**** BEGIN LOGGING AT Fri Sep 28 02:59:57 2012
Sep 28 09:49:47 <tasslehoff> koen: sgx-pain when running 3.2.18 on my board: http://pastebin.com/5WBWbg9H. seen this before?
Sep 28 09:50:25 <koen> looks like you are trying to access a block that is turned off
Sep 28 11:05:18 <tasslehoff> huh. I'm crawling away at a louse 500MHz, when I have 720MHz available.... didn't u-boot use to tell the kernel which freq the userspace governor should use?
Sep 28 11:06:19 <ogra_> iirc uboot can only set the mpurate (highest freq)
Sep 28 11:06:31 <koen> and you shouldn't do that
Sep 28 11:06:44 <koen> since it bypassed a huge of the kernel infra
Sep 28 11:07:10 <koen> depending on the kernel version mpurate=foo will no scale the voltage before the change
Sep 28 11:17:08 <tasslehoff> so booting with mpurate=auto is the right thing to do. why is 500MHz chosen for a 720MHz omap? I've read that running at 720 MHz can shorten the lifespan, but couldn't it at least do 600MHz?
Sep 28 11:19:00 <tasslehoff> seems linux-mainline_3.2 uses 'performance', so I suspect that one runs at 720MHz.
Sep 28 11:23:12 <av500> tasslehoff: look into cpufrey folder in sysfs
Sep 28 11:23:22 <av500> there are max and min values and governors etc..
Sep 28 11:24:02 <tasslehoff> av500: I know. I just thought/feared there was a reason 500 was chosen by default.
Sep 28 11:24:17 <av500> with what governor?
Sep 28 11:24:18 <tasslehoff> since its tempting to crank it up to 720 :)
Sep 28 11:24:24 <tasslehoff> userspace
Sep 28 11:24:52 <av500> no idea about current defaults
Sep 28 11:25:14 <av500> and no idea about your needs
Sep 28 11:26:00 <koen> tasslehoff: so set it to 'ondemand' or 'performance'
Sep 28 11:27:22 <tasslehoff> koen: yeah. my ramblings really led to: no reason to not run at 720MHz? friday + musb issues makes me more incoherent than usual..
Sep 28 11:27:37 <av500> tasslehoff: do you need 720mhz?
Sep 28 11:28:26 <av500> does it blink the cursor faster?
Sep 28 11:29:25 <tasslehoff> av500: if I'm gonna run from musb, I figure it could make me run faster.
Sep 28 11:29:43 <av500> remember, you can run, but you cannot hide
Sep 28 11:29:58 <tasslehoff> at 720MHz, maybe I could stay ahead :)
Sep 28 11:30:45 <tasslehoff> koen: have you ran any 3D applications on 3.x kernels? This guy seems to have the same sorrows I do http://permalink.gmane.org/gmane.linux.embedded.yocto.meta-ti/551
Sep 28 11:31:23 <koen> I have
Sep 28 11:31:35 <koen> xbmc on 3.0.x and 3.2.x for beagle adn 3.2.x for bone
Sep 28 11:32:04 <tasslehoff> on a C3?
Sep 28 11:32:33 <koen> c5 and xM
Sep 28 11:36:33 <bradfa> I think I figured out my tps65217 woes
Sep 28 11:36:47 <av500> pray tell
Sep 28 11:37:03 <bradfa> SEQDWN simply runs the sequencer for shutdown but doesn't actually change the mode from ACTIVE to either SLEEP or OFF.
Sep 28 11:37:29 <bradfa> So, the controller is still in ACTIVE mode, just it turned off all the DC rails, so it's slightly hard for anything to write to the I2C registers to turn everything back on
Sep 28 11:37:48 <bradfa> and while in ACTIVE mode, it auto switches to what ever input is preferred and available
Sep 28 11:38:02 <bradfa> just setting the interrupt registers when things happen, like AC power insert
Sep 28 11:38:09 <bradfa> but it won't call the sequencer again
Sep 28 11:38:38 <bradfa> so it seems the only reliable way to enter OFF state is using the PWR_EN line
Sep 28 11:38:49 <bradfa> which means fun with the RTC module in am335x has to stay
Sep 28 11:38:57 <bradfa> at least on the bones...
Sep 28 12:17:21 <woglinde> tasslehoff something new about the usb-problem?
Sep 28 16:28:09 <HeadAche> I did it, I
Sep 28 16:28:15 <HeadAche> I'm so awesome
Sep 28 16:28:27 <HeadAche> Now hopefully I never have to touch a kernel again :D
Sep 28 16:35:20 <_av500_> did what?
Sep 28 17:27:57 <mranostay> _av500_: best not to ask
Sep 28 17:31:07 <mru> http://xkcd.com/763/
Sep 28 17:57:02 <_av500_> alas, my dad is on xkcd
Sep 28 17:57:43 <Crofton|work> my wife ran the computer until the hard drive failed
Sep 28 17:57:51 <Crofton|work> Now I need to save the uicken data files
Sep 28 17:58:11 <_av500_> there alsways a floppy
Sep 28 18:02:36 <Crofton|work> heh
Sep 28 18:42:38 * ds2 tosses scooby snacks around the channel
Sep 28 18:49:17 <_av500_> \o/
Sep 28 18:49:23 <_av500_> long time no snacks
Sep 28 18:54:39 <dm8tbr> *burp*
Sep 28 18:55:53 * ds2 ignites it
Sep 28 19:01:29 <_av500_> flambee
Sep 28 19:02:04 * djlewis farts and lights that
Sep 28 19:05:39 * _av500_ rage quits
Sep 28 19:06:23 <mranostay> djlewis just went too far
Sep 28 19:06:41 <ds2> is A3 bigger then a sheet of A4 or is it A5 that is bigger then A4?
Sep 28 19:18:44 <koen> ds2: a3 is 2x a4
Sep 28 19:18:49 <koen> a5 is 2x a4
Sep 28 19:18:50 <koen> etc
Sep 28 19:19:21 <koen> a0 is about half a door
Sep 28 19:19:23 <janne> a sheet of Din A X is 1m^2 / 2^x
Sep 28 19:19:35 <ds2> so smaller the A number, large the paper?
Sep 28 19:20:05 <janne> yes, starting at A0
Sep 28 19:20:26 <ds2> would a roll of paper be an A-999 or something large like that?
Sep 28 19:21:21 <ds2> thanks. A3's are not bigger then tabaloid :(
Sep 28 19:27:00 <mranostay> koen: so it i like shotgun guages :)
Sep 28 19:28:02 <janne> ds2: not really a roll, it would be still an awful large sheet with ratio of 1:2^0.5
Sep 28 19:33:03 <janne> oh, sizes larger than DIN A0 are actually specified. by using a multiplier in front of of the A0. so two sheets of A0 are 2A0
Sep 28 19:36:03 <spacecolonyone> With oebb.sh I SHOULD be on the denzil branch for meta-oe, oe-core, and meta-ti (as opposed to master), right?
Sep 28 19:51:40 <mranostay> spacecolonyone: probably
Sep 28 19:55:16 <denix> spacecolonyone: depends on the branch of oebb.sh
Sep 28 20:01:37 * spacecolonyone checks
Sep 28 20:02:32 <mdp> mranostay, or AWG
Sep 28 20:02:45 <spacecolonyone> denix: Im on master
Sep 28 20:03:26 <spacecolonyone> 877a087
Sep 28 20:03:59 <spacecolonyone> a 9/7/12 commit by koen
Sep 28 20:05:30 <spacecolonyone> I've noticed when I do oebb.sh update it is almost always pulling changes for those layers master branch, but I can't recall seeing any updates to denzil in a long time, which got me worried I was out of sync or something
Sep 28 20:57:03 <Canaimero-Marig1> fcjfvjhfjf
Sep 28 20:57:06 <Canaimero-Marig1> vhvvjvhv
Sep 28 20:57:08 <Canaimero-Marig1> bcfvhmfvhgfv
Sep 28 21:03:19 <mranostay> nice
Sep 28 21:08:14 <djlewis> but whatr are the passwords to?
Sep 28 21:49:53 <mranostay> _av500_: shouldn't you be out drinking for two weeks?
Sep 28 21:51:08 <djlewis> umm, drinking :)
Sep 28 21:53:37 <mranostay> need to go one of these years
Sep 28 22:13:03 <jkridner__> hi merosage
Sep 28 22:13:07 <merosage> hi!
Sep 28 22:13:47 <jkridner__> koen: we are pulling the Ethernet MAC address from the device fuses, right?
Sep 28 22:15:45 <jkridner__> koen: merosage is saying that u-boot saves off the MAC address somehow.
Sep 28 22:15:57 <jkridner__> onto the SD card or EEPROM or something like that...
Sep 28 22:16:07 <jkridner__> I'm not familiar with any solution like that.
Sep 28 22:16:33 * jkridner__ wonders when rcn-ee will be around.
Sep 28 22:19:30 <jkridner__> merosage: http://www.petalogix.com/resources/documentation/petalinux/userguide/Bootloaders/UBoot/UBMacStorage does not apply to the BeagleBone, since we don't provide in u-boot on BeagleBone a way to save the environment variables.
Sep 28 22:32:23 <ds2> jkridner_: someone should make it so the bone saves that to the EEPROM
Sep 28 22:37:18 <merosage> http://pastebin.com/202hz7Bd
Sep 28 22:37:45 <jkridner__> ds2: it would be handy. you gonna do it? :)
Sep 28 22:38:12 <ds2> jkridner_: would you do the upstreaming? :D
Sep 28 22:38:28 <ds2> actually, the read from EEPROM code will be implemented soon (being used in another project)
Sep 28 22:38:41 <jkridner__> well, I'd be more likely to do it if you sent me something that worked as a starting point.
Sep 28 22:39:05 <ds2> you'll probally have to port it to the AM33x code base as this is being done for the AM35x
Sep 28 22:40:03 <ds2> and it is in UB2
Sep 28 22:41:05 <jkridner__> merosage: I wonder why the network interface ID was changed from eth0 to eth1 by udev
Sep 28 22:41:52 <jkridner__> ds2: shouldn't the I2C EEPROM code be common and this be more a matter of having the I2C routines implemented?
Sep 28 22:42:11 <jkridner__> merosage: I think that udev event is likely to tell you why the Ethernet isn't working.
Sep 28 22:42:42 <ds2> jkridner_: the I2C code is there in UB2... it is just a matter of connecting the read or write part to it
Sep 28 22:43:14 <ds2> and the i2c code is universal among the AM35x stuff... donno off hand if the AM33x uses the exact same block
Sep 28 22:47:28 <merosage> thanks again jkridner__!!!!!
Sep 28 22:47:48 <jkridner__> merosage: glad we were able to find your issue and work around it, if not figure out why udev is changing the adapter name.
Sep 28 22:48:21 <jkridner__> The serial console and 'dmesg' are friends! :-)
Sep 28 22:48:28 <jkridner__> along with pastebinit. :)
Sep 28 22:48:54 <merosage> indeed
Sep 28 22:50:29 * mranostay yawns
Sep 28 22:51:29 <jkridner__> mranostay: you showing us a geiger counter at Maker Faire NY?
Sep 28 22:52:04 <ds2> and it will be going nbuts? ;)
Sep 28 22:52:08 <ds2> nuts
Sep 28 22:52:25 <mranostay> not in NYC so no :)
Sep 28 23:21:01 <merosage> jkridner:  so udev is writing a persistent MAC to /etc/udev/rules.d/70-persistent-net.rules
Sep 28 23:21:13 <merosage> I'm just going to add a line that wipes that out in /etc/rc.local
Sep 28 23:21:15 <merosage> that seems to work
Sep 28 23:23:11 <hiredman> are the ain pins mapped correctly in the srm?
Sep 28 23:23:17 <hiredman> for beaglebones
Sep 28 23:26:20 <hiredman> also, they are 0 indexed in /sys/kernel/debug/omap_mux/, but seem to be 1 indexed in /sys/devices/platform/omap/tsc
Sep 28 23:42:02 <noob_> so, what's the difference between the denzil and master branch of git://git.yoctoproject.org/meta-ti?
Sep 28 23:42:26 <noob_> and why is there git://git.yoctoproject.org/meta-ti and git://github.com/Angstrom-distribution/meta-ti
Sep 28 23:42:48 <noob_> if i'm compiling an angstrom kernel, should I be using https://github.com/Angstrom-distribution/meta-ti rather than the yoctoproject?
Sep 28 23:45:02 <noob_> I notice the angstrom meta-ti on github is a ways behind the yoctoproject repository
Sep 29 00:50:45 <denix> noob_: so if you have noticed all that, why asking questions?
Sep 29 00:51:11 <mranostay> denix: Friday for sure
Sep 29 00:51:27 <noob_> i don't understand
Sep 29 00:52:22 <denix> mranostay: where's prpplague with his "ho-ho-ho, merry freaking friday!" :)
Sep 29 00:53:01 <noob_> the readme of the github has the title "The official OpenEmbedded/Yocto BSP layer for Texas Instruments platforms."
Sep 29 00:53:20 <noob_> but, that doesn't seem to be hte case, since the official is at yoctoproject
Sep 29 00:53:29 <denix> what else does it say?
Sep 29 00:53:30 <noob_> so, is it some sort of horribly delayed mirror?
Sep 29 00:53:57 <noob_> under that, it says "it is hosted at"
Sep 29 00:54:18 <noob_> so, either the "title" of the readme is wrong
Sep 29 00:54:30 <noob_> or, there's a collection of horrible grammar.
Sep 29 00:54:44 <mranostay> denix: got a beer?
Sep 29 00:56:23 <denix> mranostay: yep
Sep 29 00:56:41 <noob_> i'm just wondering what the goal of the github is over the yocto branches
Sep 29 00:57:05 <denix> noob_: read at the very top
Sep 29 00:57:21 <noob_> "Mirror of the Texas Instruments BSP layer for OE-core" ?
Sep 29 00:58:12 <noob_> so, it is a horribly delayed "mirror", with its own collection of unrelated branches?
Sep 29 00:58:25 * mranostay raids denix's beer fridge
Sep 29 00:58:46 <noob_> if that's so, i guess the definition of mirror isn't what it used to be :-\
Sep 29 00:59:08 <noob_> and i'm still curious about what it's purpose is.
Sep 29 00:59:11 <denix> s/delayed/outdated
Sep 29 00:59:30 <denix> it
Sep 29 00:59:39 <denix> it's purpose is to confuse people :)
Sep 29 00:59:41 <spacecolonyone> Anyone had an issue with the i2c bus timing out? i2cdetect and i2cdump both timeout, and my camera driver fails with -110 (ETIMEOUT) during boot.
Sep 29 01:00:33 <noob_> so, that means it's not tailored for the angstrom distribution in any way then?
Sep 29 01:00:54 <noob_> (that assumes mirror has *any* relation to the definition)
Sep 29 01:01:25 <spacecolonyone> I've put a logic analyzer on it and when I attempt to probe the bus I see SDA and SCL go low, but no ticks on them. They just drop lot for just over 1s and then go back high.
Sep 29 01:01:40 <spacecolonyone> BBxM, btw
Sep 29 01:01:55 <mranostay> you sure you have it wired correctly?
Sep 29 01:02:12 <thurbad> which i2c bus?
Sep 29 01:02:40 <thurbad> and if not one of the preconfigured buses, have you set up the pinmux correctly?
Sep 29 01:03:25 <denix> noob_: it is koen's personal playground and scratch area - he uses it to send me pull requests for meta-ti
Sep 29 01:05:49 <spacecolonyone> i2c-2 for the cam port
Sep 29 01:06:08 <noob_> ok, so every mention of "mirror" is false? :P
Sep 29 01:06:19 <spacecolonyone> mranostay: I'm pretty sure, but I don't even see SCL toggling on a scope
Sep 29 01:08:57 <mranostay> spacecolonyone: er well that is a good sign that something is broken
Sep 29 01:09:42 <spacecolonyone> yea, I'm wondering if I've done something bad in my driver that improperly grabs the bus
Sep 29 01:11:24 <denix> noob_: the master branch is indeed an outdated mirror - it used to be current, but no more
Sep 29 01:11:55 <denix> noob_: do you expect everything on the internet to be up to date? :)
Sep 29 01:12:55 <noob_> if it's titled "mirror" then yeah :P especially considering it could be kept up to date with a cron script.
Sep 29 01:13:18 <noob_> that's why i was questioning it's purpose. someone's playground is the purpose. that's all i was wondering. :)
Sep 29 01:13:22 <mranostay> denix: http://knowyourmeme.com/photos/239112-first-day-on-the-internet-kid <--
Sep 29 01:14:45 <denix> mranostay: :)
Sep 29 01:17:49 <noob_> so what distribution do most of you guys use with the beagleboard/bone?
Sep 29 01:18:25 <denix> windows95
Sep 29 01:19:32 <spacecolonyone> I prefer me
Sep 29 01:19:41 <spacecolonyone> windows me, that is
Sep 29 01:19:48 <spacecolonyone> seriously: angstrom
Sep 29 01:20:37 <spacecolonyone> mranostay: I think I'm going to pull my custom module, reboot, and see if i2cdetect/dump work then.
Sep 29 01:21:30 <spacecolonyone> it stinks though, because the wireless won't make it upstairs into the electronics lab here
Sep 29 01:23:19 <ds2> everything just works.
Sep 29 01:28:55 <mranostay> ds2: well are out of jobs if that is the case
Sep 29 01:29:00 <mranostay> *we
Sep 29 01:29:09 <ds2> not really
Sep 29 01:29:19 <ds2> got to build the next thing
Sep 29 01:37:49 <ka6sox> ds2, what is the "next thing"?
Sep 29 01:40:29 <ds2> whatever the guy writing the check says ;)
Sep 29 01:48:47 <spacecolonyone> hmmm 1.5ms pulses every 1.006s
Sep 29 01:50:07 <ds2> someone doing gyros?
Sep 29 01:51:01 <spacecolonyone> lamb or beef?
Sep 29 01:51:18 <ds2> lamb of course
Sep 29 01:51:52 <spacecolonyone> thanks. now I want lamb.
Sep 29 02:00:30 <spacecolonyone> So as I understand i2c, if there is nothing on the bus and I attempt to probe the registers of say address 0x20: the host will issue a start condition (pull SDA & SCL low), then clock out 0x20, then fail as it won't get an ack, right?
Sep 29 02:00:46 <agmlego> Pretty much, yeah.
Sep 29 02:01:33 <spacecolonyone> I see SDA&SCL go low simultaneously, but I don't see any subsequent activity
Sep 29 02:01:52 <spacecolonyone> they just go back high
Sep 29 02:02:47 * spacecolonyone tries the validation SD card
Sep 29 02:04:41 <spacecolonyone> I see when the onboard pull-ups get enabled
Sep 29 02:09:44 <spacecolonyone> protip: make sure your logic analyzer is sampling fast enough
Sep 29 02:11:33 <agmlego> tHat too.
Sep 29 02:14:15 <spacecolonyone> I guess this just means there is yet another issue with these prototype boards: http://imgur.com/bz63W
Sep 29 02:14:33 <spacecolonyone> mine, not the BBxM :/
Sep 29 02:25:08 <spacecolonyone> later guys
**** ENDING LOGGING AT Sat Sep 29 02:59:57 2012