**** BEGIN LOGGING AT Thu Jan 16 02:59:59 2020 Jan 16 03:00:00 sudo cp ../include/*.h /usr/local/include/ Jan 16 03:00:59 (that's it) Jan 16 03:01:01 does ../include/ = usr/include? Jan 16 03:01:08 no Jan 16 03:01:10 Okay. Jan 16 03:01:33 if I meant that, I would have written that Jan 16 03:01:40 I meant _exactly_ what I wrote Jan 16 03:01:46 Otay! Jan 16 03:03:47 As usual, you do not know exactly what I am doing but you figured out my issue w/ a couple of pastes. Nice! Jan 16 03:04:06 The source in the Makefile worked and make installed the source. Jan 16 03:04:51 ./major-tom-pru: error while loading shared libraries: libprussdrv.so: cannot open shared object file: No such file or directory <<<< This is my error. Jan 16 03:05:32 that looks like it's miscompiled Jan 16 03:05:38 Okay. Jan 16 03:05:40 No issue. Jan 16 03:05:50 how or why did it link to a libprussdrv.so ? Jan 16 03:06:00 No clue. let me check. Jan 16 03:06:28 just do "make clean" and then "make" in that project directory Jan 16 03:06:36 Okay. Jan 16 03:07:16 what is this anyway? I found the project on github but it has no description or anything Jan 16 03:07:28 Okay. That may have done it. There is something called the PRU Assembler Version 0.87. Jan 16 03:07:30 Oh. Jan 16 03:07:38 oh fail, remove these lines: https://github.com/HudsonWerks/bbb-pru-sandbox/blob/master/major-tom-pru.c#L108-L112 Jan 16 03:07:45 It is from a book I read years ago. Jan 16 03:07:53 you do _not_ need to be root to use prussdrv Jan 16 03:07:57 Let me show you the actual book github.com page. Jan 16 03:07:59 It is not that one. Jan 16 03:08:21 It is hudsenwerks but not the sandbox Jan 16 03:08:35 https://github.com/HudsonWerks/bbb-pru/tree/master/simple-tests ? Jan 16 03:08:44 OH. Yep. That is it. Jan 16 03:08:56 same thing, https://github.com/HudsonWerks/bbb-pru/blob/master/simple-tests/major-tom-pru.c#L108-L112 Jan 16 03:09:29 OH. Okay. Jan 16 03:09:32 Yep. Jan 16 03:09:38 remove those lines and it should work fine without root privileges Jan 16 03:09:45 Okay. Jan 16 03:10:47 root cause :D Jan 16 03:10:57 Doing? Jan 16 03:11:01 Sorry. Jan 16 03:11:05 Diong! Jan 16 03:13:27 I erased those numbers and poopoo. Jan 16 03:13:35 Please hold. Jan 16 03:13:46 I meant, I erased those lines. Jan 16 03:14:38 and...>>>> https://pastebin.com/pzaCKKP7 Jan 16 03:15:39 Forget it. Jan 16 03:15:44 I figured out what I did. Jan 16 03:16:01 failed to remove the } or something probably Jan 16 03:16:27 Yep. Jan 16 03:16:29 But... Jan 16 03:16:39 Look here: ./major-tom-pru: error while loading shared libraries: libprussdrv.so: cannot open shared object file: No such file or directory. Jan 16 03:16:51 Same error while loading the file. Jan 16 03:17:03 either w/ ./file or sudo ./file Jan 16 03:17:13 wtf how or why is it still linking to a shared library Jan 16 03:17:30 does /usr/local/lib/libprussdrv.so exist? Jan 16 03:17:41 Let me check. Jan 16 03:18:31 it does but w/out the .so suffix. Jan 16 03:18:46 i.e. "it doesn't" Jan 16 03:18:48 there is a libprussdrv in /usr/local/lib/ Jan 16 03:18:56 Right. Does not! Jan 16 03:19:17 Right, no .so Jan 16 03:19:22 so how the fuck is it linking to a libprussdrv.so Jan 16 03:19:27 if none is around Jan 16 03:19:40 Good question. Let me investigate further. Jan 16 03:20:03 Yea. There is. Jan 16 03:20:06 did you setup weird environment variables or something? Jan 16 03:20:07 ? Jan 16 03:20:13 "There is" ? Jan 16 03:20:19 libprussdrv.so! Jan 16 03:20:22 where? Jan 16 03:20:23 Yes. Jan 16 03:20:31 In /usr/local/lib/ Jan 16 03:20:43 ... I just asked you said there wasn't Jan 16 03:20:47 how did it get there? Jan 16 03:21:06 Good question. Let me look. Jan 16 03:21:19 sudo rm -f /usr/local/lib/libprussdrv.so /usr/local/libprussdrvd.* Jan 16 03:21:28 There are four files, two .so and two .a files. Jan 16 03:21:32 Okay. Jan 16 03:21:39 there should only be libprussdrv.a Jan 16 03:22:24 Okay. Now there is only one. Jan 16 03:22:28 after that try 'make clean' and 'make' again in the simple-tests Jan 16 03:22:34 OKay. Jan 16 03:23:13 prussdrv_open() failed Jan 16 03:23:13 Segmentation fault Jan 16 03:23:18 This is what happened. Jan 16 03:23:23 lol, nice Jan 16 03:23:26 Ha. Jan 16 03:23:28 Oh. Jan 16 03:23:36 I have to set up the pru w/ config-pin! Jan 16 03:23:42 lemme see if there's an obvious bug in the code Jan 16 03:23:50 Okay. Jan 16 03:23:50 no, config-pin is unimportant Jan 16 03:23:54 Oh. Jan 16 03:24:16 Is that b/c the source is just reading words? Jan 16 03:24:27 I mean, it's important if you want pru to be able to use certain direct gpios, but failing to set it up won't cause code to crash Jan 16 03:24:38 Oh. Jan 16 03:24:39 Okay. Jan 16 03:24:54 and the code doesn't use any pins Jan 16 03:24:59 Right. Jan 16 03:25:01 so there's nothing to config-pin Jan 16 03:26:08 The README in simple-tests/ stated to use echo BB-BONE-PRU-01 >/sys/devices/bone_capemgr.9/slots Jan 16 03:26:09 can you share the output of: ls -l /dev/uio* Jan 16 03:26:12 Sure. Jan 16 03:26:18 So, it is older source. Jan 16 03:26:23 yeah ignore that Jan 16 03:26:37 Okay. Jan 16 03:26:58 cannot access /dev/uio* no such file or dir. Jan 16 03:27:13 Should I change that in uEnv.txt? Jan 16 03:28:15 if you want to use uio you need to configure that in /boot/uEnv.txt yes, I've definitely told you that before Jan 16 03:28:57 I know. Jan 16 03:29:06 I was just wondering if it was necessary right now? Jan 16 03:30:22 that entirely depends on whether it is necessary for this software to actually work instead of crash right now :P Jan 16 03:30:33 Ha. I will go and test it. Jan 16 03:31:31 Reboot! Jan 16 03:36:49 That did it. Jan 16 03:37:04 Why does rproc not work when uio does work on this source? Jan 16 03:37:43 the two are mutually exclusive and incompatible Jan 16 03:38:16 software written to use uio won't work with rproc and vice versa Jan 16 03:38:33 Oh. So, this source was typed up in regards to uio on purpose? Jan 16 03:39:02 How did you know that idea was true? Jan 16 03:39:11 If not one, then another? Jan 16 03:39:48 anything that uses pasm and/or libprussdrv requires uio-pruss Jan 16 03:40:00 rproc cannot load pru executables produced by pasm Jan 16 03:40:08 Oh. Neat. Now, I know. Okay. Jan 16 03:40:30 so, is rproc for the .c loader/linker? Jan 16 03:40:53 and uio is for the assembly? Jan 16 03:41:40 Anyway...I got it now. I found some good docs. a while back but I have been unable to find them again on ti.com's site. Jan 16 03:41:41 you can write assembly if you use rproc, but you then need to use clpru as assembler instead of pasm Jan 16 03:41:49 Oh! Jan 16 03:41:51 Okay. Jan 16 03:42:13 note that my py-uio python library also (as the name implies) requires uio Jan 16 03:42:32 So, it is about the linker and not the language. Jan 16 03:42:49 universal-io? Jan 16 03:43:23 ? Jan 16 03:43:49 rproc can only load ELF executables Jan 16 03:44:35 in case of uio the loading is done in userspace hence what you can load depends on what you use: libprussdrv can only load raw executables as produced by pasm, my py-uio library can load both formats Jan 16 03:44:55 Oh. Okay. I am thinking that this is far beyond my knowledge now. I...oh. Okay. Jan 16 03:46:22 So, there are libraries specifically for pru, these libraries are used at the head of the source in either .c files or .asm files, and the linkers are either pasm or clpru for the command line when running the source. Jan 16 03:46:25 Does that sound right? Jan 16 03:46:38 ??? Jan 16 03:46:43 Okay. Forget it. Jan 16 03:46:53 I will go and read more. Jan 16 03:47:44 Every time, when I think I can make sense out of how complicated the pruss is still, I make errors in my language (English). Jan 16 03:51:10 Where are all the gpio stuff for pruss on the am335x? Jan 16 03:51:32 ? Jan 16 03:51:48 Sorry. stuff = tutorials Jan 16 03:52:27 oh, no idea Jan 16 03:52:29 I mean. In .c files, would I just use case to handle motors? Jan 16 03:52:32 Me neither! Jan 16 03:53:35 if you use config-pin to configure pins to "pruin" mode, they will show up in R31 of the pru core. conversely, any pins that are configured to "pruout" mode will output bits of R30 of the pru core Jan 16 03:54:30 Like w/ that python script you helped me w/, instead of if state == "Blah":, would it be case state == "Blah";? Jan 16 03:54:34 Does that make any sense? Jan 16 03:54:57 overviews of which pin can be configured as which pruin/pruout signal of which pru core can probably be found in multiple places, but my pins spreadsheet ( https://goo.gl/Jkcg0w ) is one of those places Jan 16 03:55:01 no it doesn't make any sense Jan 16 03:55:12 why are you using C ? Jan 16 03:55:36 Oh. I thought those command line tools, pasm and clpru, were for .c files only. Jan 16 03:56:30 you're confusing code that runs on pru and code that runs on linux but is responsible to managing pru Jan 16 03:56:45 there's no reason to use C for the latter, you can use py-uio Jan 16 03:57:04 Oh. Hmm. Jan 16 03:57:07 why are you trying to use pru anyway? for what purpose? Jan 16 03:57:30 I wanted to learn something new and bypass the am335x to get speed of motion. Jan 16 03:57:43 ??? Jan 16 03:57:47 Right? Jan 16 03:58:11 I thought the pruss bypassed the am335x and went straight to the pins for usage. Jan 16 03:58:16 you know what, I'm done for today with getting confused by trying to understand you Jan 16 03:58:22 Okay. Jan 16 03:58:25 No issue. Jan 16 03:58:36 Lever! Never leave home w/out it! Jan 16 03:59:49 @zmatt: Please forgive me. TI.com changed their wiki pages. I am trying to refresh old ideas and bring them to new light in the 2020 decade. That is all. Jan 16 04:00:40 As I usually do, my time consists of making the new-old and the old-new. Sort of. Anyway, later for now. Jan 16 04:04:33 good night... Jan 16 04:13:33 Yep! Jan 16 05:30:03 So, is the m4, the pru on the AI? Jan 16 05:30:46 Oh and what is a mailbox on the pru core? Jan 16 05:31:45 I see it is a section of the core but does it just serve a purpose of relaying or is it an entire different portion of the core for many things? Jan 16 05:32:22 Okay, I forgot. Good evening. Dang it. Jan 16 06:49:21 set_: m4 and pru have absolutely nothing to do with each other. mailboxes are just on-chip peripherals that can be used for (inefficient) interprocessor communication if you want to (they're not all that useful), they are not part of any core Jan 16 08:27:54 Oh. Jan 16 17:02:23 the versions of python on the include eMMc image for debian are kind of old. Even for 2018_09 Jan 16 17:02:54 Thought of the moment at least :D Jan 16 23:37:48 should be 2.7.16 or 17? and 3.whateveryoufancy Jan 16 23:38:00 3.9 if you're crazy. Jan 16 23:38:05 3.7 good enough for me. Jan 17 00:47:50 Yes... seemed odd but I didn't update it to the latest I only got what was available in 2019_07 so .. hmmm 10 months old. :D Jan 17 01:26:39 heh Jan 17 01:29:39 Well yeah :D 10.2 is out and GOK what they were using for the BBB Jan 17 01:32:15 They used debian 9.5 the version of the 9 series in 2019_07 was 9.9 ... :D Jan 17 01:32:44 :eyeroll: Jan 17 01:34:25 Well the good thing is it works. Working is good. Jan 17 01:34:34 definitely! Jan 17 01:34:44 can't be overstated that :p Jan 17 01:35:21 https://www.digikey.com/eewiki/display/linuxonarm/BeagleBone+Black#BeagleBoneBlack-Debian10 lgtm though :D Jan 17 01:36:11 I cut my teeth on these instructions ~8-10 year ago Jan 17 01:43:44 billions and billions of seconds ago... Jan 17 01:44:13 yup sorta :p Jan 17 01:48:43 using the beagleboard or the beagle bone? I had the former. The DLP projector I had was always too bright and would flip out. Seems TI had a few bugs on how it treated the DPL (way back when) Jan 17 02:13:08 Speaking of TI, does anyone use CCS for projecting source to their BBB or other family of boards? Jan 17 02:13:51 This is what I am doing: https://processors.wiki.ti.com/index.php/PRU_Training:_Hands-on_Labs! Jan 17 02:14:21 I cannot figure out anything for whatever reason. Does anyone know how to use CCS v9 w/ older Labs? Jan 17 02:15:25 I am on 6. on that page online on where it says to add files under projects. The add files option is not lit. I cannot choose it for some reason. Jan 17 02:17:33 pru_cfg.h is not found in my list of resources. I just got it. Jan 17 02:19:13 Oh. I remember why now. Their git page is down. Jan 17 02:19:38 right click on the project name and click new-> then click header or source file... and you should ALWAYS import a local copy of files :D Jan 17 02:21:15 I just found the pru_cfg.h file online on github. For some reason, TI's git is down for maintenance. Jan 17 02:21:33 Got it. Jan 17 02:22:29 As soon as I ask? sheesh. Jan 17 02:24:29 I've used CCS a LOT? Jan 17 02:24:36 Okay. Jan 17 02:24:45 It's just eclipse you realize Jan 17 02:24:52 So, I am using it for the millionth time. Right. Jan 17 02:25:00 Just eclipse. Jan 17 02:25:07 <-- 500 millionth time Jan 17 02:25:14 Nice! Jan 17 02:25:23 I keep erasing it and reinstalling it. Jan 17 02:25:31 It's been over a decade :| Jan 17 02:25:35 Ha. Jan 17 02:28:05 GenTooMan: Hey. Jan 17 02:28:11 installing eclipse should be a snap you just download the installer run it and poof it installs exactly the correct (whatever) and whatever eclipse extras you might want. I use PyDev with eclipse clunky but good enough to work with PyGObject and wxPython. I even knew enough to modify nrsc5-gui (https://github.com/cmnybo/nrsc5-gui) Jan 17 02:28:34 Aw. Jan 17 02:29:47 I looked. The photo of that person walking "into" the wall is neat. Jan 17 02:30:00 I never heard of that gui. Jan 17 02:30:59 Well the guy who wrote it is in Seattle area. If you invest a few cents into an RTL-SDR 3 you can play with HD radio monitor air traffic pick up LoRa transmissions... etc. Jan 17 02:31:41 Aw. Jan 17 02:31:45 Oh listen to ham radio weather radio and well everything up to 1.7GHz you can pick up. That aside did your robot ever work? Jan 17 02:32:15 Now...I have the pru_cfg.h found but my main(): void file is failing. Jan 17 02:32:24 Yea! Jan 17 02:32:41 The robot worked and I am bringing it the bot to the faire soon. March! Jan 17 02:33:39 I am using qmake. I think this is the issue. Jan 17 02:44:41 I have an unrecognized global name. That CCS stuff has some error enhancements that are nice! Jan 17 02:57:25 CT_CFG? Look here: https://pastebin.com/Wks0f9ET. Jan 17 02:57:41 Come on, who is interested in making PRU source work still? Jan 17 02:58:11 __R30 is giving me errors too. **** ENDING LOGGING AT Fri Jan 17 02:59:57 2020