**** BEGIN LOGGING AT Wed Aug 21 02:59:59 2013 Aug 21 12:36:49 av500: ping Aug 21 13:35:52 yo Aug 21 14:01:32 vmayoral: what are your thoughts on converting bb_dc_motors to C and use the sysfs interface for pwms? Aug 21 14:01:52 vmayoral: that should make it work with every pwm controller linux supports Aug 21 14:03:10 av500: any link to Angstrom that fits into eMMC, not the flasher? Aug 21 14:03:23 the one from beagleboard.org is 4gigs in size Aug 21 14:04:32 vvu|Mobile: what is inside the flasher? Aug 21 14:04:40 it should have an emmc image inside, no? Aug 21 14:04:45 what else does it flash? Aug 21 14:05:02 it runs from sdcard the flasher and drops files on eMMC Aug 21 14:05:42 yes Aug 21 14:05:51 so it must have files to drop, no? Aug 21 14:06:34 but i want to implement something that downloads from intenret and just drops it Aug 21 14:06:39 it does fdisk ; mkfs.vfat ; cp mlo ; mkfs.ext4 ; tar -x Aug 21 14:06:47 not dd Aug 21 14:07:01 koen: no problem. I'll do it. Aug 21 14:07:36 koen: ah right Aug 21 14:08:15 koen: so no more need for native post build stuff? Aug 21 14:09:28 koen: just detected that it's not possible to install ROS Hydromedusa from the eu.feeds.angstrom-... in a clean Angstrom image because there's a circular conflict. I'm going to create an Angstrom image with ROS Hydromedusa and make it available for the community Aug 21 14:09:55 koen: some users asked for it contacting me directly Aug 21 14:10:39 vvu|Mobile: you could also just dd a working emmc image from a BBB Aug 21 14:11:39 yeah i can get one Aug 21 14:11:57 but i want to implement the download from internet function in my project Aug 21 14:14:16 right Aug 21 14:14:31 but that is independent Aug 21 14:14:37 we can host a test file anywhere for a test Aug 21 14:18:11 yep Aug 21 14:18:15 had time to test my stuff? Aug 21 14:21:46 {o\ Aug 21 14:21:48 /o\ Aug 21 14:21:55 * av500 hides Aug 21 14:22:21 av500: to compare 2 strings in bash [ $compare == "testString" ]; Aug 21 14:24:32 av500: will be back for the meeting and with the progress report online Aug 21 14:24:33 brb Aug 21 14:25:02 ok Aug 21 14:33:33 compare in sh is done with just one = e.g. if [ "a" = "b" ] Aug 21 14:52:08 av500: for v2012.12 postinst are still run, for v2013.06 they are 99% gone Aug 21 14:53:45 koen: so wrt the gsoc, would you say we should look into the mkfs/untar way? Aug 21 14:54:28 I'm leaning towards "yes" Aug 21 14:54:37 but I don't have strong opinion on that Aug 21 15:35:56 av500: except making sdcards what is the next step ? Aug 21 15:51:15 hah Aug 21 15:51:27 I finally added an alarm to the .ics Aug 21 15:51:54 now I just need to ignore my ADD for the next 9 minutes Aug 21 15:55:21 hehe Aug 21 15:56:25 gm all Aug 21 15:56:48 evening ! Aug 21 15:56:48 hello! Aug 21 15:56:56 good morning! Aug 21 15:57:13 * hatguy just blogged Aug 21 15:58:55 prpplague: look! a squirel! Aug 21 15:59:21 koen: get a gun Aug 21 15:59:45 koen: i thought you were out on honeymoon Aug 21 15:59:59 hello Aug 21 16:00:58 hi Aug 21 16:02:19 * jkridner|work takes a quick look at http://anujdeshpande.wordpress.com/2013/08/21/week-9-progress-with-spi, http://prpplague.github.io/Energia/blog/2013/08/07/how-to-use-the-wikis/, http://beagleros.wordpress.com/2013/08/21/recipes-status, http://minix-i2c.blogspot.com/2013/08/report-9-august-12-august-18.html, http://beagleboard-gsoc13.blogspot.com/2013/08/iio-list-feedback-update-short-hiatus.html and http://meanbot.wordpr Aug 21 16:02:20 * jkridner|work ess.com/2013/08/19/frustration/ Aug 21 16:02:31 vvu: is http://beagleboot.vdev.ro/?p=204 your latest? Aug 21 16:02:33 GM everyone. sorry i'm a bit late Aug 21 16:02:47 <_av500_> hello all Aug 21 16:02:51 pushed a new one couple of minutes agi Aug 21 16:02:52 ago Aug 21 16:02:58 _av500_: how dare you speak to us like that! Aug 21 16:03:06 jkridner|work: http://beagleboot.vdev.ro/?p=207 Aug 21 16:03:14 jkridner|work: that's not my latest Aug 21 16:04:19 prpplague: that's after september 14th, still 3 weeks of freedom left :) Aug 21 16:04:29 <_av500_> ah Aug 21 16:04:31 jkridner|work: http://meanbot.wordpress.com/2013/08/21/progress/ Aug 21 16:04:42 vvu: since you are looking for testers, can you make one page(wiki?)/post with a description of what the app does, what is needed to run it and the steps for acquiring running it and post to beagleboard@googlegroups.com? Aug 21 16:04:50 thanks hatguy Aug 21 16:04:57 koen: oh, for some reason i thought it was august 14th Aug 21 16:05:00 jkridner|work: yep, 2night will be there Aug 21 16:05:19 guess we can all openly pre-congratulate koen Aug 21 16:05:36 <_av500_> nah, he might chicken out Aug 21 16:06:20 like having a APK for people to test (I started a few times on the testing but it is quite a hassle to setup ) Aug 21 16:06:27 koen getting married too? Aug 21 16:06:32 we have 3 gsoc weddings right ?? Aug 21 16:06:35 thats 3 marriages this gsoc season :) Aug 21 16:06:40 keesj_: the problem is changing the kernel for people Aug 21 16:06:45 who's the third? Aug 21 16:06:52 the apk is easy to deploy with the files inside it Aug 21 16:06:58 mine is on 19th oct.. Aug 21 16:07:06 ZubairLK: ahh right... Aug 21 16:08:00 hatguy: I married last thrusday Aug 21 16:09:43 vvu, _av500_, keesj_: can it really be just an APK? can a built kernel for various phones and instructions on installing it be provided as well? Aug 21 16:09:47 keesj_: congratulations! I vaguely remember making a blunder with your name I guess Aug 21 16:10:28 i can provide the APK and instructions how to build the kernel *what configs to enable*. Aug 21 16:10:33 <_av500_> jkridner|work: yes, that can be provided Aug 21 16:10:34 GSoW Aug 21 16:11:04 +1 Aug 21 16:11:07 hahahaha Aug 21 16:11:39 thanks to those who provided updates. Aug 21 16:12:01 anybody worried about finishing their projects on time or being able to document them clearly? Aug 21 16:12:07 do i get to do some bhangra wedding dances? Aug 21 16:12:32 * vvu wants some testers :) anybody has testers to sell? 3+1 free deal ? Aug 21 16:13:16 I find it hard to track progress on that side. java code did not change over the last week and the post was also there. If a kernel mod is really needed (and I still find that hard te believe) it is going to be hard to sell Aug 21 16:13:18 if you can give me instructions for testing on a G1, Galaxy S II or Moto X since those are my Android-capable phones. Aug 21 16:13:44 jkridner|work: moto x lives up to the hype ? Aug 21 16:13:54 jkridner|work: a little tip on how to document my work. I think its mostly self documenting.. One final blog post on how to get it all working I guess.. Aug 21 16:14:00 idem here (I have cyanogemod galaxy SII) Aug 21 16:14:03 haven't tried it yet.... not even sure that is the real model # they gave me... haven't even touched it. Aug 21 16:14:25 keesj_: great. so I can follow the same instructions. Aug 21 16:14:28 keesj_: i have my changes locally Aug 21 16:14:39 <_av500_> keesj_: the kernel mod is needed on e.g. N7 Aug 21 16:14:45 <_av500_> because there is a bug in the OTG code Aug 21 16:14:50 <_av500_> on the N7 side Aug 21 16:14:55 i will be making a tutorial for 2morrow with all the thingies needed Aug 21 16:15:01 <_av500_> so it hangs with the USB code in the boot ROM Aug 21 16:15:05 same here vvu, i have a Galaxy S III, if you provide instructions i can give it a try Aug 21 16:15:06 <_av500_> which we cannot change Aug 21 16:15:21 <_av500_> its fixed in 3.4 Aug 21 16:15:28 <_av500_> which is maybe in latest android only Aug 21 16:15:30 <_av500_> 4.3 Aug 21 16:16:07 vvu: galaxy nexus and nexus 7 here Aug 21 16:16:08 I guess my GTA04 will also run Android. Aug 21 16:16:20 anujdeshpande: tested on both, i have them Aug 21 16:16:22 and it works Aug 21 16:16:23 vvu: is it possible to run Android on a BeagleBone and boot another BeagleBone from it? Aug 21 16:16:34 tcort: technically is possible Aug 21 16:17:09 <_av500_> assuming usb host works for android on BBB Aug 21 16:17:16 +1 Aug 21 16:17:45 +1 Aug 21 16:17:55 thatd be fun Aug 21 16:18:31 tcort: that's a really great idea. not sure if it'd been mentioned before. Aug 21 16:18:50 <_av500_> it can be also a BB B running linux Aug 21 16:18:56 <_av500_> booting another one Aug 21 16:19:02 <_av500_> heck, even a PC running linux :) Aug 21 16:19:08 _av500_: how so, don't you need Android? Aug 21 16:19:16 isn't the app an Android app? Aug 21 16:19:19 <_av500_> yes Aug 21 16:19:25 <_av500_> the other is a script and some c code :) Aug 21 16:19:32 <_av500_> to be written :) Aug 21 16:19:54 oh, so that is in scope for this project? Aug 21 16:20:07 <_av500_> no Aug 21 16:20:15 <_av500_> its just "possible" in the widest sense Aug 21 16:20:31 ok so what will be the next step for the project? Aug 21 16:20:37 https://code.google.com/p/rowboat/ rowboat? Aug 21 16:20:40 k, but I think it is reasonable to document the steps for testing this Android app with a BeagleBone Black running Android, no? Aug 21 16:20:42 <_av500_> vvu: make a full chain Aug 21 16:20:46 I don't know how active that still is Aug 21 16:20:47 <_av500_> jkridner|work: yes of course Aug 21 16:20:52 <_av500_> keesj_: thats the one Aug 21 16:20:59 full chain meaning ? Aug 21 16:21:04 <_av500_> well, everything Aug 21 16:21:08 <_av500_> image on a server Aug 21 16:21:10 <_av500_> app that picks it Aug 21 16:21:12 <_av500_> downloads Aug 21 16:21:14 <_av500_> flashes Aug 21 16:21:19 <_av500_> docs Aug 21 16:21:26 <_av500_> and 3d animations Aug 21 16:21:28 basically 1st part i have it Aug 21 16:21:30 now docs Aug 21 16:22:23 the easiest ways I've found to run Android on Black are: http://circuitco.com/support/index.php?title=Android and http://www.ti.com/tool/androidsdk-sitara Aug 21 16:22:32 * keesj_ thinks of xbill Aug 21 16:24:16 "like Bones infecting bones". but on a more serious note it would be good to get some more results out of the project Aug 21 16:24:26 vmayoral: thanks for all the ROS updates and the latest status report. do you have a rough idea which of those modules you plan to get running by the end of the project? Aug 21 16:25:35 keesj_: if a Bone can be used to boot another Bone, I think that's pretty useful for automating testing. good result? Aug 21 16:25:51 yes Aug 21 16:26:23 ok so i need to get a bone running android and test the thingie ? Aug 21 16:27:06 jkridner|work: do you mean the ros_control ROS package? Aug 21 16:27:12 at least not focusing on the N7 and usb-host (as we did this in the first place to simplify things) Aug 21 16:28:01 it's av500 's call (perhaps it is better to brainstorm offline) Aug 21 16:28:18 I mean stuff not tested at https://github.com/vmayoral/beagle-ros/blob/master/recipes/README.md Aug 21 16:28:53 on my side until 2morrow i will have the tutorial and the chain of doing stuff Aug 21 16:29:23 thanks vvu Aug 21 16:29:42 <_av500_> vvu: great Aug 21 16:29:54 <_av500_> lets continue after that Aug 21 16:29:54 i will let aside a bit making sdcards for rpi Aug 21 16:30:04 if that is ok for you jkridner|work ? Aug 21 16:30:13 jkridner|work: Everything marked as untested is out of the initial scope but i can put some time on them and try to make everything compile and test it. Aug 21 16:30:36 vvu, _av500_: works nicely for me. much easier to test with BBB than a phone Aug 21 16:31:40 vmayoral: can you summarise in a mail the errors you are getting in cross-compiling some of the recipes you listed there ? Aug 21 16:31:40 <_av500_> ok Aug 21 16:32:19 jkridner|work: i'll first take care of the things planned (not much left). Do you want me to remove from my reports the packages that are out of the scope of the GSOC? Aug 21 16:32:49 no, just indicate which ones are. Aug 21 16:32:49 khem: sure Aug 21 16:33:25 all right Aug 21 16:34:01 * jkridner|work is just looking for acknowledgement from vmayoral, vvu, hatguy, anujdeshpande, jj2baile, tcort and ZubairLK that what is needed before the project wraps up is well understood. Aug 21 16:34:41 jkridner|work: yepp have a pretty good understanding about that Aug 21 16:34:43 yepp. we have a realistic aim of what all things have to be completed by sept 20 Aug 21 16:34:54 jkridner|work: driver works. getting IIO list to accept it. they've sent modifications to structure of code. should be done in a week or so Aug 21 16:35:14 jkridner|work: if android on BBB behaves nicely in like 1 week we can have everything up and running Aug 21 16:35:18 vmayoral: copying beagleboard-gsoc@googlegroups on that message to khem might be nice/helpful... traffic there has been pretty low. Aug 21 16:35:42 ZubairLK: just read the oneshot feedback. Aug 21 16:35:46 jkridner|work: i think it'll be fine here as well Aug 21 16:35:47 jkridner|work: yes, I know what I need to do. BTW, I have the weathercape in hand now. Aug 21 16:36:02 tcort: great. Aug 21 16:36:16 prpplague: will you be able to ship the new bacon by next week? Aug 21 16:36:25 android should just work on the BBB Aug 21 16:36:34 get the rowboat stuff Aug 21 16:36:39 ds2: the usb host things Aug 21 16:36:45 hatguy: hopefully Aug 21 16:36:53 prpplague: okay Aug 21 16:37:02 vvu: there is a hard wired host on the BBB Aug 21 16:37:05 no OTG Aug 21 16:37:18 the new bacon cape would be a pretty good minux target at some point too, thanks to the accelerometer. Aug 21 16:37:21 ok, now the RNDIS problem Aug 21 16:37:30 also good for ROS. Aug 21 16:37:33 if that is avoided things are great Aug 21 16:38:27 yepp and I have a feeling any good ROS targets would be good Userspace targets Aug 21 16:39:53 vvu: what is the RNDIS problem? Aug 21 16:40:27 in this case you'd control the kernel Aug 21 16:40:34 in the android kernel i was needed to enable HOST RNDIS DRIVER config, to set up for me the RNDIS connection. if this was disable i couldn`t receive data from the am335x even to set up the connection Aug 21 16:41:39 ds2: any idea if i have root access with rowboat? Aug 21 16:41:52 <_av500_> vvu: should be doable Aug 21 16:42:11 <_av500_> jkridner|work: feel free to continue the meeting Aug 21 16:42:17 <_av500_> we can take RNDIS for later Aug 21 16:42:19 vvu: you'd have root access with each of the links I sent. both are built using rowboat. Aug 21 16:42:27 vvu: yes. you have full access Aug 21 16:42:38 _av500_: or to wrap it up? Aug 21 16:42:52 perfect Aug 21 16:42:56 * ds2 looks for pru jtaggers Aug 21 16:43:26 i need some dvi adapter or the uHDMI is working? Aug 21 16:43:55 ds2: Good day. Aug 21 16:44:06 uHDMI works with the http://circuitco.com/support/index.php?title=Android image. I've tried that one. Aug 21 16:44:13 jj2baile: hey... long time :D Aug 21 16:44:21 ds2: yeah....= Aug 21 16:44:24 vvu: depending on what you want as a display Aug 21 16:44:31 ds2: Getting back up to speed Aug 21 16:44:41 i have a Sony Bravia LCD Aug 21 16:44:44 720p display Aug 21 16:45:00 jj2baile: you changed blogs? Aug 21 16:45:00 k, guess there isn't any other new business, so I'll make sure folks know they can get on their way... Aug 21 16:45:07 thanks all... see you next week. Aug 21 16:45:17 vvu: then yes, you need the adapter/cable you spoke of Aug 21 16:45:21 * jkridner|work concludes official portion of meeting. Aug 21 16:45:31 av500: i think is your call for what to do next. until 2morrow i will provide things on how to work this project with a phone Aug 21 16:45:43 then switching to test it with a BBB*android* is ok? Aug 21 16:46:05 ds2: my tv has hdmi in, do i really need it? don`t really see the connection Aug 21 16:46:24 thanks all Aug 21 16:46:38 vvu: you still need a microHDMI to HDMI cable Aug 21 16:46:44 i have that :D Aug 21 16:47:05 vvu: the microHDMI connector is on the bottom,next to the uSD slot, IIRC. looks like USB but it is labeled as HDMI Aug 21 16:47:13 yep yep Aug 21 16:51:29 av500: i`m offline for now, will be back 2night working. Aug 21 16:51:33 have a good day! Aug 21 16:53:13 <_av500_> ok Aug 21 18:57:22 gm, just discovered I am about 3 hrs late :-( sorry about that, didn't miss too much fun I hope ;-) Aug 21 23:19:38 ka6sox, jj2baile - ping Aug 21 23:20:04 hiya mluckham Aug 21 23:20:17 hi Tom ... i'm baaaaack Aug 21 23:20:31 eggcellent Aug 21 23:21:02 enjoyed playing with grandson #1 "up north", and some river kayaking Aug 21 23:21:13 so I'm currently setup the same way y'all are for building Aug 21 23:21:26 there are new fresh schematics on the github Aug 21 23:21:33 I'll take a look Aug 21 23:21:42 and I'm looking for you and I to create a repo for the software you are working with. Aug 21 23:21:52 Hello. Aug 21 23:21:55 why don't we do that now Aug 21 23:21:57 jj2baile - hi Aug 21 23:21:58 hiya jj2baile Aug 21 23:22:15 jj2baile, did you see mluckham's vring demo code? Aug 21 23:22:29 I don't think i've seen that yet Aug 21 23:22:52 that's what we were going to put in the repo, initially. Aug 21 23:23:08 also need to accommodate our version of OpenOCD, and panto's kernel Aug 21 23:23:11 Ok, anywhere I can take a look at it now? Or should i just wait till the repo is up Aug 21 23:23:22 jj2baile: is there a new blog? Aug 21 23:23:36 ds2: No, there isn't Aug 21 23:23:47 or is there a suggestion on a squeegee to clear up the window for observers? Aug 21 23:24:12 ds2, hopefully after today there will be more clarity Aug 21 23:24:47 mluckham, I"m going to call it bonetag_software Aug 21 23:25:06 sounds good to me - pls send me the link Aug 21 23:25:09 give me your github ID so I can add you as a contributor. Aug 21 23:25:45 mluckham54 Aug 21 23:25:49 k Aug 21 23:26:06 ka6sox: have a sudden need to jtag a bone so... ;) Aug 21 23:26:13 scary! Aug 21 23:28:50 uh.... Aug 21 23:29:04 ds2, what you need then is the one that Russ made Aug 21 23:29:11 thats for programming a Bone Aug 21 23:29:40 mluckham, jj2baile both of you are collaborators Aug 21 23:29:53 mluckham, can you populate with what you have? Aug 21 23:29:54 ka6sox: made? u mean the cape? Aug 21 23:29:59 or? Aug 21 23:30:10 he did a cape for what you need Aug 21 23:30:22 jj2baile, after he pushes can you meld yours in too? Aug 21 23:30:32 ka6sox: ok Aug 21 23:30:42 ya, it gives you everything on the black you had on the white, plus a little more Aug 21 23:31:08 I think I can rig up something Aug 21 23:31:25 AFAIK, it just needs a connector and to a jtag pod, probally a FTDI Aug 21 23:31:42 but it would be nice to use the PRU JTAG as a comparism Aug 21 23:32:01 ds2, its not ready yet... :) Aug 21 23:32:20 ka6sox: that's where testing comes in play Aug 21 23:32:39 ka6sox: since it is 3.3V to 3.3V, something crude should be doable Aug 21 23:33:04 Russ: have you done suspend AND resume on the AM33x? Aug 21 23:33:32 tom, can you send me the url? Aug 21 23:33:53 or maybe i just search for bonetag_software on github? Aug 21 23:33:54 ds2, yes Aug 21 23:34:16 Russ: do you know by default if all the wake sources are enabled? Aug 21 23:34:29 getting it to suspend w/o problem but none of hte wake sources seems to do anything Aug 21 23:35:06 I haven't looked into how wake sources get enabled or disabled, but I think its IP specific Aug 21 23:35:17 Russ: what did you use for a wake source? Aug 21 23:35:20 uart0 Aug 21 23:35:27 uart0 is the console, right? Aug 21 23:35:33 or is that the pri_uart0 pins? Aug 21 23:35:39 found it - now have to figure out how to upload ... Aug 21 23:36:01 * vvu just woke up...time to work now Aug 21 23:36:07 yes, its the console Aug 21 23:36:12 first do a pull Aug 21 23:36:12 then move your stuff in Aug 21 23:36:12 then git add * Aug 21 23:36:13 git commit Aug 21 23:36:18 then a git push Aug 21 23:36:28 tx Aug 21 23:36:39 actually clone is the right answer Aug 21 23:36:40 sorry Aug 21 23:36:41 d0h Aug 21 23:36:45 ah...so that confirms there is something F'ed then Aug 21 23:36:48 but use your key Aug 21 23:36:52 Russ: BBB or BBW? Aug 21 23:36:57 or PG1.x or PG2.x Aug 21 23:36:59 git@github: Aug 21 23:37:15 bbw Aug 21 23:37:25 but I've also tested on am335x-evm with pg2.0 Aug 21 23:38:21 jj2baile, mluckham i think we want to create directories by username till we can create a merged setup Aug 21 23:38:26 I'll add panto now too. Aug 21 23:38:41 the am335x pm firmware is in a bit of flux right now, so you need to make sure you have the right version Aug 21 23:38:58 I see Aug 21 23:39:11 you recommend the devkit for the pm firmware file or the rowboat repo one? Aug 21 23:39:17 or some other source? Aug 21 23:40:02 grab this one if you are using dave gerlach's patchset http://arago-project.org/git/projects/?p=am33x-cm3.git;a=tree;f=bin;h=0e5fe9907aa49c82ff9896eff38c9428ce060f42;hb=refs/heads/next2 Aug 21 23:40:26 patchset? Hmmmm the devkit sources claim it works so I been using them as a baseline Aug 21 23:40:45 3.2? Aug 21 23:41:04 yeah Aug 21 23:41:17 the devkit notes claims it supports suspend and resume Aug 21 23:41:33 so I tried console, gpio0, and USB Aug 21 23:41:53 then use this firmware http://arago-project.org/git/projects/?p=am33x-cm3.git;a=tree;f=bin;h=6361b0c1f42f49438513ee7851d7e005a16dd98d;hb=750362868d914702086187096ec2c67b68eac101 Aug 21 23:42:23 ka6sox: Got it. Makes things simpler for now Aug 21 23:42:23 is the JTAG of the M3 reachable? Aug 21 23:42:34 and if that works, if you could test with this firmware http://arago-project.org/git/projects/?p=am33x-cm3.git;a=tree;f=bin;h=ed041997fbc3630b217e36e1303ed9f4d894da46;hb=next2-cleanup I'd appreciate it Aug 21 23:42:46 yes, but I've never had any luck getting anything productive done with it Aug 21 23:42:50 ds2, iirc all of them hare on the bus Aug 21 23:43:22 oh... so I would not be able to break it on exceptions like hitting the wfi instruction? :/ Aug 21 23:43:31 ok... I'll try those Aug 21 23:43:49 I'm not really sure, its probably just that I suck at code composer Aug 21 23:44:00 oh. i was going to openocd it Aug 21 23:44:18 you'd have to add support, if you do, I'd appreciate seeing patches Aug 21 23:44:27 figure....try to break on wfi on the A8 side. if that works, trap it on the M3 side Aug 21 23:44:41 it does have m3 support, just not whatever options need to be plugged into the bone board file and am335x target file Aug 21 23:44:50 maybe I wasn't clear - does the connection to the M3 require another magic key? Aug 21 23:45:04 the m3 is just another thing on the jtag chain if that's what you mean Aug 21 23:45:10 wfi is a stock M3 instruction Aug 21 23:45:20 not looking at debugging anything else Aug 21 23:45:25 wfi is a stock instruction on m3 and a8 Aug 21 23:45:36 but it does something funky, so it may not work well in the debugger Aug 21 23:45:39 all I want to know is where am I dying during resume Aug 21 23:45:45 Ohhh gotcha Aug 21 23:45:58 I ended up just debugging by gpio Aug 21 23:46:09 this is all assuming JTAG isn't killed during "suspend" Aug 21 23:46:21 hard to gpio debug the M3 w/o docs....sucks to be on the outside Aug 21 23:46:55 I don't even have docs on a how a lot of that stuff works Aug 21 23:47:05 but don't you have M3 source? Aug 21 23:47:11 at some point I need to setup clearcase on my corporate machine so I can get the RTL Aug 21 23:47:18 M3 firmware source, that's on arago-project Aug 21 23:47:21 ohhh Aug 21 23:47:24 it is? Aug 21 23:47:26 http://arago-project.org/git/projects/?p=am33x-cm3.git;a=summary Aug 21 23:47:40 mluckham, I added a cheap readme that I'll mod when you and jj2baile are done. Aug 21 23:47:51 you can build it too, but you need the amsdk compiler Aug 21 23:48:24 cloning Aug 21 23:48:33 is it in C or ASM? Aug 21 23:48:50 c Aug 21 23:48:58 if you want something that might be easier to read, you can try http://arago-project.org/git/projects/?p=am33x-cm3.git;a=shortlog;h=refs/heads/next2-cleanup Aug 21 23:49:21 ah Aug 21 23:49:27 be sure to read http://arago-project.org/git/projects/?p=am33x-cm3.git;a=blob;f=License.txt;h=03b1f482c83a7edbf31af45d68ffbbe23aabdb57;hb=HEAD Aug 21 23:49:38 is there a gotcha in the license? Aug 21 23:50:14 doesn't seem like it, but it isn't a standard one I recognize Aug 21 23:50:43 Oh Aug 21 23:51:20 I'll assume it is similar to the license for simpliciti Aug 21 23:51:50 oh my...this sure looks like the similar code on the O3 running on the A8 side Aug 21 23:52:02 03? Aug 21 23:52:06 or O0? Aug 21 23:52:12 ack, O3... Aug 21 23:52:52 i get an error when trying to push Aug 21 23:53:12 the dm37x :) Aug 21 23:53:20 "refs not found: did you run git update-server-info on the server?" Aug 21 23:53:53 thanks..this source is helpful even if I cannot build it Aug 21 23:53:53 mluckham, did you clone first? Aug 21 23:54:08 traced it on the A8 side and this answers what is going on at the other end Aug 21 23:54:15 first I did 'git clone git://github.com/ka6soc/BoneTag_Software' Aug 21 23:54:23 okay thats the issue Aug 21 23:54:30 git@github Aug 21 23:54:35 you are in RO mode Aug 21 23:54:51 git clone git@github/ka6sox/... ? Aug 21 23:55:02 corret Aug 21 23:55:06 ok Aug 21 23:55:14 I think you might need a colon instead of the first slash Aug 21 23:55:52 ds2, it builds with "arm-arago-linux-gnueabi-gcc (GCC) 4.5.3 20110311 (prerelease)" Aug 21 23:55:53 git@github.com:ka6sox/BoneTag_Software.git Aug 21 23:56:03 so it may build with any other 4.5 gcc Aug 21 23:56:28 "arm-linux-gnueabi-gcc (Ubuntu/Linaro 4.7.3-1ubuntu1) 4.7.3" does not build a working version Aug 21 23:56:37 Russ: does it need a new gcc? Aug 21 23:56:53 old (4.5) Aug 21 23:56:59 I don't know why Aug 21 23:57:06 hmmm Aug 21 23:57:16 the built binary is checked into the repo though Aug 21 23:57:18 I have an older toolchain for the STM32s Aug 21 23:57:42 you can always try, its probably some section change or new optimization Aug 21 23:57:52 haven't had the bandwidth to debug it Aug 21 23:57:52 the goal is to figure what went wrong...and it sounds like there are enough gotchas that I should not do jtag Aug 21 23:58:05 so I'll have to figure out how to talk to gpio and wiggle stuff with that Aug 21 23:58:12 the first thing is to try a firmware release that has pg2.x support Aug 21 23:58:38 afaik, there is nothing special needed in userland on the A8, right? Aug 21 23:58:46 http://arago-project.org/git/projects/?p=am33x-cm3.git;a=commitdiff;h=30d0fc60191b343b9e36406fd2cc3fe744c9d057 Aug 21 23:58:56 no, I don't think so Aug 21 23:58:58 i.e. boot kernel with init=/bin/sh; mount sysfs; echo mem > /sys/power/state should work Aug 21 23:59:11 right, but be sure that the firmware loads Aug 21 23:59:31 yes, firmware is linked into kernel Aug 21 23:59:41 not distributing it so there should be no GPL issues yet Aug 21 23:59:41 ds2, you are trying to mod the PM code in the M3? Aug 21 23:59:55 ka6sox: no, I am trying to find out why I cannot resume Aug 22 00:00:04 kk Aug 22 00:00:04 but that might make me need to mod the PM code on the M3 to debug that Aug 22 00:00:12 ok Aug 22 00:00:20 brb Aug 22 00:00:33 permission denied (public key) - the remote end hung up unexpectedly Aug 22 00:00:38 suspend isn't too useful w/o resume working :( Aug 22 00:00:53 alternative is to try to get Joel's snapshot boot working Aug 22 00:02:46 mluckham, I think I need a deploy key from you...dunno why... Aug 22 00:02:55 do you have your key uploaded? Aug 22 00:03:08 (and are you on the one with that key?) Aug 22 00:05:12 ds2, is this part of your GSoC project? Aug 22 00:05:19 no Aug 22 00:05:29 it was an attempt to use the GSoC project Aug 22 00:07:22 :) Aug 22 00:09:59 are you sure it's actually suspending? Aug 22 00:14:38 it claims it is Aug 22 00:14:44 [ 96.960083] PM: late suspend of devices complete after 4.913 msecs Aug 22 00:14:49 then it is dead Aug 22 00:16:45 uploaded okay now Aug 22 00:17:10 the SSH key on the github account from previously, was generated for my Windows computer Aug 22 00:18:06 only sertest uploaded, openocd-code folder is empty - any ideas? Aug 22 00:18:49 jj2baile, sertest was compiled directly on the BBB Aug 22 00:19:16 ok Aug 22 00:19:47 ds2, right, I'm curious how far the pm firmware is getting Aug 22 00:20:06 but since it's on a bone, its not like you can measure vdd mpu current Aug 22 00:20:54 openocd-code is intended to be a clone of openocd + my changes (haven't done any yet) Aug 22 00:21:29 Russ: *nod* Aug 22 00:22:10 but since you pointed me to the sources, Ican poke around and infer things. Used to not have that many blackboxes to the PRCM Aug 22 00:22:43 i wasn't able to build openocd on bbb (lacking libutil) so I have only built it on Ubuntu so far Aug 22 00:23:11 i intended to get openocd running first on ubuntu, preferrably with jtag target the Stellaris Launchpad ICD Aug 22 00:30:33 why would you build openocd on bbb? Aug 22 00:30:53 maybe I'm missing context Aug 22 00:31:34 gdbserver mode? Aug 22 00:31:43 for pru/jtag project Aug 22 00:32:01 ultimately it will be cross-compiled and run ON BBB Aug 22 00:32:23 ah, sounds like a cool project Aug 22 00:32:41