**** BEGIN LOGGING AT Tue Sep 14 02:59:57 2010 Sep 14 06:42:25 will be back on 30th, c'ya. Sep 14 06:43:16 :-O Sep 14 06:43:38 Holiday trip? :) Sep 14 09:44:58 morning Sep 14 09:53:15 what version of vala does the shr buildhost use? Sep 14 12:04:57 anyone see dos1 lately? Sep 14 12:05:19 he's at school propably ;) Sep 14 12:10:13 soltys: good guess Sep 14 12:13:26 ;) Sep 14 12:20:26 gitorious is down :( Sep 14 12:21:55 is there any way to find out which bootloader image I'm using? Sep 14 12:22:43 I'm trying to find out which one I'm using from those at this link http://dos.openmoko.pl/overclock/ Sep 14 12:23:25 badcloud: you can cat it Sep 14 12:23:49 what's the location if I flashed it? Sep 14 12:24:34 badcloud: /dev/mtdblock1? Sep 14 12:26:55 lindi-: that gives me lots of giberish Sep 14 12:30:02 badcloud: that's what binaries look like Sep 14 12:32:53 yeah, I got that from 'cat /dev/mtdblock1|grep qi' ;P Sep 14 12:32:58 so what do I do then? Sep 14 12:34:09 do I need to translate it to ascii somehow (intuition+ignorance=funny and dumb questions)? Sep 14 12:49:02 lindi-: found it. Qi Bootloader s3c2442 sowa master_c38b062a609f1442 Sep 14 12:49:26 lindi-: but 'c38b062a609f1442' corresponds with all of the images on the link Sep 14 12:51:20 badcloud: the link? Sep 14 12:52:06 http://dos.openmoko.pl/overclock/ Sep 14 12:52:26 well, ask dos? ;) Sep 14 12:52:38 badcloud: you can just compare the binaries instead of looking at plaintext strings Sep 14 12:52:42 waiting for him to show :) Sep 14 12:54:24 badcloud: again, i recommend you to use glamo 2-4-2 instead Sep 14 12:54:33 badcloud: if you have omhacks you can use it to see the timings Sep 14 12:54:48 ah it was about overclocking Sep 14 12:54:48 ah Sep 14 12:54:51 yes Sep 14 12:55:06 gena2x: didn't you say there isn't a qi image with that? Sep 14 12:55:25 you can use overcked image + lindi's omhacks tool Sep 14 12:55:32 this will be fastest Sep 14 12:55:46 badcloud: 'om screen glamo-bus-timings 2-4-2' Sep 14 12:55:51 but it's not in debian unstable yet Sep 14 12:55:57 (that particular option) Sep 14 12:56:03 lindi-: aaah Sep 14 12:56:20 squeeze is kind of frozen now Sep 14 12:56:51 gena2x: it's just that my u-boot is broken Sep 14 12:57:05 so I'm using qi Sep 14 12:57:06 broken? Sep 14 12:57:12 what yes Sep 14 12:57:17 *yes Sep 14 12:57:24 what's wrong with it? Sep 14 12:57:34 if I flash uboot Sep 14 12:57:55 I see nand boot text Sep 14 12:57:59 "poweroff" Sep 14 12:58:02 nothing but that Sep 14 12:58:19 you should prepare and flash environment Sep 14 12:58:30 gena2x: how Sep 14 12:58:38 use devirginator too Sep 14 12:58:40 *tool Sep 14 12:58:52 http://svn.openmoko.org/trunk/src/host/devirginator/ Sep 14 12:59:01 it has envedit.pl Sep 14 12:59:10 I tried that. got stuck along the way. I'll tell you exactly at which stage in one sec Sep 14 12:59:10 and sample config environment.in Sep 14 13:00:15 gena2x: oh wait, don't I need jtag for that? Sep 14 13:00:43 edit it, ./envedit.pl -f environment.in -s 262144 > env.new Sep 14 13:01:06 then sudo dfu-util -a u-boot_env -D env.new Sep 14 13:01:13 and that's all. Sep 14 13:01:26 no of course Sep 14 13:01:37 no jtag Sep 14 13:01:54 this is just one tool to cook uboot env Sep 14 13:01:56 I'll try now Sep 14 13:03:57 gena2x: done Sep 14 13:04:08 so which u-boot image did you recommend Sep 14 13:04:10 should work Sep 14 13:04:18 wait a bit... Sep 14 13:05:11 sure Sep 14 13:06:43 badcloud: www.bsdmn.com/openmoko/glamo/242/u-boot_440_glamo242.udfu Sep 14 13:06:48 for overclocked Sep 14 13:07:23 and if it will not work (kernel crashes, no menu at all), www.bsdmn.com/openmoko/glamo/242/u-boot_glamo242.udfu Sep 14 13:07:52 440 has unchanged memory voltage Sep 14 13:08:24 and cpu too Sep 14 13:09:53 gena2x: same deal, "poweroff" at the top, horizontal colorful segmented line along the bottom third of screen Sep 14 13:10:18 ah, try POWER+AUX Sep 14 13:10:42 while booting default it try to load picture from nand Sep 14 13:10:55 with power+aux you should be menu Sep 14 13:12:23 *get Sep 14 13:12:50 DocScrutinizer, GNUtoo|laptop: I have SHR running without r&d mode :) Sep 14 13:12:56 gena2x: any idea what's wrong? Sep 14 13:13:03 mrmoku: :-D Sep 14 13:13:09 kick wd? Sep 14 13:13:26 correct watchdog.conf... Sep 14 13:13:35 actually two confs Sep 14 13:13:45 err, that's a SHR specific? Sep 14 13:13:48 and starting watchdog and wd_keepalive Sep 14 13:14:06 don't think so... as I got it from ele...'s site ;) Sep 14 13:14:59 hmm, I gather watchdog and wd_keepalive are /etc/init.d ? Sep 14 13:15:06 mrmoku, nice! Sep 14 13:15:26 DocScrutinizer: yep Sep 14 13:16:11 yep, nothing like that in maemo Sep 14 13:16:15 gena2x: how would I do that, load picture from nand? Sep 14 13:17:07 I'd be highly interested in the one interesting line in those scripts - is it "echo foobar > /dev/watchdog_foo"? Sep 14 13:17:46 DocScrutinizer: ahh, then I misunderstood your /etc/init.d... watchdog and wd_keepalive are two daemons that are started Sep 14 13:17:51 badcloud: see people.om.org/joerg/dontpanic Sep 14 13:18:14 mrmoku: no scripts? o.O Sep 14 13:18:20 no Sep 14 13:18:27 blah Sep 14 13:18:31 http://www.matthew.ath.cx/cgi-bin/man/man2html?8+watchdog Sep 14 13:18:50 DocScrutinizer: bad link Sep 14 13:19:11 badcloud: it's an insider hint, no link :-D Sep 14 13:20:04 kkk, won't fsck with u. mompl Sep 14 13:20:27 http://people.openmoko.org/joerg/bootsplash/ Sep 14 13:20:56 has very terse howto Sep 14 13:21:32 ? Sep 14 13:21:45 btw, now some guys in maemo are picking up ACI to get rid of maemo PA madness :-D Sep 14 13:22:43 so after 2+ years it seems there'll be at least a minimal result from my ALSA digging the mud Sep 14 13:22:54 :) Sep 14 13:22:55 aci ? Sep 14 13:23:09 http://people.openmoko.org/joerg/ALSA/ACI/ Sep 14 13:23:50 esp acihooklib there Sep 14 13:25:16 which allows to found a completely integrated better-maybe audio management on it, than all PA policy enforcer and libcanberra and whatnot Sep 14 13:25:54 DocScrutinizer: ah, just got it, *wink wink* Sep 14 13:27:17 GNUtoo|laptop: hmm... there are no /dev/mmcblk* nodes ? Sep 14 13:27:22 badcloud: you should just place it to special nand partition Sep 14 13:27:38 mrmoku, there are Sep 14 13:27:52 mrmoku, /dev/mmcblk1p1 for instance Sep 14 13:27:57 or mmcblk0p1 Sep 14 13:27:59 but beware Sep 14 13:28:05 we should use bootloader stuff Sep 14 13:28:19 because /dev/mmcblk1 and /dev/mmcblk0 are often interverted Sep 14 13:28:21 badcloud: i saw exact instruction on wiki Sep 14 13:28:37 mrmoku, I'll push my kernel config later on Sep 14 13:29:01 idea being the user apps using ALSA snd_open() and snd_close() don't do this with "ALSA:default" but come with their own .alsarc.d/* alsaconfig files to include to ALSA .alsarc and defining a unique ALSA audio device for the app. Then with acihooklib and ALSA hooks plugin you define arbitrary commands that get executed whenever such a unique individual audio dev is snd_open()'ed or close()'ed Sep 14 13:29:53 GNUtoo|laptop: I don't have them... cat /proc/partitions lists them though Sep 14 13:30:01 I have Sep 14 13:30:14 mrmoku, but I've also that in my kernel: Sep 14 13:30:19 *omap mmc driver Sep 14 13:30:25 *block devices for mmc cards Sep 14 13:30:30 *ext2/3 Sep 14 13:30:37 and similar stuff necessary for booting Sep 14 13:31:31 hmm Sep 14 13:32:41 GNUtoo|laptop: you are using multiboot? Sep 14 13:33:06 I tried everything Sep 14 13:33:09 :) Sep 14 13:33:18 currently I've a flashed kernel Sep 14 13:33:23 I'll retry uboot Sep 14 13:33:29 but later Sep 14 13:33:45 to *not* share the kernel with maemo... that would be possible with multiboot? Sep 14 13:33:54 yes indeed Sep 14 13:34:49 also with uboot but multiboot is more user friendly Sep 14 13:41:02 flashed the splashscreen Sep 14 13:41:18 still getting nor boot menu text Sep 14 13:41:42 nor is what I meant instead of 'NAND ' (beforehand) Sep 14 13:47:08 badcloud: how do you boot? Sep 14 13:47:27 lindi-: long power press Sep 14 13:48:12 badcloud: maybe your AUX is stuck? Sep 14 13:48:33 lindi-: I don't get nor menu when qi is flashed Sep 14 13:48:55 badcloud: how do you know it is in NOR? Sep 14 13:48:57 lindi-: btw, when uboot is flashed, it's not that I get nor menu, only the firsl line of text Sep 14 13:49:38 if uboot is flashed and I press power for ~3 sec Sep 14 13:49:39 badcloud: to see menu, hold power for 1 sec and press and hold AUX Sep 14 13:50:09 gena2x: one sec, reflashing uboot to try this out Sep 14 13:50:14 i mean power for 1 sec, then power+aux Sep 14 13:51:23 gena2x: same deal, the text 'poweroff' is at the top left corner of screen Sep 14 13:51:37 now without the colorful horiz line at the bottom Sep 14 13:51:50 I can wait a while to see what happens Sep 14 13:52:11 i think you just didn't flash boot env properly Sep 14 13:52:59 I can download devirginator again Sep 14 13:53:20 you have no need in whole thing Sep 14 13:53:36 just download envedit and environment.in Sep 14 13:53:41 ok Sep 14 13:53:43 and use my instructions Sep 14 13:54:12 so far this worked for everyone Sep 14 13:55:11 Undefined subroutine &main::crc32 called at ./envedit.pl line 79 Sep 14 13:55:46 ops Sep 14 13:56:07 ok, anyway, i g2g Sep 14 13:56:19 ok Sep 14 13:56:20 it should work Sep 14 13:56:34 do I need to do it as root or reg user? Sep 14 13:56:46 ^ was as root Sep 14 13:57:19 dos1: just the man I wanted to see Sep 14 13:57:44 do you have a copy of that priv msg conversation we had about me testing your qi overclk images? Sep 14 13:58:22 badcloud: hi Sep 14 13:59:09 badcloud: hmm... did you have "badcloud" nick back then, or with some suffix? Sep 14 13:59:30 could be '_' as a suffix Sep 14 13:59:31 "badcloud" log is empty for me in konversation :x Sep 14 13:59:31 not sure Sep 14 13:59:40 dag Sep 14 13:59:51 I don't remember which of the images worked for me Sep 14 14:11:12 badcloud: could you change your nick to badcloud_ for a while? i can't find how to open custom log here ;x Sep 14 14:15:26 GNUtoo|laptop: did you find out what is causing dbus-daemon to run havoc? Sep 14 14:15:39 mrmoku, later, I'm working for now Sep 14 14:15:47 s/working/studying Sep 14 14:15:47 GNUtoo|laptop: ok :) Sep 14 14:18:14 mrmoku, maybe fsogsmd Sep 14 14:18:34 mrmoku, try dummy modem Sep 14 14:20:35 GNUtoo|laptop: even after killing phonefsod, phoneuid and fsogsmd it behaves bad Sep 14 14:21:56 that's because of cron Sep 14 14:22:07 try putting a dummy modem Sep 14 14:22:43 hmm... ok Sep 14 14:26:16 cat /proc/interrupts Sep 14 14:26:18 oops Sep 14 14:28:46 dos1: ok Sep 14 15:05:20 dos1: locally building fixed watchdog now :) Sep 14 15:08:27 mrmoku: nice :) Sep 14 15:12:40 dos1: looking into xserver-common now Sep 14 15:28:50 dos1: any idea how we get a WiFi resource for n900 into fsodeviced? Sep 14 16:01:54 dos1: watchdog and xserver-common pushed Sep 14 17:30:46 GNUtoo: do you have a pointer for me where to find multiboot? Sep 14 17:31:35 mrmoku, apt-get it Sep 14 17:31:43 mrmoku, it's in the extra feeds Sep 14 17:31:53 my computer uses too much ram and freeze Sep 14 17:31:56 ahh, ok, thanks Sep 14 17:32:01 so I'm trying to fix that now Sep 14 17:32:08 so I can build again Sep 14 18:11:36 GNUtoo: heh, forgot to add an entry for maemo :/ Sep 14 18:11:39 can only boot shr now :P Sep 14 18:34:40 mrmoku, lol Sep 14 18:34:48 mrmoku, maybe it's the way to go? Sep 14 18:34:56 aka eat your own dog fod Sep 14 18:34:58 *food Sep 14 18:39:57 GNUtoo: hehe, wanted to get still some inspirations from the maemo ui though :) Sep 14 18:40:12 ok nice Sep 14 18:42:07 ahh... 0 is default to maemo in multiboot menu Sep 14 18:47:57 mrmoku, any new commits? Sep 14 18:48:04 should I try to make a new image? Sep 14 18:48:10 dos11, hi Sep 14 18:48:26 dos11, hi Sep 14 18:49:08 GNUtoo|laptop: I pushed xserver-common and watchdog Sep 14 18:49:13 ok Sep 14 18:49:19 should I push udev? Sep 14 18:49:20 should obsolete the use of r&d Sep 14 18:49:24 ok nice Sep 14 18:49:36 what did you fix in udev? Sep 14 18:49:40 nothing yet Sep 14 18:49:50 I want to add it for certain machines only,in the image Sep 14 18:49:58 but you could do it too Sep 14 18:49:59 ahh... yep Sep 14 18:50:10 I wanted to find out how to handle wifi Sep 14 18:50:17 ah wifi Sep 14 18:50:22 indeed we have a kernel issue Sep 14 18:50:24 it works fine, but fsodeviced does not discover it as resource Sep 14 18:50:28 ifconfig wlan0 up Sep 14 18:50:36 works fine Sep 14 18:50:38 if: Sep 14 18:50:42 *you have the firmware Sep 14 18:50:45 *you use udev Sep 14 18:50:57 *you have modified udev rules Sep 14 18:50:59 but: Sep 14 18:51:06 ifconfig wlan0 down doesn't Sep 14 18:51:12 maybe that's because I'm using PSM? Sep 14 18:51:18 psm? Sep 14 18:51:22 power saving mode Sep 14 18:51:26 iwconfig wlan0 power on Sep 14 18:51:31 ahh, ok Sep 14 18:51:32 huge power savings Sep 14 18:51:42 basically the router buffer the packets for you Sep 14 18:51:52 and then send them all at once when the phone request them Sep 14 18:51:54 and other stuff Sep 14 18:52:08 * mrmoku installs wireless-tools Sep 14 18:53:53 GNUtoo|laptop: how do you see if ifconfig down works? Sep 14 18:54:03 if I do ifconfig down wlan0 and wmaster0 disappear Sep 14 18:54:04 if it segfault Sep 14 18:54:09 then look at dmesg Sep 14 18:54:10 still shown with ifconfig -a Sep 14 18:54:20 does it segfaults Sep 14 18:54:21 ? Sep 14 18:54:46 [ 535.265747] wl1251: down Sep 14 18:54:48 works Sep 14 18:54:50 does not segfault Sep 14 18:55:09 I have the following on ifconfig up though Sep 14 18:55:11 [ 105.859588] wl1251: ERROR Power save entry failed, giving up Sep 14 18:55:21 strange Sep 14 18:55:24 I'll try Sep 14 18:55:41 hmm... not on ifconfig up Sep 14 18:55:52 probably maemo trying to connect to my AP Sep 14 18:56:13 btw. I have no modified udev rules Sep 14 18:56:19 but I have udev and the firmware installed Sep 14 18:56:45 mrmoku, http://pastebin.com/E0EnDzu5 Sep 14 18:57:00 mrmoku, look here for modified dev rules: Sep 14 18:57:18 SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1f:df:*", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="wlan*", NAME="wlan0" Sep 14 18:57:20 in: Sep 14 18:57:24 /etc/udev/rules.d/70-persistent-net.rules Sep 14 18:57:32 note the 00:1f:df:* Sep 14 18:57:37 I'll explain Sep 14 18:57:42 basically it's a wl1251_spi Sep 14 18:58:04 GNUtoo|laptop: ok, that udev rule is only for persistent naming of the interface Sep 14 18:58:05 the wl1251 don't have the memory(flash,eeprom etc...) to store the mac address Sep 14 18:58:10 indeed Sep 14 18:58:19 else the interface do: Sep 14 18:58:29 wlan0->wlan1->wlan2->wlan3... Sep 14 18:58:48 heh, ok Sep 14 19:08:49 GNUtoo|laptop: even when associated ifconfig down does not segfault for me Sep 14 19:09:19 strange Sep 14 19:09:28 maybe different kernel config Sep 14 19:09:29 here's mine Sep 14 19:09:38 I'm using maemo default kernel Sep 14 19:10:07 http://pastebin.com/5nZYnm0T Sep 14 19:10:10 ah ok Sep 14 19:10:11 not me Sep 14 19:10:21 GNUtoo|laptop: (mac address) i figured it out as i have to click "reg" button on my router every time i want to connect to wifi with SHR/N900 :D Sep 14 19:10:36 ok Sep 14 19:10:37 lol Sep 14 19:11:01 I've a GNU/Linux x86 DIY router Sep 14 19:11:14 ath9k Sep 14 19:11:18 hostapd Sep 14 19:11:26 stock distro kernel Sep 14 19:11:28 etc... Sep 14 19:11:38 but all the config were made by me Sep 14 19:11:48 what kernel source are you using? Sep 14 19:11:51 like hostapd.conf, named.conf etc... Sep 14 19:11:53 on what? Sep 14 19:11:57 n900 Sep 14 19:11:58 the ones in oe Sep 14 19:12:15 kernel_2.6.28-20101501+0m5.diff.gz Sep 14 19:12:20 did somebody already integrate the patches from meego? Sep 14 19:12:21 I don't know if I pushed it tough Sep 14 19:12:26 no Sep 14 19:12:31 we should do that probably Sep 14 19:12:34 ok Sep 14 19:12:54 http://meego.gitorious.org/meego-os-base/kernel-source/trees/master/patches Sep 14 19:13:08 there is a config for n900 too in that repo Sep 14 19:13:29 ok Sep 14 19:14:18 and maybe there is rfkill in that kernel too Sep 14 19:14:26 can old meego scan while associated btw? Sep 14 19:14:33 old meego? Sep 14 19:14:36 you mean maemo? Sep 14 19:14:37 s/meego/maemo Sep 14 19:14:38 yes Sep 14 19:15:29 yup, I can do iwlist scan while associated Sep 14 19:16:15 ok Sep 14 19:16:21 what year is the kenrel from? Sep 14 19:16:30 if it's 2009 I'm very interested Sep 14 19:16:38 the maemo one? Sep 14 19:16:43 yes Sep 14 19:16:50 Nokia-N900-51-1:/etc/multiboot.d# cat /proc/version Sep 14 19:16:50 Linux version 2.6.28-omap1 (bifh6@maemo-bifh-19) (gcc version 4.2.1) #1 PREEMPT Thu Dec 17 09:40:52 EET 2009 Sep 14 19:17:06 because scan while associated has been proposed and pushed very recently Sep 14 19:17:10 in compat-wireless Sep 14 19:17:13 and linux-next Sep 14 19:17:27 hmm... interesting Sep 14 19:17:32 indeed Sep 14 19:17:35 for the dream Sep 14 19:17:38 but dream is sdio Sep 14 19:17:55 maybe they patched it... Sep 14 19:18:33 indeed if it makes PSM work on dream.... Sep 14 19:19:02 I should ask dos11 for advises on GUI for dream Sep 14 22:44:15 is our ALSA dmix plugin optimized for Arm? **** ENDING LOGGING AT Wed Sep 15 02:59:57 2010