**** BEGIN LOGGING AT Tue Aug 28 03:00:03 2012 Aug 28 09:26:13 ogra_: https://launchpad.net/~rsalveti/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=quantal Aug 28 09:26:14 ogra_: for the new pvr-omap4 package, removing the dos2unix dependency and also fixing the packaging for quantal Aug 28 09:26:31 great, thanks Aug 28 09:26:35 infinity: problem is that the builder is now caling debian/rules build-arch instead of debian/rules build Aug 28 09:26:40 that's why it failed Aug 28 09:26:54 (stop doing all these nightshifts, it will kill you some day !) Aug 28 09:27:00 and which you already fixed in other packages, by changing the build to depend on build-arch Aug 28 09:27:30 pbuilder/debuild didn't get the problem because it's still calling with debian/rules build Aug 28 09:27:36 ogra_: :-) Aug 28 09:27:48 ogra_: wanted to finish this today still, as tomorrow is conference day Aug 28 09:27:52 hard to get something done Aug 28 09:28:13 oh, indeed, i forgot about plumbers Aug 28 09:28:25 I'm at utc-7 today Aug 28 09:28:28 take a pillow to the talks though Aug 28 09:28:35 :-) Aug 28 10:32:56 rsalveti: You're not calling pbuilder/debuild with -B, then. Aug 28 11:37:55 * ogra_ wonders why he gets such a bad throughput from cdimage.u.c today Aug 28 11:38:23 zsnc goes rarely over 100kB Aug 28 11:38:33 *zsync Aug 28 11:44:14 ogra_: have you seen bug 1041607 ? =)))) my panda is now stuck on top of microwave oven near the router to get LAN instead of WiFi =) Aug 28 11:44:16 Launchpad bug 1041607 in linux-meta-ti-omap4 "cannot install linux-ti-omap4 3.5.0-209-omap4" [Undecided,New] https://launchpad.net/bugs/1041607 Aug 28 11:46:20 hmm, weird Aug 28 11:47:21 the bug should be against linux-ti-omap4 btw, not meta ... Aug 28 11:48:26 ogra_: i filed it against omap4, somebody else moved it to meta.... Aug 28 11:48:39 and it looks like an issue with update-initramfs, does it work if you do: sudo aupdate-initramfs -c -k 3.5.0-209-omap4; sudo dpkg --configure -a Aug 28 11:49:12 *update-initramfs indeed :P Aug 28 11:50:38 looks like that is teh issue: update-initramfs: deferring update (hook will be called later) Aug 28 11:50:57 since it directly tries to copy around the initrd.img right afterwards ... Aug 28 11:52:41 let me try Aug 28 11:52:54 * xnox ssh panda.local Aug 28 11:53:18 i think the kernel package postinst scripts actually should force the update-initramfs run instead of using a trigger at that point Aug 28 11:53:47 (or alternatively the copy code shoudl be omitted ...) Aug 28 12:02:14 ogra_: XFCE became alternative of Unity 2D? Aug 28 12:03:14 there is no "alternative" per se Aug 28 12:03:29 for ac100 is swithced the images to lubuntu Aug 28 12:03:47 its up to the people maintaining the images what they use i guess Aug 28 12:10:27 I'm finding lxde glitchy on precise with the tf201 Aug 28 12:11:04 ogra_: was the WiFi missing on pandas bug fixed? Aug 28 12:11:10 ogra_: what was the bug #? Aug 28 12:11:40 Bug 1032779 ? Aug 28 12:11:42 Launchpad bug 1032779 in linux-firmware "Missing firmware for wireless chipset on pandaboard" [Medium,Fix released] https://launchpad.net/bugs/1032779 Aug 28 12:18:51 ogra_: ok. so 3.5 supposedly doesn't use that firmware, but I still don't have wifi with 3.5 kernel this time. Which package should i file the bug against? firmware? linux-ti-omap4? Aug 28 12:19:07 getting same thing in dmseg Aug 28 12:19:08 [ 23.202850] wl12xx: ERROR could not get nvs file ti-connectivity/wl1271-nvs.bin: -2 Aug 28 12:19:24 ogra_: do you have wifi on your panda in quantal? Aug 28 12:19:43 no idea, havent tried it :) Aug 28 12:19:52 * ogra_ has a dedicated switch for all his arm boards Aug 28 12:20:25 i'll try once i'm done with PVR and can bear to trash the panda while playing with it :) Aug 28 12:23:13 ogra_: can I bribe you with bamboo? =) Aug 28 12:23:28 lol, i'm not a panda ! Aug 28 12:23:48 ogra_: but you have so many of them =)))) Aug 28 12:26:02 http://plus.url.google.com/url?sa=z&n=1346156762908&url=http%3A%2F%2Fbit.ly%2FRUbLqy&usg=OLZd5hl2msGAN9KAoBgn0GpCCcM. Aug 28 12:26:34 or better https://lh4.googleusercontent.com/-XsraXa6V0VY/UDya87ScR3I/AAAAAAAADMc/dQsqSah_KzM/w497-h373/ae0016-4.jpg Aug 28 12:26:36 :) Aug 28 12:28:35 ogra_: from the intermediate photos, I would have never guessed the final result will be panda. I'd guess something else... Aug 28 12:29:54 haha, yeah ... pretty "else" ... Aug 28 12:47:09 * ogra_ twiddles thumbs waiting for the archive mail for the upload ... Aug 28 12:47:20 ... and considers to tra an ac100 isntall test Aug 28 12:47:31 *try Aug 28 12:48:40 i wonder what /lib/udev/mtp-probe is ... Aug 28 12:48:50 always generates warnings on boot Aug 28 12:49:00 ogra_: mpt probe? Aug 28 12:49:09 yep Aug 28 12:49:26 ac100 doesnt use a splash atm... so i see these boot messages Aug 28 12:49:39 and udev seems to be unhappy with that script Aug 28 12:53:58 https://launchpad.net/ubuntu/+source/pvr-omap4/1.9.0.5.1.1-0ubuntu3/+build/3748952 Aug 28 12:54:02 ah, that looks better Aug 28 12:54:17 (1.6M vs 47 bytes) Aug 28 12:54:32 * ogra_ waits for the publisher Aug 28 13:45:24 rsalveti, pvr loads but i still see some errors: Aug 28 13:45:29 PVR:(Warning): PVRDRMSetFD: could not get plugin ioctl base: -22 [69, /pvr_bridge_u.c] Aug 28 13:45:42 (EE) Couldn't get PVR Services status Aug 28 13:45:56 PVR:(Error): OpenServices: drmOpen failed [120, /pvr_bridge_u.c] Aug 28 13:46:10 (EE) OMAP(0): ERROR: PVR2DCreateDeviceContext failed Aug 28 13:47:59 and es2_info doesnt work :( Aug 28 13:49:24 rsalveti, http://paste.ubuntu.com/1171824/ Aug 28 14:21:04 rsalveti, so the issue seems to be that the header cant be found anywhere linux/omap_drv.h ... and it seems your fix to linux didnt work, Aug 28 14:31:13 rsalveti: btw, what exactly did you have to change in live-build to use it? why can't we just use the default live-build? and when are you planning to merge your changes upstream? Aug 28 14:40:10 * ogra_ sighs Aug 28 14:40:42 so after shuffling headers around for 1h it now fails on omap_priv.h Aug 28 14:42:03 rsalveti, your headers fixes went into the wrong package :( ... (linux vs linux-ti-omap4) Aug 28 14:43:04 and omap_priv.h seems to be nowhere on my filesystem Aug 28 14:50:25 it's private ... ;-) Aug 28 14:51:28 heh, well, looks like it should be in include/drm/omap_priv.h Aug 28 14:51:32 but it isnt indeed Aug 28 14:52:38 so that header situation is completely messed up :/ Aug 28 14:52:57 maybe XavB knows about that... Aug 28 14:53:03 *sadpanda* Aug 28 14:55:19 ogra_: ndec: we do have a "hack" to add omap_drv.h and omap_drm.h to headers package but I think that Ricardo already picked that one Aug 28 14:55:44 XavB, he forgot omap:priv.h Aug 28 14:55:56 s/\:/_/ Aug 28 14:57:06 me copy pastes omap_priv.h and omap_drm.h from robclark's lwn post :P Aug 28 14:57:59 Error! Bad return status for module build on kernel: 3.5.0-209-omap4 (armhf) Aug 28 14:57:59 Consult /var/lib/dkms/pvr-omap4/1.9.0.5.1.1/build/make.log for more information. Aug 28 14:58:01 *sniff* Aug 28 15:00:12 ogra_: can you pastebin your log please? Aug 28 15:01:05 if my panda ever comes back up again after the reboot i just did ... Aug 28 15:01:39 ogra_: are you sure about omap_priv.h? I only have omap_dmm_priv.h... Aug 28 15:02:05 http://paste.ubuntu.com/1171936/ Aug 28 15:03:50 (sorry for the german :) ) Aug 28 15:07:26 ogra_: in fact we do not have "include omap_priv.h" anymore on our kernel. Aug 28 15:07:27 http://dev.omapzoom.org/?p=integration/kernel-ubuntu.git;a=commit;h=0428bbdcdb53255b8b42a8ed73ab6ec1d8975054 Aug 28 15:07:40 great Aug 28 15:08:16 so i guess the omap_drv.h i have is broken Aug 28 15:20:13 ubuntu@panda:~$ grep omap_drm /usr/src/linux-headers-3.5.0-209/include/linux/omap_drv.h|grep include Aug 28 15:20:14 #include "omap_drm.h" Aug 28 15:20:20 yeah. looks like Aug 28 15:20:32 now i wonder how rsalveti managed to get that workin Aug 28 15:20:33 g Aug 28 15:21:15 * ogra_ needs a break Aug 28 15:34:48 ndec: afaik the live-build we're using is just a rev from upstream Aug 28 15:34:54 just that we tested it and it worked fine Aug 28 15:35:23 infinity: but why the previous builders (pecise) were also calling build first? Aug 28 15:35:45 this affected as well the ati and nvidia drivers I believe Aug 28 15:35:59 at least I saw you fixed the ati one Aug 28 15:42:33 ogra_: the warning is ok, but the second error it seems to be related with the lack of module Aug 28 15:43:09 rsalveti, well, yeah, because it doesnt build, since the header file ended up in linux, not linux-ti-omap4 Aug 28 15:43:25 ogra_: it is at linux-ti-omap4! Aug 28 15:43:29 that's where I sent my first patch Aug 28 15:43:30 and the header we have seems to be outdated and missing http://dev.omapzoom.org/?p=integration/kernel-ubuntu.git;a=commit;h=0428bbdcdb53255b8b42a8ed73ab6ec1d8975054 Aug 28 15:43:38 rsalveti, i dont have it Aug 28 15:44:11 weird, let me check that Aug 28 15:44:18 it is the in -209 but not in 209-omap4 Aug 28 15:44:26 *it is in Aug 28 15:47:42 ubuntu@panda:~$ find /usr/src/ -name omap_drv* Aug 28 15:47:42 /usr/src/linux-headers-3.5.0-209/include/linux/omap_drv.h Aug 28 15:47:42 /usr/src/linux-headers-3.5.0-209/drivers/staging/omapdrm/omap_drv.h Aug 28 15:47:42 /usr/src/linux-headers-3.5.0-12/drivers/staging/omapdrm/omap_drv.h Aug 28 15:48:07 dkms tries to use /usr/src/linux-headers-3.5.0-209-omap4 Aug 28 15:48:30 rsalveti, ^^^ Aug 28 15:48:56 let me check locally Aug 28 15:51:40 ogra_: you only need omap_drm.h and omap_drv.h, don't need the private one Aug 28 15:52:07 rsalveti, the omap_drv.h i have here has an include omap_priv.h Aug 28 15:52:18 i copied it from -209 to -209-omap4 Aug 28 15:52:37 which makes dkms work but then fail on this include Aug 28 15:53:34 * ogra_ pulls both headers out of the above commit and retries Aug 28 15:55:12 Error! Bad return status for module build on kernel: 3.5.0-209-omap4 (armhf) Aug 28 15:55:14 GRRR ! Aug 28 15:56:42 /var/lib/dkms/pvr-omap4/1.9.0.5.1.1/build/services4/srvkm/env/linux/private_data.h:95:28: fatal error: linux/omap_drv.h: Datei oder Verzeichnis nicht gefunden Aug 28 15:56:46 * ogra_ curses Aug 28 15:59:14 #if defined(SUPPORT_DRI_DRM_EXTERNAL) Aug 28 15:59:14 #include Aug 28 15:59:17 hmm Aug 28 15:59:46 that the only trace i find in private_data.h Aug 28 16:06:00 http://paste.ubuntu.com/1172053/ Aug 28 16:06:04 the full log again Aug 28 16:06:37 * ogra_ has to prepare for his ubuntu developer week talk about arm now Aug 28 16:08:18 rsalveti, i clearly see an "#include "omap_priv.h"" in my omap_drv.h Aug 28 16:08:30 ogra_: I'm checking that now Aug 28 16:08:35 not sure why you dont Aug 28 16:08:44 ok Aug 28 16:09:13 rsalveti, i'll leave that to you then, rtg asked for a bu with a fix attached if we need kernel changes Aug 28 16:09:18 *bug Aug 28 16:09:27 ogra_: sure Aug 28 16:09:43 * ogra_ wonders what to talk about in UDW **** ENDING LOGGING AT Tue Aug 28 16:11:53 2012 **** BEGIN LOGGING AT Tue Aug 28 16:12:23 2012 Aug 28 16:18:42 ogra_: root@ubuntu-desktop:/usr/src/linux-headers-3.5.0-209-omap4/include/linux# ln -s ../../../linux-headers-3.5.0-209/include/linux/omap_drm.h omap_drm.h Aug 28 16:18:42 root@ubuntu-desktop:/usr/src/linux-headers-3.5.0-209-omap4/include/linux# ln -s ../../../linux-headers-3.5.0-209/include/linux/omap_drv.h omap_drv.h Aug 28 16:18:46 to get your module to build Aug 28 16:18:57 I'm checking now why this header is not linked at the original package Aug 28 16:19:15 rsalveti, right, thats how i got to the omap_priv.h error :) Aug 28 16:19:29 ogra_: just did that, and installed pvr-omap4 Aug 28 16:19:45 for me that fails with the above error Aug 28 16:20:11 and obviously there is an #include "omap_priv.h" in my omap_drv.h Aug 28 16:20:48 I don't have any here Aug 28 16:20:56 are you sure you just linked the files? Aug 28 16:21:02 or you downloaded it from another source? Aug 28 16:21:09 from LP Aug 28 16:21:59 i downloaded teh linux-image package for -209-omap4, dpkg -i'ed it and its headers, then rebooted and apt-get installed pvr-omap4 Aug 28 16:22:21 to fail with the missing omap_drv.h ... Aug 28 16:22:39 then i linked the files as above and got to the point where it started failing on _priv.h Aug 28 16:23:17 *now* i have the files replaced by the ones from the omapzoom commit for a test, but that doesnt help either Aug 28 16:23:41 infinity, hello, the armadaxp bitss look clean to me, Marvell also again said they believe it sould be all GPL2 (one of their 3 options in HAL code) Aug 28 16:23:56 so could the package be moved from PPA to archives ? Aug 28 16:24:12 janimo, they said they belive that last time too ... and NCommander uploaded a horridly illegal package Aug 28 16:24:27 (because he belived them) Aug 28 16:25:07 ogra_, do you recall what type of bits were illegal there? Aug 28 16:25:28 I know the original patch had extra scripts in net and elsewhere where I would not have though there should be something arch specific Aug 28 16:25:41 janimo, not from the top of my head, and the discussion about it wasnt in a publically logges channel iirc Aug 28 16:25:54 *logged Aug 28 16:26:32 well, anyway to me it looks like any other kernel code :) Aug 28 16:26:46 no blobs or other strange bits Aug 28 16:27:07 well, infinity might surely know more but it will be hard to catch him this week Aug 28 16:27:14 apart from some C HAL code that is wrapped in linux specifics but that too has an ok license Aug 28 16:27:15 (plumbers) Aug 28 16:27:23 ok Aug 28 16:29:03 ogra_: http://paste.ubuntu.com/1172106/ Aug 28 16:29:54 I'm now checking why the link is not done properly at the flavour specific headers package Aug 28 16:30:06 rsalveti, what io also dont get is, how do your headers end up in include/linux/ ... mine are definitely under drivers/staging/omapdrm originally Aug 28 16:30:31 ogra_: this is the original path, but we have links at include/linux to drivers/staging Aug 28 16:31:18 ubuntu@panda:~$ ls -l /usr/src/linux-headers-3.5.0-209/include/linux/omap_drv.h Aug 28 16:31:19 lrwxrwxrwx 1 root root 40 Aug 22 18:37 /usr/src/linux-headers-3.5.0-209/include/linux/omap_drv.h -> ../../drivers/staging/omapdrm/omap_drv.h Aug 28 16:31:25 right, i missed that one, sorry Aug 28 16:32:50 ogra_: ok, will investigate here what is happening with the flavor specific headers package Aug 28 16:33:59 thanks a lot ! Aug 28 16:34:36 * ogra_ really doesnt get how you have teh patience to dig through all this every day ... Aug 28 16:34:51 after fiddling with it for even half a day i got enough :) Aug 28 16:43:19 ah, I get the same usb ethernet oom with my mcs7830 on ac100 as on omap Aug 28 16:43:35 3.6-rc2 kernel Aug 28 16:44:42 great Aug 28 16:50:47 yeah, everything works as expected ;-) Aug 28 16:51:54 now use u-boot to load the kernel Aug 28 16:52:35 * ogra_ still fails to see the advantage of usin u-boot Aug 28 16:55:42 hi, using gcc 4.7.2 I try to build for a target system based arch = armv5te. does anyone have a clue why It can't compile via gcc 4.7 ? and I get |error: bad value (armv5te) for -march= switch Aug 28 16:56:27 ogra_: he, too soon for apri jokes Aug 28 16:58:19 we can use initrd's > 2 MB Aug 28 16:58:34 offt, who wants that :) Aug 28 16:58:41 *pfft even Aug 28 17:16:11 yeah, boots up - but no backlight yet :-( Aug 28 18:14:32 i would like to try porting ubuntu on arm where do i start? Aug 28 18:14:52 well, what HW do you have ? Aug 28 18:15:03 (read: where do you want to port it to) Aug 28 18:19:56 janimo: It wasn't blobs last time, it was a mess of drivers that just plain weren't free. Aug 28 18:20:30 * ogra_ is out now ... dinner etc Aug 28 18:20:33 janimo: Diff between mainline and the Marvell sources, then rgrep 'opyright', and check headers for telltale signs like "all rights reserved" and such. Aug 28 18:21:33 @ogra_ http://www.gsmarena.com/micromax_funbook_p300-4701.php Aug 28 18:22:36 rsalveti: The behaviour of dpkg-buildpackage -B changed in Quantal to call build-arch if it exists. Aug 28 18:23:04 rsalveti: Hence why this will work in precise, it will work in quantal with -b, but will fail in quantal with -B. Aug 28 18:24:05 janimo: Anyhow, I'm still hopping around between airports, I can try to have a look later, but I really don't want to be the one to do a full audit again. Aug 28 18:24:21 infinity, thanks, I understand. Aug 28 18:24:23 janimo: If I do a partial audit (and I will, if you ask me to copy it), and find non-free bits, I'll be annoyed. :P Aug 28 18:24:39 one copyright header I see is the one mentioned above, with a slightly verbose boilerplate Aug 28 18:24:45 but looks ok nonetheless Aug 28 18:24:56 @ogra_ can i work with that hardware Aug 28 18:24:57 infinity, I do not want to annoy you good sir Aug 28 18:25:12 because you know my IP address Aug 28 18:25:33 I'll sneak into your house in the middle of the night and force-feed you meat. Aug 28 18:25:41 That's a vegan's worst nightmare, right? Aug 28 18:25:59 infinity, I do have disturbing dreams of eating meat indeed Aug 28 18:26:11 not nightmares but I feel disappointed while dreaming Aug 28 18:26:17 so yes, you're pretty close Aug 28 18:26:19 Hah. Aug 28 18:26:42 raki1, with regard to bootloader and kernel this is hard to tell, there inst much info about the CPU "1.22 GHz Cortex-A8" does not say much ... but from the specs you should definietly be able to run ubuntu userspace on it though Aug 28 18:26:56 Holy crap, we might finally be boarding my 1.5h-late flight. Aug 28 18:27:09 \o/ Aug 28 18:27:15 which airport are you now in? Aug 28 18:27:18 huh ? Aug 28 18:27:22 you arent there yet ? Aug 28 18:27:38 maybe there's a conference internal airport for lazy people Aug 28 18:27:48 LOL Aug 28 18:27:51 to travel between wings of the same building Aug 28 18:28:00 wings eh ? Aug 28 18:28:13 oh, you found an unintended pun Aug 28 18:28:23 :) Aug 28 18:28:23 I do not always pun but when I do they are really unintended Aug 28 18:28:29 anyway, really dinner now Aug 28 18:28:36 * ogra_ -> out Aug 28 18:49:55 infinity: got it, is this option only used for the archs !i386 at launchpad? Aug 28 18:50:10 afaik i386 is the one responsible for the arch indep packages Aug 28 20:22:16 ogra_: janimo what horribly illegal package? Aug 28 20:22:34 NCommander, just assumptions. But armadaxp kernel Aug 28 20:22:57 janimo: oh, probably because I forgot to strip out the firmware >.<; Aug 28 20:22:58 infinity, says there were illegal bits in previous drops Aug 28 20:23:02 Oh Aug 28 20:23:06 No, I stripped those all out Aug 28 20:23:18 * NCommander went through file by file to make sure there wasn't anything illegal Aug 28 20:23:48 there's a 3.5 drop now and infinity says we need to make sure it's not like that time, when it was also said to be all clean Aug 28 20:24:11 so the firmware is now in linux-firmware or was not even needed for the armadaxp? Aug 28 20:47:05 janimo: the later I believe. I only did the initial 3.0 and 3.2 drops from Marvella nd those were both clean and were crosschecked to make sure the illegal bits were stripped oput Aug 28 20:49:31 iirc, the firmware is for Marvel Wifi (which is not on the board). Their kernel blob has ALL of their development work in it for every product they are working on. Makes stripping and cleaning...fun. Aug 28 20:50:19 From the 3.2 kernel, I seem to remember keeping ~10% of their mods. Aug 28 20:50:28 Maybe a little more. Aug 28 21:00:47 rsalveti: Right, i386 builds with -b, everything else builds with -B Aug 28 21:21:12 infinity: yeah, that explains better, thanks **** ENDING LOGGING AT Wed Aug 29 02:59:58 2012