dokoGrueMaster: please document your testing in the bug report00:55
GrueMasterok, will do.00:55
GrueMasterUpdated.  And I'm calling it a day.  Ping me if you need me.01:03
=== ericm-Zzz is now known as ericm
=== ogra is now known as Guest2736
=== Guest2736 is now known as ogra_
=== asac_ is now known as asac
=== Jack87|Away is now known as Jack87
sveinseAnyone tried running rootstock installing natty? I'm getting a fatal from qemu and a subsequent lockup08:31
rsalvetisveinse: are you trying with qemu from maverick?08:31
rsalvetithat could be the issue08:31
rsalvetinewer qemu is a lot better, but I believe only available for natty atm08:31
rsalvetimaybe linaro is also providing it at a ppa08:32
sveinseok, thanks. I'll try natty then08:32
rsalvetiit's now called qemu-linaro08:32
rsalvetithe src package08:32
sveinseOT: Is there installers for natty (i386) available? I need to setup a new virtualbox with natty then08:33
rsalvetisveinse: you can download and install our alpha 3 image08:34
rsalvetithat was released last week08:34
sveinsersalveti: uhm, where exactly can I find the alpha 3 images? I cant seem to find them08:46
rsalvetisveinse: http://www.ubuntu.com/testing/natty/alpha308:48
ppisatiericm_: ping09:08
AudioHI All, I have used latest  Maverick image and booted the panda board. But none of the sound drivers are installed. How can i get audio drivers installed?09:09
ppisatiericm_: i managed to get the dove board working (it was a dumb mistake on my side... :P)09:16
ppisatiericm_: now i'm building a new kernel for it (old way make uImage ARCH=arm CROSS_COMPILE...)09:17
ppisatiericm_: but i noticed that i used on dove, contrary to omap that uses rootstock and a plain fs, a squashfs09:17
ppisatiericm_: can you tell me how to recreate it? or, how to get a new kernel there?09:18
ppisatiericm_: i mean, besides copying uImage over the stick09:18
=== Jack87 is now known as Jack87|Away
ppisatiericm_: is there something else that i should do? i.e. what about the modules? and so on...09:19
ericm_ppisati, I normally use fdr binary-dove09:41
ericm_ppisati, for the kernel package, but you need something like below first09:41
ericm_export $(dpkg-architecture -aarmel); export CROSS_COMPILE=arm-linux-gnueabi-09:41
ericm_ppisati, and then do a 'fdr clean; fdr binary-dove'09:41
ppisatiericm_: fdr? let me try09:42
ericm_ppisati, fakeroot debian/rules09:42
ppisatiericm_: and excuse me if i hammer you with stupid questions, but i want to get started (boot the board, compile custom kernels for it, install new kernel, etcetc)09:43
ericm_ppisati, dude I'm completely fine, you are doing good09:43
ppisatiericm_: k09:43
ppisatiericm_: and with the binady-dove i get... a .deb?09:45
ericm_ppisati, yep09:46
ericm_ppisati, in the parent directory09:46
ppisatiericm_: ok, and how do i install it in the dove image?09:46
ppisatiericm_: it's a squashfs image, i can't copy there and dpkg -i ...09:46
ericm_ppisati, copy it over - through wire or usb09:46
ericm_ppisati, oh - you may want to install that image first09:47
ericm_ppisati, you need a SATA hard drive09:47
ppisatiericm_: no no, wait... :P)09:47
ppisatiericm_: we have 2 things (actually 3)09:47
ppisatiericm_: to update here09:47
ppisatiericm_: the uImage in casper/ (and i can just copy over it) and that's ok09:48
ppisatiericm_: but then i've to update the modules and uInitrd09:48
ppisatiericm_: i already tried to copy an uImage i produced with make uImage but at boot the board was not so happy09:49
ericm_ppisati, the casper/ image is the installer image, so you just want to use that image as a test base instead of a full system?09:49
ppisatiericm_: ok09:49
ericm_ppisati, yeah - probably modules missing due to uInitrd not matching09:49
ppisatiericm_: right09:49
ppisatiericm_: how do i recreate uInitrd and the lib/modules/... stuff?09:49
ericm_ppisati, to cross generate the uInitrd is much more painful09:50
ericm_ppisati, if you insist doing that - you may need a chroot environment09:50
ppisatiericm_: uhm09:50
ppisatiericm_: so i guess you compiled it natively?09:50
ppisatifrom the arm board itself09:50
ppisatii guess09:50
ericm_ppisati, but I personally prefer to install the system and get it generated natively09:50
ericm_ppisati, yes09:50
=== ericm__ is now known as ericm-dinner
=== Jack87|Away is now known as Jack87
=== Jack87 is now known as Jack87|Away
=== Jack87|Away is now known as Jack87
=== Jack87 is now known as Jack87|Away
ppisatiericm-dinner: so let me see if i got it right, you attached a sata hd to the mvl dove10:28
ppisatiericm-dinner: booted via a usb stick10:28
ppisatiericm-dinner: installed everything on sata disk10:28
ppisatiericm-dinner: and then used that for all the work, right?10:29
=== ogra_ is now known as ogra
sveinseHi. I'm getting a lot of "include/qt4/QtCore/qstring.h:187:17: note: the mangling of ‘va_list’ has changed in GCC 4.4" while cross compiling Qt for armel. Anyone familiar with a fix?11:54
ografix the code :)11:55
sveinseThe source has been patched with 92_armel_gcc43_valist_compat.diff from qt4-x11 sources, but it doesn't seem to stop the nagging11:55
sveinseinterestingly I'm getting no response from the qt-developers, they don't have the issue11:55
ppisatiogra: can i ask you for the dove board too?11:56
ograwell, are they using 4.5 and the gold toolchain ?11:56
sveinseI've noticed that qt and (linaro/ubuntu) gcc isn't always best of friends11:56
ograppisati, you mean hos i installed my dove back then ?11:56
sveinseI've tried both 4.4(.4) and 4.5(.2)11:57
ppisatiogra: yep, more or less11:57
ograiirc i used usb to install to the sata disk11:57
ppisatiogra: booted from the disk and did all the compilation from there, right?11:57
ograhaving my full system on sata in the end11:57
ogracompilation ?11:58
ppisatiogra: yep11:58
ppisatiogra: new kernel and stuff like that11:58
ppisatiogra: actually it's the kernel, kernel modules and uInitrd11:58
ograwell, i only compile userspace packages usually, but yes11:58
ppisatiogra: uhm... k11:58
ograi dont touch kernel if i dont have to, thats what we have you guys for ;)11:59
* ogra is a consumer of the binaries here 11:59
ograsveinse, well, i can only point you to the toolchain specialists in #linaro11:59
ograbut i would think that gcc and binutils gold are way more strict than whats out there elsewheer12:00
sveinsewell, thanks12:01
ogras/gcc/linaro gcc/12:01
sveinseyeah, I got that.12:05
sveinseExcept I'm not running linaro gcc yet, since my target system still is maverick. And I can't install natty via rootstock since qemu fails *sigh*12:06
ogramaverick uses linaro gcc12:11
ograall over the place12:11
loolsveinse: You could try with qemu-linaro12:14
loolsveinse: We have a backport in the ~linaro-maintainers tools PPA; it's fairly solid12:14
loolit does break versatile kernels though, but we can switch to vexpress instead  :-)12:14
loologra rsalveti: ^12:15
ogrado we have a netinst setup for vexpress ?12:15
ogra(to have an easy place to download vmlinuz)12:16
sveinselool, I found this: https://code.launchpad.net/~linaro-maintainers/ubuntu/maverick/qemu-linaro/ppa - Are there any prebuilt packages available, or do I need to build from source?12:22
loolsveinse: https://launchpad.net/~linaro-maintainers/+archive/tools is the one12:23
sveinselool, aaah. Thanks a lot12:24
ppisatiogra: do you know who created the dove image?13:04
ppisatiogra: i'm trying to replace the kernel there with a new one13:04
=== ericm-dinner is now known as ericm|ubuntu
ericm|ubuntuppisati, it's auto generated I guess13:10
ogradesign and initial creation was a team effort, all further maintenance after bringing it to existance is done by NCommander13:10
ppisatiogra: k, all ping ncommander then13:10
ppisatiwe are rallying in Millibank13:11
ppisatiand i don't have any sata disk to attach to that board13:11
ppisatielse i would install the system on the disk, compile a new .deb kernel13:11
ograyou will have to re-roll the whole initrd if you replace the kernel13:11
ppisaticopy over there, and install13:11
ppisatiogra: right13:11
ppisatiogra: that's one of my problem :)13:11
ograsince its a casper initramfs13:12
ppisatii'ts a squashfs13:12
ograalso note that there were massive issues (screen/console related iirc) with the recent set of kernel patches (GrueMaster knows more)13:13
ppisatiogra: that's what i'm trying to reproduce :)13:13
ppisatiogra: and i hope i got these13:13
ograto re-roll your initrd, just unpack the squashfs ion an intel machine, use qemu-arm-static to chroot into it and do the changes you need13:14
ppisatiogra: uhm13:14
* XorA needs a new cool device :-D13:15
=== dmart_ is now known as dmart
ppisatiogra: that's a good idea13:15
ppisatiogra: can you walk me to that process?13:15
ograinstall qemu-kvm-extras.static on your host pc13:16
ograthen unpack your squashfs somewhere13:16
ogracopy /usr/bin/qemu-arm-static into the unpacked root to usr/bin/13:16
=== dmart_ is now known as dmart
ograthen just chroot <your unpacked root>13:17
* ogra needs to work on a machine without internet access for a while ... afk for a few ...13:20
sveinseare there a gdb-armel-cross available now? There were some talk of it a while ago13:22
hrwsveinse: https://launchpad.net/~linaro-maintainers/+archive/toolchain/+packages13:25
hrwsveinse: grab gdb-multiarch13:25
loolThat reminds me I should do the same thing for gdb-multiarch as for binutils-multiarch13:26
sveinseSo I install the gdb-multiarch on the host and install gdbserver on target, right?13:31
loolsveinse: Yup13:32
loolsveinse: There are other gdb servers out there, like the one in qemu or kgdb13:32
sveinselool, yes, but the biggest issue in my experience is to find a native gdbserver for the target system. CSL provides target gdbserver, but it's doesn't run without hazzle due to different runtime libs configs.13:35
loolsveinse: well we provide a gdbserver package13:41
loolsveinse: It's linux+eglibc13:41
sveinselool, ok. I planned on using the gdbserver package in armel ubuntu... Or simply the gdbserver armel package in the linaro PPA13:42
ppisatiogra: the idea is good13:42
ppisatiogra: but update-initramfs complains about missing files in /proc&c13:43
loolsveinse: Sure, either is fine13:43
ppisatiroot@longino:/# update-initramfs -u -k 2.6.32-204-dove13:43
ppisatiupdate-initramfs: Generating /boot/initrd.img-2.6.32-204-dove13:43
ppisaticryptsetup: WARNING: could not determine root device from /etc/fstab13:43
ppisati/bin/cat: /proc/cmdline: No such file or directory13:43
ppisati/bin/grep: /proc/cpuinfo: No such file or directory13:43
sveinselool, excellent. Thanks!13:43
ppisatiogra: what scares is that, even if i mount /proc inside the new chroot environment13:47
ppisatiogra: stuff like cpuinfo will still reflect the host system13:47
ppisatiogra: and that's what update-initramfs is looking for13:47
sveinseHmm. Are there any side-effect to qemu-linaro?13:54
sveinseIt's eating my cpu cycles while doing apt-get...13:55
ppisatiogra: it seems initrd is ok14:06
ppisatiogra: even if it complained about /proc14:06
=== ericm|ubuntu is now known as ericm-Zzz
GrueMastermorning all.15:50
janimoGrueMaster, hello15:54
GrueMasterjanimo: What's up?  (other than me).15:54
rsalvetijanimo: how is the qt neon bug going?15:57
* ogra has to report that testing neon on the ac100 is impossible atm15:58
ogramy upgrade to natty failed miserably15:58
ograalignment traps, SIGILLs and segfaults all over the place after rebooting16:00
ograno input at all in X ... all GTK apps fail16:00
janimoGrueMaster, nah, it's just you :)16:02
janimorsalveti, waiting for Qt build to finish16:03
GrueMasterPffft.  Figures.16:03
* GrueMaster needs his own private barista.16:04
ograi'll try to do a piecemeal upgrade later this evening to identify the package that introduces the issue16:04
ograprobably libc but could be higher level16:04
GrueMasterogra: I should be able to tell you later today how my maverick->natty upgrade went.  Spent all day downloading packages (it didn't like my mirror).  Just noticed it was stuck on sudoers file.16:17
=== zyga is now known as zyga-afk
=== XorA is now known as XorA|gone
=== doko_ is now known as doko
ppisatiNCommander: ping17:25
ppisatiNCommander: are you the guys who roll the mvl-dove images?17:25
NCommanderppisati: were mostly17:28
ppisatiNCommander: ok17:29
ppisatiNCommander: can you tell me how do you create it?17:30
ppisatiNCommander: i'm trying to recreate one with a newer kernel17:30
ppisatiNCommander: s/^/^cause /17:30
NCommanderppisati: why are you trying to respin with a new kernel? Spinning a new image requires a considerable amount of work and trouble17:31
ppisatiNCommander: i know17:31
ppisatiNCommander: we are sprinting in Millibank, i've the dove board but no sata disk17:31
ppisatiNCommander: so the only way to get a new kernel on that board, is to recrate the usb image17:31
ppisatiNCommander: this is what i did17:32
ppisatiNCommander: i unsquashed the fs on my fs17:32
ppisatiNCommander: copied over the modules17:32
ppisatiNCommander: and recreated the squashfs17:32
ppisatiNCommander: then i chrooted in it17:32
NCommanderYou need to use livecd-rootfs to respin the squashfs, and modify it to grab your new kernel17:33
NCommanderppisati: then use the files it generates17:33
ppisatiand create the uImage and uInitrd17:33
ppisati-update-initramfs -u -k 2.6.32-214-dove17:33
ppisati-mkimage -A arm -O linux -T ramdisk -C none -a 0 -e 0 -n initramfs -d /boot/initrd.img-2.6.32-214-dove /boot/uInitrd17:33
ppisati-mkimage -A arm -O linux -T kernel -C none -a 0x00008000 -e 0x00008000 -n 2.6.32-214-dove -d /boot/vmlinuz-* /boot/uImage17:33
NCommandersomething like that17:34
ppisatithen i copy the uImage and uInitrd over the usb stick17:34
ppisatiand boot with that17:34
ppisatiNCommander: uhmmm... so it should be ok17:34
NCommanderppisati: ?17:34
ppisatiNCommander: i mean, i wanted to double check that i didn't do anything stupid on my side17:34
NCommanderppisati: no, that sounds right17:35
ppisatiNCommander: k17:35
ppisatiNCommander: do you do that by hand everytme?17:37
NCommanderppisati: no, there's an entire infrastructure to do and handle it automatically17:37
amelimAre there any additional steps required to boot a preinstalled OMAP3 image for a Gumstix Overo?17:52
rsalvetiprobably correct x-loader and u-boot17:54
rsalvetiyou can already find the packages at the archive17:54
rsalvetiand it should work with natty17:54
rsalvetijust need to replace the original x-loader and u-boot from the image to the ones for overo17:55
NCommanderjanimo: w.r.t. to the headless stuff, what seed are you using?17:56
janimoNCommander, minimal+standard17:57
janimoavoiding complicating it to see it work17:57
janimothen if we need more custom stuff we can create a seed17:57
janimoat least I assumed it is less complicated by not creating a specific seed17:57
janimoless files needing change etc17:57
GrueMasterOk, wrt upgrading frm Maverick to Natty, the only issue I have run into so far is that the current default session is for une-efl to run in Maverick, and upgrading reverts to ubuntu desktop (unity + clutter) which fails.  Need to figure out a way to autodetect and select unity-2d.18:04
amelimThanks rsalveti, I'll try that18:10
GrueMasterGrrr.  ubiquity is broken on current images.19:05
amelimhow long should it take to pop up on my monitor on the first boot?19:12
rsalvetiGrueMaster: how broken?19:15
GrueMasterBah.  Appears to be a simple typo in the latest ubiquity-dm.  I edited /usr/bin/ubiquity-dm:385 and removed one ")" and now it works.19:16
GrueMasterBug filed with patch attached.  Sad thing is vi picked it up immediately in auto-syntax check.19:27
GrueMasterAha, looking at the bzr diff, I see how it came about.  4 lines previous changed from "extras.append(subprocess.Popen(" to "proc = subprocess.Popen(" which is why the extra parenthesis was missed.19:30
GrueMasterStill, a simple code check would have picked it up.19:30
=== AndyJ82 is now known as AndyJ
loolGrueMaster: bug #?19:50
loollp #73153619:50
ubot2Launchpad bug 731536 in ubiquity "Typo in ubiquity-dm causes ubiquity/oem-config to not launch" [High,New] https://launchpad.net/bugs/73153619:50
GrueMasteryep, that's the one.19:51
GrueMasterI'm just glad it wasn't uploaded last week.19:52
loolHmm can't commit to it20:00
loolsend a mp though20:01
loolGrueMaster: Vcs-Bzr is owned by ~ubuntu-installer, so I've sent a merge request with your patch20:07
GrueMasterAh, ok.  I was getting to that next, but had to deal with a system running the latest image.  Multitasking can be fun.20:10
=== ogra is now known as Guest59853
GrueMasterjanimo: Your boundary limits fail on an 8G class 10 card.  partition 1 file system is unrecognizable.20:15
GrueMasterSame with partition 2.20:15
=== Jack87|Away is now known as Jack87
janimoGrueMaster, ouch. Thanks20:19
janimodid you test on another card where it works?20:19
janimoGrueMaster, omap3 or omap4?20:21
GrueMasterI'm doing a write-timing test atm.  Compare flashing yesterday's image with todays.  It appears to be taking a very long time.20:21
GrueMasterAnd this was with omap4.20:22
janimoI wonder why the omap4 image is 78M20:22
janimoinstead of 500M+20:22
janimomaybe not fully built?20:22
GrueMasterHmm.  Didn't even notice that.  sigh.20:24
janimoGrueMaster, so only omap4 was respun I guess, and it came out only slightly larger than the VFAT partition20:28
GrueMasterThat means zsync clobbered my 20110308 image.  Ok.20:28
GrueMasterI'll smack NCommander.20:29
=== Jack87 is now known as Jack87|Away
NCommanderGrueMaster: bah, it was a case of extreme stupidity on antimony's part.20:44
NCommanderjanimo: sfdisk: ERROR: sector 0 does not have an msdos signature /srv/cdimage.ubuntu.com/scratch/ubuntu-netbook/daily-preinstalled/debian-cd/armel+oma$20:45
NCommanderyour fix broke it, I'll back it out and do a .2 to confirm its been unbroken20:46
janimoNCommander, hmm, maybe dropping the initial 63 from the first partition20:48
janimoI am puzzled why the boot sector was not written20:49
NCommanderjanimo: I'm not really interested in madly guessing to fix this :-/20:49
janimoNCommander, do you know why the image came out only 78 M ?20:49
janimowas the above error given by the script without >/dev/null ?20:49
NCommanderjanimo: it was what was in the log20:49
amelimhmm, so I replaced the u-boot.bin and MLO for the Gumstix. However, once I start it up nothing happens. The console outputs that it's booting the kernel but nothing pops up on my screen. :/20:59
GrueMasteramelim: What video output are you using?  DVI?20:59
amelimHDMI -> DIV21:00
amelimsorry DVI21:00
GrueMasterWell, that *should* work ok.  Can you try adding console=ttyO2,115200 earlyprintk=tty)2,115200 and seeing if there is something in the kernel output?21:01
GrueMasterUnfortunately I don't have one of these devices to test with.21:02
amelimto boot.src?21:02
GrueMasteryes.  Add them to the bootargs.  You will need to convert the boot.scr to text first (dd bs=64 skip=1 if=boot.scr of=boot.script), make the mod, then reconvert to u-boot script (mkimage -A arm -O linux -T script -d boot.script boot.scr21:04
=== Guest59853 is now known as ogra
ograjanimo, NCommander, the script only fails because of the missing redirect to /dev/null, the sfdisk error is normal21:09
ograoh, jani is gone21:09
NCommanderogra: I backed it out, and I'm not going to run it down ATM21:10
ograsfdisk always complains if no partition table exists at all21:10
ograWarning: partition 1 does not end at a cylinder boundary21:10
ograSuccessfully wrote the new partition table21:10
ograRe-reading the partition table ...21:10
ograthe script runs "set -e", thats why sfdisk fd1 and 2 were redirected to null21:11
amelimGrueMaster I'm not getting any output on my console other than what I've seen previously. Here is my boot.script http://pastebin.com/6ZiFYTwU21:13
GrueMasterThese console lines should give you output on the serial port.  I am assuming you have a serial debugging cable hooked up to your pc?21:15
amelimdoh, I'm connected through ttyUSB021:16
amelimshould I change the script accordingly?21:16
GrueMasterIs ttyUSB0 on the desktop system or the overo?21:16
amelimhost side21:17
rsalvetiamelim: it could be a bug at the display code for overo21:17
rsalvetiamelim: can you try linaro's kernel?21:17
amelimrsalveti: Where can I find that?21:18
rsalvetiamelim: package linux-image-linaro-omap21:19
rsalvetiif you're testing with natty21:19
amelimI was working with maverick21:20
rsalvetiamelim: oh, not sure if it was ever tested with overo21:21
amelimI don't suppose there are any pre-installed images for Natty yet?21:23
rsalvetiamelim: yup, and we just had alpha 3 last week21:24
amelimrsalveti: Any chance you can direct me to them? I'm having difficulty finding them21:27
GrueMasteramelim: http://cdimage.ubuntu.com/ubuntu-netbook/releases/natty/alpha-3/21:27
amelimmany thanks21:27
GrueMasteruse the omap image.21:27
amelimGrueMaster, rsalveti: I've gotten the console to output on boot and have encountered the following error with the installer http://pastebin.com/UWBqtxcX22:50
GrueMasteramelim: Which image is this?  Natty Alpha 3?22:57
amelimIs there a recent version of Ubuntu which is verified to work on an Overo that I can use in the meantime?23:17
GrueMasterNot having said hardware, I couldn't tell you.  I am looking into this issue now though to see what may be the cause.23:18
GrueMasterBut it appears there is a difference with the image I tested vs the image they posted.  Not good.23:20
GrueMasterAt the console prompt that oem-config dumps you to, try "apt-get remove ubiquity-slideshow-*" and see if it produces anything.23:21
amelimneither ubiquity-slideshow-kubuntu nor slideshow-ubuntu were installed23:27
GrueMasterI'm flashing an SD with that image now.  I should have it ready for boot soon.23:30
GrueMasterGood thing I keep a mirror of all images between release milestones.23:31
GrueMaster(well, all images I actively work with).23:31
GrueMasterHmm.  Ok, the install finished here without errors.23:53

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!