**** BEGIN LOGGING AT Sun Oct 28 03:00:03 2012 Oct 28 03:11:45 ICE on stdarg.h Oct 28 03:11:47 arrrrrrrgh Oct 28 03:12:09 * alan_o bangs head Oct 28 03:23:48 whoa - tsunami warning due to a 7.1 quake off the coast of B.C. Oct 28 03:58:09 emeb: get out the surf board? Oct 28 03:58:34 mranostay: any chance you're on the 3.7 kernel on bone? Oct 28 03:58:52 yes Oct 28 03:59:00 something up? Oct 28 03:59:01 so anything special there? Oct 28 03:59:04 yeah, won't boot :) Oct 28 03:59:09 I'm running old u-boot Oct 28 03:59:12 should that matter? Oct 28 03:59:22 where does it fail? Oct 28 03:59:27 U-Boot SPL 2011.09-00010-g81c8c79 (Feb 13 2012 - 14:48:03) Oct 28 03:59:36 Starting kernel ... Oct 28 03:59:42 no output after that Oct 28 03:59:48 using the dtb uImage? Oct 28 03:59:54 I'm using my same stuff to boot as I did for 3.6 and 3.5 Oct 28 04:00:00 those still boot no problem Oct 28 04:00:06 yes Oct 28 04:00:10 uImage with dtb glued on Oct 28 04:00:13 who's tree? Oct 28 04:00:20 koen's 3.7 Oct 28 04:00:32 branch? Oct 28 04:00:36 I wonder if the dtb gluing isn't working right...... Oct 28 04:00:37 the 3.7 branch Oct 28 04:00:40 stand by Oct 28 04:00:46 https://github.com/beagleboard/kernel/tree/3.7 Oct 28 04:01:31 i'm using https://github.com/koenkooi/linux/tree/3.7-for-panto-rebase Oct 28 04:02:06 hmmm Oct 28 04:02:12 the one I'm on was updated yesterday Oct 28 04:02:29 and I had to fix one of the patches (fuzz). Oct 28 04:02:44 I wonder if he didn't test the one that's on there right exactly now. Oct 28 04:02:51 but, I had this the other day Oct 28 04:02:59 couldn't boot with 3.7 Oct 28 04:03:11 I turned off thumb2_kernel and it got farther Oct 28 04:03:18 no clue. haven't built an imae in a week or so Oct 28 04:03:22 this is new toolchain too (codesourcery) Oct 28 04:03:29 alan_o: yeah i never use the thumb2 kernel crap Oct 28 04:03:45 I'm just building the stock config that he has checked in Oct 28 04:03:50 trying to follow the exact steps :) Oct 28 04:03:57 but Oct 28 04:04:05 the one exact step I haven't followed is u-boot version Oct 28 04:04:22 i'm using the u-boot that came on the SD card Oct 28 04:04:40 yeah, that's what I've got too Oct 28 04:04:42 haven't touched it Oct 28 04:04:47 I pull my kernels from tftp Oct 28 04:05:00 but u-boot is stock from when I bought it Oct 28 04:05:26 ok, so the link you sent is an actual kernel tree Oct 28 04:06:38 Looks like it's reflected from one of the patch.sh repos Oct 28 04:07:13 ok, I think I'm giving up for today. Thanks for the info Oct 28 04:58:58 hi emeb_mac Oct 28 05:05:33 mranostay: heya Oct 28 05:06:25 mranostay: how's things? Oct 28 05:19:13 not too bad Oct 28 05:23:26 getting psyched for ELC-E Oct 28 05:23:28 ? Oct 28 06:51:25 * mranostay rides the split Oct 28 07:00:15 yeeeeehaw Oct 28 07:00:41 sounds... uncomfortable. Oct 28 07:35:50 <_av500_> gm Oct 28 07:45:49 <_av500_> mranostay: yo Oct 28 08:16:37 <_av500_> winter tires me Oct 28 08:19:55 * dm8tbr enjoys the winter wonderland outside his window Oct 28 08:20:09 too bad I have to go to copenhagen and barcelona ;) Oct 28 09:50:02 summer tires tired Oct 28 13:17:44 * dm8tbr has boarded for CPH Oct 28 16:09:51 * dm8tbr arrives at bella center Oct 28 16:09:59 anyone here yet? Oct 28 16:11:15 they all went somewhere else. didn't you get the memo? Oct 28 16:14:07 aside from mru, he just got delayed Oct 28 16:38:59 tsk, the hotel 'TV' sucks. it offers an 'external device' option but doesn't react at all Oct 28 16:40:12 it works if you rip out the HDMI cable from the 'smart concierge menu though' and plug into that Oct 28 16:42:46 * dm8tbr goes to scout out some prime trolling locations for mru Oct 28 16:50:30 haha Oct 28 18:29:04 mranostay: I built that panto-rebase kernel, and still no love :( I think I need new u-boot. That's the next step Oct 28 18:39:14 alan_o: or your doing it wrong :) Oct 28 18:40:55 could be. I built 3.6 and 3.5 to its satisfaction though. Oct 28 18:41:17 What toolchain do you use? Oct 28 18:41:46 I just picked up a CodeSourcery one. had been using a ct-ng one that I built, but it ICEd on 3.7 Oct 28 19:22:12 alan_o: Koen's angstrom ones Oct 28 19:22:49 ok, I may check that out. Another data point anyway. building u-boot-2012.10 Oct 28 19:29:08 hmm, I get streaming: "wait_for_user_access Timeout" . Looking at the code, It says: This timeout definition is a worst-case ultra defensive measure against Oct 28 19:29:08 * unexpected controller lock ups. Ideally, we should never ever hit this Oct 28 19:29:08 * scenario in practice. Oct 28 19:29:16 arrrrgh Oct 28 19:29:32 nothing is ever easy..... Oct 28 19:37:22 hallo everybody Oct 28 20:17:34 any recommendations for small powered usb hubs? Oct 28 20:21:03 one that works? Oct 28 20:22:06 yeah, I am planning on upgrading from one that doesn't (fried it somehow) Oct 28 20:23:22 mranostay: Followed the exact steps. new u-boot. loading dtb from u-boot like the instructions. Nothing.... If you don't mind.. could you send me the git tag of the kernel you have, and the version of the toolchain download? I'm not sure what else to try at this point. Oct 28 20:23:32 I have a little robot with a beaglebone, and I am using the hub for wifi and a webcam, so I am looking for something smallish, the current hub I have is a little belkin 2 inch square thing Oct 28 20:26:19 alan_o: linux-3.7.0-rc2-58212-g20db468 Oct 28 20:28:39 mranostay: I don't have that tag. I guess they rebased it away.... The branch name does have "rebase" in it. Oct 28 20:28:51 oh well.. thanks anyway. Oct 28 20:35:43 ugh.... 12k/s on the toolchain download Oct 28 20:37:19 yeah koen has bad pipes :P Oct 28 20:37:41 Nobody ever says what the "popular android download" is either Oct 28 20:39:38 trying to see if I can get a toolchain any faster from the us.feeds mirror of Narcissus. Oct 28 20:40:03 sure Oct 28 20:57:34 Well... with the narcissus toolchain.... ICE Oct 28 20:57:54 41 minutes till the other one is downloaded Oct 28 20:58:24 (same ICE as I got from my ct-ng toolchain) Oct 28 20:58:53 alan_o seems you are bullied down Oct 28 20:59:20 Haven't heard it put quite like that before, but yes :( Oct 28 20:59:31 I envision it as the stack of problems Oct 28 20:59:37 stack depth is currently about..... Oct 28 20:59:57 well it depends what we're counting :) Oct 28 21:01:58 1. boot 3.7 kernel, 2. get the SPI comm with my device working (I am assuming SPI works on 3.7 bone kernel based on what I read), 3. start soldering up my cape Oct 28 21:02:35 right now it all "works" with 3.6 using spi-gpio, because when I started on the mainline stuff, SPI wasn't working yet. spi-gpio is of course slower than I want Oct 28 21:03:47 I failed to get spi working on 3.6 last night. Could have been any number of things, but I thought moving to 3.7 would be easier, getting onto a version where it for sure works. Oct 28 21:05:12 I'm hoping the new u-boot isn't necessary, because the cpsw doesn't seem to work at all in it. Maybe that's toolchain also. Once you don't trust your toolchain it's hard to say anything for certain. Oct 28 21:27:38 hey koen Oct 28 21:47:57 <_av500_> mranostay: give the man a break Oct 28 21:48:32 <_av500_> alsa, he is at the systemd gala dinner ... Oct 28 21:49:02 ah need pics from him for showcase poster :P Oct 28 21:49:15 oh nice another dead SD card Oct 28 21:49:33 <_av500_> throw it on the pile Oct 28 21:50:55 kingston no surprise Oct 28 21:51:27 Who do you like for SD cards? Oct 28 21:53:39 <_av500_> sandisk Oct 28 21:53:41 <_av500_> and sandisk Oct 28 21:53:49 <_av500_> if not available Oct 28 21:53:52 <_av500_> use sandisk Oct 28 21:54:05 and never have to scandisk? Oct 28 21:54:25 <_av500_> I run DEFRAG.EXE twice daily Oct 28 21:54:33 i usually use sandisk Oct 28 21:58:30 <_av500_> whar do you chant while formatting? Oct 28 21:58:33 <_av500_> what* Oct 28 21:58:43 <_av500_> I settled on something runic Oct 28 21:59:35 good night Oct 28 22:00:25 <_av500_> woglinde: you have 1h more! Oct 28 22:00:26 <_av500_> use it! Oct 28 22:00:59 heh Oct 28 22:03:24 aw man... I'm gonna lose my extra hour on the plane to ELC.... Oct 28 22:04:08 av500 I used it last night Oct 28 22:08:19 <_av500_> I used only 30 minutes last night Oct 28 22:10:12 mranostay: ok.... something seriously wrong Oct 28 22:10:39 Assembler messages: Error: unrecognized option -mauto-it Oct 28 22:10:50 argh Oct 28 22:12:46 that is thumb2 Oct 28 22:12:54 turn off the kernel option Oct 28 22:13:08 koen swears that works Oct 28 22:15:33 ok, it got past that... Seems like on my ct-ng compiler there was something else related to thumb. different error. Oct 28 22:21:50 <_av500_> ctng? Oct 28 22:21:57 crosstool-ng Oct 28 22:25:41 ok, here's a really dumb question...... I'm not putting the modules on there, just the uImage. There's no way that can make me see _nothing_ after "Starting Kernel," right? I mean it never has before. Oct 28 22:25:55 unless serial port is compiled in as a module Oct 28 22:26:09 early printk turned off? Oct 28 22:28:21 <_av500_> you should see something Oct 28 22:28:27 <_av500_> modules or not Oct 28 22:28:47 yeah, that's what I'd tell someone who asked me..... but I'm pulling my hair out here. Oct 28 22:28:52 <_av500_> unless you are running HURD and Linux as a module... Oct 28 22:29:02 I'm lost... different u-boots, different compilers Oct 28 22:29:08 same kernel Oct 28 22:29:16 <_av500_> and same command line? Oct 28 22:29:21 <_av500_> print the uboot env Oct 28 22:29:37 yeah, good point Oct 28 22:29:46 I'm using koen's uEnv.txt Oct 28 22:31:21 http://pastebin.com/Zz665Ah3 Oct 28 22:31:34 ok, so it's koen's uEnv.txt with one change :) Oct 28 22:31:54 The change is to load stuff from the FAT partition instead of the ext4 partition Oct 28 22:32:36 I changed the env in the u-boot source a bit too (for the same purpose). But it shouldn't matter. The files are getting loaded properly Oct 28 22:32:39 and it's the correct ones Oct 28 22:32:49 (because I renamed /boot in the ext4 partition) Oct 28 22:35:26 ok Oct 28 22:35:34 turned on earlyprintk and now I see output Oct 28 22:35:42 I'm wondering now about the command line Oct 28 22:35:50 why didn't something set the console.... Oct 28 22:36:05 (btw it panics with no init found, so there's probably no command line Oct 28 22:36:10 arrgh. how did that happen.... Oct 28 22:38:30 oh no! I broke the command line when I was screwing with the u-boot env...... Oct 28 22:39:07 hmm. still says console=ttyO0.... Oct 28 22:40:15 * _av500_ wonders if DT would help here Oct 28 22:40:36 hehe, it's the problem and the solution :) Oct 28 22:40:46 kind of like that Microsoft virus scanner Oct 28 22:46:46 * mranostay stabs _av500_ Oct 28 22:51:27 <_av500_> not you too Oct 28 22:51:38 et tu mranostay ? Oct 28 22:52:26 * mranostay goes to talk in the mirror 25-35 minutes :) Oct 28 22:52:38 sweet! I'll be doing some of that this week. Oct 28 22:52:46 but I don't use a mirror Oct 28 22:52:51 too distracting to look at myself. Oct 28 22:53:07 "Man that guy is good looking" etc. Oct 28 22:53:25 alan_o: record youself and count the ums instead Oct 28 22:53:40 <_av500_> um, right! Oct 28 22:53:48 yeah, that too.... I try hard to add silence instead of "um" Oct 28 22:55:01 surprisingly not nervous.. but i guess that only matters the day of Oct 28 22:55:11 yeah Oct 28 22:55:21 I'm booting ! Oct 28 22:55:33 3.7 tftp nfs, woot! Oct 28 22:55:42 <_av500_> so what was it? Oct 28 22:55:43 same old u-boot Oct 28 22:55:49 someone needs to get out more :) Oct 28 22:56:09 Not sure yet. It's either the toolchain, or the lack of earlyprintk.... Oct 28 22:57:21 <_av500_> its one of these where after debuggin when you revert everything back is still works Oct 28 22:57:22 mranostay: you said it sir.... I'll be really happy about 12pm CET on Nov 5 Oct 28 22:58:08 oh, thumb is off Oct 28 22:58:20 <_av500_> put it on ice Oct 28 22:58:31 of course it wouldn't build with thumb on with this angstrom toolchain Oct 28 22:58:43 yeah, that's probably not a bad idea Oct 28 22:58:57 too many other things to do now... Oct 28 23:29:18 alan_o, I managed to damage my beaglebone last night...possibly overcurrented Oct 28 23:30:21 excellent timing as I try to finish some code for next week :( Oct 28 23:31:37 mdp: no!! Oct 28 23:31:55 I've actually been thinking about that.... worrying about destroying it Oct 28 23:32:32 yeah, lovely weekend...I have to identify the culprit before attaching my backup bone Oct 28 23:33:16 yeah no doubt. the old sofware-guy's "that'll never happen again" doesn't really play here. Oct 28 23:33:48 So do you know about this capebus stuff? Oct 28 23:34:09 I've gotten 3.7 booting and I need to attach my spi driver Oct 28 23:34:20 so far found one 6mA capable output that was shorted to gnd Oct 28 23:34:31 nothing in dmesg about spi0 or spi1 but "capebus.11: Registered spi-dt platform driver" Oct 28 23:34:45 That's a little more than 6mA Oct 28 23:35:10 * alan_o calculates the resistivity of copper wire Oct 28 23:35:23 well, I know about it from a general review sense as panto was working on it. but I'm still working on a non capebus code base Oct 28 23:35:45 I'm trying to figure out if I need to use it Oct 28 23:36:11 I want to make sure there weren't two problems like that Oct 28 23:36:41 I'm happy to make a little board-alan.c file (which is what I had with 3.6) if that will work. Before though, I wasn't using spi1, but using spi-gpio instead because spi wasn't ready when I started. Oct 28 23:36:53 mdp: so was it a solder mistake? Oct 28 23:37:50 no. I left of hardwired assertion to gnd wired on a bus xcvr after I hooked up a control signal from the bone Oct 28 23:38:28 so internal pullip on. driven high and gnded externally Oct 28 23:39:21 the hardwired assertion was for testing, I assume then. Oct 28 23:39:40 Some hardware guys I know put resistors _everywhere_ Oct 28 23:40:01 I expected more like isolated buffer death but I have a board that only starts 1/10 times now and then quits Oct 28 23:40:05 they don't really have good answers for why always, but certainly would have helped you here. Oct 28 23:40:40 yes, this was a missed wire in my nest of wires. thought I pulled it :( Oct 28 23:40:40 yeah, it seems like I've read people on the mailing lists do seemingly simple things and the whole thing smokes. Like 5v on a 3.3v pin Oct 28 23:42:20 seems like there's not a lot of protectoin Oct 28 23:43:09 contrasted with when I put my PIC18 directly onto an rs232 serial line (-12v) while forgetting the max232, and it still worked great afterward. Oct 28 23:43:43 I'm sure all that stuff is easier on a micro. These SOCs have a lot more to fit inside. Less space for protection Oct 29 00:05:56 Unhandled fault: external abort on non-linefetch (0x1028) at 0xfa1a Oct 29 00:05:56 e140 Oct 29 00:06:00 *sigh* Oct 29 00:06:08 registering for my interrupt Oct 29 00:08:00 oh Oct 29 00:08:04 I bet I know what that is..... Oct 29 00:18:58 it wasn't that :( Oct 29 01:46:17 out of beer! oh noes! Oct 29 01:46:39 alan_o: i've seen that in my output. sure it isn't good :) Oct 29 01:47:12 I'm pretty sure it's not good Oct 29 01:47:14 stack dump Oct 29 01:48:09 http://pastebin.com/BzvUyfSC Oct 29 01:48:41 Looks like it's trying to do some configuration for the GPIO Interrupt Oct 29 01:49:53 nothing like doing searches on the 4500 page TRM Oct 29 01:50:29 i get the beachball of death on those :) Oct 29 01:50:59 I had to do it again to remind myself what that looked like :) Oct 29 01:51:10 I was a big fan of the Windows 95 animated hourglass Oct 29 01:51:20 the one with the blue on it that turned over Oct 29 01:55:11 clippy! Oct 29 02:09:36 alan_o: what they have these days? Oct 29 02:09:54 * mranostay hasn't used Windows more than probably 2 hours in the last 7 years :) Oct 29 02:12:04 I wish I could say the ame Oct 29 02:12:05 same Oct 29 02:12:41 at least i can avoid windows IT calls from family Oct 29 02:12:53 Yeah, that's always a bonus Oct 29 02:13:10 the problem is.... even though I'm not a Windows guy, I still know more than they do. Oct 29 02:13:13 alan_o: "my favorite 'how are you getting a tech job if you don't know Windows IT questions'?" Oct 29 02:13:21 yeah Oct 29 02:13:33 i don't even think my mom knows what i actaully do :) Oct 29 02:13:58 A friend's dad, shortly after I graduated from college asked me to look at his virus-ridden PC, and said "Let's see if your education was worth a damn" Oct 29 02:14:51 "stop downloading attachments from porn sites" :) Oct 29 02:15:05 would be my reply Oct 29 02:28:21 the answer should be - 'no, it wasn't' Oct 29 02:32:05 ds2: touche Oct 29 02:36:05 alan_o: false alarm. I isolated it to an improper termination of the sysboot pins I'm using Oct 29 02:36:07 yay! Oct 29 02:39:56 oh man. That must be a huge relief **** ENDING LOGGING AT Mon Oct 29 03:00:01 2012