**** BEGIN LOGGING AT Mon Nov 15 02:59:58 2010 Nov 15 06:29:37 JaMa: nice. question is if that depends on the target kernel to be execed - or if booting into maemo should work too Nov 15 06:54:51 mrmoku: maybe it was because I did it "right" with / remounted RO and every service stoped (including ssh/xserver) Nov 15 06:55:36 and I've also used kexec-klibc instead kexec built against eglibc.. will try with same kexec from maemo too Nov 15 07:56:54 morning Nov 15 07:59:31 moin Nov 15 08:08:47 moin Nov 15 08:13:28 morning JaMa Nov 15 08:13:44 mokohome is crashing since latest efl upgrade :( Nov 15 08:19:56 daniele_athome: did you try to rebuild it against newer ecore? Nov 15 08:20:21 JaMa, no i just took the shr package you've rebuilt Nov 15 08:20:33 daniele_athome: ah it's part of mokosuite2? Nov 15 08:20:53 yes JaMa Nov 15 08:20:56 for now :) Nov 15 08:21:05 it's all a (app)bunch of apps :) Nov 15 08:21:12 like the repository name says ;) Nov 15 08:23:02 well I just bumped PR to rebuild it, didn't test it here Nov 15 08:24:09 JaMa, don't worry i'll figure it out ;) Nov 15 08:24:57 I know and I don't worry about it already :) Nov 15 08:26:33 lol Nov 15 09:06:09 mrmoku, GNUtoo: n900-2.6.35 rebased on top of OE (for linux-2.6.35 changes) Nov 15 09:11:06 freesmartphone.org: 03morphis 07msmcomm * r6dc487553685 10/libmsmcomm/ (31 files in 3 dirs): libmsmcomm: rename and add a lot of stuff ... maybe it's unusable now ... Nov 15 09:20:01 JaMa: hmm... maybe we should try to add the DVFS patches from http://dev.omapzoom.org/?p=thara/omap-dvfs.git;a=shortlog;h=refs/heads/pm-dvfs Nov 15 09:26:28 freesmartphone.org: 03morphis 07msmcomm * r7137154f38e7 10/libmsmcomm/ (2 files in 2 dirs): Nov 15 09:26:28 freesmartphone.org: libmsmcomm: fix bug in process-all-hci-messages.py script Nov 15 09:26:28 freesmartphone.org: We have to be carefull with the results from the all_hci_messages files as their order is Nov 15 09:26:28 freesmartphone.org: not the right to 100%. Nov 15 09:26:31 freesmartphone.org: 03morphis 07msmcomm * r16ad655c5a9e 10/libmsmcomm/src/internal.h: libmsmcomm: define more group ids Nov 15 09:28:49 mrmoku: I don't see them merged to meego kernel, do we want to be so bleeding edge? :) Nov 15 09:30:18 freesmartphone.org: 03morphis 07msmcomm * r3a53bd8d8624 10/libmsmcomm/src/ (7 files): libmsmcomm: fix build error in response_audio.c Nov 15 09:30:19 freesmartphone.org: 03morphis 07msmcomm * red19e7325299 10/libmsmcomm/src/Makefile.am: libmsmcomm: add pdsm files to build process Nov 15 09:31:19 JaMa: only if meego kernel is moving too slow :P Nov 15 09:31:36 :) Nov 15 09:37:30 freesmartphone.org: 03morphis 07msmcomm * rce810a98bd3e 10/libmsmcomm/specs/messages/ (7 files): libmsmcomm: add empty files for messages definitions Nov 15 09:53:14 freesmartphone.org: 03morphis 07msmcomm * r97d4d8dd2c5b 10/libmsmcomm/ (4 files in 2 dirs): libmsmcomm: implement some more phonebook events Nov 15 12:17:58 mickey|office: hey :) he's the proof there's something wrong with fsousaged stuck in SUSPEND state: http://paste.debian.net/99879/ Nov 15 12:18:39 PaulFertser: hi :) Nov 15 12:18:50 JaMa: hi Nov 15 12:19:09 JaMa: no news yet, but i'm going to work on it soon. Nov 15 12:20:14 ok :) Nov 15 12:20:31 JaMa: but "Stanger than Paradise" is highly adviced. Nov 15 12:21:56 ok I'll move it higher on my todo :) Nov 15 12:23:56 JaMa: it's not weird enough to be totally incomprehensible and is humourous enough to be enjoyable without bad/sad/sorrow aftertaste. Nov 15 12:26:07 that's why it's way higher on todo then Pierrot_le_fou Nov 15 12:31:38 PaulFertser: I took the power-kernel patches from git Nov 15 12:32:02 there is one which is activated in the quilt series file Nov 15 12:32:12 which does not apply Nov 15 12:32:21 armthumb.diff Nov 15 12:32:37 any opinion on that one? Nov 15 12:36:06 mrmoku: was power-kernel updated to include all fixes which were part of pr1.3 upgrade? Nov 15 12:36:31 mrmoku: ok, this is essential only if you have some thumb code. Nov 15 12:36:53 mrmoku: yesterday I was thinking about switching from default kernel in meamo (pr1.3) but then I wasn't sure about it and went to bed instead Nov 15 12:36:55 JaMa: loks like, yes Nov 15 12:37:02 mrmoku: do you enable thumb interworking? Nov 15 12:37:41 JaMa: yes, the latest power kernel has all the fixes from pr1.3 (and none of the bugs/misfeatures it introduced comparing to pr1.2). Nov 15 12:38:04 PaulFertser: yup, as far as I know we have thumb enabled Nov 15 12:39:04 mrmoku: this patch would probably impact performance of that negatively. Nov 15 12:39:32 so we don't want it anyway? Nov 15 12:40:03 mrmoku: i guess it's better to disable thumb altogether. Nov 15 12:40:33 JaMa: do we have it enabled for n900? Nov 15 12:40:43 mrmoku: to avoid needless flushing of branch prediction state on every context switch. Nov 15 12:41:06 I'm checking distro config but seems like it's disabled for armv6+ Nov 15 12:41:17 ahh, ok Nov 15 12:41:31 then all is fine :) Nov 15 12:42:06 it's enabled by default only for armv4t armv5te armv5teb Nov 15 12:42:42 You might want to ask lindi- whether thumb interworking is a good thing to have at all. Nov 15 12:42:59 As i do not know anything about it. Nov 15 12:44:02 this is old, but I guess that still valid http://www.mail-archive.com/angstrom-distro-devel@linuxtogo.org/msg02585.html Nov 15 12:44:39 JaMa: btw, have you managed to found anything new about that horrible X crash? Nov 15 12:46:15 I've tried to rebuild everything related with -fstack-protector-all Nov 15 12:46:33 but it doesn't fail as it should Nov 15 12:47:20 then I can see how is stack overwritten (in 8th iteration of same function with same params :/) Nov 15 12:47:58 and best part is that I know better in which function it happens Nov 15 12:48:19 but maybe it's still not the real problem and that function just reveals it :/ Nov 15 12:49:07 JaMa: complex stuff indeed Nov 15 12:50:02 it would be easier without every "function" in Xorg being macro which calls lookup function to find real implementation for it :) Nov 15 13:04:48 hi there Nov 15 13:05:30 JaMa, hmmm no one responded at [oe] binconfig.bbclass optional last slash handling Nov 15 13:07:00 PaulFertser: I don't have much experience with thumb really Nov 15 13:09:57 GNUtoo|laptop: and pb didn't respond to my ping (he added that trailing slash) Nov 15 13:10:08 ok Nov 15 13:10:15 let's ping him again then Nov 15 13:10:22 maybe he wasn't there on the week-end Nov 15 13:11:29 lindi-: ok, thanks nevertheless :) Nov 15 13:12:25 PaulFertser: mostly avoided it to keep things simple Nov 15 13:13:11 lindi-: sidenote: on STM32 (cortex-m3) only thumb2 is available, no arm set :) Nov 15 13:13:47 but cortex-m3 is a microcontroller right? Nov 15 13:14:18 PaulFertser: got any theory for the following: Nov 15 13:14:25 GNUtoo|laptop: sure Nov 15 13:15:52 it seem that the thumb discussion is resurfacing Nov 15 13:16:04 PaulFertser: I have an openmoko as an rtk-gps base station. it is connected to gsm network for remote administration but also uses an usb ethernet dongle that says "current draw: 850mW" Nov 15 13:16:04 what's the advantage apart size on armv4? Nov 15 13:17:04 PaulFertser: when I look at the logs I see that every time the GSM says %CPRI: 1,2 the usb ethernet dongle disconnects from the usb bus: http://paste.debian.net/99882/ Nov 15 13:18:45 GNUtoo|laptop: just size (not only in NAND/uSD but also in cache) Nov 15 13:19:01 ok Nov 15 13:19:20 so it's not faster? Nov 15 13:19:29 (because of cache size stuff) Nov 15 13:19:40 were there any benchmarks Nov 15 13:19:58 lindi-: probably the easiest to try would be to connect a large capacitor in parallel to the dongle's power supply. Nov 15 13:20:57 lindi-: what cable do you use to connect to the dongle? is it long and does it have ferrite bead? Nov 15 13:21:10 lindi-: so sorry, no real suggestions except for the obvious... Nov 15 13:21:12 GNUtoo|laptop: yes sometimes it could be faster thanks to size, but I haven't see any benchmarks Nov 15 13:21:23 PaulFertser: the original openmoko(tm) usb cable :P Nov 15 13:21:26 s/could/could (in theory)/ Nov 15 13:21:27 JaMa meant: GNUtoo|laptop: yes sometimes it could (in theory) be faster thanks to size, but I haven't see any benchmarks Nov 15 13:21:50 PaulFertser: 'om usb charger-limit' said 1000. I reduced that to 500 now Nov 15 13:22:08 ok Nov 15 13:23:26 PaulFertser: it's this http://www.dealextreme.com/details.dx/sku.13526 hub. openmoko is connected to the upstream port and wall charger to the red power port Nov 15 13:23:47 lindi-: hm, probably right, if the sudden current draw by GSM makes your charger output less. Nov 15 13:26:00 PaulFertser: I guess 2A wall charger might help? ;) Nov 15 13:27:16 PaulFertser: this %CPRI seems to come exactly every four hours Nov 15 13:27:38 PaulFertser: so I guess I could prepare for it and "om usb charger-limit 0" just before it... Nov 15 13:29:25 lindi-: if that's really the reason... Nov 15 14:18:12 GNUtoo|laptop: Hi! I have added some more informations on the pre 2 Nov 15 14:18:23 ok nice Nov 15 14:18:30 GNUtoo|laptop: by the way, why did you remove the type of CPU for the pre 2? Nov 15 14:18:31 mickey|office: ayt? Would you like to discuss the timeout issue? The only thing i can think of is not sending the stupid "wake up" sequence in time. Is it possible that the time spent in suspend is not counted and on resume you think you were talking with the modem "just a moment ago" so sending the wakeup sequence is not necessary? Nov 15 14:18:36 not usefull here? Nov 15 14:19:03 GarthPS, soc != CPU Nov 15 14:19:13 GNUtoo|laptop: ah yeah sure :p Nov 15 14:19:14 I kept cpu but removed the soc type as it was not relevant Nov 15 14:19:25 it's mentioned at some other places Nov 15 14:20:17 GNUtoo|laptop: yeah . other topic, who work on the pre actualy ? Nov 15 14:20:26 GarthPS, morphis Nov 15 14:20:30 and some other people Nov 15 14:20:39 I don't remember who were the other people Nov 15 14:20:53 GNUtoo|laptop: ok so he is not alone :) Nov 15 14:21:13 no Nov 15 14:21:25 I even helped with wifi Nov 15 14:21:53 GNUtoo|laptop: and so for the pre 2 the only differrent thing to change is the kernel right? Nov 15 14:22:01 mickey|office: i'll try to get a decent strace log now since i have this bug reproducing fairly often. Nov 15 14:22:09 s/is/should be Nov 15 14:22:13 GarthPS, I don't know, I think the kenrel is the same Nov 15 14:22:29 or mostly the same Nov 15 14:23:04 I will check something with morphis before trying shr-u on it Nov 15 14:23:14 yes you should Nov 15 14:25:00 anyway if you guys want me to ding for more infos from my pre 2 , don't hesitate. Nov 15 14:25:04 GarthPS, maybe not necessary if it uses the same machine ID Nov 15 14:25:23 PaulFertser: could you try setting the default timeout to something more reasonable? Nov 15 14:25:35 playya: how can I know that ? Nov 15 14:25:43 it'd be interesting to see e.g. how it behaves with 10 seconds Nov 15 14:25:47 palm did not release the source yet Nov 15 14:25:49 mickey|office: yes, but the timeout is a bad thing on its own. Nov 15 14:26:08 mickey|office: i'll set it to 10 seconds after i'll get my strace result most probably. Nov 15 14:26:08 i'm referring to the modem timeout here Nov 15 14:26:11 dbus timeout is something else Nov 15 14:26:13 GarthPS, do you have a working novaterm setup? Nov 15 14:26:15 playya: ok Nov 15 14:26:41 playya: eu it is the free versionbecause the palm one does not yet include new usb id Nov 15 14:26:50 but yeah I can novaterm or ssh Nov 15 14:27:52 ok Nov 15 14:27:59 mickey|office: i know. Nov 15 14:28:05 could you paste /proc/cpuinfo? Nov 15 14:28:26 mickey|office: but do you remember seeing modem timeouts ever with the old codebase once the "wakeup-wakeup" sequence was properly working? I do not. Nov 15 14:28:46 mickey|office: so the frequency with which i get timeouts is suggesting something's wrong. Nov 15 14:29:28 PaulFertser: do you also get timeouts during normal operation or only near suspend/resume cycles? Nov 15 14:30:18 mickey|office: no logs, no proof, too many nights passed; but i do not remember any during normal operation. Nov 15 14:30:34 ok, in that case your theory could make sense Nov 15 14:30:44 playya: http://www.webos-internals.org/wiki/PalmPre2HardwareAndDriversInformations#cat_.2Fproc.2Fcpuinfo Nov 15 14:31:04 and with the default timeout being 2 minutes, the dbus timeout hits us even if we detect the modem not answering and retry sending the command :D Nov 15 14:31:13 mickey|office: sure :) Nov 15 14:31:54 mickey|office, it seem that FSO is frozen when in suspend state Nov 15 14:32:03 I mean it doesn't accept commands Nov 15 14:32:05 if that helps Nov 15 14:32:15 playya: (oh , do know how to use my laptop connection with the pre by its usbnet? I am already doing all needed things from the laptop side Nov 15 14:32:27 GarthPS, ok. i think our kernel shouldn't work :( Nov 15 14:32:30 mickey|office: btw, i filed a ticket about fsousaged, it seems to be stuck forever in the wrong state, not sure if that's enough information, i haven't tried digging anything about it: http://trac.freesmartphone.org/ticket/594 Nov 15 14:32:51 I've also the impression that under htcdream sms(but not calls) comes when the phone goes out of suspend Nov 15 14:32:57 GarthPS, no. but there's a free hotspot app in the app catalog Nov 15 14:33:01 playya: .. :/ Nov 15 14:33:11 htcdream is interesting because you can interact with the phone while in suspend mode Nov 15 14:33:39 playya: yah but I want tne other way too Nov 15 14:33:48 morphis just connected Nov 15 14:34:29 playya: not 3g>pre2>wifi>laptop but laptop>usbnet>pre2 as I was doing with the Freerunner Nov 15 14:34:58 playya: about the kernel does this mesn a huge amount of work ? Nov 15 14:36:16 we need the ID set by booty to detect the correct machine in the kernel Nov 15 14:36:54 this is what I'm doing on my laptop to share the connection with the freerunner: iptables -A POSTROUTING -t nat -s 192.168.0.0/24 -j MASQUERADE Nov 15 14:36:54 up echo 1 > /proc/sys/net/ipv4/ip_forward Nov 15 14:37:10 s/^\s*up// Nov 15 14:37:42 what does your last line ? Nov 15 14:37:57 the 2 first lines I did already Nov 15 14:38:28 playya: sorry, ping. Nov 15 14:38:34 the last line was to remove the up in the second line Nov 15 14:38:49 playya: ah , k; yeah I was wondering Nov 15 14:39:20 are you able to ping the pre over usbnet? Nov 15 14:39:59 playya: but the pre 2 still does not use this interface for internet. I think it miss something from its side Nov 15 14:40:11 did you setup a default route and dns on you laptop? Nov 15 14:40:41 playya: I think it is ok as it works for th FR Nov 15 14:40:54 hmm. ok Nov 15 14:41:35 so the kernel ,the simpliest is to way their release right ? Nov 15 14:42:05 yes Nov 15 14:42:14 s/way/wait Nov 15 14:42:47 ok. it should not be to long anymore no? did they anouce an ETA? Nov 15 14:43:09 afaik no Nov 15 14:43:40 could you paste the result of gunzip < /proc/config.gz , too? Nov 15 14:44:44 bbl. need to got to the grocery Nov 15 14:47:02 freesmartphone.org: 03morphis 07msmcomm * ra6b10b47c781 10/libmsmcomm/ (7 files in 4 dirs): libmsmcomm: implement response message for get_home_network_name command Nov 15 14:52:45 playya: http://www.webos-internals.org/wiki/PalmPre2HardwareAndDriversInformations#gunzip_.3C_.2Fproc.2Fconfig.gz Nov 15 17:49:26 mickeyl: ping Nov 15 18:18:45 playya: by "not compatible" what do you mean? what in particular ? Nov 15 18:19:03 oups wrong chanel (but good person :) ) Nov 15 18:19:47 libeflvala? Nov 15 18:26:19 mickeyl: hi. will you have few seconds to answer my mail (about gps)? Nov 15 18:34:57 leviathan, hi, did you managed to start geeksphone hacking? :) Nov 15 19:23:10 freesmartphone.org: 03seba.dos1 07cornucopia * r32e225fd6ecd 10/fsodeviced/ (4 files in 3 dirs): fsodeviced: add gpio_input module for devices using gpio-switch interface (like sliders, camera shutter and buttons on N900) Nov 15 19:24:20 JaMa, i can't commit my changes on eflvala, because my mainboard is broken :/ Nov 15 19:24:34 ah :/ Nov 15 19:25:10 btw did Lukasz contact you? Nov 15 19:25:27 In last email about ffphonelog he said: Those "updated vala bindings" introduce api inconsistency between Nov 15 19:25:29 Toolbar.append vs Genlist.item_append (and others) Nov 15 19:26:30 iirc those elementary changes were for consistency, so it's a bit strange if it created inconsistency in vala bidings (but maybe you haven't yet updated) Nov 15 19:46:47 JaMa, no elm update yet Nov 15 19:47:16 i need some updates in the lower layer's first Nov 15 19:52:58 playya: I'm ok with current state (because ffphonelog compiles fine), just wanted to relay that message from Lukasz :) Nov 15 20:14:26 freesmartphone.org: 03seba.dos1 07cornucopia * r7579b6eda797 10/fsodeviced/src/plugins/gpio_input/plugin.vala: fsodeviced: gpio_input: don't use the same dbus path for different gpio devices Nov 15 20:34:55 YES :) Nov 15 20:36:04 anyone seen GNUtoo today? Nov 15 20:36:38 YES :) Nov 15 20:38:37 freesmartphone.org: 03seba.dos1 07cornucopia * r727fca27fd90 10/fsodeviced/src/plugins/gpio_input/plugin.vala: fsodeviced: gpio_input: allow to specify, which string indicates button being pressed Nov 15 20:46:04 freesmartphone.org: 03seba.dos1 07cornucopia * r0972b945a27f 10/fsodeviced/src/plugins/gpio_input/plugin.vala: fsodeviced: gpio_input: fix compilation error. Bad dos, bad! Nov 15 20:51:42 dos1: autocriticism as sign of a good character :P Nov 15 20:52:06 s/as/is/ Nov 15 20:52:07 mrmoku meant: dos1: autocriticism is sign of a good character :P Nov 15 20:53:02 * JaMa has fix for crashing Xorg Nov 15 20:53:27 * JaMa doesn't belive he was looking on it whole time and then was mislead somewhere else.. Nov 15 20:53:51 imho a "SPANK SPANK SPANK" is missing :) Nov 15 20:54:29 JaMa: who is guilty? :) Nov 15 20:54:42 author of course :) Nov 15 20:54:52 name of the package? :P Nov 15 20:55:02 xf86-video-glamo glamo-drm.c Nov 15 20:55:29 ok :) Nov 15 20:57:15 * JaMa wants fast graphics without all debugs back! :) Nov 15 21:07:47 freesmartphone.org: 03seba.dos1 07cornucopia * r8fa36faa70f8 10/fsodeviced/conf/nokia_n900/fsodeviced.conf: fsodeviced: conf: use gpio_input on N900 Nov 15 21:07:49 guess what! :) Nov 15 21:08:15 all buttons and sliders on N900 should be supported by FSO now :) Nov 15 21:09:16 dos1: goooooooood :D Nov 15 21:09:50 dos1: I need to build something on buildhost, can I drop your cornucopia SRCREV bump with make update? Nov 15 21:10:03 JaMa: yup, i just used it for testing Nov 15 21:10:22 JaMa: though i'm sending now patch for bumping it ;) Nov 15 21:11:46 did you test it on all cornucopia recipes or just fsodeviced? Nov 15 21:13:45 JaMa: fsodeviced and libfsoframework Nov 15 21:14:37 JaMa: there were also small changes to fsogsmd (but for other modems) and mdbus2 in between, i'm checking if they compile ATM Nov 15 21:15:21 ok, let me know and I'll apply then Nov 15 21:16:40 JaMa: they compile :) Nov 15 21:23:21 ok applied :) Nov 15 21:47:17 Weiss: http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=a02d3d0eefe03576c33294b4201fdbe6bde21cde Nov 15 21:58:50 JaMa: fixes the GTK problems? Nov 15 21:59:07 yes Nov 15 21:59:37 orrery/dictator/gpe-filemanager/gpe-sketchbook/foxtrotgps Nov 15 22:01:00 everytime it was drawing > 256 (512) Solid Fills for 1 PrepareSolid Nov 15 22:21:35 Lopi, you were looking for me? Nov 15 23:04:38 JaMa|Zzz, i've fixed the bug, tomorrow i will post bump rev patch Nov 15 23:04:40 :) Nov 15 23:06:26 goodnight Nov 16 00:33:47 Hai guys. I just tested a recent SHR-unstable and have been reporting a few bugs, but one very obvious one is that the top bar in the lock screen is just broken noise, and also covering a part of the lock screen clock. Since this is so obious I wonder if I might have missed something while searching for it in the trac. Nov 16 00:34:06 So what is the "official" term for the lock screen? :) Nov 16 00:45:11 pipe: all the core devs are asleep by now, so do not expect a fast answer ;) Nov 16 00:46:32 Well wake them up! ;) Nov 16 00:47:05 I'll continue my exploration. Nov 16 01:50:59 PaulFertser, would you be able to help me out with something? **** ENDING LOGGING AT Tue Nov 16 02:59:58 2010