 <brycecater> sorry, i am not really sure what you mean by properties of /..
Nov 11 08:29:26 <brycecater> still working on learning the syntax
Nov 11 08:30:00 <zmatt> brycecater: the overall structure of the overlay is   / { property = value; property = value; /..; node { ... }; node { ... }; ... };
Nov 11 08:30:16 <zmatt> sorry for that / in the middle, typo
Nov 11 08:30:41 <Super_guest> https://github.com/jadonk/validation-scripts/tree/master/test-capemgr
Nov 11 08:30:45 <Super_guest> there is an example ;x
Nov 11 08:31:01 <zmatt> in this case, the child nodes of / are  fragment@0 { ... }; fragment@1 { ... };  etc
Nov 11 08:31:13 <zmatt> also, I just realized there's really no excuse for this horrible syntax
Nov 11 08:31:28 <Super_guest> thats why u just build from the example
Nov 11 08:31:34 <Super_guest> isntead of writing from scratch
Nov 11 08:31:41 <zmatt> dtc could easily have supported writing normal fragments
Nov 11 08:32:41 <zmatt> i.e.    &ocp { CONTENT };    instead of this    / { fragment@5 { target = <&ocp>; __overlay__ { CONTENT }; }; };    crap
Nov 11 08:32:57 <zmatt> what moron came up with this
Nov 11 08:33:01 <zmatt> jeez
Nov 11 08:33:17 <zmatt> that's yet another reason I hate overlays
Nov 11 08:33:47 <zmatt> oh, I see a potential reason why you might have trouble with multiple overlays
Nov 11 08:34:10 <zmatt> you make a new node &ocp/helper
Nov 11 08:34:21 <zmatt> ... do they all use that same name?  "helper" ?
Nov 11 08:35:01 <Super_guest> u guyz have influence on what will be fixed on the bbb ;x?
Nov 11 08:35:02 <zmatt> since that means the second and later overlays just try to change the properties of the "helper" made by the first overlay (which is probably ignored by the kerne)
Nov 11 08:35:46 <brycecater> zmatt: they did all use helper, that gist that i uploaded was my attempt at combining them into one overlay
Nov 11 08:36:27 <brycecater> but even with removing the properties of / i still get the same error on trying to load it
Nov 11 08:36:27 <zmatt> brycecater: if you use bone-pinmux-helper at all, at least you need to pick unique node names for each overlay
Nov 11 08:36:39 <zmatt> there's no need to label it though (the "test_helper:")
Nov 11 08:36:50 <zmatt> since you never reference it
Nov 11 08:37:36 <zmatt> brycecater: the dtbos you loaded successfully, you also compiled those yourself?
Nov 11 08:37:48 <brycecater> yeah
Nov 11 08:37:54 <zmatt> same dtc?
Nov 11 08:38:15 <zmatt> since afaik 3.8 and 4.1 use slightly different dtbo formats and corresponding dtc versions
Nov 11 08:38:39 <brycecater> yeah
Nov 11 08:38:44 <brycecater> same dtc
Nov 11 08:38:49 <zmatt> weird
Nov 11 08:39:08 <brycecater> here is one the the seperate dts files that worked
Nov 11 08:39:09 <brycecater> https://gist.github.com/brycedcarter/cc996382b15a60c320c7
Nov 11 08:40:02 <zmatt> btw normally you'd associate the pinmux of a peripheral with that peripheral rather than using bone-pinmux-helper
Nov 11 08:43:49 <brycecater> zmatt: how would i do that
Nov 11 08:45:43 <brycecater> how would i go about setting unique node names?
Nov 11 08:47:29 <Super_guest> hmm i got mesages over tty 04 02 but 05 not working
Nov 11 08:47:34 <Super_guest> i mean O
Nov 11 08:48:00 <Super_guest> good enough for now ;x?
Nov 11 08:48:46 <zmatt> http://gerbil.xs4all.nl/rgb-led.dtsi.html   ... so, that's using normal syntax instead of overlay-horror (since I use a custom dts) and uses macros instead of hex numbers for the pinmux... also I use different label names for the pwm peripherals, so some adaptation is needed ;P
Nov 11 08:49:44 <Super_guest> hmm those adafruit overlays might also work from the firmware ;x
Nov 11 08:49:46 <Super_guest> for uart ;x
Nov 11 08:52:46 <zmatt> brycecater: so I actually use a _custom_ dts, not stock-with-some-changes, but I accept that I'm a nutcase
Nov 11 08:53:23 <Super_guest> could be that tty O5 is not working because its hdmi standard
Nov 11 08:53:26 <zmatt> (I just happen to like short, consistent labels and sensible structure)
Nov 11 08:53:28 <Super_guest> but overlay means overlay right ;x
Nov 11 08:54:04 <Super_guest> it changes the mode so o.O
Nov 11 08:54:58 <zmatt> kernel will refuse to load conflicting overlays
Nov 11 08:56:07 <zmatt> if you want to check which pinmux options are available, you can use my spreadsheet (in particular the P9 and P8 tabs, or the BBB tab) -> https://goo.gl/Jkcg0w
Nov 11 08:57:10 <zmatt> uart5 indeed conflicts with the lcdc pins (hence with hdmi)
Nov 11 09:02:08 <zmatt> veremit: lol, I had some inexplicable crashes in my baremetal app that appeared and vanished with seemingly unrelated code changes...
Nov 11 09:02:48 <zmatt> veremit: just noticed I still had a 512-byte stack configured, and absolutely nothing to prevent a stack overflow from scribbling over the code
Nov 11 09:03:05 <zmatt> :$
Nov 11 09:09:27 <zmatt> I guess it may be a good idea to add support for page tables and use separate non-adjacent pages with appropriate protection for text/rodata/data/stack
Nov 11 09:12:44 <Super_guest> hmm the bootup with armhf isnt working with uEnv
Nov 11 09:12:59 <Super_guest> so 2015-11-3 will fix this
Nov 11 09:22:39 <stt_michael> zmatt,  nice stack use :P
Nov 11 09:22:54 <zmatt> 512 bytes should be enough for anyone right
Nov 11 09:23:47 <stt_michael> like 1024k is plenty of ram ;)
Nov 11 09:24:21 <stt_michael> all these greedy buggers with 16gb .. wtf you fillin it with? prn?!
Nov 11 09:24:22 <stt_michael> lol
Nov 11 09:27:34 <zmatt> I use about 40k ram atm (code + data), but I see it's including the alignment padding required for the section table in that count
Nov 11 09:27:49 <zmatt> (16 KB table required to be aligned to its size)
Nov 11 09:27:51 <brycecater> any insight on what thees errors might mean? https://gist.github.com/brycedcarter/09cbbcf8d51c7db11938
Nov 11 09:29:04 <Super_guest> i am looking at this list and i want jessi so i have glic with 2.19
Nov 11 09:29:06 <Super_guest> http://elinux.org/Beagleboard:BeagleBoneBlack_Debian#2015-11-03
Nov 11 09:29:13 <Super_guest> but confused of what the flasher: means
Nov 11 09:35:48 <tbr> Super_guest: flasher means it will install it on the eMMC
Nov 11 09:36:17 <tbr> the name usually indicates also if it's suitable only for the 4GB model or also for the revB 2GB models
Nov 11 09:36:58 <zmatt> the 2gb versions also have the benefit of including less crap, hence less time to flash and more free eMMC space available afterward
Nov 11 09:37:26 * tbr prefers the console images, zero crap
Nov 11 09:37:43 <zmatt> I usually still find some pages to purge of them actually
Nov 11 09:37:59 <zmatt> but the console images are... not for everyone :P
Nov 11 09:38:04 <zmatt> *some packages
Nov 11 09:38:12 <tbr> mimiimiimi no hdmi!
Nov 11 09:40:13 <zmatt> they lack a lot more than that of course... it's unfortunate that right now if you want to be able to follow any tutorial without finding files missing, you're forced to download an lxde/lxqt image even if you don't use HDMI at all
Nov 11 09:41:12 <tbr> would be good if there were metapackages or such
Nov 11 09:43:50 <zmatt> right now a lot of what's on the lxde/lxqt images lives completely outside package management... I wouldn't even know how to install it on a console image
Nov 11 09:44:10 <zmatt> (granted, if I cared even a little bit I would probably find out quickly, but still)
Nov 11 09:51:53 <brycecater> zmatt: i am getting this error "pin 44e109b4.0 already requested by 44e10800.pinmux; cannot claim for ocp:helper" but i dont think that i am requesting that pin... any idea what is going on
Nov 11 09:54:42 <zmatt> hmm, it's not a pin used for anything by default
Nov 11 09:54:42 <Super_guest> but i can install it from the sd card to the eMMC
Nov 11 09:55:19 <zmatt> Super_guest: flashers automatically do that; you just write the flasher image to SD, boot it, and it'll flash eMMC
Nov 11 09:56:34 <zmatt> lemme see if there's an easy way to see who's hogging what
Nov 11 10:22:46 <zmatt> brycecater: could you pastebin the content of /sys/kernel/debug/pinctrl/44e10800.pinmux/pinmux-pins ?
Nov 11 10:23:58 <ag_melbourne> Hello :) has anyone got a uio_pypruss.ko driver for Jessie?
Nov 11 10:37:17 <brycecater> zmatt: ended up getting it working... turns out i was just being dumb... i did have a pin conflict
Nov 11 10:40:55 <zmatt> I'm still curious to see the content, since you're using different pinmux mechanisms than I am and I want to make my script robust against that :)
Nov 11 10:52:22 <brycecater> here it is:
Nov 11 10:52:23 <brycecater> https://gist.github.com/brycedcarter/2af7a2c41c3dd6f1a519
Nov 11 10:52:31 <brycecater> one again, thanks for all the help
Nov 11 11:01:04 <stt_michael> zmatt, Super_guest .. there are standalone version of the lxqt/console images available for evaluation IIRC
Nov 11 11:01:14 <stt_michael> once rcn-ee gets online .. we'll poke him :D
Nov 11 11:01:22 <Super_guest> o.O?
Nov 11 11:01:46 <stt_michael> so you can try before flashing
Nov 11 11:01:56 <stt_michael> right .. coffeee ..
Nov 11 11:02:01 <Super_guest> qumu?
Nov 11 11:08:08 <Super_guest> so we tackled the device tree overlay problem today
Nov 11 11:08:24 <Super_guest> what other problems might we come across?
Nov 11 11:16:46 <stt_michael> yea.. don't ask those kinda questions .. lol ..
Nov 11 11:55:11 <Super_guest> my beaglebone black is still flashing?
Nov 11 11:55:24 <Super_guest> whats the default led pattern, only hartbeat?
Nov 11 11:56:05 <Super_guest> he has gotten some cpu activity also but is this default ;x or is this becaus he is flashing
Nov 11 11:59:59 <Super_guest> it did flash 4 times or something 4 leds at start
Nov 11 12:04:03 <Super_guest> using the usb port instead of barrel jack\
Nov 11 12:28:38 * zmatt created a spiffied-up http://gerbil.xs4all.nl/show-pins-v2.pl.gz
Nov 11 12:30:00 <zmatt> should my color choices not be to one's taste, you can customize them
Nov 11 12:30:05 <zmatt> (by editing the source)
Nov 11 12:40:13 <av500> https://www.hackster.io/beagleboard/projects
Nov 11 12:44:44 <zmatt> random link?
Nov 11 12:46:12 <Super_guest> still flashing
Nov 11 12:46:20 <Super_guest> should have used 2A barrel jack no ;(?
Nov 11 12:47:21 <zmatt> heartbeat means it's not flashing, i.e. either you used a non-flasher or failed to hold the SD button properly
Nov 11 12:48:54 <zmatt> the flasher makes a distinctive KITT led pattern while it's flashing
Nov 11 12:48:59 <zmatt> and all leds on when it's done
Nov 11 12:49:25 <zmatt> if the leds look at all like they do when you boot a normal system, you've booted a normal system
Nov 11 12:57:59 <tbr> I think the march flasher had a different pattern?
Nov 11 12:58:15 <zmatt> he's flashing the latest now I thought?
Nov 11 12:58:22 <av500> it flashes a 4/4 beat?
Nov 11 12:59:18 <zmatt> oh well who cares, as long as it makes a distinctive led pattern that cannot be mistaken for a normal running BBB
Nov 11 13:22:49 <Super_guest> its still flashing???
Nov 11 13:23:09 <zmatt> read what I said a few lines ago
Nov 11 13:23:45 <zmatt> which file did you download?
Nov 11 13:26:18 <Super_guest> egh
Nov 11 13:27:03 <Super_guest> microSD/Standalone: (lxqt-4gb) (BeagleBone/BeagleBone Black/BeagleBone Green)
Nov 11 13:27:12 <Super_guest> http://elinux.org/Beagleboard:BeagleBoneBlack_Debian#Jessie_Snapshot_lxqt
Nov 11 13:27:45 <zmatt> Standalone
Nov 11 13:27:47 <Super_guest> i wrote it to sd card after i unpack the image
Nov 11 13:27:48 <zmatt> i.e. not a flasher
Nov 11 13:28:10 <Super_guest> but then i configure the uEnv.txt file to start to flash from sd card
Nov 11 13:28:18 <Super_guest> it wont work?
Nov 11 13:28:20 <Super_guest> ??
Nov 11 13:28:37 <zmatt> why would you attempt such a thing instead of downloading an actual flasher?
Nov 11 13:28:49 <Super_guest> cmdline=init=/opt/scripts/tools/eMMC/init-eMMC-flasher-v3.sh\
Nov 11 13:28:52 <Super_guest> i dont know what a flasher is
Nov 11 13:28:56 <Super_guest> i follow the tutorial
Nov 11 13:29:13 <Super_guest> read it form sd card
Nov 11 13:29:16 <Super_guest> so i flash the sd card
Nov 11 13:29:16 <zmatt> first two links at the url you linked to
Nov 11 13:29:21 <zmatt> Flasher: (lxqt-4gb) (BeagleBone Black/Green 4GB eMMC)
Nov 11 13:29:23 <zmatt> Flasher: (lxqt-2gb) (BeagleBone Black 2GB eMMC)
Nov 11 13:29:29 <Super_guest> what can i do with that?
Nov 11 13:29:33 <Super_guest> flash to sd card?
Nov 11 13:29:52 <Super_guest> with win32diskimager?
Nov 11 13:30:13 <Super_guest> ;(
Nov 11 13:30:15 <zmatt> then boot the BBB from the sd card and it'll automatically reflash your eMMC
Nov 11 13:30:19 <zmatt> I explained this already
Nov 11 13:30:23 <Super_guest> i am dumb
Nov 11 13:30:26 <Super_guest> i didnt press the button
Nov 11 13:30:39 <Super_guest> can i flash it with win32diskimager over usb
Nov 11 13:30:50 <Super_guest> instead of flashing the sd card, i select the drive of bbb?
Nov 11 13:31:08 <zmatt> not recommended
Nov 11 13:31:18 <Super_guest> ok so i flash the sd card?
Nov 11 13:31:25 <zmatt> yes, with a flasher image
Nov 11 13:31:27 <Super_guest> i have done with with the standalone image
Nov 11 13:31:38 <Super_guest> u say do the flasher image, and then try procedure again?
Nov 11 13:31:54 <Super_guest> ok
Nov 11 13:31:56 <Super_guest> thank u
Nov 11 13:32:05 <Super_guest> <(^_^<)
Nov 11 13:32:06 <zmatt> yes.  you can choose between the 4gb and the 2gb image... I think I'd personally go for the 2gb image since it'll have less crap on it
Nov 11 13:32:14 <Super_guest> i need crap man
Nov 11 13:32:24 <zmatt> then go ahead and use the 4gb image
Nov 11 13:32:55 <Super_guest> ok i unplug
Nov 11 13:32:56 <Super_guest> ;(
Nov 11 13:33:07 <Super_guest> wait
Nov 11 13:33:08 <Super_guest> that fucker
Nov 11 13:33:15 <Super_guest> isnt flashing leds anymore
Nov 11 13:33:56 <Super_guest> its doing led 1 for micro sd now
Nov 11 13:34:25 <Super_guest> is it possible to flash the standalone?
Nov 11 13:34:31 <Super_guest> or am i just imaging stuff?
Nov 11 13:34:44 <zmatt> I don't know, I don't care
Nov 11 13:35:03 <Super_guest> i have it plugged in for 4 hours -.
Nov 11 13:35:53 <zmatt> I think it takes at most 45 minutes or so for a large image
Nov 11 13:35:53 <Super_guest> o god
Nov 11 13:36:00 <Super_guest> it has new ID
Nov 11 13:36:06 <Super_guest> it must have worked
Nov 11 13:36:49 <zmatt> "it" ?  keep in mind you're just staring the the standalone system running from sd
Nov 11 13:37:04 <Super_guest> ??
Nov 11 13:37:09 <Super_guest> i didnt press the button to boot from sd
Nov 11 13:37:47 <zmatt> your eMMC has an old image, those were prone to use the SD card as rootfs if inserted at boot
Nov 11 13:37:48 <Super_guest> i just brouwse to the bbb drive on my pc
Nov 11 13:38:01 <Super_guest> i am so confused right now
Nov 11 13:38:03 <Super_guest> i cant putty to it
Nov 11 13:38:08 <Super_guest> ill just unplug it
Nov 11 13:38:10 <Super_guest> and see what it does
Nov 11 13:38:11 <zmatt> it's the standalone image
Nov 11 13:38:20 <Super_guest> i dont know what that means
Nov 11 13:38:29 <zmatt> just a normal system
Nov 11 13:38:34 <zmatt> but installed on SD instead of on eMMC
Nov 11 13:38:47 <zmatt> it accidently booted from it due to a bootloader/kernel issue
Nov 11 13:39:31 <zmatt> linux still assigns /dev/mmcblk0 to the first MMC/SD card it sees
Nov 11 13:39:45 <Super_guest> so if i remove the sd card
Nov 11 13:39:47 <Super_guest> it wouldnt work
Nov 11 13:39:57 <zmatt> you'd boot to your previous system again
Nov 11 13:40:48 <zmatt> booting from sd in this way is a bug and not recommended, since it uses the bootloader and kernel from eMMC but root filesystem from SD
Nov 11 13:41:35 <zmatt> it's just an icky known issue, primarily caused by some obnoxious people on the linux-mmc mailing list refusing to apply a patch
Nov 11 13:41:56 <Super_guest> ok it didnt work
Nov 11 13:42:03 <zmatt> afaik rcn-ee worked around it in later images
Nov 11 13:42:08 <zmatt> but your eMMC still has an old one
Nov 11 13:42:23 <Super_guest> its no longer showing up as a beaglebone
Nov 11 13:42:26 <Super_guest> without sd card
Nov 11 13:42:38 <zmatt> afte reboot?
Nov 11 13:42:43 <Super_guest> unplug
Nov 11 13:42:48 <Super_guest> same thing yes
Nov 11 13:43:11 <zmatt> just download the flasher
Nov 11 13:43:56 <Super_guest> so when i have the flasher on the sd card, i put in the bbb, and i don thave to press any boot buttons?
Nov 11 13:44:22 <zmatt> boot button is still absolutely required.   if you want to be absolutely sure you properly held down the SD button during power on, do it without SD card inserted -- the BBB won't boot at all (only power led goes on), you can then insert the SD card and press the reset button to boot it
Nov 11 13:44:51 <Spidler_> zmatt: Is there a good way to upgrade only the bootloader and related config on the eMMC without reflashing completely new image?
Nov 11 13:45:01 <zmatt> Spidler_: suer, dd
Nov 11 13:45:03 <zmatt> *sure
Nov 11 13:45:13 <Spidler_> I haven't found one that doesn't involve compiling it yourself.
Nov 11 13:46:19 <zmatt> Spidler_: how old is your current filesystem?  do you have a separate FAT boot partition (old style) ?
Nov 11 13:46:31 <Spidler_> Yea
Nov 11 13:46:49 <Spidler_> Separate FAT partition.  Means that dding into the mmc block is going to cause me slight issues.
Nov 11 13:46:54 <Super_guest> zmatt i think i dont need to press the boot button, since what u describe alreayd happens, since the old eMMC has that flasher line inabled in the uEnv.txt
Nov 11 13:47:39 <zmatt> Super_guest: go away.  you continuous try random things instead of paying attention to instructions, and then you're surprised when things don't work
Nov 11 13:48:18 <zmatt> Spidler_: if you have a separate FAT partition then no dd'ing is involved, you can just update the bootloader files on it
Nov 11 13:48:50 <zmatt> Spidler_: though, it's possible both systems are present at the same time, in which case the new (non-FAT) way takes precedence
Nov 11 13:49:22 <zmatt> you can check that by doing   dd if=/dev/mmcblk0 skip=$(( 0x20000 / 512 )) count=1 | hexdump -C
Nov 11 13:49:25 <Spidler_> I'm aware, MLO & uboot.img . The question is more, is there a pre-built binary package that I can download,  or am I forced to chase down a git revision that's in use?
Nov 11 13:49:52 <zmatt> just download the most recent console image and copy from it
Nov 11 13:50:24 <Spidler_> That just returns  0, so there seems to be nothing there.
Nov 11 13:50:36 <zmatt> ok, then it's just the FAT partition
Nov 11 13:50:39 <zmatt> hold on
Nov 11 13:50:41 <Spidler_> Okay, manually copying the files in, goodie.
Nov 11 13:50:47 <zmatt> yeah except you can't
Nov 11 13:50:56 <zmatt> since the latest images don't use a boot partition
Nov 11 13:51:04 <Spidler_> ah. okay
Nov 11 13:51:12 <Spidler_> I'll just stick to the old uboot
Nov 11 13:51:15 <zmatt> I was about to extract the files from the image for you
Nov 11 13:51:22 <Spidler_> It's not hassling me and I'm not having a huge problem with them now
Nov 11 13:51:23 <zmatt> you can also switch to the new system
Nov 11 13:51:36 <Spidler_> I'd rather not do that on 50+ devices that run headless
Nov 11 13:51:41 <zmatt> can you do fdisk -l /dev/mmcblk0 ?
Nov 11 13:51:52 <Spidler_> sure
Nov 11 13:51:59 <zmatt> well, just be sure to try it once with one you have within recovery distance :)
Nov 11 13:52:00 <Spidler_> Device         Boot   Start     End Sectors  Size Id Type
Nov 11 13:52:00 <Spidler_> /dev/mmcblk0p1 *       2048  198655  196608   96M  e W95 FAT16 (LBA)
Nov 11 13:52:00 <Spidler_> /dev/mmcblk0p2       198656 1222655 1024000  500M 83 Linux
Nov 11 13:52:00 <Spidler_> /dev/mmcblk0p3      1222656 2246655 1024000  500M 83 Linux
Nov 11 13:52:00 <Spidler_> /dev/mmcblk0p4      2246656 7553023 5306368  2.5G 83 Linux
Nov 11 13:52:08 <zmatt> heuh
Nov 11 13:52:22 <Spidler_> p2 and p3 are squashfs filesystems
Nov 11 13:52:30 <Spidler_> So I need kernel to live on vfat, sadly.
Nov 11 13:52:41 <Spidler_> With all the love that entails for me and the stability of vfat.
Nov 11 13:52:44 <zmatt> it's a very customized things?
Nov 11 13:52:49 <Spidler_> Yeaps.
Nov 11 13:52:58 <Spidler_> Well, not the bootloader. :P
Nov 11 13:53:33 <Spidler_> And since I'm looking at uboot right now, and adjusting the boot scripts & kernel to support failover / boot count
Nov 11 13:53:34 <zmatt> why do you *need* the kernel on vfat?  instead of grabbing it from p4 ?
Nov 11 13:53:48 <Spidler_> Because p4 isn't persistent in the way you think of it
Nov 11 13:53:52 <zmatt> ah ok
Nov 11 13:54:09 <Spidler_> I'd _really_ want uboot to support squashfs
Nov 11 13:54:17 <zmatt> you could make the boot partition ext also
Nov 11 13:54:28 <Spidler_> And if I get the funding I'm hoping for, I'll probably end up supporting that development
Nov 11 13:54:34 <Spidler_> as it would make my life a fair bit easier
Nov 11 13:54:44 <Spidler_> Yea.
Nov 11 13:55:14 <Spidler_> But yes, I'd really want to reduce the amount of scary things that need to be updated separately
Nov 11 13:55:28 <Spidler_> Having to tie kernel + filesystem + dtbs together and always keep them in sync bugs me.
Nov 11 13:55:39 <zmatt> btw, you realize your filesystems are misaligned w.r.t. the NAND erase blocks?
Nov 11 13:55:40 <Spidler_> (filesytem is tied to kernel mostly because of modules)
Nov 11 13:55:53 <Spidler_> These days,  yea.
Nov 11 13:55:58 <zmatt> 4 MB
Nov 11 13:56:03 <Spidler_> Mmm.
Nov 11 13:56:07 <Spidler_> "Too late to fix" :/
Nov 11 13:56:39 <zmatt> hehe, it eh, would indeed require a bit more creativity then you'd be comfortable with probably
Nov 11 13:58:42 <zmatt> but you can update u-boot on FAT or install it in its proper place for the new method (which lies before the first partition)
Nov 11 14:00:01 <zmatt> the latter method is relatively easy: copy the space between the partition table and the first partition from a recent console image
Nov 11 14:01:10 <zmatt> if there's trouble, boot it from some SD card and zeroize that space again
Nov 11 14:01:29 <zmatt> it's not part of any partition so risk is low
Nov 11 14:01:50 <zmatt> and wouldn't require touching your existing stuff initially
Nov 11 14:03:13 <zmatt> if it boots from the new u-boot, it no longer needs to live on the FAT partition, and you can consider whether you want to replace it with a non-FAT filesystem of your preference to hold the kernel and such
Nov 11 14:04:23 <zmatt> you can also try Falcon booting
Nov 11 14:09:41 <zmatt> this is the layout in current u-boot as far as I gathered -> http://pastebin.com/wSnVCpnR
Nov 11 14:10:15 <Spidler_> zmatt: okay, nice. Thanks
Nov 11 14:10:28 <Spidler_> zmatt: any documentation on Falcon booting?
Nov 11 14:10:51 <zmatt> the MLO and u-boot.img indicated in that map are btw exactly the same as the MLO and u-boot.img you'd put on a FAT partition
Nov 11 14:11:14 <zmatt> (for some reason they compile One U-Boot to Rule Them All... that's why it's so hideously bloated)
Nov 11 14:11:32 <zmatt> it'll even boot an am335x-evm
Nov 11 14:11:51 <Spidler_> Yeah. I've seen that. had some issues with search-paths for dtb's. Which is why I have my current excursion into how this hangs together.
Nov 11 14:12:17 <zmatt> the short summary is: like a mess
Nov 11 14:12:19 <zmatt> ;)
Nov 11 14:13:10 <Spidler_> Yep!
Nov 11 14:13:44 <zmatt> it's absurd they need 2 stages for the bootloader because they couldn't fit it in the max space for an MLO (which is slightly above 100 KB iirc)
Nov 11 14:13:56 <Spidler_> If it's only a few boards I have on my hands I could probably figure out how to revamp the partition table and adjust it nicely.
Nov 11 14:13:59 <zmatt> you could fit a multitasking OS in there :P
Nov 11 14:14:03 <Spidler_> Hahaha
Nov 11 14:14:22 <Spidler_> Well, I've got 500 megs of space for those squashfs filesystems... which are  ~120 megs or so
Nov 11 14:14:30 <Spidler_> So I can adjust things like that
Nov 11 14:15:11 <Spidler_> And 120 megs is huge, but there's X, fair pieces of GNOME and more in there.
Nov 11 14:15:28 <zmatt> ew
Nov 11 14:15:30 <zmatt> :)
Nov 11 14:15:31 <Spidler_> Enough to run a touch screen kiosk with a full screen SVG app.
Nov 11 14:15:49 <zmatt> can't you run like webkit directly on the fb ?
Nov 11 14:15:59 <Spidler_> touch <.<
Nov 11 14:16:06 <zmatt> so?
Nov 11 14:16:17 <Spidler_> And sure, but the development cost for adapting it is _way_ more costly than a few megs of space
Nov 11 14:16:21 <zmatt> point
Nov 11 14:16:30 <Spidler_> I mean, seriously, I'm _drowning_ in flash here with 4 Gigs :)
Nov 11 14:17:06 <zmatt> yeah, adds to boot time though, our customer basically wants nearly instant-on
Nov 11 14:17:11 <Spidler_> Ah
Nov 11 14:17:23 <Spidler_> We don't really care about boottime.
Nov 11 14:17:29 <zmatt> that helps
Nov 11 14:17:39 <Spidler_> Ours isn't really functional until you have network anyhow
Nov 11 14:17:55 <Spidler_> And webbrowsers aren't the fastest to load
Nov 11 14:18:27 <Spidler_> total boot time is ~1 min 20s
Nov 11 14:18:28 <zmatt> network up would not be a limiting factor if u-boot and kernel would STOP RESETTING THAT PHY
Nov 11 14:18:33 <zmatt> !@#&
Nov 11 14:18:38 <Spidler_> Heh
Nov 11 14:19:00 <Spidler_> I'm more concerned that we have at least one board where the PHY comes up and says "disconnected" all the tim
Nov 11 14:19:03 <Spidler_> e
Nov 11 14:19:05 <Spidler_> until you power cycle it.
Nov 11 14:19:07 <zmatt> seriously, I'm working on some baremetal code, for testing purposes and hopefully I can grow it into a bootloader
Nov 11 14:19:08 <Spidler_> Which isn't okay.
Nov 11 14:19:32 <Spidler_> mmm. I've been itching to find a reason to use the Real Time Units.
Nov 11 14:19:41 <Spidler_> But I just can't find one <.<
Nov 11 14:19:46 <zmatt> if I press "r" it'll reset the processor, ROM will use bootp and tftp and download the application from my laptop
Nov 11 14:19:49 <zmatt> it's instant
Nov 11 14:19:56 <Spidler_> Oh, nifty.
Nov 11 14:20:05 <zmatt> it has booted before my finger has let go of the 'r' key basically
Nov 11 14:20:25 <Spidler_> That's quite sweet. What are you building, if I can ask?
Nov 11 14:21:26 <zmatt> right now I'm just slowly growing a basic but comfy environment.  I also use it as a harness to perform tests that are awkward to run on linux
Nov 11 14:21:36 <zmatt> like "how exactly does the LCD controller behave" ?
Nov 11 14:23:52 <Spidler_> Ah, right.
Nov 11 14:24:19 <zmatt> but an explicit goal is to make a minimalistic rescue loader that allows remote (but secure) recovery via network
Nov 11 14:24:28 <Spidler_> I mostly build SCADA type stuff, the display is part of a cabinet standing in a boiler room that shows status of heat pumps/elevators/fans
Nov 11 14:24:52 <Spidler_> So, basically hooking up and putting a ton of junk that shouldn't be online, online ;P
Nov 11 14:25:17 <Spidler_> ( And attempting to do it without  telnet, modbus-tcp or other annoying idiotic protocols)
Nov 11 14:25:28 <zmatt> have you ever heard of Internet Census 2012 ?
Nov 11 14:25:30 <ol_> hi
Nov 11 14:26:37 <Spidler_> Yea
Nov 11 14:27:02 <zmatt> just checking ;)  it's a useful precautionary tale for any embedded systems developer
Nov 11 14:27:14 <ol_> is it ok to use http://beagleboard.org/latest-images with freedombox-setup? because the freedombox-unstable-0.6 have read/write errors.:-)
Nov 11 14:27:51 <Spidler_> Oh yeah.
Nov 11 14:27:54 <zmatt> note that http://beagleboard.org/latest-images is ancient crap, for some reason that page lags seriously behind
Nov 11 14:29:01 <ol_> hm
Nov 11 14:29:20 <zmatt> the latest release is 2015-11-03, though it's also ancient crap (wheezy) and I recommend a "testing" image (actually contains jessie, which is debian stable)
Nov 11 14:29:32 <zmatt> no idea about the freedombox thing
Nov 11 14:30:20 <ol_> zmatt: is there a good alternative to the freedombox project?
Nov 11 14:30:22 <zmatt> Spidler_: the nice thing about the bootloader environment is the ability to freely use the hardware available, no rules to abide to, no other drivers or whatever to take into consideration
Nov 11 14:30:27 <zmatt> ol_: I don't know it
Nov 11 14:30:31 <zmatt> never heard of it
Nov 11 14:32:57 <ol_> zmatt: thx.:-)
Nov 11 14:34:34 <ol_> zmatt: where i get the debian images for the beaglebone?
Nov 11 14:34:48 <zmatt> http://elinux.org/Beagleboard:BeagleBoneBlack_Debian
Nov 11 14:35:11 <Spidler_> zmatt: Yea, feels a bit like being back in DOS ;)
Nov 11 14:35:16 <zmatt> you can find both the (wheezy) releases and the jessie snapshots
Nov 11 14:35:53 <zmatt> Spidler_: never done any DOS programming, I find the x86 architecture abhorrent
Nov 11 14:36:29 <zmatt> but I'm pretty sure under DOS you didn't have the level of comfort you have on a modern ARM-based SoC
Nov 11 14:37:26 <Spidler_> No, most definitely not.
Nov 11 14:37:41 <Spidler_> You had to tip-toe around things to avoid breaking stuff that just happened to be in memory at strange places
Nov 11 14:37:46 <zmatt> hehehe
Nov 11 14:37:50 <Spidler_> Old x86 was a very weeeird architecture
Nov 11 14:38:02 <zmatt> http://www.pagetable.com/?p=600
Nov 11 14:38:07 <Spidler_> With a ton of magic memory areas and other places
Nov 11 14:38:14 <Spidler_> hah
Nov 11 14:38:19 <zmatt> that's not really "old x86" though
Nov 11 14:39:03 <Spidler_> DOS, or the 80286 I used to do assembly on?
Nov 11 14:39:33 <zmatt> no I mean, it's still a weeeird architecture with magic memory areas and stuff
Nov 11 14:39:47 <Spidler_> oh, Yes indeed
Nov 11 14:39:51 <zmatt> if I understand correctly, gate A20 is still there, as a virtualized wire to the chipset
Nov 11 14:39:57 <zmatt> (not a physical one anymore)
Nov 11 14:40:03 <Spidler_> Used to be _more_ weird in the sense that you started in Real mode
Nov 11 14:40:07 <Spidler_> Noone does real mode anymore
Nov 11 14:40:16 <zmatt> still boots up in it doesn't it?
Nov 11 14:40:25 <Spidler_> And these days you don't really have a magic memory area for the video.. wait you do
Nov 11 14:40:28 <Spidler_> Meh
Nov 11 14:40:33 <zmatt> and then there's "big real mode"
Nov 11 14:40:35 <Spidler_> Mmmm
Nov 11 14:40:37 <Spidler_> And more
Nov 11 14:40:41 <zmatt> and magic memory areas of the SMM
Nov 11 14:41:04 <zmatt> I used to think SMM was the most horrific part of x86 but they just keep piling on more and more horror
Nov 11 14:41:42 <zmatt> (see also http://blog.invisiblethings.org/2015/10/27/x86_harmful.html )
Nov 11 14:42:11 <Spidler_> Yea, I've read that one too :)
Nov 11 14:42:37 <Spidler_> It's one of the best modern explanations of how x86 strings together
Nov 11 14:42:47 <Spidler_> It just also happens to explain that it sucks horribly for security
Nov 11 14:43:36 <zmatt> yeah, x86 is imho irrecoverable from a security pov.  not that the am335x is fully transparent, but I have a pretty good idea what secrom does, which on a GP device is fortunately "not much"
Nov 11 14:43:58 <Spidler_> MMm
Nov 11 15:52:33 <zmatt> wow, adding detailed fault reporting, while useful, has suddenly made my little baremetal test prog HUGE
Nov 11 15:52:39 <zmatt> it's almost 6 KB now
Nov 11 15:52:41 <zmatt> o.o
Nov 11 15:54:07 <Spidler_> Hehe
Nov 11 15:54:26 <Spidler_> Just now I'm chasing for a box that I can't find
Nov 11 15:54:38 <Spidler_> Well, couldn't. Turns out that it was one that wasn't supposed to be on that network.
Nov 11 15:55:10 <zmatt> heh, that helps
Nov 11 16:14:46 <Spidler_> My fixups have fixups...
Nov 11 16:14:48 * Spidler_ mutters
Nov 11 16:33:49 <jhuk> hello
Nov 11 16:34:26 <jhuk> is a power supply 5V with 1A enough for a bbb rev c. with an wifi-usb-dongle? or is it better 2A? thanks a lot!
Nov 11 16:38:25 <tbr> 2A is certainly better
Nov 11 16:38:34 <tbr> 1A is on the margin at best
Nov 11 16:39:35 <tbr> if price is a concern, buy it from aliexpress. they have a powersale today. you'll get the stuff in 2-4 weeks by china post ;)
Nov 11 16:40:18 <jhuk> tbr: :-)
Nov 11 16:41:06 <tbr> also power supplies don't like being operated near/above the limit, especially not the cheap chinese crap
Nov 11 16:41:30 <jhuk> tbr: ok, i take > 1A and no crap
Nov 11 16:41:34 <Xu> is this some sort of customer service?
Nov 11 16:41:42 <jhuk> tbr: thanks.:-)
Nov 11 16:42:04 <tbr> Xu: it's people helping each other
Nov 11 16:42:32 <Xu> i see. Has anyone have had experience with defective BBB within warranty?
Nov 11 16:43:09 <Xu> My BBB died on me all of sudden. I was using it yesterday.
Nov 11 16:43:53 <Xu> now when I connect it to either USB to my MAC, or 5V adapter, the power LED will blink once and nothing after that...
Nov 11 16:44:51 <Xu> or does anyone know the contact information of the manufacturer?
Nov 11 16:54:23 <kaak> is there a straightforward way to toggle the USB VBUS line on the USB host?
Nov 11 16:54:54 <kaak> I've got a misbehaving USB device (shockingly not related to musb) that may benefit hard power cycle for recovery.
Nov 11 16:55:53 <kaak> I.E. Does musb have hooks into the PM subsystem, and if so, what are those hooks?
Nov 11 17:28:03 <zmatt> vbus can be controlled yes
Nov 11 17:29:02 <zmatt> the cpu has a dedicated vbus-control output for that, presumably controlled from somewhere in the usb subsystem, but you can also just temporarily mux it to gpio and drive it low
Nov 11 17:42:35 <Abdullah> hi, has any tried running ardupilot on BBB
Nov 11 18:19:21 <zmatt> kaak: I just tried it, the gpio route works ... but after that the usb driver is in a state of complete confusion and I have to unload and reload the musb_dsps kernel module before it works again
Nov 11 18:23:37 <zmatt> kaak: feel free to try it: http://gerbil.xs4all.nl/vbus.tar.gz
Nov 11 18:23:52 <zmatt> afk, shopping
Nov 11 18:35:57 <[Butch]> Dumb question, how do you change the editor font in Cloud9?
Nov 11 18:43:15 <[Butch]> Failing that, does anyone know if there’s a Cloud9 IRC channel?
Nov 11 18:51:38 <stt_michael> https://c9.io/support - #cloud9ide
Nov 11 19:01:38 <[Butch]> stt_michael: Thank you!
Nov 11 19:10:40 <[Butch]> I’m back with another silly question, which y’all have mostly answered already in the past.
Nov 11 19:11:13 <[Butch]> I’m trying to get a flow meter working with the BBB using interrupts.  I’m using the sample code provided in the attachInterrupt call, but now it doesn’t want to work.
Nov 11 19:11:32 <[Butch]> I’m using exactly the same wiring and flow meter that has worked in the past, but today it has decided to do nothing.
Nov 11 19:12:22 <[Butch]> Actually, that’s not quite true.  It sometimes prints one interrupt’s worth of data.  Other times nothing.  In the past, it has happily spewed data while I was blowing in the meter until the interrupt detached.
Nov 11 19:12:27 <[Butch]> Any ideas why that might be?
Nov 11 19:12:45 <[Butch]> And no, I’m not ruling out, “because you’re stoopid(tm)."
Nov 11 19:13:58 <tbr> use scope with signal?
Nov 11 19:14:08 <[Butch]> If only I had one . . .
Nov 11 19:14:16 <[Butch]> This is a very low budget operation.  :-(
Nov 11 19:14:38 <tbr> aliexpress has a sale going on today...
Nov 11 19:14:58 <tbr> any cheap crap scope would probably work
Nov 11 19:15:31 <[Butch]> tbr: Work no pay for anything.
Nov 11 19:17:31 <tbr> likely an hour or two of work would already pay for more than one cheap crap scope
Nov 11 19:19:38 <[Butch]> tbr: See, there you go applying logic.  :-)
Nov 11 19:24:06 <tbr> merely stating the obvious (does not apply to damagement)
Nov 11 19:24:50 <[Butch]> trb: Agreed.  If only logic could apply to management.
Nov 11 19:29:02 <zmatt> kaak: never mind the gpio crap, you can do it via sysfs
Nov 11 19:29:24 <zmatt> though I will admit, in a not entirely obvious way :P
Nov 11 19:31:58 <zmatt> kaak: unloading the kernel driver (musb_dsps) also kills bus power btw
Nov 11 19:32:07 <tbr> btw: just to show that I mean it: http://www.aliexpress.com/item/Good-quality-1PCS-DSO138-2-4-TFT-Digital-Oscilloscope-Kit-DIY-parts-1Msps-with-probe-hot/32505358646.html
Nov 11 19:32:53 <[Butch]> tbr: OK, that’s just crazy.  But where do I get the ‘scope to debug the ‘scope that I’m building?
Nov 11 19:38:55 <tbr> [Butch]: for 10 more you can apparently get one ready assembled
Nov 11 19:39:05 <[Butch]> tbr: I just rebuilt the circuit using an Arduino and some old interrupt code I had lying around.  It works fine.  The flow meter is working.
Nov 11 19:39:18 <[Butch]> tbr: Ah, the chicken and egg problem solved for only $10.  :-)
Nov 11 19:39:40 <tbr> http://www.aliexpress.com/item/DSO138-2-4-TFT-DIY-Digital-Display-Oscilloscope-200KHz-Tester-1Msps-Analog-Bandwidth-w-Probe/32387248598.html?spm=2114.01020208.3.322.UEajDs&ws_ab_test=searchweb201556_3_79_78_77_92_91_80,searchweb201644_5,searchweb201560_9
Nov 11 19:42:45 <[Butch]> Any idea why the same flow meter with the same wiring wouldn’t work on the BBB using the sample interrupt code?  I know I had it working before using this code . . .
Nov 11 20:41:39 <zmatt> veremit: turned out it wasn't a stack overflow at all
Nov 11 20:42:18 <zmatt> in fact 512 seems to be plenty, I don't even reach 256 bytes
Nov 11 20:42:55 <veremit> what happened?
Nov 11 20:43:04 <zmatt> however, I now hate linkers more than ever before -.-
Nov 11 20:44:33 <zmatt> I had reserved a specific memory location for a specific purpose (long story), but the linker clearly felt working around it was too much effort and put code/data right on top of it
Nov 11 20:45:59 <zmatt> hence every boot one, essentially arbitrary, word of code or data would get clobbered when I wrote to that location, hence the completely random erratic behaviour
Nov 11 20:46:44 <zmatt> -.-
Nov 11 20:49:49 <zmatt> linkers can be fun like that... like a while ago already I also found it deciding that adhering to 4 KB alignment is clearly more important than the fact I told it memory started at 0x402f0400, and it put code at 0x402f0000 instead
Nov 11 20:53:34 <zmatt> but, everything now works again, dispite the heavy changes throughout my codebase the tests I wrote earlier also still work, so   git commit -m 'omg huge commit'    and pushed \o/
Nov 11 21:26:06 <opl> hi, i have further question to the configuration of the power supply. i have a 18Watt with the mods 3.0/4.5/5.0/6.0/7.5/9.0/12.0 V. which should I set? i've understand to set it on 5V. thanks!
Nov 11 21:29:50 <zmatt> for the beaglebone ?  5V yes
Nov 11 21:30:39 <opl> zmatt: yes, because there are read/write errors with another power supply. ok, and is 6.0 or 7.5 also ok. or do I destroy the beaglebone with these settings?
Nov 11 21:38:34 <opl> zmatt: maybe i've asked stupid stuff.
Nov 11 21:48:36 <opl> :-)
Nov 11 22:03:59 <opl> zmatt: thanks, bye! i've understand. the box runs on 5V!
Nov 11 22:53:03 * GenTooMan makes quiet capacitor poping sounds in the background.
Nov 11 23:05:24 <uudruid74> What's up my beagles?   I've got an external USB HD on my BBB.  Its getting HOT.  Is there an hdparm command that will get it to spin down and cool off?   Its running full tilt all the time
Nov 11 23:13:50 <veremit> uudruid-74 .. usually drives have their own power management .. eg. western digital 'green' drives
Nov 11 23:14:53 <veremit> uudruid-74 .. could try items from this Arch article .. https://wiki.archlinux.org/index.php/Hdparm#Power_management_configuration
Nov 11 23:15:33 <uudruid-74> This is a normal desktop drive with a usb drive.   Will check the link.  Thanks
Nov 11 23:36:25 <ninja232> Hi, how does the open source hardware with Beagle Bone work? Are the hardware designs GPL'd on github? Can I copy/modify it for my own design? If so, what are the recourses? Attribution?
Nov 11 23:39:37 <veremit> ninja232 .. http://elinux.org/Beagleboard:BeagleBoneBlack#Terms_of_Use
Nov 11 23:48:23 <ninja232> Thanx
Nov 12 02:11:16 <Set_> Hello...
Nov 12 02:11:42 <Set_> Does anyone follow the book, "Make: Getting Started with Beaglebone?"
Nov 12 02:12:08 <Set_> I am on page 68 to 73.
Nov 12 02:12:45 <Set_> I know that python does not accept the "@" symbol but I have to use it says the book for email purposes.
Nov 12 02:13:14 <uudruid-74> accept it for what?
Nov 12 02:13:33 <GenTooMan> you can use it in a string without issue.
Nov 12 02:13:40 <Set_> I am sending an email to myself and then alerting me if the door is open via calling "alertTime = 0."
Nov 12 02:15:13 <Set_> Oh.
Nov 12 02:15:20 <Set_> That is what I am trying to do.
Nov 12 02:15:33 <Set_> I am using it in a string but I am getting a syntax error.
Nov 12 02:15:49 <uudruid-74> pastebin
Nov 12 02:16:09 <Set_> Can I show you the line that is getting called a syntax error?
Nov 12 02:16:27 <Set_> it is line (8)...I will type the line here.
Nov 12 02:16:29 <Set_> Please hold.
Nov 12 02:17:05 <uudruid-74> the error is likely in the previous line
Nov 12 02:17:10 <Set_> msg['From'] = myAddress where myAddress is my google.com email address.
Nov 12 02:17:12 <Set_> Oh.
Nov 12 02:17:25 <Set_> Okay...please hold.
Nov 12 02:18:22 <Set_> see...the myAddress is not in quotation marks. Do you think this is the issue?
Nov 12 02:18:44 <Set_> so...
Nov 12 02:18:46 <uudruid-74> definately
Nov 12 02:18:59 <uudruid-74> strings should be quoted
Nov 12 02:19:27 <Set_> msg['From'] = "myAddress" <<<<<<< like this!
Nov 12 02:19:56 <Set_> I see...because my previous myAddress was in quotes.
Nov 12 02:20:15 <Set_> The book does not use quotes. This is what has me baffled.
Nov 12 02:20:36 <uudruid-74> Yup.  I don't use Python much, but .... it groks
Nov 12 02:21:00 <Set_> I will go for it and see what happens.
Nov 12 02:21:03 <uudruid-74> Probably a typo.  There are tutorials online
Nov 12 02:21:26 <Set_> I have two .py files and I have to call a part to my first .py file in my second amount of code.
Nov 12 02:21:38 <Set_> ...from my second .py file.
Nov 12 02:21:59 <Set_> Yea...I noticed all the tutorials out there. Thank you.
Nov 12 02:22:29 <Set_> You "da" man!
Nov 12 02:22:38 <Set_> I will be right back. I will check it out.
Nov 12 02:25:01 <Set_> I got all kinds of problems. Dang!
Nov 12 02:26:04 <uudruid-74> Indeed
Nov 12 02:26:08 <Set_> Hey <uudruid-74>
Nov 12 02:26:17 <uudruid-74> Yes?
Nov 12 02:26:23 <Set_> I know I am new but here goes it.
Nov 12 02:26:46 <Set_> How do I make a recommendation to Python 2.xx if they already moved to 3.xx?
Nov 12 02:26:56 <Set_> Impossible?
Nov 12 02:27:07 <uudruid-74> You don't.  They've moved on
Nov 12 02:27:12 <Set_> Dang.
Nov 12 02:27:16 <Set_> Okay.
Nov 12 02:27:45 <Set_> Can I show you two lines of what I got back in my code application?
Nov 12 02:28:00 <Set_> I ran it and this is what I got back...can I show you?
Nov 12 02:28:16 <uudruid-74> OK
Nov 12 02:28:19 <Set_> Cool!
Nov 12 02:28:22 <Set_> Hold please...
Nov 12 02:28:58 <Set_> Traceback (most recent call last):   File "emailer.py", line 18, in <module>     alertMe("Hey!", "You are the mickidity!")
Nov 12 02:29:19 <Set_> Do you get that?
Nov 12 02:29:42 <Set_> It goes on...there are about eight lines of text that repeated.
Nov 12 02:29:50 <uudruid-74> One line, but it doesn't tell me much
Nov 12 02:30:21 <Set_> Oh.
Nov 12 02:30:52 <Set_> root@beaglebone:~#   File "emailer.py", line 18, in <module> -bash: syntax error near unexpected token `newline'
Nov 12 02:30:57 <Set_> How about that?
Nov 12 02:30:58 <uudruid-74> And I'm not really a python guy, so I don't know all the ins and outs if its standard library.  I can fake it when I need to, but that's rare.
Nov 12 02:31:14 <Set_> Okay...no problem.
Nov 12 02:31:35 <Set_> I will rewrite it and see what happens.
Nov 12 02:31:39 <uudruid-74> Sounds like it expects more on the line.  You might wanna try a python channel :-)
Nov 12 02:31:59 <Set_> I think I have indented in the wrong places. I will check that.
Nov 12 02:32:08 <Set_> What is python channel?
Nov 12 02:33:01 <uudruid-74> I dunno, try #python ?
Nov 12 02:33:31 <uudruid-74> yeah
Nov 12 02:33:35 <Set_> I have some books and I go to python.org for their doc files.
Nov 12 02:33:49 <uudruid-74> type /join #python
Nov 12 02:33:52 <Set_> They have unlimited docs on python.org.
Nov 12 02:34:49 <uudruid-74> well, maybe not unlimited :-)
Nov 12 02:35:05 <Set_> I know but a bunch...
Nov 12 02:35:11 <Set_> bunches.
Nov 12 02:35:35 <uudruid-74> yup
Nov 12 02:37:26 <Set_> I got a "gai" error. I have to look this up. It says server not known.
Nov 12 02:37:50 <Set_> I probably need a new server code for smtp.gmail.com.
Nov 12 02:38:19 <uudruid-74> server code??
Nov 12 02:38:46 <Set_> yea...it says 587 is what to use for gmail. I use it but they have others.
Nov 12 02:39:19 <uudruid-74> thats a port number
Nov 12 02:39:26 <uudruid-74> not a code
Nov 12 02:39:31 <Set_> Oh.
Nov 12 02:39:33 <Set_> Thank you.
Nov 12 02:39:59 <Set_> I need a port number other than 587.
Nov 12 02:40:03 <Set_> Most likely.
Nov 12 02:40:15 <GenTooMan> are you trying to do snmp connection?
Nov 12 02:40:26 <uudruid-74> Hmmm ... what are trying to do?
Nov 12 02:40:44 <Set_> Yea...I am using my BBB to send me an e-mail via smtp.google.com.
Nov 12 02:40:47 <uudruid-74> 587 is smtp over ssl
Nov 12 02:41:03 <uudruid-74> You need ssl set up
Nov 12 02:41:09 <Set_> See...they want me to use tls.
Nov 12 02:41:29 <Set_> Oh...I wonder why they have tls here.
Nov 12 02:41:47 <uudruid-74> TLS and SSL is the same thing
Nov 12 02:42:12 <uudruid-74> more or less
Nov 12 02:42:19 <Set_> Basically.
Nov 12 02:42:34 <Set_> I am going to type ssl instead of tls and see what happens.
Nov 12 02:42:42 <uudruid-74> no
Nov 12 02:43:12 <GenTooMan> do much network relayed code Set_?
Nov 12 02:43:17 <Set_> I tried to type ssl instead of tls before but I had my myAddress without quotes.
Nov 12 02:43:19 <Set_> Why no?
Nov 12 02:43:19 <GenTooMan> related even.
Nov 12 02:43:44 <Set_> No.
Nov 12 02:43:47 <Set_> I do not.
Nov 12 02:43:53 <Set_> I am just starting.
Nov 12 02:43:56 <uudruid-74> Are you using a GMail account?  Like sending to yourself?  If so, just use port 25
Nov 12 02:43:58 <GenTooMan> ok SSL is a protocol not a program... so you have to use SSL library calls to use it ... the documentation should say that clearly.
Nov 12 02:44:13 <Set_> Okay.
Nov 12 02:44:32 <GenTooMan> anyone try the SIMPLE approach like uudruid-74 suggests. It's clear text.
Nov 12 02:44:48 <GenTooMan> anyone == anyhow ... argh getting tired.
Nov 12 02:45:12 <uudruid-74> me 2 .. head is killing me
Nov 12 02:45:17 <Set_> Hold on! I am using my BBB via online via PuTTY to save and run python code.
Nov 12 02:45:26 <Set_> I am using nano.
Nov 12 02:45:33 <Set_> I run my code via PuTTY.
Nov 12 02:45:41 <Set_> Bash script.
Nov 12 02:45:43 <GenTooMan> and ... your point?
Nov 12 02:45:52 <Set_> Nothing. I just wanted to say that.
Nov 12 02:46:28 <Set_> Hey uudruid? Why should I use 25 as a port?
Nov 12 02:46:31 <GenTooMan> oh ... hehehe. Yeah unix is nifty you could use a serial port connect to your windows machine using teraterm and do the same.
Nov 12 02:46:32 <uudruid-74> putty?   Are you using Windows?
Nov 12 02:46:46 <Set_> Ya...
Nov 12 02:47:12 <Set_> I have a couple of different systems but I am using Win 10 right now.
Nov 12 02:47:29 <uudruid-74> Set_  port 25 is smtp without ssl
Nov 12 02:47:30 <GenTooMan> He doesn't have the BBB hooked to a monitor and hub with mouse and keyboard obviously.
Nov 12 02:47:36 <Set_> I have some Linux OS on VMware.
Nov 12 02:47:45 <Set_> Nope.
Nov 12 02:48:11 <uudruid-74> best to wipe Windows, go 100% Linux
Nov 12 02:48:13 <Set_> I just got what happened. My Ethernet is not plugged in.
Nov 12 02:48:16 <Set_> Damn!
Nov 12 02:48:38 <uudruid-74> face palm
Nov 12 02:48:49 <Set_> Hhahahahaha.
Nov 12 02:48:58 <Set_> I plugged it in the wrong device.
Nov 12 02:49:02 * GenTooMan hands frying pan adjuster to Set_
Nov 12 02:49:08 <Set_> Ha!
Nov 12 02:49:25 * GenTooMan whispers "it's a precision adjustment tool really".
Nov 12 02:49:33 <Set_> Let me run this code again and see what happens now.
Nov 12 02:49:46 * uudruid-74 hands Set_ a ViewMaster
Nov 12 02:50:08 * GenTooMan watches the toaster explode noisily and hmmms.
Nov 12 02:50:10 <Set_> I got issues man. i am sorry.
Nov 12 02:50:40 <Set_> I got it!
Nov 12 02:50:42 <Set_> It works!
Nov 12 02:50:46 <Set_> It is ALIVE!
Nov 12 02:50:58 <GenTooMan> Don't worry we are just laughing with you. I've done "is it plugged in" after spending hours of building something. Yeah plug it in!
Nov 12 02:51:14 <Set_> Yep. I get it man. It is cool.
Nov 12 02:51:36 <GenTooMan> So Set_ will now work on his BBB based spam engine.
Nov 12 02:51:47 <Set_> my e-mail ran and it says and I quote, "Alert" "You are the mickidity!"
Nov 12 02:51:52 <Set_> Hhahahaha.
Nov 12 02:52:16 <uudruid-74> yes, you are
Nov 12 02:52:18 <Set_> That is cool.
Nov 12 02:52:27 <GenTooMan> You could do DOS attacks on yourself too now.
Nov 12 02:52:30 <Set_> Now...I have to use it to notify me when someone opens the door.
Nov 12 02:52:35 <Set_> Damn!
Nov 12 02:52:59 <Set_> I have this window sensor plugged into my BBB via breadboard now.
Nov 12 02:53:03 <uudruid-74> by camera?
Nov 12 02:53:10 <Set_> By e-mail!
Nov 12 02:53:13 <uudruid-74> oh ... sensor
Nov 12 02:53:15 <Set_> Yea.
Nov 12 02:53:29 <Set_> I could use a camera if needed.
Nov 12 02:53:34 <Set_> I would have to learn again.
Nov 12 02:53:43 <Set_> Oh and druid...it was a typo in the book.
Nov 12 02:53:46 <Set_> You were right.
Nov 12 02:53:50 <GenTooMan> You could set it up so that it snaps a shot when the door is ajar.
Nov 12 02:54:03 <uudruid-74> I used a webcam inside a server cabinet'  If someone opened the door, it would send the picture to my phone
Nov 12 02:54:09 <Set_> Ya...I was going to play, "No woman, No cry!"
Nov 12 02:54:18 <Set_> Ha!
Nov 12 02:54:45 <Set_> The door opens and No woman, No cry comes barreling out of the loud speaker!
Nov 12 02:54:51 <uudruid-74> No door sensor used, just monitored the camera for changes
Nov 12 02:54:56 <Set_> Oh.
Nov 12 02:54:59 <Set_> That is cool.
Nov 12 02:55:41 <Set_> Let me run this dooralert.py file and see if it works now!
Nov 12 02:55:50 <uudruid-74> low tech security but effective.  Just turned it off via ssh if I needed to open the cabinet
Nov 12 02:55:51 <GenTooMan> that would be funny if it shows the door open and then a big sledge hammer ...
Nov 12 02:56:03 <uudruid-74> LOL
Nov 12 02:56:46 <Set_> Hey...I got it.
Nov 12 02:56:50 <Set_> It works too.
Nov 12 02:57:09 <Set_> Now...I can apply a battery and scare someone!
Nov 12 02:57:19 <Set_> Well...a battery and a speaker.
Nov 12 02:57:44 <GenTooMan> you could use the "that was easy" message.
Nov 12 02:57:55 <GenTooMan> That should be "Creepy" for people.
Nov 12 02:58:25 <Set_> Creep town, USA is in Set_'s place!
Nov 12 02:59:57 <uudruid-74> OK
**** ENDING LOGGING AT Thu Nov 12 02:59:58 2015