**** BEGIN LOGGING AT Tue May 31 02:59:58 2016
May 31 07:20:47 <henrix_> Hi, anybody here?
May 31 07:53:45 <Wormo> henrix_: hi still there?
May 31 08:16:27 <henrix_> Hi, yeah. Currently I have some problems setting up the TI Linux SDK. I created a new SD Card but it's not booting on the X15. I want to compile and test the opencl examples in the SDK. Is there another way to do this?
May 31 08:18:07 <Wormo> What's the symptom for the new SD card? It's really best to get that working, since otherwise you'll be stuck if anything gets messed up while developing on the emmc so that won't boot right anymore.
May 31 08:18:50 <Wormo> In particular can you do a pastebin of the serial console output when it fails to boot
May 31 08:26:19 <henrix_> Ah ok, so a serial connection is mandatory? the sdk guide says "There are many ways to connect the host development platform and the target board. These connections will vary depending on how you like to develop and what you are trying to do. Here is an example of a common set up with a serial connection for console and ethernet for networking (TFTP, NFS, etc.):" So I thought, it also works with network only
May 31 08:26:55 <Wormo> serial connection is what you want as soon as you have a boot issue
May 31 08:27:20 <Wormo> you can only see what is going on with u-boot and early kernel messages if you have serial connection hooked up
May 31 08:27:29 <henrix_> Ok, thanks a lot!
May 31 08:27:36 <Wormo> welcome
May 31 08:31:26 <henrix_> One more thing. Can i use the serial connection via USB or via expansion header only?
May 31 08:33:13 <Wormo> the header pins have the serial port for u-boot
May 31 08:33:28 <Wormo> I take it you don't have a FTDI serial cable handy?
May 31 08:33:46 <Wormo> one of those usb-to-serial adapters for your dev box
May 31 08:35:46 <henrix_> yes, but I ordered yesterday (https://www.amazon.de/gp/product/B00AFRXKFU). It should arrive today;-)
May 31 08:35:53 <Wormo> ok good!
May 31 08:40:36 <henrix_> The problem is the expansion header. I also can't test our soundcard on X15 yet due to hw interface. I think I will have the same problem with serial adapter. I there a kind of breakout board for X15?
May 31 08:55:23 <Wormo> Maybe? I'm not sure if it's compatible with Beagleboard xM expansions
May 31 08:55:30 <Wormo> http://www.tincantools.com/Trainer-xM-Board.html
May 31 09:10:52 <henrix_> ok, I just talked to my mentor rma. He will create a breakout board on his own. Can you give me the exact UART port for u-boot?
May 31 09:12:37 <henrix_> is it UART1 on P18?
May 31 09:17:39 <Wormo> Um I've not actually used a x15 yet, the old xM had a physical RS232 connector, would need to look up x15 board diagram
May 31 09:19:06 <ZeekHuge> Hey Wormo !
May 31 09:19:23 <Wormo> hi ZeekHuge, sorry to say I'm about to head out :/
May 31 09:19:34 <ZeekHuge> oh .. ahh okay.
May 31 09:19:37 <ZeekHuge> :)
May 31 09:19:48 <Wormo> henrix_: http://dev.iachieved.it/iachievedit/a-look-at-the-beagleboard-x15/
May 31 09:20:30 <Wormo> on beagleboard x15 debug port is a separate little connector like on bbb
May 31 09:21:45 <Wormo> you can fit an ftdi cable or an adafruit PL2303 on there without interferring with other header pins
May 31 09:24:14 <henrix_> ok, thanks. I'll have a look
May 31 09:24:34 <Wormo> ZeekHuge: any quick questions or comments?
May 31 09:25:19 <ZeekHuge> just that .. I have, for now, stopped with the kernel and rproc thing and focusing on other things in the week's goals list.
May 31 09:26:08 <ZeekHuge> like ... compiling the work and putting them up on the repo readmes , docs , travis
May 31 09:26:16 <ZeekHuge> and other things
May 31 09:26:25 <Wormo> good to get that done
May 31 10:06:57 <Wormo> hm my idea of 'soon' is kind of loose sometimes ;)
May 31 10:06:59 <Wormo> gnight
May 31 10:31:06 <kiran4399> alexhiam: should I enable the i2c bus in am33xx.dts if I enable the mpu6050 driver?
May 31 12:31:25 <jic23> gah - odd discovery of the day - the linear diode arrays I'm messing with actually output a higher voltage as they come out of saturation than when they are saturated.  Just spent a good half day chasing vhdl bugs before sticking the scope on the amp input.
May 31 12:31:35 <jic23> This is my new venting location ;)
May 31 12:49:04 <chanakya_vc> hey bradfa,are you there?
May 31 12:49:21 <chanakya_vc> hey mdp,are you there?
May 31 12:55:07 <bradfa> chanakya_vc: hi
May 31 12:55:20 <bradfa> chanakya_vc: I saw you sent an email recently but I haven't had time to read it yet
May 31 12:55:28 <bradfa> chanakya_vc: I will get to it this morning
May 31 12:57:08 <chanakya_vc> bradfa, Just wanted to update you.I installed the latest version on my BBB
May 31 12:57:17 <chanakya_vc> Debian 8.4
May 31 12:57:51 <chanakya_vc> bradfa, I am working on the code tonight. Will you be online today?
May 31 12:58:22 <chanakya_vc> bradfa, If not I can ask you about sigrok tomorrow?
May 31 12:59:38 <bradfa> chanakya_vc: I'll be around today, yes
May 31 13:00:20 <bradfa> chanakya_vc: good to hear you got your bbb flashed
May 31 13:01:00 <bradfa> chanakya_vc: be sure to send your weekly status report today
May 31 13:04:05 <chanakya_vc> bradfa,  Yes I will..I am currently in the lab soldering my ttl converter : )
May 31 13:05:39 <bradfa> chanakya_vc: ok, if it smells like chicken, you're holding the soldering iron wrong ;)
May 31 13:06:16 <bradfa> https://i.redditmedia.com/kKObF46m_lZ9ibg_bpHH1s1cqSp1J-sgh7rTPjNKex4.jpg?w=600&s=331a0c15fa22a90b78bfac6c2c0b32b2
May 31 13:06:45 <chanakya_vc> bradfa, Haha. SMD soldering is a bit difficult. ; )
May 31 13:21:13 <bradfa> chanakya_vc: you'll get the hang of it.  Once you use hot air and solder paste for SMD you'll be spoiled, though :)
May 31 13:30:59 <chanakya_vc> bradfa, Done.Tested it too
May 31 13:31:07 <chanakya_vc> It's working.
May 31 13:39:04 <bradfa> chanakya_vc: what version of ubuntu are you using? 14.04 trusty?
May 31 13:46:43 <chanakya_vc> On my computer?Yes Ubuntu 14.04
May 31 13:47:03 <chanakya_vc> Trusty
May 31 13:53:00 <bradfa> chanakya_vc: ok, thanks
May 31 13:54:19 <bradfa> chanakya_vc: can you pastebin what happens when you run 'sigrok-cli --scan'  does it find your saleae logic clone?
May 31 13:56:43 <chanakya_vc> bradfa, Should I attach the logic analyzer and then do it? Right now I have the ttl attached to my BBB so I guess it is detecting that?
May 31 13:57:30 <bradfa> chanakya_vc: use your saleae on your main PC, not on the BBB
May 31 13:57:41 <bradfa> just plug it in via USB
May 31 13:57:46 <bradfa> and run 'sigrok-cli --scan'
May 31 13:58:09 <bradfa> if your PC doesn't find it, you may need to install the sigrok-firmware package which supports the FX2
May 31 13:58:45 <chanakya_vc> Okay bradfa,I need to go afk right now.I will be back in about an hour an half
May 31 13:58:51 <bradfa> chanakya_vc: ok
May 31 13:59:14 <alexhiam> jkridner: you have a boneblue dts you're working on?
May 31 13:59:29 <chanakya_vc> bradfa,  And do check out the code whenever you find the time.
May 31 14:04:53 <bradfa> chanakya_vc: I'm looking at your makefile
May 31 14:07:15 <jic23> alexhiam, did you and kiran see that note I dropped here last night about the inv_mpu driver?
May 31 14:09:23 <alexhiam> jic23: yeah, trying to find the thread on lkml now...
May 31 14:09:37 <jic23> probably never on lkml. - give me a sec.
May 31 14:10:59 <jic23> alexhiam - http://marc.info/?l=linux-iio&m=146358387211983&w=2
May 31 14:11:45 <jic23> iio is high enough churn of emails that most stuff only ever gets posted on linux-iio unless there is a reason to gather wider review.  LKML is too much of a firehose ;)
May 31 14:11:58 <alexhiam> ah, makes sense
May 31 14:12:49 <jic23> Few outstanding bits in there such as my arguement (discussion) with Leonard about how to handle some of the data munging (annoyingly he was right ;)
May 31 14:12:56 <jic23> + some device tree queries on the bindings.
May 31 14:14:44 <jic23> Clever approach that hangs the slave sensors off an i2c master provided by the device - then provides enough extra info for the mpu to access them itself to fill its ring buffers.
May 31 14:15:39 <jic23> I'll admit I thought no one would ever do a nice job of this particular bit of hardware and we'd have to accept a dirty bodge with separate slave drivers for each possible slave (which is what I think the android driver does?  Earlier drivers certainly did)
May 31 14:16:08 <jic23> But we proposed a while ago how it should be done 'in theory' and Leonard stepped up and did it.  A pleasant surprise!.
May 31 14:16:37 <jic23> Latest ST parts have the same sort of handling of slave i2c buses so some poor person will have to do the same for that driver at some point.
May 31 14:17:30 <alexhiam> sick, this does look much better
May 31 14:17:39 <jic23> Anyhow, I doubt it will change much before merging and as you can see in patch 5 binding example, should do the mpu9050 ak part directly.
May 31 14:18:34 <jic23> So I'd suggest a backport of any inv_mpu changes to whatever kernel Kiran is working with + those patches on top but keep an eye on the thread to get the final version once it's stabilised (probably a couple of weeks).
May 31 14:18:57 <jic23> Anything odd on the blue or is it simply an mpu9050?
May 31 14:19:08 <jic23> Any of those slave i2c buses pinned out for example?
May 31 14:19:45 <alexhiam> it's a 9250. afaik that bus is not pinned out
May 31 14:20:07 <jic23> *laughs* new part every week sometimes (or it feels like it ;)
May 31 14:20:40 <alexhiam> and that's a patch on the inv_mpu6050 driver that's already in bb.org/linux, so that makes life even easier
May 31 14:21:18 <alexhiam> thanks for pointing that out jic23!
May 31 14:22:28 <jic23> You are welcome. Always interesting to see what is going on for these projects :)
May 31 14:22:36 <alexhiam> the android one is ugly for sure, then on top of that there's at least a couple other hacky version floating around out there
May 31 14:24:45 <jic23> yeah. Invensense are better than they were at supporting their own products, but still some way to go.  Right now the Intel guys are interested in it which is 'curious'.  Honestly I've never been entirely sure why they are pumping serious resources into other peoples devices rather than just their sensorhubs...
May 31 14:24:56 <jic23> I guess it's skill development to some degree.
May 31 14:25:30 <alexhiam> oh yeah, didn't notice the @intel address, weird
May 31 14:26:15 <jic23> A lot of mainline sensor drivers these day are provided by intel.  Several different teams that I know of working on it (or at least they are in different countries).
May 31 14:26:51 <jic23> I get the feeling some are doing it on the side for fun, but not most of them.
May 31 14:27:38 <jic23> Also, Daniel Baluta has an outreachY intern every year (sometimes twice a year) which gets us a few new parts supported each time and a lot of ugly corners cleaned up ;)
May 31 14:28:37 <alexhiam> well that's handy
May 31 14:28:41 <jic23> On that note, the beagle gsoc student who did the adc driver a few years back did a good job on the mainlining side. Something for the new guys to live up to ;)
May 31 14:31:27 <alexhiam> I didn't realize that was a gsoc project (before my time)
May 31 14:32:20 <jic23> http://beagleboard-gsoc13.blogspot.co.uk/search?updated-min=2013-01-01T00:00:00-08:00&updated-max=2014-01-01T00:00:00-08:00&max-results=39
May 31 14:33:57 <alexhiam> oh wowsers, they documented the whole mainlining process! that should be required reading for all our students
May 31 14:34:54 <jic23> Yeah, wasn't exactly smooth riding  (can't imagine who changed his mind about how to suggest doing some of it twice during the process....)
May 31 14:36:28 <jic23> Was funny - most people were assuming he was a TI contractor with decent amount of experience (as was I for at least the first 3 revisions) until Greg KH posted a nice note on G+ about how well the guy he was mentoring was getting on ;)
May 31 14:36:41 <jic23> Actually that might have been after we took the first few patches.
May 31 14:38:21 <alexhiam> oooh right, that was Greg KH's project
May 31 14:40:07 * alexhiam is doing the weekly cleaning - down to 7 tab filled browser windows from 11...
May 31 14:40:24 <jic23> yeah - Greg KH effectively mentors about half the kernel community ;)
May 31 14:41:28 <jic23> In theory he's the contact for Kernel GSoc projects still in general I think, but they haven't taken any for a few years.
May 31 14:44:13 <alexhiam> hmm, turns out building Linux inside Dropbox is not exactly an efficient use of CPU...
May 31 14:44:53 <alexhiam> cc1: ~10% CPU, Dropbox: 125.4%
May 31 15:10:01 <jic23> alexhiam *laughs*
May 31 15:13:41 <alexhiam> I moved the kernel source right after posting that, and dropbox is still using a whole core just deleting it :|
May 31 15:16:39 <alexhiam> jic23: you have any thoughts on the 'proper way' for students to be workign on kernel stuff? e.g. https://github.com/RobertCNelson/bb-kernel vs just working in a clone of https://github.com/beagleboard/linux
May 31 15:17:04 <alexhiam> and just pushing to a repo with an individual driver vs a clone of bb.org/linux, etc.
May 31 15:17:48 <jic23> I'd always work within a full tree - can generate patches later for RCN tree from that.
May 31 15:17:55 <jic23> Much less fiddly!
May 31 15:18:05 <jic23> + you always have the kernel you think you have running.
May 31 15:18:24 <alexhiam> yeah. The bb-kernel repo is nice for deploying kernels to an sd card, but that can also just be done manually
May 31 15:39:08 <alexhiam> wait, what is the best way to deploy the uImage to an sd card when building from the rcn tree?
May 31 15:39:46 * alexhiam has always used the bb-kernel tools
May 31 15:51:49 <jkridner> nerdboy: Digi-Key SO# 46737976 has your ADC boards on the way.
May 31 15:51:53 <jkridner> alexhiam: have you seen amr?
May 31 15:52:03 <alexhiam> jkridner: not for a while...
May 31 15:52:03 <kiran4399> alexhiam: where should I submit the weekly-report?
May 31 15:52:18 <alexhiam> kiran4399: the beagleboard-gsoc ml
May 31 15:52:31 <kiran4399> OK..
May 31 15:54:18 <alexhiam> kiran4399: jic23 pointed out that there's a patchset on the iio ml that updates the inv_mpu6050 driver in a much nicer way than that android one
May 31 15:54:23 <alexhiam> kiran4399: http://marc.info/?l=linux-iio&m=146358387211983&w=2
May 31 15:54:46 <alexhiam> so we probably just want to apply that to bb.org/linux
May 31 15:55:11 <alexhiam> kiran4399: also, you don't want to change am33xx.dts, you'll want to make a new file called something like am33xx-boneblue.dts (to be consistent with the current naming). It'll probably look a bit like am335x-bonegreen-wireless.dts, I think its got the same wifi/bt chipset
May 31 15:55:44 <kiran4399> alexhiam: yeah.. exactly..
May 31 15:55:47 <kiran4399> I've done like that..
May 31 15:56:17 <kiran4399> but I made the am33xxboneblue.dts from strawson cape dts..
May 31 15:56:21 <kiran4399> is that ok
May 31 15:56:34 <alexhiam> kiran4399: I'm looking at this commit: https://github.com/kiran4399/beaglebone_kernel/commit/fe5747e33e4943b912ea4165dcd2d2a3ee6703ad
May 31 15:56:44 <alexhiam> where you changed am33xx.dtsi
May 31 15:57:27 <alexhiam> (need better commit messages! "defconfig for blue added" doesn't describe that change)
May 31 15:58:06 <kiran4399> alexhiam: https://github.com/kiran4399/beaglebone_kernel/blob/4.4/arch/arm/boot/dts/am335x-boneblue.dts
May 31 15:58:23 <kiran4399> alexhiam: I am sorry.. I'll write more meaningful commit messages
May 31 15:58:24 <alexhiam> and adding the new dts under the commit message "mpu.h added to include/linux" is no good either, be careful with your commits
May 31 15:59:13 <alexhiam> are you just doing `commit .`? it's best to explicitly `git add` individual files to make sure your committing separate changes separately
May 31 16:01:15 <alexhiam> kiran4399: that overlay won't work as is, overlays are different than the base dts. It should look more like https://github.com/kiran4399/beaglebone_kernel/blob/4.4/arch/arm/boot/dts/am335x-bonegreen-wireless.dts
May 31 16:01:17 <kiran4399> alexhiam: https://github.com/kiran4399/beaglebone_kernel/blob/4.4/arch/arm/boot/dts/am335x-boneblue.dts#L395
May 31 16:01:21 <kiran4399> look at this..
May 31 16:01:38 <kiran4399> alexhiam: ok.. got your earlier message.
May 31 16:02:24 <alexhiam> jkridner: do we want the base boneblue dts muxing and enabling everything?
May 31 16:02:33 <alexhiam> I guess there's no reason not to...
May 31 16:02:49 <jkridner> yeah.
May 31 16:02:57 <jkridner> but, the default should match the schematic.
May 31 16:03:07 <alexhiam> right
May 31 16:03:28 <alexhiam> for now we just go by the Strawson cape
May 31 16:03:50 <alexhiam> kiran4399: check out the way the new driver on iio does it: http://marc.info/?l=linux-iio&m=146358381111940&w=2
May 31 16:03:53 <kiran4399> alexhiam: so you want me to write the mpu9150 in the base dts for beaglebone blue..
May 31 16:04:04 <jkridner> mpu9250
May 31 16:04:30 <alexhiam> kiran4399: yeah, that'll go in the base dts, along with encoders, pwm, etc.
May 31 16:06:08 <kiran4399> alexhiam: do you have the base dts for blue?
May 31 16:06:24 <alexhiam> no.. that's what you're writing ;)
May 31 16:06:50 <alexhiam> jkridner may have it started
May 31 16:07:14 <jkridner> yeah, but it might take me longer to dig up my work.
May 31 16:07:21 <alexhiam> jkridner: does the blue have the same wifi/bt chipset as the green wireless?
May 31 16:07:26 <jkridner> yes.
May 31 16:07:52 <alexhiam> cool, so kiran4399, you can copy the greenwireless dts and build off that
May 31 16:08:51 <alexhiam> jkridner: is there a script somewhere for deploying a kernel to an sd card built from rcn's tree (not bb-kernel)?
May 31 16:09:17 * alexhiam is still a kernel newb in many ways...
May 31 16:09:26 <jkridner> which rcn kernel are you interested in that isn't on http://github.com/beagleboard/linux ?
May 31 16:09:39 <jkridner> http://builds.beagleboard.org provides built .deb files..
May 31 16:09:42 <kiran4399> alexhiam: but would'nt the base dts for blue be like the base dts for beaglebone black?
May 31 16:09:48 <jkridner> the scripts merged into the tree build .deb files.
May 31 16:09:59 <jkridner> you just use 'dpkg -i XXX.deb' to install.
May 31 16:09:59 <alexhiam> jkridner: I mean that one, just not using https://github.com/RobertCNelson/bb-kernel
May 31 16:10:27 <alexhiam> ohhh, I'm just building it wrong..
May 31 16:10:35 <alexhiam> I was building a uImage
May 31 16:12:23 <kiran4399> alexhiam: you got my earlier messae?
May 31 16:12:49 <alexhiam> kiran4399: https://github.com/kiran4399/beaglebone_kernel/blob/4.4/arch/arm/boot/dts/am335x-bonegreen-wireless.dts#L10
May 31 16:12:56 <alexhiam> it includes it
May 31 16:15:51 <kiran4399> alexhiam: so.. when I was building the kernel using bb-kernel..
May 31 16:16:04 <kiran4399> alexhiam: it clones the linux trovalds kernel..
May 31 16:16:18 <kiran4399> but I want it to clone the beaglebone kernel..
May 31 16:17:01 <alexhiam> kiran4399: the bb.org/linux tree is mainline with patches applied, bb-kernel clones mainline and applies those patches
May 31 16:17:34 <kiran4399> alexhiam: yeah.. as I expected.. so do you want me to generate the patch and then apply it?
May 31 16:18:08 <kiran4399> alexhiam: also is it like the bb-kernel will apply the beaglebone patch to the linux trovalds kernel and I have to then apply my patch(where I did my modifications)?
May 31 16:19:21 <alexhiam> kiran4399: bb-kernel gives you a patched tree to work in, then you can generate patches from changes you make there
May 31 16:20:49 <alexhiam> kiran4399: but jic23 pointed out that it's better to work directly in a clone of the rcn tree (as you're doing now)
May 31 16:24:43 <jic23b> Yeah work messily on a git branch then create a clean patch set later. Editting history is always fun :)
May 31 16:26:01 <jic23b> Though in gsoc case keep the messy branch as well as will show the path as well as the destination. Normally you just drop the messy branch and pretend it never existed :)
May 31 16:32:07 <jkridner|work> kiran4399: patches you want in torvald's tree should be submitted to the appropriate maintainer's list.... bb.org won't do that for you. you can use the bb.org kernel to debug, but you want to submit the patch upstream before declaring yourself done. You should debug against mainline as well. When you submit the patch upstream, bb.org has no problem backporting.
May 31 16:33:57 <alexhiam> kiran4399: looks like there's already a driver for the pressure sensor: https://github.com/beagleboard/linux/blob/4.4/drivers/iio/pressure/bmp280.c
May 31 16:34:42 <kiran4399> alexhiam: yeah.. that's right.. I saw that yesterday.. bmp280 is for both bmp180 and bmp280
May 31 16:35:06 <alexhiam> so you might not have any drivers to upstream after all ;)
May 31 16:35:45 <jic23b> Is there a blue schematic out there?
May 31 16:35:48 <alexhiam> kiran4399: so next tasks are the boneblue dts, and grabbing that patchset for the inv_mpu6050 driver
May 31 16:36:25 <jic23b> Not the inv patch set is going through at least one more rev.
May 31 16:36:27 <jic23b> Note!
May 31 16:36:54 <jic23b> Won't change too much though probably.
May 31 16:37:21 <kiran4399> alexhiam: why should we need the patchset for inv_mpu6050?
May 31 16:37:28 <alexhiam> hopefully good enough for getting the blue APIs up and running, then eventually the final version will be backported from mainline
May 31 16:38:09 <alexhiam> kiran4399: it updates it to support all the invensense parts in a much better way than that android one
May 31 16:38:32 <alexhiam> and it's already on it's way to being upstreamed
May 31 16:56:04 <kiran4399> alexhiam: but we don't have any other invesense drivers apart from the android ones??
May 31 16:57:11 <alexhiam> kiran4399: it's a set of patches for the inv_mpu6050 driver that's already in mainline: https://github.com/beagleboard/linux/tree/4.4/drivers/iio/imu/inv_mpu6050
May 31 16:57:42 <kiran4399> alexhiam: but that is only for mpu 6050.. we need for mpu 9150 right?
May 31 16:58:54 <alexhiam> it updates it to work with any mpu6050 based device, e.g. the 9150, 9250, etc.
May 31 16:59:20 <alexhiam> and does it in a cleaner way than that one from android
May 31 16:59:43 <kiran4399> alexhiam: ok.. so you want me to leave the android driver?
May 31 17:00:04 <alexhiam> yeah, I'd say so
May 31 17:00:40 <kiran4399> OK.. alexhiam: well.. mpu 9150 is just 6050 and the magnetometer.. so it should'nt be a problem..
May 31 17:03:03 <alexhiam> kiran4399: yeah, one of the example dt configs is for the 9150: http://marc.info/?l=linux-iio&m=146358381111940&w=2
May 31 17:08:26 <kiran4399> alexhiam: I did'nt get it.. what exactly is the one which you sent me?
May 31 17:08:49 <alexhiam> one of the patches, scroll down to see the Device Tree examples
May 31 17:11:15 <kiran4399> alexhiam: why isn't this patch there in the mainline?
May 31 17:11:45 <alexhiam> kira4399: that's on the iio ml, which is separate from the mainline ml
May 31 17:12:23 <kiran4399> can I get that somewhere?
May 31 17:12:26 <kiran4399> alexhiam: ^^
May 31 17:12:43 <alexhiam> get what?
May 31 17:13:16 <kiran4399> iio ml
May 31 17:16:01 <alexhiam> kiran4399: http://vger.kernel.org/vger-lists.html#linux-iio
May 31 17:18:11 <kiran4399> alexhiam: so right now.. you want me to generate the patchset which when applied to mpu6050, makes the driver to work even for mpu9150..
May 31 17:18:19 <kiran4399> alexhiam: right?
May 31 17:19:00 <alexhiam> not generate it, it's already there on the ml. You want to pull the patches adn apply them to bb.org/linux, then build and test
May 31 17:19:45 <kiran4399> alexhiam: you mean it is already there in the iio ml?
May 31 17:20:04 <alexhiam> yeah, that link is the patches
May 31 17:20:30 <alexhiam> starting here: http://marc.info/?l=linux-iio&m=146358387211983&w=2
May 31 17:20:41 <alexhiam> look at the subject, "0/7"
May 31 17:20:45 <kiran4399> alexhiam: ok..
May 31 17:20:59 <alexhiam> you can click the "Download messageRAW" link to get the raw patch
May 31 17:21:17 <jic23c> You *might* need some precursors as iirc there have been some other updates since 4.4
May 31 17:21:17 <alexhiam> and the "next in list" link to get the next one
May 31 17:21:51 <alexhiam> jic23c: I was wondering...
May 31 17:22:29 <alexhiam> jic23c: changes to inv_mpu6050? or core iio stuff?
May 31 17:23:49 <jic23c> Possibly both though not that much core...
May 31 17:24:41 <kiran4399> alexhiam: you want me to apply all 7 patches(I mean are these 7 patches called as a patchset?)..?
May 31 17:24:50 <alexhiam> yeah
May 31 17:26:30 <jic23c> Shucks 4.4 might be pre spi support.. quite a few precursors...
May 31 17:26:55 <jic23c> https://git.kernel.org/cgit/linux/kernel/git/jic23/iio.git/log/drivers/iio/imu/inv_mpu6050?h=testing
May 31 17:29:20 <alexhiam> yeah, the rcn tree is at 2015-07-20 - Merge 4.2-rc3 into staging-next
May 31 17:29:26 <alexhiam> https://github.com/beagleboard/linux/commits/4.4/drivers/iio/imu/inv_mpu6050
May 31 17:29:34 <alexhiam> :|
May 31 17:29:57 <jic23c> Yeah spi series was 4.6... ah well good experience for Koran :)
May 31 17:30:17 <jic23c> Gah auto correct!
May 31 17:30:20 <jic23c>  sorry
May 31 17:30:32 <jic23c> Kiran !
May 31 17:30:43 <alexhiam> it does have a spi file though https://github.com/beagleboard/linux/blob/4.4/drivers/iio/imu/inv_mpu6050/inv_mpu_spi.c
May 31 17:31:06 <nerdboy> the bb-kernel is manual deploy
May 31 17:31:19 <nerdboy> same as yocto/oe build
May 31 17:32:02 <nerdboy> if you're building it yourself with build script then copy the deploy dir stuff to your bb
May 31 17:32:18 <alexhiam> does this seem right? https://gist.github.com/alexanderhiam/29fcf814f731dd94f8850961f1a61ae3
May 31 17:32:30 <alexhiam> cc jkridner|work ^
May 31 17:32:41 <chanakya_vc> Hey bradfa are you there?
May 31 17:32:48 <jic23c> Alexhiam. RCN has back ported current mainline IIO!
May 31 17:33:08 <jic23c> Doing all Kiran's work for him!
May 31 17:33:19 <alexhiam> oooohhh, that last commit there
May 31 17:33:37 <alexhiam> it's a big one :P
May 31 17:33:40 <chanakya_vc> bradfa,  Sorry for the delay.Had to come back home from the lab and I got stuck in traffic :(
May 31 17:33:59 <jic23c> Quite he cut and paste the directory I think... nasty.
May 31 17:34:08 <alexhiam> lol
May 31 17:34:32 <jic23c> Easy enough if more time consuming to cherrypick.
May 31 17:34:52 <jic23c> Anyhow off for a bit.
May 31 17:35:56 <chanakya_vc> alexhiam, jic23c  Do you use Jekyll for Blogging?
May 31 17:36:03 <nerdboy> alexhiam: looks reasonable, except i don't usually use debs
May 31 17:36:36 <chanakya_vc> nerdboy, mdp bradfa ^^
May 31 17:36:38 <alexhiam> nerdboy: debs seem like the best way to deploy to an existing bone fs
May 31 17:37:19 <alexhiam> chanakya_vc: I use wordpress + a markdown plugin (though I also don't do much blogging these days...)
May 31 17:38:05 <nerdboy> use .rst or .md in github wiki pages
May 31 17:38:07 <chanakya_vc> Okay alexhiam So you would suggest Wordpress? Is it better than Gitpages?
May 31 17:38:26 <nerdboy> if you use rst you can pasted it into stuff like pelican
May 31 17:39:01 <nerdboy> chanakya_vc: you mean gh-pages branch or the auto stuff?
May 31 17:39:20 <alexhiam> chanakya_vc: if it's for your gsoc project you might as well use gh-pages
May 31 17:39:35 <chanakya_vc> alexhiam, gh pages
May 31 17:39:38 <chanakya_vc> nerdboy, ^^
May 31 17:39:45 <nerdboy> gh-pages just displays what you put there
May 31 17:40:01 <chanakya_vc> But wait I have used the automatic template and stuff
May 31 17:40:04 <nerdboy> meaning it has to be valid html, etc
May 31 17:40:11 <alexhiam> which supports Jekyll templates
May 31 17:40:24 <nerdboy> if the template stuff dumps the right format
May 31 17:40:32 <nerdboy> shoudl be good to go
May 31 17:40:58 <alexhiam> I should say I use wordpress.org, which is self hosted so you'd need a server to use it. Not a fan of wordpress.com
May 31 17:41:02 * nerdboy usually throws a bunch of static html at gh-pages
May 31 17:41:04 <alexhiam> gh-pages is free
May 31 17:41:41 <chanakya_vc> Okay alexhiam .But setting up jekyll is so difficult.I compiled ruby from source and then when I tried to install jekyll into local host,it said it needed something called Open ssl\
May 31 17:41:46 <chanakya_vc> whatever that is
May 31 17:42:16 <nerdboy> http://sarnold.github.io/gwocss/  <=  https://github.com/sarnold/gwocss/tree/gh-pages
May 31 17:42:35 <alexhiam> first of all, you should certainly read up on openssl ;)
May 31 17:43:12 <chanakya_vc> Anyway alexhiam nerdboy I am working on my weekly progress report right now.Should post it in about an hour or so
May 31 17:43:19 <chanakya_vc> Okay I will alexhiam
May 31 17:43:29 * nerdboy enjoys the simplicity/portability of reStructuredText
May 31 17:43:44 * alexhiam wants markdown for all things
May 31 17:43:59 <nerdboy> silly markdown people...
May 31 17:44:08 <alexhiam> until I want to do something that it doesn't support....
May 31 17:44:30 <chanakya_vc> I need to set up a blog.The thing is that when I will set up my own site,I might use wordpress for that.So won't wordpress be better in that sense?
May 31 17:44:57 <chanakya_vc> alexhiam,nerdboy  ^^
May 31 17:45:18 <chanakya_vc> Or are gh pages integrable with wordpress
May 31 17:45:46 <nerdboy> gh-pages is just a branch that does we rendering of what you put there
May 31 17:45:52 <nerdboy> *web even
May 31 17:46:08 <alexhiam> you can just point a domain (or subdomain) at gh-pages
May 31 17:49:41 <chanakya_vc> Okay got it alexhiam nerdboy Thanks : )
May 31 17:57:11 <nerdboy> if you have a machine to do your own blog-y thing then try installing pelican
May 31 17:57:56 <nerdboy> "make html" from .rst to build your site, theme it with bootstrap
May 31 17:58:17 <kiran4399> alexhiam: there?
May 31 17:58:21 <alexhiam> yo
May 31 17:58:21 <nerdboy> alexhiam: can you do this with markdown?  https://github.com/VCTLabs/foss-rst-presentations/blob/master/hardened/hardened_system_profile.rst
May 31 17:58:39 <kiran4399> if you see in the patch 5/7
May 31 17:59:28 <alexhiam> nerdboy: ah, certainly not as cleanly
May 31 17:59:40 <alexhiam> I think a toc requires extra tools
May 31 17:59:41 <nerdboy> https://github.com/VCTLabs/vct-web  <= readme for pelican setup
May 31 17:59:45 <kiran4399> it is said that mpu 9150 is 6050 and ak9875 on chip.
May 31 17:59:48 <kiran4399> alexhiam: ^^
May 31 18:00:06 <alexhiam> kiran4399: yeah
May 31 18:00:14 <nerdboy> alexhiam: rst2pdf that file
May 31 18:01:04 <nerdboy> pelican also does markdown
May 31 18:02:15 <alexhiam> nerdboy: very snazzy
May 31 18:02:30 <kiran4399> alexhiam:  example containing mpu9150 dts should be added to the base dts for blue right??
May 31 18:02:34 <alexhiam> pandoc does have some mardown support
May 31 18:02:39 <nerdboy> which part?
May 31 18:02:48 <alexhiam> nerdboy: the pdf
May 31 18:02:48 <kiran4399> alexhiam: +Example describing mpu9150 (which includes an ak9875 on chip)
May 31 18:03:02 <kiran4399> alexhiam: i am talking about this ^^
May 31 18:03:38 <alexhiam> kiran4399: you can use that config with the cape for now. Remember the blue will have the 9250, so we'll want to see if there's anything that'll need to be different
May 31 18:04:21 * nerdboy builds has new resume format, slides, documents, and top-hats all with a Makefile
May 31 18:06:09 <nerdboy> make "new car"
May 31 18:06:20 <chanakya_vc> nerdboy, What is this pelican thing?And could you also guide me regarding rst
May 31 18:06:33 <chanakya_vc> nerdboy, I only know Markdown
May 31 18:06:58 <nerdboy> see the readme
May 31 18:07:06 <nerdboy> https://github.com/VCTLabs/vct-web  <= readme for pelican setup
May 31 18:07:22 <chanakya_vc> nerdboy, Currently in the process of downloading open ssl for installing jekyll
May 31 18:07:25 <nerdboy> ignore the vct parts and use the references
May 31 18:07:29 <chanakya_vc> Okay will do
May 31 18:08:43 <Wormo> chanakya_vc: http://www.unexpected-vortices.com/doc-notes/markdown-and-rest-compared.html
May 31 18:11:09 <nerdboy> rst is kinda python-centric
May 31 18:11:13 <kiran4399_> alexhiam: alright then.. looks like I am all set for the next task..
May 31 18:11:27 <nerdboy> docutils, sphinx, rst2pdf, etc
May 31 18:11:48 <alexhiam> kiran4399_: great
May 31 18:17:20 <chanakya_vc> nerdboy, So from What I understood it converts markdown to Html which you can upload?
May 31 18:17:39 <chanakya_vc> Wormo, Will go through it.
May 31 18:18:53 <kiran4399_> alexhiam: I am getting this error when I am running the rebuild command..
May 31 18:19:04 <Wormo> chanakya_vc: pelican does that, yes, pregenerated html for given content instead of cgi scripts creating the menus and so forth dynamically
May 31 18:19:13 <kiran4399_> alexhiam: /bin/sh: 0: Can't open /home/kiran/bb-kernel/tools/tools/host_det.sh
May 31 18:19:38 <kiran4399_> alexhiam: oh.. sorry.. got it.. looks like a bug..
May 31 18:20:11 <chanakya_vc> Wormo, Okay. Got it. But for the lack of time,I guess I will stick with Markdown and Jekyll
May 31 18:20:24 <chanakya_vc> Maybe try it out later
May 31 18:21:48 <nerdboy> yeah, markdown, rst, maybe other formats
May 31 18:48:24 <bradfa> chanakya_vc: sorry, very busy right now with another engineer, but if you write up an email I can reply later tonight
May 31 18:49:27 <chanakya_vc> bradfa, I am working on the code as of now.Perhaps you could help me with sigrok tomorrow.I would prefer to sort it out on the IRC.
May 31 18:58:51 <pmezydlo1> av500: hi, are you there?
May 31 18:59:08 <pmezydlo1> _av500_:^^
May 31 19:36:04 <bradfa> chanakya_vc: yes, I can help with sigrok
May 31 19:36:27 <chanakya_vc> bradfa, Are you free now?
May 31 19:36:29 <bradfa> chanakya_vc: please don't forget to send your second status report soon so that it arrives on Tuesday for at least some mentors
May 31 19:36:43 <bradfa> chanakya_vc: no, sorry, not free now, have to finish up a few things :(
May 31 19:36:45 <chanakya_vc> bradfa,working on that now
May 31 19:37:12 <bradfa> chanakya_vc: I'll help you with your makefile and sigrok tomorrow, ok?
May 31 19:37:27 <chanakya_vc> bradfa, Np we can talk tomorrow.AS it is I am quite busy with the report and the code : )
May 31 19:37:42 <chanakya_vc> bradfa, Okay .No problem : )
May 31 20:02:49 * nerdboy has visiting gentoo dev so is a little distracted...
May 31 20:44:17 <_av500_> pmezydlo1: hi
May 31 20:44:20 <_av500_> I am
May 31 20:44:23 <_av500_> not for long
May 31 20:45:22 <pmezydlo1> hi
May 31 20:49:46 <pmezydlo1> _av500_:i'm wondering what should i write tomorrow. should be something about me, my project or something?
May 31 20:57:28 <_av500_> during the meeting
May 31 20:57:35 <_av500_> about your project
May 31 20:57:47 <_av500_> it does not have to be big
May 31 20:57:55 <_av500_> just what you are working on right now
May 31 20:58:08 <pmezydlo1> ok thanks
May 31 20:58:18 <_av500_> you dont need to explain all the basics
May 31 20:58:21 <_av500_> dive in
May 31 20:58:28 <pmezydlo1> it is a little boring
May 31 20:58:55 <pmezydlo1> bye
Jun 01 02:25:30 <Abhishek_> Hi
Jun 01 02:25:49 <Abhishek_> ZeekHuge: what are the errors you get when you try to compile BeagleLogic?
Jun 01 02:29:01 <ZeekHuge> Hi Abhishek_ ! How are you ? :)
Jun 01 02:29:11 <ZeekHuge> here it is : http://paste.debian.net/713094/
Jun 01 02:29:58 <Abhishek_> Ok, so it seems you are able to compile the module but it's not gonna work
Jun 01 02:30:15 <ZeekHuge> yep.
Jun 01 02:31:29 <ZeekHuge> What can be the problem according to you ?
Jun 01 02:31:58 <ZeekHuge> Abhishek_: ^
Jun 01 02:32:04 <Abhishek_> So BeagleLogic needs to "plug in" to the remoteproc module so that it can start/stop/reset the PRUs and handle interrupts.
Jun 01 02:32:20 <Abhishek_> There are a few "glue" functions that you'll have to implement
Jun 01 02:32:52 <Abhishek_> The kernel API has changed, so this may require changes to the BeagleLogic module itself
Jun 01 02:33:28 <Abhishek_> https://github.com/abhishek-kakkar/BeagleLogic/blob/master/beaglelogic-kernel-driver/beaglelogic_glue.h
Jun 01 02:34:36 <Abhishek_> The 3.8.13 remoteproc module had defined a concept of "upcalls" and "downcalls" if you look through the code
Jun 01 02:34:55 <Abhishek_> (pru_rproc.c in the 3.8.13 kernel tree)
Jun 01 02:35:11 <ZeekHuge> okay, so these undefined symbols were basically defined by remoteproc driver and then beaglelogic request to bind/unbind using it ?
Jun 01 02:35:17 <Abhishek_> yeo
Jun 01 02:35:19 <Abhishek_> *yep
Jun 01 02:36:09 <Abhishek_> Look at the 3.8.13 kernel tree in the beagleboard kernel, you will find bind/unbind functions in the pru_rproc.c file
Jun 01 02:38:55 <ZeekHuge> i can see it .
Jun 01 02:39:04 <ZeekHuge> https://github.com/beagleboard/linux/blob/3.8/drivers/remoteproc/pru_rproc.c#L2452
Jun 01 02:40:55 <ZeekHuge> and as far as 4.4.11 is concerned, I am unable to see anything LIKE a bind function
Jun 01 02:41:51 <ZeekHuge> https://github.com/RobertCNelson/linux-stable-rcn-ee/tree/4.4.11-ti-r29/drivers/remoteproc
Jun 01 02:43:49 <Abhishek_> yep, there's no downcalls and upcalls either
Jun 01 02:44:15 <Abhishek_> ( I miss them )
Jun 01 02:46:29 <ZeekHuge> downcalls and upcalls basically used to generate and interrupt right ? and the PRU used to respond according to the what interrupt it is ? correct ?
Jun 01 02:47:04 <Abhishek_> It was a way to interact with PRU firmware'
Jun 01 02:47:26 <ZeekHuge> okay interrupt = PRUSS INTC
Jun 01 02:47:31 <Abhishek_> yep
Jun 01 02:49:00 <Abhishek_> so basically on a very high level a pru_downcall in the kernel would lead to a function name handle_downcall in the PRU being called with the *exact same parameters* that you passed when you did pru_downcall in the kernel
Jun 01 02:49:03 <ZeekHuge> and driver was like, it just reserved some space on the ram and tell about it to one of the PRUs and the pru would then just write on that location ?
Jun 01 02:49:36 <Abhishek_> on a high level, yes
Jun 01 02:49:58 <Abhishek_> the devil is in the details :)
Jun 01 02:50:04 <ZeekHuge> :)
Jun 01 02:50:15 <ZeekHuge> can we use a single pru for this project ?
Jun 01 02:50:44 <ZeekHuge> I mean communicating over the interconnect. probably using DMA ?
Jun 01 02:51:02 <Abhishek_> explore yourself :)
Jun 01 02:51:02 <ZeekHuge> going to the reference manual .
Jun 01 02:51:28 <ZeekHuge> okay. I just wanted to know if you tried this ?
Jun 01 02:51:49 <Abhishek_> I've not tried it so far.
Jun 01 02:52:25 <ZeekHuge> Right, so experimenting, but once the project is gets some pace.
Jun 01 02:52:43 <ZeekHuge> So what do you think can be a workaround ?
Jun 01 02:52:48 <Abhishek_> for what?
Jun 01 02:53:18 <ZeekHuge> I need the beaglelogic's firmware, and the kernel side to support that firmware.
Jun 01 02:54:14 <ZeekHuge> I'll probably have to get into the latest rproc drivers first.
Jun 01 02:56:53 <ZeekHuge> Abhishek_: yes ?
Jun 01 02:58:10 <Abhishek_> Workaround for what, is my question.
**** ENDING LOGGING AT Wed Jun 01 02:59:59 2016