**** BEGIN LOGGING AT Tue Oct 18 02:59:57 2011 Oct 18 05:42:31 pespin: yay :-D Oct 18 05:54:29 JaMa|Off: gah... I really have a hard time to find recipes these days :/ Oct 18 06:07:37 I stumbled upon "ERROR: '/OE/shr-core/meta-openembedded/meta-oe/recipes-connectivity/openssl/openssl_1.0.0e.bb' failed" while trying to build shr-core with "bitbake -k shr-image", any tips on how to fix it? Oct 18 06:33:10 taruti: how does it fail precisely? Oct 18 06:34:08 mrmoku: http://pastebin.com/1ErP3fYP Oct 18 06:35:46 taruti: hmm... there is missing something before? Oct 18 06:36:44 mrmoku: there are many screenfulls of the NOTE lines Oct 18 06:38:44 taruti: look for those talking about openssl Oct 18 06:39:27 taruti: to see for what task it is failing Oct 18 06:41:35 hmm, I'll try to decrease parallelism and get more error information, too short buffer in the terminal Oct 18 06:43:46 taruti: or don't run with -k Oct 18 06:43:54 it should stop then after failure Oct 18 06:46:31 mrmoku: http://pastebin.com/raw.php?i=nm9AfjdX now this has the error Oct 18 06:49:49 taruti: the real error is still a bit before :) Oct 18 06:50:14 taruti: do you use the shr chroot to build? Oct 18 06:50:26 yes Oct 18 06:50:55 there are some nice alias for building images Oct 18 06:51:02 bbl = shr light Oct 18 06:51:05 bbf = shr fat Oct 18 06:51:07 bbi = both Oct 18 06:51:18 the nice thing is it will store the log into a file Oct 18 06:51:21 log.om-gta02 Oct 18 06:51:55 errr Oct 18 06:51:59 bbf is *not* the fat image Oct 18 06:52:19 it's the feed Oct 18 06:53:14 leviathan: what modules are essential to be loaded on start? Oct 18 06:54:45 mrmoku: yet again http://pastebin.com/tsJ3XiRN Oct 18 06:56:56 Cannot create directory /OE/shr-core/tmp/work/armv4t-oe-linux-gnueabi/openssl-1.0.0e-r13.3/image/usr: File exists Oct 18 06:56:59 is strange Oct 18 06:57:21 taruti: try 'bb -c cleansstate openssl; bb openssl' Oct 18 06:57:52 ok Oct 18 07:12:31 mrmoku: like which one? Oct 18 07:12:49 taruti: there should be fix for openssl already Oct 18 07:13:12 hmm. this worked. Oct 18 07:13:31 http://patches.openembedded.org/patch/13369/ Oct 18 07:18:17 JaMa|Off: in general Oct 18 07:18:36 JaMa|Off: like X being spread over three layers :P Oct 18 07:18:53 JaMa|Off: btw. do you have xserver-common patches in git? Oct 18 07:21:40 mrmoku: g_ether Oct 18 07:21:44 :-) Oct 18 07:22:33 mrmoku: only moving them to oe-core.. but they needs to be reworked somehow better Oct 18 07:22:48 mrmoku: to split those machine related changes to BSP etc.. Oct 18 07:24:18 leviathan: ok, what about hso? Oct 18 07:24:33 JaMa|Off: I want to add gta04 there Oct 18 07:26:13 leviathan: btw. hanging was caused by suspend Oct 18 07:26:24 I guess there is just nothing configured to resume it Oct 18 07:29:45 mrmoku: for now you can.. I'll migrate it later Oct 18 07:29:57 JaMa|Off: ok Oct 18 08:01:16 leviathan: sound is still missing, right? Oct 18 08:48:05 good morning Oct 18 12:36:04 SHR: 03Martin.Jansa 07shr-chroot * recfb2cd6c761 10/ (240 files in 28 dirs): system upgrade Oct 18 12:36:44 * JaMa needs more swap or ram on SHR buildhost to build webkit-efl/webkit-gtk :/ Oct 18 12:37:28 JaMa: time to send morphis ssh keys? Oct 18 12:54:58 I did already to mickey iirc Oct 18 13:07:28 [ 226.810164] mmc2: new SDIO card at address 0001 Oct 18 13:07:34 hi btw Oct 18 13:09:46 wlan0 no wireless extensions. Oct 18 13:09:49 lol Oct 18 13:09:55 that's because I lack WEXT Oct 18 13:10:07 I just added cfg80211's wext emulation Oct 18 13:15:30 else since there is still WIRELESS_EXT we could port the same bcm4329 that dcordes uses Oct 18 13:15:35 maybe it's more complete Oct 18 13:17:04 mrmoku: http://paste.pocoo.org/show/494343/ Oct 18 13:22:19 root@crespo:~# ifconfig wlan0 up Oct 18 13:22:19 ifconfig: SIOCSIFFLAGS: Operation not permitted Oct 18 13:22:22 hmmm Oct 18 13:22:28 and rfkill is at 1 Oct 18 13:22:33 so unblocked Oct 18 13:39:14 I just pushed an activator and I'll go Oct 18 13:39:24 I think the activator works but something else doesn't work.... Oct 18 13:39:28 maybe a firmware issue? Oct 18 13:39:51 I must find the new firmware anyway since kenrel.org is beeing restored and that there is not yet the firmware tree there Oct 18 13:39:55 I'll go bye Oct 18 13:49:44 JaMa|Off: ouch Oct 18 13:56:06 mrmoku: I'll try with -b, because bitbake takes quite a lot of ram lately too and with -b it's much less Oct 18 13:59:13 ok Oct 18 14:02:00 hmm... we're doomed to use the new kernel anyway on gta04 Oct 18 14:02:17 as the old working one does not build with our toolchain Oct 18 14:21:06 * JaMa|Off found jackpot :) shr-chroot/OE/shr-core/tmp/old.. but removing it from SHR buildhost seems to take few hours under load > 6, weird Oct 18 14:21:42 not even half of it.. 0.0 0.1 0:58.18 | | `- rm -rf ../old/ armv4t-oe-linux-gnueabi/ Oct 18 14:22:52 hehe :) Oct 18 15:23:00 2:10.67 and still removing :) Oct 18 15:48:42 damn... u-boot does not follow symlinks when loading a file :/ Oct 18 15:49:23 anyway... school reunion now Oct 18 15:49:24 bbl Oct 18 15:49:54 enjoy Oct 18 16:09:40 Good Afternoon people Oct 18 16:15:54 pespin: please fix your git config to include right name to your patches Oct 18 16:47:59 ok sorry Oct 18 16:48:02 bbl Oct 18 17:58:09 mrmoku, xrandr -o 1 's input doesn't rotate on om-gta02, it did before, JaMa|Off is the person to report to right? Oct 18 17:58:12 hi btw Oct 18 18:00:17 GNUtoo: hi, please check oe-devel thread "[meta-oe] X randr-support.patch breaks geode" Oct 18 18:00:23 ok Oct 18 18:00:32 that thread is old.... Oct 18 18:01:05 yes, but that's the reason why this patch is not in oe-core (at least yet) Oct 18 18:02:04 I can add it for our devices in BSP for now.. but in long run we need to fix it properly, that it will work for both Oct 18 18:04:51 GNUtoo: JaMa|Off always can point to some mail in some mailinglist ... no matter what error you throw at him :-) Oct 18 18:05:04 no idea how he does that Oct 18 18:05:33 ok Oct 18 18:05:39 lol Oct 18 18:05:44 radekp: ping Oct 18 18:06:06 mrmoku: pong :) Oct 18 18:06:17 radekp: what are your gta04 kernel plans? :-) Oct 18 18:06:39 mrmoku: hmm i dont know yet... Oct 18 18:07:00 we're doomed to use the master branch Oct 18 18:07:10 because our toolchain does not buiild the old one Oct 18 18:07:21 mrmoku: maybe i will start with 2.6.32 then maybe move to 3.1 Oct 18 18:07:34 ok Oct 18 18:08:00 I have no idea what I should do... Oct 18 18:08:11 fix 2.6.32 to build with the new toolchain Oct 18 18:08:15 your toolchain does not build 2.6.32? Oct 18 18:08:19 or help leviathan to make 3.1 work Oct 18 18:08:33 radekp: yeah... some arm thumb mode issue Oct 18 18:08:49 it got fixed in the kernel at some point in time Oct 18 18:09:06 mrmoku: and if you force ARM mode for that recipe? Oct 18 18:09:20 maybe disabling thumb in kernel could help? Oct 18 18:09:27 http://comments.gmane.org/gmane.comp.handhelds.openembedded/40650 Oct 18 18:09:36 maybe... but I don't want that Oct 18 18:10:20 JaMa|Off: to build without thumb? Oct 18 18:10:43 | /tmp/cc8Q5hju.s: Assembler messages: Oct 18 18:10:44 | /tmp/cc8Q5hju.s:299: Error: selected processor does not support ARM mode `smc #0' Oct 18 18:10:46 | /tmp/cc8Q5hju.s:337: Error: selected processor does not support ARM mode `smc #0' Oct 18 18:10:49 is the error Oct 18 18:10:59 smc is some security extension I think Oct 18 18:11:07 which the OMAP3 does not have Oct 18 18:11:36 i dont know in what state is 3.1, of course it would be better then 2.6.32 Oct 18 18:11:37 mrmoku, 2.6.32 ouch that's as old as the htcdream kernel Oct 18 18:11:56 radekp: it boots... but is still missing lots of stuff Oct 18 18:12:15 ok Oct 18 18:12:17 * mrmoku wonders if he should learn something about omap kernels Oct 18 18:12:23 and help leviathan Oct 18 18:12:40 maybe I'm the one that should do that stuff tough Oct 18 18:12:45 but I don't have a gta04 Oct 18 18:14:11 I really hope there will be another batch still this year Oct 18 18:14:22 ok Oct 18 18:15:21 leviathan: ping Oct 18 18:22:40 JaMa|Off: you mean setting ARM_INSTARUCTION_SET = "arm" in the kernel recipe would help? Oct 18 18:32:00 mrmoku: worth trying Oct 18 18:33:08 mrmoku: or you should be able to find patch for that (I was using backported patch on spitz for a while) Oct 18 18:44:10 heyho Oct 18 18:46:49 evening Oct 18 18:46:57 still fighting with libeflvala build Oct 18 18:47:23 JaMa|Off: yeah, will try to find a patch Oct 18 18:47:33 I managed to narrow it down to this: for some reason on my box, ecore_evas does not make it to libtool invocation when building eflvala: Oct 18 18:47:36 http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/make-eflvala.log Oct 18 18:47:53 while for others, like pespin, with whome I discussed it yesterday, it does Oct 18 18:48:29 I wonder how... looking at elementary.pc, it lists there only in Requires.private, which, pkg-config doc says, is used only for static linking Oct 18 18:49:07 (libeflvala.so builds, but then examples fail with: undefined reference to symbol 'ecore_evas_init') Oct 18 18:49:12 morphis, hi I'm progressing with wifi Oct 18 18:49:18 now it loads etc...but It says: Oct 18 18:49:24 GNUtoo: I already saw your commits Oct 18 18:49:26 [ 2031.994409] mmc2:0001:2: pending IRQ with no handler Oct 18 18:49:34 yes but I just got furthurer Oct 18 18:49:47 I didn't push the firmware before(to the device) Oct 18 18:49:52 so it failed Oct 18 18:50:25 ok Oct 18 18:50:30 but now it works? Oct 18 18:50:38 see the above message Oct 18 18:50:43 it's an error not a warning Oct 18 18:50:47 ah I thought you solved it Oct 18 18:50:57 I'll try to solve it now Oct 18 18:51:04 ok Oct 18 18:51:13 btw. can we include the firmware in our images? Oct 18 18:51:16 I solved the return -ENOPERM by adding the firmware Oct 18 18:51:18 I think so Oct 18 18:51:30 great Oct 18 18:51:36 but let's see that later Oct 18 18:52:17 ok Oct 18 18:52:34 let's see how to pass that bcm4329_wlan_irq Oct 18 18:52:35 I didn't read license agreement for the firmware from google so I am asking :) Oct 18 18:52:47 it's from broadcom not google Oct 18 18:52:55 linux-firmware tree Oct 18 18:53:12 redistributing google's firmware is not allowed Oct 18 18:53:39 ok Oct 18 18:53:57 so we can't redistribute the firmware from their nexuss page? Oct 18 18:54:02 no Oct 18 18:54:06 I don't think so Oct 18 18:54:16 anyway it's useless with that new driver Oct 18 18:55:32 yes Oct 18 18:55:36 I just asking Oct 18 18:55:55 but what does cyanogen? do they cut the firmware files from the original rootfs? Oct 18 18:56:29 they do illegal stuff Oct 18 18:56:40 they have a illegal redistribution method Oct 18 18:56:46 rapideshare or something like that Oct 18 18:57:50 ieee Oct 18 18:58:46 anybody with a working, by-hand-built libeflvala pls? Oct 18 19:00:43 JaMa|Off: what about shr-core and a toolchain? Oct 18 19:02:23 zub: where is the error in your paste? Oct 18 19:03:24 GNUtoo: I hate this kind of work, put something somewhere nobody finds it Oct 18 19:04:06 ? Oct 18 19:04:35 no I meant official cyanogenmod images are redistributed trough sharing websites Oct 18 19:05:04 mrmoku: the library builds ok, so it's not in *that* log... it explodes when linking the examples, see here: http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/make-all.log Oct 18 19:06:41 zub: /usr/bin/ld: note: 'ecore_evas_init' is defined in DSO /home/zub/efl-trunk/lib/libecore_evas.so.1 so try adding it to the linker command line Oct 18 19:06:45 this one, right Oct 18 19:06:57 (the thing is my libeflvala.so does not include libecore_evas DT_NEEDED entry... then what links to libelementary explodes, unless it also explicitly links to libecore_evas) Oct 18 19:06:58 while it seems that your (as for: you for who it works) libeflvala.so does include the dependency on libecore_evas Oct 18 19:07:05 that helps Oct 18 19:07:09 of course Oct 18 19:07:18 adding ecore_Evas into PKG_CHECK_MODULES helps Oct 18 19:07:20 GNUtoo: hm, but they offer official images for downloading on their side Oct 18 19:07:26 I have one on my phone Oct 18 19:07:30 no Oct 18 19:07:38 that why I back then produiced a patch for this, thinking it's some EFL change that requires this, only to find. yest. that it just works for you w/o Oct 18 19:07:45 they point to them Oct 18 19:07:48 but that's all Oct 18 19:08:02 ohha Oct 18 19:08:06 but I'm a a loss why it works for you and explodes for me :( Oct 18 19:08:21 zub: that might depend on your distro if the toolchain requires you to explicitely give all libraries Oct 18 19:08:25 ah right Oct 18 19:08:30 it might depend on the model Oct 18 19:08:52 btw. suspend is now working as I expect it to work Oct 18 19:08:54 ah no it has changed Oct 18 19:09:04 mrmoku: yes, but lookin at pespin's log (http://paste.pocoo.org/show/494141/ vs. mine http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/make-eflvala.log) I see that for him ecore_evas *somehow* makes it in, I don't know how :-( Oct 18 19:09:11 before you were going trough rapideshare or something like taht Oct 18 19:09:20 pkg-config doesn't list the dependency unless oyu use --static... and static is not used Oct 18 19:09:35 hmm... pkg-config for ecore? Oct 18 19:09:43 for elementary Oct 18 19:10:27 http://www.fpaste.org/97wj/ Oct 18 19:10:55 so I wonder why/how does lecore_evas get to the build for him (and possibly for you all for whome it just works) Oct 18 19:11:03 zub: I have the same output Oct 18 19:11:21 I think pespin also had, IIRC... yet for him -lecore_evas somehow makes it to the build Oct 18 19:11:25 * zub is going bananas Oct 18 19:12:10 i.e. I mean /usr/local/lib/libecore_evas.so in http://paste.pocoo.org/show/494141/ (the libtool invocation) Oct 18 19:13:57 pespin: btw. cool thing :-D (tp) Oct 18 19:14:19 if I do e.g. http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0001-ecore-split-into-ecore-and-ecore-evas-to-match-EFL-c.patch - so I have ecore and ecore-evas separately in libeflvala, it works for me, OTOH if it's not needed then... Oct 18 19:14:29 then why would anyone care for integrating it :) Oct 18 19:14:37 pespin: I would really like to start that fso tp backend thing... just the bindings are hindering me Oct 18 19:15:20 zub: hmm... no idea :/ Oct 18 19:15:25 na ja Oct 18 19:15:31 thanks anyway for having a look Oct 18 19:15:35 np Oct 18 19:15:49 separately I'd like to push for this patch :-P http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0002-elementary-rename-elm-to-elementary-to-match-EFL-cha.patch Oct 18 19:16:17 thing is: --pkg elm means: valac looks for elm.vapi (and deps), but also asks pkg-config for elm.pc Oct 18 19:16:27 but elementary renamed it to elementary.pc Oct 18 19:16:39 -> you get potentially wrong cflags and linker flags Oct 18 19:17:07 it works for you because usually elementary headers path is already in include path, and for linking vala is not used and pkg-config is used explicitely Oct 18 19:17:24 but I beliueve that this wrong naming would break valac --pkg elm test.vala Oct 18 19:17:29 (i.e. compile + link by valac) Oct 18 19:17:36 ic Oct 18 19:18:10 (when valac doesn't get any info from pkgconfig (i.e. no .pc file), valac doesn't add any extra settings from that...) Oct 18 19:18:31 so I think it works by accident Oct 18 19:18:35 playya_: wasn't it you taking care of eflvala? Oct 18 19:19:24 morphis, or maybe it was for the radios, I don't remember well sorry Oct 18 19:21:06 GNUtoo: no problem Oct 18 19:22:18 anybody knows how to build a toolchain for shr-core? Oct 18 19:24:07 freesmartphone.org: 03morphis 07cornucopia * re76cd2cb9548 10/fsogsmd/src/plugins/modem_samsung/ (Makefile.am channel.vala mediators_sim.vala unsolicited.vala): Oct 18 19:24:07 freesmartphone.org: fsogsmd: modem_samsung: huge steps towards working sim card authentication Oct 18 19:24:07 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:24:08 freesmartphone.org: 03morphis 07cornucopia * r5ff40e94de21 10/fsogsmd/src/plugins/modem_samsung/channel.vala: Oct 18 19:24:08 freesmartphone.org: fsogsmd: modem_samsung: first send request message should have id 1 not 0 Oct 18 19:24:08 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:24:09 freesmartphone.org: 03morphis 07cornucopia * ra26f3367ba35 10/fsogsmd/src/plugins/modem_samsung/plugin.vala: Oct 18 19:24:09 freesmartphone.org: fsogsmd: modem_samsung: register the first two mediators for sim commands Oct 18 19:24:10 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:24:10 freesmartphone.org: 03morphis 07cornucopia * rd49c41ebea9e 10/fsogsmd/src/plugins/modem_samsung/mediators_sim.vala: Oct 18 19:24:11 freesmartphone.org: fsogsmd: modem_samsung: fix pin validation Oct 18 19:24:11 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:24:12 freesmartphone.org: 03morphis 07cornucopia * ra88be002ad7e 10/fsogsmd/src/plugins/modem_samsung/mediators_sim.vala: Oct 18 19:24:41 freesmartphone.org: fsogsmd: modem_samsung: set correct length of data array for message we want to send Oct 18 19:24:41 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:24:41 freesmartphone.org: 03morphis 07cornucopia * r5962ffeaa856 10/fsogsmd/src/plugins/modem_samsung/channel.vala: Oct 18 19:24:41 freesmartphone.org: fsogsmd: modem_samsung: dump request message when sending it to the modem Oct 18 19:24:42 freesmartphone.org: Signed-off-by: Simon Busch Oct 18 19:25:14 sorry for the noise Oct 18 19:25:48 is there still demand for dbus-ish iliwi-like app? Oct 18 19:26:18 or has someon meanwhile steppe dup and wrote something? :) Oct 18 19:28:55 zub: meanwhile connman + the connman e gadget works fine Oct 18 19:29:10 ah, ok Oct 18 19:29:19 well almost fine Oct 18 19:29:26 connman needs some fso resource integration Oct 18 19:29:27 so no more demand for my "killer ap" that's been in the making for last 6 months Oct 18 19:29:36 hehe Oct 18 19:29:37 (snail-speed-development) Oct 18 19:30:11 no idea... I like the gadget way... as one does not have to keep an app started Oct 18 19:30:16 now I got stopped on this eflvala thing... in fact even if I build it I get crash in genlist :-( Oct 18 19:30:40 and maybe someday it will integrate usb + wifi + gprs alltogether Oct 18 19:30:42 even in the eflvala demo, I guess my box is cursed Oct 18 19:31:09 btw. anybody using debian here? maybe it's some distro-thing (dunno how though) Oct 18 19:35:44 * mrmoku arch linux Oct 18 19:41:48 JaMa|Off: task-core-sdk ? Oct 18 19:42:34 hmm... no... does not look like Oct 18 19:43:18 task-sdk-host-nativesdk.bb Oct 18 19:46:37 isn't that native toolchain? Oct 18 19:46:50 at least I believe it used to be bac in the good old days Oct 18 19:47:24 there also used to be some meta-sdk that pulled the host and target tasts for sdk Oct 18 19:47:32 meta-toolchain? Oct 18 19:48:11 it used to be meta-toolchain-*.bb Oct 18 19:48:23 zub: yeah... nothing is like in the good old days anymore :/ Oct 18 19:48:35 * zub doesn't want to live on this planet any more Oct 18 19:49:14 nah, not that dramatic :-P Oct 18 19:51:17 ah, here he comes :) Oct 18 19:51:40 zub, hi, did you solve the issue? :P Oct 18 19:51:48 no, I'm giving up Oct 18 19:51:53 on life :-P Oct 18 19:52:18 zub, you could try building with my script :P Oct 18 19:52:22 but before that I'd like to at least get the two patches: http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0002-elementary-rename-elm-to-elementary-to-match-EFL-cha.patch + http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0003-eflvala-get-rid-of-a-warning.patch in Oct 18 19:52:35 hm, ok Oct 18 19:53:30 zub, http://espeweb.net/misc/pespin/scripts/get_e3.sh Oct 18 19:54:52 JaMa|Off, hmm as far as I see my name was correct in last patch I sent before you told me it was wrong :S Oct 18 19:55:43 pespin: http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/patch2-rant.txt Oct 18 19:55:49 thx Oct 18 19:55:52 I'll give it a try Oct 18 20:00:56 zub, ok, but have in mind that i don't have commit access so I can't push anything :P Oct 18 20:01:39 so I'll just keep pestering everyone ontill I get a ban :) (well, not really, I give up soon :) Oct 18 20:02:49 pespin: MODULES="eina eet evas ecore embryo edje elementary" ./get_e3.sh checkout clean configure make install <- like this? Oct 18 20:03:06 (well, clean sound superfluous, I just tossed it in because I felt like :) ) Oct 18 20:04:24 zub, ./get_e3.sh whatever_no_used Oct 18 20:04:30 it will output all the stuff needed Oct 18 20:04:36 you can remove last "e" module Oct 18 20:04:40 as you don' use e17 afaik Oct 18 20:04:46 I don't Oct 18 20:04:50 then ake the list of modules Oct 18 20:04:57 and I don't want to build more than needed... sigh Oct 18 20:05:02 and MODULES="here the modules" ./get_e3.sh checkout Oct 18 20:05:32 ok Oct 18 20:05:33 when you are finsished, I usually do: ./get_e3.sh clean update configure make install Oct 18 20:05:44 to rebuild everything to last rev Oct 18 20:05:54 what's "engage" btw? Oct 18 20:05:55 zub, you can remove "engage" module too Oct 18 20:05:58 heh Oct 18 20:05:59 gut :) Oct 18 20:06:04 it's a module for e17 Oct 18 20:06:10 like a OSX dock Oct 18 20:06:18 * pespin dinner, bbl Oct 18 20:10:28 ciao Oct 18 20:10:53 list! Oct 18 20:11:06 ciao a tutti Oct 18 20:15:01 morphis, maybe I should try the driver dcordes uses....no SDIO IRQ in that driver Oct 18 20:15:14 hm Oct 18 20:15:40 which one is he using? Oct 18 20:15:46 one that does plain WEXT Oct 18 20:15:53 that rules out connman Oct 18 20:16:03 hm Oct 18 20:16:24 but iliwi should work Oct 18 20:17:09 there is still a plain wext in the kenrel so it could work Oct 18 20:19:10 for a first start it should be an option Oct 18 20:22:31 pespin: I've fixed it in my jansa/pull branch, but see From: here: http://patches.openembedded.org/patch/13419/ Oct 18 20:23:34 morphis, anyway we do not have a lot of hope for upstream Oct 18 20:24:38 pespin: hmm you're right.. it was broken by patchwork.. here it seems right http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-October/035561.html Oct 18 20:25:06 to get an upstream kernel running on our device or getting the upstream wifi driver working? Oct 18 20:26:00 mrmoku, if i'm in mood, yes Oct 18 20:27:24 morphis, I mean everything upstream with reguards to the kernel Oct 18 20:27:32 with gta02 it's possible tough Oct 18 20:28:35 GNUtoo: ok Oct 18 20:29:08 but can we integrate the missing wext command in our driver with comes with our kernel? Oct 18 20:29:45 we do the countrary Oct 18 20:30:01 I don't know WEXT enough to do that Oct 18 20:30:09 if someone knows it it might be easier Oct 18 20:38:16 pespin: built... got the same error :-/ Oct 18 20:38:38 that is, building libeflvala by hand, as that is not handled by the get_e3 script, right? Oct 18 20:39:08 thought I'm installing efl into ~ (~/efl-pespin in this case), not in the default /usr/local Oct 18 20:39:41 but that should not be an issue, I believe the env is Ok, even the linker is helpful with telling me Oct 18 20:39:50 /usr/bin/ld: note: 'ecore_evas_init' is defined in DSO /home/zub/efl-pespin/lib/libecore_evas.so.1 so try adding it to the linker command line Oct 18 20:42:30 a distro issue? but how? oh well... Oct 18 20:42:34 that's frustrating Oct 18 20:43:49 splitting ecore.vapi into ecore.vapi and ecore-evas.vapi handles it (because then valac explicitly asks pkgconfig for ecore-evas)... but it'd be a change that would be considered unneeded around here I guess Oct 18 20:44:06 OTOH it doesn't sound that bad as there is ecore-evas, so it could be just split off... Oct 18 20:44:47 zub, MODULES="vala" ./get_e3 checkout :) Oct 18 20:44:56 aaah ok, I try that Oct 18 20:45:54 zub, basically it looks for dirs in sv.enlightenment.org/SOME_DIRS/module_name Oct 18 20:47:03 pespin: anyway, get the error again :-/ Oct 18 20:47:21 I'm running git vala, but they should not have influence on the linking, which is not done via vala anyway Oct 18 20:47:26 s/they/that/ Oct 18 20:47:27 zub meant: I'm running git vala, but that should not have influence on the linking, which is not done via vala anyway Oct 18 20:50:33 pespin: what pkg-config version you have, pls? Oct 18 20:50:43 0.26 here Oct 18 20:56:10 pespin: could you pls. paste somewhere your vala/eflvala/Makefile? Oct 18 20:57:32 zub, moment please Oct 18 20:58:19 pespin: can you send some dummy patch to with ie meta-efl to oe-devel ML? Oct 18 20:58:25 pespin: just to test patchwork Oct 18 20:58:29 zub, 0.26 too Oct 18 20:59:46 pespin: http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-October/035662.html Oct 18 20:59:51 zub, http://paste.pocoo.org/show/494568/ Oct 18 21:00:11 pespin_: danke schön Oct 18 21:00:42 JaMa|Off, moment Oct 18 21:01:29 DEPS_LIBS is the same for me and for you Oct 18 21:02:37 JaMa|Off, to try if the name problem is fixed? Oct 18 21:02:49 yes Oct 18 21:03:17 because from Kerr reply I guess someone has to update names in patchwork database Oct 18 21:03:23 ah, the libtool cmdline is the same, only what libtool does differs Oct 18 21:07:26 JaMa|Off, sent Oct 18 21:07:57 freesmartphone.org: 03angelo 07aurora * rd3f7b0ca07e0 10/aurora-daemon/src/ (4 files): aurora-daemon: first implementation of key events handling Oct 18 21:07:58 freesmartphone.org: 03angelo 07aurora * r21faae04ae2b 10/ (6 files in 5 dirs): Merge branch 'master' of git.freesmartphone.org:aurora Oct 18 21:08:09 maybe some .la stuff? anyway, libtool is evil magic Oct 18 21:08:31 yeah hehe Oct 18 21:09:27 freesmartphone.org: 03angelo 07aurora * r0ea22da34528 10/aurora-daemon/data/keyboard/MainKeyboard.qml: Oct 18 21:09:27 freesmartphone.org: aurora-daemon: update keyboard Oct 18 21:09:27 freesmartphone.org: * - Replace Rectangles by Buttons components Oct 18 21:09:27 freesmartphone.org: * - Implement signal sending by QML side Oct 18 21:09:32 freesmartphone.org: 03angelo 07aurora * rcf73643db604 10/aurora-daemon/src/ (4 files): Merge branch 'master' of git.freesmartphone.org:aurora Oct 18 21:11:19 and whoever invented the .la files should be shot... (ok, maybe they have some use on some exotic systems, but for me they were always a pain...) Oct 18 21:11:35 and adding there the "# Please DO NOT delete this file! Oct 18 21:11:36 " Oct 18 21:11:40 blah Oct 18 21:17:03 the .la files (e.g. elementary) seems to include the ecore_evas dependency, so why does not libtool pull the lib in? :-( Oct 18 22:20:11 mrmoku: sryx Oct 18 22:20:15 *sry Oct 18 22:20:26 I had a pointless discussion with Richard Stallmen Oct 18 22:20:37 and had to drink 3liters of beers afterwards Oct 18 22:20:51 to kill of my depression about an RL meeting with him -.- Oct 18 22:21:00 I'm still drunken, soo... Oct 18 22:21:06 but at least you know why your Oct 18 22:21:12 mrmoku: pong Oct 18 22:21:19 had such a long delay Oct 18 22:21:20 -.- Oct 18 22:21:30 1.3°% Oct 18 22:21:32 btw Oct 18 22:21:34 >_< Oct 18 22:21:59 and screw the endorcement Oct 18 22:22:02 we don't need it Oct 18 22:22:09 screw the fsf Oct 18 22:22:36 if they wanna prevent us from using the term "free" to describe our phone Oct 18 22:22:46 I'm not scared of using my fist Oct 18 22:22:48 >_< Oct 18 22:23:05 anyway Oct 18 22:23:10 I've got a lot of agressions Oct 18 22:23:15 maybe you read it >_> Oct 18 22:23:21 so Oct 18 22:23:24 pespin_: I found out what causes the genlist (anything using genlist) to crash for me: git valac -> crash (sigsegv), valav 0.12.1 (debian testing repo) -> genlist works Oct 18 22:23:36 sigh... another roadblock, /me is going to cry Oct 18 22:23:48 pespin_: what valac are you using? Oct 18 22:23:50 zub: git valac isn't a valid commant Oct 18 22:24:07 zub: and you mean release number Oct 18 22:24:08 right? Oct 18 22:24:29 * leviathan zub: is an IRC command Oct 18 22:24:30 zub, I was using 0.13 since some days ago Oct 18 22:24:35 then 0.14 appeared Oct 18 22:24:44 and with it, 2 bugs with struct properties Oct 18 22:24:55 one of them is fixed now in git, so I'm using git atm Oct 18 22:25:01 valac 3d153c752e3a1c561d96294ced17003db9b577d5 Oct 18 22:25:14 zub: you can just checkout from git and make dh_make Oct 18 22:25:15 pespin_: and genlists work for you? Oct 18 22:25:29 leviathan: ehm, what are you talking about (the whole time)? Oct 18 22:25:59 zub: you can just checkout the newest vala from git Oct 18 22:26:04 and built a debian package of it Oct 18 22:26:08 by using dh_mak Oct 18 22:26:09 zub, no idea, I'm not using genlist in my apps ;) (I'm using elm_list atm, but I'll pass to genlist in the future) Oct 18 22:26:11 *dh_make Oct 18 22:26:27 pespin_: could you run the libeflvala example (for elementary) + click the genlist demo button? Oct 18 22:26:33 I get immediatelly sigsegv... Oct 18 22:26:39 ok Oct 18 22:27:03 leviathan: I could, though I don't see how that would help me Oct 18 22:27:12 sry Oct 18 22:27:13 I just build by hand and install in ~/vala Oct 18 22:27:16 nah, n/p Oct 18 22:27:21 that's what richard stallman made of me xD Oct 18 22:27:23 no Oct 18 22:27:26 lol :) Oct 18 22:27:31 had today a RL meeting with him Oct 18 22:27:38 collision of opinions Oct 18 22:27:49 normally, meeting an idol is positiv Oct 18 22:28:03 but if you get cutoff by the director of the discussion Oct 18 22:28:11 because you can't find a consense Oct 18 22:28:29 and he still is refusing to accept gta04 as a "free" project Oct 18 22:28:32 screw him Oct 18 22:28:38 and then I got depresses Oct 18 22:28:41 *depressed Oct 18 22:28:50 btw. I discussed the EFL linking stuff in #workingset... if I got it correctly, the result was: don't do this transitive app uses liba, liba puls libb, and app then happily uses libb stuff... Oct 18 22:28:55 hmm Oct 18 22:29:11 but the EFL guys have different opinion, if I got that I was in #e yest. morning Oct 18 22:29:24 leviathan: I used to like him more than I do now... Oct 18 22:29:56 at least In now understand the relationship between him and linus torwalds Oct 18 22:29:59 >_< Oct 18 22:30:08 btw Oct 18 22:30:12 how did you get to meet that VIP anyway? Oct 18 22:30:20 well Oct 18 22:30:38 question time + friends as moderator ;-) Oct 18 22:30:44 BUT Oct 18 22:30:47 hmm Oct 18 22:30:59 I still have to meet linus torwalds now Oct 18 22:31:08 zub, it segfaults here too Oct 18 22:31:19 pespin_: debian or OE? Oct 18 22:31:33 leviathan, archlinux Oct 18 22:31:36 hmm Oct 18 22:31:49 vala from few days ago, same for EFL Oct 18 22:31:55 vala on arch? O_o Oct 18 22:31:58 probably the vapi file is a bit outdated Oct 18 22:32:02 or there's a bug in the example code Oct 18 22:32:03 pespin_: oh, I *love you*! at least something is reproducible across boxes! :) Oct 18 22:32:12 otherwise I'd just go banans Oct 18 22:32:13 for sure Oct 18 22:32:15 hehe Oct 18 22:32:40 leviathan, yeah, it's nice, I have latest valac from git just calling yaourt -S vala-git Oct 18 22:32:41 pespin_: my app also crashes with genlist Oct 18 22:32:41 :P Oct 18 22:32:54 thought I could well have it wrong, but it was working at some point in the past Oct 18 22:33:03 zub, is your app public somewhere? Oct 18 22:33:06 also! building the sam thing with older valac -> it works Oct 18 22:33:10 as I said Oct 18 22:33:18 my app is shit... and gui is not really committed Oct 18 22:33:19 zub, then it might be a valac error hehe Oct 18 22:33:54 ssh://zub.lamer.la/var/scm/git/shr_supplicant.git (pardon the name :) ) Oct 18 22:34:27 but the gui is probably not commited in that as I only recently started adding that... and then hit all the libeflvala + crashing genlist fun Oct 18 22:35:16 ok, sry, that's drunken IRC within a logged channel, and I certainly will regret it tomorrow Oct 18 22:35:16 -.- Oct 18 22:35:19 aaah, you can't get to the ssh:// url Oct 18 22:35:20 mmnt Oct 18 22:35:20 so brb Oct 18 22:35:24 until I'm clean again Oct 18 22:35:31 ^_^" Oct 18 22:35:32 leviathan: heh Oct 18 22:35:36 'night, then Oct 18 22:35:39 thx Oct 18 22:35:41 ^^ Oct 18 22:35:42 gn8 Oct 18 22:35:55 at least I could tell what I was feeling Oct 18 22:35:58 that really helped Oct 18 22:35:59 thx Oct 18 22:36:04 even if it was really offtopic Oct 18 22:36:07 nix zu danken Oct 18 22:36:10 leviathan, gnight hehe Oct 18 22:36:11 in respect to this channel Oct 18 22:36:14 ^^ Oct 18 22:36:16 gn8 Oct 18 22:36:30 pespin_: git://zub.lamer.la/shr_supplicant.git this should work Oct 18 22:36:39 leviathan, write about it in your blog, I follow it! :) Oct 18 22:36:47 ok! Oct 18 22:37:00 pespin_: I wanted to change something in iliwi and when I looked at the code, I... didn't like it Oct 18 22:37:22 lol I already have it in my git dir, I don't know how Oct 18 22:37:23 xD Oct 18 22:37:37 I was hold aback by lazyness, but also by stretching valac to its limis w.r.t. generics etc. :-/ Oct 18 22:37:55 pespin_: at some point I noticed that in the ML Oct 18 22:38:05 but the progress was terribly slow - see log Oct 18 22:39:22 anyway nothing you can really tun in the app... this libgsupplicant is something on top of wpa_supplicant, has some hardcoded-basic tast, that does some scanning and tries to connect to a hardcoded network Oct 18 22:39:33 libthreadconnector again has some crude test Oct 18 22:39:39 and gui is... not really anywhere Oct 18 22:39:54 well, first step to have a working app hehe Oct 18 22:39:57 libthreadconnector was the part that was stress-testing valac :) Oct 18 22:39:58 I started like this with etalk Oct 18 22:42:12 I decided to make the supplicant part in one thread, gui in another (yes, async dbus call could be ok with just one thread, though if there's some dbus mess, I think they can block at least a while) Oct 18 22:42:20 and that made for some challenges :) Oct 18 22:42:54 pespin_: any ideas about the genlist? I don't really know where to look, I saw the backtrace, but that doesn't help much Oct 18 22:43:02 some strncat with NULL, IIRC Oct 18 22:43:20 so possibly some NULL being fed when elementary doesn't like it (?) Oct 18 22:43:24 zub, first would be to check Elm C API and see that the vapi is still correct Oct 18 22:43:35 next, check example code Oct 18 22:43:45 then generated C example code Oct 18 22:43:55 helped with backtrace Oct 18 22:44:09 but anyway, first thing would be checking the vapi Oct 18 22:44:27 sounds reasonable Oct 18 22:44:30 almost, that is :) Oct 18 22:44:36 let's hope I find some time during next weeks to update them Oct 18 22:44:37 first thing: get some sleep! Oct 18 22:44:44 at least for me Oct 18 22:45:00 damn it's again so late Oct 18 22:45:05 hehe yeah Oct 18 22:45:39 anyway thanks a lot for consulting stuff Oct 18 22:45:41 'night Oct 18 22:47:18 zub, np, I wish you could get it working, would be nice to have more people writing stuff with vala+efl :) Oct 18 23:11:38 zub: if the #e guys have a different opinion regarding indirect linking, they will have a fun time fighting it out with Fedora and Debian... Oct 18 23:12:50 antrik: Im on debian btw Oct 18 23:13:39 though I did not figure the precise issue; it's libtool behaving differently... dunno why Oct 18 23:14:27 the indirect stuff (if I make libeflvala depend on ecore-evas) works... Oct 18 23:14:47 strange, complex and demotivating :( Oct 18 23:15:05 also, Im sleeping already :) Oct 18 23:15:38 I don't know the exact connection; but it sounds very much like the --no-add-needed stuff (or whatever the option is called) Oct 18 23:15:53 Fedora changed the default a while ago, and Debian fairly recently Oct 18 23:16:18 antrik: have you seen the discussion in #workingset? Oct 18 23:16:23 no Oct 18 23:17:11 but I saw several discussions relating to this issue, and to my uneducated ears it sounds very much like the same... Oct 18 23:17:54 Im not convinced :( Oct 18 23:18:29 I believe that is a dangerous stuff, but my build breaks a bit earlier than that Oct 18 23:19:17 same libtool invocation makes libtool use diff libs when linking on my vs. pespin's box Oct 18 23:21:03 so, once gcc (asked to link) is given the extra libs, once it's not Oct 18 23:21:25 the actual gcc command line is different? Oct 18 23:21:31 yes Oct 18 23:21:35 hm Oct 18 23:21:41 Im on phone now Oct 18 23:21:48 but let me fid it Oct 18 23:22:03 I have no idea what libtool does, so no more guesses from me ;-) Oct 18 23:22:34 it's http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala Oct 18 23:23:07 and there's a log with some 'diff' or so in name Oct 18 23:23:43 libtool-difference.log Oct 18 23:23:52 I dont know why Oct 18 23:25:07 on #workingset there was a knowledgable and helpful guy, he'd explain why he thinks it's wrong, but he'd stubbornly refuse to look at that difference Oct 18 23:25:42 and I finally doze off, this time for real Oct 18 23:25:50 'night Oct 19 00:50:42 good night all **** ENDING LOGGING AT Wed Oct 19 02:59:57 2011