**** BEGIN LOGGING AT Fri Mar 23 02:59:58 2012 Mar 23 13:37:01 rsalveti, pvr drivers uploaded to NEW, talk to infinity for an archive admin review to get them out there ;) Mar 23 14:55:21 ogra_: Know any logical reason why the arm team was assigned bug 856988? Mar 23 14:55:22 Launchpad bug 856988 in gstreamer0.10 "totem cannot play quicktime file after upgrade to oneiric" [High,Confirmed] https://launchpad.net/bugs/856988 Mar 23 14:56:34 Bug was filed against amd64. Not a single arm reference in the bug that I saw. Mar 23 15:00:10 just unassign it Mar 23 15:00:19 seriously not an arm prob Mar 23 15:04:46 GrueMaster: with apt-setup, my test failed, and I'm not sure why, I'll stab it more in my today Mar 23 15:07:53 GrueMaster, i pinged kate about it, she will look after the meeting why she assigned it to arm Mar 23 15:08:22 She probably wanted the ubuntu-audio team. Mar 23 15:08:48 or -desktop since i see that seb did set the importance already Mar 23 15:08:59 we'll hear after the meeting i guess ;) Mar 23 15:12:09 Oh, beta 2 has opened up in the iso tracker. Guess I can shut down the daily image tests I have automated. Mar 23 15:42:58 not getting any hdmi output after todays update. Mar 23 15:56:05 ogra_, GrueMaster, Daviey: bug 961133 updated with logs from today's image Mar 23 15:56:07 Launchpad bug 961133 in linux "Ubuntu Desktop ARM Images do not boot on my Pandaboard ("EXT3-fs (mmcblk0p2): error: couldn't mount because of unsupported optional features (240)")" [Medium,Confirmed] https://launchpad.net/bugs/961133 Mar 23 16:00:14 Riddell, looks all proper, are you sure you have the monitor plugged in to the right port ? Mar 23 16:00:27 (try changing them for a test) Mar 23 16:01:02 it definitely boots and also seems to do the reboot after resizing the FS just fine ... theoretically it should now sit with ubiquity on the screen Mar 23 16:01:13 ogra_: no I'm not sure because I have had inconsistent advice from this channel and from tests, it works with the DVI port with oneiric images. ubuntu server requires the HDMI port Mar 23 16:02:04 ubuntu server is headless. Doesn't care about video ports. Mar 23 16:02:06 well, just pulg it over to the other port, i'm never sure which one the kernel driver supports in which TI kernel Mar 23 16:02:16 right, server runs on serial Mar 23 16:02:16 (kernel might). Mar 23 16:02:55 Riddell, in any case that looks like a successful boot Mar 23 16:03:49 if there are issues i would blame the graphics driver ot the monitor Mar 23 16:03:58 s/ot/or/ Mar 23 16:04:17 ogra_: and yet it's not Mar 23 16:04:25 ?? Mar 23 16:04:29 not what ? Mar 23 16:04:30 ogra_: the graphics driver works fine with ubuntu server precise and ubuntu desktop oneiric Mar 23 16:04:35 not successful Mar 23 16:04:56 so what do you see on the screen then between the two u-boot runs i see there ? Mar 23 16:05:08 ogra_: there is no output to the monitor Mar 23 16:05:58 how long does it take approximately beweeen the two boot runs ? and does the left LED blink in a consant ~1sec fequency after it sits there on the serial ? Mar 23 16:07:00 I don't know, I haven't timed it, would it help you if I did? Mar 23 16:07:19 the reszizing run usually takes about 1-2min, then it reboots (note that only happens once if you first boot a new image)... if after that you see the left LED blink constantly it means that everything is fine but the screen doesnt output Mar 23 16:07:21 I don't know which is the left LED but one LED is currently doing two blinks in rapid succession then about 1 sec off Mar 23 16:07:35 sounds about right Mar 23 16:07:58 it has definuitely to do with the graphics/monitor Mar 23 16:08:17 hi, i need to edit boot.script and then get the boot.scr. What should I do? Thank you Mar 23 16:08:20 what kind of monitor is that ? Mar 23 16:08:26 ogra_: acer DVI Mar 23 16:08:33 Spider-Pork, already installed image ? Mar 23 16:08:51 Riddell, is it capable of doing 720p and 1080p modes ? Mar 23 16:09:03 ogra_: I don't know what that means Mar 23 16:09:18 do you know which modes it supports ? Mar 23 16:09:21 ogra_: no Mar 23 16:09:34 ogra_: yep Mar 23 16:09:43 ogra_: i need to enable DVI output Mar 23 16:09:52 as reported here http://omappedia.org/wiki/PandaBoard_FAQ#How_do_I_enable_DVI_output.3F Mar 23 16:09:55 the graphics driver reads the monitors EDID usually, if it doesnt find modes it can support it can actually fall over Mar 23 16:10:17 Spider-Pork, after you edited /boot/boot.script, just run sudo flash-kernel Mar 23 16:10:25 it will update boot.scr Mar 23 16:10:31 cool thank you Mar 23 16:10:56 my problem now is that i can't log in with X :) Mar 23 16:11:06 i can't see no output on my DVI monitor Mar 23 16:12:12 Riddell, when you used the precise server image, did you actually see anything before X came up ? (plymouth ... console etc) Mar 23 16:12:19 before apt-get install omap4-extras was connected to HDMI Mar 23 16:12:33 but after upgrade i can't see anything on HDMI Mar 23 16:12:39 ogra_: no I had to start X from the serial command line Mar 23 16:12:51 Riddle; What monitor do you have? What are the specs? Mar 23 16:13:00 Riddell, aha, sounds like an issue with the framebuffer driver and your monitor Mar 23 16:13:55 GrueMaster: this one http://search.acerdirect.co.uk/acer/V223hqv Mar 23 16:15:20 Riddell, between oneiric and precise the graphics driver was cvompletely redone by TI, that might explain your issues (though i would still love to know where your weird bootmessages came from when you fiddled with the image) Mar 23 16:16:24 Riddell, talk to robclark if he knows anything about issues with that monitor (he maintains the graphics stack and we actually know that some monitors and brands dont work well at all) Mar 23 16:16:46 ogra_: robclark is a TI guy? Mar 23 16:16:53 yes Mar 23 16:17:31 Looking at the specs for that monitor, It is 1920x1080 (HD 1080). But, port wise it only lists vga. Does it have DVI & HDMI? Mar 23 16:17:33 Riddell, and about your complaint that it is hard to install ubuntu on panda, 100s of other users dont have any probs at all ;) Mar 23 16:17:57 GrueMaster: DVI yes, HDMI no Mar 23 16:18:37 right, intrestingly http://www.acerdirect.co.uk/ACER_55cm_21.5_Wide_16_9_FHD_5ms_20000_1_ACM_black_MPRII_EURO-UK_EMEA_200ni_ET.WV3HE.017/version.asp says 0 for DVI and HDMI ports Mar 23 16:18:48 As to today's image, I am not getting video either. Until that gets sorted, please stick with beta 1. Mar 23 16:19:09 Riddell, the display part should be largely similar.. but I'd start w/ just trying to get fbcon (ie. text) up, without yet starting x.. Mar 23 16:19:26 drm.debug=7 in bootargs, and pastebin the boot log would be helpful Mar 23 16:19:40 robclark: what is fbcon? Mar 23 16:19:47 ie. text console Mar 23 16:19:55 You can use today's ubuntu-server image for this test. Mar 23 16:20:12 GrueMaster: why is that better? Mar 23 16:20:15 or just put "text drm.debug=7" in your bootargs Mar 23 16:20:23 right, it shoudl come up with a framebuffer console after install and show you a login prompt onm the screen Mar 23 16:20:26 GrueMaster: am not getting any hdmi output after todays update either. Mar 23 16:20:44 Riddell, because you cant run the desktop installer without a screen Mar 23 16:21:40 boot args are set like that? http://omappedia.org/wiki/PandaBoard_FAQ#How_do_I_saveenv_for_boot_args_on_a_PandaBoard.3F Mar 23 16:22:17 Riddell, just run through the server install, then edit /boto/boot.script and change the bootargs there ... then you run sudo flash-kernel and reboot Mar 23 16:22:21 robclark: THis might be an indicator: sysfs: cannot create duplicate filename '/devices/platform/omap/omapdss_dss' Mar 23 16:22:28 oh Mar 23 16:22:43 hrm? Mar 23 16:22:47 s/boto/boot/ Mar 23 16:22:53 how can that happen? Mar 23 16:23:14 wasnt there a fix from rsalveti for exactly such an issue ? Mar 23 16:23:27 One sec and I'll pastebin a dmesg. Mar 23 16:23:30 maybe, I've not seen that before Mar 23 16:24:08 * debian/patches/111_armel-drv-fallbacks.patch: Mar 23 16:24:08 - Avoid loading the driver to test if it's available. Xorg will later load Mar 23 16:24:08 and validate the module, and if it's already loaded it'll trigger an Mar 23 16:24:08 error and invalidate the driver (LP: #959928) Mar 23 16:24:25 though i think that was focusing on pvr Mar 23 16:24:34 http://paste.ubuntu.com/896634/ Mar 23 16:25:05 ogra_: I done the steps you suggest to me. Now the board won't boot Mar 23 16:25:18 dmesg log from my ubuntu-server install (preseeded btw - woot!) Mar 23 16:25:48 Spider-Pork, well, i have no idea about the omappedia stuff, we otften have people that follow outdated stuff from there, what exactly did you change in your boot.script ? Mar 23 16:26:01 hmm, do we somehow end up with two omapdss platform devices? That would cause all hell to break loose... Mar 23 16:26:50 ogra_: as suggested here http://omappedia.org/wiki/PandaBoard_FAQ#How_do_I_enable_DVI_output.3F I added "omapdss.def_disp=dvi" at the end of the file Mar 23 16:27:07 no idea if that works or not with the current images Mar 23 16:27:19 the problem is that don't boot Mar 23 16:27:26 the SD card led are dead Mar 23 16:27:39 SD card and run led Mar 23 16:28:05 and you definitely only added that arg and properly ran flash-kernel after that ? Mar 23 16:28:25 yep Mar 23 16:28:26 sounds like your boot partition is trashed somehow ... Mar 23 16:28:31 cool Mar 23 16:28:40 this was an official ubuntu image ? Mar 23 16:28:51 fortunately I got a safe copy of boot.scr Mar 23 16:28:52 (downloaded from cdimage.ubuntu.com) Mar 23 16:28:55 ogra_: yep Mar 23 16:29:09 fwiw, I'm not sure, all the legacy omapfb bootargs are ignored, omapdss bootargs, not sure if they could have some unintended result.. Mar 23 16:29:11 https://wiki.ubuntu.com/ARM/OMAP Mar 23 16:29:13 i dont think its your boot.scr Mar 23 16:29:25 omapdrm will just use any/all displays that it detects a connection on Mar 23 16:30:00 if the LEDs dont light up the bootloader (MLO) is not found ... if your boot.scr would be trashed it would still boot into a bootloader session Mar 23 16:30:05 oh well, 3 hours of ubuntu installation and upgrade in the trash Mar 23 16:30:21 did you install any third party stuff ? Mar 23 16:30:33 only omap4-extras Mar 23 16:30:34 an ubuntu install usually doesnt take more than 20min on the panda Mar 23 16:30:38 for video acceleration Mar 23 16:30:46 the installation yep Mar 23 16:30:53 but the upgrade ... Mar 23 16:31:01 using the icon on the desktop ? or following some weird omappedia docs again ? Mar 23 16:31:27 http://omappedia.org/wiki/Ubuntu_PPA Mar 23 16:31:31 sigh Mar 23 16:31:31 with shell command Mar 23 16:32:06 I apologize if I followed wrong guide. What link should follow? Mar 23 16:32:36 you shoudl just click the TI icon on the desktop Mar 23 16:33:01 (note that this only is available on released images, for 12.04 there is no working GLES driver yet) Mar 23 16:33:14 I'm using 11.10 Mar 23 16:33:42 well, there you have the TI icon on the launcher after first boot Mar 23 16:33:51 it does all you need Mar 23 16:34:25 OK, I'll reinstall the whole prebuild image Mar 23 16:34:31 (and the person who added the sed command to that wikipage to trash your sources.list should be smacked ... this should rather link to the launchpad documentation for enabling PPAs) Mar 23 16:35:01 ah ok :) Mar 23 16:35:11 image i took here are ok? https://wiki.ubuntu.com/ARM/OMAP Mar 23 16:35:22 *images Mar 23 16:35:41 https://wiki.ubuntu.com/ARM/OMAP has links to ubuntu wikipages that document enabling the PPA and the like Mar 23 16:35:58 or the respective subpages have, cant remember ... Mar 23 16:36:15 yeah, folllowing the ubuntu wiki should generally get you everything working Mar 23 16:36:37 at least for released images we test and verify the docs and images every time before release Mar 23 16:37:53 ok so I'll follow that guide, thank you Mar 23 16:38:05 :) Mar 23 16:50:53 Daviey: today's ubuntu server image gives me 100s of lines like this: Use of uninitialized value $value in substitution (s///) at /usr/share/perl5/Debconf/Format/822.pm line 65, <$__ANONIO__> line 23345. Mar 23 16:57:50 Riddell: Where are you seeing that? Mar 23 16:58:24 GrueMaster: "today's ubuntu server image" Mar 23 16:59:03 can you be a little more specific? As in "I am not seeing this anywhere, but would like to try to reproduce it". Mar 23 17:00:31 Riddell: Can you provide more before/after logging. Mar 23 17:00:42 on a call.. will investigate shortly after Mar 23 17:00:44 GrueMaster: booted the server image, it did the resizing fine, on second boot I get that message repeated indefinately Mar 23 17:00:57 ogra_: btw: using your boot.scr & preseed.cfg, system booted straight through to login w/o human intervention. Mar 23 17:01:21 Riddell: I am not seeing that at all. Mar 23 17:01:35 Can you double check your md5sum on the image? Mar 23 17:01:53 Also, try blanking and reflashing the SD (in case it has some corrupted bits). Mar 23 17:02:05 GrueMaster: 4b7d83d50edd57024bcc370ecb248224 precise-preinstalled-server-armhf+omap4.img.gz Mar 23 17:11:07 GrueMaster, cool, well, i tested it, so i would be surprised if it would stop working within two days ;) Mar 23 17:11:52 ogra_: Only issue I have is it doesn't pull the hostname. I'll double check my preseed though. Mar 23 17:12:09 did you put in a hostname actually ? Mar 23 17:12:22 the original pressed you added didnt have one Mar 23 17:12:59 (it used the system fallback in the preseed ... which would be localhost) Mar 23 17:13:30 d-i netcfg/get_hostname string unassigned-hostname is what netboot uses. Mar 23 17:13:53 right, that means "use the default" Mar 23 17:14:19 if you set foo instead of unassigned-hostname it should set it to foo Mar 23 17:15:03 I thought it would get it from dhcp. Maybe I am using a boot cmdline for that. Mar 23 17:15:51 hmm, probably another bug we need to hunt down Mar 23 17:16:11 i thought it uses the system default, but dont quote me on that Mar 23 17:18:10 Might be a d-i thing. Mar 23 17:53:33 robclark: here's the serial output from adding drm.debug=7 http://paste.kde.org/445214/ Mar 23 17:53:42 is that what you mean by boot log? Mar 23 18:01:34 Riddell, hmm, you might need some loglevel=10 or something like that.. so that we actually see the traces.. Mar 23 18:01:42 or just try and run dmesg Mar 23 18:02:03 (if the kernel is built w/ large enough log buffer we might not loose the traces we want to see) Mar 23 18:02:15 robclark: http://starsky.19inch.net/~jr/tmp/dmesg Mar 23 18:03:02 oh, and you need to add drm.debug=7 in bootargs Mar 23 18:03:16 looks like that is still missing Mar 23 18:12:55 robclark: http://starsky.19inch.net/~jr/tmp/dmesg any better now? Mar 23 18:16:10 Riddell, yes, much.. Mar 23 18:19:00 * Riddell wanders off in hunt for other monitors to try Mar 23 18:19:15 hmm, what was the issue now? Looks like it finds and configs hdmi for 1080p.. Mar 23 18:19:47 robclark: the monitor does not show anything Mar 23 18:20:01 hmm.. Mar 23 18:20:15 robclark: I have a DVI-D monitor which is plugged into the HDMI port Mar 23 18:20:28 oh, I think you need to add one more bootarg.. Mar 23 18:20:35 you are missing console=tty0 Mar 23 18:20:43 (you should have two console='s Mar 23 18:20:44 ) Mar 23 18:20:55 works fine manually installing ubuntu-desktop and starting lightdm from login Mar 23 18:21:20 Also, what image is this? Beta 1? It isn't the same issue as the daily image. Mar 23 18:21:26 (from an ubuntu-server image) Mar 23 18:21:32 GrueMaster: it is ubuntu server beta 1 Mar 23 18:21:36 works fine with oneiric plugging the same monitor into the DVI port Mar 23 18:22:11 well, at least the display part looks ok.. if you add the second console= you should see boot console on screen as well.. Mar 23 18:22:39 I'm not sure about lightdm, etc.. I'm using 12.04 but hadn't updated for a week or so, so I guess I'm not on latest Mar 23 18:23:14 well that just starts X Mar 23 18:23:34 anyway starting again with an ubuntu server install to try the console= thing Mar 23 18:24:10 fwiw, might be quicker to just change bootargs.. Mar 23 18:24:31 ctrl-alt-f1, see if that can switch you to another VT Mar 23 18:24:46 (if you have no other way to login and edit boot.script and flash-kernel) Mar 23 18:24:57 alas I already started wiping the card in preparation for going hunting monitors Mar 23 18:25:04 heheh, ok Mar 23 18:29:11 robclark: is DVI the right port for oneiric and HDMI right for precise? Mar 23 18:29:32 Riddell: I only test on HDMI. Since Maverick. Mar 23 18:29:57 GrueMaster, Riddell, in theory which ever one is plugged in should work.. Mar 23 18:30:00 (I do test DVI occasionally, but not as default). Mar 23 18:30:16 maybe not a bad idea to test.. but it works for me w/ upstream (linus's tree) kernel Mar 23 18:30:44 why does it have two if DVI monitors can work with HDMI ports? Mar 23 18:30:44 hdmi is known to work Mar 23 18:30:46 ogra_: thanks Mar 23 18:31:02 Dual monitor support maybe? Mar 23 18:31:08 Riddell: is this the issue that is happening just after the latest upload? Mar 23 18:31:13 Riddell, to confuse people about which one to plug in :-P Mar 23 18:31:17 (just kidding) Mar 23 18:31:17 might be some conflict with the upstream changes Mar 23 18:31:24 I need to check with the latest kernel Mar 23 18:31:36 one upload was to sync with ti lt's 3.2 tree Mar 23 18:31:45 and another to sync with ubuntu's changes and stable updates Mar 23 18:31:53 rsalveti: I think he is seeing other issues. But the daily desktop does have a separate issue. Mar 23 18:32:15 rsalveti: See http://paste.ubuntu.com/896634/ Mar 23 18:32:20 No video out. Mar 23 18:32:36 rsalveti: no Mar 23 18:32:46 why would minicom be giving me this corruption? http://paste.kde.org/445262/ Mar 23 18:32:47 GrueMaster: yeah, that's the one I saw a bug for Mar 23 18:32:51 GrueMaster: can check that Mar 23 18:33:51 Riddell: That looks like bug 924399. Mar 23 18:33:52 Launchpad bug 924399 in ubiquity "oem-config-debconf window size issue on keyboard selection screen" [Low,Confirmed] https://launchpad.net/bugs/924399 Mar 23 18:33:54 did you get the same behavior with screen? Mar 23 18:34:37 rsalveti: who me? Mar 23 18:35:57 GrueMaster: my screen looks not very much like that screenshot, no colours and the non US ascii characters are all garbled Mar 23 18:36:28 Riddell: yeah Mar 23 18:36:34 Riddell: Minicom doesn't look the same as screen. Mar 23 18:37:14 rsalveti: I don't run screen, are you suggesting I run minicom inside a screen session? Mar 23 18:37:29 no, just open screen directly instead minicom Mar 23 18:37:31 GrueMaster: well no, what does screen have to do with it? Mar 23 18:37:34 Riddell: screen /dev/ttyUSB0 115200. Mar 23 18:37:45 ah Mar 23 18:37:51 for one, it displays differently. Mar 23 18:38:19 this experience really does remind me of linux on a PC a la 1999 Mar 23 18:51:29 robclark: added console=tty0 nothing on monitor Mar 23 18:53:02 Riddell, if you still have "quiet splash" in bootargs, drop that and replace w/ "debug" (and possibly "text" if you don't want x to start).. Mar 23 18:54:12 how do I save the output with screen? Mar 23 18:54:24 robclark: Is there a way to see what his system reads for edid? Mar 23 18:54:38 Riddell: -A, H Mar 23 18:54:53 GrueMaster, drm.debug=7.. will enable enough traces to see the edid parsing.. (which looked fine, btw) Mar 23 18:55:02 That should toggle logging on/off. Mar 23 18:56:17 robclark: ok. Mar 23 18:57:04 GrueMaster: that only saves what is currently on my terminal not the scrollback Mar 23 18:58:22 Riddell: What do you need in scrollback that you can't reproduce with your console history? Mar 23 18:58:34 (i.e. running dmesg again). Mar 23 18:58:46 GrueMaster: with screen I have no console history Mar 23 18:58:58 up arrow? Mar 23 18:59:27 um that depends on the context in which it is pressed Mar 23 18:59:42 there is no ability to use konsole's scroll function when running screen Mar 23 19:00:12 This I am painfully aware of. Mar 23 19:02:23 robclark: http://starsky.19inch.net/~jr/tmp/dmesg updated with quiet splash changed to debug Mar 23 19:02:30 * Riddell goes in hunt for monitors again Mar 23 19:05:08 well, looks like it is detecting the monitor, and parsing edid fine.. Mar 23 19:05:15 so I don't think another monitor would help.. Mar 23 19:05:34 from text console, try (possibly as root): cat /dev/urandom > /dev/fb0 Mar 23 19:05:45 should produce snow/noise on the display Mar 23 20:38:40 robclark: well another monitor did not help Mar 23 20:38:41 http://starsky.19inch.net/~jr/tmp/dmesg-chris Mar 23 20:39:14 robclark: cat /dev/urandom > /dev/fb0 does nothing Mar 23 20:44:20 oh, btw, if dvi doesn't work, I expect this is the reason: Mar 23 20:44:22 omapdrm omapdrm.0: dvi has no driver.. skipping it Mar 23 20:44:49 for hdmi, looks like there is some PM issue: Mar 23 20:44:50 [ 5.430938] Failed to start PHY Mar 23 20:44:50 [ 5.434265] omapdss HDMI error: failed to power on device Mar 23 20:45:05 not sure if there were recently some kernel updates? Mar 23 20:45:52 I do get something like that the very first time I try to power on HDMI, but on upstream kernel don't have all the other hdmi errors.. Mar 23 20:46:16 my guess is that tomi fixed something, and the fix isn't backported to ubuntu kernel.. Mar 23 20:46:49 at omapdrm level, all looks as it should be, no EDID or hot plug detect or similar issues Mar 23 20:47:19 mm Mar 23 20:47:41 ogra, GrueMaster: is there any arm guy in canonical in britain that I could meet up with to debug in person? Mar 23 20:47:45 e.g. in millbank? Mar 23 20:52:27 Riddell, maybe give a try w/ the linux-ti-omap from ppa:tiomap-dev/omap-trunk ? Mar 23 20:52:57 it is 3.3 based.. and the shape of omapdss has been steadily improving over last few kernel release cycles Mar 23 20:53:48 robclark: so I just add that PPA and install linux-ti-omap and that's all it needs? Mar 23 20:53:58 I think so Mar 23 20:54:25 I'm usually rebuilding kernel manually myself, but I think the one from the ppa should work (ie. not depend on anything else) Mar 23 20:56:12 http://starsky.19inch.net/~jr/tmp/dmesg-hdmi-projector dmesg when plugged into hdmi projector Mar 23 20:57:07 this stuff is very funny looking: sysfs: cannot create duplicate filename '/devices/platform/omap/omapdss_dss' Mar 23 20:57:33 something is a bit odd about that kernel, not sure if a bad merge of some patch or something?? Mar 23 20:57:46 Even back in 3.2 days, I didn't see stuff like that Mar 23 21:20:19 linux-ti-omap4: Installed: 3.3.0.1481 Mar 23 21:21:30 http://starsky.19inch.net/~jr/tmp/dmesg-linux-ti-omap4-3.3.0.1481 Mar 23 21:21:31 no change Mar 23 21:21:37 I do believe I have now lost the will to live Mar 23 21:21:56 oh my god Mar 23 21:22:00 it works Mar 23 21:22:04 just in time to save my life Mar 23 21:22:19 of course it doesn't work on the HDMI port, that would be too consistent Mar 23 21:22:23 it only works on the DVI port Mar 23 21:24:28 ogra, GrueMaster: well I have found a solution to bug 961133 Mar 23 21:24:29 Launchpad bug 961133 in linux "Ubuntu Desktop ARM Images do not boot on my Pandaboard ("EXT3-fs (mmcblk0p2): error: couldn't mount because of unsupported optional features (240)")" [Medium,Confirmed] https://launchpad.net/bugs/961133 Mar 23 21:25:09 ogra, GrueMaster: who do I poke to get that backported? Mar 23 21:27:09 Not sure. Part of the problem is that yours is a very isolated situation. Have you been able to get video from the HDMI port on any previous image (Oneiric, Angstrom test image, etc)? As I said earlier, I test the HDMI port almost exclusively as it is the default port. Mar 23 21:27:51 Now there is a possibility that the precise kernel has issues with the dvi port, I just haven't had the cycles to test tat use case. Mar 23 21:28:17 I am more concerned that you don't have HDMI, as that could indicate a defective board. Mar 23 21:31:00 well I'm out of lifeforce for today, if you have suggestions on how I can help more do let me know (such as finding other people in britain I can meet to debug with) Mar 23 21:32:27 If you can't get HDMI working with the images I suggested, my next suggestion would be for you to bring your board to UDS where I can test it directly, and possibly swap it for one of mine that is known to work. Unfortunately, My status of UDS is uncertain at this moment. Mar 23 21:39:07 GrueMaster: yeah that's what I'll do but it would be nicer to be able to meet up with someone with the skills and hardware sooner Mar 23 21:39:54 slangasek: sladen is suggesting you will know such people ^^ Mar 23 21:40:30 Outside of IS, I don't know of any internally. Externally, I'm sure Linaro has people. Mar 23 21:42:08 Riddell: hi, skills and hardware for what? Mar 23 21:47:43 slangasek: for getting a working pandaboard Mar 23 21:47:51 Riddell: I've not been paying attention here, but how on earth did we decide that your inability to mount a rootfs was a video driver ug? Mar 23 21:47:55 s/ug/bug/ Mar 23 21:48:28 infinity: because ogra decided my previous advice on editing the boot files was incorrect and that was his next idea Mar 23 21:48:37 That error looks like what you'd see if you try to mount ext4 as ext3, though no idea how that would affect you, and only you. Mar 23 21:49:08 Riddell: Or, was that error with some whacky modified bootfiles that forced rootfstype to ext3? Mar 23 21:49:17 maybe Mar 23 21:49:34 * Riddell goes to make dinner with the last scrapes of his lifeforce Mar 23 22:00:37 infinity: I think what happened is he somehow clobbered his boot.scr, and that was the default boot environment from u-boot. Mar 23 22:01:11 I would have to reboot a system here and poke around in u-boot to confirm. Mar 23 22:06:18 As to his hdmi not working, up until today's image I would suspect a hardware issue. Everything up to today has worked fine for me and others on this channel. Mar 23 22:20:26 rsalveti: I filed a bug for today's video issue. Didn't see anything in LP prior to mine. Bug 963512. Mar 23 22:20:27 Launchpad bug 963512 in linux-ti-omap4 "Latest kernel updates broke video on omap4" [Critical,Confirmed] https://launchpad.net/bugs/963512 Mar 23 22:21:00 GrueMaster: ok, thanks Mar 23 22:21:14 need to run a bisect Mar 23 22:25:38 Well, looks like it was uploaded into 20120322 image. 20120320 had 3.2.0-1409.12 kernel. Mar 23 22:26:14 And I know it worked. Mar 23 22:27:47 yeah Mar 23 22:27:59 the difference between 09 and 10 is the "stable" fixes :-) Mar 23 22:28:27 Pfft. Figures. Mar 23 22:28:51 Last time a kernel guy told me that, they fried two of their babbages. Literally. Mar 23 22:29:06 That was in Dublin iirc. Mar 23 22:29:08 lol Mar 23 22:29:09 true Mar 23 22:29:33 quit Mar 23 22:30:15 resume Mar 23 22:39:26 suspend Mar 23 22:58:54 rsalveti: Do you know if someone will fix bug 943058 before 12.04 release? This is becoming critical, as the start-r rev boards replaced earlier versions and started shipping in December. Mar 23 22:58:55 Launchpad bug 943058 in linux-linaro-lt-mx5 "Kernel doesn't support usb on newer rev quickstart boards." [High,New] https://launchpad.net/bugs/943058 Mar 23 23:39:09 GrueMaster: I discussed that with the freescale landing team, and they said they would try to fix these issues in the next following days Mar 23 23:39:15 so I hope to get a better kernel before release Mar 23 23:39:50 cool. It would be a shame not to support the newer rev hw. Mar 23 23:41:29 rsalveti: What are the odds on that "better kernel" also being 3.2? :/ Mar 23 23:41:51 infinity: I believe that's the goal Mar 23 23:41:59 as currently 3.2 only works for mx6 Mar 23 23:42:27 rsalveti: If that goal is met, I'll buy you ridiculous amounts of beer at UDS (and you have no obligation whatsoever to take them back to the LT if you don't want to) Mar 23 23:42:46 infinity: hahah, that's easy ;-) Mar 23 23:43:02 heh. I'd add to that, but I'm not sure if I am going. Mar 23 23:43:23 GrueMaster: It's a short trip, drive down. Mar 23 23:43:28 well, it's a quite short travel for you Mar 23 23:43:30 yeah Mar 23 23:43:47 Then what, sleep in NCommander's car? Have you seen his car? Mar 23 23:44:04 Get a cot in Oli's room. Mar 23 23:44:11 He'll love that. Mar 23 23:44:31 Darth Vader sleeping mask and all. Mar 23 23:44:35 But I just meant swing down for a day and say hi. Mar 23 23:44:40 Ideally a party day. ;) Mar 23 23:45:01 Erm, it is an 8 hour drive, one way. Mar 23 23:45:17 That's short! Mar 23 23:46:06 Not for me. I'd have to stop and piddle every 2 hours, just to keep up with the coffee I'd need. Mar 23 23:48:51 infinity: While you are here, care to look at https://code.launchpad.net/~gruemaster/bacula/960761/+merge/99137 and tell me if I'm doing it right? It can wait until post beta thaw. Mar 23 23:50:59 * GrueMaster is learning the ropes of bug fixing where he can. Mar 23 23:52:22 GrueMaster: Erm, yeah. _Description implies that there are pofiles being merged into a final template. I assume that's not the case here? :P Mar 23 23:53:31 I have know idea what the implication is. In the case of the mysql package, I get a dpkg error on the one line with _Description. On the pgsql, all lines had the error. Mar 23 23:54:35 I haven't looked to see if there are pofiles for these, but since the error is being generated, I would say no. Mar 23 23:58:13 At the very least, the templates should be consistent. Either all _Description with po or not (s my patch changes). Mar 24 00:41:41 GrueMaster: templates should have _Description only if they're being postprocessed with dh_installdebconf to merge in translations from debian/po at build time Mar 24 00:42:01 and if you change something to use _Description, you should also be running debconf-updatepo as part of preparing the source Mar 24 01:09:53 slangasek: You're reading the diff backwards, he's changing it to NOT use _Description because the maintainer originally did, but doesn't appear to have a debian/po. Mar 24 01:10:08 I'm not reading a diff, I'm reading scrollback :) Mar 24 01:10:12 slangasek: Ahh. Mar 24 01:10:16 https://code.launchpad.net/~gruemaster/bacula/960761/+merge/99137 Mar 24 01:10:58 slangasek: Looks like the templates for bacula itself are right, but some cargo-culting oops happened on the *sql plugins, and they got _Description. Mar 24 01:11:08 slangasek: Which, I assume, means no templates at all. Mar 24 01:11:14 well, using templates without enabling translation support is not "right" Mar 24 01:11:28 * slangasek takes a look at the source Mar 24 01:11:30 Well, FSVO "right". Mar 24 01:12:02 slangasek: To be fair, I'm not sure what the expected behaviour would be here. dh_installdebconf is being called. But yeah, one set of templates uses Description, the other _Description, and no debian/po. Mar 24 01:12:06 it's even using dh_installdebconf in debian/rules Mar 24 01:12:09 slangasek: My change is exactly the opposite. Check the mysql.template and you will see. Mar 24 01:12:22 slangasek: And I haven't actually looked at the output, just that Tobin claims there are no templates for *sql. Mar 24 01:12:23 so the right solution here is to run 'debconf-updatepo' and *populate* debian/po Mar 24 01:12:41 slangasek: And to fix bacula.templates to use _? Mar 24 01:12:48 yes Mar 24 01:12:51 Fair enough. That is something I didn't know. Mar 24 01:12:51 Check. Mar 24 01:13:08 GrueMaster: Now we know. (And knowing is half the battle). Mar 24 01:13:16 pork chop sandwiches! Mar 24 01:13:26 Who's a body massage MACHINE? Mar 24 01:13:36 * GrueMaster thinks infinity sounds like a kids commercial. Mar 24 01:13:58 interesting, why does 'debconf-updatepo' fail here for me with an xgettext complaint Mar 24 01:14:08 oh yes Mar 24 01:14:16 you have to run 'debconf-gettextize' for the initial poing Mar 24 01:14:17 oh baby? Mar 24 01:16:04 slangasek: Okay, it seriously took me two minutes and many re-readings of "poing" before I realized it was neither a typo, nor a word I didn't know, but a verbification of "po". Mar 24 01:16:08 slangasek: Ouch. Mar 24 01:16:26 sed -i -e's/_Description/Description/' debian/*.templates && debconf-gettextize debian/*.templates Mar 24 01:16:29 infinity: you're welcome Mar 24 01:16:49 GrueMaster: ^^ that's the "right" fix here Mar 24 01:17:12 I always have great confidence when "right" is in quotes. Mar 24 01:17:19 Ok. Noted. Mar 24 01:17:24 although, here's a fun changelog entry Mar 24 01:17:29 bacula (3.0.3-3) unstable; urgency=medium Mar 24 01:17:33 [...] Mar 24 01:17:34 * This change means that all Bacula debconf templates are finally removed. Mar 24 01:17:34 As a result, no need for translations remains. Closing open translation Mar 24 01:17:34