**** BEGIN LOGGING AT Fri Feb 24 02:59:58 2012 Feb 24 03:01:28 the exact path is /home/orlcp440/setup-scripts/sources/meta-openembedded/meta-oe/recipes-connectivity/dhcp/dhcp_4.2.0.bbappend Feb 24 03:04:08 BRockRoll: your sd is bad Feb 24 03:05:17 aholler: what's an SD? Feb 24 03:05:29 sd-card Feb 24 03:05:58 are you just guessing? I should at least see some messages about the X-Loader right? Feb 24 03:06:32 I'm not guessing. where does the x-loader come from? Feb 24 03:07:44 aholler: I assumed that some code on the board must be fixed, and able to initialize the ability to speak over SDIO and read from card Feb 24 03:08:28 and if it doesn' find a loader there it tries the serial, therefor the 60 Feb 24 03:08:56 so build a card with a proper MLO Feb 24 03:09:17 aholler: I've been trying, but this script everyone uses requires a bunch of junk be installed like kpartx Feb 24 03:09:23 aholler: trying now to do it manually Feb 24 03:13:48 try resizing the first partition or similiar. the rom-code is like a diva and doesn't like every fat mkfs produces Feb 24 03:13:59 hehe Feb 24 03:14:27 this one guide I'm following now is having me do a bunch of sector/head/cylinder stuff and the values that came by default were all wrong; hopefully this will lead to success Feb 24 03:14:43 I know you're anxiously waiting; will let you know as soon as I have something Feb 24 03:39:31 go go sync Feb 24 03:42:16 ...got 60 Feb 24 03:46:40 try the user button ? Feb 24 03:49:23 yea user just does nothing Feb 24 03:49:27 every reset prints "60" though Feb 24 03:51:13 which model ? Feb 24 03:51:25 60= xm Feb 24 03:51:31 hmm Feb 24 03:51:42 yea I'm seeing the google hits now, people having similar problems Feb 24 03:52:01 aholler set you straight Feb 24 03:52:17 sd card is not right Feb 24 03:52:50 When building Angstrom for Beaglebone I get this error ERROR: No recipes available for: recipes-connectivity/dhcp/dhcp_4.2.0.bbappend Feb 24 03:52:54 does anybody know how I can fix it? Feb 24 03:52:55 that's like the FAA declaring "the plane was messed up" at the end of a crash investigation Feb 24 03:53:25 your plane never left ground Feb 24 03:55:33 runway incidents happen Feb 24 03:56:47 but you are not alone, the diva is responsible for an uncountable amount of hours people had spend formatting sd-cards Feb 24 03:57:56 and it's a source of many myths. ;) Feb 24 03:59:05 you people are too smart for me to believe that you're all just repeating the partition/reformat steps until it works Feb 24 03:59:12 somebody's gotta know wtf's going wrong Feb 24 03:59:46 ansearch the ml, there is mail from one named vladimir whichexplains whats going on Feb 24 04:00:32 http://groups.google.com/group/beagleboard/browse_thread/thread/ae8e64e6be02baae Feb 24 04:05:06 I compiled his program and it actually prints PASS :( Feb 24 04:05:31 and the partition is marked active? Feb 24 04:06:25 and a MLO is there? Feb 24 04:06:40 dunno what active is; it's marked bootable (I used 'a' in fdisk) Feb 24 04:07:07 yep got MLO, u-boot.bin u-boot.img (copied it just to be sure) and uImage Feb 24 04:07:52 play a bit more, it's something like the beagle-school to take that hurdle ;) Feb 24 04:08:28 http://pastebin.com/6QLR2MBK Feb 24 04:08:36 any pitfalls in there? what about my start at 2048? Feb 24 04:09:16 i just started using the mkcard scripts for my cards Feb 24 04:10:04 a long time ago Feb 24 04:11:30 ok wait Feb 24 04:11:51 there was no check if open() succeed in vladimir's code Feb 24 04:12:06 without root privileges open() would fail and junk values could result in "PASS" Feb 24 04:12:21 with sudo, values that match with fdisk shows come up and I get "FAIL!" Feb 24 04:18:06 go go hexcurse Feb 24 04:18:15 start: 2048 mbr: 497953 bpb: 497953 -> PASS Feb 24 04:19:03 woohoo look at all these green lights! Feb 24 05:07:34 hi all! I have a question about beaglebone ain. I got standard build, and I don't see /sys/devices/platform/tsc/ directory to monitor AIN values in the shell. In fact, subdir "tsc" is totally missing. What am I doing wrong? Feb 24 05:24:09 anybody reading this channel? Feb 24 05:46:43 kosta: I am reading but unfortunately do not know how to help you Feb 24 05:57:53 kosta: maybe modules not loaded or other kernel config issue Feb 24 07:15:21 I'm trying to use ttyO4 and ttyO5, but it does not seem to work. Is there software configuration that must be done before using it? Feb 24 07:15:44 beaglebone A3 Feb 24 07:19:05 first thing would be checking the kernel config if they get provided. besides, "it does not seem to work" is the second most useless error description. the most useless one is "does not work"... Feb 24 07:25:02 Thanks Russ. Connecting to the bone using tx=9,13, rx=9,11 with known good serial system at 3.3V. No input. Using vanilla angstrom image. Feb 24 07:25:13 hey, no problem! Feb 24 07:25:32 wait, who are you? Feb 24 07:25:40 hrhrhr Feb 24 07:25:54 as i said, check kernel config and possibly pinmux Feb 24 07:26:11 ah, sorry Russ, got you confused with LetoThe2nd Feb 24 07:26:17 thanks LetoThe2nd Feb 24 07:26:27 Russ: hey, you also look like a sandworm? Feb 24 07:26:37 ? Feb 24 07:27:01 ah, atreides Feb 24 07:29:59 Leto, device is in /dev/ttyO4. Shows up in /proc/tty/driver/OMAP-SERIAL Feb 24 07:30:52 just doesn't seem to allow input. Disabled flow control. Seems to have received one byte, according to OMAP-SERIAL Feb 24 07:31:19 i'd still vote for checking pinmux Feb 24 07:31:57 Thanks Feb 24 07:54:41 i tried booting my bealgeboard xm using the latest 2012 angstrom demo images..but it gives an error saying that it cannot read uEnv.txt. Then it says that it cannot get the Kernel image Feb 24 07:57:27 Jayneil: how did you set it up Feb 24 07:57:31 do you have a boot partition? Feb 24 07:59:00 yes Feb 24 07:59:08 one boot partition and one angstrom Feb 24 07:59:23 Here is the output I get from minicom when i boot my board Feb 24 07:59:35 http://pastebin.com/7e2u4cqa Feb 24 08:11:50 Leto, you were right, mux was set incorrectly. thanks Feb 24 08:13:26 anyone have any ideas why my beagleboard is not able to find the kernel image..?? I even tried pushing the user button, but even that did not solve the problem.. Feb 24 08:13:35 Jayneil: lookd like it tries to boot the kernel from the ext2 partition Feb 24 08:13:39 looks Feb 24 08:13:44 in /boo/uimage Feb 24 08:16:48 i see. it should actually look in the fat partition . so what should i do know..? do i have to create a uEnv.txt file? Feb 24 08:18:41 or place the kernel in /boot/ on the ext2 Feb 24 08:18:59 ok..will try it now Feb 24 08:26:27 Kernel panic - not syncing: No init found. Try passing init= option to kernel. Feb 24 08:26:33 got this error Feb 24 08:27:26 then i'd check the bootargs, and eventually look out for the rootwait arg (or howsitcalled nowadays) Feb 24 08:28:17 + make sure that there's an init waiting there :) Feb 24 10:08:45 sweetness Feb 24 10:08:58 2 weeks of holiday, starting now Feb 24 10:09:09 have fun! Feb 24 10:09:28 this 2 weeks notice thing is working out to my advantage Feb 24 10:10:10 you mean, its over now? Feb 24 10:10:15 *now* Feb 24 10:10:24 in 3 hours and 50 minutes Feb 24 10:10:28 \o/ Feb 24 10:10:36 * koen can't count Feb 24 10:10:39 5 hours 50 Feb 24 10:10:46 today is my last day at TI Feb 24 10:10:49 i was about to say :) Feb 24 10:10:58 koen: make sure to rm -rf all your work Feb 24 10:11:11 but make a copy 1st Feb 24 10:11:25 hrhrhr Feb 24 10:11:55 as much as I want to, I'm not going to copy the pvr source tree Feb 24 10:12:10 the code makes my eyes bleed anyway Feb 24 10:12:12 no need, I have it Feb 24 10:12:49 * koen hands av500 a wipe Feb 24 10:12:57 I think if one crosses the pvr source code with flash lite source code, a black hole appears instantly Feb 24 10:13:45 av500: hehe - and out comes singularity Feb 24 10:13:54 * koen finds a huge box and starts packing up TI gear Feb 24 10:14:00 no, pandemonium Feb 24 10:14:09 av500: http://en.wikipedia.org/wiki/Singularity_(operating_system) Feb 24 10:14:31 av500: oh, you mean http://en.wikipedia.org/wiki/Pandemonium_(Pet_Shop_Boys_album) Feb 24 10:17:40 "Elvis has left the Silo" Feb 24 10:18:58 he's still alive? Feb 24 10:19:29 according to bruce campbell, not anymore. Feb 24 10:19:53 "OElvis has left the Silo" Feb 24 10:20:17 * koen makes a note to slip 'silOE' into an email somewhere Feb 24 11:04:48 silOE like silly OE or for what does that stand for? Feb 24 11:05:18 just get gentoo ;) Feb 24 11:07:41 gentoe Feb 24 11:17:21 http://www.youtube.com/watch?v=_kgPlsczhZg Feb 24 11:18:29 omg, while playing an mp3 file - with all the 10mhz that needs.... Feb 24 11:19:08 * av500 could drive a bike and listen to his walkman at the same time - in the 80s..... Feb 24 11:19:21 with single-brain Feb 24 11:19:34 and nobody mentioned the time need to unlock the device(s), thus making them usable Feb 24 11:20:32 the walkman? Feb 24 11:20:42 I still have one ;) Feb 24 11:20:52 rooted? Feb 24 11:21:27 yes, can play every tape Feb 24 11:22:25 amazing technic Feb 24 11:22:53 it even has a button to record things. Feb 24 11:23:16 wow Feb 24 11:23:32 a pirate device Feb 24 11:24:02 I'm awaiting letters from some lawyers Feb 24 11:25:50 so the dark side got me long time ago Feb 24 11:30:23 I should spend this to some museum, "this were the (dark) times where you could made private copies to hear music when and where you wanted without getting known by some company" Feb 24 11:31:14 in fact, you paid a few cents per blank tape for that Feb 24 11:31:50 yes, now i'm paying more and get less Feb 24 11:32:43 hey i've got a beaglebone rev4 and my ethernet doesn't work. i remove the r219 but with my debian demo the ethernet dosen't work either what can i do now Feb 24 11:33:04 write a better problem description Feb 24 11:34:10 and/or ask in #debian Feb 24 11:35:32 so i've got a beaglebone rev4 with a running debian distribution. now i can not connect via ethernet to the beaglebone Feb 24 11:36:28 i searched in the interent and i found a thread that i should remove the R2190 resistor. i remove it. but the ethernet port dosen't work Feb 24 11:37:10 aholler: maybe there is a sw solution Feb 24 11:37:15 or hardware Feb 24 11:37:27 you've tried using a cable? Feb 24 11:37:54 usb cable? Feb 24 11:38:15 preferable an ethernet cable Feb 24 11:38:39 some IPs are needed too and ... Feb 24 11:39:56 might become hard to get the old ones, they've become rare Feb 24 11:40:11 yes i've tried using a cable Feb 24 11:41:11 when i'm using the ifconfig command i got only this lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:8 errors:0 dropped:0 overruns:0 frame:0 TX packets:8 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:560 (560.0 B) TX bytes:560 Feb 24 11:41:42 ifup -v eth0 Feb 24 11:41:44 but in my /etc/netstart/interfaces i enable a static ip adress Feb 24 11:43:21 Configuring interface eth0=eth0 (inet) run-parts --verbose /etc/network/if-pre-up.d ifconfig eth0 192.168.101.45 netmask 255.255.255.0 broadcast 192.168.101.255 up SIOCSIFADDR: No such device eth0: ERROR while getting interface flags: No such device SIOCSIFNETMASK: No such device SIOCSIFBRDADDR: No such device eth0: ERROR while getting interface flags: No such device eth0: ERROR while getting interface flags: No such d Feb 24 11:44:10 dmesg | grep eth0 Feb 24 11:44:34 [ 4.697632] udev[196]: renamed network interface eth0 to eth2 Feb 24 11:44:36 and please use pastebin Feb 24 11:44:57 there you have it Feb 24 11:47:00 mv /etc/udev/rules.d/70-persistent-net.rules /root Feb 24 11:47:24 mattzz: pastebin please Feb 24 11:47:36 that rule doesn't make much sense on a device which always uses a new mac Feb 24 11:47:41 av500: ? Feb 24 11:49:48 av500: my understanding is that pasting multiple lines in a chat is discouraged, no? Feb 24 11:50:30 indeed Feb 24 11:50:33 so use pastebin Feb 24 11:50:38 that's why he said 'pastebin' Feb 24 11:53:04 I'm lost ;) Feb 24 11:53:19 there is always google Feb 24 11:53:27 #1 irc rule, google it Feb 24 11:53:34 s/irc/life/ Feb 24 11:54:32 I'm lost regarding your comments. Feb 24 11:54:40 it's friday Feb 24 11:54:45 # This file was automatically generated by the /lib/udev/write_net_rules # program, run by the persistent-net-generator.rules rules file. # # You can modify it, as long as you keep each rule on a single # line, and change only the value of the NAME= key. # Unknown net device (/devices/platform/cpsw.0/net/eth0) (cpsw) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="d4:94:a1:39:10:44", ATTR{dev_id}=="0x0", ATTR{type}= Feb 24 11:54:50 ;) Feb 24 11:54:56 http://www.extremetech.com/computing/119571-mozilla-partners-up-with-lg-to-combat-apple-and-google-with-its-own-device Feb 24 11:55:04 LG of all companies Feb 24 11:55:23 kate123: use pastebin Feb 24 11:56:10 err ... is there a reason why I can fopen/fprintf /sys/class/gpio/gpiox/direction and not .../value ? Feb 24 11:59:36 http://pastebin.com/zp7SsPGW Feb 24 11:59:47 sorry i forgot Feb 24 12:01:17 FRIDAY! Feb 24 12:01:25 \\oo// Feb 24 12:01:28 or as Crofton|work would experience it: FIRday! Feb 24 12:01:58 IIRday never ends? Feb 24 12:03:11 kate123: you can either modify the rules file regarding eth0/eth2 or /etc/network/interfaces Feb 24 12:03:25 kate123: that stuff basically renames your ethernet device on every boot because you allways get a different mac. find out how to always use the same mac or how to disable that mechanism Feb 24 12:04:34 heh Feb 24 12:08:40 maybe just "mv /lib/udev/write_net_rules /root" helps Feb 24 12:15:24 av500 that looks more like a misionary position than a hooray :D Feb 24 12:15:47 its motion blur... Feb 24 12:16:52 :) Feb 24 12:20:17 mattzz: thanks a lot i made a stupid mistake Feb 24 12:20:54 i configure debian with another beaglebone so its a different mac Feb 24 12:21:44 embarrassing Feb 24 12:23:03 np ;) Feb 24 12:35:47 Crofton|work: vhost on emerald is active, you should be able to point www. to there now Feb 24 12:37:05 in any language I can use the GPIO but in C for some reason I can't open /sys/class/gpio/gpiox/value in a proper manner Feb 24 12:37:20 using fopen buffers the changes to the file until the end of the program Feb 24 12:37:24 how do you open it in another language? Feb 24 12:37:33 I just write to it Feb 24 12:37:41 echo ... > ... Feb 24 12:37:54 get rid of the f Feb 24 12:38:00 yes Feb 24 12:38:16 you mean the open syscall ? Feb 24 12:38:36 fopen != open Feb 24 12:38:57 doesn't work either, O_SYNC fails the same (bufferized) and O_DIRECT makes the open call fail directly Feb 24 12:41:15 the first one guessing your open call will win a rma for his crystal ball Feb 24 12:42:33 fd = open("/sys/class/gpio/gpio70/value", O_WRONLY | O_CREAT | O_DIRECT, 0666); Feb 24 12:42:37 fails at call Feb 24 12:42:44 create? Feb 24 12:42:51 direct? Feb 24 12:43:25 open(foo, O_WRONLY) Feb 24 12:43:39 create is in every call that works (I used strace) and direct because otherwise it's buffered Feb 24 12:44:03 well, the close it Feb 24 12:44:08 it will be written Feb 24 12:44:26 open(foo, O_WRONLY); write(...); close(); Feb 24 12:44:28 done Feb 24 12:44:36 yay suboptimality :) Feb 24 12:44:52 ? Feb 24 12:45:19 I do that in a loop, so lots of unnecesssary calls Feb 24 12:46:13 given that you are talking to a virtual filesystem, I'd dare say why care? Feb 24 12:47:22 but yes, optimize early Feb 24 12:49:34 checking the rc of open might help sometimes too ;) Feb 24 12:52:36 but av500 usually leaves that out as exercise for the reader ;) Feb 24 12:54:45 av500: I confirm it works if you close the file each time ;) Feb 24 12:55:16 aholler: what do you mean by "rc" ? Feb 24 12:55:19 welcome to sysfs :) Feb 24 12:55:28 return code Feb 24 12:55:29 return code Feb 24 12:55:35 strerror(errno) Feb 24 12:55:48 ah you mean the nice errno code Feb 24 12:55:53 stuff like that Feb 24 12:56:12 yeah I have exit(errno) in case I read <0 or NULL Feb 24 13:26:02 I found better than fopen fclose each loop, I can fflush the file which buffers Feb 24 13:26:13 I assume it's a tad faster Feb 24 13:27:09 Hi guys, I guess 'm not first who's asking; anyone know what happen to angstrom-distribution.org server? Feb 24 13:27:33 holidays Feb 24 13:27:45 servers on holidays ? :D Feb 24 13:28:03 everyone and thing needs freetime ;) Feb 24 13:29:45 got it, any info about when its back from holiday? Feb 24 13:31:01 you could always use the cache from google Feb 24 13:31:54 or http://beagleboard.org/angstrom Feb 24 13:32:52 I've just got mine beaglebone, and would like to connect it with MMA7260 accelerometer, the tsc/ainN is good but its more for reading temperature or so, but for accelero I would like to have a stream of data? anyone can advice, please. Feb 24 13:33:59 if your accelerometer is analog and connected to ainx then you can only poll at the desired frequency Feb 24 13:34:51 I'd go for a spi or i2c accelerometer Feb 24 13:35:01 the ADC drivers on the bone are ehm Feb 24 13:35:03 ehm Feb 24 13:35:09 "unfinished" Feb 24 13:35:37 great news, koen, are they going to more like a stream? Feb 24 13:36:10 yet get an event device, so "yes" Feb 24 13:36:24 and most of them have an interupt line to signal drops and taps Feb 24 13:36:42 option b is to hack the tsadc driver Feb 24 13:37:38 yah, that's the reason I was looking for angstrom server ...till you suggested the mirror Feb 24 13:39:10 ... which is kind a dead, too or I can't get in. Feb 24 13:41:33 thx a lot koen,spow ... for the valuable info, will get to it next week, got to go now.... Feb 24 13:42:05 first time i'm being thanked here :P Feb 24 13:59:47 koen, when will you finish the codezz? Feb 24 14:00:00 never Feb 24 14:02:07 it's like me working on the adafruit driver in my copious spare time Feb 24 14:02:21 as I keep stacking other distractions in front of it Feb 24 14:03:06 mdp: do you have a .config for your st7735fb tree on the githubzz? I am struggling to build the kernel as I am getting undefined references Feb 24 14:03:36 mattzz, the omap2plus_defconfig is updated in that branch Feb 24 14:03:46 did you pick the st7735fb branch? Feb 24 14:03:52 mdp: yep Feb 24 14:04:25 let me surf over there and take a look to refresh my memory…I kinda left stuff hanging in november :) Feb 24 14:04:49 mdp: I will check for the omap2plus, thanks! Feb 24 14:07:54 well, I lied Feb 24 14:07:57 I do that a lot Feb 24 14:08:06 mattzz, apparently I didn't push a happy defconfig Feb 24 14:08:42 I'll push that in a few Feb 24 14:08:54 mdp: cool, thanks a lot. Feb 24 14:09:28 mattzz, if it helps at all, I intend to rebase this support to vaibhav's github beaglebone tree that he announced on the list..then continue cleanup from there Feb 24 14:09:42 but won't have time until next weekish Feb 24 14:10:45 koen, that reminds me…at one time, there were pmic changes that would smoke an old A1 board…do you know if current trees are A1 safe? Feb 24 14:10:49 thats great news. I just want to learn about kernel modules development so I think your tree is still a good starting point for playing with the adafruit display Feb 24 14:11:34 mattzz, it's at least known working if totally out of date, I'll fix that Feb 24 14:12:46 mattzz, you might also look at kkeller's post on the list…he did some neat stuff with driving one of those nokia displays from userspace…obviously not performance oriented like a kernel driver, but very cool from javascript Feb 24 14:13:33 mdp: no idea, my mdp: Thx, I will check it out. Were you referring to https://github.com/hvaibhav/am335x-linux? Feb 24 14:15:54 no, ohporter Feb 24 14:16:30 ah, ohporter will be rebased to vaibhav? Feb 24 14:16:37 koen, oh, well, I have no eeprom programmed on the A1s, none of that ever existed Feb 24 14:16:46 mdp: then it should be ok Feb 24 14:17:32 mattzz, yeah, I'll rebase the st7735fb related commits to hvaibhav's tree RSN Feb 24 14:18:01 got it, thx. Gotta get more used to git lingo ;) And to git... Feb 24 14:18:03 that's the canonical best working most feature complete, closest to mainline tree Feb 24 14:18:07 mdp: with my new found spare time the next weeks I'll work on a cape design for 2x st7735fb Feb 24 14:18:12 for bone, that is Feb 24 14:18:18 koen, nice! :) Feb 24 14:18:35 sweet. Feb 24 14:18:46 make sure, you make the codez to have the two additional bitbanged uSD slots work too :P Feb 24 14:19:00 koen, though that's a freebie as we know Feb 24 14:19:18 actually, it called "teh codez" Feb 24 14:19:23 everybody will want 3 functional uSD slots Feb 24 14:19:25 oops, sorry Feb 24 14:19:37 mdp: I'm going to make a cape that uses the bare lcds: https://www.adafruit.com/products/618 Feb 24 14:19:39 I'm so old I fsck that l88t speak up Feb 24 14:19:54 hehe Feb 24 14:19:56 and both on one spi port Feb 24 14:20:12 koen, oh yeah…I assume you'll just hot bar solder the fpc then? Feb 24 14:20:33 assuming you try this on a proto cape first... Feb 24 14:20:42 best to skip the slow 4050 Feb 24 14:21:17 mdp: I'm going to leverage my new employer Feb 24 14:21:19 koen, keep in mind that I set up the driver generic enough to handle any configuration..it's actually partially ready even for a completely different lcd hooked to that controller Feb 24 14:21:24 but not quite there yet Feb 24 14:22:32 your cape will only different in number of instances though Feb 24 14:22:47 koen, gotcha..that's going to be nice Feb 24 14:24:00 koen, is anybody doing an xbee cape? I was thinking about bone as an aggregation point for some remote sensors Feb 24 14:24:17 I've only heard rumours Feb 24 14:24:25 pretty simple one, of course Feb 24 14:24:34 though there's usb stuff too Feb 24 14:24:44 * koen checks the official cape registrar Feb 24 14:24:47 so maybe pointless for my purposes come to think of it Feb 24 14:24:52 there's a cape registrar? Feb 24 14:25:12 yeah, boris houndleroy runs it Feb 24 14:27:51 * av500 proposes a jumpsuit cape to store unused jumpers Feb 24 14:30:34 not bad, it's like the area in vehicle fuses boxes to store spare fuses Feb 24 14:30:49 av500, you are bubbling with great ideas Feb 24 14:31:39 maybe there would be room for a first aid kit on that cape too Feb 24 14:31:58 I'd like on with a lib where I can stroe small parts and screws Feb 24 14:32:01 lid Feb 24 14:33:11 also a cape with a 8x1 char LCD that shows the current git hash of the kernel Feb 24 14:33:17 hehe Feb 24 14:33:36 1.x8 in fact Feb 24 14:33:37 powered by bonescript Feb 24 14:33:40 1x8 Feb 24 14:34:45 I was thinking of a doomsday clock cape, counting down the time on a nice OLED display until we run out of time to implement capes later this year Feb 24 14:35:17 and were all busy shooting zombies or however this will go down Feb 24 14:35:49 I thought we treck to the himalayas? Feb 24 14:35:59 to catch a crouse ship... Feb 24 14:36:03 criuse even Feb 24 14:36:06 cruise Feb 24 14:36:08 i give up Feb 24 14:36:14 spellink is not for me Feb 24 14:38:05 np, I'm not much into reedin an rytin myself Feb 24 14:39:22 mdp: would you expect the 7735 module to work with a more recent tree or would you expect problems due to the missing DMA safeness? Feb 24 14:42:48 I think it will fail since I heard that koen/vaibhav's tree added working edma support to mcspi Feb 24 14:43:16 extra dma? Feb 24 14:43:17 * av500 wonders how working with Edna is... Feb 24 14:43:23 as opposed to silent dma? Feb 24 14:43:29 koen: super dma Feb 24 14:43:35 correction: it *will* fail if dma is engaged..how's that for confidence? Feb 24 14:43:54 anyway, minor issues I'll address when I work on it next week Feb 24 14:44:02 thx Feb 24 14:45:30 unfortunately, some work distractions chewed into my hacking time over the last 3 months Feb 24 14:45:48 pesky distractions…. Feb 24 14:45:50 mdp: you have to stand strong Feb 24 14:45:57 dont let work distract you Feb 24 14:46:14 av500, it does suck…work tends to ruin the important coding tasks Feb 24 14:46:39 * koen reads https://github.com/blog/1056-blob-contributions-box Feb 24 14:46:39 if I could be strong enough to limit their time to 8.0 hours per day... Feb 24 14:46:45 instant git blame! Feb 24 14:48:21 jkridner___: can you fix up the index.html on the s3 mirrors? Feb 24 14:48:38 I'll give it a shot. Feb 24 14:49:00 "coundn't connect to host" Feb 24 14:51:45 ho ho ho Feb 24 14:51:49 merry freakin friday! Feb 24 14:52:06 * prpplague reaches into his red bag to hand out gifts of prozac and rum Feb 24 14:52:23 one each please Feb 24 14:53:01 jkridner___: right, it's down, that;s why we need the mirrors Feb 24 14:53:03 prpplague: heh Feb 24 14:53:13 prpplague: two please, I have a friend that needs those gifts too Feb 24 14:53:21 not much to be able to do about that retractively. Feb 24 14:53:51 * prpplague gives out the rum and prozac with dobles to mdp Feb 24 14:53:53 jkridner___: for i * ; echo > index.html ; done? Feb 24 14:54:10 * prpplague throws yocto thumbs drives at jkridner___ since jkridner___ has been bad this week Feb 24 14:54:34 jkridner___ has been working on world peace Feb 24 14:54:38 ehm, device tree Feb 24 14:54:49 hard to see the difference with all that DT propaganda Feb 24 14:55:10 * prpplague is totally frustrated with DT Feb 24 14:55:38 they should have done mv arch/arm arch/arm-old and repopulted arch/arm with converted drivers Feb 24 14:56:00 I thought it was about converting developers? Feb 24 14:56:01 * wmat patiently awaits PDFs of ELC presentations ;) Feb 24 14:56:11 prpplague: DT didn't bring you world peas? Feb 24 14:56:12 wmat: javascript! Feb 24 14:56:28 mdp: it looks like your cable mishap broke video and audio Feb 24 14:56:41 koen: eh? Feb 24 14:56:49 mdp: so I only get 13 minutes of awesomeness in that elc vid Feb 24 14:56:53 koen, bummer Feb 24 14:56:54 wmat: my presentation was in js :) Feb 24 14:57:18 mdp: right now i would characterize it somewhat like using a jalapeno as a suppository Feb 24 14:57:23 koen, I will send a nastygram to LF, "you #$!@@#!$@#, I only got my 13 minutes of fame!!!" Feb 24 14:57:45 hehe Feb 24 14:58:09 koen: you have to be different eh? No matter, tar it up and send it and I'll post it as an archive with a note that it's eveil. Feb 24 14:58:17 evil even Feb 24 14:58:27 koen, are you saying that during my time in the champagne room, I cause so much damage that later visitors to the champagne room could not fully take part in what it had to offer? Feb 24 14:58:31 wmat: it's up at the LF site already Feb 24 14:58:49 mdp: I'm saying there's no video evidence of what went on Feb 24 14:58:54 funny i can't get any of the videos to play behind TI's proxy Feb 24 14:59:01 koen, excellent! my job is safe Feb 24 14:59:23 koen: ah, i see Feb 24 14:59:26 prpplague: I will fedex a usb stick with the videos to your desk Feb 24 14:59:35 hehe Feb 24 15:00:51 * prpplague does not understand why some people think ICS is a good environment for doing hardware validation and bringup Feb 24 15:01:24 prpplague: android 6.0 will be called 'Kool-Aid' Feb 24 15:01:35 prpplague: just had that conversation with a friend that's doing some contracting over in your world Feb 24 15:01:58 koen: do we all have to gather at a camp in south america in order to install it? Feb 24 15:02:04 prpplague: somebody originally suggested he use ICS for some upstream scheduler work and testing..hehe Feb 24 15:02:10 mdp: do i need to deliver some some rum and prozac? Feb 24 15:02:16 mdp: hehe Feb 24 15:02:28 wmat: no, mountain view Feb 24 15:02:49 you slide over the briefcase with rum and prozac and my people will put the usb sticks of ELC videos in your trunk Feb 24 15:02:51 same difference Feb 24 15:02:59 it's all south of canada ;) Feb 24 15:03:08 wmat: hehe Feb 24 15:04:07 prpplague: because android is awesome Feb 24 15:04:16 and iOS5 is even harder to use for bringup Feb 24 15:04:35 wmat: whats that saying... something like "America thinks of Canada as their eccentric brother to the north, but Canada thinks of America as their mentally handicapped cousin to the south" Feb 24 15:05:06 Canadians think the USians are related to them? Feb 24 15:05:16 I use OSX when I bringing up new x86 h/w Feb 24 15:05:59 prpplague: sounds about right Feb 24 15:06:56 prpplague: except that we realize that cousin can beat us up whenever they please Feb 24 15:06:57 prpplague: going to try getting a visit to my new HQ soon Feb 24 15:07:13 koen: where? Feb 24 15:07:27 koen: dandy Feb 24 15:07:35 wmat: dallas, TX Feb 24 15:08:41 koen: seems like RANTday: http://lists.linaro.org/pipermail/linaro-dev/2010-December/001939.html Feb 24 15:13:26 that's a goodie Feb 24 15:13:58 temporary relief: http://www.beagleboard.org/angstrom-mirror/www.angstrom-distribution.org/demo/beagleboard/ Feb 24 15:14:46 if anyeone knows the magic with apache to make the paths to CSS/JS/etc. work, I can fix http://www.beagleboard.org/angstrom-mirror/www.angstrom-distribution.org/ Feb 24 15:14:56 prpplague: why are you picking on me? Feb 24 15:17:29 http://lists.linaro.org/pipermail/linaro-dev/2010-December/001935.html Feb 24 15:17:45 "It appears that this discussion ended up on phoronix.com [1]...." Feb 24 15:17:50 of course it did Feb 24 15:18:03 http://beagleboard.org/angstrom/ Feb 24 15:21:52 "Can't we all just be happy that we actually have 3D drivers?" "Can't you just use Windows?" Feb 24 15:21:59 hehe Feb 24 15:22:37 which 3d drivers? Feb 24 15:23:11 arm gpu Feb 24 15:23:15 whichever Feb 24 15:23:19 the mali-stuff? Feb 24 15:23:20 since all are closed Feb 24 15:23:23 all of them Feb 24 15:23:29 arm soc Feb 24 15:23:30 16:08:38 < av500> koen: seems like RANTday: http://lists.linaro.org/pipermail/linaro-dev/2010-December/001939.html Feb 24 15:23:34 it's from this thread Feb 24 15:23:48 http://beagleboard.org/angstrom/ is updated again. Feb 24 15:24:40 ynezz: your time is wrong Feb 24 15:25:01 only 3s off Feb 24 15:25:02 ah, no, sorry Feb 24 15:25:36 got confused, but those two mails just ranting generaly. Feb 24 15:25:56 oh no! 3 seconds! alarm! Feb 24 15:26:11 what should I do? Feb 24 15:26:19 run Feb 24 15:26:29 home, good idea Feb 24 15:27:38 koen: note http://beagleboard.org/angstrom Feb 24 15:28:17 anyone know if eglibc has NEON versions of sin/cos Feb 24 15:28:30 funny, someone from mcafee had a talk about m68k at the elc. Feb 24 15:28:50 jkridner___: thanks Feb 24 15:29:12 did the even exist when m68k was used last? ;) Feb 24 15:29:27 heee Feb 24 15:30:19 What was the M68K based QUICC processor Feb 24 15:30:19 that thing lingered on way longer than the days of the actualy m68k Feb 24 15:30:30 In the days before ARM (gasp, it's horrifying considering that), I did a lot of stuff with MPC850 and MPC860 Feb 24 15:30:44 * SilicaGel stabs windows in the crank Feb 24 15:31:50 Is that too american-colloquial? Feb 24 15:32:12 * SilicaGel stabs windows in the johnson Feb 24 15:34:29 maybe we all have already got infected by some undisclosed evil arm trojan, eating sources of graphic-drivers Feb 24 15:35:12 you mean ics? Feb 24 15:35:27 I Crush Source? Feb 24 15:35:31 and mcafee comes too help after they finished their m68k-project Feb 24 15:35:42 wait what are you saying bad abuot ICS Feb 24 15:36:19 I got in on one of those $99 HP Touchpads, and I put ice cream sandwich on it, and I am in love with it Feb 24 15:37:08 koen: I was able to get on the beagle list, so that means my patch was able to make it out Feb 24 15:37:59 jkridner___: koen: see the ML, somebody installed ubuntu because angstrom.org was down!!!!!!!!!!!! Feb 24 15:38:06 jwinnebeck: I have it applied locally already Feb 24 15:38:17 av500: and then asked how to use OE to build stuff for it :) Feb 24 15:38:20 on top of ics? Feb 24 15:38:28 koen: OK so I patched the proper tree then? Feb 24 15:38:39 I was afraid after your comments yesterday I started from the wrong place Feb 24 15:38:53 I also have a patch for the meta-ti layer Feb 24 15:39:09 because that's originally how I did it. I don't know if that is useful Feb 24 15:39:28 I needed it because I want to use bitbake systemd-image to get an image Feb 24 15:39:39 that's actually the proper way Feb 24 15:39:40 and your repo isn't used in that... I don't quite understand what the plan is long-term Feb 24 15:39:58 read my mail to the beagle list Feb 24 15:40:00 so should I post the patch for the meta-texasinstruments layer? Feb 24 15:40:07 Ok I will read it again Feb 24 15:40:11 my repo is only there for people to dumb to use 'patch' or 'git am' Feb 24 15:40:30 or too lazy Feb 24 15:40:39 "The long term goal is to manage everything with devicetree, but till that actually works on beaglebone use the cape method in the repo mentioned above." Feb 24 15:40:41 * dwery loves lazyness Feb 24 15:40:43 I'm not sure what "devicetree" is Feb 24 15:40:49 patch is evil, it eats comments Feb 24 15:40:56 jwinnebeck: have you seen starwars? Feb 24 15:41:01 yes? Feb 24 15:41:06 jwinnebeck: with the jedi hand wave? Feb 24 15:41:08 yes Feb 24 15:41:11 more importantly, has he seen that it is not as good as star trek? Feb 24 15:41:19 imagine that every time someone mentions devicetree Feb 24 15:41:26 ok Feb 24 15:41:42 SilicaGel, MC68360 Feb 24 15:41:49 it some magical fairyland solutions that will take some time to actually work Feb 24 15:41:50 and MC68302 before that Feb 24 15:41:51 panto: *snap* that's it! Feb 24 15:41:59 that didn't seem to help me understand anything except to say that I'm not supposed to understand it, because devicetree isn't ther kernel I'm looking for :) Feb 24 15:42:04 302 yes. 302 was the first one I remember... Feb 24 15:42:09 * panto had put alot of time on those Feb 24 15:42:24 OK, so then koen but your kernel source isn't used in bitbake layer, so who is using this? Feb 24 15:42:34 Shouldn't I also post the patch to the bitbake part? Feb 24 15:42:37 extra bonus points if your remember what the differences between MC68302 & MC68QH302 was :) Feb 24 15:42:37 jwinnebeck: lazy and/or dumb people Feb 24 15:42:48 jwinnebeck: DT is ACPI Feb 24 15:42:56 jwinnebeck: that repo is just the git version of the recipe Feb 24 15:43:03 av500, wut? Feb 24 15:43:14 Yeah I understand, it's the kernel with the patches applied Feb 24 15:43:18 which is good Feb 24 15:43:30 But that's not what you use with bitbake systemd-image... Feb 24 15:43:38 So I need to add the patch to the meta-ti layer Feb 24 15:43:51 right Feb 24 15:43:52 so that everyone can use it, not just the lazy/dumb people as you say Feb 24 15:43:58 OK then I should post that to beagle list too? Feb 24 15:44:05 if you want Feb 24 15:44:17 ok I will do that now, it wasn't clear to me. That was actually the first commit I produced Feb 24 15:44:21 the one for meta-ti Feb 24 15:44:45 why not shorten meta-ti to meti? Feb 24 15:45:04 meti@home Feb 24 15:45:07 yetu! Feb 24 15:45:47 klendathu! Feb 24 15:46:30 klatu Feb 24 15:47:33 cthulu Feb 24 15:48:16 must be really evil that virus Feb 24 15:48:54 * emeb waits for news of a good virus Feb 24 15:49:09 stuxnet Feb 24 15:49:28 depends on your perspective I guess Feb 24 15:59:16 av500: that's "klaatu". :-) Feb 24 16:10:00 mdp: do you have a link to that LO you demoed? Feb 24 16:11:08 is it http://www.seeedstudio.com/depot/open-workbench-logic-sniffer-p-612.html?cPath=174 ? Feb 24 16:12:59 have this in my shopping basket ;) Feb 24 16:15:36 koen: yea Feb 24 16:16:55 koen: if you are looking to buy one for work, i'd suggest either the saelie or the chronovu Feb 24 16:17:11 koen: http://sigrok.org/wiki/Main_Page Feb 24 16:20:38 * emeb has a salae. Seems to work OK. Feb 24 16:20:47 my next toys: bus birate, open logic sniffer, coolrunner II and infrared toy ;) Feb 24 16:23:39 did you say bus pirate? you want a bus pirate ? are you sure you mean bus? Feb 24 16:23:56 looks handy Feb 24 16:25:04 * koen realizes he can actually expense that stuff now Feb 24 16:25:35 SilicaGel: http://www.watterott.com/index.php?page=product&xee00c=59c04def1b070bee41c5d658c2c04b17&info=1470&xee00c=59c04def1b070bee41c5d658c2c04b17 Feb 24 16:25:57 koen: i like the chronovu8 for most 3.3V stuff Feb 24 16:26:01 oh neat, and cheap Feb 24 16:26:03 gee, HD44780 LCD Feb 24 16:26:29 is this site like the sparkfun of europe? Feb 24 16:26:34 yes Feb 24 16:26:42 at least for me Feb 24 16:27:04 because they have so much of the embedded toys on offer Feb 24 16:27:21 that's cool Feb 24 16:27:49 my friend found limor sells http://www.adafruit.com/products/126 and I just spent $24 for a board that does the same thing on sparkfun >:( Feb 24 16:29:51 oh no, High Wind Warning Feb 24 16:29:55 isn't xbee dead? Feb 24 16:36:22 SilicaGel: they even have prpplague's stuff ;) Feb 24 16:49:35 hrm my eagle license is single user Feb 24 16:49:39 but it doesn't say single isntallation Feb 24 16:49:42 and I can only be in one place at at ime Feb 24 16:49:48 so I wonder if I can install my license here too Feb 24 16:50:12 * SilicaGel checks the license Feb 24 16:53:22 yay it says it may be installed on as many computers as I want Feb 24 16:53:28 as long as I make sure only one can be used at a time Feb 24 16:55:59 you have to remove the CPU from the non-used one Feb 24 16:56:08 like you do with your starter cap Feb 24 16:58:06 get a "useless machine" ;) Feb 24 16:59:18 but installing eagle on it makes it useful Feb 24 16:59:40 eagleboard? Feb 24 17:00:27 cadsoft eagle yeah Feb 24 17:04:06 the useless box is such a great idea. have to build such a thing: http://www.youtube.com/watch?v=gKsP4vuTg2c ;) Feb 24 17:14:14 I concur with prpplague, if you want something that's just a COTS tool, Salae Feb 24 17:14:46 if you want to play with open hardware and, in my experience, pretty much it just works like a Salae…then OBLS is it Feb 24 17:15:03 keep in mind that Salae is limited to 24Mhz Feb 24 17:15:26 chronovu can do more Feb 24 17:15:46 sigrok wiki implies OBLS support is only so-so, no 200Ms/s mode Feb 24 17:15:48 haven't tried that..I really liked Salae when I used it Feb 24 17:16:02 sigrok was not my recommended client :) Feb 24 17:16:14 the java crap one is Feb 24 17:16:17 koen: the chronovu can do more and is more robust, but it is limited to a min voltage of 2.0V Feb 24 17:16:32 koen: so if you are working with some of the omap stuff that will be an issue Feb 24 17:17:20 ceil(1.8) = 2 Feb 24 17:18:09 koen: yea i prefer to have something that can do as low as 1V Feb 24 17:18:26 koen: whatever you do, don't get a logicport Feb 24 17:19:10 koen: the asix one is top of the line, but probably out of your budget Feb 24 17:30:05 * mranostay scrolls uup Feb 24 17:30:07 *up Feb 24 17:37:38 up up and away! Feb 24 17:40:50 <_av500_> lacie dumped a 2x2tb nas on me Feb 24 17:41:01 <_av500_> well, not on me directly Feb 24 17:41:38 ouch Feb 24 17:41:42 esata? Feb 24 17:42:22 <_av500_> koen: says so on the box Feb 24 17:43:32 <_av500_> one ex coworker made me beta tester Feb 24 17:44:53 <_av500_> it comes with EU, UK, UK and "strange" plug Feb 24 17:44:58 <_av500_> US Feb 24 17:45:09 hey! Feb 24 17:45:33 <_av500_> US is not the "strange" Feb 24 17:45:35 _av500_: 2big quadra? Feb 24 17:45:45 <_av500_> koen: 2big NAS Feb 24 17:45:57 <_av500_> mranostay: US is the "joke" one Feb 24 17:48:01 <_av500_> ah, seems australasian Feb 24 17:49:28 we prefer the term, "special", thank you very much Feb 24 17:50:19 <_av500_> mdp: when i saw a US mains plug for the 1st time, i thought it was for a model railway set Feb 24 17:50:47 how is it any more of a "joke" than two round pins? Feb 24 17:51:15 oh lord, flame war of outlets.... Feb 24 17:51:24 ridiculous… i know Feb 24 17:51:39 we will convert you to amuricun! Feb 24 17:51:52 and 60Hz ftw! Feb 24 17:52:07 christ... we will bitch about anything won't we? :) Feb 24 17:52:24 mranostay: go surfing Feb 24 17:52:53 not in that part of CA :) Feb 24 17:54:05 <_av500_> wifes 1st remark: where are you going to put that? Feb 24 17:54:16 <_av500_> its a design NAS! Feb 24 17:54:31 exporting poweroutages is still a market niche Feb 24 17:55:24 aholler: touche Feb 24 17:57:10 and environmental friendly, could become a green logo Feb 24 17:57:59 http://en.wikipedia.org/wiki/2003_Italy_blackout Feb 24 17:58:00 and a price from astronomers ;) Feb 24 17:58:50 mranostay, it's all the same just like AZ is a giant barren desert Feb 24 18:01:09 jay6981: that page is missing pictures like here: http://en.wikipedia.org/wiki/Northeast_blackout_of_2003 ;) Feb 24 18:05:12 aholler: yes yes i know i recall that Feb 24 18:06:55 mdp: Ohio had a nice part in that outage :) Feb 24 18:09:01 yep Feb 24 18:09:18 neighbor is a firstenergy grid op Feb 24 18:10:26 if he was in the IT world, the job would be called, "Gridmaster" Feb 24 18:12:43 sounds like a mildly stressful job Feb 24 18:13:25 perhaps a bit like being an "AirTrafficMaster" :) Feb 24 18:13:38 heh Feb 24 18:13:48 when you eff up everyone knows :) Feb 24 18:14:19 probably why they make $150k a year :) Feb 24 18:14:31 yeah Feb 24 18:15:18 most people near me have a generator anyway, we are on well water..so if the grid goes down, there's no running water otherwise Feb 24 18:15:45 not that this happens regularly..but just in case Feb 24 18:16:11 well no compound is complete without a generator Feb 24 18:16:22 true that Feb 24 18:16:36 unless it's an amish compound Feb 24 18:16:56 I also learned that archery equipment is important to avoid attracting the hordes of zombies with all the gunshots Feb 24 18:17:14 amazing what you can learn from popular TV series Feb 24 18:17:32 jay6981: well it is going to be hard for the Amish to ward off the zombies with muzzleloaders :) Feb 24 18:18:09 the neo-nazi dude on the walking dead is like a one man seal team 6 with his crossbow :) Feb 24 18:19:33 neo nazi zombies? Feb 24 18:19:41 no no Feb 24 18:20:05 nm, have to see the show…the group of characters is humorous Feb 24 18:20:07 that guy just seems like a redneck… his brother was the neo nazi Feb 24 18:20:33 yeah, they are definitely casting him as the more tolerant guy Feb 24 18:23:13 hmm, I've just decided to build me an usb-rtc using a launchpad and a ds1307 Feb 24 18:24:21 so the launchpad lying around here will get finally a usage ;) Feb 24 18:27:13 cool Feb 24 18:27:30 I've only used mine for usb serial type feed applications Feb 24 18:27:43 how capable is it for programming for other classes? Feb 24 18:28:10 don't no, I'm just starting to install gcc Feb 24 18:28:13 or are you just using ttyACM to read the data? Feb 24 18:28:18 ok Feb 24 18:28:28 mdp: yes that is the idea Feb 24 18:28:32 they are so cheap, they're great to use for simple tasks Feb 24 18:28:34 cool Feb 24 18:29:06 yeah, I have one with a reed switch on my detached garage door..hooked into the usb on an openwrt router out there Feb 24 18:29:19 great for $5 + a $.75 reed switch Feb 24 18:29:45 <_av500_> too expensive Feb 24 18:29:50 <_av500_> that 1mio units Feb 24 18:29:55 <_av500_> you are killing the company Feb 24 18:29:58 the only downside is I can't write those 20 lines of code in javascript Feb 24 18:30:34 heh Feb 24 18:31:02 aholler, if you used beaglebone for an extra $84, you could implement this in bonescript Feb 24 18:31:43 ;) Feb 24 18:33:03 I don't like programming languages starting with a j ;) Feb 24 18:34:27 aholler, that's what Boris Houndleroy made you a new one that starts with b Feb 24 18:34:32 embrace the future man! Feb 24 18:35:14 s/what/why/ Feb 24 18:35:21 uh, my fuzzy has read bonescript as js ;) Feb 24 18:35:39 so sorry ;) Feb 24 18:36:21 it could have been .pl with somebody like srostedt at the helm :) Feb 24 18:36:40 * mdp is cursing ktest.pl on and off while hacking on it Feb 24 18:36:48 <_av500_> PL1? Feb 24 18:39:14 mdp: not a fan of 2500+ line perl files :) Feb 24 18:39:30 might as well go COBOL once you venture toward PL1 Feb 24 18:39:57 that's when the old guys start remembering the good 'ol days of submitting jobs with JCL Feb 24 18:40:16 mranostay, it's another adventure Feb 24 18:40:34 I have read the presentation about ktest(.pl) in the morning Feb 24 18:40:54 "embedded edition"? Feb 24 18:40:56 could be a use case for the beagle Feb 24 18:41:37 lots of use cases for auto bisect / test Feb 24 18:42:20 could be not written in perl though :) Feb 24 18:42:25 yes, but I liked that automatic power on/off through powering through usb Feb 24 18:42:27 I have to finish cleaning up support for it deploying and image+modules on the host for nfsroot usage…it can't do that with some of the hardcoded ssh/scp desktop-isms atm Feb 24 18:43:01 aholler, oh yeah…an issue for bone only if we're talking beagle Feb 24 18:44:02 <_av500_> mdp: CICS ftw Feb 24 18:44:06 mdp: scp localhost: Feb 24 18:44:07 ? Feb 24 18:44:23 i have an c4, and somewhere I should have a kernel which runs that powered only through usb. but of course the limited range of working kernels when powered through usb limits it's usage for ktest.pl Feb 24 18:44:35 ;) Feb 24 18:44:45 the angstrom kernel works when usb powered Feb 24 18:44:51 on rev c and xm Feb 24 18:44:51 koen, hardcoded assumptions that his module tarball is to be unpacked at "/" Feb 24 18:44:54 <_av500_> "usb powered kernel" Feb 24 18:45:09 mdp: I know the answer to that Feb 24 18:45:16 mdp: "just use an initramfs" Feb 24 18:45:16 there are hacks Feb 24 18:45:21 LOL Feb 24 18:45:36 and my favourite: Feb 24 18:45:38 "device tree should help out with that" Feb 24 18:45:38 the current mainline kernel still has not enough pm for that? Feb 24 18:46:01 which angstrom kernel runs on usb power? Feb 24 18:46:03 <_av500_> aholler: there is a reason it is called "mains" line kernel Feb 24 18:46:30 aholler, my only use case involves a power switch, so I power from mains on everything in my local board farm that gets regression tested Feb 24 18:46:32 thurbad: a kernel stripped down Feb 24 18:46:34 2.6.29 did I think ,2.6.32 did not Feb 24 18:46:49 well on the BBxm/BB Feb 24 18:46:50 aholler, so beaglebone's switching smarts screws that up Feb 24 18:46:53 thurbad: 3.0.17 and 3.0.22 Feb 24 18:48:14 dt does power managment? nice to hear ;) Feb 24 18:49:14 <_av500_> yes Feb 24 18:49:23 <_av500_> it produces its own power even Feb 24 18:53:40 aholler: DT does everything! Feb 24 18:56:37 I've read that presentation too, but for me it would be enough if it enables me to update an arm-kernel without having knowledge about the used SoC Feb 24 19:01:08 Hmmm Feb 24 19:01:39 what is DT? Feb 24 19:01:54 Dumb Turd? Feb 24 19:02:51 it's a jedi mind trick Feb 24 19:02:56 <_av500_> Dev Terror Feb 24 19:03:52 <_av500_> DT helps the move graphics cards between mobile phones Feb 24 19:03:55 ah, down to 20 commits in xbmc Feb 24 19:04:32 VESA or PCI graphics cards? Feb 24 19:04:39 I gut feeling is that https://github.com/xbmc/xbmc/commit/bf9dd471187f56eeb35a6f5ce4a3c39faaa7e72e breaks the build Feb 24 19:05:04 <_av500_> your gut? Feb 24 19:05:12 the new star will be esdt, encrypted signed DT Feb 24 19:05:47 <_av500_> protected, encrypted: pesdt Feb 24 19:06:27 <_av500_> of course dt must be singed, or hackers could declare drive A: removable Feb 24 19:06:40 <_av500_> and steal the boot floppy Feb 24 19:06:42 yes, have forgotten that, companies have to be protected from its customers Feb 24 19:06:54 <_av500_> customers are the worst! Feb 24 19:07:02 Are the requests you make over L3 all a fixed size, like "read 32 bits" ? Feb 24 19:07:13 Or can it "burst" ? Feb 24 19:08:34 if someones searches marketing pro, I know some sekrets how to get rid of customers ;) Feb 24 19:08:49 SilicaGel: what did you used to do with MPC850's and MPC860's? Feb 24 19:09:38 I was at this company Microwave Data Systems and we made radios mostly for the SCADA markets ... oil and gas production and distribution, electric distribution automation, municipal water and wastewater, that sort of thing. We just made the radios. Feb 24 19:09:59 These radios all had diagnostic channels but via RS-232 Feb 24 19:10:07 nice Feb 24 19:10:21 dealt with those same chips before Feb 24 19:10:25 so what I did was made a device with the MPC860 Feb 24 19:10:39 beacuse the MPC860 not only had ethernet (pretty cool for its day) but it also had a shitload of serial ports Feb 24 19:11:08 I have a picture of the device I made with an MPC860 but the picture is really huge Feb 24 19:11:18 but do you want to see it anyway? Feb 24 19:11:24 the MPC855 is an interesting beast Feb 24 19:11:29 built in UTOPIA Feb 24 19:11:37 then after I left the company they made a version with an MPC850 built right on to the radio. Feb 24 19:12:22 Basically what this device did was be a protocol converter between our radio proprietary diagnostics and "standard" diagnostics. Mainly: SNMP, Modbus, and this electric company protocol called UCA which became a major factor in electric utility deregulation. Feb 24 19:12:33 http://www.autofrog.com/~chrisp/nx100/cimg0219.jpg Feb 24 19:12:48 that's my mpc860 board, expect it to be pretty slow. Feb 24 19:13:07 Serial ports on the left of that picture, and those plastic things are light pipes that project to the little labeled windows above where the connectors go Feb 24 19:14:39 did you guys program the aux risc processor? Feb 24 19:14:51 no Feb 24 19:15:03 do you know how T.I. says that the PRU isunsupported? Feb 24 19:15:10 SilicaGel, Are you in Rochester, NY? Feb 24 19:15:18 well back then, Motorola said "if you try to use the risc communications processor, we are going to kill you. dead." Feb 24 19:15:29 I didn't think it was even possible for an end-user to write software for that thing. Feb 24 19:15:34 bradfa: yup Feb 24 19:15:48 In fact, you know, I forgot that RISC comm processor was even there until you said that. Feb 24 19:16:01 SilicaGel, Mind if I ask where you work now? Still at MDS? Feb 24 19:16:10 SilicaGel: the how is easy, the why is harder Feb 24 19:17:12 SilicaGel: the details on L3/OCP aren't talked about much in the TRMs Feb 24 19:17:17 eh? they didn't say that to us Feb 24 19:17:28 no. I left MDS to be a director of I.T. at Adaptive Broadband, beacuse they paid me a lot of money to do that for a year. But that was enough, and I said screw I.T. I'm an engineer. I am a research engineer at R.I.T. now Feb 24 19:17:39 the risc processor there was suppose to have support for dsp like features, IIRC Feb 24 19:17:45 Please don't use this information to hunt me down and murder me. Feb 24 19:18:07 SilicaGel, neat. Don't worry, I won't :) Feb 24 19:18:16 jay6981: yeah. We were just wondering about the performance of random writes vs. sequential Feb 24 19:18:34 How about this question Feb 24 19:18:34 SilicaGel: wondering here as well Feb 24 19:18:35 ds2, did you have that talk with your h/w guy yet? :) Feb 24 19:19:12 I'm also in Rochester, I work: http://www.omni-id.com If you're interested in neat RF stuff, we've got quite a bit Feb 24 19:19:14 SilicaGel: i did some benchmarks against reading/writing DRAM from cortex-a8 and was surprised by some results Feb 24 19:19:18 mdp: yep. had a fine little chat with them Feb 24 19:20:00 surprised happily? Feb 24 19:20:02 I hope? Feb 24 19:20:08 unfortuantely no Feb 24 19:20:14 the memory bus seems to suck Feb 24 19:20:15 there is something else i'm wondering. Feb 24 19:20:24 ds2, cool Feb 24 19:20:37 If I write something from PRU to DDR, and then I write from PRU to DMEM0 Feb 24 19:20:47 at the same time, the MPU is reading those Feb 24 19:21:11 are there synchronization issues here? Like the PRU thinks it wrote to DDR but when the MPU reads from it, it's not there yet? Feb 24 19:21:28 well PRU is much simpler than MPU Feb 24 19:21:34 ok but i means this Feb 24 19:21:48 but MPU has very deep pipeline and can defer writes too Feb 24 19:21:48 when you write from the PRU to DDR you're sending a mesasge over the L3 to something. OCMR or OMCR or something. Feb 24 19:22:12 over L3 to DRAM controller Feb 24 19:22:22 OCM is for SRAM Feb 24 19:22:28 oh sorry Feb 24 19:22:41 but DRAM can't sit directly on the L3, DRAM is DRAM ... so there's something in between there Feb 24 19:22:45 there can be bus contention sure Feb 24 19:23:40 yeah, you're talking to the DRAM controller which talks to the DRAM Feb 24 19:24:32 i'm talking about whatever takes things OFF the L3 and puts them into DRAM (and the other way around) Feb 24 19:24:35 let me pull up the trm Feb 24 19:25:15 I'm looking at Figure 10-1 L3 Topology on page 1320 of the rev B trm Feb 24 19:25:25 oh, the OCP ports Feb 24 19:25:30 i guess what I am asking is where the hell on that picture is the DRAM Feb 24 19:26:03 also, apparently the MPU subsystem is the MPUSS. That's kind of disgusting. Feb 24 19:26:37 i think i know the answer, I think it's the EMIF Feb 24 19:26:52 that's the am335x name for the dram controller Feb 24 19:26:55 or is it the OCMC Feb 24 19:27:05 the EMIF says it's only 16 bits Feb 24 19:27:10 emif=external memory Feb 24 19:27:15 yeah the am335x bus is even worse :) Feb 24 19:27:53 i'm looking at am335x trm Feb 24 19:28:36 i meant worse than the am35xx Feb 24 19:29:14 emif=external memory inertface controller=dram Feb 24 19:29:23 ocmc=on chip memory controller=sram Feb 24 19:30:36 looks like L3 is 128 bits wide to EMIF, but EMIF is 16 bites wide to SDRAM module Feb 24 19:33:00 even though dram is 16 bits, 32-bit writes/reads are still atomic ? Feb 24 19:33:33 it's DDR so yeah Feb 24 19:34:49 how much more crippled is the am33x's emif compared to am35x's? Feb 24 19:35:56 the 35x is 32bit to DRAM and 64bit to L3 Feb 24 19:36:20 prpplague: so I think I'm done getting acclimated to the openocd cmdline interface…confirmed that although xM loads u-boot spl and then loads u-boot while FS2 is connected, something kills uart output while FS2 is connected Feb 24 19:36:24 so each 64bit L3 read is a single DRAM transaction Feb 24 19:36:58 bbiab Feb 24 19:37:04 did LPDDR1 support go away on the 33x? Feb 24 19:37:08 prpplague: my silly guess atm is that a clock is shared somewhere and it whacks the UART fck in some way so I get no serial output Feb 24 19:37:16 mdp: doh Feb 24 19:37:58 prpplague: because all jtag halt/resume/step/bp functionality works and I'm seeing everything executed end to end otherwise in the loading chain Feb 24 19:38:07 no, 33x still has lpddr1 support (per trm) Feb 24 19:39:25 prpplague: so anyway, reporting here in case somebody else has seen console output die with jtag connected Feb 24 19:39:31 mdp: thanks Feb 24 19:39:48 mdp: feel free to start a wiki page on elinux.org to document your work Feb 24 19:39:52 tbh, this is one board I never had occasion to connect bdi to Feb 24 19:40:01 will do Feb 24 19:40:16 Tartarus is following close behind me and may have more to add soon Feb 24 19:40:56 mdp: question relating to your elc stuff - isn't there a SPI fb accelerator block on the am33x that could have helped with some of what you did? Feb 24 19:41:16 there is? Feb 24 19:41:17 I dunno Feb 24 19:41:30 suppose to be some DMA engine Feb 24 19:41:42 McSPI is hooked to EDMA, you mean? Feb 24 19:41:44 but I can't figure out if it is tied to the McSPIs or has another interface Feb 24 19:41:53 no, some guy posted about it... let me dig up the name Feb 24 19:42:18 posted where? Feb 24 19:42:23 beagle list Feb 24 19:42:35 ok, haven't made it to that list today yet Feb 24 19:42:51 it's last in my folder list due to the low SNR Feb 24 19:43:02 if you know what I mean Feb 24 19:43:12 LIDD controller I think Feb 24 19:43:17 it was a while ago Feb 24 19:43:30 on my list to look at Feb 24 19:43:42 wow, ok, never heard of it..I'll look into it Feb 24 19:44:02 I haven't yet worked much at all with am335x, unfortunately Feb 24 19:44:06 looks like I am wrong Feb 24 19:44:15 that supports a parallel interface only, not serial Feb 24 19:44:40 ahh, ok Feb 24 19:44:41 but I suppose some hw glue... Feb 24 19:46:09 tbh, with a proper spi display controller..and working linux spi controller driver with dma support..you do achieve zero copy from userspace to controller with what I described Feb 24 19:46:52 the st7586fb driver I referenced implements that…it's just stuck on a 2.6.32-based branch in my ohporter/linux tree Feb 24 19:46:57 *nod* Feb 24 19:47:07 got other spi displays in mind... much larger ones Feb 24 19:47:13 ahh, ok Feb 24 19:47:23 but for some reason they are available cheap cheap cheap...like 7" displays for sub $10 in quantity Feb 24 19:47:30 hrm Feb 24 19:47:35 link? Feb 24 19:48:00 or I suppose that's a disty quote Feb 24 19:48:05 yes Feb 24 19:48:21 min. purchase is either 10K or 100K units Feb 24 19:48:27 consider that this is exactly how you do large res eink support in linux too Feb 24 19:48:31 so that's not unusual Feb 24 19:48:51 I think the twist is doing high refresh video etc. via deferred i/o Feb 24 19:49:10 ultimately you are limited by how fast your SPI can go Feb 24 19:49:35 that's something the displaylink guys don't really do due to usb and the high resolutions they are pushing over their similar approach Feb 24 19:49:37 I was a little surprised you didn't go with the infamous i2c based nokia displays Feb 24 19:50:17 well, 7735 was purely me playing around for fun…not TI supported Feb 24 19:50:43 the i2c displays seemed a bit more common Feb 24 19:50:46 the 7586 on an am1808 part was/is a production customer design dropped on my desk Feb 24 19:50:48 rip them out of old phones Feb 24 19:50:53 sure Feb 24 19:51:07 there's a couple fb drivers floating around for the nokia displays Feb 24 19:51:17 this isn't a hard topic...the hard part is getting decent frame rates Feb 24 19:51:19 but that's no fun if the codez already exist Feb 24 19:51:26 *nod* Feb 24 19:53:48 speaking of nokia displays..there's KKeller now :) Feb 24 19:54:00 hehe Feb 24 19:54:10 hello mdp... Feb 24 19:54:25 KKeller: rockin' the 3G at the cabin I see :) Feb 24 19:55:09 yes + have BBone on network and playing with cloud9 Feb 24 19:55:21 All using my iPad :) Feb 24 19:56:01 we need fewer clouds...clear sky computing!! Feb 24 19:57:17 <_av500_> skynet Feb 24 19:57:45 <_av500_> ds2: these 7" spi, keep some for me Feb 24 19:58:07 _av500_: you probally have a pile of them in the backroom ;) Feb 24 19:58:26 next to your mediatek and telechip reels Feb 24 20:00:30 <_av500_> not spi Feb 24 20:00:35 ds2, I will answer more questions for a 7" spi display :) Feb 24 20:01:48 okay, just buy the other 9,999 units ;) Feb 24 20:04:07 anyone going to the MWC? Feb 24 20:04:19 whia is MWC? Feb 24 20:04:29 mobile world Feb 24 20:04:50 Crofton|work: going for the tapas? Feb 24 20:04:59 or something more fermented? Feb 24 20:05:05 no Feb 24 20:05:08 I can't Feb 24 20:10:18 what's the point of using the 7" SPI display anyway? Feb 24 20:11:52 in the end it's always lvds/parallel isn't it? I mean, there's always some IC which is doing SPI -> lvds or parallel Feb 24 20:16:02 when all you have is spi? Feb 24 20:17:18 well, if we talk in beagle context Feb 24 20:17:43 otherwise I understand the reasons of course Feb 24 20:17:58 maybe lcd io pads muxed to something else? Feb 24 20:20:24 why incur the extra cost of all those connections if you don't need the performance? Feb 24 20:20:48 again…the eink use case Feb 24 20:20:50 oh yeah it takes more energy to drive more pins right? Feb 24 20:21:03 it costs more money to produce that board Feb 24 20:22:13 those pins are driven by that IC on the LCD anyway (well it might be done more effeciently) Feb 24 20:22:21 yes money, another point Feb 24 20:22:23 there's a whole world beyond displaying 1080p 120hz Feb 24 20:22:29 Yay my PRU is on the L3 bus finally!!!!!!!!!!!!!!!! Feb 24 20:22:41 I wish I understaood what the F smart standby / smart idle modes were Feb 24 20:22:51 I just turned them off and set SYSCFG to "no standby" and "no idle" Feb 24 20:22:52 now it works Feb 24 20:23:03 …using the latest whizbang OMAP8 (The Ocho) display subsystem :) Feb 24 20:24:18 what? Feb 24 20:24:51 mdp you could not resist could you... "The Ocho" Feb 24 20:26:34 ynezz: nm, point is that there are always reasons to use "only what you need" in a design Feb 24 20:27:49 6 Feb 24 20:28:00 that's why people don't design in an OMAP4430 as a gas well pump controller, it could do the job, but the controller would be needlessly expensive Feb 24 20:28:06 routing Feb 24 20:28:24 hello everybody Feb 24 20:28:30 KKeller: I knew you'd like that Feb 24 20:29:30 mdp: agreed. Feb 24 20:30:14 mdp: people might be doing it Feb 24 20:30:32 gas pumps these days are hot beds for ad pushers Feb 24 20:30:44 _well_ Feb 24 20:30:51 :D Feb 24 20:30:55 no argument there Feb 24 20:32:01 besides, the next CA rules may insist gas pumps measure things down to the nearest molecule and account for it all Feb 24 20:32:01 can I ask you guys a question? Feb 24 20:32:30 no. you broke rule #1 Feb 24 20:32:31 well, buy a bike while it's still cheap :) Feb 24 20:33:18 oh, a bike would be fun year round for my 50 mile daily work drive Feb 24 20:33:30 if you do that then they'll impose a road tax on food Feb 24 20:33:31 ;) Feb 24 20:33:38 "do you remember those times where there was enough gas even for private cars" Feb 24 20:33:44 and start regulating bathroom emissions Feb 24 20:34:33 he says it used to be a farm, before the motor law Feb 24 20:35:18 does anyone know from a benchmarking point of view how does the beaglebone compare to the beagleboard (using only the ARM side of course)? Feb 24 20:36:15 in cache? out of cache? Feb 24 20:36:25 doing what ops? Feb 24 20:37:04 running bonescript? Feb 24 20:37:07 just give him those nice numbers! Feb 24 20:37:10 well i`m interested in porting a speech recognition program (Sphinx) Feb 24 20:37:42 so pretty memory and computational intensive application Feb 24 20:38:32 the bone has a narrower memory bus Feb 24 20:38:53 has no audio inputs Feb 24 20:39:01 pittbull, I'm a big fan of trade studies that involve actually running some real code..if only the core cpu intensive fragments on a ref board Feb 24 20:39:32 i have an omap3530evm from mistral Feb 24 20:39:40 which i`m using now Feb 24 20:39:45 since all benchmarks lie, except your own…it's best to make your own and try it Feb 24 20:39:52 but it`s an older revision (Rev D) Feb 24 20:39:52 may have mercy on your soul Feb 24 20:40:01 ds2, LOL Feb 24 20:40:32 mdp: i did that, and I couldn't understand why single memory fetch was taking ~100ns out of cache Feb 24 20:40:34 * mdp looks for #mistral Feb 24 20:41:27 and it has smaller memory (128 MB) Feb 24 20:41:43 that`s why i was looking to move to a beagleboard Feb 24 20:42:01 i know the bone has no audio inputs but i thought i`d use a USB adapter Feb 24 20:42:17 well, considering that beagle* boards are nearly free from a commercial cost POV…it's easy to move Feb 24 20:42:22 seems like bb memory performance is capped around 350MB/s Feb 24 20:42:28 seems a lot slower than it ought to be Feb 24 20:42:56 i guess that's AXI/OCP's fault? Feb 24 20:46:17 I didn`t mention that i am a student and right now money is not my strongest point Feb 24 20:46:36 so that`s why i`m asking you for some pointers Feb 24 20:47:12 provide compilable code and you would get the output, easy Feb 24 20:47:24 I can run the test on my C4 Feb 24 20:47:42 if it's not some retarted makefile :) Feb 24 20:49:16 there was even some benchmark already available for different boards Feb 24 20:49:36 but I don't remember what was it exactly Feb 24 20:49:55 i'm sorry but i can`t provide the code right now Feb 24 20:50:11 and it`s a whole library actually Feb 24 20:50:15 i haven`t found Feb 24 20:50:20 any benchmarks Feb 24 20:50:38 and i figured maybe someone around here knows Feb 24 20:50:56 then I don't get it, why do you want benchmarks if you don't know what would you need :) Feb 24 20:52:00 what do you mean by i don`t know what i need? Feb 24 20:52:01 besides, what porting would you have to do? both are the same cpu arch Feb 24 20:52:18 I think mru/aholler and "the others" here were running some benchmarks (I think it was for memory bw) on different animal boards Feb 24 20:52:46 the toolkit is written for x86 Feb 24 20:53:00 and the results were posted to the mailing list, so try to search there Feb 24 20:53:55 (if I'm not mistaken completly) Feb 24 20:55:01 there is already speech tools in angstrom Feb 24 20:55:21 but angstrom site looks odd today Feb 24 20:55:42 seized by fbi Feb 24 20:56:15 <_av500_> the warez Feb 24 20:56:21 malwarez Feb 24 20:56:45 crap, where willI get the codez now? Feb 24 20:56:53 need my feedzzz Feb 24 20:57:00 * mranostay deletes all knowledge of koen Feb 24 20:57:02 opkg-torrent Feb 24 20:57:07 i need to do it as an university project.. Feb 24 20:57:22 ah, there we go :) Feb 24 20:57:25 mranostay, his code is never going into space now! Feb 24 21:00:19 mdp: "i don't recall" Feb 24 21:00:44 so my question put simple was if a bone is faster than the beagle in intensive cpu and memory apps Feb 24 21:00:57 heh no Feb 24 21:01:09 no Feb 24 21:01:17 <_av500_> nas setup started Feb 24 21:01:21 get the beagle and use built in audio Feb 24 21:01:48 <_av500_> pittbull: both have cortex a8 Feb 24 21:01:52 a usb audio dongle might add nore headache to the bone Feb 24 21:02:11 <_av500_> so i dont see why at same clock one should be significantly faster Feb 24 21:02:25 <_av500_> specially not for speech processing Feb 24 21:03:34 http://groups.google.com/group/beagleboard/browse_thread/thread/8a02388772a1d177/f979ffbfc5f34ac3?lnk=gst&q=povray#f979ffbfc5f34ac3 Feb 24 21:03:46 it was actually some povray benchmark Feb 24 21:03:52 but bone is probably missing Feb 24 21:04:03 <_av500_> that is hardfp vs soft Feb 24 21:04:07 <_av500_> one the same cpu Feb 24 21:04:52 he wanted numbers :p Feb 24 21:05:22 <_av500_> 54 Feb 24 21:06:01 thanks, now I've to change my passwords :( Feb 24 21:06:19 <_av500_> 55? Feb 24 21:06:19 thanks for the link Feb 24 21:06:36 <_av500_> the link does not help to answer your question Feb 24 21:07:02 i know Feb 24 21:07:19 but i appreciate the fact that he took the time Feb 24 21:08:09 I speak google Feb 24 21:13:49 hmm, hit again by ti's registration obsession. Seems I'm either to old to learn to avoid their side or I'm to naive and still have hope Feb 24 21:15:40 yes, it's usually easier to get the samples from them then softwarez :p Feb 24 21:16:18 your registration info is sold for an additional revenue stream Feb 24 21:16:31 times are tight, my chilluns need to eat Feb 24 21:19:27 * _av500_ loves free samples from TI Feb 24 21:19:41 <_av500_> 2 day fedex delivery inlcuded Feb 24 21:21:44 * _av500_ needs an ip address table for his tiny home network Feb 24 21:21:53 <_av500_> hm, 42 and 54 are already taken Feb 24 21:22:29 I'm using dhcpd.conf for that Feb 24 21:23:01 _av500_: i made sure i coughed in the box for you Feb 24 21:23:10 _av500_: compliments of ds2 Feb 24 21:23:40 <_av500_> aholler: yes, but my 50c dsl router does not have that :) Feb 24 21:23:53 <_av500_> prpplague: thx Feb 24 21:29:24 av500: just one of your environmental friendly arm devices Feb 24 21:32:18 <_av500_> i need a proper dsl router Feb 24 21:32:22 <_av500_> not that telekom shit Feb 24 21:33:35 still using my original speedstream ( a manageable unit). Feb 24 21:33:50 <_av500_> well, this one I can manage Feb 24 21:33:55 <_av500_> its still crap Feb 24 21:34:08 I have two spares Feb 24 21:34:24 hmm, I would consider registration to shop a free omap5-dev-board sample ;) Feb 24 21:34:40 hmm, havent seen those Feb 24 21:35:54 av500: aren't these avm-devices, so hackable? Feb 24 21:36:14 avm devices Feb 24 21:36:18 * koen coughs Feb 24 21:36:20 seen lots of little chips I would struggle with mounting to a board. Feb 24 21:36:28 usb works great on avm devices Feb 24 21:36:52 that's where fifmode=5 comes from Feb 24 21:36:57 fifomode* Feb 24 21:38:33 <_av500_> yes Feb 24 21:40:05 <_av500_> aholler: but they suffer from feature bloat Feb 24 21:40:09 <_av500_> I only want a dsl router Feb 24 21:40:25 <_av500_> i dont need the wifi and nas and whatever part of it Feb 24 21:40:53 <_av500_> but the dsl only is no more Feb 24 21:41:04 <_av500_> so i need to pay like 150 for the dsl part Feb 24 21:41:47 <_av500_> koen: there was that av guy in gst-ti Feb 24 21:41:49 <_av500_> avm Feb 24 21:43:48 mine are dsl router with only one ethernet port, no wifi. My linksys wifi router takes care of that. Feb 24 21:44:19 <_av500_> yes Feb 24 21:44:36 <_av500_> but avm has a neat dsl line quality display Feb 24 21:44:40 _av500_: you are right on that most are loaded these days. Feb 24 21:44:43 <_av500_> that cheap ones lack Feb 24 21:45:15 I have to load the builtin web page for that info Feb 24 21:45:51 ther eis a dsl link light on front Feb 24 21:47:34 <_av500_> damn, now I need a gige router Feb 24 21:47:41 <_av500_> the dree nas is costly :) Feb 24 21:47:45 <_av500_> free Feb 24 21:48:01 how did you get a free nas? Feb 24 21:48:13 <_av500_> I#m beta tester for LaCie Feb 24 21:48:17 ah Feb 24 21:48:18 <_av500_> ex coworker is there now Feb 24 21:48:23 ex them to send me one Feb 24 21:48:31 <_av500_> sure Feb 24 21:48:59 <_av500_> Crofton|work: they also have a test program for their cloud drives Feb 24 21:49:05 <_av500_> seems more fitting for you Feb 24 21:49:10 <_av500_> cyberbits Feb 24 21:50:08 I have a semi nutty email from a "Cloud Computing Commander" Feb 24 21:54:18 hmm, ccs for linux will come soon ;) Feb 24 21:54:39 <_av500_> css, no? Feb 24 21:55:56 code composer? Feb 24 21:56:04 that has been out for linux for a while Feb 24 21:58:36 ah, yes, just 5.1.1 will come soon Feb 24 22:04:51 I wonder how that looks like, but again not without registration. ;) Feb 24 22:06:36 what is it good for? Feb 24 22:06:52 don't know, I'm just curious Feb 24 22:08:33 <_av500_> ynezz: dsp debugging Feb 24 22:10:16 * _av500_ used CCS like 10ys ago Feb 24 22:10:40 <_av500_> arm7 + c54x Feb 24 22:11:35 ccs? awful experience.... Feb 24 22:14:28 _av500_: what kind of chip had a arm7 + c54x? Feb 24 22:14:35 <_av500_> dsc21 Feb 24 22:14:37 grandpa of the OSK? Feb 24 22:14:42 ohhhh Feb 24 22:14:43 <_av500_> no Feb 24 22:14:49 <_av500_> for digital cameras Feb 24 22:15:04 is that grandpa of the davinci's? Feb 24 22:15:05 <_av500_> dsc21, 24, 25, dm270, dm320 Feb 24 22:15:16 nifty Feb 24 22:15:17 <_av500_> and the dm310 which was ill fated Feb 24 22:15:33 <_av500_> they botched the mmu so linux was a nogo Feb 24 22:15:52 <_av500_> it would not traverse pages Feb 24 22:15:59 <_av500_> so it faulted every 4k :) Feb 24 22:16:14 CONFIG_NO_MMU ;) Feb 24 22:16:21 <_av500_> and rewriting linux to JMP every 4k was out of the question :) Feb 24 22:16:33 <_av500_> ds2: yes, we tried that Feb 24 22:17:14 <_av500_> but it was for a qtopia device and we needed to be binary compatible to the zaurus Feb 24 22:17:30 <_av500_> otherti customers used it with their no-OSs Feb 24 22:17:30 oh Feb 24 22:17:47 <_av500_> we got OMAP1 instead and a free DM270 on top Feb 24 22:18:05 strictly speaking, you cannot be binary compatible with the Z with any device Feb 24 22:18:10 <_av500_> so we had arm7+arm9+c54+c55 Feb 24 22:18:11 any TI device Feb 24 22:18:20 <_av500_> and 64+32mb ram Feb 24 22:18:37 <_av500_> /dev/fb was on the DM270 via 16bit HPI Feb 24 22:18:49 <_av500_> and no 8bit access Feb 24 22:18:49 none of those run iwmmx! Feb 24 22:19:06 <_av500_> no need for iwmmx Feb 24 22:19:12 <_av500_> most of zaurus stuff ran fine Feb 24 22:19:23 <_av500_> it was 90% a different path or so Feb 24 22:19:31 I see. Feb 24 22:19:33 <_av500_> which was amazin in fact Feb 24 22:19:36 <_av500_> +g Feb 24 22:19:57 I miss my Z's Feb 24 22:20:10 <_av500_> i still have one for sale Feb 24 22:20:18 which one? Feb 24 22:20:21 <_av500_> with dock *and* serial cable Feb 24 22:20:24 SL5500? Feb 24 22:20:25 <_av500_> original one Feb 24 22:20:30 SL5000? Feb 24 22:20:36 <_av500_> think so Feb 24 22:20:38 <_av500_> need to check Feb 24 22:20:41 original retail or the give away one? Feb 24 22:20:45 <_av500_> gee Feb 24 22:20:54 <_av500_> made in the morning or the afternoon? Feb 24 22:21:03 they gave away the SL5000 only... with half the memory, IIRC Feb 24 22:21:07 the SL5500 was retail Feb 24 22:21:18 <_av500_> the girl who assembled it had black hair Feb 24 22:21:34 so she was not from the great north... Feb 24 22:21:45 <_av500_> trolltech made the sw only ;) Feb 24 22:22:01 <_av500_> but in australia Feb 24 22:22:22 <_av500_> these were fun times Feb 24 22:22:28 yes they were Feb 24 22:22:38 <_av500_> like the esmertec people being literally thrown out of our CEOs office Feb 24 22:22:53 now if they only put WiFi on there so one didn't have to burn the CF slot Feb 24 22:23:01 <_av500_> when asked for the price of their JV Feb 24 22:23:04 <_av500_> JV Feb 24 22:23:06 <_av500_> damn Feb 24 22:23:10 <_av500_> the m is fubar Feb 24 22:23:12 <_av500_> JVM Feb 24 22:23:28 heheh Feb 24 22:23:35 <_av500_> they had insane ideas Feb 24 22:27:41 "This patch is intentionally incomplete to simplify the review." Feb 24 22:27:45 https://github.com/torvalds/linux/commit/d80e731ecab420ddcb79ee9d0ac427acbc187b4b Feb 24 22:28:30 <_av500_> "this tax return is intentionally incomplete to simplify the review" :) Feb 24 22:32:27 hello! any way to read beaglebone expansion cape eeprom data from userland? i2cget says device busy Feb 24 22:32:52 see my email to the beagleboard list Feb 24 22:34:11 e.g. http://groups.google.com/group/beagleboard/search?group=beagleboard&q=eeprom+hostname Feb 24 22:36:14 thanks! Feb 25 00:19:12 When building Angstrom for Beaglebone I get this error ERROR: No recipes available for: recipes-connectivity/dhcp/dhcp_4.2.0.bbappend Feb 25 00:19:28 does anybody know how I can fix it? Feb 25 00:20:32 the exact path is /home/orlcp440/setup-scripts/sources/meta-openembedded/meta-oe/recipes-connectivity/dhcp/dhcp_4.2.0.bbappend Feb 25 00:20:37 OrlandoT: well there is no recipe named dhcp_4.2.0.bb Feb 25 00:24:06 I think I have an alternative to the no space heater in cube policies Feb 25 00:27:07 i was told I could listen to the radio at a reasonable volume from nine to eleven Feb 25 00:28:04 trying to argue 150db is a reasonable volume? :D Feb 25 00:28:19 what? :) Feb 25 00:28:59 it's not okay, because if they take my stapler again I'll set the building on fire Feb 25 00:29:00 where 0db is threshold of hearing ;) Feb 25 00:30:34 former cube mate of mine had a mirrored drive door G4 mac. that functioned as both a space heater and source of irritating noise. Feb 25 00:30:41 power to heat and noise conversion device Feb 25 00:32:28 ah, the windtunnel Feb 25 00:32:33 * mru has one of those Feb 25 00:32:37 ds2: you mentioning silly office policies just reminded me of office space Feb 25 00:32:57 haha windtunnel is definitely right Feb 25 00:33:30 jay6981: that was leading in another direction Feb 25 00:33:48 cuz I am getting more heat out of this core i7 then out of most space heaters :D Feb 25 00:34:20 my i7 sits at a rather constant 45C Feb 25 00:34:23 regardless of load Feb 25 00:34:36 desktop i7 or laptop? Feb 25 00:34:40 desktop Feb 25 00:34:48 this is a laptop Feb 25 00:34:56 with a hot air duct on the side Feb 25 00:35:01 laptop has an i5 and there I certainly notice a difference based on load Feb 25 00:35:22 the i7 in my MBP runs amazingly cool Feb 25 00:36:10 maybe this is a badly designed laptop *shrug* Feb 25 00:36:52 at high load my laptop doubles as hair dryer Feb 25 00:37:06 nifty for traveling Feb 25 00:37:15 yeah those little fans are annoying when they rev up Feb 25 00:38:09 wonder if that sucker is thermothrottling... feeling slower then the atom Feb 25 00:39:30 then this other crappy laptop with the AMD Neo always feels like it's about to melt Feb 25 00:41:09 I swear I willnever try again downloading something from ti ... Feb 25 00:41:28 try downloading from other folks Feb 25 00:41:40 then ti stuff would seem like paradise Feb 25 00:43:34 I know why i prefer avr Feb 25 00:47:02 Hi Feb 25 00:47:44 I'm back.. I was here yesterday trying to compile ftdi_sio driver on my beaglebone Feb 25 00:48:15 I've download the kernel source files onto the beaglebone and tried Make.. but getting an error: Feb 25 00:48:20 make: *** No rule to make target `include/config/auto.conf', needed by `include/config/kernel.release'. Stop. Feb 25 00:48:43 it's getting further now than before, and looks promising.. Feb 25 00:48:59 but suspect I'm still missing something Feb 25 00:51:02 so I'm trying to following the instructions on here: http://support.robotis.com/en/techsupport_eng.htm#software/dynamixel_sdk/usb2dynamixel/usb2dxl_linux.htm Feb 25 00:51:20 but getting errors on step 4, when doing the Make Feb 25 00:51:30 you have to configure the kernel and than call make modules_prepare Feb 25 00:52:39 so I use the ./configure command Feb 25 00:52:54 then "make modules_prepare" Feb 25 00:52:55 ? Feb 25 00:53:24 hah no Feb 25 00:53:26 from the /usr/src/ftdi_sio folder Feb 25 00:53:36 no, get a .config for your board, mabye make beagle_foo_defconfig works Feb 25 00:53:58 I don't know the exact name Feb 25 00:54:03 you can get the .config by saying zcat /proc/config.gz > .config Feb 25 00:54:11 if you're trying to do it on your board Feb 25 00:54:18 yep Feb 25 00:54:31 where do I put the .config file when it's created? Feb 25 00:54:44 top of kernel tree Feb 25 00:54:47 kernel source root Feb 25 00:55:49 ok Feb 25 00:55:51 ta Feb 25 00:56:50 looks like I'm getting further Feb 25 00:56:58 but I've got another error now Feb 25 00:56:59 ERROR: Kernel configuration is invalid. Feb 25 00:57:15 include/generated/autoconf.h or include/config/auto.conf are missing. Feb 25 00:57:26 Run 'make oldconfig && make prepare' on kernel src to fix it. Feb 25 00:57:49 did you run the prescribed commands? Feb 25 00:58:01 trying now Feb 25 00:59:01 is that not going to compile the whole kernel? Feb 25 00:59:51 ST7735 framebuffer support (FB_ST7735) [N/m/y/?] (NEW) Feb 25 01:00:10 do I just press enter past these? Feb 25 01:00:25 sounds like the source you're compiling might not match the kernel you're running Feb 25 01:00:38 sure Feb 25 01:00:44 oh no Feb 25 01:00:46 lol Feb 25 01:01:18 new kernel wants to know if you want certain new stuff Feb 25 01:01:56 ah ok Feb 25 01:03:36 it's not so simple to compile one driver then.. Feb 25 01:03:45 I only modified 1 file Feb 25 01:04:17 just the ftdi_sio.c file Feb 25 01:04:19 why? Feb 25 01:04:52 it's required for the USB2dynamixal device I'm using Feb 25 01:05:11 you don't just need to add a new vendor id do you? Feb 25 01:05:32 no it's slightly more modifications Feb 25 01:05:40 k Feb 25 01:05:58 about 12 lines of code Feb 25 01:06:08 there detailed here: http://support.robotis.com/en/techsupport_eng.htm#software/dynamixel_sdk/usb2dynamixel/usb2dxl_linux.htm Feb 25 01:06:15 you oughta know this stuff is never simple :) Feb 25 01:06:36 well yeah, I kind of guessed.. being new to linux etc Feb 25 01:06:53 I'm used to programming microcontrollers Feb 25 01:07:06 not rebuilding linux kernels Feb 25 01:07:10 :) Feb 25 01:08:08 mranostay: if there is no such recipe then how can I get that recipe? or is that a bug? Feb 25 01:08:32 I followed the instructions on their website Feb 25 01:09:00 does it take a while to do the 'make prepare' on a linux kernel Feb 25 01:10:11 maybe on the beagle it does Feb 25 01:10:30 seems to be repeating the same steps over and over Feb 25 01:12:42 think I killed the system with heat Feb 25 01:14:28 scripts/kconfig/conf --silentoldconfig Kconfig Feb 25 01:15:30 HOSTCC scripts/kconfig/conf.c Feb 25 01:16:14 what kind of filesystem are you doing things on? Feb 25 01:17:33 it's just the default beaglebone filesystem Feb 25 01:17:51 did you set the clock? Feb 25 01:20:00 no Feb 25 01:20:06 how do I set the clock? Feb 25 01:20:13 date Feb 25 01:20:15 it keeps warning about that Feb 25 01:20:22 for a reason Feb 25 01:20:39 Fri Nov 18 02:58:44 GMT 2011 Feb 25 01:21:01 RTFM Feb 25 01:21:19 doing that ds2 :) Feb 25 01:23:21 dfsadf Feb 25 01:23:31 Sharpyuk: how did you make out Feb 25 01:23:42 with your ftdi problem Feb 25 01:23:46 scroll back Feb 25 01:23:49 still trying Feb 25 01:23:54 :) Feb 25 01:24:23 can someone help me with s-video setup on beagleboard-xm rev-c Feb 25 01:25:15 in Ubuntu 11.10 Feb 25 01:27:41 ok I've set the date and time now, and retrying the make prepare on the kernel source Feb 25 01:27:42 is it just me or has the Angsrom website been down for 2 days now? http://www.angstrom-distribution.org ? Feb 25 01:27:48 looking better now Feb 25 01:29:25 Or point me in the right direction ? Feb 25 01:30:07 seth123: you probably need to figure out how to set the modeline for 480i video Feb 25 01:32:59 do i need to make a boot.cmd script ? Feb 25 01:33:21 probably Feb 25 01:33:29 or set it manually and try it first Feb 25 01:33:46 ftdi_sio.c:68:22: error: field 'chip_type' has incomplete type Feb 25 01:33:48 through the set env variables Feb 25 01:33:52 right Feb 25 01:34:04 thanks will give it a try Feb 25 01:35:15 ftdi_sio.c:864:3: error: 'FT232H' undeclared here (not in a function) Feb 25 01:55:02 alright this is what i came up with setenv bootargs omapfb.mode=tv:ntsc omapdss.def_disp=tv Feb 25 01:55:13 didn't work Feb 25 01:55:56 it looks reasonable but i don't know what they should be exactly Feb 25 01:56:28 okay, thanks for your help Feb 25 01:57:48 seth123: have you seen this? http://www.mjmwired.net/kernel/Documentation/arm/OMAP/DSS Feb 25 01:58:31 No, will take a look Feb 25 02:32:44 x **** ENDING LOGGING AT Sat Feb 25 02:59:59 2012