**** BEGIN LOGGING AT Thu Jul 02 02:59:59 2015
Jul 02 05:01:32 * nerdboy thinking of fun project ideas for next year...
Jul 02 05:49:44 <nerdboy> got a folder now even  https://bpaste.net/show/a7dd18162762
Jul 02 05:50:37 <nerdboy> good practical stuff
Jul 02 07:07:39 <av500> nerdboy: heh
Jul 02 07:10:03 <ebadawy> av500: I'm sorry for not respodning during the meeting, how will it go for the next week?
Jul 02 07:13:32 <av500> we have two for next week, so youre up the one after
Jul 02 07:14:44 <ebadawy> av500: OK, thanks.
Jul 02 09:05:22 <shubhangi> Abhishek_: pushed the binaries
Jul 02 09:18:06 <av500> gently I hope
Jul 02 11:25:55 <Abhishek_> shubhangi: I meant not to push them into GitHub, just maybe upload them on a drive folder.
Jul 02 11:26:11 <Abhishek_> shubhangi: Do you recognize the error message I posted yesterday?
Jul 02 12:53:45 <jkridner> hi rohith_
Jul 02 12:56:29 <vmayoral> rohith_: ping
Jul 02 13:00:51 <rohith_> jkridner vmayoral: pong
Jul 02 13:01:34 <jkridner> rohith_: seems you've been able to get in some last-minute contributions. that's good. hope you and other mentors can interact here on #beagle-gsoc more...
Jul 02 13:01:51 <jkridner> rohith_: even if your direct mentors aren't around, it is good to just come in here and chat in off hours....
Jul 02 13:02:02 <rohith_> jkridner: Will do
Jul 02 13:02:14 <jkridner> the other students can often help in blocking issues and many of the mentors are also around to help out.
Jul 02 13:02:41 <jkridner> rohith_: if nothing else, getting a bit of moral support (motivation) can be a good thing to keep up the day-to-day...
Jul 02 13:03:02 <jkridner> I know how hard it can be to get started working when on your own and not feeling the pressure or motivation...
Jul 02 13:03:17 <jkridner> often, just asking some silly questions can be rather useful to get your mind back on the right track.
Jul 02 13:03:51 <jkridner> sharing a commit here or there, even with people not working on your project, can add a lot of clarity, even if they don't have great/directed suggestions.
Jul 02 13:04:15 <rohith_> jkridner: I understand. Will be more regular
Jul 02 13:04:22 <jkridner> for me, it matters a *lot* that I see the students and mentors hanging out in here.
Jul 02 13:04:40 <jkridner> hi pdp7_...  have you gotten familiar with any of the projects?
Jul 02 13:05:03 <jkridner> panto-cloud: tried reaching you repeatedly about your board.... I've told Fedex to go ahead and destroy it.
Jul 02 13:05:22 <jkridner> pdp7_: did you figure when you are coming in for Maker Faire?
Jul 02 13:05:42 <jkridner> Still Friday night or perhaps earlier?
Jul 02 13:06:04 <rohith_> vmayoral: Just saw the email. Thanks a ton for another chance. Appreciate it and will follow a schedule as suggested by you and get on track
Jul 02 13:06:05 * jkridner would love to throw a small get-together at the house if there were people in town.
Jul 02 13:19:14 <Abhishek_> shubhangi: ping
Jul 02 13:40:30 <jkridner> ebadawy, arianepaola: how's things going?
Jul 02 13:41:14 <ebadawy> jkridner: currently I'm working the preview card.
Jul 02 13:41:45 <ebadawy> jkridner: I've made a PR on what I have done in 'create' page.
Jul 02 13:44:32 <Abhishek_> jkridner: Any news on the BeagleBone Green?
Jul 02 13:44:39 <jkridner> no
Jul 02 13:44:59 <Abhishek_> is it already released for sale yet?
Jul 02 13:55:56 <jkridner> ebadawy: I merged the create pull-request....
Jul 02 13:56:25 <jkridner> ebadawy: I'd really like to see you remove the title change and put the description and any other fields into a .JSON file.
Jul 02 13:57:20 <ebadawy> jkrider: sounds more reasonable, I'll do it.
Jul 02 13:57:28 <jkridner> I also question if "New Tutorial" should be there for all the pages.
Jul 02 13:57:31 <ebadawy> jkridner* :)
Jul 02 13:58:11 <jkridner> ebadawy: seems we should create a menu that matches http://beagleboard.org and add the new tutorial entry in there somewhere reasonably.
Jul 02 13:58:46 <jkridner> ebadawy: also, I'm not sure if the LOGIN works at http://beagleboard.github.io/bone101, but it seems not to.
Jul 02 13:59:28 <ebadawy> jkridner: you mean it just showes in the home page?
Jul 02 13:59:47 <ebadawy> jkridner: I'll check it ^^
Jul 02 14:09:01 <neemo> m
Jul 02 14:09:03 <neemo> 12:53  * nerdboy thinking of fun project ideas for next year...
Jul 02 14:09:33 <ebadawy> jkridner: that's weird, it throughs an error for not able to read access_token! https://github.com/beagleboard/bone101/blob/master/static/github-login.js#L14
Jul 02 14:11:51 * jkridner has to drop off.... 8% on battery
Jul 02 14:12:12 <jkridner> ebadawy: might be due to the return address for the token you used.
Jul 02 14:13:59 <ebadawy> jkridner: found it, I didn't add beagleboard domain in oauth whitelist!
Jul 02 14:15:26 <neemo> some funky actions this middle click has in irssi
Jul 02 14:16:31 <neemo> anyway, don't mind my random interruptions
Jul 02 14:16:46 <neemo> also, nice list nerdboy :)
Jul 02 14:40:29 <arianepaola> jkridner: good, catching up though
Jul 02 14:55:46 <shubhangi> Abhishek_: ping
Jul 02 14:56:44 <alexanderhiam> shubhangi: have you started working on 4.1 support yet?
Jul 02 14:58:25 <shubhangi> alexanderhiam: Abhishek_ is trying to test it on 4.1 these days. will probably get it running in a day or two.
Jul 02 14:58:41 <alexanderhiam> awesome
Jul 02 14:59:15 <shubhangi> alexanderhiam: will ping you as soon as it gets functional on 4.1
Jul 02 15:00:31 <Abhishek_> shubhangi: pong
Jul 02 15:00:50 <Abhishek_> does the error message look familiar to you?
Jul 02 15:01:20 <Abhishek_> [   52.687522] pruss-rproc 4a300000.pruss: platform data (reset configuration information) missing
Jul 02 15:01:29 <shubhangi> Abhishek_: i am working on it right now. will give you the solution by tonight
Jul 02 15:01:51 <Abhishek_> is it on the 3.14 kernels as well?
Jul 02 15:02:04 <shubhangi> Abhishek_: never came up for me
Jul 02 15:02:12 <Abhishek_> I see
Jul 02 15:02:33 * Abhishek_ is going afk for dinner
Jul 02 16:14:42 <shubhangi> Abhishek_: the platform data is missing for you. How are you instantiating your driver. is the driver getting probed when the overlay is applied or do you insert the module manually
Jul 02 16:40:04 <Abhishek_> shubhangi: Gets probed when the overlay is applied
Jul 02 16:40:14 * Abhishek_ is back
Jul 02 16:58:41 <shubhangi> Abhishek_: okay .. let me figure it out
Jul 02 17:00:05 <shubhangi> Abhishek_: till what time are you awake ? we often seem to miss each other's pings
Jul 02 17:09:18 <Abhishek_> shubhangi: today, till 2pm or so
Jul 02 17:09:29 <Abhishek_> *2am
Jul 02 17:09:33 <shubhangi> ok
Jul 02 17:09:54 <Abhishek_> I'm travelling this weekend
Jul 02 17:10:08 <Abhishek_> I'll post the patches so that you could try it out on your side
Jul 02 17:10:27 <shubhangi> okay
Jul 02 17:10:53 <shubhangi> i'll look into it .. i might set up 4.1 on my BBB tonight
Jul 02 17:21:16 <azizulhakim> vvu: av500: any point or suggestion on lazzy update plan?
Jul 02 17:27:30 <vvu> azizulhakim: 21:09 < _av500_> vvu: he says the number of bytes transfered matches the video pixel data
Jul 02 17:27:37 <vvu> 21:09 < _av500_> so that part looks good to me
Jul 02 17:27:37 <vvu> 21:09 < _av500_> I still want the raw transfer speed numbers
Jul 02 17:27:38 <vvu> 21:09 < _av500_> so FB involved, just pump data as fast as you can over ADK
Jul 02 17:27:43 <vvu> from the logs
Jul 02 17:30:24 <azizulhakim> I remember that, but that still won't solve the problem. I checked the code and I don't see any possible delay between new data-transfer start
Jul 02 17:30:51 <azizulhakim> as soon as there is new page, we pass it using urb
Jul 02 17:32:25 <vvu> av500: _av500_ ^
Jul 02 17:36:31 <azizulhakim> vvu: av500: In 3.69MB/s transfer speed, it'll take 23 seconds to pass a 7 second video, which is too much
Jul 02 17:39:47 <azizulhakim> may be I can do some work on lazy update and see the result and then we can decide?
Jul 02 17:45:22 <shubhangi> Abhishek_: ping
Jul 02 17:45:43 <shubhangi> i think i found the cause
Jul 02 17:46:22 <shubhangi> 4.1 doesnt has this https://github.com/beagleboard/linux/blob/3.14/drivers/remoteproc/pruss_remoteproc.c#L33
Jul 02 17:46:43 <shubhangi> take it from here : https://github.com/beagleboard/linux/blob/3.14/include/linux/platform_data/remoteproc-pruss.h
Jul 02 17:47:12 <shubhangi> have*
Jul 02 17:51:48 <vvu> azizulhakim: yep
Jul 02 17:51:56 <vvu> keep us updated via github/irc
Jul 02 17:52:13 <vvu> i am following the beagle_usb repo on github so i get emails when something new is in there
Jul 02 17:52:44 <azizulhakim> sure
Jul 02 18:04:22 <Abhishek_> shubhangi: That's already there, won't have compiled otherwise
Jul 02 18:09:01 <Abhishek_> This may have something to do with the device tree
Jul 02 18:09:33 <nerdboy> "platform data (reset configuration information) missing" sounds like maybe a platform support kernel option was missed?
Jul 02 18:10:41 <Abhishek_> nerdboy: here: https://github.com/beagleboard/linux/blob/3.14/drivers/remoteproc/pruss_remoteproc.c#L1030
Jul 02 18:10:46 <nerdboy> not ported/not enabled or maybe stale code was not updated fully for new DT stuff?
Jul 02 18:11:11 <Abhishek_> Hmm, I think so\
Jul 02 18:11:27 <nerdboy> something missed porting to 4.1
Jul 02 18:16:10 <Abhishek_> Commenting out that part gives a kernel panic
Jul 02 18:16:32 <Abhishek_> Cannot SSH into the board, so examining logs (or might have to use a serial cable)
Jul 02 18:18:47 <Abhishek_> shubhangi: Post me a dmesg log of the pruss module loading.
Jul 02 18:18:54 <Abhishek_> (Gist it)
Jul 02 18:19:08 <Abhishek_> Nope, its actually an oops
Jul 02 18:19:16 <shubhangi> ok
Jul 02 18:20:01 <Abhishek_> https://gist.githubusercontent.com/abhishek-kakkar/c11105a249c3ef6148cd/raw/47cc3ecf5432bdded5064ad942e28ae8e6f47d75/gistfile1.txt
Jul 02 18:20:39 <shubhangi> https://gist.github.com/shubhi1407/79cb69f7418ca4333518
Jul 02 18:50:20 <vvu> azizulhakim: if we overcome this issue what would be left to do ?
Jul 02 18:50:41 <vvu> i think some cleanup of the android app. for example if you try to connect without any usb cable plugged in it just crashes :)
Jul 02 18:54:00 * nerdboy pictures a popup "did you forget to plug your USB cable in?" <okay>  <cancel>
Jul 02 18:55:18 <azizulhakim> vvu: yes, some clean up, some issue fix, clean up on driver side
Jul 02 18:55:23 <azizulhakim> also bandwidth in audio+video together might be another issue
Jul 02 18:55:49 <azizulhakim> there are some complex issue which I couldn't fix for a long time
Jul 02 19:04:01 <vvu> why not post on the mailinglist
Jul 02 19:04:05 <vvu> and ping people
Jul 02 19:04:08 <vvu> we are here to help :)
Jul 02 19:10:57 <azizulhakim> communicating with alsa developer forum for several weeks. its during audio play in bone. if you install a driver and run a audio file, it works fine. but if you run the audio for a second time, the driver crash
Jul 02 19:11:09 <azizulhakim> happens only in bone, not pc
Jul 02 19:15:29 <vvu> can you paste a link ?
Jul 02 19:15:39 <vvu> the same audio ?  or any audio at all
Jul 02 19:16:48 <azizulhakim> any audio
Jul 02 19:34:29 <nerdboy> which alsa/audio drivers?
Jul 02 19:35:19 * nerdboy needs to build a new kernel anyway...
Jul 02 19:37:05 <azizulhakim> nerdboy: we are developing a new driver. do you mean alsa version?
Jul 02 19:37:53 <nerdboy> whatever is relevant
Jul 02 19:38:46 <nerdboy> 2nd time crash sounds like bad state
Jul 02 19:39:22 <nerdboy> do you close/release everything like you should?
Jul 02 19:39:40 <azizulhakim> I'm not sure which version. I'll let you know the version in bone when I go home.
Jul 02 19:40:05 <azizulhakim> yes I do. the same driver works on ubuntu pc without any problem
Jul 02 19:40:14 <azizulhakim> no matter how many times I play
Jul 02 19:40:55 <nerdboy> same alsa headers?
Jul 02 19:41:28 <nerdboy> is ubuntu machine 32 or 64-bit?
Jul 02 19:41:53 * nerdboy plays dumb guy like a pro
Jul 02 19:42:07 <azizulhakim> by header do you mean snd_pcm_hardware ?
Jul 02 19:42:16 <azizulhakim> tested on both 32 and 64 bit
Jul 02 19:42:37 <nerdboy> if that's your interface, yeah
Jul 02 19:42:40 <azizulhakim> tested in arch linux, debian vm using virtual box,
Jul 02 19:42:55 <azizulhakim> https://github.com/azizulhakim/beagleusb/blob/master/beagle-audio.c#L34
Jul 02 19:42:55 <nerdboy> other things like toolchain versions?
Jul 02 19:43:22 <nerdboy> recent gcc/glibc issues with < gcc 4.8.4
Jul 02 19:44:02 <nerdboy> also fpu format stuff
Jul 02 19:45:02 <Abhishek_> panto-cloud: ping
Jul 02 19:45:10 <nerdboy> things like the "__fp16" type is not defined unless you set it
Jul 02 19:45:27 * nerdboy recommends -mfp16-format=ieee
Jul 02 19:46:23 <nerdboy> neon is not ieee math compliant so you can lose precision if you're not careful
Jul 02 19:47:15 <nerdboy> azizulhakim: what cflags are you using?
Jul 02 19:47:46 <nerdboy> i wouldn't go above -O2 when debugging
Jul 02 19:53:09 <azizulhakim> nerdboy: don't have any idea on clfags stuffs :(
Jul 02 19:56:19 <Abhishek_> shubhangi: ping
Jul 02 19:56:21 <azizulhakim> how cflags might be related here. I'm not compiling alsa. I'm using the default installation that comes with linux distro
Jul 02 19:58:01 <nerdboy> you're compiling something...
Jul 02 19:58:41 <nerdboy> if you use the fpu (and you should) then you need to be a little more careful on arm
Jul 02 20:00:08 <nerdboy> math ops/guarantees of "correctness" are not the same...
Jul 02 20:03:01 <azizulhakim> ok, i'm not aware of this points. do you have any useful link or may be some keywords I should search on?
Jul 02 20:03:43 <azizulhakim> also why should we use fpu?
Jul 02 20:10:48 <nerdboy> it's the arm hardfloat thing mostly
Jul 02 20:11:38 <nerdboy> man gcc has good info on fpu options, but it stops short of fully explaining why of things
Jul 02 20:13:17 <nerdboy> i don't compile native on ubuntu arm really, but gentoo starts out with relatively generic hardfloat flags
Jul 02 20:37:58 <nerdboy> default flags for Gentoo amrv7 =>  CFLAGS="-O2 -pipe -march=armv7-a -mfpu=vfpv3-d16 -mfloat-abi=hard"
Jul 02 20:38:54 * nerdboy should not tpye with taco-hand...
Jul 02 20:48:30 <nerdboy> my less-generic flags for bbb => CFLAGS="-O3 -pipe -march=armv7-a -mtune=cortex-a8 -mfpu=neon-fp16 -mfp16-format=ieee -mfloat-abi=hard"
Jul 02 20:50:07 <nerdboy> the point is to know what flags you're using and try not to dork anything up with incorrect/overly-agressive flags
Jul 02 20:51:24 <nerdboy> like you should probably -O3 in favor of -O2 with specific vectorize flags instead
Jul 02 20:51:36 <nerdboy> *probably avoid
Jul 02 20:52:08 * nerdboy in classic "do what i say, not what i do" mode...
Jul 02 20:54:10 <Abhishek_> shubhangi: https://gist.github.com/abhishek-kakkar/c11105a249c3ef6148cd
Jul 02 20:55:07 <nerdboy> azizulhakim: you should use the fpu (safely in the case of neon) when you can because it's much faster than emulated fpu
Jul 02 21:08:08 <shubhangi> Abhishek_: pong. how was the previous error resolved ?
Jul 02 21:08:31 <shubhangi> you used the binaries in the repo ?
Jul 02 21:11:10 <Abhishek_> shubhangi: yep
Jul 02 21:11:41 <shubhangi> nd what about the platform data ?
Jul 02 21:11:52 <Abhishek_> Commented out the if block at https://github.com/beagleboard/linux/blob/3.14/drivers/remoteproc/pruss_remoteproc.c#L1094
Jul 02 21:13:03 <Abhishek_> The error is something related to has_reset which reports yes but if you look at the device tree there's no such thing in the PRU node
Jul 02 21:13:14 <shubhangi> i am updating to 4.1 right now ..
Jul 02 21:13:23 <shubhangi> yes. no such thing in the device tree
Jul 02 21:13:33 <shubhangi> what patches were you talking about?
Jul 02 21:13:46 <shubhangi> to compile driver on 4.1
Jul 02 21:13:52 <Abhishek_> Yep
Jul 02 21:14:00 <shubhangi> ok
Jul 02 21:14:29 <shubhangi> have to try this myself .. to see whats stopping the cores from booting.
Jul 02 21:15:53 <Abhishek_> shubhangi: Try to find where the error message string is (in which src file) and then compare the 4.1 one with the 3.14 one and let me know.
Jul 02 21:16:27 <shubhangi> sure.
**** ENDING LOGGING AT Fri Jul 03 02:59:58 2015