=== Jack87|Away is now known as Jack87 [04:00] 20:31 twb, did you get that? [04:00] lilstevie: I did now :-) [04:00] 19:32 What does "ld: no machine record defined" mean? [04:00] lilstevie: any idea about that one? [04:01] none [04:01] Dammit [04:01] never seen ld trip on that before [04:01] I suspect it's something hf-related [04:01] maybe [04:47] Hello, I'm having trouble getting anything on the USB bus to show up with ubuntu on a beagle board. I've tried two different versions of the prebuilt ubuntu and neither of them work. It seems to be fine under angstrom. Thoughts? [04:48] it's an error from the linxu kernel source [04:48] not ld [04:48] it's intentional, means you haven't enabled a machine [04:49] so it cant' find a certain section in the image [05:01] somebody mentioned that it might be a problem with USB_POWER_EN [05:01] anybody have any pointers? === MrCurious_ is now known as MrCurious === MrCurious is now known as 15SABCKTE [05:07] tmzt: please explain, or tell me what docs I should be reading; "enable machine" is too vague for me to work out what to do [05:08] Googling for either that or the error isn't helping, and the .S that refers to the error just says "if you get this, your binutils is too old" -- which I doubt [05:08] GNU ld (GNU Binutils for Debian) 2.21.53.20110910 [05:11] twb`: MACH_whatever isn't enabled [05:12] I don't know what that means. [05:13] You mean like MACH_TEGRA ? [05:13] Would it be getting the "whatever" part from .config, or what? [05:14] yeah [05:14] like MACH_VENTANA [05:15] http://paste.debian.net/129908/ [05:16] where is MACH_PICASSO [05:17] I turned it off because it didn't make sense to me :P [05:17] TF101 isn't an AC500 [05:17] but I told you they were mutually exclusive [05:18] How do you mean mutually exclusive? [05:18] muromec should have put depends=CONFIG_MACH_PICASSO on MACH_TF101 [05:18] board-picasso contains what is needed for the tf101 [05:18] OK, so it's just me being stupid, I will try turning PICASSO on again [05:19] you really shouldn't turn turn off things in the config without making sure :) [05:19] arm is a lot different to x86 :p some things that don't make sense are enabled for unknown reasons :) [05:19] well appear to be unknown [05:19] until you dig [05:20] I was mislead by the Kconfig help string [05:31] OK, now it passed the spot where it stopped before. Yay. === 15SABCKTE is now known as MrCurious_ [05:43] lilstevie: so what is the process to get u-boot onto the tf? I guess step #1 is to git clone u-boot? [05:43] Hmm, apt-get install u-boot u-boot-tools [05:54] correct flash config, build u-boot, correct flash config [05:54] installing u-boot like that wouldn't work :p [05:54] u-boot is not generic [05:55] Well, obviously once I apt-get install the tools I need to use them to generate an image and nvflash it onto the device or something [05:56] lilstevie: does turning VFP on mean that hard-float is off? [05:57] Oh V is for vector not virtual [05:59] VFP is central to hardfloat :p [06:00] as for the tools, they are more for making the right images for initrd and kernel image [06:00] but the kernel image is already one [06:00] done [06:00] if you make with 'make uImage' [06:01] But if I do "make uImage", how does the initrd get into the image [06:01] Or can u-boot just find the kernel and ramdisk in /boot like an x86 bootloader [06:01] when using uboot you dont put the initrd in the image [06:01] you have a uImage and a uInitrd [06:03] Aha, on amd64 "apt-get install u-boot" gives you a useless empty package, but on armhf I can see it providing stuff for other arm systems (efika, omap3, etc) [06:04] anybody know a fast way to byte swap values in c? [06:04] use the networking functions to do it? [06:04] I'm trying to dump a uimage file as it's parts (it's a multi uncompressed) [06:23] Poking around in the u-boot git tree, I can see "make harmony_config" and "make seaboard_config" but nothing else pointing at include/config/tegra2-common.h [06:24] And I see now that "make uImage" in linux will run u-boot-tools' "mkimage" internally. [06:25] make ventana CROSS_COMPILE=cross-compiler-arm DEV_SRC_TREE="tegra-tf101" [06:25] In the u-boot source? [06:26] yes [06:26] make: *** No rule to make target `ventana'. Stop. [06:29] hmbleh [06:29] i dont remember look on the xda board [06:29] somewhere in general muromecs thread has the info [06:29] I never got the hang of web forums, but I'll try [06:32] Looks like xda-developers has a robots.txt or something that gives preferential treatment to google over duckduckgo :-/ [06:33] http://forum.xda-developers.com/showthread.php?t=1202541 suggests ventana_config, /me tries [06:33] thats it [06:34] And it looks like he has a fork of u-boot so maybe I need to use that instead of mainline u-boot [06:34] I didn't do the port so I don't remember [06:34] oh yeah it isn't in mainline [06:39] Waah [06:40] /usr/src/u-boot/lib/sha1.c:228:1: internal compiler error: in maybe_add_or_update_dep_1, at sched-deps.c:845 === popey_ is now known as popey === doko_ is now known as doko === Jack87 is now known as Jack87|Away === lool- is now known as lool === chrisccoulson_ is now known as chrisccoulson === travalas_ is now known as travalas === fisuk1 is now known as fisuk [09:53] janimo, so we should check what files we actually have for mx5 [09:53] i suspect its just the publishing/html generating code that has a hiccup [09:53] and that the image is actually fine [09:53] ogra_, we should only have the .gz I guess. The first test runs did not produce iso, just the sep 14 run [09:54] which was not manual so likely the setup needs to be changed a bit [09:54] it is still an img.gz, the code generating the suffix actually uses file on the unzipped file [09:54] but it is 1.5G [09:55] is there a build of mx5 with the uboot file being found? [09:55] bah, i'm blind [09:55] we have both ! [09:55] right, when I meant we have iso I meant in addition, not instead, should have worded better :) [09:55] * ogra_ only saw the img.gz and img.gz.zsync files [09:55] well, i could have looked better :) [09:56] so get a dvd reader then ... [09:57] Calling uboot-mkimage on kernel and initrd [09:57] mkimage: Can't open /srv/cdimage.ubuntu.com/scratch/ubuntu/daily-preinstalled/preinstalled/armel+mx5.kernel-mx5: No such file or directory [09:57] make: *** [bin-preinstalled_images] Error 1 [09:57] so why do we have that .iso file there [09:57] ARGH !!!!!!! [09:57] Calling uboot-mkimage on kernel and initrd [09:57] mkimage: Can't open /srv/cdimage.ubuntu.com/scratch/ubuntu/daily-preinstalled/preinstalled/armel+mx5.kernel-mx5: No such file or directory [09:57] make: *** [bin-preinstalled_images] Error 1 [09:57] err [09:57] wrong paste [09:57] Publishing armel+mx5 ... [09:57] WARNING: Unknown file type 'Linux rev 1.0 ext3 filesystem data, UUID=2e4ade6f-20c2-4670-9475-51fe999a8d0e'; assuming .iso [09:57] Making armel+mx5 zsync metafile ... [09:58] grnbl [09:59] gah [10:00] so the live builder produces: livecd.ubuntu-mx5.kernel-linaro-lt-mx5 [10:01] http://annonaceae.buildd/~buildd/LiveCD/oneiric/ubuntu-mx5/current/livecd.ubuntu-mx5.kernel-mx5: [10:01] 2011-09-14 04:45:47 ERROR 404: Not Found. [10:01] hmpf [10:01] more hacks ahead [10:04] janimo, i assume we dont want to rename mx5 to linaro-lt-mx5 all over the place ? [10:05] no, mx5 should be fine and much less confusing [10:05] lt-linaro is just the designation of the kernel origin [10:06] right, live-build cant handle cases where the subarch name isnt identical to the kernel name suffix [10:07] though that it assumes iso for ext3 is really weird [10:09] ah, k, the iso is a fallout of the missing kernel [10:24] hmm, i really fail to find the code in live-build that actually creates livecd.ubuntu-mx5.kernel-mx5 [10:41] janimo, i fear we need infinity ... i can stare at the code for another two hours to find out where to put a proper hack, but i guess he will be faster ... [10:41] indeed [10:42] and i dont really feel comfortable to mangle the LB_LINUX_FLAVOURS variable during the build since all arches use it [10:42] and i guess thats the wrong approach anyway [10:43] * ogra_ wonders if he could instead do a hack on the cdimage side [10:56] janimo, i got a testbuild running with a hack in cdimage, lets see [11:11] lilstevie: I'm going home now, but I have typed up what I've learnt up to about last sunday: http://cyber.com.au/~twb/doc/tf101.txt (source) tf101.html (rendered) [11:11] Comments, fixes, etc. welcome. [11:13] twb: well I can tell you why the bootloader needs to be uploaded :) [11:14] No honest, I'm going home. Email me or something. /me runs [11:14] heh [11:14] sure you are [11:15] there are some fixes I will write later [11:15] I kinda have a 1500 word essay due tomorrow which I only just started [11:19] i'm trying to get ubuntu working on my beagle board xM rev c, but nothing on the USB bus seems to be working. Somebody mentioned looking at USB_POWER_EN. But I haven't been able to track down much on that issue. Does anybody have any pointers? [11:20] * ogra_ runs another mx5 build [11:39] and another one [11:42] lol [11:42] oho ! [11:43] that one looks good [11:46] janimo, i think i have something for you in a minute [11:46] it created partitions etc, now it is compressing [11:46] hmm, though it again uses a 1.5G file ... where does that come from [11:47] ogra_, that is the standard new DVD iso size for Ubuntu right? [11:48] might be [11:49] hmm, but thats not a debian-cd issue [11:49] the live builder actually generated an 1.5G ext3 [11:52] an installer for arm that can be put on an sdcard or usb stick would be awesome [11:52] lilstevie, thats how all our images work [11:53] I could make a blob for the start of the mmc that would solve the need to use nvflash :p [11:53] what bothers me is that it is still convinced it builds a CD [11:53] even though it isnt doing that [11:53] ah heh [12:28] * ogra_ will wait for infinity ... that is all odd === zyga is now known as zyga-food [13:40] NCommander: weren't you preparing a kernel patch for nic/usb? === zyga-food is now known as zyga [14:42] hmm, oss4 has a weird build failure [14:42] complaining about an unused variable (which triggers Werror) but the variable is actually used [14:43] hmm [14:43] * ogra_ doesnt get it [14:44] ossmixd.c: In function 'update_values': [14:44] ossmixd.c:165:7: error: variable 'prev_value' set but not used [-Werror=unused-but-set-variable] [14:44] cc1: all warnings being treated as errors [14:44] ... [14:44] but ... [14:44] int value, prev_value; [14:44] .... [14:44] prev_value = mixc_get_value (mixernum, i); [14:44] ... [14:45] i dont get why that would spill a warning at all [14:48] ogra_: The netinstall for omap is no longer pulling x-loader. We will need it as it is still a separate package from u-boot. [14:49] hmm, i didnt change it [14:49] i think [14:49] didnt NCommander upload a d-i change for that recently ? [14:49] * ogra_ checks [14:49] Didn't we take it out of the seed? Or is this a netinstall issue only? [14:49] He updated flash-kernel for omap. [14:50] i didnt touch any omap code at all [14:50] only omap4 [14:51] Ok, so it is the flash-kernel for d-i that needs fixing? [14:51] and i have no idea how michael pulls x-loader, that should happen by flash-kernel-installer usually [14:52] I'll pounce on him when he gets up. [14:52] He may have thought that with the x-loader being pulled in for panda that it was that way for all omap. [14:53] yeah, he was on vac. when it happened [14:53] ogra_, why do you try to compile oss4? [14:53] He just updated it Monday. [14:53] diwic, i try to make it not ftbfs :) [14:54] diwic, it looks like it is right that 'prev_value' isnt used further, but it is definitely getting a value assigned once [14:54] i wonder if i could just drop that var [14:55] ogra_, I guess you could, but is oss4 really relevant for arm? Better just mark it as "don't even try"? [14:56] diwic, not an option :) [14:56] (see cjwatsons mail to -devel) [14:56] what makes me really wonder is why it doesnt ftbfs on all arches with that error [14:56] ogra_, ok will look later (will have to go to meeting now) [14:57] no hurry [14:57] (i didnt even mean to summon you :), but if you want to take it ... ) [15:20] Hi! I am having issues with "vdr" multimedia server.. it locks up my dvbt tuner on pandaboard.. on x86 it behaves fine (natty) [15:20] Any hints what to try? It locks up some time after I stop recording/streaming [15:21] http://fpaste.org/4E4X/ is the dmesg.. I repluggged the tuner ~2 times and restarted vdr during the logged time [15:31] looks like a driver issue ... [15:31] file a bug i'd say :) [15:35] GrueMaster, so looking at NCommander's f-k changes, he doesnt seem to touch x-loader stuff at all [15:36] The x-loader package isn't installed at all. The f-k part fails when making the system bootable due to missing MLO. [15:36] That is why I thought it might be a seed issue. [15:37] right, but i see no code changes touchcing anything x-loader related [15:37] might [15:37] iirc we seed it in the installer seed [15:38] ogra@osiris:~/Devel/seeds/platform.oneiric$ grep x-loader * [15:38] supported-hardware-common: * x-loader-omap3-beagle [armel] # omap3 first stage bootloader [15:38] supported-hardware-common: * x-loader-omap4-panda [armel] # omap4 first stage bootloader [15:38] yup, we do [15:39] but i think thats only for main, that seed isnt used anywhere else [15:39] Well, not sure what to look for. The x-loader wasn't installed on this system after running netinstall. I had to go in and manually add it. [15:40] right, i'm still digging [15:40] are you sure it was ever installed in these images ? [15:41] it had to be. We had working netinstall on Maverick (although unofficially). [15:41] i'm talking about this cycle :) [15:42] NCommander completely rewriote what i did in lucid/maverick [15:42] This cycle it hasn't worked for other reasons. [15:43] right, so we dont know if the new code ever worked [15:44] I'll ping NCommanderwhen he gets up. [15:45] yeah, to me it looks like that never worked [15:46] at least there were no recent changes that could break it [16:02] Is there a kind soul who might be able to help me with the Beagle board under QEMU emulation? I keep getting "qemu: fatal: Trying to execute code outside RAM or ROM at 0x40304350" when attempting to load the Oneiric OMAP4 kernel [16:03] ogra_, sry.. I was away.. file bug agains what? everything is OK when I use mplayer.. everything is OK when using VDR on x86 [16:04] well, thses are definitely kernel messages [16:04] if a userspace app triggers such things you should file a bug against the kernel [16:04] oh.. thinking that way around.. that makes sense.. ok [16:04] ogra_, on launchpad I guess? [16:05] yep [16:05] (cause kernel bugzilla is down anyway) [16:05] or just call ubuntu-bug on the panda [16:05] it shoudl wlak you through [16:05] *walk [16:05] that will need launchpad account anyway.. right? [16:05] yes [16:06] it would be easier from here.. thanks for help.. I hope it's something easier to fix.. [16:06] ogra_, do you think booting vanilla kernel 3.0.0. and testing with that would help? [16:08] MMlosh, where would you get such a kernel ? [16:08] do you have one for omap4 around ? [16:09] ogra_, yes.. I have that ready to go [16:09] well, try it, then you can tell in the bug that its working in vanilla oamp4 [16:09] not completely vanilla.. it has some minimal patches according to the wiki.. [16:09] ah.. crap.. I think it lacks dvb support [16:09] funny though, i wouldnt think that works without all the TI patches [16:11] utlemming: The omap4 kernel is not the beagleboard kernel. That is for the pandaboard. You want an omap kernel. [16:11] ogra_, yes.. but I started with the omap4 defconfig from the wiki [16:11] GrueMaster, this is on panda [16:12] MMlosh: I was referring to utlemming's post on using the omap4 kernel in quemu emulating a beagleboard (before your post). [16:12] MMlosh, well, our kernel carries like 1700 patches from TI to make panda work in all aspects [16:12] GrueMaster: I just double-checked, I'm install linux-image-omap [16:13] I used the 3.0.0 image to get the AP support... it worked fine [16:15] ogra_, I'll get most up to date kernel then and compile.. I hope I get the config right, so the tuner works.. but first the basic report.. and finding out my launchpad password [16:15] which kernel are you currently running btw ? [16:15] we have a 3.0 omap4 kernel in oneirirc [16:18] ogra_, the default one in natty [16:18] 2.6.38-1208-omap4 [16:19] hmm.. that's sure worth trying.. [16:19] the Oneiric linux-image-omap is for OMAP3 devices [16:19] utlemming, yes [16:19] ogra_: that is the one that I am using that is failing in qemu [16:20] utlemming, with what bootloader setup do you use it ? [16:21] u-boot and MLO from the Natty Netbook images [16:21] and you have a vfat you point qemu to [16:21] ? [16:21] carrying all these bits [16:21] yeah, on a vfat [16:22] weird, that should work [16:22] ogra_, he has a point - can I keep mlo and uboot from natty, or I have to upgrade for that 3.0 kernel? [16:22] should work [16:23] ok [16:24] Oh...wait a minute...I'm using the OMAP4 MLO and uboot [16:24] I'll try the OMAP3 ones [16:29] * utlemming feels like an idiot for using the wrong MLO and u-boot === davidm` is now known as davidm === davidm is now known as Guest56598 === NekoXP is now known as Neko [17:31] oh.. the bluish color of the boot messages is gone.. it's pure gray with 3.0.0.. the dvb tuner works.. let's see if it locks up when I stop using it [17:35] and the minor issue with fluxbox is gone (it used to paint default "Workspace 1" instead of my configured name at the beginning) [17:42] taking my preference for 3.0 back.. Xv does not work.. maybe I need X11 update [17:52] ogra_: GrueMaster: I'm alive. I didn't touch the x-loader code in F-K, the onmly thing I did was make sure the proper packages get instlaled in f-k-i on omap3 [17:52] Well, missed one. :P [17:52] ? [17:53] x-loader is not installed. [17:53] there is no x-loader on omap [17:53] great [17:53] (it only moved into u-boot on panda) [17:53] am I just installing the wrong package? [17:53] ogra_: *facepalm* [17:53] no, we arent ending up with any MLO [17:54] omap3 still needs the old way [17:54] lovely [17:54] I'll smack it with a stick today [17:55] well, up to now the omap netinst wasnt usable anyway ... and nobody complained [17:55] * ogra_ wouldnt cry if it wouldnt get ready to be honest [17:56] ogra_: english parse fail [17:57] really ? [17:57] that sentence didn't make sense to me [17:57] hmm, my grammar is a bit weird today, agreed [17:58] ogra_: anyway, I'm going to resolve my nictine fix so I'll bbiab [17:58] NCommander, we havent had any working netinst omap3 images yet and nobody complained ... so it doesnt seem like anyone misses it ... and i wouldnt miss it too if you wouldnt manage [17:58] s/too/either/ [17:58] * ogra_ hopes thats more understandable :) [18:13] ogra_, the -omap4 kernel from oneiric works so far... the tuner's "channel lock" LED sometimes gets off for a second, but always comes back on, meaning the tuner is working [18:14] too bad DPMS and Xv do not work... maybe I need to update X11 from oneiric too [18:17] Is the issue still worth filling on launchpad? I guess that noone will bother fixing that for a <30 days until replacement release === Jack87|Away is now known as Jack87 [20:13] hi folks [20:14] I'm a bit puzzled. I was under the impression that armel was a supported platform for natty. however, it seems there's binary packages that are only available for x86 platforms [20:14] I'm having a bit of trouble getting text within QEMU....I'm using as my u-boot boot.scr parameters "setenv bootargs vram=1M mem=511M console=ttyS2,115200n8 serialtty=ttyS2 root=LABEL=cloudimg-rootfs fixrtc ro". All I see is "Starting kernel ...". Does anyone have any idea what I'm doing wrong here? [20:24] MMlosh: increasing it 2M seems to make things a bit happier, but not much === Jack87 is now known as Jack87|Away [21:25] hi .. i have some problems installing the prebuilt ubuntu binaries on my Pandaboard (revision A2) board .. i believe that i have followed the instructions in the "Getting Started" guide , and im trying to install Natty .. but i can't boot .. the boot sequence does not get longer than "Uncompressing Linux... done, booting the kernel." .. [21:25] After that, the two leds next to the sd slot are flashing for some minutes .. and then nothing further happens [21:25] the image that i've copied to the sd card is ' ubuntu-11.04-preinstalled-netbook-armel+omap4.img ' [21:26] :) please help === Jack87|Away is now known as Jack87 === Jack87 is now known as Jack87|Away