**** BEGIN LOGGING AT Thu Sep 29 02:59:56 2011 Sep 29 03:15:31 jcrigby: cool Sep 29 03:26:50 jcrigby: did you fire up another build already? Sep 29 03:26:58 at the package recipe Sep 29 03:27:00 yes Sep 29 03:27:05 jcrigby: great Sep 29 03:27:34 https://code.launchpad.net/~linaro-maintainers/+archive/staging-overlay/+build/2813063 Sep 29 03:32:40 GrueMaster: Nevermind, no new images because ubiquity was FTBFS on i386. Yay for skew the other direction for once. :P Sep 29 03:33:40 GrueMaster: If I'm up late, I might re-spin some armel images for testing. Sep 29 07:49:08 Hi guys, I'm trying to boot my board but it stops at uncompressing kernel … I've checked my serial line is setup properly in uboot and there is no boot.src Sep 29 07:49:46 sniperjo_: it might be working but just not printing anything Sep 29 07:50:01 That line comes from the bootloader; you then need to tell the kernel where console= to put it Sep 29 07:50:16 twb: where can i do that ? Sep 29 07:50:22 It's args passed to the kernel Sep 29 07:50:46 e.g. qemu-system-arm -nographic -M versatilepb -kernel vmlinuz-3.0.0-1-versatile -initrd initrd.gz -append 'console=ttyS0 console=tty0 console=ttyAMA0,38400n8' Sep 29 07:51:24 in uboot i have setenv linux_args setenv bootargs console=${console} nfsroot=${serverip}:/dev/mmcblk0p2 ip=${ipaddr}:${serverip}:192.168.1.254:255.255.255.0::eth0:off; tftpboot 84000000 uImage; run linux_args; bootm 84000000; Sep 29 07:52:56 sniperjo_: And what is console set to? Sep 29 07:53:16 console=ttyS2,115200n8 Sep 29 07:53:50 Your board actually has a device called ttyS2? Sep 29 07:54:42 yes, its the console thats used when i boot in angstrom Sep 29 07:55:16 Potentially renamed in later kernels. Sep 29 07:55:52 Though just a shot in the dark. It could just as easily be the kernel just not booting too. :P Sep 29 07:55:56 infinity: by the way in still o n2.6.32, do i need to update my host computer Sep 29 07:56:46 Update your "host"? Sep 29 07:57:16 The annoying thing is/.. somewhere, one of the things i have done, I've got to " Uncompressing kernel…. done" and I've just read somewhere that sometimes thats because its set up or a guy enviroment Sep 29 07:57:31 as in, what i used to compile the kernel Sep 29 07:58:56 If you're using Angstrom sources, I'd recommend using whatever toolchain they recommend (ie: I don't know). Sep 29 07:59:18 For something closer to upstream or Linaro sources, the latest and shiniest oneiric would be saner. Sep 29 07:59:31 i really don't want to use angstrom Sep 29 07:59:53 so in that case i do need to update? Sep 29 08:00:07 I'm not even sure what device you're using... Sep 29 08:00:26 But if there are newer kernels out there that support it, that would be a good place to start. Sep 29 08:00:56 infinity: an atom netbook Sep 29 08:01:03 ... Sep 29 08:01:07 I mean your ARM device. Sep 29 08:01:32 ah, technexion tsunami, omap4540, 25mb ram Sep 29 08:01:58 ... Sep 29 08:02:08 whoops… 3530 Sep 29 08:02:19 You're trying to run Ubuntu on a pseudo-embedded system? Sep 29 08:02:24 256mb ram… unless I've got some crazy prototype Sep 29 08:02:38 Oh, 256 sounds less painful. :P Sep 29 08:03:22 I'm going to guess that our "omap" kernel is too Beagle-specific to work on it, eh? :/ Sep 29 08:03:41 But, our omap sources, with the Angstrom .config shoved at it, might Just Work. Sep 29 08:03:52 yeah, it dies straight away, not the same vendor id Sep 29 08:04:00 And for that, I'd definitely recommend compiling from an oneiric system, yes. Sep 29 08:04:28 Kernel shouldn't be checking vendor id at all. You mean our images do that (as in, our uBoot)? Sep 29 08:04:39 I'm saying you could try our kernel with their uBoot. Sep 29 08:04:52 But it's probably still missing some device support, since it's fairly Beagle-ish. Sep 29 08:05:12 infinity: I've defiantly tried it, and it definitely didn't work Sep 29 08:05:20 Either way, it's 2am, I should probably not be trying to be helpful, and go to bed. Sep 29 08:06:06 But some shiny newish linaro omap sources with the .config from your Angstrom kernel, built with an oneiric toolchain, may well work. Sep 29 08:07:26 ok, so i need to update my "host to one irc" and try and compile again Sep 29 08:46:52 * ogra_ hugs infinity, thanks for the installer quietening Sep 29 10:03:37 ogra_: NP. It was annoying me in testing. Sadly, no new images today with all my fixes yesterday, since ubiquity's FTBFS on x86. :/ Sep 29 10:03:54 ogra_: Yay for archive skew in the other direction for once? Sep 29 10:07:17 infinity, well, skew is skew :) Sep 29 13:04:38 hi! Sep 29 13:05:15 hey zumbi Sep 29 13:05:18 hi Sep 29 13:05:40 ogra_: could you pull dietlibc from debian/experimental for oneiric? Sep 29 13:05:56 * ogra_ can try ... we have final freeze today :) Sep 29 13:05:58 ogra_: it currently ftbfs on *buntu Sep 29 13:06:10 but experimental should have the fixes Sep 29 13:06:31 to have a working version, at least it works on debian armel/armhf Sep 29 13:06:34 infinity, can you sync that ? or do i need a bug etc etc Sep 29 13:06:51 I told doko but he seem to have missed it Sep 29 13:07:08 very likely, i think he is really overworked Sep 29 13:07:16 sure Sep 29 13:07:22 ogra_: thanks Sep 29 13:10:00 ogra_, please test build before syncing Sep 29 13:10:22 hmm, k Sep 29 13:10:42 i dont have the time and resources to do that today, can we still sync after the freeze ? Sep 29 13:10:49 doko, ^^^ Sep 29 13:12:58 GrueMaster: yesterday, weren't you suggesting that the problem with the verbose audio messages on console were due to a driver rename? s/SDP4430/Panda/? Sep 29 14:01:20 ppisati: it is, because then the current ucm files are useless after this change Sep 29 14:03:21 ppisati, i hope that fix got in in todays build ? Sep 29 14:03:41 else we will have lost sound again Sep 29 14:05:04 which fix? Sep 29 14:05:18 s/SDP4430/Panda/ Sep 29 14:05:25 well, i did the rename Sep 29 14:05:30 good Sep 29 14:05:33 no wait Sep 29 14:05:47 i'm using the jack on the back of my panda board Sep 29 14:05:49 as long as alsactl finds the right name all is fine Sep 29 14:05:51 and i always had audio Sep 29 14:06:14 the renaming only happened recently in andys branch i think Sep 29 14:06:17 the rename was just to silence the messages on console Sep 29 14:06:21 no Sep 29 14:06:29 all alsa settings fully depend on the name Sep 29 14:06:54 so i shouldn't have audio, right? Sep 29 14:07:03 but i have it Sep 29 14:07:14 not hdmi, using the lower audio jack Sep 29 14:08:31 with the recent images and the new kernel ? Sep 29 14:08:51 i.e. on a fresh install where alsactl didnt create the defaults yet Sep 29 14:09:07 yes Sep 29 14:09:10 ah Sep 29 14:09:59 where does it store the config? i mean alsa Sep 29 14:10:19 its tricky to get rid of it Sep 29 14:10:33 doh Sep 29 14:11:04 best is to take the SD in your PC and remove /var/lib/alsa/asound.state Sep 29 14:11:11 *best Sep 29 14:11:14 ok Sep 29 14:11:17 i'll try Sep 29 14:11:58 (it gets recreated on every shutdown, so its not easy to do it on a running system without hacking around in upstart jobs and udev rules) Sep 29 14:12:15 single user and i axe it Sep 29 14:12:18 btw Sep 29 14:12:41 to check the renaming: /proc/asound/cards should be equal to 2.6.38.xx, right? Sep 29 14:12:50 yep Sep 29 14:12:59 ok Sep 29 14:13:21 btw, i thgought fix could go in even after final freeze Sep 29 14:13:24 or maybe not? Sep 29 14:13:29 sure Sep 29 14:13:34 ah ok Sep 29 14:13:44 but we still havent had much testing for the sound stuff Sep 29 14:13:50 so the earlier the better Sep 29 14:13:50 ok ok Sep 29 14:13:57 axe alsa state Sep 29 14:14:00 test again Sep 29 14:14:07 if not working, apply the rename and retest Sep 29 14:14:10 would be our first release where sound works on release day Sep 29 14:14:16 :) Sep 29 14:22:35 ogra_: i don't want to be pessimistic... but i don't have audio... granted that I am not using your kernel, but I am using the kernel which was used by ppisati to make the ubuntu kernel... Sep 29 14:22:48 hmm Sep 29 14:23:00 well, lets see what paolos test turns out now Sep 29 14:23:46 testing with an existing state file is indeed kind of moot Sep 29 14:28:59 * diwic has not understood why we need this UCM thing in the first place, whereas one could write a PulseAudio profile instead to do the same thing. Sep 29 14:30:11 UCM was supposed to help ;-) Sep 29 14:30:17 diwic, i think the idea was to use both Sep 29 14:30:41 ogra_: on one panda that I have dist-upgraded since alpha2, audio works fine ;-) Sep 29 14:31:08 hmm Sep 29 14:31:17 ogra_, no, PulseAudio profiles/paths, and UCM, are overlapping concepts. Sep 29 14:31:22 so should we roll back to the kernel we had at A2 ? Sep 29 14:31:24 :) Sep 29 14:31:37 if we use our images from yesterday evening + our kernel --> no audio Sep 29 14:31:41 ;-) Sep 29 14:31:49 ndec, iirc that was the natty kernel still Sep 29 14:32:00 here we go... Sep 29 14:32:20 it's even easier to find in the archive! Sep 29 14:32:57 for UCM to work, we need to have /usr/share/alsa/ucm/Panda/Panda.conf, is that correct? Sep 29 14:34:42 i think so Sep 29 14:35:23 ogra_, I'm ready to push the final u-boot source once someone approves Sep 29 14:35:42 oh, go ahead Sep 29 14:35:52 oh, ok Sep 29 14:36:25 i'll do the paperwork in a minute on the bug Sep 29 14:37:08 ogra_: did you submit the change (s/SDP4430/Panda/) to alsa? Sep 29 14:40:19 ndec, i think diwic added it in an SRU Sep 29 14:41:00 ogra_: can you send me (again) the link with the queue of packages waiting for entering in the archive? Sep 29 14:48:12 * diwic does not remember changing "SDP4430" to "Panda". Sep 29 14:50:24 rsalveti: ogra_: ^^ if we change the device name in the kernel and not in alsa, we might be in troubles.. Sep 29 14:53:44 ok, after deleting the Sep 29 14:53:48 yup Sep 29 14:53:51 ndec: I have the latest alsa source.. I find no reference to "Panda" there. It is all ADP4430. Sep 29 14:53:53 /var/lib/alsa/asound.state file and rebooting Sep 29 14:54:17 we have the unfamous "pulseaudio go nuts and swamp the console with lo messages" and 100% cpu usage for this Sep 29 14:54:47 renaming the driver s/Panda/SDP4430/ we get back to a "normal" logging on console Sep 29 14:54:51 nut no audio :( Sep 29 14:54:53 but Sep 29 14:55:15 ppisati: Hence why I filed it as a kernel bug originally. While the messages are directly from Pulseaudio, it is because of a kernel change. Sep 29 14:55:47 GrueMaster: ok, so i can shove the driver name change in, but still there's some alsa tweaking to do Sep 29 14:56:02 else we won't have audio after a fresh install Sep 29 14:56:29 alt!!! Sep 29 14:56:34 the volume was low Sep 29 14:56:36 @#$#@$ Sep 29 14:56:39 we have audio! :) Sep 29 14:56:50 A fresh install should just work. In my earlier test, renaming SDP4430 in all of the ucm files and rebooting worked fine. Sep 29 14:56:52 yes, we haz it Sep 29 14:56:59 Cool. Sep 29 14:57:22 ok, let me commit it then Sep 29 14:57:27 GrueMaster: ? there is already SDP4430 in all UCM files. Sep 29 14:57:42 The change to Panda is not in the upstream alsa git tree, so I don't know where it came from. Sep 29 14:57:51 ndec, right. Not sure where the rename to Panda came from. Sep 29 14:58:25 I only did that as a quick experiment to see why it wasn't working. Sep 29 14:58:45 UCM shouldn't have changed since Natty. Sep 29 15:01:15 didn't we have an omap4 audio lp bug? Sep 29 15:01:19 crap, can't find it... Sep 29 15:01:50 I've followed a tutorial for getting ubuntu on my board ,I've set ups tftp server and such so it downloads the kernel and rootfs , it stop at "uncompressing…booting kernel ".. from google lots of people have that because their console isn't set right, but my bootargs use ${console} which works when the prebuilt sdcard is in there Sep 29 15:02:21 ppisati: bug 820466 Sep 29 15:02:23 Launchpad bug 820466 in linux-ti-omap4 "No sound on omap4 pandaboard" [High,Fix released] https://launchpad.net/bugs/820466 Sep 29 15:21:11 ogra_: ppisati: diwic: if i rename all the SDP4430 to Panda in /usr/share/alsa/ucm, sound is working with a kernel that calls the sound card Panda Sep 29 15:21:54 ndec: Upstream alsa still names it SDP4430. Sep 29 15:22:34 this is for the Blaze board Sep 29 15:22:53 And changing the kernel name is easier (one line if I read the driver correctly). Sep 29 15:23:06 are you sure it's upstream btw? Sep 29 15:23:16 but wrong... Sep 29 15:23:23 diwic, I'll upload your openjdk/pulseaudio arm fix, just with a cast to jlong instead Sep 29 15:23:55 i meant easier but wrong. the real name in the kernel should be Panda Sep 29 15:23:56 ndec: I have the alsa-project git tree here and up-to-date. Sep 29 15:24:47 Why? Is it different from platform to platform? Sep 29 15:24:56 yes Sep 29 15:25:46 GrueMaster: where in alsa upstream did they put the conf files? Sep 29 15:26:26 I haven't seen any alsaucm files upstream yet, but they may be in a git tree I don't have yet. Sep 29 15:26:38 Will check after the meeting. #ubuntu-meeting. Sep 29 15:27:29 * ppisati refrains from commiting anything then... Sep 29 15:35:56 ndec: Since we have freeze today, it makes more sense to fixe the one line in the kernel than the multiple ucm files. If TI wants it renamed to Panda, we can do it next cycle, but we will need to know way before freeze. Sep 29 15:49:50 GrueMaster: I'd rather build alsa than another kernel. Sep 29 15:49:55 To be fair... Sep 29 15:50:35 Really? Alsa is a bit more intrusive. Kernel is a separate tree affecting only omap4. Sep 29 15:51:04 There's nothing intrusive about s/SDP4430/Panda/ unless you have a sound card called SDP4430 on your amd64 machine. Sep 29 15:51:27 No, but the packages are arch all. Sep 29 15:52:12 That doesn't make the fix any scarier. Sep 29 15:52:57 On the kernel, it is a single line. Sep 29 15:55:33 It's one line in alsa if I do it with sed in debian/rules instead of a patch. (which I wouldn't do, but my point is that, conceptually, this isn't a large or unreviewable change). Sep 29 15:55:58 And alsa is much faster to build/test. And the fix belongs in userspace anyway, we'll have to do this all over again in 2 weeks if we don't do it now. Sep 29 15:56:09 (Or forget for 3/4 of the cycle and whine at the end that sound doesn't work, and argue again) Sep 29 15:58:14 infinity: I don't trust it. Remember, ext4 is a simple change too. Right now I want the quick fix. Since we have to rebuild the kernel anyways. Sep 29 16:00:04 The "correct" thing to do is fix it upstream. Since alsa doesn't know about "Panda", this has not been done. Alsa does know about the SDP4430 though. Sep 29 16:00:33 GrueMaster: can you point me to where SDP4430 is reference in upstream alsa? Sep 29 16:01:38 GrueMaster: ext4 was a simple change. It landed at the same time as a bunch of mx5 stuff that make people blame ext4 for other breakages though. :P Sep 29 16:02:12 infinity: It wasn't a simple change as it required a bunch of jasper changes before we had working images. Sep 29 16:02:29 See above. Sep 29 16:02:56 And, for the record, SDP4430 isn't upstream in alsa at all. Sep 29 16:02:59 It's a local patch. Sep 29 16:03:07 there were mx5 changes *and* a completely new image type (ac100) *and* ext4 at the same time Sep 29 16:03:14 We fix it in our packages, we're done. Sep 29 16:03:32 ndec: I would, but the alsa-project gitweb interface seems to be broken. The tree is git://git.alsa-project.org/alsa-kernel.git and the file is soc/omap/sdp4430.c Sep 29 16:03:55 if you want to blame anything, blame ac100, that caused the most intrusive changes Sep 29 16:04:00 Oh, you're talking driver filenames? That's meaningless. Sep 29 16:04:06 It's only the UCM stuff that needs fixing. Sep 29 16:04:41 yeah, the filename doesnt matter, only whats in /proc/asound/cards Sep 29 16:04:50 GrueMaster: in this file you will find this: Sep 29 16:04:50 if (machine_is_omap_4430sdp()) Sep 29 16:04:50 snd_soc_sdp4430.name = "SDP4430"; Sep 29 16:04:50 else if (machine_is_omap4_panda()) Sep 29 16:04:50 snd_soc_sdp4430.name = "Panda"; Sep 29 16:04:53 ;-) Sep 29 16:05:08 yes, SDP4430 used to be for the blaze Sep 29 16:05:40 (and probably still is, i havent seen a recent blaze) Sep 29 16:05:47 what I copy/pasted is where we are doing the device name, e.g. what ends up in /proc/sounds/cards, and what is used by alsaucm config file Sep 29 16:05:50 ndec: I just ran a git pull and didn't see it when grepping for Panda. Sep 29 16:05:55 right Sep 29 16:06:52 GrueMaster: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-oneiric.git;a=blob;f=sound/soc/omap/sdp4430.c;h=424ca3d512f6b15c3f2318bbcd6644a77fd052ac;hb=ti-omap4 Sep 29 16:07:06 But just to err on the side of caution, I am recloning my tree. Sep 29 16:07:31 ndec: That is our tree. Not upstream. Sep 29 16:07:54 that will be upstream at some point. Sep 29 16:08:00 And when did the addition of Panda get there? Sep 29 16:08:09 if we were using upstream stuff only, that wouldn't too nice... Sep 29 16:08:43 If we revert our omap4 kernel to vanilla, we may as well drop the subarch. Sep 29 16:09:12 * GrueMaster is getting tired of this argument. Sep 29 16:09:25 you can't drop the subarch it will break flash-kernel Sep 29 16:09:27 I really don't care what you guys do. Just fix the thing. Sep 29 16:09:34 Yes. In the time it's been happening, libasound could have been fixed, uploaded, and built. Sep 29 16:10:10 Same with the kernel. Bad argument. Sep 29 16:10:39 infinity: by the way, don't rename the files from SDP4430 to Panda, duplicate them. so that you get support for Panda and SDP4430 board Sep 29 16:10:39 Bug was filed a while ago. If I had been doing daily testing instead of server testing, I would have been all over this weeks ago. Sep 29 16:10:51 ndec: That was the plan. Sep 29 16:11:07 * ndec remembers you talked about sed ;-) Sep 29 16:11:23 ndec: Was an example. :) Sep 29 16:11:26 so you are going to fix in user space, right? Sep 29 16:11:34 *fix it Sep 29 16:12:16 GrueMaster: 26 minutes versus 9 hours seems to support what I said. But whatever. Sep 29 16:12:31 ppisati: That's where the fix belongs. I see no point kludging it. Sep 29 16:12:32 GrueMaster: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-oneiric.git;a=commit;h=14a8827efa05606e7d8776d11ad2e030c12d427c Sep 29 16:12:39 ok Sep 29 16:12:53 it's signed-off by liam. a good sign that it might end up upstream Sep 29 16:13:36 hi! Question: is it able yet to set the screen resolution on latest ubuntu? Sep 29 16:13:42 on pandaboard Sep 29 16:13:43 :) Sep 29 16:14:02 I second that question :D Sep 29 16:14:47 festhead: yes. at least on all the screens i have seen... Sep 29 16:14:53 it picked the largest one at boot. Sep 29 16:15:20 and where do you get the options? Sep 29 16:15:28 now if you want to change the resolution, you need support for XRandR in the driver, so you need to install our DDK driver from the TI PPA Sep 29 16:15:31 and it works too. Sep 29 16:16:06 it just pick the bigger resolution supported by the screen. no option. Sep 29 16:17:13 well, you can force a resolution on kernel cmdline, no ? Sep 29 16:17:19 was that dropped ? Sep 29 16:17:34 can you? Sep 29 16:17:34 ogra_, seems so, nothing I tried worked there.. it just keeled back over to 640x480 Sep 29 16:17:56 ndec, you could in the past, yeah, hdmimode= and friends Sep 29 16:18:16 ogra_: if you use the TI DDK driver yes you can using drm Sep 29 16:18:25 indeed Sep 29 16:18:26 * ndec thinks that hdmimode has been removed Sep 29 16:18:43 * ogra_ isnt up to date with hacks :) Sep 29 16:18:55 and then i also have a proper monitor Sep 29 16:19:07 the good news is that the hack was removed in this case... in general we only add more hacks ;-) Sep 29 16:19:10 so i wouldnt need such hacks Sep 29 16:19:15 heh Sep 29 16:19:56 Yeah, I kind of get a giggle out of my Panda having the nicest monitor in the house. Sep 29 16:21:38 ogra_: infinity: i have to go... but i confirm that adding ucm/Panda fixed the audio issues on at least 2 different boards here... Sep 29 16:21:40 byw Sep 29 16:21:42 bye Sep 29 16:21:57 infinity, upload away i'd say Sep 29 16:21:58 ndec: Yeah. I'm testing and uploading within the hour. Sep 29 16:22:27 After I find coffee and a snack. Sep 29 16:23:02 Don't forget udev will also need to be changed. Sep 29 16:23:06 Possibly. Sep 29 16:25:56 GrueMaster: I should hope not. Sep 29 16:26:12 GrueMaster: This isn't about kernel driver names, it's about alsa device names. Sep 29 16:28:15 infinity: There is a udev rule to load the alsaucm files. I'll doublecheck to see how it is set up. Sep 29 16:32:06 It is currently set to "ATTRS{id} =="SDP4430", RUN+="/usr/bin/alsaucm set _verb HiFi" and "ATTRS{id} =="SDP4430", RUN+="/usr/bin/alsaucm set _verb Record" Sep 29 16:32:45 Not sure where ATTRS is set from, but I'm just saying heads up. Sep 29 17:06:24 So, entirely unrelated, but noticed in testing this. "alsaucm set _verb Hifi" works great. "alsaucm set _verb Record" fails. Sep 29 17:07:09 doko, xranby, ok, did both patches (pulseaudiodataline.patch and contextgetstate.patch) make it into openjdk-6 - 6b23~pre10-0ubuntu4 and did anybody actually test contextgetstate.patch ? Sep 29 17:08:43 infinity, did you test with an mp3 player or the like ? mics dont work Sep 29 17:09:18 infinity, its a real "line in" Sep 29 17:10:19 ogra_: No, I mean the command fails. Sep 29 17:10:36 UH ! Sep 29 17:10:43 that should definitely not happen Sep 29 17:10:54 bug it :) Sep 29 17:11:24 playback is definitely more important and if that works i'm already happy Sep 29 17:12:27 It would appear the PDM switch for AMIC no longer exists. Looking to see if there is an alternative (renamed) control. Sep 29 17:13:07 well, the HW doesnt support mics ... it probably just reflects the reality now Sep 29 17:13:20 For now, commenting out the two lines makes alsaucm work. Sep 29 17:13:33 (or does AMIC not mean analog microphone ?) Sep 29 17:13:44 It supports audio in and that PDM enabled input. Sep 29 17:13:54 is it correct that is ubuntu 11.06 the ablility to change the resolution on a pandaboard is supported? Sep 29 17:13:59 Like a mute button. Sep 29 17:14:03 right, but only high level sources Sep 29 17:14:17 festhead, there is no ubuntu 11.06 Sep 29 17:14:37 great :P Sep 29 17:14:50 there will be ubuntu 11.10 and there is 11.04 Sep 29 17:15:15 and is it supported in 10.10? Sep 29 17:15:47 well, it is supported by the closed driver in all releases afaik Sep 29 17:15:59 diwic, contextgetstate.patch: did change the cast to jlong. no, the build would have taken too long. but imo it can't get worse Sep 29 17:16:05 but not by the free framebuffer driver we ship by default Sep 29 17:16:49 doko, ok, just wanted to make sure you saw there were two patches. Let's hope it works then :-) Sep 29 17:17:12 ogra_, so no support? Sep 29 17:17:20 ?? Sep 29 17:17:40 sure, but you need the pvr driver from the TI PPA Sep 29 17:17:48 ok Sep 29 17:17:55 as ndec stated a while ago already Sep 29 17:18:03 thanks Sep 29 17:18:15 welcome := Sep 29 17:18:17 :) Sep 29 17:20:44 hmm, the log in the banshee bug looks suspiciously like an ubuntu-one issue Sep 29 17:21:38 it seems to die in a function called U1RequestChrome Sep 29 17:23:01 * ogra_ removes banshee-extension-ubuntuonemusicstore for a test Sep 29 17:24:42 hmm, sadly no Sep 29 17:29:11 Hmm. No record device according to pulseaudio. This may take a bit to figure out. Sep 29 17:30:42 might even be the kernel Sep 29 17:31:21 GrueMaster, does the device show up in "arecord -l" Sep 29 17:32:18 There are a lot of devices that the SOC can handle. Question is figuring out the right one for alsaucm so that pulseaudio works. Sep 29 17:33:27 right. Sep 29 17:39:28 ogra_, still no succes, i allready had those drivers. i'm using a pico projector and my resolution is only 848x480. no succes with xrandr Sep 29 17:40:19 how is that attached ? HDMI ? Sep 29 17:40:55 yes Sep 29 17:41:37 well, try to catch robclark, he knows about graphics drivers and might have more ideas, i'm running out of them now Sep 29 17:42:23 robclark, any ideas? :) Sep 29 17:42:39 its likely that your projector doesnt provide an EDID table the driver can handle Sep 29 17:46:28 thanks for the help anyway Sep 29 17:53:12 festhead, ogra_, no EDID for pico.. Sep 29 17:53:23 aha ! Sep 29 17:53:24 it and LCD panels would only support current resolution Sep 29 17:54:52 no ability to force resolution? Sep 29 17:57:17 boot with a monitor that has a supported resolution for the pico ... then plug over ? Sep 29 17:59:18 haha! thought of that, but i have no hdmi-monitors. i tried the standard resolution (by plugging nothing in) but that was even smaller Sep 29 18:00:12 yeah, thats 640x480 Sep 29 18:01:21 festhead: Do you have a DVI monitor? You can use a DVI<>HDMI cable. Sep 29 18:02:01 no only big vga monitors Sep 29 18:02:18 ouch Sep 29 18:02:30 hmm, oem-config-remove is just running here for whom it intrests ... Sep 29 18:02:40 on the most recent image Sep 29 18:02:43 robclark: Does the kernel module still accept parameters for video resolution? Sep 29 18:03:00 ogra_: Which image? Sep 29 18:03:06 desktop panda Sep 29 18:03:13 GrueMaster, yes, although it is different from the omapfb bootargs of old Sep 29 18:03:23 initramfs-tools is running atm Sep 29 18:03:38 other caveat: for displays that don't support EDID, there is only one valid resolution to choose from Sep 29 18:03:56 ie, you can't set arbitrary resolution on the LCD panels on blaze, for example Sep 29 18:04:06 or pico projector Sep 29 18:04:43 * robclark is referring to projector built in on blaze, not one connected over hdmi/dvi.. Sep 29 18:04:59 for something connected over hdmi/dvi, it all depends on what the EDID says Sep 29 18:05:24 ok. Is there a way to fix festhead with a kernel cmdline parameter? Is it documented somewhere? Sep 29 18:06:14 well, *if* he is using the drm driver, and connecting over hdmi/dvi, and still just gets one choice of resolutions, that probably means that it is all that the display supports Sep 29 18:06:23 no amount of bootargs would change that Sep 29 18:06:48 if he is using older omapfb stuff instead.. then there are some bootargs that would help Sep 29 18:07:57 GrueMaster, and finished, i'm at lightDM proper Sep 29 18:08:18 ogra_: What image? What platform? Sep 29 18:08:30 again pnada desktop latest image Sep 29 18:09:03 sandisk class4 4G card though ... vers slow (one of the blue ones) Sep 29 18:09:33 hehe, and the boot partition is automounted Sep 29 18:09:46 Hmm. Will have to test on other cards. My 16G Class 6 Micro Center SD fails consistently. Sep 29 18:10:10 and apport pops in my face Sep 29 18:11:14 Nice. I just got a segfault probing for recording devices on my panda. Sep 29 18:11:24 kernel oops. Sep 29 18:12:45 ogra_: The boot partition rename thing hasn't landed in current images yet. Sep 29 18:12:50 ogra_: But the fix is in. :P Sep 29 18:13:28 yep, thats why i laughed Sep 29 18:14:02 but still no TI icon Sep 29 18:14:05 grmbl Sep 29 18:15:11 Where's the TI icon meant to be coming from? Sep 29 18:16:16 jasper Sep 29 18:16:37 well, its *meant* to be coming from a ti-ppa package but that always slipped Sep 29 18:16:56 that jasper does it is one of these eternal interim solutions Sep 29 18:17:14 its there and all, the prob is that it dopesnt show up in the launcher Sep 29 18:18:24 the prob here is that we need to append to the gsettings value for the launcher favorites Sep 29 18:18:34 without trashing them Sep 29 18:23:12 ah, the favorites are defined in /usr/share/glib-2.0/schemas/com.canonical.Unity.gschema.xml Sep 29 18:23:48 ogra_: if it's too much troubles i don't care getting rid of the icon... it would be a little less user friendly, but we can live with that. Sep 29 18:24:03 of course they are. o_O Sep 29 18:24:08 ndec|home, well, i still have until 23:00 Sep 29 18:24:28 GrueMaster, in natty they were somewhere completely else :) Sep 29 18:24:30 i don't want you to feel bad if you don't make it ;-) Sep 29 18:24:51 And in P, they will change yet again. Sep 29 18:24:59 ndec|home, k, but i'll try my best at least ... its one hack less if i dont make it though Sep 29 18:25:02 what is 'they'? Sep 29 18:25:19 ndec|home, the gnome desktop defaults Sep 29 18:25:24 ah ... Sep 29 18:25:37 in gconf mangling the defaults was easy Sep 29 18:26:06 gsettings/dconf requires the config files to be compiled after changing them ... thats a bit more tricky Sep 29 18:26:50 apparently thats an improvement Sep 29 18:27:08 i havent found out for whom yet, but it surely is for someone Sep 29 18:28:17 * ogra_ wonders why favorite entires all need to have the .desktop suffix in the name ... Sep 29 18:28:31 its not that there would be any other file format it accepts Sep 29 18:31:14 sniff ... just adding it doesnt help apparently Sep 29 18:31:52 * ogra_ wonders if unity-2d has its own favorites Sep 29 18:32:15 Probably. Sep 29 18:32:30 Add something to the bar, then find/grep for it. Sep 29 18:33:29 (usually much faster than figuring out where, if any, updated documentation may be) Sep 29 18:34:43 GrueMaster, i want to change the defaults, drag/drop adding it will place it somewhere totally different Sep 29 18:34:53 (in the config) Sep 29 18:34:58 Will it? Sep 29 18:35:01 yes Sep 29 18:35:48 it just dumps the .desktop file in a certain place in your user config Sep 29 18:35:58 while i need to set a configuration key Sep 29 18:36:02 Should be somewhere in either /etc or /usr/share then (if standards are being followed). Sep 29 18:36:18 right Sep 29 18:36:26 in /usr/share/glib-2.0/schemas/com.canonical.Unity.gschema.xml Sep 29 18:37:17 But that is unity. Not unity-2d. Or do they share the same files (like they should). Sep 29 18:37:23 yes Sep 29 18:37:32 they use the same backend for the dash Sep 29 18:37:36 and lanucher Sep 29 18:55:08 * ogra_ finds it irritating that gnome-power-manager turns the battery icon red here and shows 2:12 next to it Sep 29 18:57:29 ogra_: On your panda install, did you notice the background behavior during oem-config? Sep 29 18:57:40 bah, clicking the ti icon does nothing, i guess the gnome-open apt: protocol changed too Sep 29 18:57:53 GrueMaster, you mean wallpapers ? Sep 29 18:58:07 yes, its ugly but i doubt we can do much abouot it now Sep 29 18:58:18 Well, the black screen that reveals the wallpaper when you move the mouse. Sep 29 18:58:30 ARGH !!!!!!!!!!!!! Sep 29 18:59:00 GrueMaster, yeah, thats an old xfbdev bug, we had it on beagles and babbages too Sep 29 18:59:21 great, so the TI icon doesnt work because gnome-open doesnt exist anymore Sep 29 19:02:41 ah, its xdg-open nowadays Sep 29 19:03:07 yea, gnome-open is so last cycle. :P Sep 29 19:03:42 its libgnome2 ! Sep 29 19:04:41 well, igot it working if you navigate to /usr/share/applications and click it now Sep 29 19:04:59 but still not in the launcher Sep 29 19:05:09 sigh. 20110928 daily fails oem-config on another SD card of mine. yea. Sep 29 19:10:30 oh, btw.. is there any reason when you install an optimized driver for Xorg, it's still loading fbdev in the background? Sep 29 19:10:34 Ah, second run seems to be removing oem-config now. Sep 29 19:10:39 we got some really odd performance problems down to that on the Efika Sep 29 19:11:11 * ogra_ points Neko to #ubuntu-x Sep 29 19:11:35 yeah I thought as much.... Sep 29 19:11:37 YAY ! Sep 29 19:11:42 happens on Panda too, oddly... Sep 29 19:11:42 * ogra_ got it Sep 29 19:11:46 ogra_: \o/ Sep 29 19:11:50 i see the TI icon in the guest session Sep 29 19:12:17 seems dconf is really really evil and writes the full defaults as a binary blob into the user dir Sep 29 19:12:22 ogra_: But does it blend? Sep 29 19:12:32 so later system default changes dont get noticed Sep 29 19:13:41 GrueMaster, well, its a bit brownish Sep 29 19:13:42 Given that no one needs that icon except on fresh installs (well, in theory), that seems reasonable. :P Sep 29 19:18:04 infinity, its still a bug bjut seems its a unity one Sep 29 19:18:26 * ogra_ thought the new system was that bad, but seems it only applies to the launcher defaults Sep 29 19:22:14 wow, doing a netinstall to an SD takes a loooong time. Started it at ~11:30. Still running. Just asked for task selection. Sep 29 19:22:25 And this is from a local mirror. Sep 29 19:22:47 ogra_: I'm not sure I'd call that a bug. If it is, it's a bug in pretty much every X application. Sep 29 19:23:13 ogra_: System defaults are applied on first-run and/or user creation, they don't get re-applied. Sep 29 19:25:39 infinity, gconf worked differently Sep 29 19:59:25 hi! where can i find a free compiler for the dsp (TMS320DMC64X+) on the omap4430 chip? Sep 29 20:09:21 infinity, could you take a cross check on commits 172 and 173 in the jasper-initramfs tree ? Sep 29 20:09:38 s/take/do/ Sep 29 20:16:46 ogra_: sed on an XML file, really? There's no prescribed sane way to do this? Sep 29 20:17:02 infinity, well, i stole from casper Sep 29 20:17:14 so i assumed its a blessed way Sep 29 20:17:36 the xml file is the input i dont think there is a saner wayx Sep 29 20:17:53 (teh actual config gets generated from it as a binary blob) Sep 29 20:18:32 * infinity sees no such code in casper... Sep 29 20:18:36 http://paste.ubuntu.com/699291/ Sep 29 20:19:57 i could instead put an com.canonical.Unity.gschema.override in place but i still need to pull the key out of the xml file and generate a new value Sep 29 20:20:08 oh, and .override files are .ini format :P Sep 29 20:20:39 (for consistency i guess :P ) Sep 29 20:21:06 Anyhow, it does the same thing as what you just pastebinned. So, yay for that? Sep 29 20:21:12 But I still can't find that in the casper source. :P Sep 29 20:21:30 might be in some edubuntu package that adds to casper Sep 29 20:21:38 Possibly. Sep 29 20:21:51 its all edubuntu bits and stgraber actualyl gave it to me in -desktop Sep 29 20:22:16 I'm going to assume you've tested that shell on your system in an isolated script? Sep 29 20:22:27 But, if the edubuntu bits work, this should too. Sep 29 20:22:29 so ok to upload ? (i might have it written differently, but i know the casper code has been tested already) Sep 29 20:22:44 Yeah, I'm all for cargo-culting tested code. Sep 29 20:22:49 hehe Sep 29 20:22:57 * ogra_ rolls the package Sep 29 20:23:09 ogra_: it's from edubuntu-live Sep 29 20:23:14 ah Sep 29 20:23:18 Oh. Sep 29 20:23:25 ogra_: Quoting. Sep 29 20:23:33 * ogra_ looks Sep 29 20:23:42 ogra_: [ -e $FOO ] and [ -e "$FOO" ] behave very differently. Sep 29 20:23:52 yeah, seeing it Sep 29 20:24:12 (Shouldn't be an issue, since you explicitely set it to a string, but I'm paranoid. Sep 29 20:24:13 fixed Sep 29 20:24:15 ) Sep 29 20:24:22 and you shoudl be Sep 29 20:24:32 especially on my code past 8pm local time Sep 29 20:24:38 :) Sep 29 20:24:58 ogra_: I actually learned about the override stuff later on but haven't had a chance to test them yet and IIRC you still need to update the binary db, so it's just cleaner (but I usually don't care much about cleanliness in a casper hack as long as it works ;)) Sep 29 20:25:12 yeah Sep 29 20:25:28 i dont actually think the .override file gains you much Sep 29 20:25:35 Well, the obvious problem with the hack is that it fails to work if the entry that you're anchoring your regex on goes away. :P Sep 29 20:25:45 yeah Sep 29 20:25:51 thats easier in casper indeed Sep 29 20:25:56 right, but the override would need to hardcode the whole list, not much better :) Sep 29 20:25:58 since yuo can tie to ubiquity Sep 29 20:26:08 exactly Sep 29 20:26:15 thats waht i didn in gconf though Sep 29 20:26:24 it's fairly safe to assume that something using casper will have ubiquity installed (at least it's for Edubuntu) :) Sep 29 20:26:50 right, we dont use casper Sep 29 20:27:07 and dont have an ubiquity item in the launcher Sep 29 20:27:27 ogra_: we do at the time that jasper-setup runs. :) Sep 29 20:27:51 i would actuallyx have preferred to pull the whole value out of the file, but grepping in xml is just moot Sep 29 20:28:34 infinity, we do what ? use casper or have an ubiquity item ? Sep 29 20:28:46 ogra_: The latter. Sep 29 20:28:51 i think the latter is put in place by casper Sep 29 20:28:56 ogra_: The launcher only goes away after oem-config runs. Sep 29 20:29:09 oh, really ? Sep 29 20:29:14 festhead: http://linux-c6x.org/wiki/index.php/Downloading/Installing_Software Sep 29 20:29:17 * ogra_ has never checked Sep 29 20:29:29 i always thought casper handles it Sep 29 20:29:31 ogra_: It's why we've had an open bug until 3 days ago about an "install icon on the launcher". Sep 29 20:29:48 festhead: http://omappedia.org/wiki/Syslink_Project#Ducati.2FTesla_IPC_Source_Code Sep 29 20:29:48 oh, indeed Sep 29 20:29:56 heh, i think i even commented on it Sep 29 20:30:30 ndec: thanks! Sep 29 20:30:54 jussi: saw that you got the answer on #pandaboard ;-) Sep 29 20:31:19 ogra_: Anyhow, whatever. If that works, it's good enough for me for now. It's an ugly hack regardless. Sep 29 20:31:28 yep Sep 29 20:31:40 well, all that PPA stuff should be in a package proper Sep 29 20:32:13 we should do that when implementing the "jasper needs to die in fire" spec Sep 29 20:36:05 aaand ... uploaded Sep 29 20:36:35 GrueMaster: ogra_: infinity: you knew that alsa-lib changes was done in linaro ubuntu overlay PPA? https://launchpad.net/~linaro-maintainers/+archive/overlay/+packages Sep 29 20:36:52 ndec|home, no :( Sep 29 20:37:02 rsalveti: why would such a change be done in the overlay PPA before it gets done in ubuntu? Sep 29 20:37:03 why wasnt it just fixed in the archive Sep 29 20:37:08 ;-) Sep 29 20:37:17 ndec|home, i suspect because they use natty Sep 29 20:37:51 true, they use alsa-lib from oneiric.. Sep 29 20:38:02 though we have a policy in ubuntu that fixes happening in an SRU (which such a PPA effectively is) have to happen in the dev release as well Sep 29 20:38:33 we should probably discuss how to fix that for teh future at UDS Sep 29 20:39:22 well, at least we have it in the archive too ;-) Sep 29 20:39:41 now, yes Sep 29 20:40:05 and with wasting at least 1h of three devs for discussing it Sep 29 20:40:34 i think we can do better ;) Sep 29 20:41:00 22min to freeze ... Sep 29 20:41:12 i think i'll call it a feierabend :) Sep 29 20:41:51 your first release with audio? Sep 29 20:41:54 ndec|home, so we will need an oneiric metapackage in the ppa ... do you want to do that ? Sep 29 20:42:00 yeah, that too :) Sep 29 20:42:06 i will do that. Sep 29 20:42:23 we have the gfx packages already in the ppa. Sep 29 20:42:24 thx, essentially i think its just copying the existing one and adjusting the deps Sep 29 20:42:33 for whats there atm Sep 29 20:42:34 we will upload the meta pkg Sep 29 20:42:39 great, thx Sep 29 20:42:51 icon is back too :) Sep 29 20:42:55 ;-) Sep 29 21:03:43 ndec|home: ogra_: first, we're now using the same kernel Sep 29 21:04:02 second, I said already that I had the changes needed for the tilt before the tilt was merged at ubuntu Sep 29 21:04:17 third, I fixed this *yesterday* :-) Sep 29 21:04:39 for the release, then before pushing to ubuntu I first need to check and validate with the current ubuntu kernel Sep 29 21:04:50 and due lack of time, this wasn't done already Sep 29 21:05:21 and ubuntu is going to final freeze Sep 29 21:05:42 I don't want to push changes that might fix the issues :-) Sep 29 21:05:58 only want to push things that I sure it'd be a solution for the problem Sep 29 21:07:43 and the latest ucm configs were all posted at the no sound support for omap4 bug, just that nobody saw it Sep 29 21:16:12 https://bugs.launchpad.net/ubuntu/+source/alsa-lib/+bug/746023/comments/50 Sep 29 21:16:15 Ubuntu bug 746023 in alsa-utils "No sound on omap4" [High,In progress] Sep 29 21:16:18 when I commented about the issue :-) Sep 29 21:19:50 rsalveti: are the panda config files different from the SDP4430 ones? Sep 29 21:19:52 in your pkg Sep 29 21:20:23 ndec|home: currently, no, but we're using the configs provided by the multimedia wg Sep 29 21:20:42 I guess it's not the same one ubuntu is using Sep 29 21:21:10 i will check tomorrow. sound is working with our config, at least on headset. Sep 29 21:21:33 only issue is that hdmi out doesn't seems to be working anymore Sep 29 21:21:35 just jack Sep 29 21:22:00 but didn't have time to investigate the issue yet Sep 29 21:51:26 just got an mx53 board, it's pretty sweet. i'm installing natty on it Sep 29 21:51:44 anybody know about getting the nfs kernel server going? do the natty kernel images not support that Sep 29 21:54:16 dash: Instead of trying to get natty on it (which we don't support on that platform), why not try oneiric? The latest daily image works, although you may need to create a swap file. Sep 29 21:54:38 The natty images are TI only. Sep 29 21:55:01 GrueMaster: ach! oneiric sounds fine to me, but the wiki said "works like a charm" for mx53 and natty :) Sep 29 21:55:15 i've already got a swapfile anyway, no big deal :) Sep 29 21:56:28 Erm, which wiki? Sep 29 21:56:32 GrueMaster: They ship with natty on an SD. Sep 29 21:56:58 (And a freescale kernel) Sep 29 21:57:09 Ah. Yea, they are making their own image. Ours has the latest kernel and other Ubuntu bits. Sep 29 22:00:54 The board came with lucid. Sep 29 22:01:35 GrueMaster: I was looking at this: https://wiki.ubuntu.com/ARM/DeviceSupport Sep 29 22:01:52 It doesn't say "natty", I imagined that part, I guess. :) Sep 29 22:02:25 dash: Type "cat /etc/lsb-release". Sep 29 22:02:52 GrueMaster: well it says natty /now/ Sep 29 22:03:02 'cause I upgraded Sep 29 22:03:07 but sources.list was all lucid Sep 29 22:03:58 Yea, that's what I thought. It shipped with Lucid, not Natty. At any rate, oneiric has an image for that platform now. Sep 29 22:04:06 awesome. Sep 29 22:43:45 evening dudes Sep 29 22:49:22 in /boot/tools there is a file called update_boot_files.sh, how do i run it ? nothing happens when i chmod +x it Sep 29 22:51:19 /boot/tools? Not one of our directories that I am aware of. Sep 29 22:52:31 I'm praying its the answer to all my problems, i want to update boot.scr Sep 29 22:55:56 I just did a dist-upgrade to the daily from the 28th Sep 30 00:23:48 if i connected to my board with serial, it would say last login from …. would it sound weird to have ttyS2 instead of tty02 ? **** ENDING LOGGING AT Fri Sep 30 02:59:57 2011