**** BEGIN LOGGING AT Fri Mar 23 02:59:58 2012 Mar 23 04:37:41 * mranostay waves Mar 23 04:38:26 yo Mar 23 04:39:01 Moooo Mar 23 04:40:47 * mranostay pokes around bonescript Mar 23 04:41:22 adding meat? Mar 23 04:42:35 seeing if should do my geiger app in bonescript Mar 23 04:49:34 probably will just build my HV probe Mar 23 04:57:13 mranostay: what kind of sense element are you settling with? Mar 23 05:01:12 ds2: you mean tube? Mar 23 05:03:37 mranostay: you are going with a tube? thought you might be using that term generically Mar 23 05:04:49 (vs a scitilator (sp) or other detectors) Mar 23 05:25:23 ds2: yeah plain old tube Mar 23 05:28:50 mranostay: how sensitive of a tube are you using? alpha/beta/gamma tube? Mar 23 05:36:51 ds2: SBT-9 Mar 23 05:42:41 cheap LND-712 basically Mar 23 05:45:24 electronics goldmine stuff? Mar 23 05:49:23 ebay'ed it Mar 23 06:11:19 so sakoman, do you know the status of the musb gadget stuff? Mar 23 06:32:36 dammit :/ Mar 23 06:32:54 I googled "arm conditions" to try to find a reference for the condition flags Mar 23 06:33:03 now AdSense thinks I have carpal tunnel. Mar 23 06:34:08 XMPPwocky: you will someday :) Mar 23 06:49:59 http://pastebin.com/2VBPA2nM this basically does: while (; i < j; i++) { ... }, yes? Mar 23 06:59:43 * mranostay notes he has no real sleep schedule anymore Mar 23 06:59:45 ... Mar 23 07:05:22 green13: ah, probably not Mar 23 07:05:26 woops Mar 23 07:15:59 bummer, u-boot has two usb gadget frameworks, and old one, and a new one Mar 23 07:16:10 musb is supported on the old framework Mar 23 07:16:16 ethernet is supported on the new framework Mar 23 07:17:22 isn't ethernet usbhost? Mar 23 07:17:50 drivers/usb/gadget/ether.c Mar 23 07:20:07 I stand corrected! Mar 23 08:59:34 Hello, I have a short question related to booting Ubuntu on the Beaglebone Mar 23 08:59:47 I downloaded the demo image for 11.10 and the boot seems to go fine, however, I cant log in Mar 23 09:00:17 when the login prompt is output over the serial, I seemingly cant send any keyboard strokes to the device Mar 23 09:00:24 has anyone experienced anything similar? Mar 23 09:01:58 how are you connecting to the serial device? (both software and hardware wise) Mar 23 09:02:00 http://www.buildinginternetofthings.com/2012/03/22/femtoduino-the-smallest-arduino-clone-available/ Mar 23 09:02:24 through USB Mar 23 09:02:40 I remembered wrong, I cant send any keystrokes through the serial Mar 23 09:02:48 so it seems to be read-only Mar 23 09:03:35 check your terminal emulator serial port settings Mar 23 09:03:48 flow control off Mar 23 09:03:54 ok, I just use screen Mar 23 09:04:12 meta-a-f in screen IIRC Mar 23 09:05:17 hm, no luck, it still wont accept any input Mar 23 09:09:03 there is something happening though, the device complains that getty has been hanging for too long Mar 23 09:11:45 minicom and picocom also show the same behavior Mar 23 09:20:37 and nanocom? Mar 23 09:20:45 femtocom? Mar 23 09:27:06 * dm8tbr throws a Gigacom at av500 Mar 23 09:30:33 dm8tbr: which reminds me of that pair of gigaset M101 that I have Mar 23 09:30:50 64kbit uart over dect :) Mar 23 09:31:16 teehee, yeah, back in the days(tm) Mar 23 09:31:30 worked like a charm in the dialup eara, no new to drill holes in the flat Mar 23 09:32:19 -a Mar 23 09:34:02 but also convenient to extend Uni club 'network' to the dorm ;) Mar 23 09:34:17 or that Mar 23 09:34:42 hmm, I should hook that to the nas box Mar 23 09:34:43 "back in the day" yeah. how i do NOT miss those days. Mar 23 09:35:40 LetoThe2nd: http://www.youtube.com/watch?v=cm0iQz24Aac Mar 23 09:37:09 av500: seen them three ties :) Mar 23 09:37:14 times, even Mar 23 09:37:48 was not wearing my vodka-tractors-rock n roll hoodie when we met last time, i guess ;P Mar 23 10:12:47 Gah Mar 23 10:13:09 I think push and pop are broken on my baglebone, or I'm not doing it right Mar 23 10:18:49 or both Mar 23 11:05:44 koen: any luck with these latest 3.5 LCD timings? need someone to test? Mar 23 11:29:02 WTF. Mar 23 11:29:53 ldr sp, =stack_start_address; push {r0, r1}; pop {r0, r1}; push {r0, r1, r2, r3}; pop {r0, r1, r2, r3} Mar 23 11:30:17 r2 and r3 are corrupted. Mar 23 11:30:21 I'm pretty sure push and pop work fine Mar 23 11:30:45 any special place sp has to be? it's word aligned Mar 23 11:30:52 remember the stack grows downwards Mar 23 11:30:59 yep Mar 23 11:31:07 i can actually view memory and see it Mar 23 11:31:15 and the stack must be word-aligned at all times Mar 23 11:31:19 and the second push call just seems to no-op Mar 23 11:32:08 trust me, push and pop work Mar 23 11:32:09 mru: yep, stack_start_address is 0x8200fc00 Mar 23 11:32:28 mru: i know, but i'm trying to figure out why they won't in this case Mar 23 11:33:19 * XMPPwocky tries to make a test case Mar 23 11:42:15 oh, now it works Mar 23 11:42:17 course. Mar 23 11:43:47 night Mar 23 11:43:58 falls Mar 23 11:45:16 and is trampled by his horse Mar 23 11:46:37 oh, carp. i know what it was Mar 23 11:46:44 do tell Mar 23 11:47:08 and are you sure it was a carp, not a trout? Mar 23 11:47:29 the trout is after all the fish of choice on irc Mar 23 11:47:29 boot.bin is old Mar 23 11:47:31 was probably the sole reason Mar 23 11:47:49 av500: :/ Mar 23 11:49:38 are we doing fish puns today? Mar 23 11:50:00 but in fresh bins ... Mar 23 11:50:26 mru: i don Mar 23 11:50:37 i'm not sure WHAT the halibut we're doing Mar 23 11:51:08 but the rest of us seem to have cod on quickly Mar 23 11:52:08 come on, salmon else has to come up with one Mar 23 11:52:34 but you're doing just fin on your own Mar 23 11:53:22 i really ought to go to bed, so can we resume this at a more opportuna moment? Mar 23 11:53:47 * av500 does not eel like it Mar 23 12:01:21 hello, i am trying to compile a tcl/tk application on the BBoard rev c5.. http://pastebin.com/LbuzQXGn Mar 23 12:01:53 i am not sure if this is right place to ask about this.. please dont be angry with me. Mar 23 12:01:56 tcl is not compiled Mar 23 12:01:57 :) Mar 23 12:02:17 but cannot find anything about tk-lib-dev on the web Mar 23 12:03:05 yes it is because there is no tklib listed in Angstrom kernel Mar 23 12:03:13 i mean kernel --> linux Mar 23 12:03:28 where should i get that? is it impossible? Mar 23 12:04:05 no, the tcl language is not compiled Mar 23 12:04:05 tk-lib-dev Mar 23 12:04:22 but there is a tcl library for embedding a tcl interpreter into an application Mar 23 12:04:25 what do you mean? sorry.. Mar 23 12:05:07 yes i did successfully install tcl package. Mar 23 12:05:16 good Mar 23 12:05:31 but you're missing the -dev packages, apparently Mar 23 12:05:31 i am compiling puredata on the board Mar 23 12:05:46 that needs tcl-dev Mar 23 12:05:52 tk-dev Mar 23 12:05:54 also. Mar 23 12:06:35 is there anyway that i can skip them or.. find alternative? generally speaking.. Mar 23 12:07:36 when some app needs -dev packages to be complied? Mar 23 12:08:05 sounds stupid.. but i am not really understanding.. this situation.. Mar 23 12:08:32 -dev package is something.. of raw codes.. on the things, right? Mar 23 12:08:32 you can skip them if you remove tcl from that app Mar 23 12:08:42 only then? Mar 23 12:08:44 but I guess then the app does not work any more Mar 23 12:08:44 um.. Mar 23 12:08:49 yes. Mar 23 12:08:56 if you want to compile an app to use tcl, you need tcl headers Mar 23 12:08:57 the -dev packages are header files and (sometimes) static libs Mar 23 12:09:08 thus you need the -dev package Mar 23 12:09:13 opgk Mar 23 12:09:35 then.. you mean that i can use any header files for tcltk... for BBoard? Mar 23 12:09:41 i mean.. angstrom? Mar 23 12:09:50 if the interface is same Mar 23 12:10:09 but still.. they should be configured.. properly.. right? Mar 23 12:10:35 http://www.angstrom-distribution.org/repo/?pkgname=tcl-dev Mar 23 12:10:41 why not just install that? Mar 23 12:10:48 opkg does not support.. tcl-lib-dev Mar 23 12:10:56 only.. tcl-dev Mar 23 12:10:59 then try tcl-dev Mar 23 12:11:07 is it also includes tcl-lib-dev? Mar 23 12:11:14 how would I know Mar 23 12:11:18 I just googled that Mar 23 12:11:30 oops.. sorry Mar 23 12:11:39 anyway thanks a lot Mar 23 12:11:50 sergey, larry: ^^^^ Mar 23 12:12:17 thank you very much! :) Mar 23 12:24:55 Right now my SPI is outputting on D1 but I would like it to output on D0 is that possible? Mar 23 12:25:23 change the pinmux? Mar 23 12:27:16 hey :) Mar 23 12:27:17 I changed the pin mux on the D1 to be in mode 7 (GPIO) and it still outputs data Mar 23 12:28:05 Does the pin mux need to be changed at a particular time? perhaps before spi board registration? Mar 23 12:28:19 hmmm Mar 23 12:28:40 icota: no luck Mar 23 12:29:13 the 800x480 timings work on the 3.5" as well, but as one would expect, it is cropped Mar 23 12:29:24 koen: i think there might be some modesetting via SPI to be done Mar 23 12:29:39 icota: I tried your timings in 16 and 24 bit mode, no dice Mar 23 12:30:31 koen: what is that mystery panel? Mar 23 12:30:43 av500: 3.5" lcd 320x240 for a cape Mar 23 12:31:44 av500: CDTech S035Q01 Mar 23 12:31:48 that's so ............ 2003 Mar 23 12:32:39 icota: from what I can decypher from the datasheet is that the pixclock needs to be between 8 and 33 mhz for non 24 bit modes Mar 23 12:32:59 koen: yeah, the datasheet is really bad Mar 23 12:33:28 koen: what does the whole SPI example code do? i don't really get it Mar 23 12:34:02 don't know, I'd like to avoid using the SPI setup Mar 23 12:34:25 icota: can you try the 800x480 mode and see if that gets a picture on your lcd as well? Mar 23 12:34:48 my plan was to go from there and adjust the resolution, pixclock and porches to step by step Mar 23 12:35:08 but since I destroyed my lcd it involved a roundtrip to texas to test Mar 23 12:35:32 icota: i'll let you know in half an hour if it works on my side Mar 23 12:35:39 if so, i'll try to go from there Mar 23 12:35:49 fbset should be enough? Mar 23 12:36:26 koen: you work in dallas? Mar 23 12:36:41 I work from home in the netherlands, but the office is in the dallas area Mar 23 12:37:01 icota: fbset doesn't work, the kernel needs to get rebuilt Mar 23 12:37:12 well, fbset only works to read the timings, not to set them Mar 23 12:37:27 which is why this video works is *$($*#($*(# annoying me Mar 23 12:37:53 good thing i have the next 6 hours free Mar 23 12:38:00 don't know the details of your case, but you can set the timings via sysfs Mar 23 12:38:34 tomba: not on am335x, that's da8xx-fb.c based Mar 23 12:38:53 koen: what mode are you using? Mar 23 12:38:56 ok, never mind then =) Mar 23 12:38:57 which is a dumb driver for an even dumber framebuffer Mar 23 12:39:02 av500: mode? Mar 23 12:39:11 pixclock needs to be between 8 and 33 mhz for non 24 bit modes Mar 23 12:39:13 ^^^ Mar 23 12:39:26 16bpp Mar 23 12:40:56 koen: this one: http://wenku.baidu.com/view/8426530f844769eae009edfc.html ? Mar 23 12:42:55 av500: yes, that's the one Mar 23 12:43:08 av500: koen has a more recent datasheet though Mar 23 12:43:43 its still boring 2003 stuff Mar 23 12:44:05 I'd settle for a datasheet I can actually read Mar 23 12:44:11 without pixelated diagrams Mar 23 12:44:19 use the baidu one Mar 23 12:50:28 don't recall the baidu one being all that better, hard to tell between your 6/8/9's Mar 23 12:51:36 I'm trying to access memory at 0x481A0000 (SPI1 config) but it fails although I can easily access GPMC config memory or UART config memory, is there something special about McSPI config registers? Mar 23 12:57:06 Basically I'm trying to make sure that my SPI is in master mode along with other configs Mar 23 12:58:13 From the TRM (spruh73c) figure 24-1 It shows that D0 should be an output if the SPI is in master mode Mar 23 13:07:00 jsabeaudry: AIUI you can choose what d0 and d1 are Mar 23 13:10:18 koen, yup I confirmed that in the TRM but it seems I need to enable something to be able to access the McSPI configuration registers, if I was on OMAP3530 it would be "McSPI functional (CM_FCLKEN1_CORE) and interface (CM_ICLKEN1_CORE) " but I haven't found the equivalent yet Mar 23 13:12:40 ah perhaps the CM_PER_SPI1_CLKCTRL I'll try it out Mar 23 13:16:16 I have a hardware problem, is this the right place to ask about it? Mar 23 13:16:33 depends on the hardware Mar 23 13:16:52 and the problem Mar 23 13:17:55 I've had my beagleboard for two days. This morning I plugged in the power and just got a red light, and it's been that way since. Mar 23 13:23:59 any ideas why it might do that? Mar 23 13:26:04 (and how I can make sure not to do it in the future)? Mar 23 13:27:11 bad voltage Mar 23 13:31:46 it's the adapter that came with the board though. Mar 23 13:35:35 is it a beagle-xm? Mar 23 13:36:12 Yes. Mar 23 13:36:33 (beagle classic doesn't have the overvoltage protection) Mar 23 13:36:48 but neither model comes with a psu Mar 23 13:37:07 where did you buy it? Mar 23 13:37:19 Digikey sells a power adapter, which is listed alongside the beagleboard Mar 23 13:37:30 ok Mar 23 13:37:47 are you sure you didn't plug in some other psu by mistake? Mar 23 13:37:56 I only have the one. Mar 23 13:38:20 It wasn't fully plugged in, so I pushed it all the way in, and now I get the red light. Mar 23 13:38:28 then check the voltage with a meter Mar 23 13:38:56 ok Mar 23 13:39:04 what should it be, 5 volts? Mar 23 13:39:07 yes Mar 23 13:39:32 ok Mar 23 13:39:53 will do that this afternoon when I have access to a meter Mar 23 13:45:08 Is it normal that on my first SPI transmission the CM_PER is reset along with all my settings? Mar 23 14:10:45 hello Mar 23 14:10:56 question: Mar 23 14:11:29 Is beaglebone kernel/linux compiled with v4linux Mar 23 14:11:37 or video4linux Mar 23 14:11:40 Thanks Mar 23 14:22:52 Is beaglebone kernel/linux compiled with vedo4linux Mar 23 14:23:25 zcat /proc/config.gz and find out Mar 23 14:23:33 depends on the kernel Mar 23 14:24:29 koen correct me if i'm wrong but usualy it should be compiled with it Mar 23 14:27:58 compiled with what? Mar 23 14:28:17 video4linux Mar 23 14:30:22 15:23 < koen> zcat /proc/config.gz and find out Mar 23 14:30:31 I havent recieved the board yet Mar 23 14:31:10 I was planning on buying usb cam in mean while, so I wanted to ask if it does support v4linux, Mar 23 14:31:35 if not is it easy to comile kernel for the board? Mar 23 14:31:51 for those who can read: yes, IMHO :) Mar 23 14:32:08 what if you can't read? Mar 23 14:32:23 prpplague: then its not easy. :P Mar 23 14:33:21 for compiling the kernel being able to type should be enough Mar 23 14:33:22 you could use a screenreader Mar 23 14:33:28 you talking to me leto Mar 23 14:33:28 it would read it for you Mar 23 14:34:00 ogra_: screenreaders don't count, they are just format converters. Mar 23 14:34:40 indeed, but it helps as long as you can *hear* you dont need to be able to read Mar 23 14:34:52 if you cant hear either you are lost indeed Mar 23 14:36:03 as i said. ability to read, be it using what ever format conversion you nedd/prefer. :P Mar 23 14:36:22 he still can watch youtube tutorials...but only if his kernel was compiled with video4linux :) Mar 23 14:36:47 hi, what it the meaning of omapfb.vram=0:8M,1:4M,2:4M in uEnv.txt? Mar 23 14:38:02 stuk_gen: does this help http://www.elinux.org/BeagleBoard/video Mar 23 14:38:40 Hi All, I'm new to the beaglebone... I have it booted up and attached to the network and I can ssh to it and iotherwiise treat it like another server on the network. My task it to create a custom image that can be distributed. I am going to use bitbake to compile the kernel and the console image and after that I want to add some things like rvm (ruby version manager), RUby, a web server (nginx probably) and a ferw other thi Mar 23 14:39:26 you might want to limit your sentences to 200 chars or less ;) Mar 23 14:39:39 mru: your town is on slashdot Mar 23 14:39:50 poor slashdot. Mar 23 14:41:03 So can I build the essential and install my add-on and then duoplicate that? Mar 23 14:41:12 thats good to hear kelleysislander how can we help you? Mar 23 14:41:25 apparently lots of illegal hobbits there Mar 23 14:41:37 Thanks for helping with the rules! I'm a noob here so thabnks again! Mar 23 14:41:42 nasty hobbitses? Mar 23 14:42:04 Ha Ha !@@ Mar 23 14:42:43 sure kelleysislander you can build your image and dd it Mar 23 14:42:43 I am tasked with building a custom distr which will have a web server, ruby, the padrino web framework Mar 23 14:43:11 I have a bitbake recipe for some special device drivers that need to be baked in. After that... Mar 23 14:43:17 you might even take one of the angstrom demo images and just install the needed packages Mar 23 14:43:23 koen: about what? Mar 23 14:43:44 oh ok then forget about the demo images :) Mar 23 14:43:48 can I simply add my extras and crate a single bootable image? Mar 23 14:43:49 mru: a local pub is getting sued because it's called "the hobbit" Mar 23 14:43:53 oh that Mar 23 14:43:56 that's old news Mar 23 14:44:05 I said 'slashdot', didn't I :) Mar 23 14:44:34 why "forget the demo image"? Mar 23 14:44:44 they're keeping the name Mar 23 14:44:57 if you need your own drivers in the kernel you need to bitbake it Mar 23 14:44:57 and they were not sued Mar 23 14:45:16 the lawyers sent them a somewhat threatening letter Mar 23 14:45:18 yes I have an oebb.sh for that Mar 23 14:48:41 so then go ahead and start baking :) Mar 23 14:50:13 So after bitbaking I can add my other stuff and then crteate the bootable image in one shebang? Mar 23 14:51:10 you bitbake the kernel with virtual/kernel and then you bitbake your image (e.g. console-image) Mar 23 14:52:34 OK, have that done... Mar 23 14:53:02 Then aftet that boot up on that image and install my extras on top of that??? Mar 23 14:53:21 koen: I'm not getting anything with your timings, how do you test? Mar 23 14:53:36 * mdp has to wonder if kelleysislander is local to him Mar 23 14:53:59 I'm in Dallas Mar 23 14:54:03 ok :) Mar 23 14:54:11 and you? Mar 23 14:54:17 that could be one side of the road...you can also bitbake you apllications ins the bitbake process Mar 23 14:54:19 ohio Mar 23 14:54:30 I'm from Ohio Mar 23 14:54:38 and am going to kelleys island tomorrow Mar 23 14:54:44 kelleysislander: ok, so you know why I asked then :) Mar 23 14:54:58 cool, great weather right now, as you probably know Mar 23 14:55:02 martrom how would I bitbake my extras into the process? Mar 23 14:55:17 mdp - it's been warmer there than here ha ha Mar 23 14:55:54 mdp yep us Ohioans stick together !@@ Mar 23 14:55:57 kelleysislander: yeah, everybody with a "seasonal vehicle" has them out already..it's a bit unusual to be rockin the 83F in mid-march :) Mar 23 14:56:24 Martron - is one approach betther than the other? Mar 23 14:57:03 http://www.armbedded.eu/documentation/html/linuxguide/ch07s03.html Mar 23 14:57:29 kelleysislander: same shit :) Mar 23 14:57:39 So my overall approach is to bitbake the kernel and console image, also need ssh so I can connect... Mar 23 14:58:44 martron thanks for that excellent resource Mar 23 14:59:42 always depends on how the future use will be....if you just have to create ONE image, do it your way...if you have to keep that image on a current kernel all the time you might want to just fire up one bitbake of your own recipie Mar 23 14:59:47 OK, that's great fr now, my fortst task is to bitbake kernel and console image. Then with that success under the belt add extras. Mar 23 15:00:25 one very fist step could be just to use the demo images Mar 23 15:00:35 and put your applications there Mar 23 15:00:42 How might I concoct a recipe for something custom such as "RVM - Ruby Version manager"? Mar 23 15:00:55 depends on how fast you need to prototype it Mar 23 15:00:59 Martron - sounds like a good approach Mar 23 15:01:26 I had a fellow do it initially and the first compile took 12 hrs sunsequent ones were less die to compiler caching Mar 23 15:02:32 OK, baby steps first - I will do the kernel and console and the specific oebb.sh from the client and see how that works. Mar 23 15:02:38 i dont know about ruby but you might check at angstrom what they already have http://www.angstrom-distribution.org/repo/ Mar 23 15:03:27 On RUby I have successfully compiled and am using a ruby on the BB... pretty cool it really didn't take much more time to compile than on my i7 MB Pro! Mar 23 15:04:18 blame apple Mar 23 15:04:31 use linux :) Mar 23 15:04:43 Martron FYI: the "rvm" program makes compiling rubies super easy. I tried "opkg install ruby" and it was not found so now I am upgrading to see if it shows up Mar 23 15:05:07 do a opkg udate first Mar 23 15:05:18 I had the compile done on an Ubuntu box which is what I will be suiong, just usiong the mac for comparison Mar 23 15:05:24 you did a opkg upgrade?? Mar 23 15:05:40 yes it is still running lots of packages!@@ Mar 23 15:05:53 hopfully it will boot afterwards Mar 23 15:06:01 Ha Ha! Mar 23 15:06:04 no kidding Mar 23 15:06:13 yeah ... gulp! Mar 23 15:06:32 there were problems with the systemd not doing the post installation steps Mar 23 15:06:48 took me quite a week to find that out Mar 23 15:06:53 yikes Mar 23 15:07:12 koen might has fixed that already :) Mar 23 15:07:29 you'll find out soon Mar 23 15:07:40 so if it doesn't that what? Mar 23 15:07:54 no way to revert? Mar 23 15:08:13 i put the demo image on my card Mar 23 15:08:29 you can upgrade the kernel though Mar 23 15:10:07 still running the upgrade the outout is nine liles long... Mar 23 15:10:48 removing the git stuff now Mar 23 15:12:34 Martron: If this ends up hosed where to get a new demo image? Mar 23 15:13:35 http://www.angstrom-distribution.org/demo/beaglebone/ Mar 23 15:16:01 Martron: Oh yeah, thanks Mar 23 15:17:58 Martron: Oh yeah, I'm on the beaglebone not beagleboard if that makes a diff... Mar 23 15:18:11 koen: gsoc rejection meeting? Mar 23 15:18:59 this is why i pointed you the beaglebone demos :) Mar 23 15:20:15 Martron: Yep, thanks for that Mar 23 15:38:02 What kind of thing could cause a functionnal clock to get disabled? Mar 23 15:38:17 (on the am335x) Mar 23 15:38:29 you didnt wind it up in time ? Mar 23 15:40:02 its called dysfinctional clock now Mar 23 15:44:40 av500: where, when? Mar 23 15:45:07 15 min in #gsoc Mar 23 15:45:08 icota: I send the uImage to texas and they boot and report back :) Mar 23 15:45:18 icota: note that backlight is turned off Mar 23 15:45:24 koen: is that a BAAS? Mar 23 15:45:28 koen: i did turn it on Mar 23 15:48:32 jsabeaurdry, you have a PM enabled am335x kernel and the block is being clock gated? Mar 23 15:48:44 * mdp throws out one possibility Mar 23 15:50:19 mdp, what is "PM enabled"? Mar 23 15:51:41 power management Mar 23 15:53:59 To be more specific, CM_PER_SPI_CLKCTRL gets reset when I pipe to /dev/spidev2.0 Mar 23 15:54:29 what kernel are you using? Mar 23 15:54:39 icota: I've seen the 800x480 timings work on the lcd, but nothing else :( Mar 23 15:55:33 koen: is the framebuffer 800x480 as well? i'm using your latest kernel from github and getting nothing Mar 23 15:57:03 mdp, I'm using kernel for koen's repo with a patch for spidev Mar 23 15:57:12 s/for/from/ Mar 23 15:58:03 ok, cool, wanted to be sure I'm looking at the same thing Mar 23 15:58:40 we have enough kernels to rival the number of flavors at baskin-robbins Mar 23 16:00:24 icota: yes, have a look at the lcd7 setup routines in the boardfile Mar 23 16:00:46 jsabeaudry: CM_PER_SPI2_CLKCTRL, I imagine? Mar 23 16:01:29 CM_PER_SPI1_CLKCTRL Mar 23 16:01:46 er, yes, due to the kooky mcspi instance numbering Mar 23 16:01:46 gotcha Mar 23 16:06:35 could spidev possibly reset that stuff? Mar 23 16:07:09 not directly Mar 23 16:07:46 however, it can happen indirectly via the omap clock management framework (tied to hwmod) Mar 23 16:07:53 looking at it now Mar 23 16:08:51 jsabeaudry: since you mention that CM_PER_SPI1_CLKCTRL is reset, what are the contents before/after you access spidev2.0? Mar 23 16:09:58 I set it to 0x2, after echoing to spidev the value is 0x30000 Mar 23 16:11:49 koen, can you confirm that your branch is one that pulled in the PSP PM stuffs? Mar 23 16:11:59 IIRC, you mentioned that Mar 23 16:12:11 jsabeaudry: ok, thx.. Mar 23 16:16:33 jsabeaudry: in normal operation you should not have to touch clocks regs directly, this should happen automagically for you Mar 23 16:16:45 mdp: it has PM bits Mar 23 16:16:47 so you're seeing the management framework also writing to it obviously Mar 23 16:16:55 mdp: but proper PM requires the secret m3 firmware Mar 23 16:17:22 koen, is that pretty much the latest stuff? I know there's some system test on a "formal release" happening now..wondering if it's current with what's in those arago trees Mar 23 16:17:27 koen, ahh :) Mar 23 16:17:37 s/secret/sekrit/ Mar 23 16:17:56 all TI parts have to have sekrit m3 firmware for sexy features Mar 23 16:17:58 so right now it's non working cpuidble, broken cpufreq and suspend needs that firmware Mar 23 16:18:01 no idea on clock gating Mar 23 16:18:31 mdp, What is the normal way of setting SPI configuration, for example setting D0 to be the output pin like in figure 24-1 of the TRM ? Mar 23 16:18:55 I could not find another way of setting SPI configs Mar 23 16:18:59 well, my recommendation is to ignore figure 24-1 since it's misleading Mar 23 16:19:40 it's showing one possible use case rather than how the spi-omap2-mcspi.c driver is configured Mar 23 16:19:41 I need transmit only mode on pin D0 Mar 23 16:19:53 that's currently hardcoded in the mcspi master driver :-/ Mar 23 16:19:57 ah that is the file I need to modify? Mar 23 16:20:02 yep Mar 23 16:20:18 well that should do it Mar 23 16:20:24 I'll take a look Mar 23 16:20:32 thanks a lot for your help Mar 23 16:20:56 jsabeaudry, I'll take a further look at this clock issue, that's a bit disturbing…though I'm PM dumb and certainly know nothing about the status of am335x kernel work :) Mar 23 16:21:49 could the driver you mentionned simply reset all registers to hard coded values? Mar 23 16:21:59 perhaps it's not a matter of PM Mar 23 16:23:05 well, my take is that somebody is calling clock disable on the spi1_fck for some reason Mar 23 16:24:05 the reason I asked about PM..is that in the world of runtime PM (dunno how comprehensive our implementation is), that can happen when the driver is initialized but has no work to do... Mar 23 16:24:15 so perhaps there's an implementation bug Mar 23 16:24:21 all speculation atm Mar 23 16:24:31 uneducated guesses :) Mar 23 16:25:12 jsabeaudry: take a look at omap2_mcspi_setup_transfer(), go whack DPE0/DPE1 the other way Mar 23 16:25:43 I have note here to submit a patch to add a platform data option for this Mar 23 16:26:25 but, for my own stuff, it's easy to KISS and just swap D0/D1 wiring :) Mar 23 16:26:26 oh dear, sparkfun is selling 12V carhorns Mar 23 16:26:34 which draw up to 8A Mar 23 16:26:47 koen: use a level shifter Mar 23 16:27:07 level shifters ftw Mar 23 16:27:20 a TI level shifter of course Mar 23 16:27:33 not a cheap FPGA like we do .. cough... Mar 23 16:31:41 mdp, great, I'm recompiling, can't wait to see if this will work, so quick compared to the hours I spent trying to change registers Mar 23 16:33:20 av500, yes, TI level shifter, my family needs to eat Mar 23 16:33:59 mdp: eat level shifters Mar 23 16:34:28 jsabeaudry: just to clarify, you *are* finding that the PER reg is reset before you go enable it? Mar 23 16:34:41 mdp: it's funny to get reports from people saying the bone defconfig needs to turn of the thumb2 options Mar 23 16:34:51 mdp: it's really good at finding broken toolchains :) Mar 23 16:35:05 mdp, you mean at boot time or after echoing to spidev? Mar 23 16:36:04 koen, that's like people complaining that ubifs breaks their nand driver :) heard that a few times…perspective is everything Mar 23 16:36:42 jsabeaudry: after boot…before you wrote 0x2 to enable it… Mar 23 16:37:10 I mean, you did do that because it was off to start, correct? Mar 23 16:37:39 * mdp may try this locally anyway Mar 23 16:38:18 mdp, Yes it is reset at boot, and yes I first needed to enable it because I couldnt access SPI config Mar 23 16:39:59 jsabeaudry: how does it feel to be an alpha tester? Mar 23 16:40:33 koen, are you running other spi kernel drivers on this same kernel? Mar 23 16:42:56 mdp, Can be frustrating at times but overall I'm having a good time with the bone Mar 23 16:43:17 btw, the modification to the hardcoded driver worked Mar 23 16:43:42 time to load up some code on that Spartan 6! Mar 23 16:46:04 excellent! can I go home now? :) Mar 23 16:47:29 :) Mar 23 16:50:39 jsabeaudry: I try to think of bone as having a lot of unrealized potential Mar 23 16:51:21 that's the management friendly suggested wording Mar 23 16:51:30 hahahaha Mar 23 16:51:54 It's small, relatively cheap, lots of IO, for me it does the job Mar 23 16:52:39 fits in well with the msp430 and arduino project space Mar 23 16:53:40 eh? Mar 23 16:53:47 the msp430 is in a class by itself! Mar 23 16:54:02 or has the PM gotten good enough to last 2-3 years as a RTC on a CR2032? Mar 23 16:54:06 indeed Mar 23 16:54:24 just from the hobbyist interfacing perspective Mar 23 16:54:30 blah ;) Mar 23 16:54:38 I agree, I like to use 430 as remote sensors Mar 23 16:54:55 especially since TI more or less gives them away Mar 23 16:55:01 that certainly helps Mar 23 16:55:53 but really, nothing compares to R-Pi Mar 23 16:56:03 the greatest of them all Mar 23 16:57:07 * mdp sometimes wonders if he has to use sartalics or not Mar 23 16:57:22 :) Mar 23 16:57:34 such an open and well documented platform Mar 23 16:57:40 I saw a kid on the street today wearing a t-shirt with the text "believe the hype" Mar 23 16:57:45 surely, it puts the silly pvr to shame Mar 23 16:57:47 yep Mar 23 16:57:47 lol Mar 23 16:58:06 I wish the msp430's did composite out Mar 23 16:58:26 shame to use an omap for what amounts to be a clock driving a composite LCD Mar 23 17:00:01 but then I guess the chumby does set a new baseline Mar 23 17:02:12 I'm looking forward to the spidev and i2c-dev tutorials for R-Pi Fedora Mar 23 17:06:10 mdp: think about the children! Mar 23 17:07:10 yep, I heard they don't go into science and technology because they haven't had a R-Pi in their hands :) Mar 23 17:07:27 yes Mar 23 17:07:31 it's the missing link Mar 23 17:07:33 no access to computers Mar 23 17:07:49 except their iphones Mar 23 17:07:54 on PS3s Mar 23 17:07:56 and Mar 23 17:08:04 and gaming rigs Mar 23 17:08:11 no access at all Mar 23 17:08:11 its cruel Mar 23 17:08:36 some day computers will be pervasive and allow them to pursue their science and technology career dreams Mar 23 17:08:56 yes, we can dream Mar 23 17:18:09 koen, I see the binary blob stuff now. I feel like I'm at pfefferz' talk now Mar 23 17:18:52 what, where? Mar 23 17:19:37 was just reading pm33xx.c Mar 23 17:20:28 blob? is that the cortex-m firmware? Mar 23 17:20:45 yes Mar 23 17:21:35 it's been a while since I skimmed that part of the TRM..but IIRC..all the m3 does is handle wakeup events from deepsleep..something along those lines Mar 23 17:22:03 ds2, kernel loads cm3-firmware.bin to run on the m3 Mar 23 17:22:19 I wonder if that is encrypted... Mar 23 17:22:48 hopefully Mar 23 17:22:58 like the PRU, this is to complicated for anybody outside of TI to understand Mar 23 17:23:14 rot26'ed? Mar 23 17:24:06 crap, you broke through our security measures! Mar 23 17:24:32 mdp: when I asked about the m3 firmware answers ranged from "what m3? There's no m3!" to "The firmware isn't needed, it will just delay boot by 60s and break suspend" Mar 23 17:25:10 uhhhhh, ok Mar 23 17:25:12 mdp: I suspect only a handfull of people know about the m3 firmware and even less have the actual blob Mar 23 17:25:21 then we can remove all the code Mar 23 17:27:17 I sure hope that's just some miscommunication, cause the M3 is fundamental to am335x pm transitions :) Mar 23 17:28:10 mdp: you know the deal with m3 firmwares Mar 23 17:28:25 mdp: rule #1: don't talk about m3 firmwares Mar 23 17:28:26 8.1.1.4.3 and 8.1.1.4.6 for those following along Mar 23 17:28:30 * mranostay hands av500 a tofu turkey Mar 23 17:28:49 is eating fake meat ethical? :) Mar 23 17:29:05 koen, am I going to be kicked out of M3 Firmware Club now? Mar 23 17:29:36 mranostay: is eating meat prepared to taste like vegetables ethical? Mar 23 17:30:05 save a cow, eat a vegetarian Mar 23 17:31:12 we should domesticate vegetarians to eat Mar 23 17:31:33 the correct mixture of pasture and grain feeding to get the ideal marbling in the meat Mar 23 17:31:34 mmm Mar 23 17:31:49 koen, you are thinking outside the box today! Mar 23 17:32:49 yes Mar 23 17:33:16 but I will try to visit TXLC next week Mar 23 17:33:52 or ruth's chris, but that never happens Mar 23 17:35:52 bone daddies? Mar 23 17:36:19 RC does a nice job Mar 23 17:36:59 I feel so out of touch when reading the beagle ml Mar 23 17:37:05 lol Mar 23 17:37:07 something about a new "layout" Mar 23 17:37:31 took me a while to figure out that some people use the google groups webpage as their primary means of access Mar 23 17:37:41 but you are Boris' communicator to the people!? Mar 23 17:37:53 oh yeah, I couldn't believe that thread Mar 23 17:38:20 I added it to my shocking revelations list…like Santa Claus.. Mar 23 17:39:37 koen, maybe you guys should change to phpbb so people can read on Tapatalk :) Mar 23 17:55:46 mdp: TIL people still use phpbb Mar 23 18:08:24 so i did a loop on the beagle-xm using arecord -f S16_LE -r 48000 | aplay - (who said the twl4030 generated poor sound ? imho it was suprizingly good) Mar 23 18:09:16 koen: short story is google is broken. avoid it and you'll be happier Mar 23 18:13:10 i don't think i've used Google Groups in like 5 to 7 years Mar 23 18:13:25 it isn't just that...all their shit is broq Mar 23 18:16:41 jsabeaudry: that clock behavior you are seeing is expected…when you write to spidev, the driver (by way of runtime pm hooks) will flip the fck on, do xfers in the workqueue, and then flip fck off Mar 23 18:17:11 ds2: I get the email by email, so I never look at the interface :) Mar 23 18:17:31 koen: so do I but everything else is broken..maps, search, etc Mar 23 18:17:44 jsabeaudry: before you ever write to spidev, it's also expect to have the fck off because after the device is registered (on boot, if you are using spi_board_info, clocks are disabled to spi Mar 23 18:18:19 jsabeaudry: if the fck is not getting automagically enabled by access to spidev, then that's a bug. Mar 23 18:19:01 mdp, is it also normal that my settings are reset by that operation? Mar 23 18:19:18 yes Mar 23 18:19:43 ok then everything is explained Mar 23 18:19:45 great Mar 23 18:19:49 because you're not expect to be banging directly on those PER registers, there's an entire kernel framework to manage the clock and power domains Mar 23 18:20:11 but wait, I thought it was non-functional uncless you explictly forced it to 0x2? Mar 23 18:20:37 oh no the spidev works, accessing registers was non functional Mar 23 18:21:24 and still is, but that is by design from what I could understand in what you just said Mar 23 18:21:58 jsabeaudry: are you trying to swap d0/d1? Mar 23 18:22:17 the adequate solution is to have a platform data i would guess Mar 23 18:22:22 ds2, yes I did Mar 23 18:22:43 jsabeaudry, ok, that PER clock controls the entire IP block…has nothing to do with d0/d1 Mar 23 18:22:45 jsabeaudry: are you making it configureable via platform data? Mar 23 18:23:13 ds2, no I just changed the hard coded values in the driver Mar 23 18:23:38 oh :( Mar 23 18:23:48 ds2, I'll post a patch for that in a bit Mar 23 18:23:54 * mdp runs an errand Mar 23 18:23:55 mdp: which list? Mar 23 18:31:05 Is it possible to write to a cape eeprom using the default sysfs driver or is that simply for reading? Mar 23 18:41:03 hello Mar 23 18:41:07 ? Mar 23 18:46:39 wow that guy stayed a whole two minutes Mar 23 18:46:50 * mranostay hides from jkridner Mar 23 18:46:58 oh? Mar 23 18:47:04 have you been bad? Mar 23 18:47:17 no just me :) Mar 23 18:47:45 * mranostay blames av500 Mar 23 18:47:51 I always do. Mar 23 18:48:12 you are in the Bay Area, right? I'll be out there Monday. Mar 23 18:48:36 jkridner: you are ESC'ing? Mar 23 18:48:41 yup. Mar 23 18:48:50 presenter or exhibitor? Mar 23 18:48:55 exhibitor. Mar 23 18:49:14 My only presentation is a 25-minute luncheon thing. Mar 23 18:49:25 is there an associated tech day this year? Mar 23 18:49:39 I'm not sure. Not that I'm scheduled to be at…. Mar 23 18:49:55 I'm only exhibiting at DesignWest/ESC. Mar 23 18:50:14 *nod* but the talks and stuff are better at the techday side in the previous years Mar 23 18:50:20 trying to figure out which day am I going Mar 23 18:50:31 yeah not going to ESC that I know of Mar 23 18:51:48 I think I really want to figure out how to share my Internet connection to my BeagleBone but I'm not sure what to try next. Mar 23 18:52:21 usb network gadget? Mar 23 18:52:24 via what medium? Mar 23 18:52:27 that is, over the USB, rather than having to hook up wifi. Mar 23 18:52:28 jkridner: did you plug the power in ;) Mar 23 18:52:29 yeah. Mar 23 18:52:55 this is from an ICS phone or? Mar 23 18:53:18 mac laptop has wifi. Mar 23 18:53:45 I'll probably just get lazy and plug a USB wifi dongle I have here in my bag into my BeagleBone.... Mar 23 18:53:47 does macs support NAT? Mar 23 18:53:51 Hey, anyone know what kernel arguments actually ARE? Mar 23 18:53:56 if so, that's just a few lines Mar 23 18:54:02 it does, but I'm never able to get it to work. Mar 23 18:54:08 XMPPwocky: it argues with the kernel. Mar 23 18:54:35 ds2: well, i know what they /are/ Mar 23 18:54:38 but where do they go? Mar 23 18:54:39 XMPPwocky: way to be vague Mar 23 18:54:45 tried a couple of times and the settings always get a bit wonky as I try new things and packets never move. Mar 23 18:54:54 mranostay: in memory Mar 23 18:55:17 if I load a binary image to 0x82000000 and then "go 0x82000000 foo bar Mar 23 18:55:23 Hey, anyone know what kernel arguments actually ARE? Mar 23 18:55:30 that is exactly what you asked. Mar 23 18:55:33 ahem uEnv.txt Mar 23 18:55:41 NAND Mar 23 18:55:44 :) Mar 23 18:55:46 uImage Mar 23 18:56:12 26 + 26 = 54! Mar 23 18:56:51 so I have to read that file just to get the arguments? Mar 23 18:57:13 jkridner: you have to get out and push the packets a bit to get them started :) Mar 23 18:57:31 are you saying the macs are not primed? Mar 23 18:58:01 yeah you have to press that little bubble thing twice Mar 23 19:06:43 ds2: okay, so just to get the boot arguments, I have to implement filesystems, NAND access, etc? Mar 23 19:08:13 WTF are you doing? Mar 23 19:08:45 writing a kernel. Mar 23 19:12:23 why are you writing a kernel? linux not good enough? Mar 23 19:12:30 learning about ARM Mar 23 19:13:36 then prehaps reading more FMs are in order Mar 23 19:14:30 i just can't figure out WHERE boot args go Mar 23 19:15:18 anywhere you like Mar 23 19:15:23 since you are writing a kernel Mar 23 19:15:48 but where does u-boot put them Mar 23 19:16:22 read the u-boot code if you to use u-boot Mar 23 19:20:00 Is thare a tool chain for beaglebone, suppose I want to compile the kernel for the board? Mar 23 19:22:52 seems I saw that yesterday on a link off beagleboard.org Mar 23 19:22:59 don't have a bone myself, but I assume you can use OE/bitbake with the angstrom repo just configure it for the bone Mar 23 19:23:18 under "Getting Started" Mar 23 19:23:29 probably Mar 23 19:23:51 Thanks, I will look into it. Mar 23 19:50:31 <_av500_> XMPPwocky: uboot passes boot args in some memory location Mar 23 19:53:16 R1 needs to match the kernel Mar 23 19:53:21 err Mar 23 19:53:25 platform Mar 23 19:53:52 unless arm has gone all DTS now ? Mar 23 19:54:32 jsabeaudry: it is r/w Mar 23 20:16:31 <_av500_> unsolo_: its not a linux kernel Mar 23 20:16:55 oO Mar 23 20:16:57 i see Mar 23 20:18:18 DTS is the future! Mar 23 20:18:38 <_av500_> not dolby? Mar 23 20:18:43 haha Mar 23 20:18:50 Device Tree Structure.. Mar 23 20:18:54 i recon Mar 23 20:19:03 _av500_: aha, yeah Mar 23 20:19:07 Digital Telephone Service . . . Mar 23 20:19:23 the uImage format is involved Mar 23 20:19:33 XMPPwocky: make one using mkimage Mar 23 20:19:52 ye[ Mar 23 20:19:54 from any gzipped image i think Mar 23 20:21:02 <_av500_> unsolo_: again, its not linux Mar 23 20:21:15 _av500_: would it matter Mar 23 20:21:22 <_av500_> uboot should be able to start an uncompressed binary too Mar 23 20:21:28 it can Mar 23 20:21:29 <_av500_> so gzip is not needed Mar 23 20:21:39 simplu load it into ram and do go instead of bootm Mar 23 20:22:02 eg fatload mmc 0 0x80000000 mbinary Mar 23 20:22:07 go 0x80000000 Mar 23 20:22:09 boots it Mar 23 20:22:33 <_av500_> or it gets the hose again Mar 23 20:23:10 pain that u-boot doesnt handle zimage and elf Mar 23 20:23:20 <_av500_> teach it Mar 23 20:23:21 i guess you can probably build it in if you want Mar 23 20:25:43 if you do a load using openocd at the point where u-boot stops you can also load binaries Mar 23 20:25:52 since ram etc is configured Mar 23 20:29:40 unsolo_, zimage support patches have been posted to the u-boot list.. ;) Mar 23 20:31:14 Is zImage support good? Mar 23 20:31:22 How are we compressing now? bz? Mar 23 20:31:59 SilicaGel, honestly, zImage support just removes the need for the end user to run mkimage to build a uImage, other then that it's pretty much exactly the same.. Mar 23 20:32:27 ah ok. that doesn't sound like a big deal to me, running mkimage wasn't hard Mar 23 20:32:39 It was pretty difficult when I didn't have mkimage installed Mar 23 20:32:45 once i figured that out, got easier Mar 23 20:32:47 :) Mar 23 20:33:07 it's a bigger deal for non-omap silicon that has differenet memory relocation, so they can't build a normal uImage.. Mar 23 20:33:22 ahhh Mar 23 20:33:40 a mx51 uImage won't boot on a mx53 target. ;) Mar 23 20:34:07 this is even better than the hotelicopter http://tacocopter.com/ Mar 23 20:36:09 but is it better than https://lh6.googleusercontent.com/-ciYekX2RMMU/TqnnyvWIXdI/AAAAAAAAAFE/jwk4A6A_soI/s1131/IMG_20111027_204209.jpg ? Mar 23 20:36:31 I don't know, that train full of beer is pretty awesome Mar 23 20:37:07 Imagine if that train was 1:1 scale Mar 23 20:37:34 then those would be very poor containers for such a quantity of beer Mar 23 20:37:59 <_av500_> mru: remember, that train even derailed Mar 23 20:38:19 I do remember Mar 23 20:44:51 * unsolo_ wonders if he needs a psvita Mar 23 20:45:18 you are making me thirsty Mar 23 20:45:38 <_av500_> djlewis: you have a point there Mar 23 20:45:50 <_av500_> and i have beer here Mar 23 20:46:13 is it beer o'clock Mar 23 20:46:15 ? Mar 23 20:46:15 pint Mar 23 20:55:48 um beer soon Mar 23 20:56:02 _av500_, do you have a link to your partition table tester handy Mar 23 20:56:10 HokieTux, is going nuts Mar 23 20:56:59 <_av500_> google "omap3 boot mystery" Mar 23 20:57:21 openbts|Crofton: what happened to you? did you step through that mirror into looking-glass land? Mar 23 20:59:04 ah thanks Mar 23 23:01:36 aha! Mar 23 23:01:49 "r6: start of kernel command line" Mar 23 23:01:58 and r7 is the end Mar 23 23:04:33 <_av500_> see Mar 23 23:07:35 oh, wait Mar 23 23:09:23 it's on the stack for arm Mar 23 23:11:27 <_av500_> and in r6 for x86? Mar 23 23:11:35 :) Mar 23 23:11:46 for ppv Mar 23 23:11:47 ppc Mar 23 23:15:16 yeah, linux documents uboot better than uboot :/ Mar 23 23:15:57 <_av500_> any bootloades has to implement it the same way Mar 23 23:16:26 _av500_: nope Mar 23 23:16:48 _av500_: uboot feels special. Mar 23 23:17:23 linux has CONFIG_UBOOT just for it Mar 23 23:17:58 <_av500_> ah Mar 23 23:18:21 but naked images are so much more fun Mar 23 23:20:36 ds2: i took one look at emif config Mar 23 23:20:49 and immediately learned to love uboot Mar 23 23:21:36 ~20 registers that are poorly documented and must be correct Mar 23 23:21:56 <_av500_> setting up emif is not rocket surgery Mar 23 23:22:45 provided you have the dram data sheet Mar 23 23:23:54 sure, but then it's somehow even LESS portable :p Mar 23 23:24:50 <_av500_> portable? Mar 23 23:25:19 <_av500_> dram is not portable, its soldered down Mar 23 23:25:24 <_av500_> (usually) Mar 23 23:25:37 <_av500_> where it is portable, there is a lil eeprom with its config Mar 23 23:25:51 of course the eeprom is frequently wrong Mar 23 23:25:55 at least on pc ram Mar 23 23:26:04 <_av500_> of course Mar 23 23:26:12 <_av500_> murphy etc.. Mar 23 23:26:52 and for pc ram, the data sheet is usually also wrong, or at least overly optimistic Mar 23 23:33:28 emif is easy Mar 23 23:36:10 the SDRC is a more complicated beast Mar 24 00:59:57 hello, i am trying to connect FTDI usbserial FT232R to BBoard.. http://pastebin.com/AEdXTsYD Mar 24 01:00:41 i loaded related modules and got the message that it is connected..now. but i cannot find ttyUSB0 Mar 24 01:01:54 anybody has an idea? i did install kernel-module-ftdi-sio by opkg.. Mar 24 01:02:51 does it right? Mar 24 01:04:07 FTDI usbserial should come up somewhere in /dev, isnt it? Mar 24 01:04:45 maybe i am failing to find it. or it is not working.. Mar 24 01:12:56 any idea? :( **** ENDING LOGGING AT Sat Mar 24 02:59:58 2012