**** BEGIN LOGGING AT Wed Sep 28 02:59:58 2016
Sep 28 03:02:11 <ayjay_t> actually, i get such weird behavior if use vi on my BBB over serial
Sep 28 03:02:41 <ayjay_t> i haven't noticed any patterns yet but sometimes theres still recovery files after a proper close etc
Sep 28 03:08:46 <veremit> serial can be glitchy .. I was finding that with my imx6
Sep 28 03:08:59 <veremit> I think its the lack of flow control
Sep 28 03:11:04 <ayjay_t> i don't know how i feel about it- the problem should be with vim right?
Sep 28 03:11:11 <ayjay_t> i mean if it closes, it should do all of it's closing function calls or w/e
Sep 28 03:11:47 <ayjay_t> :-/ i've actually lose a ton of owrk
Sep 28 07:48:31 <ayjay_t> <_< >_> something weird is going on between userspace and the kernel
Sep 28 07:48:50 <ayjay_t> the TSLib program somehow impacts the pull transistors on the am335x's ADCs
Sep 28 07:48:53 <ayjay_t> pullup*
Sep 28 07:49:29 <ayjay_t> and stepping through with GDB shows the behavior will repeat in really odd places- like in a framebuffer write function
Sep 28 08:10:03 <Velsiol> Has anyone here tried to do anything with DSP/MSMC on a X15 or AM57 EVM yet?
Sep 28 08:10:17 <Velsiol> Getting some weird errors trying to allocate msmc memory
Sep 28 09:02:14 <Nik_> what is difference between black and XM
Sep 28 09:06:07 <veremit> black is in production .. xm .. less so ..
Sep 28 09:30:35 <tbr> XM is ooooold
Sep 28 09:42:44 <veremit> pong
Sep 28 12:01:13 <Velsiol> Hum, even the default image doesnt work with the DSP
Sep 28 12:01:45 <jkridner|pd> there are xM-specific images.
Sep 28 12:02:05 <jkridner> Velsiol: xM is back into the weekly build cycle.
Sep 28 12:03:05 <jkridner> https://rcn-ee.com/rootfs/bb.org/testing/2016-09-18/lxqt-xm-4gb/
Sep 28 12:04:00 <Velsiol> X15 :^)
Sep 28 12:04:55 <jkridner> X15 ships with DSP support via OpenCL.
Sep 28 12:05:36 <Velsiol> If i boot up the default image and try one of the OpenCL examples i get an "Unable to allocate OCL MSMC memory" error
Sep 28 12:05:46 <Velsiol> have tried multiple weeklies as well
Sep 28 12:05:53 <Velsiol> remaking the cmem module
Sep 28 12:10:44 <Velsiol> Wondering if anyone else has gotten it to work, one guy on the beagle forum reported he had the problem but didnt post a fix
Sep 28 12:10:56 <Velsiol> and on e2e theres a guy as well but no solution
Sep 28 12:11:22 <jkridner> make your own e2e post with what you tried?
Sep 28 12:11:39 <jkridner> X15, right?
Sep 28 12:13:58 <Velsiol> Yeah,
Sep 28 12:14:22 <Velsiol> The other person has p much exactly the same issue
Sep 28 12:14:22 <Velsiol> https://e2e.ti.com/support/embedded/linux/f/354/t/541868
Sep 28 12:14:48 <Velsiol> Apart from me getting no errors on startup
Sep 28 12:15:54 <Velsiol> Disregard that, dmesg gives me the same issue
Sep 28 12:16:39 <jkridner> ah... seems they got distracted on the EVM vs. Beagle thing... I'll see if I can escalate. You could try the SDK code to see if it works.
Sep 28 12:17:14 <jkridner> total disconnect on the SD stuff.
Sep 28 12:18:43 <Velsiol> SDK code? The SDK dsp demo's work, or do you mean using the sdk FW
Sep 28 12:19:25 <jkridner> I think the firmware should be the same.
Sep 28 12:19:48 <jkridner> I'm just wondering if it all loads and works properly on the SDK.
Sep 28 12:20:02 <jkridner> I just pinged Robert and will try to get the right TI contacts.
Sep 28 12:20:40 <Velsiol> Amazing, thanks!
Sep 28 12:20:50 <Velsiol> I'm booting up the SDK to check if it loads correctly
Sep 28 12:21:52 <Velsiol> On the SDK it loads without issue
Sep 28 12:22:20 <Velsiol> no " _wait_target_disable failed"
Sep 28 12:22:21 <jkridner> that was fast. OK, that's what I expected. I'll see if Robert is aware of this integration issue.
Sep 28 12:22:47 <Velsiol> Would posting full boot logs be helpful?
Sep 28 12:22:53 <Velsiol> (in pastebin of course)
Sep 28 12:23:19 <thinkfat> Velsiol: [ 14.605258] omap-rproc 41000000.dsp: dma_alloc_coherent err: 9437184
Sep 28 12:24:32 <Velsiol> Yes, That.
Sep 28 12:24:40 <thinkfat> Velsiol: coherent memory allocator probably ran out of juice?
Sep 28 12:24:59 <thinkfat> Velsiol: maybe you need some "cma=" option on commandline?
Sep 28 12:26:46 <Velsiol> How do you mean? (<-- not that big on linux)
Sep 28 12:27:41 <thinkfat> Velsiol: I remember that the amount of memory available to dma_alloc_coherent is sort of limited, and that adding "cma=<big number>" to the kernel commandline might help
Sep 28 12:28:50 <Ragnorok> I think there's a line for that in uEnv.txt even.
Sep 28 12:29:30 <thinkfat> Velsiol: pasting the full boot log could indeed help :-)
Sep 28 12:31:26 <Velsiol> http://pastebin.com/gT4bKG5L
Sep 28 12:31:50 <Velsiol> in uEnv there is a cmdline=coherent_pool=1M quiet
Sep 28 12:32:16 <Velsiol> I guess increasing that should help?
Sep 28 12:32:18 <thinkfat> Velsiol: make that, like, 16M ?
Sep 28 12:33:33 <thinkfat> Velsiol: lines 19 and 20 look interesting
Sep 28 12:33:52 <thinkfat> Velsiol: the system creates a 8MB memory pool for DSP2
Sep 28 12:34:11 <thinkfat> Velsiol: but if I interpret the error message correctly, the image requires the allocation of 9MB
Sep 28 12:36:07 <Velsiol> i increased the coherent_pool to 16M but got the same error
Sep 28 12:36:19 <Velsiol> maybe set "cma="?
Sep 28 12:36:50 <thinkfat> Velsiol: yeah, I guess ( and it's really only a guess) you need to somehow increase the "dsp2_cma" pool, but I don't really know how to do that
Sep 28 12:39:20 <Velsiol> Allright, i'll check into that
Sep 28 12:39:47 <Velsiol> maybe in the dtb?
Sep 28 12:39:54 <thinkfat> potentially, yes
Sep 28 12:41:13 <Velsiol> oh boy, come for OpenCL, stay for device tree blob editing
Sep 28 12:49:47 <thinkfat> Velsiol: you know, there's likely the dts file around to save you from that
Sep 28 12:50:07 <thinkfat> Velsiol: but if you insist on using the magnetic needle to flip bits...
Sep 28 12:50:59 <Ragnorok> DTB isn't *that* hard.  It's just typically cryptic and poorly documented.  (grin)
Sep 28 12:53:31 <Velsiol> A.k.a. "fun"
Sep 28 12:54:32 <thinkfat> for people born with a rusty nail in their skulls
Sep 28 12:57:15 <Ragnorok> I can't say I'm a fan of DTB, but it's what we got, and it could be worse.  At least dts are sort of readable.  Ofc when I say dtb I mean compile from dts.  Direct dtb is madness.
Sep 28 13:00:18 <thinkfat> höhö
Sep 28 13:02:23 <Velsiol> Yeah i meant decompiling(?) it
Sep 28 13:03:23 <Ragnorok> I moved my BBB over to journald for all logging, and seems ... erratic.  Like today it didn't capture a boot log anywhere, but yesterday it did.  Being a journald n00b I suspect I've done something wrong, but it doesn't appear to be complicated, so I can't imagine what.
Sep 28 16:13:22 <ayjay_t> :-(
Sep 28 16:24:24 <Ragnorok> There is no joy for the ayjay type today.
Sep 28 16:27:29 <ayjay_t> i just saw the circuitco video on making beaglebone blacks and that was pretty exciting
Sep 28 17:07:24 <mhoney> ayjay_t, do you have a link?
Sep 28 17:07:56 <ayjay_t> https://www.youtube.com/watch?v=FcqQvH41OR4
Sep 28 17:08:12 <ayjay_t> they have some pretty nice equipment
Sep 28 17:20:40 <mhoney> Thank you ayjay_t
Sep 28 17:49:21 <xzcvczx> anyone know if the latest jessie flasher console image is bad? i checked shasum, use etcher which "validates" the image, and the board seems to have flashed correctly then powered down but when i restart it i just get user led 0/1 on solid and then nothing happens
Sep 28 17:49:55 <xzcvczx> jessie flash console snapshot
Sep 28 17:50:06 <tbr> hmm, not sure about which, but one of them was reported problematic
Sep 28 17:50:17 <xzcvczx> ah k, thanks :)
Sep 28 17:50:47 <tbr> if you want to know what's happening, look at the debug UART output
Sep 28 17:50:59 <xzcvczx> yeah i don't have an ftdi handy
Sep 28 17:51:08 <xzcvczx> i know i have a couple of them, but damned if i know where they are
Sep 28 17:51:21 <tbr> :)
Sep 28 17:51:32 <tbr> so, yeah, use an older one
Sep 28 17:52:44 <xzcvczx> oooo theres an updated one thats not linked from the wiki page, will try that
Sep 28 17:52:52 <tbr> or that
Sep 28 17:54:15 <xzcvczx> hmmm the bootloader doesn't get flashed eh?
Sep 28 17:54:49 <xzcvczx> i am currently flashing a big image onto the emmc and i have got bored of waiting can i just yank the power out with no issues? (well except for their being no bootable image in emmc)
Sep 28 17:55:29 <tbr> not recommended
Sep 28 17:55:34 <xzcvczx> doh
Sep 28 17:55:35 <xzcvczx> ok
Sep 28 17:55:39 <tbr> try pressing the power button
Sep 28 17:58:14 <xzcvczx> the power button is the one right next to the ethernet connector?
Sep 28 17:59:55 <tbr> IIRC, yes
Sep 28 18:00:01 <xzcvczx> cool
Sep 28 18:01:19 <tbr> btw: yanking out power in general is not recommended as over time the SoC might fail due to the way it shuts down
Sep 28 18:01:31 <xzcvczx> oh
Sep 28 18:01:33 <tbr> so doing a full shutdown is recommended
Sep 28 18:01:40 <xzcvczx> wow never knew that
Sep 28 18:01:52 <tbr> I did frequently, it never failed
Sep 28 18:02:06 <tbr> I would avoid doing it automatically and repeatedly
Sep 28 18:02:22 <xzcvczx> i shall remember that
Sep 28 18:02:30 <xzcvczx> really need to find one of my many ftdi adapters as well
Sep 28 18:02:40 <xzcvczx> i know i have a few of them floating around 3v3 ones
Sep 28 18:02:44 <tbr> it's specific to the BBB board design and SoC AFAIU
Sep 28 18:03:03 <tbr> both 3.3 and 5V Vref should work
Sep 28 18:03:08 <tbr> 3.3 is preferred
Sep 28 18:12:07 <ayjay_t> okay so when i debug tslib, i know you gusy might not care but this is weird, it calls a "put_string_center" function which writes fonts directly the framebuffer
Sep 28 18:12:33 <ayjay_t> and that function has two lines and exits, but it goes through line 216, then 217, then BACK to 216 for no reason, and if i execute that line, it breaks the ADCs
Sep 28 18:12:59 <xzcvczx> that looks healthier
Sep 28 18:16:23 <xzcvczx> so yeah seems the 09-18 console flasher is fried but the 09-21 works
Sep 28 19:05:31 <ayjay_t> having strong doubts about the tsc driver
Sep 28 19:10:00 <ayjay_t> at least it breaks when using it with any of tslibs test programs
Sep 28 19:11:40 <ogres> so I guess there's no way I can add an extra 'Full' ethernet to my beagles...
Sep 28 19:11:51 <ogres> any ideas what i can do with 21 beaglebones?
Sep 28 19:12:47 <jkridner> ogres: you mean to have 2 Ethernet adapters.... you'd have to do some crazy mods.... there are 2 Ethernet MAC interfaces on the processor.
Sep 28 19:13:07 <ogres> huh?
Sep 28 19:13:31 <ogres> I was going to add an enc28j60 but that has severe I/O limitations
Sep 28 19:13:50 <jkridner> isn't that SPI or I2C or something?
Sep 28 19:14:07 <ogres> yeah
Sep 28 19:14:07 <jkridner> BeagleBones have 2 REAL Ethernet interfaces built-in with a built-in switch.
Sep 28 19:14:08 <ogres> spi
Sep 28 19:14:17 <ogres> um
Sep 28 19:14:22 <ogres> 2 ethernets?
Sep 28 19:14:32 <ogres> then how would I add the second port?
Sep 28 19:14:46 <Ragnorok> zmatt said the second isn't usable because of either pin conflict, pins not routed out, or both.
Sep 28 19:14:50 <jkridner> example with same processor and 2 gigabit ethernets: http://www.ti.com/tool/TMDSSK3358
Sep 28 19:15:02 <jkridner> Ragnorok: sounds right.
Sep 28 19:15:12 <ogres> oh ok
Sep 28 19:15:24 <jkridner> USB would likely be faster than SPI.
Sep 28 19:15:39 <ogres> yea this has no USB port
Sep 28 19:15:53 <ogres> and there are no SPI modules that do not have TCP connection limits
Sep 28 19:16:04 <ogres> so I guess I need to figure out what other project to make
Sep 28 19:16:17 <ogres> or what I will do with all these
Sep 28 19:16:46 <jkridner> no USB?
Sep 28 19:17:15 <jkridner> great for motor control stuff.
Sep 28 19:17:33 <jkridner> coordinated quadcopters?
Sep 28 19:17:50 <jkridner> check out BBB-MINI on hackster.io
Sep 28 19:18:18 <ayjay_t> ogres ill buy a couple from you
Sep 28 19:22:11 <Ragnorok> At least the investment was small.  Heck for what you paid I'd take a few.  (grin)
Sep 28 19:22:14 <ayjay_t> especially since i know how much you paid for them haha
Sep 28 19:22:17 <ayjay_t> yeah no kidding
Sep 28 19:22:45 <Ragnorok> Long as the GPIO and the Ethernet work I'm all set.
Sep 28 19:22:56 <ayjay_t> i want serial...
Sep 28 19:24:16 <MathOnNapkins> Hrm, ayjay_t... might be an alias. *checks INTERPOL database*
Sep 28 19:25:06 <ayjay_t> <_< >_>
Sep 28 19:29:33 <ogres> gpio and ethernet work
Sep 28 19:29:40 <ogres> theres no power port tho
Sep 28 19:30:24 <ogres> I have one with a power port but im definitely keeping it for myself
Sep 28 19:39:05 <mhoney> it's a shame they didnt boost ram to 1gb on the bbb wireless ;(
Sep 28 19:49:56 <ayjay_t> the power pins are broken out anyway, if i recall
Sep 28 19:52:40 <ogres> ?
Sep 28 19:52:44 <ogres> broken out?
Sep 28 20:03:22 <Ragnorok> Available for use without significant effort.  Or search for "broken out" and see what happens.
Sep 28 20:11:41 <ogres> weird my beagles not showing on device manager
Sep 28 20:16:48 <ogres> yea nothing
Sep 28 20:26:51 <ogres> the LED's just stay on no blinking
Sep 28 20:41:45 <ogres> well removing the SD card made it boot
Sep 28 20:53:11 <ayjay_t> hmm maybe the ti driver doesn't handle all the appropriate adc errors
Sep 28 20:53:45 <ogres> maybe the SD card is empty?
Sep 28 20:54:13 <ogres> because it booted perfectly fine without it
Sep 28 22:27:47 <tomlee> Hi
Sep 28 22:58:57 <Unee0x> Has anyone experienced slow ssh connect with 8.4 Debian ?
Sep 28 23:18:03 <ogres> i do
Sep 28 23:19:22 <ayjay_t> hey
Sep 28 23:57:59 <vagrantc> Unee0x: slow ssh connection might be name resolution issues
Sep 28 23:58:16 <vagrantc> e.g. can't reach the nameserver listed in /etc/resolv.conf
Sep 29 00:08:32 <Unee0x> vagrantc: thanks, looking into it right now
Sep 29 00:29:28 <ogres> hey Unee0x
Sep 29 00:30:36 <Unee0x> ogres: hey?
Sep 29 00:31:02 <ogres> hey I was wondering something... if I insert the SD while the beagle is on, will I break anything?
Sep 29 00:35:13 <vagrantc> depends on how much force you use :)
Sep 29 00:47:50 <ogres> i meaaan
Sep 29 00:47:57 <ogres> will it reboot or something
**** ENDING LOGGING AT Thu Sep 29 02:59:58 2016