**** BEGIN LOGGING AT Wed Mar 14 03:00:01 2018
Mar 14 08:14:08 <Nbeagle> Hello
Mar 14 08:25:18 <Nbeagle> No one is here?
Mar 14 08:29:05 <Nbeagle> helo
Mar 14 09:00:27 <blathijs> Nbeagle: There probably is, but people are not often inclined to respond to hellos :-) If you have a question, best just ask it without finding someone to answer it first :-)
Mar 14 09:00:37 <blathijs> Nbeagle: Even so: Welcome!
Mar 14 10:44:05 <prabhakarlad> Hi all, I was having trouble with enabling uart10 in uboot for x15, has anyone tried enabling it ? any pointers that would help would be appreciated
Mar 14 11:26:26 <zmatt> it still weirds me out that the am572x has such a crazy number of uarts
Mar 14 11:30:54 <zmatt> prabhakarlad: you're more likely to get help if you explain what you've tried and what problems you've encountered
Mar 14 11:32:19 <zmatt> then your question wouldn't be limited to people who have tried enabling uart10 in u-boot (which is probably not many people)
Mar 14 11:58:58 <prabhakarlad> zmatt:  I have maged to get the second stage u-boot.img working on uart10, but doesn work for first stage MLO
Mar 14 12:01:48 <prabhakarlad> zmatt: following are my changes https://pastebin.com/LHEAQc5x
Mar 14 13:41:35 <prabhakarlad> zmatt: fyi managed to get it on uart 10 now :)
Mar 14 14:12:56 <config-pin> zmatt: hey, my pru is behaving really weird
Mar 14 14:13:10 <config-pin> it was working fine for generating the pwem a few days ago
Mar 14 14:13:34 <config-pin> now although it has the correct data value stored in its dram, its not giving any outputs on the servo pins
Mar 14 14:13:43 <config-pin> i checked show-pins, and they are configured as pru 1 out
Mar 14 14:14:05 <config-pin> the other pru being used for decoding a ppm signal is workin gperfectly fine
Mar 14 14:14:08 <config-pin> working perfetly*
Mar 14 14:14:23 <config-pin> also, the pwm pru blinks an led and stuff as usual
Mar 14 14:14:32 <config-pin> but it is not generating a pwm output
Mar 14 14:15:33 <config-pin> any idea why this is happening
Mar 14 14:23:49 <zmatt> config-pin: how likely do you think it is that I could possibly give an answer on that based on the information you just gave?
Mar 14 14:24:12 <zmatt> (which basically boils down to "my program doesn't work. some other program does work, but my program doesn't.  what's wrong with it?"
Mar 14 14:24:15 <zmatt> )
Mar 14 14:24:25 <blathijs> zmatt: You mean you are not psychic? Then how do you know all this other stuff!? :-p
Mar 14 14:25:45 <zmatt> :P
Mar 14 14:27:29 <zmatt> I recently also saw a funny video about asking questions... https://www.youtube.com/watch?v=53zkBvL4ZB4
Mar 14 14:42:41 <flj> you cannot ask on the beaglebone forum if you do not use google log in?
Mar 14 14:50:19 <zmatt> I think you can?
Mar 14 14:51:55 <config-pin> zmatt: nvm sorry, was loading the wrong file -_-
Mar 14 14:52:20 <flj> how? If I select ask question, a google sign in window dispalys.
Mar 14 14:53:26 <zmatt> flj: I'm not sure, but I think it should act like a normal mailing list too.  when I quickly google for it I found several suggestions on how to subscribe without a google signin but I'm not sure if they're still valid
Mar 14 14:55:06 <config-pin> zmatt: just a gen c question here, if i had say a while loop in a code, and i had a function in that whiel loop that i wanted to call at a fixed frequency
Mar 14 14:55:12 <config-pin> which library shoudl i use??
Mar 14 14:55:18 <config-pin> or what hsould i google?
Mar 14 14:55:21 <config-pin> should*
Mar 14 14:56:20 <LetoThe2nd> config-pin: while loop and fixed rate don't properly align
Mar 14 14:56:41 <config-pin> LetoThe2nd: wht do you mean?
Mar 14 14:57:32 <LetoThe2nd> while loop is idiomatic for: eat up all cpu and spin as fast as possible. fixed rate on the other hand asks for a timer that triggers.
Mar 14 14:59:28 <config-pin> so if i had to have a function being called at a fixed frequency what would i do?
Mar 14 14:59:58 <LetoThe2nd> config-pin: are you talking about linux userland, or what?
Mar 14 15:00:33 <config-pin> LetoThe2nd: yes
Mar 14 15:00:49 <flj> what email address does the beaglebone black email list have?
Mar 14 15:01:21 <config-pin> LetoThe2nd: wait userland linux refers to all the linux c libraries right??
Mar 14 15:07:21 <zmatt> config-pin: you mean you want a periodic timer
Mar 14 15:07:33 <config-pin> zmatt: yes
Mar 14 15:08:18 <zmatt> you can use one of many libraries that provide them (e.g. sd-event)
Mar 14 15:08:43 <zmatt> or if your needs are very basic and you only need one timer you could create a timerfd and block on that
Mar 14 15:09:34 <LetoThe2nd> i personally like boost::asio
Mar 14 15:09:38 <LetoThe2nd> *duckandrun*
Mar 14 15:09:44 <config-pin> right so the thing is, i dont know any of this, so what should I google lol?
Mar 14 15:09:48 * zmatt runs after LetoThe2nd with an elephant gun
Mar 14 15:10:02 * LetoThe2nd stomps on zmatt 
Mar 14 15:10:13 <zmatt> config-pin:        #include <sys/timerfd.h>
Mar 14 15:10:15 <zmatt> woops
Mar 14 15:10:21 <zmatt> sorry, paste in wrong window
Mar 14 15:10:44 <LetoThe2nd> config-pin: um. i'd suggest, look for a shopping link to k&r. after that, for kerrisks api book. :)
Mar 14 15:11:14 <config-pin> the linux programming book?
Mar 14 15:11:33 <LetoThe2nd> afaik michael kerrisk only wrote one book.
Mar 14 15:11:40 <zmatt> config-pin: if you want to use a versatile event loop, sd-daemon is a decent choice
Mar 14 15:11:51 <zmatt> sd-event sorry
Mar 14 15:11:53 <zmatt> man sd-event
Mar 14 15:12:08 <zmatt> see also /usr/include/systemd/sd-event.h
Mar 14 15:12:20 <LetoThe2nd> he said systemd!
Mar 14 15:12:30 * LetoThe2nd runs after zmatt with an elephant gun
Mar 14 15:12:31 <config-pin> zmatt: cool ill check it out
Mar 14 15:12:34 <zmatt> if you prefer html versions of manpages: https://www.freedesktop.org/software/systemd/man/sd-event.html
Mar 14 15:12:50 <zmatt> if you're a systemd hater or need portability to non-linux systems, there are plenty of alternatives
Mar 14 15:18:13 <config-pin> zmatt: umm noob q, but qhat about stuff like <signal.h> or <time.h>
Mar 14 15:18:17 <config-pin> http://man7.org/linux/man-pages/man2/timer_create.2.html
Mar 14 15:18:39 <zmatt> you don't want timer_create, those are a big headache
Mar 14 15:19:42 <zmatt> if you don't want to use any event loop library (i.e. if you only need this single timer and don't need to receive any other events of any sort) then use a timerfd: https://pastebin.com/KauR5djq
Mar 14 15:22:09 <config-pin> and are calls to ok cool
Mar 14 15:22:09 <config-pin> thanks
Mar 14 15:36:42 <Rajat> Hey there
Mar 14 15:37:07 <cyril__> hi!
Mar 14 15:37:19 <Rajat> Can i develop a startup idea while working with you
Mar 14 15:37:30 <Rajat> ?
Mar 14 15:38:27 <Rajat> Kindly acknowledge my query
Mar 14 15:41:41 <LetoThe2nd> oO( ACK? )
Mar 14 15:42:18 <tbr> RST
Mar 14 15:42:23 <LetoThe2nd> FIN
Mar 14 15:44:07 <cyril__> anyone knows a good guide to compile Linux kernel for the BeagleBone Blue?
Mar 14 15:44:28 <tbr> RCNs wiki should explain that well
Mar 14 15:44:40 <tbr> you might need to adjust the DTS
Mar 14 15:45:26 <cyril__> The official repo doesn't work?
Mar 14 15:45:33 <cyril__> https://github.com/beagleboard/linux
Mar 14 15:46:07 <cyril__> I tried to compile with make bb.org_defconfig
Mar 14 15:46:27 <cyril__> there is a boneblue.dts file
Mar 14 15:46:46 <tbr> "doesn't work"
Mar 14 15:47:13 <LetoThe2nd> tbr: "it sits on the couch all day and watches tv."
Mar 14 15:47:32 <jkridner[m]> ARCH=arm?
Mar 14 15:47:47 <jkridner[m]> which branch? which compiler?
Mar 14 15:47:48 <cyril__> I use buildroot to compile actually
Mar 14 15:47:56 <cyril__> on 4.4 branch
Mar 14 15:49:37 <jkridner[m]> 4.4, 4.9 and 4.14 are all good.
Mar 14 15:49:51 <cyril__> https://i.imgur.com/5PjGeYL.png
Mar 14 15:50:12 <cyril__> https://i.imgur.com/Pi0nMGb.png
Mar 14 15:50:17 <cyril__> here is my buildroot config
Mar 14 15:50:40 <cyril__> I have no idea about de partitions
Mar 14 15:51:25 <cyril__> I'm not sure about the bb.org_defconfig
Mar 14 15:51:37 <cyril__> or the image format (vmlinux?)
Mar 14 15:52:49 <cyril__> are patches needed?
Mar 14 15:54:06 <zmatt> look for a buildroot support channel?
Mar 14 15:54:19 <zmatt> #buildroot
Mar 14 15:54:50 <cyril__> this isn't buildroot related
Mar 14 15:55:11 <jkridner[m]> I did https://github.com/jadonk/buildroot/tree/add-pocketbeagle recently
Mar 14 15:55:14 <cyril__> fully beaglebone blue dependant
Mar 14 15:55:33 <jkridner[m]> would boot on Blue equally.
Mar 14 15:56:49 <cyril__> what's different from the official buildroot repository?
Mar 14 15:57:21 <cyril__> I already tried the beaglebone_defconfig board file from buildroot without success
Mar 14 15:57:28 <cyril__> but I base my new board config on it
Mar 14 16:03:41 <cyril__> https://github.com/cyrilfr/buildroot/tree/master/board/beagleboneblue
Mar 14 16:04:06 <cyril__> https://github.com/cyrilfr/buildroot/blob/master/configs/beagleboneblue_defconfig
Mar 14 16:09:14 <cyril__> this sd card image boots :
Mar 14 16:09:15 <cyril__> https://eewiki.net/display/linuxonarm/BeagleBone+Blue
Mar 14 16:09:37 <cyril__> but this is totally different from what you have on your buildroot
Mar 14 16:10:13 <cyril__> it doesn't use the official linux kernel by the way
Mar 14 17:03:04 <Ravi_> Hello zmatt
Mar 14 17:03:15 <Ravi_> I have my BeagleBone Black now.
Mar 14 17:03:20 <Ravi_> Are you around?
Mar 14 17:07:15 <Ravi_> I have my uEnv.txt
Mar 14 17:07:36 <Ravi_> unmodified pasted here. https://pastebin.com/LSqRL18R
Mar 14 17:07:58 <Ravi_> This is a new board. No SW update or nothing. Just as is.
Mar 14 17:08:18 <zmatt> step 1: reflash it with a recent image
Mar 14 17:08:22 <zmatt> this is something really ancient
Mar 14 17:08:25 <Ravi_> root@beaglebone:/boot# uname -a Linux beaglebone 4.1.15-ti-rt-r43 #1 SMP PREEMPT RT Thu Jan 21 20:13:58 UTC 2016 armv7l GNU/Linux
Mar 14 17:08:37 <zmatt> exactly, more than 2 years old
Mar 14 17:09:24 <Ravi_> OK. the one I used earlier is OK right? bone-debian-9.3-iot-armhf-2018-03-05-4gb.img
Mar 14 17:09:50 <zmatt> yes, that's fine
Mar 14 17:10:07 <Ravi_> I will be gone for 30 mins or something because of this new SW update.
Mar 14 17:10:18 <zmatt> is it that slow? damn
Mar 14 17:10:31 <zmatt> slow sd card?
Mar 14 17:10:43 <Ravi_> Well, the sites say about 30-45 mins.
Mar 14 17:10:54 <Ravi_> 65MBps Strontium
Mar 14 17:11:03 <Ravi_> SD card is prepared.
Mar 14 17:11:10 <zmatt> I'm only occasionally around btw.  but in general don't ask if someone is around, just ask your question (other people may be able to help)
Mar 14 17:11:38 <zmatt> I don't think it takes that long, but maybe I'm wrong
Mar 14 17:11:40 <Ravi_> I try my last known good source first :-)
Mar 14 17:12:09 <zmatt> I'm not your personal assistant
Mar 14 17:12:28 <Ravi_> Now, I just uncomment the line #cmdline=init=/opt/scripts/tools/eMMC/init-eMMC-flasher-v3.sh
Mar 14 17:12:39 <Ravi_> I know. I am sorry, I did not mean to offend you.
Mar 14 17:13:38 <Ravi_> I regard you as a person who knows things about BBB. I seek you help and advice.
Mar 14 17:16:41 <zmatt> I just meant: it's kinda pointless to ask if I'm around or not.  just ask your question
Mar 14 17:17:04 <zmatt> if you don't, there's nothing for me to help you with.  if you do, then maybe I'll see it or maybe someone else can help you
Mar 14 17:19:13 <Ravi_> I agree.
Mar 14 17:19:20 <Ravi_> I will do so.
Mar 14 17:29:23 <Ravi_> debian@beaglebone:~$ uname -a Linux beaglebone 4.9.82-ti-r102 #1 SMP PREEMPT Thu Feb 22 01:16:12 UTC 2018 armv7l GNU/Linux d
Mar 14 17:30:00 <Ravi_> So, I am no a new kernel.
Mar 14 17:33:03 <Ravi_> Uncommenting this line #disable_uboot_overlay_video=1 in /boot/uEnv.txt
Mar 14 17:33:50 <flo_od_> hey there!
Mar 14 17:34:17 <flo_od_> is it somehow possible to dump a series of "download" packages at once?
Mar 14 17:34:21 <flo_od_> :D
Mar 14 17:34:41 <flo_od_> around 2000 packages, would take a while to do that by hand xD
Mar 14 17:35:47 <flo_od_> it is an .tdc capture - using data center
Mar 14 17:37:49 <Ravi_> sudo config-pin p9.11 uart
Mar 14 17:37:57 <Ravi_> sudo config-pin p9.13 uart
Mar 14 17:41:57 <flo_od_> are you talking to me Ravi_ ? The problem is, it is already captured and i cannot do it again
Mar 14 17:42:44 <flo_od_> i am searching for something like a script which does "search for next download block" -> "export"
Mar 14 17:42:58 <Ravi_> No @ flo_od
Mar 14 17:43:06 <Ravi_> I am am seeking help here myself.
Mar 14 17:43:25 <Ravi_> I am trying to get all the 4 UARTs to function as UARTs.
Mar 14 17:43:41 <blathijs> flo_od_: What kind of download packages are you talking about exactly?
Mar 14 17:44:22 <flo_od_> it's an firmware updated :)
Mar 14 17:44:39 <flo_od_> i want to extract the firmware since i am developing a driver
Mar 14 17:44:57 <flo_od_> and I would like to dig a bit deeper to find undocumented functions
Mar 14 17:45:40 <flo_od_> https://i.imgur.com/UCpuwId.png
Mar 14 17:48:02 <zmatt> flo_od_: why are you asking about some random usb capture program here?  and more importantly, why did you think anyone would be able to guess from your question that that was what it was about?
Mar 14 17:50:07 <flo_od_> zmatt: I don't own an beagle board, but the capture is made with one - therefore i googled and found that "data center" seems to be the tool which creates those "tdc" files
Mar 14 17:50:34 <flo_od_> i thought that it is the "default" application
Mar 14 17:50:41 <flo_od_> isn't it?
Mar 14 17:52:49 <flo_od_> lol wait, beagleboard != beagle xD
Mar 14 17:53:19 <flo_od_> sorry, i was talking about the beagle analyzer
Mar 14 17:53:37 <zmatt> lol
Mar 14 17:53:44 <zmatt> okay now it makes sense
Mar 14 17:53:45 <flo_od_> ^^
Mar 14 17:53:48 <flo_od_> damn it
Mar 14 17:53:55 <zmatt> yeah, no, sorry, wrong beagle :)
Mar 14 17:53:58 <flo_od_> thank you anyway! :D
Mar 14 17:54:03 <blathijs> :-)
Mar 14 17:54:17 <zmatt> I didn't actually think of that, even though we have a beagle usb analyzer too
Mar 14 17:56:08 <flo_od_> okay, i will search for expertise somewhere else then ;D bye!
Mar 14 17:59:46 <zmatt> in case anyone happens to want an example of a periodic timer using sd-event, here: https://pastebin.com/Y8UCF8pn  (originally made it for someone who already left, so I'll just leave it for the irc logs I guess)
Mar 14 18:17:30 <Ravi_> @zmatt - the changes and the procedure you gave worked. I am now able to do a loop back on UART4. (P9.11 and P9.13)
Mar 14 18:18:13 <Ravi_> Is there a way to make the config-pin "permanent" until changed again?
Mar 14 18:18:46 <Ravi_> https://pastebin.com/83PQczfp
Mar 14 18:21:15 <Ravi_> Internet says, edit your bashrc file.
Mar 14 18:22:07 <Ravi_> any other methods?
Mar 14 18:27:39 <Ravi_> I edited the .bashrc and it worked fine for UART4. I will do the same for the other UARTs.
Mar 14 18:27:51 <Ravi_> and report back.
Mar 14 18:29:31 <zmatt> uhh .bashrc makes no sense
Mar 14 18:29:36 <zmatt> you'll want to put it in /etc/rc.local
Mar 14 18:51:08 <Ravi_> I checked out all the 4 UARTs I want to use - UART1, UART2, UART4 and UART5
Mar 14 18:51:41 <Ravi_> They all work now. Thanks for all the help.
Mar 14 18:51:53 <Ravi_> I will now move the changes to rc.local.
Mar 14 18:56:54 <Ravi_> There is no rc.local in my BBB
Mar 14 19:06:05 <Ravi_> Ok, created this file, changed the permissions and reboot. It works like I want it to.
Mar 14 19:08:28 <Ravi_> Signing off! Bye for now. Thanks for all the help zmatt.
Mar 14 19:09:45 <bien_salad> ls
Mar 14 19:09:48 <bien_salad> oops.
Mar 14 19:09:52 <bien_salad> Hello!
Mar 14 19:50:49 <bien_salad> 2nd C++ program situated on Eclipse. Boo!
Mar 14 19:50:54 <bien_salad> But...heh? I got it.
Mar 14 19:51:18 <bien_salad> That damn software gives me bad feelings. I am going to keep using it.
Mar 14 19:51:27 <bien_salad> ...
Mar 14 19:51:31 <bien_salad> 42 was the answer.
Mar 14 19:51:40 <bien_salad> ...
Mar 14 19:52:06 <bien_salad> Anyway...I 3D Printed a shaft holder for my BBG/Motor Bridge Cape mini mower.
Mar 14 19:52:34 <bien_salad> <<<< getting closer to idea formulation every, single day.
Mar 14 21:22:23 <NTQ> Hi. Somehow my PRU code does not change any values in the shared ram. What could be the problem? https://hastebin.com/mojawusori.cpp
Mar 14 21:25:12 <NTQ> here is shared_ram.h: https://hastebin.com/nemepagiri.cpp
Mar 14 21:26:10 <zmatt> I don't see anything obvious wrong with it
Mar 14 21:26:21 <NTQ> hm
Mar 14 21:26:30 <zmatt> you're sure your code successfully runs at all?
Mar 14 21:27:02 <zmatt> can you confirm the pc is at your infinite while-loop ?
Mar 14 21:27:04 <NTQ> It compiles fine in ccs so I thought it would be correct. How could I check if it runs successfully?
Mar 14 21:27:57 <zmatt> how are you loading the code onto the pru?
Mar 14 21:28:47 <NTQ> It looks like the code in the while loop does not do anything if I check the outputs with an logic analyzer.
Mar 14 21:29:15 <zmatt> what do you mean?
Mar 14 21:29:20 <NTQ> zmatt: I run the code the same way I did it before. Before I used pasm to compile.
Mar 14 21:29:46 <NTQ> Oh, sorry, my complete code toogles some bits in __R30
Mar 14 21:30:40 <zmatt> pasm produces a raw binary that people typically load using uio, clpru produces an ELF executable
Mar 14 21:30:40 <NTQ> zmatt: https://hastebin.com/vogeqogaqo.cpp
Mar 14 21:31:03 <NTQ> I use this function: prussdrv_exec_program(1, "toggle_led.out");
Mar 14 21:31:19 <zmatt> well that explains why the pru is crashing
Mar 14 21:31:55 <NTQ> oh, okay
Mar 14 21:32:18 <NTQ> what am I making wrong then?
Mar 14 21:33:38 <zmatt> clpru produces an ELF executable and therefore requires an ELF loader, such as remoteproc-pru (I've started a tiny bit of work on a ELF loader for uio-pruss but haven't had much time)
Mar 14 21:33:54 <zmatt> you may be able to extract the raw code and data sections and load it via prussdrv
Mar 14 21:38:03 <NTQ> I found a sample code here:http://catch22.eu/beaglebone/beaglebone-pru-ipc/
Mar 14 21:38:25 <NTQ> I thought compiling will be straight forward as it was with pasm.
Mar 14 21:38:34 <zmatt> you can use  readelf -l toggle_led.out  to determine the segments that need to be loaded into pru memory (the LOAD segment with flags "R E" is for iram, other LOAD segments are for dram)
Mar 14 21:38:49 <zmatt> this isn't about compiling, it's about loading it onto pru
Mar 14 21:39:15 <zmatt> and no, it isn't as straightforward
Mar 14 21:41:51 <NTQ> zmatt: readelf -l looks like this: https://hastebin.com/imiqisiqal.pas
Mar 14 21:43:10 <zmatt> okay?
Mar 14 21:44:32 <NTQ> Can I now convert the out-file to a bin file which can be loaded with prussdrv_exec_program?
Mar 14 21:47:01 <zmatt> so this is saying you need to copy 0xb0 (MemSiz) bytes from offset 0x34 (Offset) in the file to address 0x0 (PhysAddr) in iram
Mar 14 21:48:15 <zmatt> the second LOAD statement says 0x100 bytes (MemSiz) are allocated at address 0 in dram, but the FileSiz is zero for that so nothing needs to be copied from file
Mar 14 21:48:35 <zmatt> maybe it needs to be zeroed, not sure, might be a good idea to do that
Mar 14 21:49:02 <zmatt> and the entry point is at 0x0
Mar 14 21:50:26 <zmatt> I just checked the kernel: if FileSiz < MemSiz then the remaining part of that segment is indeed zeroed
Mar 14 21:53:46 <NTQ> I still don't understand why this is so complicated. Are there some compiler options to create the correct file directly? I don't wanna do these steps every time I test a new code. :-/ Or can I use an other command instead of prussdrv_exec_program?
Mar 14 22:02:21 <zmatt> TI's envisioned way of using pru is by using clpru to compile code and remoteproc-pru to load it
Mar 14 22:02:52 <zmatt> they don't care anymore about uio-pruss users
Mar 14 22:02:59 <zmatt> (nor pasm users for that matter)
Mar 14 22:03:39 <zmatt> of course you don't have to agree with either of these... I'd still use pasm and uio-pruss, but if you're going to mix the two approaches you may have to get creative
Mar 14 22:08:32 <NTQ> I switched to C because I had a problem with my ASM code where I was not able to change a value inside the shared ram on the host side and then read it in the pru. The other ways around it was no problem. I was able to send 3,200,000 Bytes per second.
Mar 14 22:09:03 <NTQ> I thought it is a problem with caches and so I decided to try C code.
Mar 14 22:09:17 <zmatt> switching to a different programming language because you have a bug in your program sounds rather radical
Mar 14 22:09:23 <NTQ> :-D
Mar 14 22:10:03 <NTQ> learning by doing and failures.
Mar 14 22:11:18 <NTQ> When I write a value to the shared ram before I start the pru binary I can read that value inside the pru. When I change a value in the shared ram while the pru is running the pru does not see the change.
Mar 14 22:11:46 <NTQ> The other way around it works flawlessly.
Mar 14 22:12:58 <zmatt> without source code I can't guess what you might be doing wrong
Mar 14 22:18:10 <NTQ> Of course. But I don't want to publish my code publicly.
Mar 14 22:47:08 <zmatt> that is your choice, then you'll just need to figure out the bug yourself
Mar 14 22:49:27 <NTQ> Maybe I try to minify my code and then paste it again. But this will take a while.
Mar 14 22:50:12 <zmatt> or make a simple test that works and then figure out what your main program is doing wrong
Mar 14 22:53:35 <NTQ> I have a structure which is mapped to the shared memory and write to it. Reading on the same structure is no problem when the pru changed something in it.
Mar 14 22:53:44 <NTQ> I will try it with a simpler program again.
Mar 14 22:57:34 <zmatt> just have pru copy from one location to another in a loop
Mar 14 22:59:58 <NTQ> I don't use any pru copy functions. I use prussdrv_map_extmem in the beginning
Mar 14 23:04:52 <zmatt> I mean, as a simple test have pru just copy some value from one shared memory location to another in an infinite loop
Mar 14 23:05:25 <zmatt> then on the arm core, while the pru is running, write the first location and verify it shows up in the other location
Mar 14 23:06:02 <NTQ> yes. That's my idea too
Mar 14 23:06:22 <NTQ> Like this: http://credentiality2.blogspot.de/2015/09/beaglebone-pru-ddr-memory-access.html
Mar 14 23:11:52 <zmatt> ddr?
Mar 14 23:12:01 <zmatt> you weren't using ddr in your C code example
Mar 14 23:13:49 <zmatt> unless I misunderstood what you meant
Mar 14 23:14:17 <zmatt> you hadn't mentioned ddr memory until now anyway
Mar 14 23:21:23 <zmatt> ddr works fine for me too
Mar 14 23:22:40 <zmatt> oh he left
Mar 15 00:30:37 <GenTooMan> I hate it when that happens
Mar 15 01:17:21 <zmatt> hah, I didn't know you could monitor for hostname changes by opening /proc/sys/kernel/hostname and sticking it into a poll()/epoll based event loop
Mar 15 01:31:51 <calculus> hmm... someone pinged me, but my scrollback doesn't go far enough to see my mention
Mar 15 01:35:24 <GenTooMan> I take it you don't have a continuous log to a file.
Mar 15 01:57:46 <zmatt> 2018-03/2018-03-08.log.bz2:02:56:35 < jkridner|p> calculus: I'm in Pasadena!
Mar 15 01:57:46 <zmatt> 2018-03/2018-03-08.log.bz2:02:56:42 < jkridner|p> calculus: you coming to SCaLE tomorrow?
Mar 15 02:03:17 <calculus> thanks zmatt, totally missed the messages, but I did see jkridner at SCaLE this past weekend
Mar 15 02:03:36 <calculus> GenTooMan: nope, no log to file
**** ENDING LOGGING AT Thu Mar 15 03:00:01 2018