**** BEGIN LOGGING AT Sun May 15 23:59:57 2005 May 16 00:13:42 hi.everyone May 16 00:45:23 morning May 16 00:50:24 morning all May 16 00:50:37 uv1: I am now May 16 00:55:03 hello May 16 00:55:16 morning hrw|work May 16 00:58:52 RP: did I see talk of a memory light version of BB last week? May 16 01:14:39 Morning RP May 16 01:15:40 As I said, if I have the corgi_power-r21 in the line, it works, but nor with r22 (just the second 2 is new, the rest stays). This is strange. May 16 01:17:48 RP: I observered one other item. If alarm comes up and the device is just in "DisplayOff" mode, the background light is not switched on automatically. The alarm is hearable though (if loud). Is this a kernel issue or a opie issue ? May 16 01:18:54 there are some opie and apm issues May 16 01:28:28 XorA: You did. zecke is working on it May 16 01:28:54 uv1: That is an opie issue May 16 01:29:10 I can't understand bitbake not working with the r22 :-/ May 16 01:32:17 if desparate, comment out the apm line in /etc/device_table (warning, alarms and a few other things wont work if you do that) dont forget to reboot May 16 01:33:05 subdue: its not #openzaurus channel May 16 01:33:14 oops May 16 01:33:18 wrong chan reply, sorry :) May 16 01:38:25 RP: Also the battery seems to be very good, or the percentage calc is NOT. My CL860 is now since approx. 45 mins in 100% CPU load, but battery still stays on 94 %. May 16 01:38:42 RP: Did you also have the probs with r22 ? May 16 01:48:24 uv1: Someone really needs to do a lock cpu at 100%, backlight on full and turn on the logging (msglevel=2 will do that now). We then need to plot the battery voltage with time May 16 01:48:48 That will show us if the percentages are accurate. If they're not, we can adjust them. May 16 01:49:03 I didn't have any problem adding the r22 patch to bitbake May 16 01:49:14 RP: how to set that msglevel? May 16 01:49:47 echo 2 > /sys/devices/platform/corgi-battery/msglevel May 16 01:49:50 RP: and can it be logged into file on cf card? I can run that test on my husky May 16 01:50:04 first will charge it to 100% May 16 01:50:19 hrw|work: It dumps the data to the console via a printk :-/ May 16 01:50:47 RP: so logread -f|tee /media/cf/dump.txt should work May 16 01:51:00 hrw|work: yes, something like that May 16 01:51:18 ok. charging May 16 01:51:24 hrw|work: If you are going to do this, could you use a kernel with the corgi_power-r22 patch pelase? May 16 01:51:39 RP: is it possible to look how charged battery is when AC plugged? May 16 01:51:41 This patch is a massive cleanup and the logging will be much clearer May 16 01:51:43 ok - will build it May 16 01:52:12 hrw|work: With corgi_power-r22, it does try and give a percentage charged estimate May 16 01:52:24 Again, it probably needs calibrating May 16 01:53:05 hrw|work: I'll update the .bbs with the new patch name... May 16 01:55:10 2.6.12-rc? May 16 01:55:59 hrw|work: It will apply to both 2.6.11 and 2.6.12-rc May 16 01:56:17 ok. which one you prefer to check? May 16 01:58:31 I'm actively using 2.6.12-rc but it makes no real difference May 16 01:58:45 ok May 16 01:59:06 will check on 2.6.11 May 16 01:59:29 RP: it will be 2.6.11-r19 with corgi-power-r22 May 16 02:01:08 hrw|work: That's fine May 16 02:01:55 03RP 07 * r1.3283.1.1 10openembedded/packages/linux/ (10 files in 2 dirs): linux-oz-2.6: Update to 2.6.12-rc3-mm3, Update corgi power patch, Updates to hx2750 code (esp tsc2101 driver). May 16 02:02:04 so where are you all located? May 16 02:02:19 mithro: geographically? May 16 02:02:31 mithro: northeast England (UK) May 16 02:02:53 mithro: Edinburgh Scotland May 16 02:03:11 mithro: Poznan in Poland May 16 02:03:32 * mithro is going to be in Europe from July May 16 02:04:11 mithro: Which bit of Europe? May 16 02:04:17 got a month to travel from Amsterdam to Stockholm May 16 02:05:30 so I'm planning to fly in and slowly travel up to Stockholm by train May 16 02:05:36 it is nice May 16 02:05:45 mithro: This is quit boaring, all the way is See. Will you do that in a small boat ? May 16 02:06:04 s/See/sea/ May 16 02:08:08 i'm thinking of heading down through belgium and germany and then back up through denmark and into sweden May 16 02:09:06 * ljp is in brissie :) May 16 02:09:27 ljp: know that :P May 16 02:10:50 mithro: If you come accross Aachen (which I assume), we possibly could drink a beer or few bottles of good red wine together ;-) May 16 02:11:24 where is Aachen May 16 02:11:28 mithro: Sounds like fun. I've never been to most of those places :) May 16 02:11:37 * mithro is hopless at Europe geography :P May 16 02:11:55 mithro: germany May 16 02:12:12 mithro: So you're US I assume ;-)) May 16 02:12:17 nope May 16 02:12:24 * mithro is in Adelaide, Australia May 16 02:12:54 One of those upsidedown people with the corks on the hat ;-) May 16 02:13:12 mithro: Nice region. Have been there for hollidays 3 times now. Good wine, good foor, australian lifestyle. Which I could be there ... May 16 02:13:23 s/foor/food/ May 16 02:13:39 s/Which/Wish/ May 16 02:14:02 it's pretty hot in the summer here however :P May 16 02:14:07 * RP has never been outside Europe May 16 02:14:22 morning all May 16 02:14:32 morning ade|desk May 16 02:14:40 RP: Still have problems to get the patch downloaded from RPSRC. Interested in investigation ? May 16 02:15:18 so where do people recommend I go :P May 16 02:15:24 uv1: I've just updated the bitbake file with that patch. I'd suggest pulling those updates and seeing whether that works May 16 02:15:29 RP: pxa_irda-r1.patch fails with that latest changeset you just put in May 16 02:15:49 XorA: Ah, try -r2 :) May 16 02:16:55 RP: Nothing to pull ? Just did it a few mins ago, but did not have linux-openzaurus_2.6.11.bb updated. So whats now ? May 16 02:17:32 * mithro is of to dinner bblr May 16 02:17:46 uv1: set your bk parent to bk://oe-devel@oe-devel.bkbits.net/openembedded May 16 02:17:54 uv1: Just don' May 16 02:17:58 uv1: Just don't clone from there May 16 02:18:24 uv1: Also check your sources directory for 0 byte sized files May 16 02:19:02 RP: How to do that (parent)? May 16 02:19:56 morning May 16 02:19:59 RP: No 0 sized files in sources. May 16 02:20:56 uv1: bk parent bk:/.... May 16 02:21:37 morning May 16 02:22:04 RP: And what then ? "bk pull; bk -r co -q" or just "bk pull" ? May 16 02:23:20 RP: oe-devels PWD ? May 16 02:23:51 uv1: bk parent bk://oe-devel.bkbits.net/openembedded May 16 02:24:00 and it wont ask for pass May 16 02:24:45 hrw|work: Thanks, just "bk pull" ? May 16 02:25:05 uv1: sorry, yes, remove the oe-devel@, then do the pull and bk -r co -q as normal May 16 02:26:27 Hi May 16 02:26:29 RP: Do you have 2.6.12-rc4 running May 16 02:26:54 do13: I have 2.6.12-rc4 running but 2.6.12-rc4-mm1 doesn't work May 16 02:27:17 I guess: Can't load libs? May 16 02:27:37 do13: No, it was crashing with some timer problem May 16 02:27:42 I posted to LKML about it May 16 02:28:36 do13: http://seclists.org/lists/linux-kernel/2005/May/2830.html May 16 02:29:49 RP: I've got some new spitz patches, both CF slots are working now too. May 16 02:30:17 BigAl: Sounds good :). If you point me at the patches, I' May 16 02:30:25 I'll add the to the .bb file May 16 02:30:43 ~lart emte for e recipes May 16 02:30:43 * ibot says "boot to the head" and knocks emte over for e recipes May 16 02:31:02 do13: I noitce rc4-mm2 is out now so perhaps I should try that... May 16 02:31:15 RP: You can also add the tosa patches: http://www.do13.in-berlin.de/openzaurus/tosa2.6.patch May 16 02:31:20 RP: http://www.orca.cx/zaurus/patches May 16 02:31:49 RP: I have the samer May 16 02:31:55 RP: The ones to add would be spitz-cf-support, spitz-base-r1 May 16 02:32:19 RP: I the same crash. Bu before there is /bin/sh: error while loading shared libraKernel panic - not syncing: Attempted ! May 16 02:32:19 ries: libc.so.6: cannot create shared object descriptor May 16 02:32:19 <3>BUG: soft lockup detected on CPU#0! May 16 02:32:33 RP: Same result. corgi_power-r22.patch is not downloaded somehow. If I change to corgi_power-r21, it works. There must be difference between r22 and r21. Pending balnk in filename, access rights? May 16 02:32:44 RP: The spitz-detection patch assumes you've got the tosa detection one installed, and there's one for spitz and tosa there too. I can make one for spitz alone if that'd be better. May 16 02:33:04 uv1: I got r22 May 16 02:33:17 uv1: Ive just downloadedf -r22 here May 16 02:33:59 BigAl: Assuming Tosa is there is the best way forward May 16 02:34:28 BigAl: I'll have a look at them shortly and add them in. May 16 02:34:54 do13: You're just using one tosa patch now? May 16 02:35:19 hrw|work: I just did a "-b ...2.6.11.bb -f - clean", "-b ...2.6.11.bb" I assume, that has to download the patch ? May 16 02:35:21 RP: No this patch is against the bb file May 16 02:35:30 do13: Does that just happen for you on rc4-mm1 or on rc4 as well? May 16 02:35:33 RP: Okies May 16 02:35:42 do13: Ah, right, that's great, thanks :) May 16 02:35:59 do13: I thought you were going to copy John for a minute there! ;-) May 16 02:36:14 ;-) May 16 02:37:01 uv1: I edited recipe to get r22 - not updated yet to latest OE May 16 02:37:59 uv1: I can't see any permissions or filename problems and can download it fine... (I use the same method you do). It must be some local problem... May 16 02:38:16 Try deleting the file from the sources directory May 16 02:38:29 RP: and the .md5 part as well May 16 02:38:29 hrw|work: Did that before with no success either. Still do not understand, why r21 works, while r22 does not. BTW, if downloaded by FireFox, I get that file as well. May 16 02:38:52 RP: I thought the patch file is not in source, but loaded on demand !? May 16 02:39:32 uv1: If it exists in sources it will use that, if not, it will download it May 16 02:39:45 RP: Delete 2.6.11.tar.gz* ? ... May 16 02:40:18 uv1: corgi_power* May 16 02:41:20 well i'm back May 16 02:48:22 RP: For some strange reason only the r22 md5 file was there. If I remember right, I did the first bitbake with a buggy edited r22 line. It works now, thanks all ... May 16 02:49:05 RP: BTW CL860 is still on 100 CPU (since approx. 2 hours now), Battery still at 83 %. May 16 02:49:32 uv1: how long does it normally last? May 16 02:49:42 ok. got 100% charge May 16 02:50:13 shit.. 96% now May 16 02:51:24 morning folks May 16 02:51:30 hey mickeyl May 16 02:51:47 RP: msglevel is WO? (write only) May 16 02:51:59 hi mickeyl May 16 02:52:37 hrw|work: i fixed y^hour problem May 16 02:52:45 mickeyl: WE V0? May 16 02:52:51 hrw|work: ya. May 16 02:53:00 cool May 16 02:53:01 hrw|work: It shouldn't be! try chmod'ing it +w... May 16 02:53:04 ~lart wifi drivers in general May 16 02:53:04 what was it? May 16 02:53:04 * ibot gets a hotmal account and SPAMs wifi drivers in general May 16 02:53:32 hostap suddenly decided to respond with invalid argument length when querying status information. May 16 02:53:33 RP: its 644. I did "echo 2>msglevel" and cat msglevel show 0 May 16 02:53:43 i had to adjust the buffer once again May 16 02:53:58 the problem is... wlan-ng needs a short buffer, hostap needs a large buffer May 16 02:53:58 hrw|work: Is it showing debug in dmesg? May 16 02:54:05 mithro: I never tested that. May 16 02:54:07 virtually impossible to get the length right to work with all drivers May 16 02:54:16 *shrug* May 16 02:54:23 lets see for which driver i broke it now :) May 16 02:54:33 s/driver/driver or kernel version/ May 16 02:54:35 RP: last one was "Charge LED Off" and battery dropped from 96 to 94 since then May 16 02:54:55 hrw|work: It should be showing something every 5 seconds or so May 16 02:55:08 Have to leave for lunch now. Will be back this afternoon. Bye .... May 16 02:55:11 (I plan to decrease that rate once the power code is stable) May 16 02:55:32 bon appetit uv1 May 16 02:55:44 heh... May 16 02:56:08 RP: got it finally May 16 02:56:18 hrw|work: What was the problem? May 16 02:57:01 maybe "echo 2>msglevel" or "cd /sys/..../corgi-battery/;echo 2>msglevel" - did "echo 2 >FULL_PATH" and it works May 16 02:57:08 have to charge to 100% back May 16 02:57:47 now have 98% May 16 03:02:43 03koen 07 * r1.3309 10openembedded/packages/thc/ (4 files in 2 dirs): Added packages/thc/ dir for stuff from http://thc.org/, moved hydra and thcrut into it. May 16 03:05:30 03mickeyl 07 * r1.3305.1.1 10openembedded/conf/distro/ (preferred-e-versions.inc openzaurus-3.5.4.conf): bump openzaurus-3.5.4 and lock down our e packages to my working versions through preferred-e-versions.inc May 16 03:06:11 anyone in charge of kstars here? May 16 03:06:20 not really May 16 03:06:44 it was merely an exercise to get it built, since it isn't usable on embedded systems at all May 16 03:07:49 it works. (a bit slow, but it works) May 16 03:07:56 only problem was in creating directories May 16 03:08:25 during installation: /opt/QtPalmtop/share/apps/kstars <= it couldn't create directories this deep May 16 03:08:30 03koen 07 * r1.3311 10openembedded/conf/distro/familiar-0.9.0.conf: familiar-0.9.0.conf: include preferred-e-versions.inc, courtesy Buildmeister Lauer May 16 03:08:39 koen: :) May 16 03:08:40 so it had problems linking May 16 03:08:50 once I created the directory, it installed fine May 16 03:08:59 opie - 64/0 - collie May 16 03:09:01 subdue: just this one ? May 16 03:09:05 yes May 16 03:09:14 ok, i'll create that in the ipk then May 16 03:09:18 it was nested too deep for it May 16 03:09:43 ok May 16 03:10:23 hmm May 16 03:10:25 wait May 16 03:10:38 that should've worked May 16 03:10:43 the directory is in the ipk May 16 03:10:50 ah May 16 03:10:55 hmm May 16 03:10:57 no May 16 03:11:01 ? May 16 03:11:17 RP: ok. got 99% of battery May 16 03:11:22 it bailed out on installation with opie-packagemanager or ipkg command line? May 16 03:11:24 I think it had the problem b/c it didnt create each directory one by one May 16 03:11:31 opie May 16 03:11:38 package manager May 16 03:11:42 can you retry with command line? May 16 03:11:46 it may be a problem in the pm May 16 03:11:49 ok May 16 03:12:42 I'll try to remove all the directories too May 16 03:13:08 03RP 07 * r1.3310 10openembedded/packages/linux/ (4 files in 2 dirs): linux-oz-2.6: Tosa updates from Dirk. Also add preliminary sptiz support under 2.6 - thanks BigAl! May 16 03:13:12 03RP 07 * r1.3309 10openembedded/packages/linux/linux-openzaurus_2.6.12-rc3-mm3.bb: linux-oz-2.6: Update to later irda patch May 16 03:13:23 RP: heh May 16 03:13:26 sptiz ? May 16 03:13:29 the new Zaurus? May 16 03:13:30 :D May 16 03:13:37 C3500 May 16 03:13:40 hehe May 16 03:13:40 bah! :) May 16 03:14:28 BigAl: They're in but it doesn't quite apply. You need to rediff the base support against Dirk's tosa base patch May 16 03:14:59 heh.. logread -f|tee /media/cf/battery.log.txt is slow in terminal May 16 03:15:46 mickeyl: why opie-sysinfo lack cpumeter? May 16 03:15:55 koen@c7x0:~$ uname -a ; uptime May 16 03:15:55 Linux c7x0 2.6.12-rc3-mm2 #1 PREEMPT Thu May 12 12:02:24 CEST 2005 armv5tel unknown May 16 03:15:55 12:15:19 up 3 days, 17:03, load average: 0.05, 0.02, 0.00 May 16 03:16:08 2.6.12 is holding up pretty good May 16 03:16:22 hrw|work: because I'm revamping it a bit May 16 03:16:26 ah May 16 03:16:28 hrw|work: new cpumeter will be part of devices tab May 16 03:16:41 btw. May 16 03:16:49 all devices correctly identified on your system? May 16 03:17:00 mickeyl: yes May 16 03:17:04 ok, good. May 16 03:17:17 ~lart CPU specific /proc/cpuinfo headers May 16 03:17:17 * ibot tries to shut CPU specific /proc/cpuinfo headers up May 16 03:17:19 mickeyl: useless anyway (20050510) May 16 03:17:47 ya, i just started it. will take a while to finish it May 16 03:18:00 koen: That's nice to know. I don't think mine ever sees 3 days as I edit the kernel before then :) May 16 03:18:28 mickeyl: info about cards should be more detailed May 16 03:18:45 Do we have any nasty gotchas under 2.6.12 apart from the keyboard rotate switches? May 16 03:19:06 hrw|work: sure May 16 03:19:32 hmm.. how to take 100% cpu in pure shell? May 16 03:19:51 gzip something May 16 03:20:07 cat /dev/random > /dev/null? May 16 03:20:18 RP: good idea May 16 03:21:10 * CoreDump|home has another go at pivot_root May 16 03:21:35 urandom might work better May 16 03:22:10 hrm. it didnt like installing to my ram May 16 03:22:19 better source of entropy May 16 03:22:21 RP: I'll check it out. May 16 03:23:06 yep, and generates a higher CPU load IIRC May 16 03:23:09 ha! 91% now May 16 03:25:33 bleh. btw - that error about libstdc++.so.6 still comes up May 16 03:27:22 subdue: "This is not a distribution support channel" - so you have to be more specific May 16 03:27:37 oh sorry, was telling him about kstars May 16 03:27:48 he wanted me to try to install it via command line May 16 03:28:11 I had mentioned that using pkgmanager it was failing b/c it couldnt create /opt/QtPalmtop/share/apps/kstars May 16 03:28:11 ArticleDiscussionEditHistoryMove May 16 03:28:25 -ignore article... May 16 03:29:13 I'm going to try to start over from scratch to make sure of all the errors May 16 03:30:17 (completely wiping out all the ram) May 16 03:31:36 k, keep me posted May 16 03:31:45 leaving for a board game day May 16 03:31:50 no computers allowed :) May 16 03:31:52 cya May 16 03:32:05 03mickeyl 07 * r1.3308.2.2 10openembedded/conf/distro/preferred-e-versions.inc: lock down embryo version May 16 03:32:07 03mickeyl 07 * r1.3308.2.1 10openembedded/packages/libpng/ (libpng_1.2.8.bb libpng-native_1.2.8.bb): repeat do_stage() in libpng-native to fix an odd packaging error on some systems May 16 03:35:02 86% May 16 03:38:34 hrw|work: why are you trying to up processor usage? May 16 03:38:51 lardman|busy: I'm testing battery May 16 03:38:55 83% May 16 03:39:16 do you want it at 100%? May 16 03:39:27 lardman|busy: brightness maxed, cpu maxed, kernel battery messages maxed and logging May 16 03:39:37 lardman|busy: I have ~100% now May 16 03:40:15 ok - I was going to suggest you run something like octave with some looping scripts to make it sit at 100% May 16 03:42:54 lardman|busy: cat /dev/urandom>/dev/null did the trick May 16 03:43:13 cool May 16 03:46:14 80% May 16 03:48:47 * RP wonderes about graphing the irc timestamps vs. hrw's percentages :) May 16 03:49:24 RP: you will get much more detailed report when my husky will shutdown May 16 03:49:41 hrw|work: I know, I'm just joking :) May 16 03:49:45 I know ;) May 16 03:49:55 ~lart openlogging.org May 16 03:49:55 * ibot executes killall -KILL openlogging.org May 16 03:50:11 zecke|sleep: ping May 16 03:50:25 takepatch: "packages/opie-calculator/SCCS/s.opie-calculator_cvs.bb" has uncommitted changes. May 16 03:50:28 Please commit pending changes with `bk commit' and reapply the patch. May 16 03:50:49 bk commit == unable to connect to openlogging.org == get off May 16 03:50:55 RP: 76% May 16 03:51:35 RP: 80% back May 16 03:51:42 bk diff packages/opie-calculator/opie-calculator_cvs.bb -Nurd > patch1 ; bk unedit packages/opie-calculator/opie-calculator_cvs.bb ; bk pull ; bk patch < patch1 May 16 03:51:43 ?! May 16 03:52:20 koen: they are checked in May 16 03:53:19 ipaq.patch breaks libsdl-qpe May 16 03:53:20 koen: pong May 16 03:53:46 CoreDump|home: ohhh May 16 03:53:49 zecke|sleep: I wanted to try bb_qa, but it bails out with pyton error May 16 03:54:00 break as in: it doesn't apply :) May 16 03:54:01 koen: Read the Fucking Manual ;) May 16 03:54:16 zecke|sleep: I did :) May 16 03:54:17 RP: 11:37:25 - 76% 11:37:30 - 80% May 16 03:54:44 CoreDump|home: will look, I've to pick up my honey from the train station later.. May 16 03:54:52 koen: paste me the error May 16 03:54:59 time for clone wars.. May 16 03:55:06 koen: you need to set BBPATH, PYTHONPATH May 16 03:55:19 koen: edit the sample file in conf/ May 16 03:55:27 koen@bitbake:/data/build/oe/bbsvn/trunk/bitbake_qa$ set PYTHONPATH=/data/build/oe/bitbake/lib/ ;set BBPATH=/data/build/oe/bbsvn/trunk/bitbake_qa:/data/build/oe/openembedded May 16 03:55:30 koen@bitbake:/data/build/oe/bbsvn/trunk/bitbake_qa$ bin/bittest May 16 03:55:32 Traceback (most recent call last): May 16 03:55:35 File "bin/bittest", line 40, in ? May 16 03:55:37 from bittest import * May 16 03:55:40 File "lib/bittest/fileparser.py", line 36, in ? May 16 03:55:42 from bb import make May 16 03:55:45 ImportError: No module named bb May 16 03:56:17 koen: use PYTHONPATH May 16 03:56:37 zecke|sleep: I did that May 16 03:56:42 see the first line May 16 03:56:47 hrw|work: I'll be interested to see the log around that jump point... May 16 03:57:01 hrw|work: Just when its finished the run though May 16 03:57:37 RP: sure May 16 03:57:43 RP: 76% now May 16 04:00:26 73 May 16 04:01:02 koen: weird May 16 04:01:22 koen: and oe/bitbake/lib does hold bb? May 16 04:01:28 yes May 16 04:01:56 aha May 16 04:02:00 s/set/export/ May 16 04:03:06 ;) May 16 04:03:09 zecke: http://pastebin.ca/11880 May 16 04:03:13 next error May 16 04:03:42 * zecke let his machine run overnight and it went from f -> g with the testcase... May 16 04:03:56 koen: yeah you need the patched bitbake May 16 04:04:25 zecke: that wasn't in the manual :) May 16 04:04:40 The ide driver model changeover made it into 2.6.12-rc4-mm2 which is nice :) May 16 04:05:02 RP: I've made a new spitz-base patch, it should apply cleanly over the tosa-base one. May 16 04:05:04 koen: that is common sense ;) May 16 04:05:15 zecke: where can I download that? May 16 04:05:17 ;) May 16 04:05:44 koen: look into your oe mailinglist archive May 16 04:06:57 To whom it may concern: some version of gtk-industrial vanished May 16 04:07:08 aargh May 16 04:07:13 *again* May 16 04:07:38 zecke: any up to date place for the patches? May 16 04:08:15 preferably one big patch May 16 04:08:15 http://page.mi.fu-berlin.de/~freyther/bitbake_patches.tar.bz2 maybe? May 16 04:08:28 RP: 70% May 16 04:08:46 hrw|work: that one has the fetcher patch ? May 16 04:09:32 koen: dont know - just got it from history May 16 04:09:43 koen: no that oen doesn't have the fetcher patch May 16 04:10:43 * koen will wait till the patches hit svn May 16 04:14:41 2.6.12-rc4-mm2 still gives the same crash :-( May 16 04:17:18 RP: 65% May 16 04:17:27 RP: and 70% now May 16 04:20:02 hrw|work: This increasing percentage has me worried a bit :-/ May 16 04:20:18 RP: I know May 16 04:20:37 65% now May 16 04:21:09 and 70% back May 16 04:24:51 RP: and again.. 70-65-70 May 16 04:25:18 RP: battery voltage=187 - 65%, =188 - 70% May 16 04:29:35 65%-70% is keeping May 16 04:31:14 So a 1 unit battery voltage difference leads to the large percentage swings. I guess that makes some sense... May 16 04:31:42 I wonder if we can increase the resolution on the max1111... May 16 04:32:50 208 iirc is 100% May 16 04:34:41 hrw|work: Its 210 May 16 04:34:49 Is it mapped linearly? May 16 04:34:56 186 = 60% and now I have 60-65-60-65 May 16 04:35:36 * koen get's annoyed with the "printk: 1 message supressed" messages May 16 04:35:48 RP: any way to surpress those messages? May 16 04:42:41 lardman|busy: No, its not linear. Its looks like we have very little range to work with :-/ May 16 04:42:55 koen: When do you seen them? May 16 04:43:32 RP: when I kill X and use the console May 16 04:44:15 koen: Do you know which messages its supressing? May 16 04:44:40 stuff like IPv6 addrconf: prefix with wrong length 48 May 16 04:44:40 corgi_mci_setpower: off May 16 04:45:06 the first one was a rogue radvd on my network May 16 04:45:38 but it seems a bit backwards to surpress 1 message and print that it has been surpressed May 16 04:45:54 koen: I've never seen it try and supress them May 16 04:46:07 I plan to remove the corgi_mci_setpower: off anyway... May 16 04:46:24 The kernel is a bit too verbose May 16 04:47:31 oooh, it has console blanking May 16 04:47:31 I updated gnu-config.patch to today config.guess May 16 04:47:32 nice May 16 04:47:45 Why did Sharp wire up this ADC withh such bad measurement range? :-( May 16 04:47:56 RP: sharp May 16 04:48:12 0-100% is 170-210 "volts" :-/ May 16 04:48:43 3.4 - 4.2 V? May 16 04:49:50 NOTE:[13:50] package gnu-config-native-0.1cvs20050516: completed May 16 04:49:52 ;) May 16 04:51:53 is there a way not to have to compile gcc-cross-initial-3.4.3+csl-arm-20050416? May 16 04:52:03 no May 16 04:52:15 per definition you need a compiler to compile stuff May 16 04:52:24 makes sense May 16 04:52:33 it will only get built once May 16 04:52:42 is it normal the compile fails on a chroot debian? May 16 04:52:54 with an internal compiler error May 16 04:52:54 not that I know off May 16 04:53:03 ARGH... May 16 04:53:10 I cannot commit from work ;( May 16 04:53:13 how do I debug that? May 16 04:53:46 no idea on that May 16 04:54:25 bummer May 16 04:54:39 Frederik: I get that sometimes, last occasion was inside gnome-vfs May 16 04:54:44 3.2 - 4.0V I think... May 16 04:55:04 55% May 16 04:55:10 XorA: and what did you do about it? May 16 04:55:14 fR May 16 04:55:34 Frederik: ICE? try again May 16 04:55:40 Frederik: killed the plugin that caused the ice, when some headers moved about a couple of days later, it went away May 16 04:57:43 okay, doesn't sound like something I'm up for... May 16 04:58:10 compile on the iPaq it is in that case May 16 04:58:58 http://hrw.one.pl/tmp/config-guess-uclibc.patch - someone can push it? it's from packages/gnu-config/gnu-config/ - updated to 20050516 snapshot May 16 05:05:50 RP: 43% May 16 05:08:38 37% May 16 05:09:13 Don't you just love the non-linear discharge profile? May 16 05:09:32 31% May 16 05:09:43 maybe his battery is worn out. May 16 05:09:46 BigAl: I reaaly love it May 16 05:10:04 btw - screen is flashing from time to time now May 16 05:10:28 I have a vaio that shows 100% and after 50sec it shuts down without any warning :> May 16 05:10:54 RP: looks like backlight drop from 100% to lower and back.. 2-3times after few seconds May 16 05:11:21 and I have 25-43-25-50-25-50-31-25-50-25-43-25 jumps May 16 05:14:36 03zecke 07 * r1.3307.1.1 10openembedded/packages/ (3 files in 3 dirs): May 16 05:14:36 Opie Media Players: May 16 05:14:36 Add two new meta packages for special tasks (playing divx, May 16 05:14:36 playing ogg+mp3) May 16 05:14:36 Add these two packages to the task-opie-multimedia in May 16 05:14:37 meta-opie.bb May 16 05:14:52 CoreDump|home: is libsdl1.2.7 new? May 16 05:15:06 zecke: dunno, didn't check May 16 05:15:53 lol May 16 05:15:57 gints|wrk: :) May 16 05:16:43 libsdl-qpe_1.2.7.bb 1.35 CSets 13 weeks schurig changed section May 16 05:17:19 ipaq.patch 1.5 CSets 64 hours zecke Patch by Gints Polis May 16 05:17:22 :) May 16 05:17:24 CoreDump|home: pssst May 16 05:17:36 hehe May 16 05:22:56 gints|wrk: how did you create that crappy patch? May 16 05:26:02 03zecke 07 * r1.3314 10openembedded/packages/libsdl/libsdl-qpe-1.2.7/ipaq.patch: May 16 05:26:02 libsdl-sqp: May 16 05:26:02 Change the ipaq.patch to be applyable. I thought times of May 16 05:26:02 manually invoking diff on two directories are over... May 16 05:26:07 CoreDump|home: please try again May 16 05:29:38 heh May 16 05:30:08 "It seems that the suspend works (with apm -s). But how resuming the ipaq ?" May 16 05:30:23 looks like 50% of s/r works on the h2210 May 16 05:30:56 hrw|work: There is a limiter in the code to drop the backlight power when the power is low :-/ May 16 05:31:44 Still, Given we'll always have that in there, I guess we can skew the percentages accordingly... May 16 05:31:57 RP: ah.. May 16 05:32:10 now have 22-37-43-22-37 May 16 05:38:15 RP: logread stopped working when opie started to complain about lack of power.. May 16 05:38:33 RP: now I have 4% May 16 05:41:26 RP: I will connect power and send you file May 16 05:41:53 but first charge must May 16 06:01:56 hrw|work: Thanks. I'll write something to extract the voltages and timestamps, plot them and see what we get :) May 16 06:07:26 ok May 16 06:07:39 ok. husky started May 16 06:13:03 RP: http://www.hrw.one.pl/tmp/battery.log.txt.bz2 May 16 06:15:45 RP: method needs changing... last entries has same time May 16 06:16:05 can anyone recommend a jtag-cable schematic? May 16 06:16:25 roh: for casio watch? May 16 06:16:41 for a simpad May 16 06:16:58 ~google simpad jtag cable May 16 06:17:06 Google says simpad jtag cable is at http://opensimpad.org/wiki/index.php/Docs/JTAGHowTo or http://externe.net/zaurus/simpad-bluetooth/ or http://www.handheld-magazin.de/board/tablett/286614.php3 or http://www.handhelds.org/hypermail/bootldr/author.html or http://www.handhelds.org/hypermail/bootldr/subject.html May 16 06:17:43 i found a passive cable. but do not know if thats recommend or if it's better do build a active one May 16 06:17:53 asked on simpad forums? May 16 06:18:03 hrw|work: thanks May 16 06:18:40 RP: no problem - hope it will help May 16 06:29:55 openoffice.org die when have something real to do... May 16 06:36:03 morning May 16 06:37:02 hi chouimat, Crofton May 16 06:37:19 morning May 16 06:39:28 03koen 07 * r1.3315 10openembedded/packages/gnu-config/gnu-config/config-guess-uclibc.patch: May 16 06:39:28 config-guess-uclibc.patch: May 16 06:39:28 update patch, courtesy Marcin 'Hrw' Juszkiewicz May 16 06:39:37 thx koen May 16 06:39:40 np May 16 06:39:45 thanks for fixing it May 16 06:40:24 np May 16 06:42:08 ljp: ping May 16 06:43:48 hrw: The result: http://www.rpsys.net/openzaurus/temp/batt-per-vs-time.gif May 16 06:44:01 That's battery percentage vs. time May 16 06:44:39 looks like the S curve I expected May 16 06:44:57 mickey|boardgame: it seems usbserial or something was added, and it messed up things. I'll have to let you know another time regarding kstars, as I'm not experiencing a bit many things May 16 06:45:33 XorA: The code is suppsed to smooth it out :-/ May 16 06:46:01 It does resemble what we've come to expect from the device though through experience May 16 06:47:01 RP: table lookup, or someone built a x^n function into the kernel? May 16 06:49:06 RP: How exactly did you measure it ? The occurence of several percentages a one time is strange, isnt it ? Also you should not think, that this is an optimized calibration curve, as batteries discharge cycles change over time and dirscharge cycles. May 16 06:50:47 RP: BTW, I let my battery drain over lunch time. After switching back to life, the battery (opie) status says 98% and NOT charging. Thats strange. May 16 06:52:10 RP: Disconnected the power jack, 65%, reconnected it, now charging ... May 16 07:03:40 XorA: Its done by a lookup table and linear interpolation May 16 07:04:10 uv1: At any one time there is only one percentage May 16 07:04:37 uv1: Its taken directly from the output of the corgi_power code May 16 07:06:44 RP: What is the time scale ? msecs ? May 16 07:06:56 uv1: jiffies May 16 07:07:29 RP: And you say that the value is bouncing between two discrete values now and then ? May 16 07:07:48 RP: Can we see a charge curve as well here ? May 16 07:08:13 03koen 07 * r1.3316 10openembedded/packages/python/python-pygtk_2.6.2.bb: python-pygtk_2.6.2.bb: courtesy Hermann Lacheiner May 16 07:09:00 RP: And would apm not show the discharge cycle as well, lets say, if you ask it every 5 secs ? May 16 07:09:35 uv1: You could use apm if you wanted to produce the same thing May 16 07:09:54 I wanted to see any other power logging with it though... May 16 07:11:15 http://www.rpsys.net/openzaurus/temp/volt-vs-time.gif May 16 07:11:30 actual battery voltage vs time this time. May 16 07:12:13 I've just realised some of the code I wiped out from the driver was to always force the percentage to the most pessimistic value and stop the bouncing :-/ May 16 07:19:03 The code also does heavy averaging for high battery voltages but not low ones :-/ May 16 07:19:43 I can certainly use this info to make the charging curve nicer May 16 07:31:29 03koen 07 * r1.3317 10openembedded/packages/python/python-pygtk_2.6.2.bb: python-pygtk_2.6.2.bb: stage python-codegen May 16 07:36:31 ~lart libtool May 16 07:36:32 * ibot flings poo at libtool May 16 07:36:33 arm-linux-libtool: link: CURRENT `1000' is not a nonnegative integer May 16 07:46:27 RP: When does the device stop charging ? What is the condition ? apm says 100% since 30 mins now, but still charging. I remember, that a buggy charging code once destroyed my ipaq3630 batteries ;-( May 16 07:46:59 uv1: It charges until it recieves a hardware interrupt May 16 07:47:08 It will then restart charging May 16 07:47:24 If it recieves the same interrupt within a certain time frame, it stops May 16 07:47:37 RP: OK, waiting ... May 16 07:47:54 uv1: Turn on msglevel2 to see what its doing May 16 07:48:04 RP: how ? May 16 07:50:47 echo 2 > /sys/devices/platform/corgi-battery/msglevel May 16 07:52:06 RP: Where does it show up then ? logread ? May 16 07:53:16 yes or dmesg etc. May 16 07:55:03 Does anyone know here, whether gnuplot work with opie (and how the package is named) ? May 16 07:55:50 hi guys May 16 07:55:55 uv1: qtplot May 16 07:55:57 hi Spyro May 16 07:56:02 Im having trouble building bootstrap-image May 16 07:56:08 | + ipkg-cl -f /home/ian/projects/openembedded/stuff/tmp/work/bootstrap-image-1.0-r0/temp/ipkg.conf -o /home/ian/projects/openembedded/stuff/tmp/rootfs install task-bootstrap May 16 07:56:08 | ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied May 16 07:56:23 Spyro: ignore May 16 07:56:23 koen: qtplot handled like gnuplot ? May 16 07:56:35 hrw|work: I would, but it bombs out at that opint May 16 07:56:36 koen@bitbake:~/OE/openembedded/packages/gnome$ ls ../gnuplot/ May 16 07:56:36 SCCS files gnuplot_4.0.0.bb qtplot_0.2.bb May 16 07:56:39 no idea May 16 07:56:48 any qt* sounds like a ugly hack to me May 16 07:57:08 |lart koen May 16 07:57:14 lets try ... May 16 07:57:58 BTW: evas-x11_0.9.9.004.bb:57:unparsed line ... Thats new. May 16 07:58:07 uv1: read the mailinglist May 16 07:58:11 uv1: add \ in --with-x line May 16 07:58:22 I have fixes for it at home May 16 07:58:32 | + ipkg-cl -f /home/ian/projects/openembedded/stuff/tmp/work/bootstrap-image-1.0-r0/temp/ipkg.conf -o /home/ian/projects/openembedded/stuff/tmp/rootfs install task-bootstrap May 16 07:58:32 | ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied May 16 07:58:32 | Nothing to be done May 16 07:58:32 | An error ocurred, return value: 4. May 16 07:58:32 | Collected errors: May 16 07:58:34 | Cannot find package task-bootstrap. May 16 07:58:36 | Check the spelling or perhaps run 'ipkg update' May 16 07:58:38 NOTE: Task failed: May 16 07:58:40 NOTE: package bootstrap-image-1.0-r0: task do_rootfs: failed May 16 07:58:42 ERROR: TaskFailed event exception, aborting May 16 07:58:44 NOTE: package bootstrap-image-1.0: failed May 16 07:58:46 ERROR: Build of bootstrap-image failed May 16 07:59:01 koen: qtplot want gnuplot, wants x11 ;-( May 16 07:59:06 Spyro: Have a look at the do_rootfs log May 16 08:00:11 uv1: only at buildtime May 16 08:00:28 uv1: read the .bb and see http://www.mneuroth.de/privat/zaurus/gnuplot.html May 16 08:00:57 RP: there are several - I am looking at the highest numbered variant May 16 08:01:12 RP: what am I looking for. May 16 08:01:28 Spyro: Each time you run bitbake you'll generate a new one. Look at the timestamp for most recent May 16 08:01:50 ok so what am I lokoing for anyhow ? May 16 08:01:58 You're looking for an error, probably at the of the file May 16 08:02:24 If you can't see it, post the logfile somewhere (pastebin?) May 16 08:03:17 wheres pastebin again ? May 16 08:03:52 http://pastebin.ca May 16 08:04:51 http://pastebin.ca/11896 May 16 08:19:04 bye all May 16 08:20:43 anyone got any idea as to the problem above ? May 16 08:22:46 Spyro: Its trying to install a package called "task-bootstrap" which doesn't sound right to me. You need to work out why... May 16 08:23:57 what makes you say its trying to install it ? it just says it cant find it? May 16 08:25:37 ipkg-cl -f /home/ian/projects/openembedded/stuff/tmp/work/bootstrap-image-1.0-r0/temp/ipkg.conf -o /home/ian/projects/openembedded/stuff/tmp/rootfs install task-bootstrap May 16 08:25:54 Its trying to install it - then fails May 16 08:26:21 why is it using ipkg anyway ? I didnt ask it for any ipkgs... May 16 08:26:37 * chouimat is away: food May 16 08:26:40 The whole thing is built with ipks May 16 08:27:00 ok I looked int he bb file for bootstrap_image and it says May 16 08:27:10 export IPKG_INSTALL = "task-bootstrap" May 16 08:27:25 is that wrong? I uncommented it and it succeeded but I still only have half a rootfs May 16 08:27:46 If its there, it needs it. but it isn't getting generated May 16 08:28:06 Have a look at the task-bootstrap logs and see if there were any errors May 16 08:30:32 koen, i put both the cvs and the targeted e bb up for you after kergoth showed me the errors of my ways May 16 08:31:01 hi May 16 08:31:28 hi Bernardo May 16 08:36:26 not sure what hrw|gone's issue is with it going to oe-devel May 16 08:54:13 RP: My CL860 is still charging. Are you sure, that you always will get an interrupt ? May 16 08:55:01 mine sometimes also charges for a long time (like all night) May 16 08:56:54 uv1: What does the voltage drop to if you take it off charge? May 16 08:57:31 I can confirm mine does stop eventually. They do seem to take an age when finally topping the charge up... May 16 08:59:40 RP: While charging 100%, Voltage : 221 May 16 09:00:07 RP: While charging 98%, Voltage : 208 May 16 09:00:28 RP: So its possibly nearly over ;-) May 16 09:01:40 RP: Hopefully they know hao to handle a battery. Possibly they want to sell new ones ? ;-( May 16 09:02:02 uv1: 210 is supposedly 100% May 16 09:02:28 RP: How is the "Status" in msglevel 2, when full ? May 16 09:02:48 so close but not 100% yet. The charging monitors temperature so I'd not expect too many problems he says crossing fingers May 16 09:02:59 uv1: Its not as simple as that. May 16 09:03:24 Full is when the CO interrupt occurs within 10 minutes of each other May 16 09:03:58 How can I see that in apm / msglevel2 ? -1% ? May 16 09:04:49 uv1: Just by watching for the CO interrupts and timestamps May 16 09:05:15 RP: So I assume -1% will never be seen ? May 16 09:06:24 I don't know what the -1% is... May 16 09:06:33 Sorry, have to leave for 30 mins ... May 16 09:07:14 RP: We had the -1% when coming to the prob with "battery level critical". You said, that this is set, when charging ... May 16 09:08:07 I think I since altered the battery code but I'm not sure May 16 09:11:33 uv1|dinner: Its not "-1%". It says -1 ? which means it can't work out how long the system has left to run May 16 09:11:45 The percentage should always be there now... May 16 10:23:04 * chouimat is away: Escaping manking stupidity for the afternoon May 16 10:24:10 * chouimat is away: Escaping mankind stupidity for the afternoon May 16 10:54:40 hmm, how do I get bitbake to fetch files and apply patches, not to configure and compile? May 16 10:55:50 bitbake -c patch May 16 10:56:17 Thanks May 16 10:57:35 I need to updagrade my osk kernel, but I want to run the kernel config by hand to check for new configure options May 16 11:04:59 good evening May 16 11:07:19 hi May 16 11:07:35 just wondering how the memory management aspects of OE are being addressed May 16 11:09:17 there are various third party patches out there that decrease memory consumption May 16 11:09:56 any recommendations for what to try? May 16 11:10:33 bitbake -c unpack -b kernelbb May 16 11:11:31 Crofton: the kernel configure will bail if it finds unselected options May 16 11:11:41 so a bitbake -c configure -b kernelbb is generally safe May 16 11:11:52 then you can re-do the .config on your own May 16 11:11:58 ah May 16 11:12:10 Note that the prior isn't always true - busybox configure will just hang May 16 11:12:27 Or at least it has in the past - I haven't screwed up a busybox config in a while May 16 11:12:47 CosmicPenguin: no infinitly looping and filling up your disk anymore? May 16 11:12:54 jjg: I'd expect the memory position to be a lot better in a week or two. Things are being done... May 16 11:13:01 treke: yeah, thats bad news... :) May 16 11:13:19 Usually, if busybox goes more then about 30 seconds on my system, thats bad May 16 11:13:41 zecke needs to wake up so that I can pester him May 16 12:05:49 Ciao all May 16 12:06:09 hey Pigi May 16 12:06:29 hi koen. Welcome back ! May 16 12:07:04 does any one know how /etc/sdcontrol is triggered on a collie? May 16 12:07:15 by sdmgr. May 16 12:08:13 * treke is about -> <- close to giving up on trying to make opie do something useful May 16 12:08:17 :p May 16 12:09:20 if you want a pda running linux, try opie, if you want to run linux on a pda, try something else May 16 12:09:37 that's the shortest explanation I encountered May 16 12:09:37 ? May 16 12:10:02 that made absolutly no sense at all May 16 12:10:07 alhtough without context it is kinda strange May 16 12:10:19 thx kergoth, sdmgr seems to be b0rked in recent build. sdcontrol isn't beeing run May 16 12:15:14 03koen 07 * r1.3318 10openembedded/packages/meta/meta-sectest-gpe.bb: add hydra, thcrut and driftnet to the sectest image May 16 12:23:46 any volunteers for upgrading gnome to 2.10? May 16 12:24:31 There's a fun three weeks May 16 12:25:23 indeed May 16 12:26:03 I tried adding gnome-python 2.10 and got stuck on a libtool thing in libgnome-2.10 May 16 12:28:28 pastebin it May 16 12:30:11 CosmicPenguin: libtool chokes on LIBGNOME_AGE=1000 complaining 1000 is not a nonnegative integer May 16 12:30:25 changing 1000 to 100 makes it compile May 16 12:30:41 but I have no idea what it's doing or supposed to do May 16 12:30:48 heh May 16 12:31:41 koen: sounds like your typical gnome issue May 16 12:41:34 ew wtf. busybox's "test" doesn't work when sdcontrol is called by sdmgr. It does work as expected when called manually. Any ideas? May 16 12:45:33 thought so :) May 16 12:46:00 building from kde-cvs is a pain May 16 12:55:38 hi May 16 12:55:40 emte: ping May 16 12:59:45 treke: good to know that kde is keeping up with gnome in that respect May 16 13:01:02 at least with gnome you can figure out how to build a single app :p May 16 13:06:36 heh May 16 13:08:42 re May 16 13:08:54 re May 16 13:09:07 yay zecke shows his face :p May 16 13:09:52 zecke: you don't happen to have some magical working kitchensync hiding anywhere, do you? May 16 13:10:07 been trying to get the svn code to build for a half an hour now :p May 16 13:10:29 ~seen ljp May 16 13:10:38 ljp is currently on #zaurus #opie #oe #elinux #handhelds.org #openzaurus May 16 13:11:05 someone tried to build qt/e with freetype support? May 16 13:14:54 I think it's been tried a number of times May 16 13:15:02 treke: oh it built yesterday May 16 13:15:10 * chouimat is back. May 16 13:15:18 treke: KitchenSync itself works, it is only kdepim being fscking broken May 16 13:15:20 zecke: Working out how to build it is the problem :p May 16 13:15:39 treke: cd kdepim && unsermake May 16 13:15:53 treke: or make instead of unsermake if you don't use unsermake May 16 13:16:02 Can you build just kitchensync somehow? May 16 13:17:50 treke: hardly you need to change some Makefile.am May 16 13:17:58 hehe May 16 13:18:03 yeah spent a while fiddling with that May 16 13:18:09 treke: to not use the built libkcal.la, libkdepim.la May 16 13:18:10 maybe I'll just rebuild full kdepim debs May 16 13:18:19 treke: look what I did with my seperate packages May 16 13:18:31 yeah thats what I was using as a reference May 16 13:18:54 my first goal is to get two instances of korganizers and kaddressbook to sync with each other May 16 13:19:00 than I'm going to take on the zaurus May 16 13:20:18 treke: if you actually start with it and run into problems, please get back to me May 16 13:20:22 treke: we will work it out ;) May 16 13:20:37 hehe May 16 13:20:53 I figured maybe the svn code works better May 16 13:21:40 for some reason syncing with a local and a remote konnector was just behaving randomly May 16 13:22:04 sometimes it would lose the text of all entries, others it wouldn't see the entries, and other times it would just lose one May 16 13:22:10 treke: it works better May 16 13:22:21 treke: until someone breaks kcal and kabc classes again May 16 13:22:25 hehe May 16 13:22:29 that a common occurance? May 16 13:22:47 treke: sadly the Resource Framework is fscking broken May 16 13:23:18 hehe, maybe it will get fixed in 4 :p May 16 13:23:52 treke: syncing is dead simple, but you need to have some knowledge... May 16 13:30:25 "dead simple" might be stretching it May 16 13:31:40 CosmicPenguin: no kidding it is May 16 13:31:58 CosmicPenguin: there are certain levels of perfection though May 16 13:32:26 In my experience, its very difficult - thats why nobody does it very well May 16 13:33:49 CosmicPenguin: I've written one simple synchronisation algorithm it works with everything May 16 13:42:18 gnu-config doesn't build today May 16 13:42:25 CosmicPenguin: bk pull May 16 13:42:36 heh May 16 13:42:53 CosmicPenguin: patch needs CVSDATE>=20050515 May 16 13:42:59 CosmicPenguin: http://bugs.openembedded.org/show_bug.cgi?id=2 May 16 13:44:41 * CosmicPenguin kicks OE May 16 13:45:41 good night all May 16 13:45:44 .nick hrw|gone May 16 13:45:45 'night hrw May 16 13:45:59 zecke: any timeframe for you bb patches to hit svn? May 16 13:46:11 mickeyl: ping May 16 13:46:16 s/you/your/ May 16 13:46:31 koen: I'm currently 'crafting' a Turing-Machine May 16 13:48:38 hrw|gone: hmm... your patch definately doesn't apply with CVSDATE-20050515 May 16 13:49:11 s/>=/>/ May 16 13:49:14 koen: and I will chat with mickeyl afterwards May 16 13:50:34 * koen reads http://www.mozilla.org/projects/tinderbox/ May 16 13:51:01 koen: look at the perl - argh *shouder* - sourcecode May 16 13:51:18 * CosmicPenguin grumbleas May 16 13:51:20 I tend not to May 16 13:51:29 koen: all you need to know there is one server holding the results May 16 13:51:32 looking at interpreted code May 16 13:51:48 koen: and clients running tests and sending the results to the server May 16 13:52:21 hmmm, I aded a new bb file for a new kernel version May 16 13:52:39 I built the kernel, cleaned task-bootstrap, and rebuilt opie-image May 16 13:52:48 now depmod -a gives May 16 13:52:57 ARNING: Couldn't open directory /lib/modules/2.6.12-rc2-omap1: No such file or directory May 16 13:52:57 FATAL: Could not open /lib/modules/2.6.12-rc2-omap1/modules.dep.temp for writing: No such file or directory May 16 13:53:07 where rc2 is the old kernel May 16 13:53:11 what have I missed? May 16 13:53:46 forget I asked the question May 16 13:53:52 Crofton: did you get my little complains about your bb files? May 16 13:53:53 * Crofton blushes May 16 13:53:56 * CosmicPenguin smacks his head May 16 13:54:01 today is the fucking 16th May 16 13:54:01 via email? May 16 13:54:10 * CosmicPenguin throws his watch out the window May 16 13:54:11 Crofton: no via irc May 16 13:54:24 I don't think so May 16 13:54:37 Crofton: besides that there is little preventing me pushing your changes May 16 13:54:41 I may have seen them, but I think I was wrapped up in something else May 16 13:54:42 ok May 16 13:54:50 Crofton: but probably some guy actual having an OMAP board should device May 16 13:55:09 I have a omap5912osk May 16 13:55:22 Crofton: it was mainly a complaint to the ORB stuff, but I don't remember it exactly May 16 13:55:28 what day did you make them? May 16 13:55:47 Crofton: one or two days ago, on this channel May 16 13:55:57 I'll search the logs May 16 13:56:11 If I change them, I'll end up with a new changeset though May 16 13:56:23 right ? In addition to the first May 16 13:57:06 yes May 16 13:57:12 ok May 16 13:57:27 Crofton: I would prefer someone else looking at the patches as well May 16 13:57:39 Crofton: but I'll push them little by little May 16 13:57:59 OK, I would like to get them in, the orb stuff won't break anything May 16 13:58:16 the machine stuff should be "safe" too May 16 13:58:38 I am a little worried about the site-config stuff, but I don't think it will lead to bustage May 16 13:59:14 there are about 3 or 4 people interested in using oe to develop images for this board May 16 13:59:20 Crofton: I won't push it for i686, or was your change intended May 16 13:59:35 Crofton: but the arm stuff looks ok May 16 13:59:46 I am thinking there was a configure issue for some of the native builds May 16 14:00:28 if you get stuff pushed, I can recreate the reason I made the 686 change May 16 14:02:15 bother rikers.org seems down May 16 14:02:21 that is were the logs are I think May 16 14:03:41 ~logs May 16 14:03:51 apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" May 16 14:07:07 ok found them May 16 14:09:23 I'll go through the comments tomorrow, need to flash the new kernel in the OSK so the modules stop barfing :) May 16 14:20:30 re May 16 14:21:05 wb mickeyl May 16 14:23:33 mickeyl: do we want to commit the first two patches tonight? May 16 14:23:50 zecke: yes! yes! yes! May 16 14:24:11 * koen should cut down on sugar intake May 16 14:24:21 zecke: yeah. you got access now? May 16 14:24:24 /nick beavis May 16 14:24:29 mickeyl: yes May 16 14:24:35 ok. May 16 14:24:43 Yeah Yeah Bitbake Patches, Yeah Yeah HeHe... May 16 14:24:43 will you handle the bitbake one and I'll do the oe ones ? May 16 14:25:08 oh do we want to make oe dependant on trunk directly? May 16 14:25:18 i thought so May 16 14:25:37 ah or do you want to put out a release before ? May 16 14:25:41 bitbake, that is May 16 14:25:45 mickeyl: exactly May 16 14:25:51 fine with me as well May 16 14:26:11 let me finish my Turing Machine first May 16 14:26:14 k May 16 14:32:27 hi mickeyl May 16 14:32:39 good evening pb_ May 16 14:32:51 zecke: heh, you are constructing a turing machine? May 16 14:40:39 pb_: Don't ask May 16 14:41:29 shit happens when you're too scared to knock the Profs Office door to get a test date fixed and take the class a second time... May 16 14:42:15 heh May 16 14:42:23 * pb_ hands zecke a large reel of tape May 16 14:42:33 Introduction to the Theory of Computation... May 16 14:42:47 my issue is I don't believe the Church Thesis about computation May 16 14:42:52 how big is an infinitly long reel of tape? May 16 14:43:14 not as big as you might think May 16 14:43:15 treke: not more poweful than two infinitly long reels of tape May 16 14:43:35 treke: start counting and get back to me tomorrow May 16 14:45:08 doh. I forgot to start my kdepim build before lunch May 16 14:46:53 oh dear. kdepim famously doesn't build properly in the afternoon. May 16 14:46:59 * koen hands treke a new lunch break May 16 14:47:14 that sounds like a winner May 16 14:47:18 'night all May 16 14:48:06 zecke: what's your disagreement with Church? May 16 14:48:23 usually too long and boring May 16 14:48:27 and all of that kneeling May 16 14:55:23 Good evening! May 16 14:55:49 ~ugt May 16 14:55:51 [ugt] Universial Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. May 16 14:56:01 please try again May 16 14:56:27 good morning !!! May 16 14:56:52 pb_: I've never seen a nondeterministic with infinite amount of memory in reality May 16 14:57:03 working with people in multiple timezones takes some getting used to May 16 14:57:19 pb_: I've seen lambda kalkül though May 16 14:57:29 I am just struggeling with flashing a WLAN-Card! May 16 14:57:41 very good May 16 14:57:46 beiss: hi May 16 14:57:59 someone please remind we what to do to fix a May 16 14:58:05 | configure.ac:119: error: possibly undefined macro: AC_DEFINE May 16 14:58:15 | configure.ac:274: error: possibly undefined macro: AM_PATH_LIBGNUTLS May 16 14:58:17 ... May 16 14:58:42 When trying to use prism2_srec I get "Are you sure you have compiled (and loaded the correct version of) hostap.o module with PRISM2_DOWNLOAD_SUPPORT definition in driver/module/hostap_config.h?" May 16 14:59:27 Does anybody know if PRISM2_DOWNLOAD_SUPPORT and PRISM2_NON_VOLATILE_DOWNLOAD is enabled? May 16 14:59:53 mickeyl: for the latter, you probably need to teach gnutls to install its .m4 files into staging/aclocal. May 16 14:59:58 I'am Using collie with DCF-660W May 16 15:00:14 I don't know what's going on with the former, though. My usual reaction is to run screaming from the room when that kind of thing happens. May 16 15:00:39 pb_: thanks. same here, i'm really trying not to learn too much about autohell May 16 15:01:08 AC_DEFINE is meant to be an autoconf builtin. May 16 15:01:22 beiss: what kind of card is that? prism3 with volatile 2nd and tertiary firmware RAM ? May 16 15:01:40 pb_: so perhaps our autoconf version is too new/too old / or something May 16 15:01:48 ahh May 16 15:01:52 a moment May 16 15:02:29 mickeyl: yah, could be May 16 15:02:42 hmm. ok i'add that package to nonworking May 16 15:02:43 :) May 16 15:03:04 hostap_diag wlan0 says: NICID: id=0x800c v1.0.0 (PRISM II (2.5) PCMCIA (SST parallel flash)) May 16 15:03:16 aha May 16 15:03:27 to my knowledge, prism2 cards don't use firmware flash May 16 15:03:37 s/flash/in RAM/ May 16 15:03:50 except the infamous socket/symbol one May 16 15:03:55 they can do. May 16 15:03:56 and the Intel pendant May 16 15:04:10 I think you can soft-load firmware into ram on all prism2s, though you don't normally do that. May 16 15:04:38 ah right. they come with secondary and tertiary in flash but you can override those versions May 16 15:05:01 One of my laptops is set up to always softload a particular version of the tertiary firmware, so I can use hostap mode without having to worry about what's in flash. May 16 15:05:09 yah, exactly. May 16 15:05:11 makes sense May 16 15:05:42 we enable volatile RAM download in our hostap build, we don't enable non-volatile download May 16 15:05:56 this one is better done on a PC or a laptop imo May 16 15:06:00 right May 16 15:06:11 non-volatile download is all a bit scary. May 16 15:07:00 but the DCF-660W ist a cf-Card, but I have no cf-card-slot in my PC May 16 15:07:26 that's unfortunate. May 16 15:07:37 what do you want to do to your dcf-660w ? May 16 15:07:48 make it work or upgrade it? May 16 15:08:01 in that case, you are doomed to either (a) spend five euros on a cf->pcmcia adapter, (b) get another card, (c) get another PC, (d) live with soft downloads, or (e) build your own hostap with the nonvolatile download anabled. May 16 15:08:15 I want to Upgrae the firmware, becaus I want to use wpa May 16 15:08:28 ah. then see pb_'s options May 16 15:08:58 how can realize soft downloads? May 16 15:09:20 see the documentation for prism2_srec. May 16 15:09:26 you just need to tell it to download to RAM, not to flash May 16 15:10:10 03mickeyl 07 * r1.3317.1.1 10openembedded/packages/ (3 files in 3 dirs): upgrade python-pysqlite2 to 2.0.0 (final). add lftp 3.2.0 to nonworking, bails out with undefined AC macros May 16 15:10:11 assuming you have 50k spare in your rootfs, you can just add an appropriate prism2_srec command to your pcmcia scripts or /etc/network/interfaces, and it will all work transparently. May 16 15:10:35 ls -la May 16 15:10:36 we do exactly that with the sandisk cards May 16 15:10:42 take a look at prism3-support package May 16 15:11:44 when I try prism2_srec -v -r wlan0 sf010704.hex I get image is not meant to be downloaded to volatile memory. May 16 15:11:56 right, you need a different image. May 16 15:12:19 go back to whereever that file came from, and select one of the ram-based images. May 16 15:15:35 zecke: i need to go to bed. if you need me to help with the upgrade, we must do it tomorrow May 16 15:15:42 g'night guys May 16 15:15:48 mickeyl: good night May 16 15:19:04 Great. Thanks for your help. May 16 15:19:54 mickey|zzZZzz: cya May 16 15:20:43 But it would be a great thing if everybody could be able to flash the firmware. Why not enable non-volatile download? May 16 15:22:36 Because, although it would be a great thing for everybody to be able to flash their firmware, it would not be so great if everyone's card was destroyed as a result. May 16 15:23:31 Nobody knows whether the firmware downloader works on these platforms, and (understandably) nobody is especially keen to be the first person to try it out. May 16 15:25:45 guys... how does one set a blank password in /etc/passwd ? May 16 15:28:42 damn mickey May 16 15:28:50 anyone know why the directfb do_stage() is this: May 16 15:28:53 oe_runmake 'DESTDIR=${STAGING_LIBDIR}/' install May 16 15:30:21 that does seem odd May 16 15:30:58 CosmicPenguin: he went the easy road and dumped everything there May 16 15:31:23 but that sort of freaks out everything else, doesn't it? May 16 15:31:43 CosmicPenguin: that depends May 16 15:32:01 well, more to the point, you'd kind of expect that it would result in creating staging/arm-linux/lib/include, staging/arm-linux/lib/lib/ and so on. May 16 15:32:33 which surely isn't what you'd want. May 16 15:33:05 there's plenty of precedent for do_stage() just being a directed "make install"; almost all the X packages do that, for example. May 16 15:33:36 indeed May 16 15:37:13 OOPS May 16 15:37:27 treke: ! May 16 15:38:10 anyone building for mipsel here? which gcc/glibc/binutils combinations are known-good? May 16 15:38:25 I forgot to give this machine swap space May 16 15:38:33 heh May 16 15:38:50 building kdepim sort of killed it May 16 15:39:49 treke: swap? I though we've a paging system :} May 16 15:43:14 guess I'l try compiling at home May 16 15:44:04 bah - screw this May 16 15:44:14 zecke: I heard that paging was overrated. May 16 15:44:35 actually, people have been talking about adding classical swapping to linux for years. May 16 15:44:42 afaik, nothing has ever come of it, though. May 16 15:44:48 Easier to figure out the headers then to try to add all the random defines that make install needs May 16 16:06:35 'night all May 16 16:12:16 * CosmicPenguin just realizes that it shouldn't matter where the directfb headers are May 16 16:29:44 * CosmicPenguin thinks there's something evil about a directfb program that uses X11 header files May 16 16:38:11 bleh May 16 16:38:27 I'm just about to walk out the door and my kde build finishes May 16 16:39:59 no going home for you May 16 16:51:22 dammit May 16 16:51:24 still broken May 16 18:29:37 well this sucks May 16 18:29:41 someone hit my nice new car May 16 18:29:55 scraped the shit out of the passenger side May 16 18:30:07 thankfully, new cars come with full coverage ... ? May 16 18:38:55 hey hrw|gone again May 16 18:49:05 yup. downstairs neighbor May 16 18:55:01 atleast you dont have to go far to break thier leggs May 16 18:55:09 :) May 16 19:09:55 * france is back (gone 20:45:31) May 16 19:36:30 * chouimat|happy is away: shower May 16 19:49:15 * chouimat|happy is back. May 16 20:11:29 night May 16 20:24:32 HA May 16 20:24:42 my jtag seems to work May 16 21:06:59 heh... done it. my jtag-rescue attempt succeeded. i have a new working bootloader now May 16 21:08:59 roh: Well Done! May 16 21:12:08 hey raster you remember which header clears up the "pointer to int comparison" warning? May 16 21:22:53 heh.. now i need to wait for serial ymodem upload of the 16mbyte-jffs2 ;) May 16 23:05:01 hi .everyone May 16 23:07:02 hey May 16 23:22:56 hi emte May 16 23:23:01 hey May 16 23:25:19 i want to use my local arm-linux-gcc.I set the PATH of arm-linux-gcc and ASSUME_PROVIDED = "virtual/arm-linux-gcc" in local.conf .is it right? May 16 23:30:04 emte .can you help me May 16 23:31:07 what do you mean by local? May 16 23:31:30 oe creates a local crosscompiler for itself May 16 23:31:44 based on device requirements May 16 23:31:55 by default May 16 23:32:37 unless your really convinced you need to bypass all the device patches you might need May 16 23:32:48 yes ,but now I want to use the one that my board use May 16 23:34:14 you will probably need to set an absolute path to the gcc you want to use May 16 23:34:20 not the virtual May 16 23:34:33 so i can't use the arm-linux-gcc oe create for me May 16 23:34:50 oh May 16 23:34:56 virtual will be the local one taht oe builds May 16 23:35:11 oh May 16 23:35:21 thanks May 16 23:36:55 like ASSUME_PROVIDED = "/root/usr/local/arm/2.95.3/bin/arm-linux-gcc" ? May 16 23:37:34 that would be my guess yes May 16 23:37:56 and it has to have permission ofcourse May 16 23:37:57 let me tyr May 16 23:37:59 try May 16 23:37:59 ASSUME_PROVIDED = "virtual/arm-linux-gcc-2.95" May 16 23:38:03 hi May 16 23:38:13 and ADD it into PATH May 16 23:38:16 hrw, for a non-oe gcc? May 16 23:38:16 hi hrw May 16 23:38:28 emte: exactly May 16 23:38:34 ~zauruskernels May 16 23:38:36 it has been said that zauruskernels is By default Zaurus use 2.4.18 and you need GCC 2.95.3 to compile them - all is described in OE wiki: http://openembedded.org/cgi-bin/moin.cgi/ZaurusKernels There is also work on 2.6: http://openembedded.org/cgi-bin/moin.cgi/ZaurusKernels26 May 16 23:38:55 thought virtual limited it to inside the oe build system May 16 23:39:28 assume_provided is for giving outside OE stuff for OE May 16 23:39:42 atleast I understand it like that May 16 23:40:27 emte: evas/evas-x11-fb_0.9.9.004.bb: --with-x May 16 23:40:37 emte: forgot \ May 16 23:41:07 forgot which? May 16 23:41:13 oh May 16 23:41:25 bad me those were copy-paste May 16 23:41:30 bitbake gives parse errors May 16 23:41:47 hrw: May 16 23:41:49 PATH=/usr/local/arm/2.95.3/bin:$PATH May 16 23:41:50 ASSUME_PROVIDED = "virtual/arm-linux-gcc" May 16 23:41:52 is that right? May 16 23:41:55 yes May 16 23:42:09 robert: ASSUME_PROVIDED = "virtual/arm-linux-gcc-2.95" May 16 23:42:33 robert: or you want external gcc 2.95.3 as default one? May 16 23:42:43 sorry about that one hrw May 16 23:42:45 yes May 16 23:43:09 emte: np - I fixed that in local tree May 16 23:43:22 robert: set DISTRO="sharprom-compatible" May 16 23:43:40 hrw:my board use its own kernel and toolchain May 16 23:43:46 ah. May 16 23:44:15 robert: so look into openembedded/conf/distro/sharprom-compatible.conf how it is done for external toolchain May 16 23:44:38 emte: will you push fixes? May 16 23:44:45 hrw:thanks May 16 23:45:44 robert: you still use gcc 2.95 for apps? May 16 23:45:52 hrw, how do you push a fix for a bk commit comment? May 16 23:46:03 yes May 16 23:46:34 emte: local one? I do bk undo and patch tree one more time May 16 23:47:10 robert: with Zaurus machines we use gcc 2.95 for their 2.4 kernel and gcc 3.4 for apps May 16 23:47:12 no the ones i pushed May 16 23:47:41 emte: it is pushed - probably only admin can change it (if can) May 16 23:47:49 we are still talking about the "evas/evas-x11-fb_0.9.9.004.bb" right? May 16 23:48:01 emte: and evas-x11_* May 16 23:48:08 08:40 hrw@home:packages$ grep "with-x" e*/*.bb May 16 23:48:09 evas/evas-x11-fb_0.9.9.004.bb: --with-x May 16 23:48:09 evas/evas-x11_0.9.9.004.bb: --with-x May 16 23:49:03 where is the with-x part comming from ... that wasnt in my commit comments May 16 23:49:41 oh wait your searching internal May 16 23:49:56 I dont know - I just know that e stuff is yours May 16 23:50:17 yeah static packages for the fd.org stuff May 16 23:50:55 tho the cvs will have the same slash issue May 16 23:51:09 since it was pushed at the same time the same way May 16 23:51:20 ok. I have to prepare to work.. May 16 23:51:26 hrw:I think it would be better to use one gcc for both kernel and apps on my board May 16 23:51:54 robert: its your board so your rules May 16 23:52:02 :) May 16 23:56:52 hrw:for my board, i think i must write new .bb and .conf in distro and .conf in machine May 16 23:57:39 robert: yes May 16 23:58:15 robert: kernel.bb (prefered linux-BOARDNAME.bb), distro.conf with all ASSUME_PROVIDED, machine.conf May 16 23:58:18 hrw would a new push be enoufh to correct the issue you think? May 16 23:58:28 emte: sure May 16 23:59:19 hrw: yes ,agree **** ENDING LOGGING AT Mon May 16 23:59:56 2005