**** BEGIN LOGGING AT Sat Aug 20 02:59:58 2016 Aug 20 04:03:32 erm, does disabling the emmc on the green wireless also take out wlan0? Aug 20 04:05:30 maybe Aug 20 04:06:07 the eMMC appears to use MMC1 on the BBGW Aug 20 04:07:01 and I'm assuming on the green (wireless) I want to use dtb=am335x-bonegreen-overlay.dtb Aug 20 04:07:19 microsd is using MMC0 Aug 20 04:07:34 although I have tried the various black options that disable emmc and wlan0 is still out Aug 20 04:08:24 so you cannot use the microsd and the wifi at the same time Aug 20 04:08:36 wait Aug 20 04:08:37 no, I can Aug 20 04:09:42 okay then you may have to create an overlay that zonks out the MMC0 for the BBGW Aug 20 04:10:14 didn't you just say microsd was on MMC0? Aug 20 04:10:57 I mean MMC1 Aug 20 04:10:59 duh Aug 20 04:11:04 sorry Aug 20 04:11:46 lets look at what the overlay that I pointed out earlier does and also look at am335x-bonegreen-overlay Aug 20 04:12:53 ##BeagleBone Green: eMMC disabled Aug 20 04:13:01 dtb=am335x-bonegreen-overlay.dtb Aug 20 04:13:44 it looks like this is the droid you were looking for afterall Aug 20 04:14:14 but note this is probably BBG not BBGW Aug 20 04:15:21 hmmmm Aug 20 04:17:57 so is bonegreen-overlay disabling mmc1 by basically not assigning it anything? (guesses based on what it has for mmc2 but not mmc1) Aug 20 04:19:54 the beaglebone kernel support is a mess Aug 20 04:22:44 trying to find the source dts for the green wireless now Aug 20 04:23:23 https://github.com/beagleboard/bb.org-overlays/blob/master/src/arm/univ-bbgw-00A0.dts is this what you mean? Aug 20 04:24:06 https://github.com/AlexShiLucky/BB-linux/blob/fb1cd5e7d9c6648415f52ae7e442938e8435668b/arch/arm/boot/dts/am335x-bonegreen-wireless.dts Aug 20 04:24:08 closest thing I've found that mentions the green wireless Aug 20 04:24:23 that is an overlay Aug 20 04:25:17 which kernel are you using? Aug 20 04:25:20 4.4? Aug 20 04:25:39 https://github.com/beagleboard/linux/blob/4.4/arch/arm/boot/dts/am335x-bonegreen-wireless.dts Aug 20 04:25:47 4.4.12, yeah Aug 20 04:26:02 the overlay must override this dts then Aug 20 04:27:37 doh that is what I just found Aug 20 04:28:20 so then next step is to figure out which lines are the problem Aug 20 04:28:41 and create an overlay that disables the appropriate interface Aug 20 04:30:34 hm, should I try using the bonegreen-wireless one first and see if that fixes it? Aug 20 04:32:55 https://github.com/beagleboard/linux/blob/4.4/arch/arm/boot/dts/am335x-bonegreen-overlay.dts Aug 20 04:32:58 yeah Aug 20 04:33:18 but you will likely loose wifi in the process Aug 20 04:35:51 am335x-bonegreen-overlay disables both mmc1 and mmc2 Aug 20 04:36:12 but it should also for the board to boot Aug 20 04:36:56 *allow Aug 20 04:37:33 I feel like I'm missing something. what's the difference between *-overlay.dts and the other *.dts? Aug 20 04:38:46 dts overlay is something that goes on top to modify, not the full one you boot with Aug 20 04:39:19 I believe that the -overlay versions of the kernel dts's require overlays to enable features Aug 20 04:39:28 okay got it. even though they all seem to sit in the same place with the same extentions Aug 20 04:40:17 the overlays dtso's are then applied over the top of these Aug 20 04:40:51 https://github.com/beagleboard/linux/tree/4.4/arch/arm/boot/dts Aug 20 04:40:59 all of these are top level Aug 20 04:41:25 all of these are overlays Aug 20 04:41:28 https://github.com/beagleboard/bb.org-overlays/tree/master/src/arm Aug 20 04:41:48 I think Aug 20 04:42:54 so that bonegreen-wireless.dts is the full thing for the board. would still need an overlay to do things like disable the emmc. as I don't see any wireless-specific overlays I've either got to muck with stuff myself or live with whatever the non-wireless-specific ones do Aug 20 04:43:32 (at least, I'm assuming it's technically possible to disable the emmc and leave the wifi intact) Aug 20 04:43:41 it is Aug 20 04:43:51 of course Aug 20 04:44:37 did you try dtb=am335x-bonegreen-overlay.dtb? Aug 20 04:45:17 hm that's in the main kernel not the overlay repo, maybe it's not really a "dts overlay"?? Aug 20 04:45:34 yep. seems to work for booting with the ADC (worked in one problematic case at least) but does kill the wifi Aug 20 04:45:42 yeah Aug 20 04:45:50 that is what I expected Aug 20 04:46:09 you can enable the wifi with an overlay now if needed Aug 20 04:47:29 is wifi needed for development? Aug 20 04:47:31 depends how much trouble it would be. it would certainly be nice and save me some grief. if enabling it though is going to be a pain then maybe not worth it Aug 20 04:47:56 how are you configuring the PRU pins now? Aug 20 04:48:31 just some calls out to pin-config (or whatever the name is, I never get it right) Aug 20 04:49:32 you can also edit the .dts (that's what I did) Aug 20 04:49:49 that will require building the kernel Aug 20 04:49:50 where it says "pick one" you uncomment one of the includes Aug 20 04:50:25 actually I got an RCN repo where it was just building dbt Aug 20 04:52:08 https://github.com/RobertCNelson/dtb-rebuilder Aug 20 04:52:44 I think that was it (workstation that I was using for development currently down) Aug 20 04:53:22 lets just make an overlay for am335x-bonegreen-overlay Aug 20 04:57:41 http://pastebin.com/MdcG1bTu Aug 20 04:58:19 create a file called wifi.dts and plop that in it Aug 20 04:59:06 alright Aug 20 04:59:28 dtc -O dtb -o wifi-00A0.dtbo -b 0 -@ wifi.dts Aug 20 04:59:36 then run this ^ Aug 20 05:00:10 cp wifi-00A0.dtbo /lib/firware Aug 20 05:00:16 then this ^ Aug 20 05:00:45 echo wifi > /sys/devices/platform/bone_capemgr/slots Aug 20 05:00:51 then this ^ Aug 20 05:01:13 then see if the wifi comes up Aug 20 05:02:03 oops forgot to add ti,beaglebone-green to the compatible list Aug 20 05:02:39 http://pastebin.com/fQhEi1vG Aug 20 05:02:51 this is the updated dts Aug 20 05:03:04 let me know if it works Aug 20 05:04:46 I would guess there is more to set up, that enabling isn't enough Aug 20 05:05:01 hm. looks like something is getting mangled. Aug 20 05:05:05 Error: wifi.dts:16.13-18 syntax error Aug 20 05:05:05 FATAL ERROR: Unable to parse input tree Aug 20 05:05:08 because non-wireless dts is missing this one https://github.com/beagleboard/linux/blob/4.4/arch/arm/boot/dts/am335x-bonegreen-wl1835.dtsi Aug 20 05:05:16 not spotting it Aug 20 05:05:35 ah Aug 20 05:06:39 if it's going to be much trouble to get up, then I should probably just move on. not critical, just (rather) nice Aug 20 05:07:24 I vote move on, USB network has been working fine here Aug 20 05:07:41 yeah Aug 20 05:07:54 project is on the line Aug 20 05:08:59 I've had some trouble with it on my laptop that makes me want to avoid it, but maybe the wireless will be a little better behaved then the non. so far it's seemed that way Aug 20 05:09:49 what are you running on the laptop side, recent ubuntu ? Aug 20 05:09:58 arch Aug 20 05:12:17 I have gentoo on this one, all the usbnet stuff as modules, both cdc and rndis Aug 20 05:12:21 actually, if it gets bad I'll just use serial. recovered my usb-serial thingy the other day Aug 20 05:12:30 That works Aug 20 05:12:44 go forth and collect data Aug 20 05:47:28 Visaoni: try this dts for the eMMC overlay Aug 20 05:47:33 http://pastebin.com/qgVJY891 Aug 20 05:49:38 should enable wifry if my calculations are correct Aug 20 05:50:24 no way for me to fully test it because I don't have a green Aug 20 05:51:12 I do, I'll save & try it later... Aug 20 05:51:44 it looks plausible this time Aug 20 05:52:02 it compiles on my BBB and mostly loads Aug 20 05:52:18 [40108.087593] bone_capemgr bone_capemgr: slot #7: Incompatible with baseboard for 'wifi' Aug 20 05:52:29 [40108.095846] bone_capemgr bone_capemgr: slot #7: Failed verification Aug 20 15:30:41 m_w: there? Aug 20 15:30:59 m_w: not sure why.. but my rproc is not working in my beaglebone black? Aug 20 15:31:12 m_w: I just tried blink led test program... Aug 20 15:31:20 m_w: can you give me quick hints? Aug 20 15:32:11 m_w: first... I loaded this overlay: https://github.com/kiran4399/bb_blue_api/blob/master/install_files/2016-05-01/overlays/BBB_SERVO-00A0.dts Aug 20 15:32:36 m_w: second I compiled the blink led program which is present in pru software package.. Aug 20 15:33:24 m_w: and copied as /lib/firmware/am335x-pru0-fw Aug 20 15:33:34 m_w: and then loaded the kernel module pru_rproc.. Aug 20 15:33:42 m_w: is there something which I am missing? Aug 20 15:53:53 kiran4399: what kernel version ? Aug 20 15:54:54 also, check your dmesg if the PRUs are actually being booted with your fw Aug 20 16:29:36 ZeekHuge: 4.4.12 Aug 20 16:53:58 ZeekHuge: did you get my previous message? Aug 20 16:54:17 ahh ... nope. Aug 20 16:54:37 ZeekHuge: I am using 4.4.12 Aug 20 16:54:53 ZeekHuge: and no.. I checked in dmesg.. it was not coming.. Aug 20 16:55:45 so that is the problem .. Aug 20 16:55:48 ahh .. Aug 20 16:56:13 probably something wrong with the way you are booting prus Aug 20 16:56:14 should try this Aug 20 16:56:14 https://www.zeekhuge.me/post/ptp_blinky/ Aug 20 16:58:43 ZeekHuge: you have such good notes!! Aug 20 17:02:46 kiran4399_: Thanks :) Aug 20 17:06:57 hey ZeekHuge you have a dead link at elinux.org for your blog, it didn't get updated after you oved it Aug 20 17:07:06 http://elinux.org/BeagleBoard/GSoC/2016_Projects#Project:_BeagleScope Aug 20 17:08:21 Wormo: okay, will remove it for now and add it back once the blog is up with all the documentation. Aug 20 17:09:20 might as well just point it to your current blog, it already has useful stuff as kiran4399_ was just noticing :) Aug 20 17:09:49 okay. Aug 20 17:11:39 ZeekHuge: thanks.. it shows in the dmesg now.. Aug 20 17:12:59 kiran4399_: Cool ! should try your code now. Aug 20 17:15:29 thanks for helping out ZeekHuge Aug 20 17:17:12 ZeekHuge: if you want to use P8.40... what should this value be? Aug 20 17:17:13 https://github.com/ZeekHuge/BeagleScope/blob/port_to_4.4.12-ti-r31%2B/examples/firmware_exmples/pru_blinky/PRU_gpioToggle/PRU_gpioToggle.c#L52 Aug 20 17:18:58 r31_7 Aug 20 17:19:17 bit 7 Aug 20 17:19:38 m_w: configuring them as output right?? so r30.. i guess... Aug 20 17:19:42 just make that 0xFFFFFFFF and any pin would do the task Aug 20 17:20:13 and then you need to configure the GPIO mux Aug 20 17:20:44 ZeekHuge: tell me one thing.. if I don't connect register... will it not glow?? Aug 20 17:20:50 *resistor Aug 20 17:21:05 oh right r30 |= 0x200 Aug 20 17:21:37 it will burn the GPIO Aug 20 17:22:09 ZeekHuge: you mean the led? Aug 20 17:22:12 that simple .. Aug 20 17:22:13 in case of blinky Aug 20 17:22:42 nope probably the PIn on your bbb Aug 20 17:22:56 connecting resistor lower than 470ohms is risky .. Aug 20 17:23:32 ZeekHuge.. hmm.. I don't have a resistor now.. but I am eager to test it.. Aug 20 17:24:06 I'll suggest you to wait .. Aug 20 17:25:03 kiran4399_: you don't have an oscilloscope? Aug 20 17:25:25 m_w: no :( Aug 20 17:25:39 kiran4399_: or since this toggle really slow a multimeter? Aug 20 17:27:53 ZeekHuge: the main suggestion I have for your wrap-up is to make a quickstart, for somebody new to to your project be able to get up to your current state quickly and then start working with you on it Aug 20 17:28:29 if you did that, then you'd want to highlight the quickstart in your final report Aug 20 17:29:57 m_w: remember you told me to use copy_to_user() function?? Aug 20 17:30:36 m_w: how can I test if some value is written to a shared memory? Aug 20 17:32:32 kiran4399_: what do you mean? Aug 20 17:33:53 Wormo: okay so what I am working now and my plan further is to add in-source comments everywhere in the source. Then add Readme at various levels in the repo, describing the content inside directories. And the README section will have a contribution section. then I will be using Issue tracker to manage a TODO list. Aug 20 17:34:35 Apart from this ... the blog will have the final report describing the progress .. Aug 20 17:34:57 and giving an overview of the project/ Aug 20 17:35:20 What else would you suggest to make it easy to start with ? Aug 20 17:36:02 m_w: did you get my earlier message? Aug 20 17:36:09 kiran4399: yeah Aug 20 17:36:23 kiran4399: what are you checking for? Aug 20 17:37:01 kiran4399: you could load a default state to the shared memory when loading the driver perhaps Aug 20 17:39:52 ZeekHuge: literally a step-by-step quickstart, pulling from some of the other stuff you have written Aug 20 17:40:46 Wormo: Hmm okay. Will do. Aug 20 17:42:54 maybe you've run across other projects where there is a nice quickstart, it can be very encouraging to get something working quickly Aug 20 17:43:56 the opposite of you guys having spent multiple days at the beginning trying to nail down which kernel to use for rpmsg, which kernel build scripts, etc Aug 20 17:44:58 if you weren't signed up for a summer project, just considering this as one possibility among others for a personal project, you might have moved on to try something else... Aug 20 17:59:46 Wormo: yeah .. that is true. Will make a good quickstart. Aug 20 18:23:44 has anyone seen Aug 20 18:23:47 alexhiam around? Aug 20 18:26:39 nope Aug 20 18:29:22 I see that you updated the servo sysfs driver Aug 20 18:29:39 does it still data abort? Aug 20 18:30:26 kiran4399 ^^ Aug 20 18:30:58 https://github.com/kiran4399/bbb_pru_firmware/blob/master/servo/servo_sysfs.c#L186 Aug 20 18:31:39 this won't compile, you have an extra } Aug 20 18:32:43 yeah.. acutally I sorted that out.. did not commit yet.. Aug 20 18:54:40 jkridner, could you check your email please Aug 20 21:21:13 Wormo, mdp, There? Aug 20 21:22:04 I just wanted to ask that in the documentation, I am putting up a diagram, so should I upload it github itself? Aug 20 21:22:22 Or should I put in drive? Which would be better? Aug 20 23:54:11 jkridner, I finished the document.. and waiting for your feedback. Thanks Aug 20 23:59:57 jkridner, please check my pull requests on bone101 and bonescript for python phase **** ENDING LOGGING AT Sun Aug 21 02:59:58 2016