**** BEGIN LOGGING AT Thu Oct 20 02:59:57 2011 Oct 20 05:37:38 leviathan: long done :) Oct 20 06:09:03 how to solve the "multiple providers" problem? Oct 20 06:09:10 any ideas? Oct 20 06:19:37 hitmoon: what does your machine conf look like? Do you have specified a preferred kernel in there? Oct 20 06:21:05 REFERRED_PROVIDER_virtual/kernel = "linux-msm7" Oct 20 06:21:15 PREFERRED_PROVIDER_virtual/kernel = "linux-msm7" Oct 20 06:21:35 i have this entry in my conf Oct 20 06:22:05 Is that correct? Oct 20 06:22:28 hitmoon: the second one looks correct, yes Oct 20 06:22:43 if it's just that one module... maybe you have disabled it in your kernel config? Oct 20 06:23:59 what that means? Oct 20 06:24:58 do you get the multiple provider error for _all_ kernel stuff, or just that one module? Oct 20 06:25:18 all kernel stuff Oct 20 06:25:23 ahh, ok Oct 20 06:26:05 then ,what should i do? Oct 20 06:26:29 and there is such a kernel recipe called linux-msm7? Oct 20 06:26:43 yes Oct 20 06:27:23 dunno then Oct 20 06:28:42 :-( Oct 20 06:30:00 hitmoon: maybe the kernel recipe is missing COMPATIBLE_MACHINE for the one your adding? Oct 20 06:30:09 COMPATIBLE_MACHINE = "htcnexus|htcleo" Oct 20 06:30:19 Oh Oct 20 06:30:21 yes Oct 20 06:30:23 maybe Oct 20 06:30:35 Let's have a try Oct 20 06:30:51 or COMPATIBLE_MACHINE = "nokia900" Oct 20 06:30:51 DEFAULT_PREFERENCE = "-2" Oct 20 06:30:51 DEFAULT_PREFERENCE_nokia900 = "1" Oct 20 06:31:10 hitmoon: DEFAULT_PREFRENCE might also be the problem Oct 20 06:40:08 mrmoku: Yes , it works! Oct 20 06:40:40 smile Oct 20 06:40:56 :) Oct 20 06:42:02 thank you Oct 20 06:44:52 mrmoku: what dou you think is key step in porting SHR? Oct 20 06:45:23 to an unsupport mobile device Oct 20 06:50:15 hitmoon: kernel support Oct 20 06:50:39 especially alsa is often a problem Oct 20 06:51:15 I have an offical kernel source for my device Oct 20 06:56:06 hitmoon: another key point is the modem of course Oct 20 06:56:39 yes Oct 20 06:59:10 Before I know SHR, I always think when can i have a free , usable mobile os running linux Oct 20 06:59:47 then android become popular Oct 20 07:01:11 But, when I actually got an android phone, I find all it's application running opon java Oct 20 07:01:29 and this is not good Oct 20 07:06:36 ah maybe 4-5 weeks to new EFL release :) Oct 20 07:10:46 JaMa|Off: ahh... init functions Oct 20 07:11:30 ? Oct 20 07:11:37 * JaMa initialized Oct 20 07:12:41 haha Oct 20 07:12:46 hitmoon: also the way interfaces are presented to applications does not look good in android.. next to the fact that all interfaces are java based Oct 20 07:13:26 everything is class/method based.. not data driven.. and because of java you have to jump to the jni loop Oct 20 07:13:43 yes Oct 20 07:13:53 09:03 < mrmoku> JaMa|Off: xserver-common is using `machine_id` in the big case to get the device... what is it? Oct 20 07:14:01 guess that one did not make it due to disconnect :) Oct 20 07:14:02 that what makes the dbus based approach so elegant.. Oct 20 07:14:25 you can easily upgrade components and interfaces.. you can easily replace stuff.. Oct 20 07:17:00 mrmoku: see /etc/init.d/functions Oct 20 07:17:18 awk 'BEGIN { FS=": " } /Hardware/ { gsub(" ", "_", $2); print tolower($2) } ' JaMa: yeah, found that Oct 20 07:18:37 thx :) Oct 20 07:29:31 freesmartphone.org: 03mok 07framework * rf1c37ff374e9 10/etc/om-gta04/frameworkd.conf: Oct 20 07:29:31 freesmartphone.org: add config for om-gta04 Oct 20 07:29:31 freesmartphone.org: Signed-off-by: Klaus Kurzmann Oct 20 07:40:33 SHR: 03Martin.Jansa 07shr-chroot * r747bc86068f2 10/ (169 files in 16 dirs): system upgrade Oct 20 11:49:46 PaulFertser: hey, going to use your gpsd activation hook for gta04 now :-) Oct 20 11:54:48 mrmoku: yo, cool :) Oct 20 11:54:58 Finally it's of any use :) Oct 20 12:19:41 PaulFertser: I wonder what to do exactly in there though... permanently activate GPS? Oct 20 12:29:51 would be interesting if fsoraw could request gps with the pid of gpsd... Oct 20 13:00:18 * JaMa changed his shr branch checkouts to track contrib/shr instead of origin/master Oct 20 13:00:35 I hope I haven't overwrite any patches in oe-core-contrib/shr in my last push :/ Oct 20 13:01:00 mrmoku: I've merged your meta-oe-contrib/shr patch for gta04 fine Oct 20 13:01:51 JaMa: ok Oct 20 13:02:09 did you push something to oe-core-contrib/shr yesterday? Oct 20 13:20:20 I'm about to gibe bq27k-detail script some maintenance. Please provide all alternative sysfs pathes to the */dump sysnode so the script can pick the right one Oct 20 13:21:02 lindi-: ^^^ Oct 20 13:21:07 mrmoku: ^^^ Oct 20 13:21:31 DocScrutinizer: you are assuming this is documented somewhere? ;) Oct 20 13:21:41 alternatively give a shell glob that expands to the right sysnode Oct 20 13:22:07 DocScrutinizer: I can not boot older kernels to test it, udev needs 2.6.34 with some patches :/ Oct 20 13:22:12 I'm not exactly assuming this is documented somewhere Oct 20 13:22:16 and I can't boot newer kernels since SD does not work there Oct 20 13:22:48 so the path I mentioned works for 2.6.34 but I'm afraid I can't provide more paths Oct 20 13:22:52 lindi-: revert those SD patches as we do :) Oct 20 13:23:00 JaMa: which ones? Oct 20 13:23:07 well, I got the original path of http://people.openmoko.org/joerg/battery/bq27k-detail Oct 20 13:23:09 err Oct 20 13:23:18 err /sys/devices/platform/s3c24xx_pwm.0/hdq.0/hdq/dump Oct 20 13:24:36 * DocScrutinizer wonders if a `find /sys/devices/ -name dump | grep hdq` would work Oct 20 13:24:57 lindi-: https://gitorious.org/shr/linux/commits/shr-2.6.39-nodrm first 10 commits (9 are reverts) on top of git.om.org HEAD Oct 20 13:25:54 btw: for maemo: i2cdump -y 2 0x55 b|tail -n +2|cut -d " " -f 2-17|./bq27k-detail " " Oct 20 13:26:29 JaMa: aha hmm Oct 20 13:26:40 FWIW Oct 20 13:26:55 DocScrutinizer: maybe the path finding and parsing could indeed be kept separate Oct 20 13:27:04 I could add "om battery hdq-dump" Oct 20 13:27:19 I won't go that path Oct 20 13:27:26 but I will :) Oct 20 13:27:29 that script is meant to be complete Oct 20 13:27:55 you're free to call it with correct path as $1 Oct 20 13:28:04 yeah but maintaining sysfs paths in 10 different places is not maintainable... Oct 20 13:28:55 lindi-: reported here http://lists.openmoko.org/pipermail/openmoko-kernel/2011-May/011350.html and seems that there is easier way to fix this misbehavior http://lists.openmoko.org/pipermail/openmoko-kernel/2011-August/011352.html Oct 20 13:29:01 I don't care what's maintainable, I want the path to sysfs for the 2 or 3 versions that emerged with newer kernels on FR lately Oct 20 13:29:31 DocScrutinizer: 2.6.39 /sys/devices/platform/s3c2440-i2c/i2c-0/0-0073/hdq/dump Oct 20 13:29:38 JaMa: thanks Oct 20 13:30:51 seems a `find /sys/devices/platform/s3c24* -name dump` should do? Oct 20 13:31:17 JaMa: hmm Oct 20 13:31:46 DocScrutinizer: well here I've used your `find /sys/devices/ -name dump | grep hdq` and this was unique result.. so for 2.6.39 it seems fine Oct 20 13:32:27 fair enough, probably hdq is a more specific parameter for a match Oct 20 13:33:05 JaMa: thanks a lot :-) Oct 20 13:34:18 DocScrutinizer: maybe add -path \*/hdq/*\ to be more picky about hdq :) Oct 20 13:34:36 to find? Oct 20 13:34:45 s#*\#\*#g Oct 20 13:34:49 yup to find Oct 20 13:35:15 yup, will do, saves the grep Oct 20 13:35:23 well don't know if it's supported by busybox version Oct 20 13:35:32 meh Oct 20 13:35:32 but I guess you don't care about that possibility :) Oct 20 13:35:56 not really, but then otoh having bash doesn't mean find is sane Oct 20 13:36:21 that's why I've put \ Oct 20 13:37:00 because otherwise it could expand to something like -path something-strange/hdq/something-even-stranger and find nothing Oct 20 13:37:40 or syntax error if there is more strange things in that directory :) Oct 20 13:37:41 IroN900:~/bin# busybox find /sys/ -path '*/backlight/*' -name brightness Oct 20 13:37:43 /sys/class/backlight/acx565akm/brightness Oct 20 13:39:14 so it works at least on maemo messybox Oct 20 13:39:55 JaMa: thanks a lot Oct 20 13:40:03 JaMa: is there some known list of regressions from 2.6.34 to 2.6.39 that I should know? ;) Oct 20 13:43:11 lindi-: I'm using it since May and ie accels seems to work better then with .34 (no hard lockups) Oct 20 13:43:22 lindi-: but check the rest of patches we have :) Oct 20 14:21:23 JaMa: https://gitorious.org/shr/linux/commits/shr-2.6.39-nodrm does not have arch/arm/configs/*gta* Oct 20 14:21:31 JaMa: or .config Oct 20 14:22:08 freesmartphone.org: 03mok 07cornucopia * r3f3d9a6f6247 10/fsodeviced/ (4 files in 3 dirs): Oct 20 14:22:08 freesmartphone.org: fsodeviced: add gta04_powercontrol plugin Oct 20 14:22:08 freesmartphone.org: Signed-off-by: Klaus Kurzmann Oct 20 14:22:09 freesmartphone.org: 03mok 07cornucopia * r9aee573422d3 10/fsodeviced/ (4 files in 3 dirs): Oct 20 14:22:09 freesmartphone.org: fsodeviced: add config for gta04 Oct 20 14:22:09 freesmartphone.org: Signed-off-by: Klaus Kurzmann Oct 20 14:22:28 mickeyl: ping Oct 20 14:24:05 lindi-: read https://gitorious.org/shr/linux#more says they are at http://git.shr-project.org/git/?p=meta-smartphone.git;a=tree;f=meta-openmoko/recipes-kernel/linux/linux-2.6.39 and really they are :) Oct 20 14:38:49 JaMa: aha Oct 20 14:39:41 JaMa: with http://git.shr-project.org/git/?p=meta-smartphone.git;a=blob_plain;f=meta-openmoko/recipes-kernel/linux/linux-2.6.39/om-gta02/defconfig;hb=HEAD and shr-2.6.39-nodrm branch I get asked about things like kernel compression mode and "Patch physical to virtual translations at runtime (EXPERIMENTAL) (ARM_PATCH_PHYS_VIRT) [N/y/?] " Oct 20 14:41:51 only those 2 or more? Oct 20 14:42:17 because at least compression mode is handled by kernel.bbclass (or linux.inc) in oe Oct 20 14:42:25 JaMa: more Oct 20 14:42:27 JaMa: like "Support for paging of anonymous memory (swap) (SWAP) [Y/n/?] (NEW) Oct 20 14:42:29 and only the result is stored with "make defconfig" Oct 20 14:42:37 JaMa: "POSIX Message Queues (POSIX_MQUEUE) [N/y/?] (NEW) Oct 20 14:42:57 when you do make oldconfig right? Oct 20 14:43:18 JaMa: BSD Process Accounting (BSD_PROCESS_ACCT) [N/y/?] (NEW) Oct 20 14:43:27 JaMa: "silentoldconfig" yes Oct 20 14:43:52 JaMa: could you test http://people.openmoko.org/joerg/battery/bq27k-detail2 please? Oct 20 14:44:17 JaMa: and definitely at least BSD_PROCESS_ACCT is old stuff but still not in http://git.shr-project.org/git/?p=meta-smartphone.git;a=blob_plain;f=meta-openmoko/recipes-kernel/linux/linux-2.6.39/om-gta02/defconfig;hb=HEAD Oct 20 14:45:15 lindi-: savedefconfig is saving only minimal defconfig.. so the rest is on default values Oct 20 14:45:51 JaMa: hmm Oct 20 14:45:55 JaMa: that sounds useful Oct 20 14:46:02 JaMa: but how should I use that? Oct 20 14:46:04 lindi-: so something like `yes '' | make oldconfig` Oct 20 14:46:15 JaMa: that won't work to questions that want an integer Oct 20 14:46:29 JaMa: how is shr building the kernels? Oct 20 14:47:12 lindi-: read do_configure_prepend in http://git.shr-project.org/git/?p=meta-smartphone.git;a=blob_plain;f=meta-openmoko/recipes-kernel/linux/linux.inc;hb=HEAD Oct 20 14:47:22 lindi-: and yes it works for integers too Oct 20 14:47:51 uhm Oct 20 14:48:00 JaMa: it might be easier for me to just get a /proc/config.gz from a running system really Oct 20 14:48:36 lindi-: mmt Oct 20 14:50:17 hello Oct 20 14:50:41 lindi-: http://build.shr-project.org/tests/jama/kernel-nodrm/config-2.6.39.gz Oct 20 14:51:28 DocScrutinizer: http://paste.pocoo.org/show/495526/ something I should try with it? Oct 20 14:52:50 DocScrutinizer: cannot try with i2cdump -y 0 0x55 b, because I have only /dev/i2c-0 which returns lot of XX only ;) Oct 20 14:53:36 freesmartphone.org: 03mok 07cornucopia * rbe97abd9eb73 10/fsotdld/ (4 files in 3 dirs): Oct 20 14:53:36 freesmartphone.org: fsotdld: add config for gta04 Oct 20 14:53:36 freesmartphone.org: Signed-off-by: Klaus Kurzmann Oct 20 14:54:09 JaMa: thanks. I already tetsed on N900, was just interested on result on OM platform Oct 20 14:54:47 DocScrutinizer: should I try with 3.1 on n900? Oct 20 14:55:03 should work exactly same way Oct 20 14:55:10 ok Oct 20 14:55:28 as on N900 it *should* use i2cdump which is kernel agnostic Oct 20 15:06:57 burn in he^H^H^H RIP Gaddafi Oct 20 15:09:22 :) Oct 20 15:13:40 wb apt Oct 20 15:33:51 JaMa: lindi@ginger:~$ uname -r Oct 20 15:33:51 2.6.39-38861de6192e78df Oct 20 15:33:53 JaMa: woot! Oct 20 15:44:37 JaMa: "om wifi power 1; sleep 10; ifconfig -a" does not show wifi interface Oct 20 15:44:43 JaMa: and omhacks needs a few new sysfs paths Oct 20 15:49:04 wifi works here.. but I'm not using omhacks to enable it Oct 20 15:50:23 SHR root@gjama ~ $ fsoraw -r WiFi -- sleep 600 & Oct 20 15:50:24 19836 Oct 20 15:50:24 SHR root@gjama ~ $ ifconfig -a Oct 20 15:50:24 eth0 Link encap:Ethernet HWaddr 00:12:CF:8E:FE:34 Oct 20 15:51:54 lindi-: and om wifi power reports 1 when it's enabled and also when I disable it Oct 20 15:54:02 JaMa: there are several different ways to disable it I think Oct 20 15:54:13 JaMa: which one does fsoraw use? Oct 20 15:55:03 JaMa: "om wifi power 0" completely unbinds the driver by writing to /sys/bus/platform/drivers/s3c-sdi/unbind Oct 20 15:56:06 it's talking to fsodevice which is using rfkill afaik Oct 20 15:56:16 JaMa: ok Oct 20 15:56:36 JaMa: I doubt this is causing the issue though Oct 20 15:56:57 JaMa: another small issue: /sys/class/power_supply/battery/ has voltage_now but current_now is missing Oct 20 15:57:14 the bind / unbind thing is no more I think Oct 20 15:57:22 at least it is disabled in fso config Oct 20 15:57:49 mrmoku: yes fso does not do it Oct 20 15:57:55 mrmoku: but I'm not using fso :) Oct 20 15:58:10 yeah, but IIRC we changed that for a reason :-) Oct 20 15:58:21 it might just not work anymore Oct 20 15:59:33 mrmoku: rfkill unblock 1 helps Oct 20 16:01:14 mrmoku: any idea about lack of "current_now" file? Oct 20 16:07:03 also usb hostmode path has changed somewhere Oct 20 16:12:13 mrmoku: frameworkd seems to use bind/unbind still Oct 20 16:12:22 mrmoku: at least if I do it clone git://git.freesmartphone.org/framework.git Oct 20 16:12:43 mrmoku: and read NeoUsbHostPowerControl Oct 20 16:15:26 lindi-: yeah, nobody taking care of frameworkd anymore Oct 20 16:15:33 * mrmoku dinner Oct 20 16:15:35 bbiab Oct 20 16:17:52 mrmoku: well there were recent changes :) Oct 20 16:18:41 mrmoku: cornucopia "class WiFiPowerControl : FsoDevice.BasePowerControl" also pokes "unbind" Oct 20 16:44:43 lindi-: yeah, but it's disabled in the config :-) Oct 20 16:45:40 http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsodeviced/conf/openmoko_gta/fsodeviced.conf;h=8e1aadb2c9acfa1aa19ccffb7941374072fc9904;hb=HEAD#l66 Oct 20 16:46:46 mrmoku: ah. so there's some other program that controls wifi power? Oct 20 16:47:06 the rfkill plugin I think Oct 20 16:47:36 mrmoku: found it Oct 20 16:47:46 mrmoku: I'll update omhacks to do both unbind and rfkill Oct 20 16:48:13 great Oct 20 16:48:14 mrmoku: next problem :-) do you have usb hostmode working with 2.6.39? Oct 20 16:52:10 good question Oct 20 16:52:18 lindi-: never tried it Oct 20 16:52:51 ok Oct 20 16:53:14 I can try in some mins if it works for me Oct 20 16:53:27 mrmoku: well I don't see the hostmode sysfs node Oct 20 16:53:33 mrmoku: so I'd be surprised if it worked Oct 20 16:53:45 good afternoon people Oct 20 16:53:58 mrmoku: more critical issue is that "aplay -l" does not list any soundcards Oct 20 16:55:06 oh Oct 20 16:55:12 hmm Oct 20 16:55:30 mrmoku: lsmod shows snd_soc_s3c24xx Oct 20 16:55:39 lindi-: I tink there is an additional module neede Oct 20 16:55:52 mrmoku: could you share lsmod output? Oct 20 16:57:09 lindi-: in 10 mins Oct 20 16:57:40 mrmoku: I guess I could try some SHR image too Oct 20 17:00:32 mrmoku: I think I need snd_soc_neo1973_gta02_wm8753 Oct 20 17:00:37 but it does not seem to get built Oct 20 17:02:02 http://build.shr-project.org/shr-unstable/images/om-gta02/ seems to have modules tarball only for 2.6.37 Oct 20 17:02:38 lindi-: unstable is obsolete :) Oct 20 17:02:48 check in shr-core Oct 20 17:03:09 will be back to kbd in 3 min :) Oct 20 17:04:34 http://build.shr-project.org/shr-core/images/om-gta02/modules-2.6.39.4-r5-oe12-om-gta02.tgz does not have snd-soc-neo1973-gta02-wm8753 either Oct 20 17:05:19 lindi- kernel-module-snd-soc-dfbmcs320 Oct 20 17:05:22 was the one IIRC Oct 20 17:05:51 mrmoku: loaded it but "aplay -l" still does not list any cards Oct 20 17:05:56 hmm Oct 20 17:06:15 what is dfbmcs320? Oct 20 17:06:27 "Driver for the DFBM-CS320 bluetooth module" Oct 20 17:06:54 lindi-: http://pastebin.com/AXKaSLQA Oct 20 17:07:08 lindi-: hmm... maybe my memory sucks Oct 20 17:07:14 mrmoku: and "aplay -l" prints the card? Oct 20 17:07:36 lindi-: http://pastebin.com/s2MHzvqh Oct 20 17:08:42 mrmoku: you seem to have snd_soc_neo1973_wm8753 -- where did you get it? Oct 20 17:08:59 mrmoku: it is not in http://build.shr-project.org/shr-core/images/om-gta02/modules-2.6.39.4-r5-oe12-om-gta02.tgz Oct 20 17:10:21 lindi-: let me check which kernel module packages are installed Oct 20 17:10:58 kernel-module-snd-soc-core - 2.6.39-r5-oe12 Oct 20 17:10:58 kernel-module-snd-soc-dfbmcs320 - 2.6.39-r5-oe12 Oct 20 17:10:58 kernel-module-snd-soc-neo1973-wm8753 - 2.6.39-r5-oe12 Oct 20 17:10:58 kernel-module-snd-soc-s3c24xx - 2.6.39-r5-oe12 Oct 20 17:10:58 kernel-module-snd-soc-s3c24xx-i2s - 2.6.39-r5-oe12 Oct 20 17:11:01 kernel-module-snd-soc-wm8753 - 2.6.39-r5-oe12 Oct 20 17:11:16 oh so it is a separate package in SHR Oct 20 17:11:35 mrmoku: is that package still built from the same kernel repository? or is it an out-of-tree module? Oct 20 17:12:27 lindi-: no out-of-tree module Oct 20 17:13:07 so it is in the kernel git tree? Oct 20 17:13:47 well it should be in that tarball for sure.. Oct 20 17:13:54 JaMa: aha Oct 20 17:14:47 lindi-: lib/modules/2.6.39.4/kernel/sound/soc/samsung/snd-soc-neo1973-wm8753.ko Oct 20 17:15:02 just downloaded that tarball to be sure and it's there Oct 20 17:15:09 iirc it changed directory.. Oct 20 17:16:09 JaMa: is the openssh-sshd problem a known problem? Oct 20 17:16:26 (sshd user not being created and thus sshd not starting) Oct 20 17:16:40 JaMa: yes it has snd-soc-neo1973-wm8753.ko Oct 20 17:17:57 whaat. Oct 20 17:18:55 mrmoku: yes.. it's that useradd.bbclass :/ Oct 20 17:19:30 ok Oct 20 17:20:02 lindi-: I've just found possible problem in usbhost.patch.. I'll test it and push patch Oct 20 17:21:01 JaMa: yeah got sound working :) Oct 20 17:21:17 it's indeed snd_soc_neo1973_wm8753 while it used to be called snd-soc-neo1973-gta02-wm8753 Oct 20 17:23:46 lindi-: yes after f5c4ffbd65892829f7ec503a89ea24eb0fc952dc ASoC: Samsung: Merge neo1937_wm8753 and neo1973_gta02_wm8753 sound board driver Oct 20 17:26:56 JaMa: ok Oct 20 17:27:36 JaMa: so it's only battery current_now, battery temp and usb hostmode that don't work Oct 20 17:28:04 freesmartphone.org: 03mok 07cornucopia * rc98d3121f9d6 10/fsodeviced/src/plugins/gta04_powercontrol/plugin.vala: Oct 20 17:28:04 freesmartphone.org: fsodeviced: fix sysfs path construction in gta04_powercontrol Oct 20 17:28:04 freesmartphone.org: Signed-off-by: Klaus Kurzmann Oct 20 17:30:40 hmm, there does not seem to be a library for using rfkill Oct 20 17:32:24 mrmoku: hmm just checked that here it does create sshd user now, but daemon is still running under root Oct 20 17:35:26 mrmoku: the problem from september http://www.mail-archive.com/openembedded-core@lists.openembedded.org/msg08498.html should be resolved now Oct 20 17:35:54 mrmoku: by http://git.openembedded.org/openembedded-core/commit/?id=6b7f399d595ef58e759dab211f4ece155119a680 Oct 20 17:36:39 but there wasn't PR bump since then.. so maybe you still have version without this change Oct 20 17:36:57 JaMa: hmm... I don't think so as I did a fresh build for armv7 Oct 20 17:37:18 ah strange Oct 20 17:37:24 JaMa: btw. the user gets added when I remove and reinstall openssh-sshd Oct 20 17:37:36 there is groupadd problem, but don't know about current useradd problem Oct 20 17:40:54 aargh, accidentally booted shr-u :) Oct 20 17:41:31 JaMa: hmm... just to be sure I will cleansstate openssh before building the next image Oct 20 17:42:53 lindi-: usbhost fixed.. I'll push rebased repo now.. just take newer usbhost.patch from it Oct 20 17:43:20 JaMa: I can rebase too Oct 20 17:46:57 lindi-: pushed Oct 20 17:56:32 JaMa: "git remote update" says "+ df40067...74c481b shr-2.6.39-nodrm -> shr/shr-2.6.39-nodrm (forced update)" Oct 20 17:56:39 JaMa: I guess this means it did the rebase? Oct 20 17:57:55 yes usbpatch was rebased as I said Oct 20 17:58:39 lindi-: here is the diff if you want http://paste.pocoo.org/show/495683/ Oct 20 18:00:17 SHR: 03Martin.Jansa 07meta-smartphone * r0486af3cccb1 10/meta-openmoko/recipes-kernel/linux/ (linux-2.6.39/shr.patch linux-openmoko_2.6.39.bb): meta-openmoko: linux-openmoko: fix usbhost Oct 20 18:11:13 nano keyboard back online :) [ 639.985000] usb 1-2: Product: SCISSORS Keyboard Oct 20 18:16:11 damn Oct 20 18:16:24 I get no /dev/ttyHS0-3 :/ Oct 20 18:19:01 what's that? Oct 20 18:21:53 JaMa: devices for the gta04 modem Oct 20 18:22:04 probably udev would create them :-P Oct 20 18:25:06 mrmoku, what kernel version? Oct 20 18:25:12 because the name changed Oct 20 18:25:22 between unofficial kernel and mainline Oct 20 18:26:02 like ttyON Oct 20 18:26:05 to something else Oct 20 18:26:15 but here it's ttyHS Oct 20 18:26:25 that means the high speed serial from the omap.... Oct 20 18:26:34 no idea how it's called now Oct 20 18:27:17 GNUtoo|laptop: you're talking about omap... but I'm talking about the hso module Oct 20 18:27:35 ok Oct 20 18:27:54 maybe read its sources? Oct 20 18:27:58 do you use devtmpfs? Oct 20 18:28:02 or mdev or udev? Oct 20 18:29:49 GNUtoo|laptop: hmm... /sys/dev/char should show it in any case... with or without udev, right? Oct 20 18:29:55 it does not though Oct 20 18:30:05 nothing changes there when I load the module Oct 20 18:30:08 (or unload it) Oct 20 18:30:12 hmmm Oct 20 18:32:01 ahh... rfkill maybe Oct 20 18:32:12 rfkill on serial? lol Oct 20 18:32:18 possible tough Oct 20 18:32:40 hmm Oct 20 18:33:04 Depends on: NETDEVICES [=y] && NET [=y] && USB [=y] && RFKILL [=m] Oct 20 18:33:10 that's why I thought rfkill Oct 20 18:33:40 ok Oct 20 18:33:50 everything is possible in code tough Oct 20 18:33:57 that's why I said that Oct 20 18:37:06 angelox|laptop: heyho Oct 20 18:37:22 GNUtoo|laptop: it seems like the modem in the nexuss is always registered with the network Oct 20 18:37:36 hi Oct 20 18:37:39 strange Oct 20 18:37:53 even with no SIM? Oct 20 18:38:07 morphis, can I work a bit on om-gta02? Oct 20 18:38:21 GNUtoo|laptop: yes Oct 20 18:38:30 GNUtoo|laptop: why not, it's your choice :) Oct 20 18:38:47 because I've tried to put a spinlock_irqsave and it said scheduling while atomic Oct 20 18:39:05 so I'll have to send a mail to the wireless kenrel ml Oct 20 18:39:58 ok Oct 20 18:40:02 no problem for em :) Oct 20 18:40:07 s/em/me/ Oct 20 18:40:07 morphis meant: no problme for em :) Oct 20 18:40:08 morphis: hi! Oct 20 18:41:13 morphis: Nokia didn't want to exchange the N8 by a N9 :( Oct 20 18:41:21 ok Oct 20 18:43:32 angelox|laptop, is getting your money back an option? Oct 20 18:44:04 GNUtoo|laptop: nope..they said "We can't do anything if you did open your phone box" Oct 20 18:47:44 hm Oct 20 18:47:47 thats bad Oct 20 18:47:57 then get another phone :) Oct 20 18:48:17 lol Oct 20 18:48:27 phones are expensive in Brazil Oct 20 18:48:36 really? and on ebay? Oct 20 18:48:39 s/phones/everything/ Oct 20 18:48:39 :) Oct 20 18:48:40 angelox|laptop meant: everything are expensive in Brazil Oct 20 18:48:42 hmm... Oct 20 18:48:56 I never been in brazil :) Oct 20 18:49:03 yes it's like twice the price Oct 20 18:49:13 wow Oct 20 18:49:18 exactly Oct 20 18:49:19 thats a lot Oct 20 18:49:21 like a 500$ phone costs 1000$ Oct 20 18:49:25 and on ebay the same? Oct 20 18:49:44 PaulFertser: gah... we have gpsd 2.95... which is 15 months old :-( Oct 20 18:49:54 no device-hook being executed :/ Oct 20 18:50:05 taxes makes ebay bad too for buying stuff Oct 20 18:50:15 mrmoku: time to upgrade then ;) Oct 20 18:50:31 * mrmoku checks what oe-core has to offer... Oct 20 18:51:06 mrmoku: nothing :/ Oct 20 18:51:06 hmm... guess I have to do that then Oct 20 18:51:10 yeah :) Oct 20 18:52:07 i didn't upgrade it when I was moving stuff from oe-classic Oct 20 18:55:13 angelox|laptop: hm, you have a om-gta02? Oct 20 18:55:31 morphis: nope Oct 20 18:55:48 hm Oct 20 18:56:02 a palm pre is even too much on ebay? Oct 20 18:56:40 which model? Oct 20 18:57:09 palm pre 2 or palm pre plus? (are them the same thing)? Oct 20 18:57:17 no Oct 20 18:57:27 palm pre 2 has a better cpu + case Oct 20 18:57:48 pre plus is better pre (which means more ram) Oct 20 18:58:08 but if you can get a pre or pre plus both are great devices Oct 20 18:59:23 552 dollars a palm pre plus Oct 20 18:59:31 angelox|laptop, http://wiki.freesmartphone.org/index.php/HardwareComparison should have the hardware details Oct 20 18:59:36 100$ a pre 3. Oct 20 18:59:45 at MercadoLivre (Brazillian's ebay copy) Oct 20 19:00:22 GNUtoo|laptop: i didn't see that there were comparason between different model versions; Thanks Oct 20 19:00:45 np Oct 20 19:01:34 btw I've no idea of the status of the various ezx Oct 20 19:01:49 you have an ezx, what's wrong with it? Oct 20 19:02:09 It's a poor hardware i guess Oct 20 19:02:12 should be faster than the freerunner still Oct 20 19:02:24 maybe no wifi on your particular model Oct 20 19:02:31 if no wifi that's not great Oct 20 19:02:41 hasn't wifi.. Oct 20 19:02:58 the calls speaker is broken.. Oct 20 19:03:50 JaMa: hello i have a configure problem, please see: http://shr-project.org/trac/ticket/1517 Oct 20 19:04:34 nschle85, hi Oct 20 19:04:40 GNUtoo|laptop: hi Oct 20 19:05:27 I cannot do every devices at once, I'm sorry, so I wonder what do do for the n900 Oct 20 19:05:34 we need proprer audio during calls Oct 20 19:05:36 and sms Oct 20 19:06:34 *what to do Oct 20 19:06:43 we must find a solution to get n900 working fine Oct 20 19:06:49 because I really care about n900 Oct 20 19:07:06 nschle85: your subversion or elementary checkout is too old Oct 20 19:07:30 nschle85: http://wiki.shr-project.org/trac/wiki/Building%20SHR#subversion1.7inshr-chroot Oct 20 19:12:03 JaMa: i started a clean build this morning , so do you mean the build host svn needs an upgrade too ? Oct 20 19:12:29 yes Oct 20 19:15:13 btw I fixed I typo in bq27k-details2 Oct 20 19:15:50 DocScrutinizer: same link? Oct 20 19:15:51 JaMa: but i thought bitbake builds a newer svn-native which is used later ? is it true ? Oct 20 19:16:07 no it's using svn from host Oct 20 19:16:50 JaMa: building a svn-native like git would solve the problem too ? Oct 20 19:17:17 maybe yes Oct 20 19:17:43 not sure if fetch2 solved this but with older bitbake it wasn't using svn-native Oct 20 19:18:12 well we even didn't have subversion-native to use :) Oct 20 19:18:23 JaMa: yup Oct 20 19:19:15 JaMa: calculation of plaintext AP Average Power been wrong (c&p error) Oct 20 19:19:52 been a dup of SAE above Oct 20 19:20:06 JaMa: In my opinion it would be nice to have subversion-native this would minimize the dependcy to the host a little bit. Oct 20 19:20:41 nschle85: agreed Oct 20 19:24:40 JaMa: is it useful to add a sanity check for svn version number not to frustrate other developers ? Oct 20 19:25:37 yes, would be nice Oct 20 19:26:02 but all shr-chroot users are fine :) Oct 20 19:26:13 if they update Oct 20 19:27:00 JaMa: i love your Makefile and i like software which compiles on several platforms :-) Oct 20 19:27:47 JaMa: so i need to update my Ubuntu LTS :-) Oct 20 19:32:07 JaMa: ok ill close the bug as "invalid" ? Oct 20 19:34:25 hmmm svn issues Oct 20 19:34:38 | Requested 'elementary >= 0.7.0.52890' but version of elementary is 0.7.0.0 Oct 20 19:34:51 I know there is a fix pending Oct 20 19:34:57 I saw it maybe on irc or ml Oct 20 19:36:09 GNUtoo|laptop: what will the fix do ? use subversion-native ? Oct 20 19:36:15 no idea Oct 20 19:36:27 I just remember people talking about that Oct 20 19:36:58 GNUtoo|laptop, hi, did you have time to push mplayer to oe-core? :) Oct 20 19:37:14 no I didn't finish it yet Oct 20 19:38:09 GNUtoo|laptop: JaMa: ok please leave #1517 open until a decision is made 1. add sanity check for svn version or 2. use subversion-native for builds Oct 20 19:38:44 GNUtoo|laptop: do you have the problem too ? Oct 20 19:38:45 * JaMa working on sanity check Oct 20 19:39:09 pespin, instead I tried the missing dependency of enjoy Oct 20 19:39:14 (lightmediascanner) Oct 20 19:39:28 GNUtoo|laptop, that should be pushed already right? Oct 20 19:39:33 I sent the patch at least... Oct 20 19:39:40 or JaMa resent it Oct 20 19:39:51 yes I saw Oct 20 19:40:06 what should I do for python-elmentary Oct 20 19:40:16 JaMa: hmm... does oe support scons? Oct 20 19:40:26 looks like gpsd switched build system Oct 20 19:40:40 lol ouch Oct 20 19:40:52 waf scons etc... are hmmm Oct 20 19:40:53 mrmoku: i hadf the same problem Oct 20 19:41:06 there is scons.bbclass Oct 20 19:41:10 mrmoku: scons seems not to supported Oct 20 19:41:30 maybe it was in oe-dev? Oct 20 19:41:40 check oe-dev too Oct 20 19:41:52 midori seem to use waf btw Oct 20 19:42:11 nschle85: hmm... lets see :) Oct 20 19:42:28 GNUtoo|laptop: jama told me that you need at least subversion 1.7.x ? Oct 20 19:42:35 on host? Oct 20 19:42:42 GNUtoo|laptop: yes Oct 20 19:43:07 GNUtoo|laptop: there is not bootstrapping build of subversion-native Oct 20 19:43:15 ouch Oct 20 19:43:49 GNUtoo|laptop: for git its done but for subversion always the host version is used Oct 20 19:44:03 | /OE/shr-core/tmp/sysroots/x86_64-linux/usr/include/curses.h:41:23: fatal error: curses-32.h: No such file or directory Oct 20 19:44:17 Version: 1.6.12dfsg-4ubuntu2.1 Oct 20 19:44:24 that's my subversion version Oct 20 19:44:42 ii subversion 1.6.12dfsg-4ubuntu2.1 Advanced version control system Oct 20 19:44:49 nschle85: can you try http://paste.pocoo.org/show/495736/ ? Oct 20 19:44:54 GNUtoo|laptop: welcome in the club :-) Oct 20 19:45:08 it means that I cannot build anymore Oct 20 19:45:13 I cannot upgrade svn.... Oct 20 19:45:22 trisquel is at the last version Oct 20 19:45:56 mrmoku: do not really remember much about scons Oct 20 19:46:12 it starts building fine Oct 20 19:46:15 JaMa: ok ill try Oct 20 19:46:17 no idea if correct though Oct 20 19:46:28 | This is a Gentoo system. Oct 20 19:46:36 | prefix = /usr (default /usr/local): installation directory prefix Oct 20 19:46:42 | Adjust your PYTHONPATH to see library directories under /usr/local/lib Oct 20 19:46:51 looks all not nice Oct 20 19:47:55 -I/OE/shr-core/tmp/sysroots/x86_64-linux/usr/include Oct 20 19:48:01 is wrong too I think Oct 20 19:49:50 mrmoku: looks like -native build :/ Oct 20 19:51:12 yeah :/ Oct 20 19:52:02 JaMa, will you do a subversion-native or should someone else do it? Oct 20 19:54:09 * JaMa already working on it Oct 20 19:54:15 ;( Oct 20 19:54:17 err Oct 20 19:54:18 morphis: aurora-keyboard is adjusted only for the Palm pre resolution. There's some problem to use that for this release? Oct 20 19:54:18 :) Oct 20 19:55:02 ok nice Oct 20 19:58:20 gah... don't tell me I have to reinstall the lxde root for gta04 to find out about hso :/ Oct 20 19:58:24 leviathan: ping Oct 20 20:00:15 mrmoku: pong Oct 20 20:00:31 leviathan: hey :) Oct 20 20:00:41 leviathan: I get no /dev/ttyHSx Oct 20 20:00:49 possible Oct 20 20:00:53 under linux-3.1 Oct 20 20:00:54 which makes it difficult to start to work on the modem :-) Oct 20 20:00:55 no Oct 20 20:01:02 I'm back to the old one for now Oct 20 20:01:03 uhh? O_o Oct 20 20:01:08 though we're *not* using udev Oct 20 20:01:27 but /sys/dev/char shows no changes either Oct 20 20:01:33 on module load/unload Oct 20 20:02:01 Oct 20 00:51:59 om-gta04 user.info kernel: [ 9385.963378] hso: drivers/net/usb/hso.c: 1.2 Option Wireless Oct 20 20:02:04 Oct 20 00:52:00 om-gta04 user.info kernel: [ 9385.986694] usbcore: registered new interface driver hso Oct 20 20:02:07 is all I get in dmesg Oct 20 20:02:54 JaMa: no, the build starts normally ? do i have something to do that sanity checks are done again ? Oct 20 20:03:16 leviathan: looking at hso.c that should all be fine Oct 20 20:03:25 any further printks are on error Oct 20 20:03:41 nschle85: nope it should fail Oct 20 20:04:14 i checked the sanity.bb file and the patch was applied Oct 20 20:04:34 sanity.bbclass Oct 20 20:04:46 mrmoku: that's really strange Oct 20 20:05:01 uhm Oct 20 20:05:20 http://lxr.free-electrons.com/source/drivers/net/usb/hso.c#L3334 Oct 20 20:05:28 leviathan: also... it is connected to usb, right? Oct 20 20:05:32 yes Oct 20 20:05:34 is lsusb supposed to show it? Oct 20 20:05:38 yes Oct 20 20:05:41 ouch Oct 20 20:05:42 I think so Oct 20 20:05:46 hmm Oct 20 20:05:52 root@om-gta04:/sys/bus/usb/drivers/hso# lsusb Oct 20 20:05:52 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Oct 20 20:05:52 Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Oct 20 20:06:50 mrmoku: you got gta04? nice! Oct 20 20:06:57 dos1: :-) Oct 20 20:07:08 dos1: I can show it to you... if you come to Essen ;) Oct 20 20:07:28 mrmoku: but well, why does it have "om-" prefix? :D Oct 20 20:07:48 dos1: because goldelico decided to keep the openmoko Oct 20 20:08:05 ok Oct 20 20:08:12 nice Oct 20 20:08:19 mrmoku: hmm Oct 20 20:08:23 mrmoku: yes, out Oct 20 20:08:36 mrmoku: uh, ouch Oct 20 20:08:47 * mrmoku will reinstall the lxde rootfs and check there Oct 20 20:08:59 basically it should initialize Oct 20 20:09:31 except you made a modified build Oct 20 20:09:35 with different config Oct 20 20:09:39 and forgot something Oct 20 20:09:56 Nikolaus testes all the boards Oct 20 20:09:58 is there anything else supposed to show up in lsusb? Oct 20 20:10:01 before shipping Oct 20 20:10:03 yeah... got the test sheet :) Oct 20 20:10:04 hmm Oct 20 20:10:17 no Oct 20 20:10:23 not as far as I know :-) Oct 20 20:10:51 ok, will try the proven working thing then... lxde Oct 20 20:12:48 When did the Free Masons contact you? Oct 20 20:12:54 uhh Oct 20 20:12:57 wrong windows Oct 20 20:12:58 :-) Oct 20 20:13:15 I do psychological stuff Oct 20 20:13:17 :-) Oct 20 20:13:58 someone got written to by some idiots Oct 20 20:14:06 -.- Oct 20 20:14:14 and he thinks it's serious Oct 20 20:14:23 while they don't use facebook, as far as I know Oct 20 20:14:30 and NOT some fantasy names Oct 20 20:14:41 which are obviously not their own Oct 20 20:14:54 and they don't assemble in RPG groups -.- Oct 20 20:14:57 so Oct 20 20:15:02 sry Oct 20 20:15:05 damn focus Oct 20 20:15:47 mrmoku: back to OUR topic Oct 20 20:15:59 reinstall the lxderootfs Oct 20 20:16:02 unpacking squeeze-lxde... brb Oct 20 20:16:05 kk Oct 20 20:17:21 dos1: did you see my flamewar with RMS? xD Oct 20 20:17:25 RL flamewar Oct 20 20:17:26 :-) Oct 20 20:17:35 leviathan: no? Oct 20 20:18:47 dos1: ok Oct 20 20:19:14 dos1: I had an RL flamewar with him during his lecture here in Zürich Oct 20 20:19:23 leviathan: i'm not really up-to-date with everything recently ;x Oct 20 20:19:25 about the GTA04 >_< Oct 20 20:19:29 leviathan: oh Oct 20 20:19:43 http://blog.the-leviathan.ch/ Oct 20 20:20:44 latest entry Oct 20 20:20:52 gives you a nice impression about how bad it went -.- Oct 20 20:21:04 leviathan: did he argumented his opinion somehow? Oct 20 20:21:14 I know rms opinion Oct 20 20:21:21 it's complicated to understand tough Oct 20 20:21:26 yup Oct 20 20:21:54 basically we should differentiate hardware and software Oct 20 20:22:01 leviathan: "As a result of this discussion FSF still refuses to call our project a “free project”, which PISSES ME ON Rich, because it’s just stupid!" - i agree... Oct 20 20:22:09 all software must be free but for the hardware it's pointless: Oct 20 20:22:32 for the hardware even if you have the source code of that CPU, you can't be sure it's the right source code Oct 20 20:22:44 and to recompile it you must spend billions of dollars Oct 20 20:22:58 roughly you don't own a CPU factory Oct 20 20:23:28 then where to draw the line between hardware and software Oct 20 20:23:40 because your washing machine may be hardware or software? Oct 20 20:23:43 what about a phone? Oct 20 20:23:47 or a feature phone? Oct 20 20:24:02 well rms drawn the line at the installation of software Oct 20 20:24:20 then following this logic gives the kind of result you know about Oct 20 20:24:59 that is to say if it's hardware you don't really need its source code etc.... Oct 20 20:25:08 so applicate that to the wifi firmware Oct 20 20:25:13 firmware is software Oct 20 20:25:20 but if it's not installed it's hardware Oct 20 20:25:47 altough I think the best solution is to free that firmware Oct 20 20:26:02 many people are interested and a project started then abandonned: Oct 20 20:26:12 *freedombox people are interested Oct 20 20:26:27 *the olpc project started then abandonned misteriously Oct 20 20:26:52 *some people have docs on the libertas capable chips(I wonder where to find such docs) Oct 20 20:26:57 there is free project, so what's the problem? Oct 20 20:27:13 free software is not only about being sure of what you run, there is also the whole knowledge transmission issue Oct 20 20:27:33 is non-free firmware distributed with the phone? Oct 20 20:28:18 dos1, I think so Oct 20 20:28:35 altough in my opinion the best way would be to merge libertas-tf-sdio in mainline Oct 20 20:28:41 and get a free firmware for that Oct 20 20:28:53 yup Oct 20 20:29:17 JaMa: i started a fresh build but sanity check will not fail :-( Oct 20 20:29:25 anyway Oct 20 20:29:26 * mrmoku searching his guitar pick to attach the serial cable Oct 20 20:29:45 if non-free firmware will be "soldered" with this proposed chip Oct 20 20:29:52 then it's really bad solution... Oct 20 20:30:20 as it makes development of that free firmware harder Oct 20 20:31:26 nschle85: sorry I have read your reply but I'm trying to get subversion-native working instead of that check Oct 20 20:31:30 now Oct 20 20:32:03 Imagine a free firmware available after units were shipped with the non-free firmware "soldered": user would be stuck with the non-free firmware while a free one could be used instead Oct 20 20:32:35 even though according to RMS if it's soldered, it doesn't need to be free Oct 20 20:32:44 it's still the same piece of code Oct 20 20:33:11 paulk: yup Oct 20 20:33:32 JaMa: ok if you have a solution i ll start a fresh build on my build server if you want it (if you want i can download a patch before build) Oct 20 20:33:42 paulk: and - if firmware is easly replacable and revertable, then development of free version is MUCH easier and more developers may try to work on it Oct 20 20:33:51 of course Oct 20 20:35:04 so soldering it *will* harm free firmware project Oct 20 20:35:09 indeed.... Oct 20 20:35:49 basically I think rms is wrong on one thing: the device lifetime Oct 20 20:36:06 can anybody explain me why closed firmware is bad ? what is the difference between a closed firmware and a closed chip ? Oct 20 20:36:08 he assume that reversing the firmware would take longer than the device lifetime Oct 20 20:36:12 also, I think it's a bit of nonsense to say: it's ok to restrict people from the right to load another firmware because if we do it that way, it doesn't need to be free… then if android phones have bootloaders you can't rewrite, they don't need to be free? where am I wrong? Oct 20 20:36:57 paulk, for that it must be the hardware that prevent to rewrite the bootloader Oct 20 20:37:33 GNUtoo|laptop, did he mention that that "soldered" chip would be a one-time writable one? Oct 20 20:37:40 no Oct 20 20:37:45 then it's the same Oct 20 20:37:49 no hardware restriction Oct 20 20:38:07 ah unless you mean it can't be flashed from the SoC Oct 20 20:39:10 then the hardware doesn't permit to easily reflash it, so it's no more software? Oct 20 20:39:42 then what's the difference between the freerunner wifi chip and the rms solution Oct 20 20:39:52 or even the modem of the freerunner Oct 20 20:40:11 it loads its code from nand Oct 20 20:40:13 leviathan: gta04:~# ls /dev/ttyHS Oct 20 20:40:13 ttyHS0 ttyHS1 ttyHS2 ttyHS3 ttyHS4 ttyHS5 Oct 20 20:40:15 or something like taht Oct 20 20:40:26 there is a bootloader inside the modem Oct 20 20:40:32 gta04:~# lsusb Oct 20 20:40:32 Bus 002 Device 002: ID 0af0:8800 Option Oct 20 20:40:34 you can even load stuff in it from serial Oct 20 20:40:38 like that for instance: Oct 20 20:40:41 yes… Oct 20 20:40:48 http://bb.osmocom.org/trac/wiki/OpenMoko Oct 20 20:41:15 so RMS solution is different in the fact that it's not connected to the SoC Oct 20 20:41:43 rms may have tough that the microcontroller loading the firmware was hardawre Oct 20 20:41:51 because you don't usually install software in it Oct 20 20:42:01 ok Oct 20 20:43:00 I see the difference Oct 20 20:44:33 anyway it's hard for me to think it's ok to consider some piece of code as software in a case and hardware in another, while it's exactly the same piece of code Oct 20 20:45:13 leviathan: [ 13.271820] driver: '2-2:1.0': driver_bound: bound to device 'hso' Oct 20 20:45:19 I don't have that in SHR Oct 20 20:45:25 gotta leave, bye! Oct 20 20:46:32 mrmoku: are you seriously working on scons support in shr-core ? Oct 20 20:47:31 nschle85: no Oct 20 20:47:47 nschle85: 21:48 * JaMa already working on it Oct 20 20:48:33 aaah :) Oct 20 20:48:37 mrmoku: hmm Oct 20 20:48:38 okee Oct 20 20:48:39 I meant subverion-native Oct 20 20:48:45 check the init scripts Oct 20 20:48:46 JaMa: ahh... hehe :P Oct 20 20:49:06 nschle85: yup, will seriously have to work on it then :) Oct 20 20:49:09 * JaMa hates apr* now Oct 20 20:49:22 JaMa: yes this i also understood Oct 20 20:49:25 * mrmoku needs a newer gpsd Oct 20 20:49:40 yes, this i also understood Oct 20 20:49:59 test Oct 20 20:50:14 is there robin? Oct 20 20:50:33 mrmoku: I see Oct 20 20:50:39 modprobe hso Oct 20 20:50:48 then go to /sys/class/tty Oct 20 20:51:25 for i in ttyHS*; do echo "/dev/$i: $(cat $i/hsotype)"; done Oct 20 20:51:59 if there is something Oct 20 20:52:00 hmm Oct 20 20:52:03 ok Oct 20 20:52:05 on the otherhand Oct 20 20:52:07 I should sleep Oct 20 20:52:15 because it's pointless what I was suggesting Oct 20 20:52:17 hmm Oct 20 20:52:18 mom Oct 20 20:52:43 mrmoku: why do you need a newer gpsd ? Oct 20 20:53:45 mrmoku: basically it should work Oct 20 20:54:07 the only thing I found reads out ttyHS* data from sys Oct 20 20:54:12 and I'll go sleep Oct 20 20:54:13 gn8 Oct 20 20:54:16 nschle85: for the device-activation hook Oct 20 20:54:20 gn8 leviathan Oct 20 20:54:24 thx Oct 20 20:54:25 :-) Oct 20 20:54:34 at least my hw is fine :) Oct 20 20:54:50 mrmoku: ahh this is what PaulFertser. added ? Oct 20 20:54:53 yup Oct 20 20:55:47 i wanted a newer version because i wanted to integrate N900 gps support Oct 20 20:56:07 but now i decided to use 2.95 Oct 20 20:56:10 nschle85: I will look into scons tomorrow Oct 20 20:56:38 but if scons will work no work is loose because a had low time to work on it Oct 20 20:56:59 so i would update to gpsd3.xx Oct 20 20:57:31 good Oct 20 21:04:43 mrmoku: hope we can meet for some hours with GNUtoo|laptop: in munich in december Oct 20 21:05:06 yes I hope too Oct 20 21:05:33 I'll try to book all etc...as soon as possible Oct 20 21:06:36 GNUtoo|laptop: i think i cannot come to essen because its the last weekend before the stressing time Oct 20 21:06:52 ok Oct 20 21:09:48 GNUtoo|laptop: different question: Oct 20 21:10:25 GNUtoo|laptop: is it possible to buy a prepaid sim card in italy with internet flat ? Oct 20 21:10:30 yes Oct 20 21:10:45 with wind internet is 1E/day without activation Oct 20 21:11:01 it's less by day if you take longer + activation Oct 20 21:12:06 GNUtoo|laptop: we will stay in italy for 5 days what does activation mean ? Oct 20 21:12:17 and where can i buy such card ? Oct 20 21:12:49 nschle85, I think you need some papers like ID card for such cards Oct 20 21:13:07 activation means: Oct 20 21:13:31 for instance for one week it costs bar to activate the option + foo/week Oct 20 21:13:39 GNUtoo|laptop: i dont have any papers i am illegal in europe :-) Oct 20 21:14:56 GNUtoo|laptop: i was born in GDR but this country does not exists till 1990 Oct 20 21:15:13 ah? Oct 20 21:15:36 GNUtoo|laptop: ok thats german history Oct 20 21:16:26 ok Oct 20 21:16:36 it's the acronym in english that I didn't know Oct 20 21:16:50 wikipedia helped.... Oct 20 21:18:41 NOTE: package subversion-native-1.7.0-r0: task do_rm_work_all: Succeeded Oct 20 21:18:46 nschle85: ^ ready to test? Oct 20 21:19:06 I'm ready to test too Oct 20 21:19:12 can you push to the shr branch Oct 20 21:19:12 JaMa: ok what do i have to do ? Oct 20 21:19:34 GNUtoo|laptop: JaMa: no Oct 20 21:20:14 GNUtoo|laptop: JaMa: pleaso do not push anything which may fail Oct 20 21:20:45 GNUtoo|laptop: JaMa: developers should test as best as they can Oct 20 21:21:08 so we should not destroy shr branch Oct 20 21:21:23 then maybe a patch that I can apply Oct 20 21:21:26 like git am Oct 20 21:21:40 JaMa: GNUtoo|laptop: i agree Oct 20 21:22:32 nschle85: http://paste.pocoo.org/show/495776/ Oct 20 21:23:30 nschle85: GNUtoo|laptop: or cherry-pick it from contrib/jansa/test Oct 20 21:23:40 ok Oct 20 21:24:49 task-shr-feed built for all 4 devices on SHR buildhost :) finally! Oct 20 21:24:52 JaMa: is it enough to apply the patch and rebuild or do i have to clean something ? Oct 20 21:25:06 nschle85: you have to build subversion-native manually Oct 20 21:25:31 nschle85: and there is newer sanity check (which is not needed now, but if you can please confirm it fails with it applied now) Oct 20 21:26:30 JaMa: ok1. ill check the sanity check (please do not apply if svn-native works) Oct 20 21:26:45 where is new sanity check patch ? Oct 20 21:26:56 also contrib/jansa/test HEAD -2 Oct 20 21:27:03 <[Rui]> hi all Oct 20 21:27:12 hi Rui Oct 20 21:27:29 * [Rui] hopes to have time to move to shr-core this weekend Oct 20 21:27:34 [Rui], hi there :) Oct 20 21:28:20 [Rui], hi I used the gta02 with shr-core for a while Oct 20 21:28:27 and I still uses it Oct 20 21:29:15 so I think it's usable if you kill shr_elm_softkey and launch it again when it crashes Oct 20 21:29:28 <[Rui]> GNUtoo|laptop, yeah, you told me it works, I just need vacations from the baby ;) Oct 20 21:29:28 altough usable is still subjective as usual Oct 20 21:29:43 when will that happen? Oct 20 21:29:50 <[Rui]> GNUtoo|laptop, if it's as usable as shr-unstable, then great :) Oct 20 21:29:50 because it's really annoying Oct 20 21:29:58 if it happens when there is a call..... Oct 20 21:30:00 I miss the call Oct 20 21:30:11 no idea what you mean by as-usable Oct 20 21:30:17 it's usable at least Oct 20 21:30:17 <[Rui]> sorry :( really hoping to jump into core this weekend :| Oct 20 21:30:53 JaMa: please send me the patch (based on shr branch) Oct 20 21:30:53 if it's this week-end it's nice Oct 20 21:31:11 nschle85, he already did: http://paste.pocoo.org/show/495776/ Oct 20 21:31:27 but it didn't work for me Oct 20 21:31:34 nschle85: git cherry-pick db199d968b7308eab1504cb017e7cd131b796235 should work for you Oct 20 21:31:52 nschle85: maybe do git remote update first Oct 20 21:31:57 it seems that svn-native is not dragged as a dependency Oct 20 21:32:05 I'll bitbake it alone Oct 20 21:32:17 it's not 23:20:32 < JaMa> nschle85: you have to build subversion-native manually Oct 20 21:32:38 ok thanks Oct 20 21:32:47 I'm not even sure that bitbake fetches have STAGING_BINDIR_NATIVE before system subversion in PATH.. Oct 20 21:33:06 I'll try Oct 20 21:33:08 thanks a lot Oct 20 21:33:27 JaMa: stop, first i want to test failing sanity check Oct 20 21:33:29 as work-around you can put full path to FETCHCMD_svn Oct 20 21:33:58 nschle85: continue, with git commands I've written here :) Oct 20 21:41:03 still failing Oct 20 21:41:18 JaMa: unfortunately it seems resume does not work here for 2.6.39 Oct 20 21:41:29 JaMa: it shows blinking cursor until watchdog resets the system Oct 20 21:42:52 I'll retry to reset the env Oct 20 21:43:22 GNUtoo|laptop: what is failing? Oct 20 21:43:35 | Requested 'elementary >= 0.7.0.52890' but version of elementary is 0.7.0.0 Oct 20 21:43:39 JaMa, I'm getting libc errors while using shr-chroot :S Oct 20 21:43:41 wget: res_query.c:251: __libc_res_nquery: Assertion `hp != hp2' failed. Oct 20 21:43:43 GNUtoo|laptop: did you rebuild elementary? Oct 20 21:43:45 I got it using git too Oct 20 21:44:02 ah I have to? Oct 20 21:44:02 pespin_: did you update today or yesterday? Oct 20 21:44:06 I'll do that Oct 20 21:44:07 JaMa, now Oct 20 21:44:14 I rebuilt python-elmentary only sorry Oct 20 21:44:16 GNUtoo|laptop: yes.. it's wrong in elementary.pc Oct 20 21:44:31 pespin_: I haven't seen this here or on SHR buildhost :/ Oct 20 21:44:38 ah ok Oct 20 21:45:02 JaMa, ok, maybe some corruption when my PC got a kernel panic while building Oct 20 21:45:21 I usually have kernel panics lately (since using nvidia twinview I think) Oct 20 21:46:46 GNUtoo|laptop: OE om-gta02@shr ~/shr-core $ cat tmp/sysroots/om-gta02/usr/lib/pkgconfig/elementary.pc | grep Version Oct 20 21:46:50 Version: 0.7.0.64150 Oct 20 21:47:19 GNUtoo|laptop: but subversion-native built fine? Oct 20 21:47:32 GNUtoo|laptop: if yes I'll push it to shr branch as it's optional anyway Oct 20 21:47:42 JaMa: OE om-gta02@shr ~/build-core-shr/shr-core/openembedded-core $ git cherry-pick db199d968b7308eab1504cb017e7cd131b796235 [shr 999b8c3] sanity: add check for subversion >= 1.7.0 Author: Martin Jansa 1 files changed, 6 insertions(+), 0 deletions(-) OE om-gta02@shr ~/build-core-shr/shr-core/openembedded-core $ cd .. OE om-gta02@shr ~/build-core-shr/shr-core $ bitbake sh Oct 20 21:48:13 JaMa: but the build starts and did not stop :-( Oct 20 21:50:35 nschle85: ok.. forgot sanity check and try subversion-native Oct 20 21:50:50 * JaMa is not good with python subprocess Oct 20 21:51:04 omfg, I think I get senile. This lil bit of bitjoggling in bash wore me out now x.x Oct 20 21:51:38 like 4 working lines of bash script / hour Oct 20 21:52:01 JaMa: new version, ststus reg bit decoded now Oct 20 21:52:16 prlly quite useful Oct 20 21:53:31 JaMa: ok ill modify my build server build process... to test git-native Oct 20 21:54:17 subversion-native, right? Oct 20 21:54:27 DocScrutinizer: thx Oct 20 21:54:49 JaMa: subversion-native :-) Oct 20 21:55:24 * JaMa switching partitions.. bye bye shr-u :) Oct 20 21:58:11 JaMa: i loved shr-u it was simple and working :-) Oct 20 21:59:11 not worth pressing AUX on every boot to shr-core :) Oct 20 22:02:32 JaMa: there's still something weird going on with audio. lsmod definitely shows snd_soc_neo1973_wm8753 now but "aplay -l" does not show any cards Oct 20 22:02:48 lindi-: you need more modules than that Oct 20 22:03:08 JaMa: and rmmod shows a backtrace Oct 20 22:03:10 lindi-: I don't have it booted now, but IIRC mrmoku sent the list in the afternoon Oct 20 22:03:26 WARNING: at /tmpfs/linux-2.6/kernel/irq/manage.c:1131 __free_irq+0xfc/0x174() Oct 20 22:03:26 Oct 21 00:58:02 ginger kernel: [ 660.550000] Trying to free already-free IRQ 48 Oct 20 22:03:34 NOTE: package shr-lite-image-2.0-r18: task do_rootfs: Started Oct 20 22:03:36 thanks a lot Oct 20 22:03:56 JaMa: to this channel? Oct 20 22:04:44 btw is it possible to operate a gta04 without sacrificing a gta02, for instance as a devboard? Oct 20 22:04:48 lindi-: yes Oct 20 22:05:03 GNUtoo|laptop: yw Oct 20 22:09:04 JaMa: and btw, tested the usbhost fix Oct 20 22:09:25 JaMa, can you try fetching libxml-parser-perl to see if you get the error? Oct 20 22:10:15 pespin_: what type of error Oct 20 22:10:23 pespin_: I've already downloaded it before Oct 20 22:10:32 JaMa, do you have any idea for fixing the shell btw: Oct 20 22:10:36 *home is wrong Oct 20 22:10:42 *no export DISPLAY=:0 Oct 20 22:10:49 wget -t 5 -q --passive-ftp --no-check-certificate -P /OE/shr-core/downloads 'http://www.cpan.org/modules/by-module/XML/XML-Parser-2.41.tar.gz' failed with signal 134, output: Oct 20 22:10:49 | wget: res_query.c:251: __libc_res_nquery: Assertion `hp != hp2' failed. Oct 20 22:10:54 I've looked in vain on om-gta02 Oct 20 22:11:12 like in the subway etc...looking at the possible files Oct 20 22:16:20 pespin_: works fine here Oct 20 22:16:37 JaMa: ok will you have a live picture ? like BBC ? then look here : http://norman-schleicher.de/jenkins/job/shr-core-n900/51/console Oct 20 22:16:43 GNUtoo|laptop: I've already looked too and found only small issue but that wasn't enough Oct 20 22:17:09 ok Oct 20 22:17:13 issues are: Oct 20 22:17:33 *a program that emtooth2 depend on doesn't work without DISPLAY=:0 Oct 20 22:17:38 I think it was obexd Oct 20 22:17:49 *huges issues with HOME Oct 20 22:17:59 nschle85: you can just update now Oct 20 22:18:18 JaMa: mrmoku: what mailing list would be good for me to send information on 2.6.39 issues? Oct 20 22:18:25 hmmm it compiles me pulseaudio when bitbaking enjoy Oct 20 22:18:36 JaMa: ok i just blamed me :-) Oct 20 22:18:43 and of course pulseaudio fails (known to fail for armv4) Oct 20 22:19:00 please have a look :-) at jenkins Oct 20 22:19:08 selected processor does not support Thumb mode `swp r1,r2,[r3]' Oct 20 22:19:15 that sounds very familiar.... Oct 20 22:19:21 GNUtoo|laptop, yeah obexd Oct 20 22:19:32 I guess I need to use cmpxchg.... Oct 20 22:20:30 lindi-: openmoko-kernel I guess but not much traffic there :/ Oct 20 22:20:50 lindi-: as ie nobody replied on my report that reverting Per's patches helps Oct 20 22:22:05 ah it's dragged in by gst-plugins-good Oct 20 22:22:47 GNUtoo|laptop: there are patches for this in jansa/test too Oct 20 22:22:59 ok Oct 20 22:23:03 thanks a lot Oct 20 22:23:07 JaMa: ok now building again http://norman-schleicher.de/jenkins/job/shr-core-n900/52/console Oct 20 22:23:12 so what are the patch doing? Oct 20 22:23:14 cmpxchg? Oct 20 22:23:15 GNUtoo|laptop: mmt I'll push them to shr branch now Oct 20 22:23:23 GNUtoo|laptop: forcing arm mode Oct 20 22:23:28 ok Oct 20 22:23:39 I guess cmpxchg is overkill for that Oct 20 22:27:12 nschle85: did you build subversion-native already? Oct 20 22:27:25 nschle85: and cleansstate elementary Oct 20 22:27:28 JaMa: shit noooooo Oct 20 22:27:38 ok ill do Oct 20 22:28:13 btw there should be elementary release soon, so hopefully this issue will be resolved soon :) Oct 20 22:28:39 I did Oct 20 22:28:41 it works Oct 20 22:30:55 JaMa, the problem is now on target: Oct 20 22:31:24 http://www.pastie.org/private/t1vqjtzogbciypvqmxag Oct 20 22:31:27 will try to fix Oct 20 22:31:41 ah I can't Oct 20 22:31:44 it's not in python Oct 20 22:32:23 JaMa, did I miss something? Oct 20 22:32:24 JaMa: right Oct 20 22:32:43 GNUtoo|laptop: what's not in python? Oct 20 22:33:10 GNUtoo|laptop: I cannot test here now as I have python-2.7 Oct 20 22:33:50 JaMa, on target? Oct 20 22:34:08 JaMa, see the last line, it's some native calls Oct 20 22:35:00 yes on target and we had similar issue in shr-settings before Oct 20 22:35:13 ok Oct 20 22:35:21 don't remember exact patch which was needed in python bindings Oct 20 22:35:25 ok Oct 20 22:36:53 what should I do then? Oct 20 22:37:35 check python-evas log for similar patch Oct 20 22:39:08 ok Oct 20 22:39:10 in oe? Oct 20 22:39:19 I guess so Oct 20 22:39:21 I'll look Oct 20 22:39:42 I guess rebuilding python-evas won't magically solve it(I didn't rebuild that) Oct 20 22:42:45 GNUtoo|laptop: no upstream Oct 20 22:43:44 ah ok Oct 20 22:50:52 SHR: 03Martin.Jansa 07meta-smartphone * rfbd6e9e8517e 10/meta-shr/conf/distro/include/preferred-shr-versions.inc: meta-shr: subversion: prefer 1.7.0 Oct 20 23:06:31 GNUtoo|laptop: I've found other way to reproduce your python-e* issue http://paste.pocoo.org/show/495808/ Oct 20 23:06:45 GNUtoo|laptop: so if you're not able to fix it I'll try tomorrow or on weekend Oct 20 23:06:51 now I'm going to bed Oct 20 23:06:54 gnight all Oct 20 23:07:05 ok Oct 20 23:07:08 JaMa: night and thanks for the kernel tips Oct 20 23:07:11 good night Oct 20 23:07:23 lindi-: btw resume works here fine Oct 20 23:07:40 JaMa: is your rootfs on SD? Oct 20 23:08:42 lindi-: and aplay too http://paste.pocoo.org/show/495811/ Oct 20 23:08:44 yes Oct 20 23:09:03 JaMa: maybe the usbhost fix broke this? ;)) Oct 20 23:09:05 just moved it from p2 to p1 but I was never using NAND Oct 20 23:09:19 I just resumed few times with this patch :) Oct 20 23:09:24 odd Oct 20 23:09:47 but I don't have watchdog in my shr-core images yet Oct 20 23:10:00 but resume is really quick Oct 20 23:10:03 I guess I should try without it first Oct 20 23:10:48 JaMa: hmm you have snd_soc_dfbmcs320 Oct 20 23:10:56 JaMa: is that the bluetooth stuff? Oct 20 23:11:21 modinfo says so Oct 20 23:11:31 but don't really know details Oct 20 23:11:53 hmm, now aplay works **** ENDING LOGGING AT Fri Oct 21 02:59:57 2011