**** BEGIN LOGGING AT Sat Aug 02 03:00:00 2014
Aug 02 11:47:43 <Laurenceb__> does BBB support wake on lan?
Aug 02 12:09:53 <tbr> Laurenceb__: I don't think so, but the answer will be in the SRM and the TRM
Aug 02 12:15:39 <zmatt> Laurenceb__: do I understand your remark of earlier right in that linear regulator U4 stays enabled after shutdown?
Aug 02 12:17:08 <zmatt> I'm guessing you have BBB rev A6 or later?  (I have A5A which predates changes made to U4 enable)
Aug 02 12:19:22 <zmatt> (and that you fixed it by pulling down U4's EN input?)
Aug 02 12:24:55 <Laurenceb__> yes
Aug 02 12:25:00 <Laurenceb__> i have A6
Aug 02 12:25:35 <Laurenceb__> i was wondering if it was a deliberate decision to allow WOL
Aug 02 12:25:40 <Laurenceb__> but it doesnt look like it
Aug 02 12:26:31 <Laurenceb__> so ive got VDD_AUX -> 100nF and 100k in parallel -> U4 EN pin -> 200k ground
Aug 02 12:27:31 <Laurenceb__> there is lots of hysteresis on U4 enable but VDD_AUX only swings from 3.3v to 1.2v
Aug 02 12:27:37 <zmatt> since the implications are pretty bad:  it means that 3V3B stays on while all other power rails are off by leaking enough current through I/Os back into the 3V3A supply to keep it high enough for the regulator to stay enabled, even with the LDO's discharge to ground
Aug 02 12:27:47 <zmatt> VDD_AUX isn't used as EN since A6
Aug 02 12:28:01 <Laurenceb__> hmmm
Aug 02 12:28:12 <zmatt> 3V3A is the enable
Aug 02 12:28:24 <zmatt> that's the main 3V3 supply for the CPU
Aug 02 12:28:38 <zmatt> that 1.2V is leakage through protection diodes and such
Aug 02 12:28:45 <Laurenceb__> i should check my board, its not here atm
Aug 02 12:29:35 <zmatt> I'll measure my board in a bit, but VDD_AUX has no reason to stay at 1.2V: it only connects to the power led and (pre-A6) U4's enable
Aug 02 12:29:35 <Laurenceb__> im looking at BBB revision C schematic here
Aug 02 12:29:48 <Laurenceb__> yeah
Aug 02 12:29:54 <Laurenceb__> same on my board
Aug 02 12:30:11 <zmatt> and the LDO supplying it has 360 ohm resistor to ground when disabled
Aug 02 12:30:13 <Laurenceb__> but the way the PMIC works makes it hard to turn off aiui
Aug 02 12:30:26 <zmatt> no
Aug 02 12:30:27 <Laurenceb__> it looks like its outputting regulated 1.2v
Aug 02 12:30:29 <zmatt> no
Aug 02 12:30:30 <Laurenceb__> hmmm
Aug 02 12:30:45 <Laurenceb__> i dont know whats wrong then, but my board works ok now
Aug 02 12:31:00 <Laurenceb__> oh
Aug 02 12:31:11 <Laurenceb__> just checked a photo of my board, its rev B6
Aug 02 12:31:25 <zmatt> I'm looking at rev C schematic, it has VDD_3V3A as EN for U4, not VDD_3V3AUX
Aug 02 12:31:44 <zmatt> change log notes for A6: 1. Moved the enable for the VDD_3V3B regulator to VDD_3V3A rail.
Aug 02 12:31:55 <Laurenceb__> ah
Aug 02 12:32:02 <Laurenceb__> that sounds like a good change
Aug 02 12:32:09 <zmatt> except it isn't
Aug 02 12:32:14 <Laurenceb__> oh?
Aug 02 12:32:31 <zmatt> since now 3V3B is kept alive even when all other power is off
Aug 02 12:32:31 <Laurenceb__> what is the problem with that?
Aug 02 12:32:37 <Laurenceb__> ah
Aug 02 12:32:39 <Laurenceb__> lolz
Aug 02 12:32:46 <zmatt> due to power leaking through protection diodes back into the 3V3A rail
Aug 02 12:32:54 <zmatt> which is a sure way to violate a chip's absolute maximum ratings :P
Aug 02 12:33:00 <Laurenceb__> so it ends up having the same problem
Aug 02 12:33:05 <Laurenceb__> very interesting
Aug 02 12:33:29 <zmatt> so if my assessment is right, this is a very serious hardware bug
Aug 02 12:33:48 <Laurenceb__> yeah just a bit lol
Aug 02 12:33:55 <Laurenceb__> well i have a fix for B6 at least
Aug 02 12:34:40 <Laurenceb__> U4 has lots of hysteresis, needs <0.4v to go low (disabled) and >2v to go high (enabled)
Aug 02 12:34:49 <Laurenceb__> the R C network achieves that
Aug 02 12:35:04 <zmatt> but basically you connected it to VDD_AUX again?
Aug 02 12:35:25 <Laurenceb__> it was already on VDD_AUX as i was modding a B6 board
Aug 02 12:35:34 <zmatt> B6 is already 3V3A as EN
Aug 02 12:35:47 <zmatt> A5C was the last to have it on AUX
Aug 02 12:36:02 <Laurenceb__> oh shit
Aug 02 12:36:08 <Laurenceb__> ive confused myself totally
Aug 02 12:36:18 <Laurenceb__> lolz
Aug 02 12:36:33 <zmatt> I'll measure my A5A to verify it shuts down properly there
Aug 02 12:37:27 <Laurenceb__> maybe i have an A6 board
Aug 02 12:37:35 <Laurenceb__> its at work lol :P
Aug 02 12:37:35 <zmatt> also, wake-on-lan is definitely not available
Aug 02 12:37:43 <Laurenceb__> ok
Aug 02 12:37:51 <Laurenceb__> i did wonder if this was a feature :D
Aug 02 12:38:25 <zmatt> or well, the cpu could enter some low-power state of course, but the ethernet subsystem and most of the device would need to remain powered
Aug 02 12:38:37 <Laurenceb__> yeah... as it does :P
Aug 02 12:39:03 <zmatt> ok, let's see.. where's that U4...
Aug 02 12:39:21 <zmatt> ah there
Aug 02 12:39:22 <Laurenceb__> yeah i think i got confused, i'll make some organised notes on monday, hardware is at work
Aug 02 12:39:44 <zmatt> well it seems pretty obvious to me, I don't see any other explanation
Aug 02 12:39:57 <Laurenceb__> ok
Aug 02 12:40:02 <zmatt> especially since a disabled LDO activates a 360 ohm discharge resistor to ground
Aug 02 12:40:20 <zmatt> so no way that rail is staying at 1.2V without significant current injection from elsewhere
Aug 02 12:40:35 <Laurenceb__> i see
Aug 02 12:40:50 <Laurenceb__> so you measure 1.2v on pin 1 of U4 ?
Aug 02 12:41:17 <zmatt> no, that's what you said
Aug 02 12:41:19 <Laurenceb__> in shutdown
Aug 02 12:41:32 <zmatt> I'm about to measure my board, but I have A5A so I'm expecting a clean shutdown there
Aug 02 12:41:37 <Laurenceb__> if you shutdown what is the voltage on pin1 of U4
Aug 02 12:41:42 <Laurenceb__> ok
Aug 02 12:42:04 <Laurenceb__> i never realised there were so many board revisions
Aug 02 12:43:22 <Laurenceb__> will take a look on monday :P, R C fixed my revision, whatever it is
Aug 02 12:48:51 <zmatt> interesting... my regulator's definitely shutdown
Aug 02 12:49:12 <zmatt> but
Aug 02 12:49:48 <zmatt> my SYS_5V measures at 0.777 V
Aug 02 12:51:38 <Laurenceb__> the battery power switch hardware is probably partially on
Aug 02 12:52:03 <zmatt> I'm checking whether it's supposed to be on or off during shutdown
Aug 02 12:52:08 <Laurenceb__> my SYS_5v is always on
Aug 02 12:52:17 <zmatt> either way I'd expect it to be 0V or around 5V, not 0.78V
Aug 02 12:52:44 <zmatt> I'm pretty sure that's the way it's supposed to be
Aug 02 12:52:45 <Laurenceb__> yeah, but im guessing you dont have V_Bat shorted to V_Bat_Sense ?
Aug 02 12:53:07 <Laurenceb__> TP5 and TP6
Aug 02 12:53:22 <Laurenceb__> i shorted them
Aug 02 12:56:08 <zmatt> I'm measuring: TP5 0.777  TP6 0.3  TP7 2.24
Aug 02 12:56:52 <Laurenceb__> sounds like the power switch is behaving strangely
Aug 02 12:57:07 <Laurenceb__> ive got a lipo cell permanently installed
Aug 02 12:58:49 <zmatt> well, it seems consistent with it trying to charge the battery with SYS_5V... only question is why the fuck is it at 0.777V
Aug 02 12:58:56 <Laurenceb__> and see V_lipo on SYS_5V
Aug 02 13:03:37 <zmatt> ah wait, it's supposed to be off
Aug 02 13:03:51 <zmatt> but "Battery voltage always supplies the system (SYS pin)"
Aug 02 13:03:59 <Laurenceb__> ah
Aug 02 13:04:07 <Laurenceb__> i should read the datasheet :P
Aug 02 13:06:37 <Laurenceb__> ive also taken a wire jumper from the MUX pin on the TPS65** PMIC to AIN7
Aug 02 13:06:59 <zmatt> well it's not exactly short, but yeah if you're doing stuff with battery power I'd say you may want to read it
Aug 02 13:07:00 <Laurenceb__> so the voltages can be monitored using i2cset on /dev/i2c0
Aug 02 13:07:09 <Laurenceb__> heh :P
Aug 02 13:13:48 <zmatt> connecting TP5 and TP6 has some effect
Aug 02 13:14:33 <Laurenceb__> aiui thats a "battery detect" function
Aug 02 13:14:56 <zmatt> yeah, and not just in a binary fashion
Aug 02 13:15:33 <zmatt> sense is very sensitive indeed: just touching with a wire (other end unconnected) causes SYS_5V to change significantly
Aug 02 13:15:47 <zmatt> in the upward direction
Aug 02 13:16:02 <zmatt> connecting it to VBAT drops the voltage to 0.4V
Aug 02 13:16:03 <Laurenceb__> maybe it should have pulldown or pullup
Aug 02 13:16:18 <zmatt> so I think I'm just seeing the battery charger trying to do its thing
Aug 02 13:16:21 <Laurenceb__> on TP6
Aug 02 13:16:28 <zmatt> no definitely not
Aug 02 13:16:35 <zmatt> it needs to measure the battery voltage
Aug 02 13:16:47 <Laurenceb__> yeah i see
Aug 02 13:17:02 <zmatt> Battery voltage sense input, connect to BAT directly at the battery terminal.
Aug 02 13:17:18 <Laurenceb__> i meant for when no battery is connected
Aug 02 13:17:49 <zmatt> depends on what's desired for SYS_5V I guess
Aug 02 13:18:45 <zmatt> datasheet says for batteryless system: "The battery pins (BAT, BATSENSE, TS) are floating"
Aug 02 13:19:53 <zmatt> alternative is connecting VBAT and VSENSE to 5V supply, and 10k resistor from TS to ground
Aug 02 13:20:30 <zmatt> which has the advantage of "higher power-efficiency because the internal power-path is by-passed and power-loss across the internal switches is avoided."
Aug 02 13:20:39 <zmatt> but you lose overvoltage protection
Aug 02 13:21:21 <Laurenceb__> interesting
Aug 02 13:21:41 <Laurenceb__> im seeing lower idel current from Debian then Angstrom
Aug 02 13:21:52 <Laurenceb__> about 100mA lower from my lipo
Aug 02 13:22:00 <zmatt> ah wait, with the former case they do connect BAT to BAT_SENSE in their example diagram
Aug 02 13:22:07 <Laurenceb__> 150mA versus 250mA
Aug 02 13:23:03 <zmatt> given angstrom's goals of being specifically targeted at embedded systems versus debian primarily desktops and servers, that's pretty sad
Aug 02 13:23:33 <tbr> given that the image was NOT power optimized, I am not surprised
Aug 02 13:23:44 <zmatt> the debian image is?
Aug 02 13:23:49 <tbr> it runs a full graphical environment, etc
Aug 02 13:24:01 <zmatt> fair enough
Aug 02 13:24:09 <tbr> zmatt: no, but there is so much cruft enabled on both
Aug 02 13:24:17 <zmatt> right, so it's pretty random
Aug 02 13:24:17 <tbr> makes it impossible to comare
Aug 02 13:24:43 <zmatt> Laurenceb__: you can try running powertop maybe
Aug 02 13:24:53 <tbr> heh, same thought
Aug 02 13:25:03 <Laurenceb__> ok
Aug 02 13:26:45 <Laurenceb__> can ethernet, gpu and hdmi be disabled?
Aug 02 13:27:41 <zmatt> the am335x has fine-grained power management... question is what opinion the board and linux have on it
Aug 02 13:27:42 <Laurenceb__> im trying to assembled an embedded device that runs an ultra high fps camera on gpio signals, then grabs h.264 over usb and does some CV stuff
Aug 02 13:27:53 <Laurenceb__> right
Aug 02 13:28:00 <Laurenceb__> *assemble
Aug 02 13:28:45 <zmatt> ethernet phy can be put in powerdown mode via mdio
Aug 02 13:28:56 <tbr> you can disable the virtual capes for hdmi
Aug 02 13:28:58 <zmatt> typical 5.6 mW consumption
Aug 02 13:32:20 <Laurenceb__> the big NXP IC, its a "HDMI framer"  ?
Aug 02 13:34:47 <veremit> wow .. relying on usb for h264 .. on the am335x .. thats brave.
Aug 02 13:34:57 <veremit> you did know the driver was broken?!
Aug 02 13:36:21 <veremit> is the cpu really fast enough to do the encoding too?
Aug 02 13:36:50 <veremit> I'm using a dm368 with optimised codecs for some hd video capture
Aug 02 13:37:06 <veremit> but a different scale project I would imagine :)
Aug 02 13:37:35 <Laurenceb__> oh dear...
Aug 02 13:37:45 <Laurenceb__> i thought it was less broken than Rpi
Aug 02 13:37:59 <Laurenceb__> this isnt for live video, the camera has internal flash
Aug 02 13:38:00 <veremit> different kinda broken .. the usb is the onyl real weakpoint
Aug 02 13:38:12 <veremit> and its been well patched .. but it is patched lol
Aug 02 13:38:14 <Laurenceb__> does mass storage work?
Aug 02 13:38:25 <Laurenceb__> the camera mounts as mass storage
Aug 02 13:38:27 <veremit> I'm using mine with a usb ide hdd .. works :)
Aug 02 13:38:37 <zmatt> at least the am335x doesn't cause random data corruption on I2C if a slave dares to use clock stretching (the rpi does)
Aug 02 13:38:43 <Laurenceb__> cool, thats all i need
Aug 02 13:38:43 <veremit> omg .. never try to network share a usb hdd with a Pi. Ever.
Aug 02 13:38:50 <Laurenceb__> hehe
Aug 02 13:38:51 <tbr> Laurenceb__: so you're not doing real time?
Aug 02 13:38:54 <Laurenceb__> no
Aug 02 13:39:00 <veremit> should be ok then
Aug 02 13:39:01 <veremit> :)
Aug 02 13:39:14 <tbr> should work then
Aug 02 13:39:27 <Laurenceb__> its for monitoring machinery
Aug 02 13:39:43 <veremit> when I first started working with the beagle . the usb wouldn't work unless a device was plugged when the board was powered .. hotplug has improved
Aug 02 13:39:49 <Laurenceb__> i get a gpio signal a few seconds before the event, set the camera to record at a few k fps
Aug 02 13:40:03 <veremit> ah .. low scale control .. should be fine
Aug 02 13:40:05 <Laurenceb__> then grab the video later and save only statistics from some basic CV
Aug 02 13:40:24 <veremit> its a good cheap solution :)
Aug 02 13:40:34 <Laurenceb__> yeah :D
Aug 02 13:40:35 <veremit> although the arduino/etc is just as good for basic io with usb ;D
Aug 02 13:41:06 <veremit> there's quite a good set of low cost solutions for different purposes comeing out
Aug 02 13:41:25 <zmatt> veremit: you don't want to do any video processing on an arduino though, also not offline video processing :P
Aug 02 13:41:40 <Laurenceb__> yeah, but h.264 decode to framebuffers is not feasible on most other hardware
Aug 02 13:41:41 <veremit> nope .. not on anything less than the stm32 ones ;P
Aug 02 13:41:48 <Laurenceb__> stm32f4 would work
Aug 02 13:41:57 <veremit> the new arduinos look very cool
Aug 02 13:41:59 <Laurenceb__> but i dont fancy hacking h.264 decode
Aug 02 13:42:06 <veremit> nope .. if its decode that be fine
Aug 02 13:42:15 <zmatt> Laurenceb__: for power management I think you'll be reading lots of datasheets... the phy has powerdown mode, the HDMI framer " can be switched to very low power Standby or Sleep modes to save power when HDMI is not used", etc etc
Aug 02 13:42:24 <zmatt> but I'm guessing you will need to configure them into those modes yourself
Aug 02 13:42:29 <Laurenceb__> ok
Aug 02 13:42:33 <veremit> yea won't be an easy ride
Aug 02 13:42:42 <veremit> I was gonna say 'custom board' but that doesn't usually come easily :)
Aug 02 13:42:43 <Laurenceb__> ill check the schematic and have a think
Aug 02 13:42:48 <Laurenceb__> heh :D
Aug 02 13:42:51 <veremit> sources are all out there
Aug 02 13:43:09 <Laurenceb__> thanks for the help
Aug 02 13:43:09 <zmatt> veremit: custom board, with a chip like the am335x... have fun
Aug 02 13:43:17 <zmatt> stuff like ddr3 layout
Aug 02 13:43:33 <veremit> yeah you gotta know what ya doin .. but if you do ..
Aug 02 13:43:45 <Laurenceb__> custom with stm32f4 would be doable
Aug 02 13:43:48 <veremit> we're about to start on imx6 layout
Aug 02 13:43:51 <Laurenceb__> dunno if itd be less or more work
Aug 02 13:43:57 <veremit> based on the imx rex designs
Aug 02 13:44:10 <Laurenceb__> most of my projects are stm32f4 based
Aug 02 13:44:13 <veremit> I do like my wandboard though :)
Aug 02 13:44:19 <Laurenceb__> but video transcoding looked like a pita
Aug 02 13:44:35 <zmatt> we have a dm814x-based board... the company that usually does our electronics design had to partially outsource it because it was kinda over their heads
Aug 02 13:44:35 <veremit> you don't want to transcode on anything that doesn't have some grunt
Aug 02 13:44:47 <veremit> encoding is bad without dedicated hardware .. decode is fine
Aug 02 13:44:47 <zmatt> (it ended up a 14-layer PCB iirc)
Aug 02 13:44:55 <veremit> 14?! holy cow
Aug 02 13:45:00 <veremit> rex is only 6
Aug 02 13:45:07 <zmatt> yeah that was my reaction too
Aug 02 13:45:23 <veremit> what on earth happened there? besides someone wanted to make some money lol
Aug 02 13:45:30 <Laurenceb__> well i should have said decode then encode at jpeg only
Aug 02 13:46:10 <zmatt> how many balls is that iMX of yours?
Aug 02 13:46:15 <veremit> mjpeg or still ?
Aug 02 13:46:35 <veremit> can't remember .. project's been shelved for now whilst we do video over IP work :)
Aug 02 13:47:13 <veremit> I remember the pin descriptions run to about 4/5 pages of a4 :D
Aug 02 13:47:23 <veremit> I was comparing the quad/single cores
Aug 02 13:47:26 <veremit> they are compatible
Aug 02 13:49:15 <zmatt> http://www.imx6rex.com/download/design-files/imx6-rex-module/v1i1/docs/iMX6%20Rex%20Module%20V1I1%20PCB%20STACKUP.pdf
Aug 02 13:49:18 <zmatt> that's 12 layers
Aug 02 13:49:55 <zmatt> if I understand what I'm looking at
Aug 02 13:50:06 <zmatt> (I just clicked something, wasn't looking for this :P )
Aug 02 13:50:19 <veremit> silkscreen's don't count :P lol
Aug 02 13:52:30 <zmatt> ah about comparable number of balls
Aug 02 13:52:45 <veremit> its got enough lol .. its just above our expertise
Aug 02 13:54:39 <veremit> but manufacture is subcontract anyway ..
Aug 02 13:54:51 <veremit> most likely just lift the design and add bits :)
Aug 02 13:55:20 <zmatt> and I don't know enough about pcbs to know what you meant with "silkscreen's don't count", I also don't know how the "14" of our pcb was counted (or even if I remember it right)
Aug 02 13:56:08 <veremit> lol .. usually you refer to the 'copper' layers .. yes there are technically possibly up to 16 or so 'physical' layers , but its the ones that need photo etching and drilling that count :)
Aug 02 13:57:12 <zmatt> it looks to me they only numbered copper layers, otherwise there'd be 28 :P
Aug 02 13:57:19 <veremit> ouch :)
Aug 02 13:57:27 <veremit> fun stuff
Aug 02 13:57:41 <zmatt> (talking about the iMX6 rex pdf I linked to)
Aug 02 13:59:07 <veremit> yea I'm with ya
Aug 02 13:59:15 <veremit> I haven't looked in much detail tbh
Aug 02 13:59:53 <tbr> Laurenceb__: to briefly touch back on the custom board. look at blue-steel. it's a stripped down BBB design.
Aug 02 14:03:08 <veremit> lol thats awesome hehe .. stripped :)
Aug 02 14:04:06 <zmatt> anyhow, time to get some groceries
Aug 02 14:05:51 <veremit> I like it "you can actually get it now, which is smoething you can't say about the beaglebone black" lol
Aug 02 17:06:51 <acovrig> I have a black rev c running debian, can I boot off SD and dd the internal partition to usb (or over ethernet) as a backup, flash a different OS, then if I decide to go back, I can dd the image back?
Aug 02 17:21:36 <tbr> acovrig: I don't think it's that simple
Aug 02 17:22:04 <acovrig> that's what I'm afraid of...
Aug 02 17:22:56 <acovrig> tbr, I got a screen that I can't get working in debian, but want to keep debian; they have an angstrom image I could flash, but would want to switch back to debain and continue working on getting it working if I see it work in angstrom
Aug 02 17:25:02 <veremit> you can download a stock debian image
Aug 02 17:28:54 <tbr> acovrig: run angstrom from sd-card then
Aug 02 17:29:52 <veremit> .. or that .. ^^ :)
Aug 02 17:29:55 <Laurenceb__> veremit: BBB is widely available on ebay
Aug 02 17:30:11 <Laurenceb__> ive ordered 3 board, all shipped in Farnell/element 14 boxes from china
Aug 02 17:30:13 <veremit> eww not payin those grossly inflated prices .. besides .. I have one .. two ..
Aug 02 17:30:19 <Laurenceb__> i suspect someone is diverting their stock
Aug 02 17:30:23 <veremit> oh definitely
Aug 02 17:30:28 <veremit> easy market
Aug 02 17:30:49 <veremit> don't think the foundation got a handle on that
Aug 02 17:30:57 <Laurenceb__> if i was feeling more helpful i would let farnell know
Aug 02 17:31:04 <acovrig> tbr, I'm not sure the image I downloaded from the display manufacture is standalone bootable, or if it installs itself on boot.
Aug 02 17:31:06 <Laurenceb__>  but they should be able to work this out by themselves
Aug 02 17:31:11 <veremit> ah not el14 too?! I thought that was just cco
Aug 02 17:31:25 <veremit> farnell won't care .. they just shift boxes .. fast :D
Aug 02 17:31:38 <Laurenceb__> farnell is owned by element 14
Aug 02 17:31:46 <veremit> yes, and cpc and onecalls too
Aug 02 17:31:46 <tbr> acovrig: oh, is it a custom display cape then. those are tricky
Aug 02 17:32:32 <veremit> ok afk a bit
Aug 02 17:32:57 <acovrig> tbr, *sigh* yea... I downloaded the ~900MB zip (with a stock angstrom and their image) from their server at ~3KB/s (my connection isn't the issue, I can download at ~5MB/s)...
Aug 02 17:45:34 <tbr> acovrig: chalkboard? tigal?
Aug 02 17:47:25 <acovrig> tbr, you asking brand? waveshare (cheep on ebay from china)
Aug 02 17:49:34 <tbr> never heard
Aug 02 17:49:40 <acovrig> tbr, this is the 'guide' I am trying to follow: http://www.wvshare.com/downloads/accBoard/DVK53X/Documentations/Porting_Drivers.pdf
Aug 02 17:56:18 <tbr> acovrig: your best chance is if you can get the sources for their kernel
Aug 02 17:56:24 <tbr> and extract the patches
Aug 02 17:56:30 <tbr> or the patches themselves
Aug 02 17:56:56 <tbr> the really really ugly way is ripping out the kernel and modules out of their image and transplanting them into the debian image
Aug 02 18:06:52 <acovrig> tbr, I compiled the kernel from their source, but when I coped the uImage over and rebooted, it wouldn't boot (I don't remember the error I saw via my TTL cable)
Aug 02 18:07:21 <tbr> lovely
Aug 02 18:20:08 <mani_> Hi
Aug 02 18:20:18 <mani_> Need a help on BeagleBone Black
Aug 02 18:21:37 <mani_> my arm gcc compiler throws error for this code.
Aug 02 18:21:47 <mani_> ******************************************************************************/ /**  * \brief     This API can be used to switch from user mode to privileged mode  *            The priviledge mode will be system mode. System mode will share  *            the same resources as user mode, but with privileges.  *  * \param     None.  *  * \return    None.  *  * Note : All the access to system configuration which needs privileged access  *
Aug 02 18:31:57 <veremit> pastebin pls
Aug 02 19:28:00 <notserpe> re: imx6rex pcb stackup... I believe the SOM is 12 layer and the carrier board is 6 layer, hence the confusion.
Aug 02 19:28:52 <veremit> I don't remember seeing 12 layers in the gerber .. but I didn't look in a lot of detail :)
Aug 02 19:29:13 <veremit> could be I guess
Aug 02 19:29:29 <veremit> thats six boards
Aug 02 19:38:12 <notserpe> veremit: what do you mean by "that's six boards?", it's quite clear on their website documentation for the SOM versus DBB, there are seperate documents just for the stackup, no need to look at gerbers.
Aug 02 19:38:57 <notserpe> anyway this channel is about BB so I go back to lurking.
Aug 02 19:39:00 <veremit> I haven't looked for a few months .. I don't honestly know .. but 12 layers to me .. is 5/6 thinknesses of board :p
Aug 02 19:39:14 <veremit> sandwiched :)
Aug 02 19:40:46 <veremit> anyhow .. time to afk .. bbl
Aug 02 19:41:59 <notserpe> there is standard terminology for this stuff, and you are not using it.
Aug 02 23:45:45 <pzn> if I have a BBB with remote connection (ssh), how can I read its revision A, B or C?
Aug 02 23:46:30 <veremit> look at the label on the p8/9 header
Aug 02 23:46:39 <veremit> there is no software versioning afaik
Aug 02 23:58:31 <thurgood> fdisk could tell you if you're on a rev C as it has additional eMMC
**** ENDING LOGGING AT Sun Aug 03 03:00:00 2014