**** BEGIN LOGGING AT Mon Jun 28 02:59:57 2010 Jun 28 05:36:53 morning Jun 28 06:02:06 how long does it take for this "not-WSoD" to go away? Jun 28 06:03:44 I have flashed latest qi, kernel and rootfs and all I get after turning the device on is a white screen Jun 28 06:04:17 no boot screen or anything :( Jun 28 06:06:47 blindcoder: I think a patch from ThibG yesterday fixes that (at least on boot), but it hasn't found its way into SHR just yet Jun 28 06:07:30 hrm, great :( any workaround? Jun 28 06:07:37 it did boot once after all Jun 28 06:07:50 only after reinserting (empty) uSD card it appeared Jun 28 06:08:01 weird Jun 28 06:08:18 I found it would often work if left to sit for a while before booting Jun 28 06:26:44 Weiss: which one? that 3/5? I've built it and mrmoku said it's the same so I didn't push it Jun 28 06:28:17 yeah... the same at boot? (wouldn't expect much change on unblank) Jun 28 06:28:17 Weiss: this one? http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/commit/41d32b16473f81ec94cd51ed51fdc6f54e1f76d6 Jun 28 06:28:29 that's the one Jun 28 06:28:57 ok, I'll push it (it doesn't make things worse here (no WS before and after) and maybe it will help someone Jun 28 06:29:37 I've sort of pushed it yesterday (accidentaly with some other patch) http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=28018dda9df0ea95584d1ca49ee904b32fc3d5ff Jun 28 06:30:32 Weiss: have you seen that backtrace from failing Xorg on non-KMS kernels? If you have any hint I'll be glad to try it Jun 28 06:59:02 who needs emacs ;) Jun 28 06:59:09 Paul :) Jun 28 06:59:37 and Paul is good guy so I'll try to fix it first :) Jun 28 07:01:50 ERROR: '/home/shr/shr-unstable/openembedded/recipes/boost/boost_1.40.0.bb' failed Jun 28 07:01:52 JaMa: yeah, right you are :) Jun 28 07:01:53 ERROR: '/home/shr/shr-unstable/openembedded/recipes/emacs/emacs-x11_23.1.bb' failed Jun 28 07:01:56 ERROR: '/home/shr/shr-unstable/openembedded/recipes/fbreader/fbreader_0.12.1.bb' failed Jun 28 07:01:59 shit Jun 28 07:02:02 Makefile:145: *** commands commence before first target. Stop. Jun 28 07:03:47 mrmoku: do I remember right that you tried "possible WS fix" and reported that it wasn't enough, right? Jun 28 07:04:16 JaMa: the initialization on boot? yeah Jun 28 07:04:32 JaMa: tried the proposed mdelay(200) too this morning... does not help Jun 28 07:05:14 mrmoku: this patch http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/commit/41d32b16473f81ec94cd51ed51fdc6f54e1f76d6 Jun 28 07:05:31 or this image http://jama.homelinux.org/org.openembedded.shr.images/om-gta02/uImage-2.6.32.13-oe3.1+gitr0+a9254be10ac2294ea20165a87c09ea6afcf66d94-r0-om-gta02.bin Jun 28 07:10:33 JaMa: the one with "invalid pointer"? Jun 28 07:11:41 Weiss: yes Jun 28 07:12:00 Weiss: I tried to remove that free() from OutputDestroy and it's not the one Jun 28 07:12:29 Weiss: and no idea why, but gdb refuses to stop on free() breakpoint (to allow me easily get backtrace from faulting one :/) Jun 28 07:23:44 JaMa: hmm... I thought so... but that one seems not to be in om-gta02-2.6.34 Jun 28 07:23:52 JaMa: what branch is that commit from? Jun 28 07:25:41 JaMa: any idea when it started happening? non-KMS worked fine just a week ago.. Jun 28 07:28:57 mrmoku: it's from ML Jun 28 07:29:38 Weiss: after mesa-dri/xserver upgrade and that patch to xf86-video-glamo (needed because of newer xserver) Jun 28 07:30:00 Weiss: but I cannot try without that patch, because build fails and build against older xserver has different ABI Jun 28 07:30:21 Weiss: and IIRC I've also tried older with -ignoreABI and it failed even sooner Jun 28 07:30:44 Weiss: I'm checking newer libdrm (needed for newer intel driver..) Jun 28 07:31:44 JaMa: trying your uImage now Jun 28 07:32:47 mrmoku: it's already pushed... and building on buildhost too Jun 28 07:37:30 JaMa: still white Jun 28 07:51:12 JaMa: hi, new SHR version with 2.6.32 is really good, btw do you know now why gms power_on sysfs node is missing now? Jun 28 07:51:30 JaMa: in 2.6.34 Jun 28 07:53:15 radekp: it was in wrong place (before pcf patch) and then get lost completely.. but now there is more important issue, resume broken after http://git.openmoko.org/?p=kernel.git;a=commit;h=62a9c01da2fd32986e45525e5ced0f1718aeab10 Jun 28 07:54:10 ahh oki Jun 28 07:54:26 thanks Jun 28 07:58:03 radekp: 2.6.32 actually works? Jun 28 07:58:30 radekp: suspend, gsm, gps, all that really works? Jun 28 08:00:46 ThibG: morning... mdelay(200) did not help Jun 28 08:03:03 lindi-: i tried SHR and suspend was ok, gsm too, havent tried gps but i expect it works too... Jun 28 08:04:12 radekp: that is wonderful news Jun 28 08:05:41 lindi-: for me this is first SHR that actually works, but i have installed it yesterday, so i cant say i have tested everything Jun 28 08:06:48 well i don't expect SHR to work but if 2.6.32 works then i'm happy :) Jun 28 08:07:06 really odd how a release which is the first working one for some people is a disaster for others :S Jun 28 08:07:14 "works" is too much, but works a lot better :) Jun 28 08:07:19 or at least faster Jun 28 08:07:51 * Weiss suspects his NAND has died Jun 28 08:08:17 outch Jun 28 08:08:19 gibberish screen after resume (fixed by switch to another program and back) and frozen X with my game :( Jun 28 08:08:20 that sucks Jun 28 08:09:59 currently building shr-unstable so I can try to find out the reason Jun 28 08:18:35 blindcoder: anything in Xorg.log/dmesg? KMS? Jun 28 08:18:42 mrmoku, ok Jun 28 08:18:57 well, I have no idea, that did the trick for gena2x :/ Jun 28 08:19:43 Weiss: haven't had a time for a detailed inspection this morning. Will have a look later. What is KMS? Jun 28 08:23:15 blindcoder: kernel modesetting... you can tell whether you have it or not from Xorg.log, and it affects whose "fault" X bugs are :) Jun 28 08:23:57 Weiss: ah, okay. it's the latest shr-unstable from this morning, so I guess it's in :) Jun 28 08:24:48 mrmoku, what was your problem again: let the screen blanked for too long made you enable to unblank it properly? Jun 28 08:28:56 ThibG: have it blanked for some time and then unblank give me WS Jun 28 08:29:20 blanking and unblanking again before 2mins pass makes it recover Jun 28 08:29:40 ThibG: I never had WS on boot Jun 28 08:33:20 ok Jun 28 08:33:29 WS on boot is a completly different issue Jun 28 08:34:38 then, it's odd, you have the exact symptoms gena2x have, and yet, the change that "fixes" it for him doesn't work for you :( Jun 28 08:35:14 ThibG: it _feels_ like some component is loosing some state if powered off for too long... Jun 28 08:35:46 hm Jun 28 08:37:18 since my simplification, blank/unblank and suspend/resume (also boot) should all be the same.. clearly that's not the case.. Jun 28 08:37:36 i.e. the idea is to completely power everything off then on again each time Jun 28 08:37:57 oh yes, I forgot you made change in this area... Jun 28 08:38:25 BIG changes as far as display handing is concerned.. any problems are my fault alone Jun 28 08:38:32 but I don't know how to fix them... Jun 28 08:38:43 mrmoku, you are using KMS, right? Jun 28 08:39:31 Weiss, the way I've "fixed" unblank for gena2x is to add a (large) delay between powering the jbt again and reinitializing it Jun 28 08:39:47 ThibG: I tried both Jun 28 08:40:05 ThibG: and that worked? Jun 28 08:40:14 Weiss, unblank, no WS Jun 28 08:40:24 ThibG: plain om-gta02-2.6.34 with the mdelay(200) for example Jun 28 08:40:45 the idea was given in an old ticket Jun 28 08:41:00 ThibG: datasheet demands at least 120 ms between power on and wakeup.. maybe I didn't take enough account of PMU wibbles Jun 28 08:41:48 is the JBT datasheet available somewhere? Jun 28 08:42:17 not publically, I'm afraid Jun 28 08:42:51 mrmoku, and that didn't work? the mdelay is between "ret = regulator_bulk_enable(ARRAY_SIZE(jbt->supplies), jbt->supplies);" and the reg writes, right? Jun 28 08:43:31 Weiss, ok Jun 28 08:43:44 ThibG: yep Jun 28 08:44:20 ThibG: http://pastie.org/1021536 Jun 28 08:44:22 then your issue is different, or gena2x was lucky for some reason, yesterday :/ Jun 28 08:44:43 ThibG: you made this one 200ms instead of 120? http://git.bitwiz.org.uk/?p=kernel.git;a=blob;f=drivers/video/backlight/jbt6k74.c;h=91651e7ea5f72ba0da8d01a33225ba2eadc79d4d;hb=b3164fafb57d29bb16957e3f6f4fc9d3cbc738ff#l272 Jun 28 08:45:04 heh?! Jun 28 08:45:09 ok Jun 28 08:45:27 we don't have any mdelay, Weiss, in om-gta02-2.6.34 Jun 28 08:45:44 hmm.. that's a different driver version (there's no standby_to_sleep with KMS) Jun 28 08:45:54 with KMS, there is only "on" and "off" Jun 28 08:46:07 ok Jun 28 08:46:42 so there are two possibilities.. either "on" is not "on enough", or "off" is not "off enough" :) Jun 28 08:47:52 https://docs.openmoko.org/trac/ticket/2274 Jun 28 08:48:18 the last two comments are what leaded me to add a huge mdelay between powering and initialization Jun 28 08:48:42 ThibG: is it possible to revert only "glamo-mci: Use threaded irqs", I tried but not sure if I got it right (after resolving resulting conflict with later patches) Jun 28 08:49:08 JaMa, I'm trying to figure out that issue, too :p Jun 28 08:49:38 but I can't see how threaded IRQs can affect this :/ Jun 28 08:53:23 JaMa, any way, if you want me to check your patch, I can ;) Jun 28 08:53:40 jftr, the certificate for https://docs.openmoko.org/ is expired. Jun 28 08:54:41 * Weiss doesn't think he's ever seen the certificate for any OM servers *not* expired... Jun 28 08:55:05 lekernel: hi Sébastien ;) Jun 28 08:55:16 also, SSL warnings having a false positive rate of 100% doesn't help.. Jun 28 08:56:02 I'd rather like a warning on http... "this connection is not encrypted at all! anyone might do anything with it!" Jun 28 08:56:06 hi Jun 28 08:56:57 hi Jun 28 09:00:06 regarding shr-lite-eglibc-ipk--20100627-om-gta02.rootfs.tar.gz: /bin/chattr is a dead symlink Jun 28 09:10:32 morning Jun 28 09:40:59 another build error :) Jun 28 09:40:59 http://pastebin.com/jLWKVeMi Jun 28 09:41:05 on eet Jun 28 09:41:23 something about libtool not found... Jun 28 09:41:28 trying to rebuild libtool now... Jun 28 09:42:03 but I think it's crosstools' libtool... Jun 28 09:42:05 not the recipe one Jun 28 09:43:57 gnuchess,ipkg-utils,emacs-x11 fixed Jun 28 09:44:04 only boost and fbreader remaining Jun 28 09:53:59 still nothing :( Jun 28 09:54:11 it halts there on libtool Jun 28 09:54:35 mmm... libtool is actually not present on cross/ directory Jun 28 09:54:57 JaMa|Wrk, what is the recipe to rebuild toolchain? Jun 28 09:55:37 <[Rui]> hi Jun 28 09:55:43 <[Rui]> good morning! Jun 28 09:56:03 <[Rui]> anyone else having weird hangups with wifi on? Jun 28 10:06:45 <[Rui]> strange, no weird hang up when plugged... Jun 28 10:09:40 hi JaMa|Wrk Jun 28 10:09:45 I've still the libdrm bug Jun 28 10:09:49 I've just pulled Jun 28 10:10:19 JaMa|Wrk, should I rebuild from scratch? Jun 28 10:14:51 libdrm bug? Jun 28 10:16:50 Weiss: with ThibG patch i put beast blanked to fridge for 30 mins and it unblanked. in prev. attempt i got 50% unfixable 50% WS fixable at some blank/unblank later. Jun 28 10:23:38 Weiss, the image doesn't find libdrm Jun 28 10:23:54 do_rootfs to be more exact Jun 28 10:27:50 GNUtoo|laptop: no, cd deploy/ipk; grep -R libdrm and find which package is still depending on libdrm (not libdrm2) Jun 28 10:28:27 GNUtoo|laptop: or as I said before check which recipes depends on it on target (from older image) and -c clean all of them Jun 28 10:29:55 GNUtoo|laptop: few were bumped yesterday http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=ba84df15f597045c0b321577de693cc7a636dc61 and few more patches around Jun 28 10:31:08 ok Jun 28 10:31:09 daniele_athome: you have to rebuild it one-by-one Jun 28 10:31:34 thanks a lot Jun 28 10:32:40 YW, let me know which recipe it was - It's strange I missed that.. Jun 28 10:34:30 ah newer boost built ok, only fbreader remaining, someone for that? :) Jun 28 10:35:24 http://tinderbox.openembedded.net/packages/623853/ Jun 28 10:36:06 JaMa|Wrk, you mean gcc-cross* ? Jun 28 10:36:12 i've started building gcc-cross only Jun 28 10:36:19 should I clean them all and restart again? Jun 28 10:37:02 libtool is not part of gcc-cross Jun 28 10:37:22 JaMa|Wrk, but it was arm-oe-...-libtool Jun 28 10:37:31 is not in the cross tools? Jun 28 10:37:47 libtool-cross_2.2.6b.bb Jun 28 10:37:57 oh :) Jun 28 10:38:25 JaMa|Wrk, clean command on gcc-cross also deletes binaries in cross? Jun 28 10:38:33 so I may stop the build right now... Jun 28 10:39:02 daniele_athome: if you clean gcc-cross then you have to clean all other gcc-* recipes Jun 28 10:39:23 daniele_athome: because gcc-cross is staging same files as ie gcc-cross-initial/intermediate Jun 28 10:39:34 ok so now Jun 28 10:39:39 let's clean all, gcc and libtool Jun 28 10:39:41 and start again Jun 28 10:39:45 so removing gcc-cross will make initial/intermediate unusable for building gcc-cross Jun 28 10:40:14 daniele_athome: maybe also all eglibc* (also shares something with it IIRC) Jun 28 10:40:21 argh :S Jun 28 10:46:52 JaMa|Wrk: Configuring kernel. Jun 28 10:46:52 update-alternatives: Linking //boot/uImage to uImage-2.6.29-rc3 Jun 28 10:46:55 ?????? Jun 28 10:47:28 lol :) Jun 28 10:48:14 Q-Master: remove old.. for some reason new one has lower u-a priority see /usr/lib/opkg/alternatives/uImage Jun 28 10:48:52 Q-Master: ah, prio is decided from last part of PV Jun 28 10:49:12 uImage-2.6.29-rc3 29 Jun 28 10:49:12 uImage-2.6.32.13 13 Jun 28 10:49:21 I have no 2.6.29 installed Jun 28 10:49:22 8) Jun 28 10:49:45 and plz, update fsogsmd with PaulFertser's fixes Jun 28 10:50:15 wtf... shr just unsuspended without a message or call coming in? Jun 28 10:50:57 blindcoder: that's normal after update Jun 28 10:50:58 Q-Master: Paul warned me yesterday, that current is badly broken.. I won't update it until he says it's resolved Jun 28 10:51:13 damn Jun 28 10:51:35 I've got today that dbus stuck bug Jun 28 10:51:47 Q-Master: opkg list-installed | grep 2.6.29; shows nothing for you? Jun 28 10:52:09 then remove that 2.6.29 line from that file.. Jun 28 10:52:16 Q-Master: what's the issue? reacting to secretserviceprobes on the GSM? :) Jun 28 10:53:25 JaMa|Wrk: what file? Jun 28 10:53:33 blindcoder: dunno. 8) Jun 28 10:53:38 12:47:37 < JaMa|Wrk> Q-Master: remove old.. for some reason new one has lower u-a priority see /usr/lib/opkg/alternatives/uImage Jun 28 10:53:46 ah Jun 28 10:54:25 removed Jun 28 10:54:42 reinstall package "kernel" Jun 28 10:55:47 why? Jun 28 10:55:55 ln -s helps Jun 28 10:57:23 because you don't want u-a to change that with next upgrade Jun 28 11:03:10 fun fact: sms only arrives when not suspended Jun 28 11:03:23 O.o Jun 28 11:03:43 yeah, I just sent myself an SMS 10 minutes ago, no notification or anything Jun 28 11:03:47 I'd expect an unsuspend on incoming short message... and alert as configured... Jun 28 11:03:49 when pressing the power button Jun 28 11:03:57 immediately "ding ding" new message Jun 28 11:04:07 happened twice now Jun 28 11:06:19 blindcoder: do you have any lights on buttons while on USB charge? Jun 28 11:06:35 Q-Master: let me check Jun 28 11:07:03 no Jun 28 11:07:07 heh Jun 28 11:07:09 * Q-Master too Jun 28 11:07:21 but the top bar indicates charging Jun 28 11:07:22 GNUtoo|laptop: ĥey Jun 28 11:07:29 blindcoder: yep Jun 28 11:07:29 GNUtoo|laptop: got the version mismatch problem solved ? Jun 28 11:07:29 device also powers up immediately after plugin Jun 28 11:08:33 dcordes, I'm upgrading the freerunner right now.I'll look at libdrm issue just after that Jun 28 11:11:54 fbreader fixed Jun 28 11:12:14 can somebody test it for me? Jun 28 11:15:14 wow Jun 28 11:15:18 fbreader fixed!!! Jun 28 11:15:25 JaMa|Wrk, are you sure f37686260161644129601ba19880abe48c56cc8d works? I've tried this version of the glamo-mci module, and still fails to suspend... Jun 28 11:15:35 shouold I test on freerunner or on htcdream? Jun 28 11:15:53 ThibG: to suspend? I'm 99% sure Jun 28 11:15:56 also I won't test now Jun 28 11:16:07 I'm upgrading my freerunner right now Jun 28 11:16:12 huh, odd, odd, odd Jun 28 11:16:18 GNUtoo|laptop: build issue if fixed and 0.12.10 version added Jun 28 11:16:32 JaMa|Wrk, 0.12.10 is not machine arch? Jun 28 11:16:34 you had some FS mounted, right? Jun 28 11:16:46 JaMa|Wrk, I looked yesterday and it wasn't Jun 28 11:17:06 ThibG: I can give you image to confirm yourself http://jama.homelinux.org/org.openembedded.shr.images/om-gta02/ Jun 28 11:17:11 and it still uses variables comming from om-gta02.conf Jun 28 11:17:18 ThibG: yes booted from mmcblk0p1 with Qi Jun 28 11:17:41 ok Jun 28 11:17:51 GNUtoo|laptop: I remember reading it.. but I didn't change arch for that Jun 28 11:17:58 ok Jun 28 11:18:07 should I do it when I'll have some oe time? Jun 28 11:18:26 Q-Master: it's just something with oeventsd<->fsodeviced communication, it can't access LEDs probably due to changed name in dbus path Jun 28 11:18:34 Q-Master: kernel's LED interface works Jun 28 11:18:41 GNUtoo|laptop: om-gta* are only 2 machines in OE overriding it.. I'm not sure it's right place Jun 28 11:19:11 JaMa|Wrk, ok but best is the enemy of good Jun 28 11:19:19 we must fix it somehow Jun 28 11:19:32 machine arch could be a temporary fix Jun 28 11:19:43 GNUtoo|laptop: It would be IMHO better in fbreader recipe, as no other recipe is using those definitions (but I agree that can be usefull somewhere) Jun 28 11:20:03 JaMa|Wrk, ok, thanks Jun 28 11:20:15 ThibG: confirmed to work? Jun 28 11:20:20 JaMa|Wrk, what bother me is the resolution, the other bits could be distro specific no? Jun 28 11:20:24 no, I have yet to do some black magick Jun 28 11:20:32 GNUtoo|laptop: ok was just curious if you had it fixed Jun 28 11:20:45 GNUtoo|laptop: xf86-video-msm drives me crazy Jun 28 11:20:58 * ThibG has no installed system on his FR Jun 28 11:21:01 dcordes, ah? Jun 28 11:21:08 ThibG: ok Jun 28 11:21:37 GNUtoo|laptop: meanwhile I tried to build it manually - the README really sucks - and tried to grab your msm_drv.so but of course ran into version mismatch Jun 28 11:21:41 GNUtoo|laptop: I haven't checked why cannot it fbreader get on runtime and behave sane on all devices Jun 28 11:22:10 JaMa|Wrk, last one some time ago was segfaulting Jun 28 11:22:14 randomly Jun 28 11:22:23 GNUtoo|laptop: but with all those platforms supported I guess you don't have to rebuild it on windows if you switch LCD resolution :) Jun 28 11:22:23 I hope this new version fixes that Jun 28 11:22:29 JaMa|Wrk, does your image require some modules, or bare minimum is built-in? Jun 28 11:22:43 ThibG: all built-in Jun 28 11:22:54 dos1: possibly Jun 28 11:22:58 hm Jun 28 11:23:04 emtooth don't work too. Jun 28 11:23:05 JaMa|Wrk, I'll read the readme Jun 28 11:23:08 ThibG: ah maybe not Jun 28 11:23:22 ThibG: this old revision has ie g_ether not built-in, sorry Jun 28 11:23:38 ok Jun 28 11:24:30 image too big for u-boot...? Jun 28 11:24:33 Y Jun 28 11:24:43 adjust env Jun 28 11:24:48 ok Jun 28 11:25:24 JaMa|Wrk, it seem that README.buil doesn't mention resolution but mention things like: Jun 28 11:25:32 TARGET_ARCH = zaurus UI_TYPE = qtopia-240x320 Jun 28 11:25:45 for us that's Jun 28 11:25:47 TARGET_ARCH = desktop UI_TYPE = gtk Jun 28 11:27:15 dcordes, which image are you using? mine? Jun 28 11:27:16 does gitorious work for someone today? Jun 28 11:27:34 it's refusing my key (no change and checked on web) Jun 28 11:27:34 JaMa|Wrk, did it ever worked with oe? Jun 28 11:27:38 ah ok Jun 28 11:28:00 GNUtoo|laptop: I never used fbreader.. Jun 28 11:28:15 ok Jun 28 11:28:30 the did it ever worked with oe? was for gitorious Jun 28 11:28:36 ok, booting, thanks, JaMa|Wrk Jun 28 11:28:42 GNUtoo|laptop: no.. and actually I don't really want to run prebuilt driver without neon support Jun 28 11:28:49 GNUtoo|laptop: this must be fixed in org.oe.dev Jun 28 11:28:51 GNUtoo|laptop: yep. desktop gtk is for us Jun 28 11:29:05 dcordes, ok Jun 28 11:30:42 hm, WS on resume Jun 28 11:30:46 well, not sure it suspends Jun 28 11:31:10 ThibG: did you download drm or nodrm version? Jun 28 11:31:20 nodrm Jun 28 11:31:32 ok Jun 28 11:31:40 Q-Master: if only my "fix" was able to solve that stuck dbus bug :/ Jun 28 11:31:41 that's without any our patches Jun 28 11:31:53 want me to try drm one? Jun 28 11:31:56 PaulFertser: so no> Jun 28 11:31:58 ? Jun 28 11:31:59 Q-Master: i was talking about another issue that was introduced by using flowcontrolled node. Jun 28 11:32:10 Q-Master: yesterday Jun 28 11:32:13 ThibG: no, it should be same (mci wise) Jun 28 11:32:37 PaulFertser: so still not good time to upgrade fsogsmd right? Jun 28 11:32:43 JaMa|Wrk: indeed Jun 28 11:33:07 ping me when it's fixed (I'll update fsogsmd.conf and bump SRCREV) Jun 28 11:33:24 JaMa|Wrk: sure thing Jun 28 11:33:46 oh, WS on boot, this time... Jun 28 11:34:20 Somehow that innocent change exposed a bigger problem. I'm not sure i'll be able to reproduce it while stracing though :/ Jun 28 11:35:45 PaulFertser: printf is our friend. also debugging optimized application is great fun Jun 28 11:35:48 :) Jun 28 11:35:48 <[Rui]> Q-Master: gtk apps are a lot slower in the freerunner than efl ones Jun 28 11:36:17 [Rui]: run oprofile and see why :) Jun 28 11:36:20 gena2x: yeah, indeed. Too bad compiling fsogsmd takes considerable time on gta02. Jun 28 11:36:35 <[Rui]> gena2x: yeah Jun 28 11:36:45 PaulFertser: i did setup debian-cross :) Jun 28 11:37:02 [Rui]: there's no efl backend for fbreader IMO Jun 28 11:37:29 <[Rui]> Q-Master: frontend you mean? but yeah, one should be written Jun 28 11:37:36 PaulFertser: so now can install ant dependency in one apt-get. Jun 28 11:37:51 s/ant/any/ Jun 28 11:37:52 gena2x meant: PaulFertser: so now can install any dependency in one apt-get. Jun 28 11:38:06 [Rui]: 8) Jun 28 11:38:15 <[Rui]> if there were going to be decent graphics cards with open drivers in the near future, maybe it would be less of an issue (then we'd get battery life drain rather than slowness) Jun 28 11:38:58 PaulFertser: i mean it's easy to compile on host. Jun 28 11:40:08 gena2x: cross-compiling all FSO2 deps without OE is not that fun anyway. Jun 28 11:40:48 PaulFertser: deps like that? Jun 28 11:41:08 PaulFertser: i mean you can install any arm lib to host with one apt-get. Jun 28 11:41:21 PaulFertser: and so with deps. Jun 28 11:43:36 <[Rui]> unplugged from power sources, using wifi, it seems like a lock-up is reached, somehow Jun 28 11:43:44 <[Rui]> plugged == no problem Jun 28 11:44:24 GNUtoo|laptop: I cannot push newer fbreader to gitorious now, so if you have time later, please try http://jama.homelinux.org/org.openembedded.shr/armv4t/fbreader_0.12.10-r0.5_armv4t.ipk Jun 28 11:44:31 <[Rui]> or not... Jun 28 11:44:47 <[Rui]> I turned wifi off , then unplugged, and already seems locked up Jun 28 11:44:57 gena2x: cross-vala and many libfso* that are part of the cornucopia. Jun 28 11:45:08 JaMa|Wrk, I can't confirm it works :/ Jun 28 11:45:26 <[Rui]> can't ssh, but pings. shr-idle locks the screen normally, but I can't launch new apps Jun 28 11:45:34 <[Rui]> it looks like fork() is borked Jun 28 11:45:42 ok Jun 28 11:45:59 <[Rui]> definitly looks like that Jun 28 11:46:09 PaulFertser: oh, vala..... thay people love exotic things? Jun 28 11:46:09 ThibG: strange, I did suspend with echo mem > /sys/power/state, can it make any difference? Jun 28 11:46:17 I do that too :) Jun 28 11:46:18 s/thay/why Jun 28 11:46:43 <[Rui]> ssh doesn't fork a child to handle my connection, and the only apps that seem to work are those that are running already Jun 28 11:46:51 ThibG: hm.. something in userspace? Jun 28 11:46:53 <[Rui]> and don't need to fork Jun 28 11:47:05 <[Rui]> I'll have to do a brute force poweroff Jun 28 11:47:11 <[Rui]> see ya later Jun 28 11:47:13 gena2x, I was running a SHR snapshot from a while ago Jun 28 11:47:43 ThibG: and without Xorg at all? (I'm not running it with nodrm) Jun 28 11:48:04 well, I have first to retrieve the g_ether module... Jun 28 11:48:37 is ar600 built-in? Jun 28 11:48:41 ThibG: I swear that with this image and modules I did suspend (no ping response confirmed sleeping) and resumed with ping response and ssh-in Jun 28 11:49:07 ThibG: yes, see coresponding modules- in that dir Jun 28 11:49:10 ok Jun 28 11:49:13 thanks Jun 28 11:49:25 ThibG: just untar them and run depmod -a 2.6.34 Jun 28 11:52:52 hm, annoying Jun 28 12:00:06 * opkg_download: Failed to download http://build.shr-project.org/shr-unstable/ipk//armv4t/evas-engine-buffer_0.9.9.060+svnr49824-r7.5_armv4t.ipk: Failed writing received data to disk/application. Jun 28 12:00:11 I bet it's because it's on ram Jun 28 12:00:17 the tmp dir for opkg Jun 28 12:00:24 I'll use a tarball then Jun 28 12:04:38 GNUtoo|laptop: see tmpdir option in opkg.conf Jun 28 12:04:53 I know there is aso -t Jun 28 12:05:08 but tarball would be faster Jun 28 12:05:17 and I already started Jun 28 12:05:31 GNUtoo|laptop: opkg didn't work cuase I had file locking disabled in kernel Jun 28 12:05:58 dcordes, ah ok Jun 28 12:06:28 dcordes, I'll push a new rootfs+ipk set soon Jun 28 12:06:53 dcordes, what's file locking option? CONFIG_FILE_LOCKING ? Jun 28 12:08:33 GNUtoo|laptop: exactly< Jun 28 12:08:51 GNUtoo|laptop: I will update the qsd kernel in OE accordingly. also needs urgent srcrev bump Jun 28 12:09:20 lol ubuntu works perfectly well on hd2 Jun 28 12:09:29 it would be nice to install FSO on it Jun 28 12:09:59 i doubt that anything linux-based runs "perfectly well" on a reverse engineerd mobile phone Jun 28 12:10:10 dcordes, ok Jun 28 12:11:39 JaMa|Wrk, i rebuilt everything Jun 28 12:11:40 hi mickey|office Jun 28 12:11:42 still no libtool :( Jun 28 12:11:58 mickey|office: if you have any ideas how can i try to help resolve the latest issue, please share, as i can probably try it in the evening. Jun 28 12:12:05 And hi there :) Jun 28 12:12:21 mickey|office: this sounds pessimistic. 3 days ago we couldn't do anything. Now I a have a full desktop environment working on it Jun 28 12:13:31 JaMa|Wrk, should I remove my /home/root it segfaults again (om-gta02) Jun 28 12:14:06 I'll do that Jun 28 12:16:02 glamo: http://pastebin.com/pS75GkVU for kernel hackers Jun 28 12:16:41 result: white screen Jun 28 12:16:46 at boot Jun 28 12:17:09 GNUtoo|laptop: do you have that "3/5" patch? Jun 28 12:17:42 not shure Jun 28 12:17:45 I've that image: Jun 28 12:17:55 daniele_athome: /OE/tmpdir-dev-shr/old/sysroots/x86_64-linux/usr/bin/arm-oe-linux-gnueabi-libtool Jun 28 12:17:57 shr-full-eglibc-ipk--20100627-om-gta02.rootfs.tar.gz Jun 28 12:18:12 JaMa|Wrk, I guess that binary is for x86-64 Jun 28 12:18:15 ah blank again Jun 28 12:18:16 I have normal x86 Jun 28 12:18:21 anyway I found it, in another directory Jun 28 12:18:22 I'll opkg update;opkg upgrade Jun 28 12:18:40 actually JaMa|Wrk the problem is in the eet Makefile, referencing $(top_srcdir)/arm-eo-linux-gnueabi-libtool Jun 28 12:18:41 daniele_athome: http://paste.pocoo.org/show/230871/ Jun 28 12:18:48 thinking that libtool is in source directory Jun 28 12:19:06 yes JaMa|Wrk i see Jun 28 12:19:23 daniele_athome: mmt I'll give you my eet build dir to compare Jun 28 12:19:47 ok Jun 28 12:20:50 JaMa|Wrk, for now I workarounded it by manually modifying the generated Makefiles Jun 28 12:20:51 daniele_athome: http://jama.homelinux.org/org.openembedded.shr.images/eet.tar.bz2 Jun 28 12:21:08 but I don't understand how this can happen to me and not to the buildhost too Jun 28 12:22:13 I have $(top_srcdir)/arm-eo-linux-gnueabi-libtool created Jun 28 12:22:25 compare logs from temp why it's not created for you Jun 28 12:24:54 JaMa|Wrk, ok, it works ok... Jun 28 12:25:01 (except, WS on resume) Jun 28 12:25:28 ok Jun 28 12:25:42 hm, wait a moment... Jun 28 12:28:04 JaMa|Wrk, ehm... i'm an idiot Jun 28 12:28:08 :S Jun 28 12:28:25 Hiii!!! Jun 28 12:28:36 JaMa|Wrk: No space left on device Jun 28 12:28:38 lol :S Jun 28 12:28:49 chichi Jun 28 12:29:06 I have a question not directly related to openmoko, but to eglibc instead (and no, I can't join the official channel since it requires invite) Jun 28 12:29:25 but... the question is small - for most of you it may sound stupid i think... Jun 28 12:30:18 In short - I would upgrade my embedded distro to glibc 2.12; can I do that without re-building? I ask this because the name of the main interpreter link has changed: first it was ld-linux.so.2, now it's ld-linux.so.3 so that executables will not find it Jun 28 12:31:11 *eglibc*, not glibc Jun 28 12:31:56 in which version it was ld-linux.so.2? Jun 28 12:32:29 2.11.2 Jun 28 12:33:01 2.9->2.10->2.11->2.12 upgrades were done without rebuild (DISTRO_PR) bump but I don't remember ld-linux soname change.. Jun 28 12:33:56 JaMa|Wrk: ... sorry, it is 2.12.90 - a development snapshot... to be precise Jun 28 12:34:40 JaMa|Wrk: and... when the soname changes, does this mean you should re-build ? (I deduced so, not finding documentation) Jun 28 12:35:00 SHR root@gojama /lib/modules/2.6.34 $ opkg search /lib/ld-linux.so.3 Jun 28 12:35:01 libc6 - 2.12-r11.2+svnr10689.5 Jun 28 12:36:08 I'm not sure, maybe ld-linux is handled different, ldd output is also different.. Jun 28 12:36:11 SHR root@gojama /lib/modules/2.6.34 $ ldd /usr/bin/time libc.so.6 => /lib/libc.so.6 (0x40024000) /lib/ld-linux.so.3 (0x40000000) Jun 28 12:36:21 grr newlines.. Jun 28 12:36:25 ldd /usr/bin/time Jun 28 12:36:28 libc.so.6 => /lib/libc.so.6 (0x40024000) Jun 28 12:36:31 /lib/ld-linux.so.3 (0x40000000) Jun 28 12:38:52 JaMa|Wrk: sorry if I'm stressing you... if I'm disturbing tell me... :) ; can you tell me what's the interpreter of that executable? Using something like readelf? Jun 28 12:41:07 INTERP 0x000134 0x00008134 0x00008134 0x00013 0x00013 R 0x1 Jun 28 12:41:07 [Requesting program interpreter: /lib/ld-linux.so.3] Jun 28 12:41:42 mrkiko: http://pastebin.ca/1891067 Jun 28 12:42:04 JaMa|Wrk: ehehe... on my debian sid system executables search for the interpreter ending in .2; Jun 28 12:42:15 JaMa|Wrk: thank Jun 28 12:43:00 mrkiko: this executable was built after upgrading eglibc.. so I'm not sure how it looked before.. Jun 28 12:43:31 ok ok ... thank you very much aniway Jun 28 12:43:39 JaMa|Wrk: ... you have been helpful I think Jun 28 12:43:55 :) OK Jun 28 12:43:58 JaMa|Wrk: can you direct me to some source of infos & irc channel? Jun 28 12:44:16 #oe if you want to use, well OE :) Jun 28 12:45:56 GNUtoo|laptop: fbreader pushed with PACKAGE_ARCH set only for om-gta*, I'll fix it if there is some new issue.. I want to finish the feed now :) Jun 28 12:45:59 thank you... I'll ask... Jun 28 12:46:14 ok Jun 28 12:46:16 YW Jun 28 12:46:16 thanks Jun 28 12:46:25 * JaMa|Wrk is going to help his boss now for a while :) Jun 28 13:08:52 dos1, how do you like 533 version? now i have even faster 533/CLK2-2-2 version! Jun 28 13:09:26 gena2x: didn't try yet, i had to have suspend working Jun 28 13:09:28 dos1, which is without overclocking. Jun 28 13:09:54 dos1, oh, only a bit overclocking 500->533. Jun 28 13:10:37 dos1, suspend... oh, my suspend is not working anyway on .34. Jun 28 13:10:55 dos1: so can't even check it now. Jun 28 13:14:46 NOTE: package task-shr-feed-1.0-r58: task do_rm_work_all: Succeeded :) whole feed rebuilt Jun 28 13:16:39 suspend... don't talk about that :( Jun 28 13:20:36 ThibG it's pleasure to solve hard problem, yeah? ;) Jun 28 13:28:00 yes, yes, a pleasure Jun 28 13:31:06 gena2x: oh, so you tweaked the timings for good? Cool stuff. Jun 28 13:32:28 PaulFertser: our memory in specs working on 133/CLK3-smth-smth or 83/CLK2-smth-smth Jun 28 13:32:45 so it works, without any voltage tuning Jun 28 13:33:17 according tests, 533/88/clk2 has same mem speed to 400/100. Jun 28 13:33:27 gena2x: but the cpu needs higher voltage anyway, right? Also does 133 feel stable, it's also glamo and nor on the same bus. Jun 28 13:33:28 but much faster cpu :) Jun 28 13:34:27 our CPU part is '500mhz 1.7v' according to BOM. Jun 28 13:34:51 so running it at 1.7 volts instead or 1.5v is in no way wrong IMO Jun 28 13:35:08 i have no 133. Jun 28 13:35:10 PaulFertser: got you problem. Jun 28 13:36:03 i have 88/CLK2 (a bit more than 83/CLK2 in spec) or something like 116/CLK3 (133/CLK3 in spec) Jun 28 13:36:55 but 116 version require memory voltage of 1.9 (instead of 1.8) Jun 28 13:37:15 but it also has peripherial clock of 116/2 Jun 28 13:37:18 not 88/2 Jun 28 13:37:36 but system 'feels' really faster. Jun 28 13:37:52 idle consumption change is nigligible Jun 28 13:38:42 gena2x: sounds uber-cool indeed Jun 28 13:39:09 default -274312 -> 533/88/CL2/1.7/1.8 -239625 (usb connected, no X, screen blanked) Jun 28 13:39:39 system have nothing running ^^^ Jun 28 13:39:59 once i got very strange: 500/83/CL2/1.7/1.9: -320437 Jun 28 13:40:10 not thinking, how that were possible. Jun 28 13:40:17 s/not/now/ Jun 28 13:40:17 gena2x meant: now thinking, how that were possible. Jun 28 13:40:42 and, as our part is 1.7V, i'll try 566 with 1.75V now Jun 28 13:41:32 as 1.75V should be in-spec for 1.7V part (this is really risky, but i have spec only on 400mhz/1.5 part) Jun 28 13:43:20 also, i guess 460/116 config not really need 1.7V on cpu. think 1.6V will be ok. Jun 28 13:50:55 i don't know is it ok to measure diff of power while usb is on :( Jun 28 13:51:44 so all power consumtion is preliminary. Jun 28 13:51:58 JaMa|Wrk: 2010-06-28T13:26:24.785414Z [WARN] fsousaged : Can't read-open /sys/class/i2c-adapter/i2c-0/0-0073/neo1973-resume.0/resume_reason: No such file or directory Jun 28 13:52:02 ?? Jun 28 13:52:59 Q-Master, lol Jun 28 13:53:18 /sys/devices/platform/s3c2440-i2c/i2c-0/0-0073/resume_reason but it's not changed in fousaged yet Jun 28 13:53:29 I think htcdream has some resume reason too Jun 28 13:53:38 seems that everything is broken Jun 28 13:54:37 what's so broken by this warning? Jun 28 13:54:55 I'm waiting with updating this until sysfs paths in 2.6.34 are resolved Jun 28 14:00:13 gena2x, 533MHz, seriously? :D Jun 28 14:00:30 ThibG: want to try? Jun 28 14:00:34 not sure Jun 28 14:00:41 it's in spec. Jun 28 14:00:46 and works ok here. Jun 28 14:01:04 no problems found atm. Jun 28 14:01:05 I thought 500MHz was the max? Jun 28 14:01:18 533 is a tune of course. Jun 28 14:01:28 but you may try 500, which is fully in spec. Jun 28 14:02:07 well, maybe later, I have this infamous glamo-mci bug to resolve Jun 28 14:03:14 ThibG: we all waiting for the fix :) Jun 28 14:10:19 ThibG: whenever you have something to test... I'm here :) Jun 28 14:11:02 mrmoku, with the msleep, I weren't able to reproduce your issue (I were without it) Jun 28 14:12:00 om screen power 0 did work fine, but waiting a bit before doing om screen power 1 didn't Jun 28 14:12:06 with the msleep, it works fine Jun 28 14:15:14 hmmm grrr Jun 28 14:15:16 Xorg: /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/armv6-novfp-oe-linux-gnueabi/usr/include/xorg/privates.h:122: dixGetPrivateAddr: Assertion `key->initialized' failed. Jun 28 14:15:17 still Jun 28 14:15:22 after rebuilding from scratch Jun 28 14:15:26 I didn't have that before Jun 28 14:15:31 it happens only in rotated mode Jun 28 14:15:38 with xf86-video-fbdev Jun 28 14:15:43 on htcdream Jun 28 14:15:59 do you rotate with xrandr or in xorg.conf directly? Jun 28 14:16:26 xorg.conf Jun 28 14:16:35 xrandr rotation don't work because it's unimplemented Jun 28 14:17:40 I also rotate input etc... Jun 28 14:17:47 I'm using CC rotate in xorg.conf with fbdev and it seems to work Jun 28 14:17:58 ok Jun 28 14:18:05 I think it's an input issue Jun 28 14:18:09 xf86-input-tslib Jun 28 14:18:28 JaMa|Wrk: here rotation/resolution change is ok (Xserver/glamo, non-kms). Jun 28 14:18:29 JaMa|Wrk, do you have a debug board? Jun 28 14:18:48 ThibG: no Jun 28 14:18:55 ok Jun 28 14:20:05 ThibG were some patch so see kernel logs without dboard (store somewhere, then reboot(?) with watchdog(?) and see logs) Jun 28 14:21:19 what should I do, it worked perfectly in landcape yesterday and it suspended/resumed without killing xorg for the first time(new rebased kernel) Jun 28 14:40:27 gena2x: memconsole? it was sth like that Jun 28 14:40:38 dos1: ramconsole Jun 28 14:40:53 yep :D Jun 28 14:43:37 dos1: yes, still never need to use that, just told to ThibG it exist. Jun 28 14:45:21 nice Jun 28 14:45:33 fltkspacetrader just gives "X I/O error" and dies Jun 28 14:45:44 X.log.0 doesn't say anything Jun 28 14:46:33 xserver can be restarted just fine Jun 28 14:52:59 true overclockers setup: S3C244X: core 549.000 MHz, memory 91.500 MHz, peripheral 45.750 MHz Jun 28 14:55:07 with CLK2 Jun 28 14:57:58 heh, crazy, as if the FR wasn't unstable enough ;) Jun 28 14:58:15 it's stable, really. Jun 28 14:58:24 i am lmbenching it right now. Jun 28 14:58:43 i see absolutely no change in stability. Jun 28 14:59:11 (of course, not testing all stuff) Jun 28 14:59:27 glamo alive, and this enough for me :) Jun 28 15:01:26 may be different timings even defeat some bugs. Jun 28 15:01:31 gena2x: run fltkspacetrader, see it die :P Jun 28 15:03:10 blindcoder: hm why not? lets try. Jun 28 15:03:52 * blindcoder .oO( now if it does NOT die, I'm so overclocking this baby ) Jun 28 15:04:28 blindcoder: i already tried different video stuff on 533 and 465. a bit nervous with 550, as it is 1.75V. Jun 28 15:04:46 gena2x: actually, it just produces an "X I/O error" here and dies Jun 28 15:05:32 don't know where to break on that error, though Jun 28 15:07:40 dcordes, repo updated Jun 28 15:07:45 with images Jun 28 15:07:52 that are consistents Jun 28 15:08:00 that is to say no mismatch with xorg Jun 28 15:08:16 erm... short question Jun 28 15:08:22 what s3cFOO is in the freerunner? Jun 28 15:09:46 blindcoder: 2410 Jun 28 15:09:49 ubuntu network manager says it's connected to a s3c2410 Jun 28 15:10:01 lindi-: okay, but I had to flash Qi*2442*udf file Jun 28 15:10:13 blindcoder: had to? Jun 28 15:10:15 because for the Qi*2410*udf it said "not for this hardware" Jun 28 15:10:37 the file qi-s3c2410-1.0.2-r2+gitr0+c38b062a609f1442e6a9e13005cfbdfd59a5ac0d.udfu Jun 28 15:10:50 when doing dfu-util -a u-boot -D qi-s3c2410* Jun 28 15:11:19 the GTA02 is a s3c2442, nor? Jun 28 15:15:00 The main Processor (CPU) of the Neo1973 GTA02 is a Samsung S3C2442B B54 (running at 400 MHz) Jun 28 15:15:06 from http://wiki.openmoko.org/wiki/Neo_FreeRunner_Hardware Jun 28 15:15:40 okay, I now know X is choking on Jun 28 15:15:47 Fl::visual(FL_DOUBLE|FL_INDEX); Jun 28 15:15:57 that's when I get an "X I/O error" Jun 28 15:17:06 blindcoder: lets try your game Jun 28 15:20:00 gena2x: don't bother, it'll probably just kill your X server anyway Jun 28 15:35:54 interesting Jun 28 15:36:03 it only crashes when I'm using double buffered windows Jun 28 15:36:09 not when using single buffered ones Jun 28 15:36:41 but single buffered windows suck :( Jun 28 15:36:49 especially with frequent whole-screen updates Jun 28 15:43:29 hrm... Jun 28 15:43:29 bitbake -i Jun 28 15:43:30 ERROR: Nothing to do. Use 'bitbake world' to build everything, or run 'bitbake --help' for usage information. Jun 28 15:43:36 bug? intention? Jun 28 15:46:33 uwargh, eye cance >_< Jun 28 15:46:43 s/cance/&r/ Jun 28 15:46:45 blindcoder meant: uwargh, eye &r >_< Jun 28 15:47:04 pff Jun 28 15:47:16 well, at least it works again Jun 28 15:49:29 so, can someone tell me why there's no working DBE on SHR anymore? Jun 28 15:55:19 hrm, or maybe I should check if double buffering is actually available Jun 28 15:55:31 instead of relying on it to be Jun 28 16:03:42 hmm, it already crashes on the check Jun 28 16:04:08 seems to me like a problem deeper in the system than in my programs Jun 28 16:04:16 X.org/glamo-driver/kernel Jun 28 16:04:23 fscking awesome :/ Jun 28 16:11:47 blindcoder: gon't go crazy, just report bug :) Jun 28 16:18:33 dcordes: http://www.pocketpc.ch/attachments/htc-touch-pro/3612d1221028638-videokabel-tv-out-kabel-ueber-extusb-htc-cable-small.png Jun 28 16:18:39 is that front or back view? Jun 28 16:20:03 leviathan: I have a HTC Hero with that kind of plug, it's front view Jun 28 16:20:33 leviathan: if I look at the front (display) of the phone and turn it upwards, I see the plug as shown there Jun 28 16:27:45 blindcoder: then this is from behind? http://shop.kineteka.com/Products/93-htc-extusb-11-pin-usb-connector-style-1.aspx Jun 28 16:28:04 * leviathan wants to be sure, that he doesnt burn his smartphone+adapter Jun 28 16:28:47 oh wait... Jun 28 16:28:57 weird Jun 28 16:29:14 on my hero the edge is on the bottom RIGHT Jun 28 16:29:18 not bottom LEFT Jun 28 16:29:30 ah, yes, of course Jun 28 16:29:32 stupid me Jun 28 16:29:42 that plug would go into the phone Jun 28 16:36:49 where are gnutoo and mickeyl Jun 28 16:36:51 hmm Jun 28 16:36:56 ok Jun 28 16:37:01 where is gnutoo Jun 28 16:37:01 ?? Jun 28 16:55:52 hmm, htc touch, das ist prä-qsm Ära IIRC, keine Ahnung, wie und ob der Support für Linux da überhaupt ist Jun 28 17:02:22 hi I read that udev was removed Jun 28 17:02:29 and that there is now devtmpfs Jun 28 17:02:30 but Jun 28 17:02:48 alsamixer: function snd_ctl_open failed for default: No such file or directory Jun 28 17:03:01 so... what should I do? Jun 28 17:03:36 also no /dev/smd0 Jun 28 17:04:03 so sound and modem don't work anymore on dream Jun 28 17:04:03 Jun 28 17:04:23 who removed udev? Jun 28 17:04:24 I bet there is some tricks to make it work right? Jun 28 17:04:25 from where? Jun 28 17:04:36 from shr-image Jun 28 17:05:03 is /etc/init.d/devices present? Jun 28 17:05:09 I'll look Jun 28 17:05:10 if so, remove it and reboot Jun 28 17:05:38 last time i checked, the devices script was from devfs era Jun 28 17:05:53 and did not know about more modern scenarios Jun 28 17:05:58 thanks a lot Jun 28 17:06:04 it was there Jun 28 17:06:20 i have been removing it in my post-install Jun 28 17:06:38 ok Jun 28 17:06:49 (obviously that's not the correct fix) Jun 28 17:06:52 btw mickeyl I've srtarted rebasing on top of android msm 2.6.32 Jun 28 17:07:20 first good thing: Jun 28 17:07:42 static rotation was viable yesterday but newer xorg broke xf86-input-tslib's rotation Jun 28 17:08:02 ouch sound is bad Jun 28 17:08:06 mplayer doesn't quit Jun 28 17:11:30 mickeyl: have you removed suspend/resume debug messages from log? I am not seeing it... Jun 28 17:11:31 do the alsa patches still apply? Jun 28 17:11:46 Q-Master: no, are you on DEBUG? Jun 28 17:11:51 mickeyl: yes Jun 28 17:12:05 mickeyl: they were till new kernel Jun 28 17:12:14 mickeyl: and now they're gone. Jun 28 17:12:20 how does it look now? Jun 28 17:12:31 mickeyl: everything is working horrible. Jun 28 17:12:55 mickeyl: there's nothing now about suspends Jun 28 17:13:27 mickeyl: so seems that fsogsmd doesn't know about suspend... Jun 28 17:15:29 JaMa|Wrk, fbreader works well for htcdream(I bitbaken it) Jun 28 17:15:36 I'll eat Jun 28 17:33:40 freesmartphone.org: 03morphis 07msmcomm * r225eefec7f15 10/msmcommd2/src/ (Makefile.am remoteclient.vala remoteclienthandler.vala): msmcommd2: add dummy remote client class Jun 28 17:33:41 freesmartphone.org: 03morphis 07msmcomm * rea6c93528566 10/msmcommd2/src/ (7 files): msmcommd2: don't use event-based-components anymore - they make class layout less readable Jun 28 17:33:43 freesmartphone.org: 03morphis 07msmcomm * r764bb295505a 10/msmcommd2/src/ (4 files): msmcommd2: start and use remote client handler from worker class Jun 28 17:33:45 freesmartphone.org: 03morphis 07msmcomm * rc172962e62d7 10/msmcommd2/src/flowcontrolhandler.vala: Jun 28 17:33:45 freesmartphone.org: msmcommd2: cleanup flow control handler Jun 28 17:33:45 freesmartphone.org: - we should not remove list items while iterating over the same list Jun 28 17:33:45 freesmartphone.org: - check frame seq number if it is the one we are expecting to recieve from the modem Jun 28 17:33:46 freesmartphone.org: - add some more comments to clarify things Jun 28 17:33:46 freesmartphone.org: 03morphis 07msmcomm * rfaf45997b999 10/msmcommd2/.gitignore: msmcommd2: add project local .gitignore file Jun 28 17:33:47 freesmartphone.org: 03morphis 07msmcomm * r83fa0d311069 10/msmcommd2/src/ (5 files): msmcommd2: rework handling of incomming frames with set ack number Jun 28 17:33:48 freesmartphone.org: 03morphis 07msmcomm * r58a8cbcac34c 10/msmcommd2/src/ (flowcontrolhandler.vala linkcontext.vala): Jun 28 17:34:05 freesmartphone.org: msmcommd2: move ack timer from flow control handler to link context Jun 28 17:34:05 freesmartphone.org: We need to trigger the ack timer for different places. We need to stop it within the flow Jun 28 17:34:05 freesmartphone.org: control handler when a frame recieves and to start in the transmission control when we Jun 28 17:34:05 freesmartphone.org: send a frame to the modem. Jun 28 17:54:35 * [Rui] waves Jun 28 18:00:26 hi! Jun 28 18:00:32 there is no /etc/ld.so.conf in the image Jun 28 18:00:40 I created my own but ldconfig complains about a python .so file (??) Jun 28 18:01:29 there is no /etc/ld.so.conf in the image Jun 28 18:01:30 I created my own but ldconfig complains about a python .so file (??) Jun 28 18:01:38 (reposting due to split) Jun 28 18:22:11 ehlo Jun 28 18:22:15 the dialer used in shr-t; where is the source code for that? Jun 28 18:22:35 it doesn't seem to be phoneui-apps.git Jun 28 18:23:44 b0ef: libphone-ui-shr.git Jun 28 18:23:50 <[Rui]> libphone-ui-shr.git Jun 28 18:25:47 thanks Jun 28 18:37:11 guys do you have some working fb.modes for qvga in 2.6.32? Jun 28 18:37:54 ehm fbset doesn't work, ok :S Jun 28 18:40:41 it's too bright :S Jun 28 18:42:54 if my NAND is apparently misbehaving, are the instructions at the bottom of http://wiki.openmoko.org/wiki/NAND_bad_blocks likely to sort it out? Jun 28 18:43:01 .. or should I just switch to running from NAND? Jun 28 18:43:06 s/NAND/SD/ Jun 28 18:48:41 lol for the too bright Jun 28 18:50:56 :( Jun 28 19:07:23 Weiss: sd seem faster btw. Jun 28 19:08:38 I was running UBI before.. it feels a lot faster than JFFS2 Jun 28 19:11:38 Weiss: hmm.. need to compare rw speed. Jun 28 19:12:19 with Qi, is it simply a case of putting the rootfs .tar.gz onto the SD card, appropriately formatted, using an SD reader? Jun 28 19:12:38 Weiss: yup Jun 28 19:12:44 Weiss: i am happy uboot user :) Jun 28 19:13:15 uboot+DS? Jun 28 19:13:16 SD* Jun 28 19:13:36 ok, I need not so much to be happy, in rare cases it not working :) Jun 28 19:14:07 Weiss: all dev on sd (debian, .34), all phoning on qtmoko (nand) Jun 28 19:15:17 mrmoku: thanks Jun 28 19:17:30 GNUtoo|laptop: thanks for test Jun 28 19:17:45 np thanks for the recipe Jun 28 19:17:59 btw I will have to patch xf86-input-tslib Jun 28 19:18:36 "patch it to work with the devPrivates changes in new Xorg servers" Jun 28 19:18:41 like they say in #xorg Jun 28 19:21:52 ah sorry I didn't notice that while setting new 1.9 xserver as default for whole SHR Jun 28 19:22:03 I'm using only evdev.. so I didn't notice Jun 28 19:26:51 JaMa, ok np but someone will have to patch it Jun 28 19:27:16 mrmoku: ping Jun 28 19:27:17 or to bugreport to penguintronics Jun 28 19:30:43 JaMa|AFK: --disable-kms should be fixed now Jun 28 19:33:33 hi, what are the voltages of the serial on the openmoko debug board Jun 28 19:39:08 onen|openBmap: pong Jun 28 19:42:54 AWK appeared in 1977 and it's still mad useful, at least for those who doesn't want to learn perl, heh. Jun 28 19:48:35 * JaMa|AFK likes AWK more then perl Jun 28 19:48:54 Weiss: thanks, cz_jc can you test it please? Jun 28 19:49:13 JaMa: hmm... lots of alignment traps I have :/ Jun 28 19:49:25 JaMa: with stuff like dmesg for example Jun 28 19:50:07 mrmoku: read ML 14:25 Re: [Shr-Devel] New SHR-unstable images with 2.6.32 kernel Jun 28 19:50:24 oh, ok :) Jun 28 19:50:27 mrmoku: khem bumped eglibc SRCREV I'm building it here to test Jun 28 19:51:08 JaMa: another from scratch one? Jun 28 19:53:35 mrmoku: I hope not, but don't know if they fixed something important in gcc for us Jun 28 19:54:16 mrmoku: as I don't know about any pending issue from GCC in our packages (except old QtXml/libxml which can be bumped locally) Jun 28 19:54:30 JaMa: phew... ok :P Jun 28 19:58:54 mrmoku: sorry. re-ping Jun 28 19:59:01 onen|openBmap: re-pong :P Jun 28 19:59:14 mrmoku: how is it about last shr image? Jun 28 19:59:28 mrmoku: is it safe to flash it to fix openbmap? Jun 28 19:59:43 onen|openBmap: I think so... GSM and GPS work both for me Jun 28 20:00:14 mrmoku: no issue with x server, u sd gets mounted automatically (I have read something on the ML about auto mounting being disabled...) Jun 28 20:00:48 onen|openBmap: x is fine... if you have an _extra_ partition on SD you want to be mounted automatically you have to add it to fstab Jun 28 20:00:53 mrmoku: ok. I leave on thursday evening. so this is quite a tight schedule thought. I 'll try Jun 28 20:01:01 :) Jun 28 20:01:20 mrmoku: thanks for your help Jun 28 20:01:31 onen|openBmap: np Jun 28 20:07:01 leviathan: hm I don't know. we have micro usb Jun 28 20:07:10 hmm Jun 28 20:07:11 ok Jun 28 20:07:14 leviathan: did you manage to find it out ? Jun 28 20:07:36 It seems that I'll need pullup resistors Jun 28 20:07:38 :-) Jun 28 20:07:39 you can just connect an measure Jun 28 20:07:43 I did Jun 28 20:07:54 1kΩ Jun 28 20:08:46 hm Jun 28 20:09:32 what a coincident that I've got 20 1kΩ resistors on my table Jun 28 20:09:34 :-) Jun 28 20:09:49 they are standart pullup resistors for 5V Jun 28 20:09:55 and I already worked with ft232 Jun 28 20:10:02 so I'm sure now Jun 28 20:10:10 I'll need to pull up the levels Jun 28 20:10:21 what funny stuff Jun 28 20:10:46 that I always come to the same point, that what I'm using at home is used in industry too Jun 28 20:10:48 ^^ Jun 28 20:11:01 means inventing ways Jun 28 20:15:50 anyway... I guess I'll smoke now Jun 28 20:15:56 and then I go to bed Jun 28 20:15:58 :-) Jun 28 20:16:41 * mrmoku wonders if that sounds like a good plan ;) Jun 28 20:17:17 leviathan: can you update wiki with your dream cable experiences ? Jun 28 20:18:48 uhm Jun 28 20:18:56 dcordes: as soon as it works I will :-) Jun 28 20:19:54 ok nice Jun 28 20:40:11 Mirv: you could be also interested weiss's patch for --disabled-kms Jun 28 20:41:27 JaMa, Weiss good to hear, I'm on it :) Jun 28 20:41:47 JaMa, btw. your rev works Jun 28 20:42:45 rev? Jun 28 20:43:21 JaMa, 'Remove x*alloc/xfree calls deprecated in xserver 1.9 RC1' Jun 28 20:43:44 JaMa, this is the official stable snapshot for gentoo now, once I push it :) Jun 28 20:43:49 cz_jc: without KMS enabled kernel right? but with old xserver 1.8* Jun 28 20:44:11 JaMa, nope, I tested that one with kms.. I'm trying the new one without kms Jun 28 20:44:35 it'll take 7 minutes or so on the FR to compile (aclocal takes forever) Jun 28 20:44:39 cz_jc: here it fails only when booted kernel without KMS support Jun 28 20:44:54 cz_jc: with same rootfs but KMS enabled kernel it works normally Jun 28 20:45:21 JaMa, dunno if I have it, I still have 2.6.29 Jun 28 20:45:52 and both drivers are built with --enable-kms, only on non-KMS kernel it fallback to swrast automatically Jun 28 20:46:08 and the fallback is now broken Jun 28 20:46:19 JaMa, how so ? Jun 28 20:46:38 cz_jc: probably not, it's not in andy-tracking branch (you would need to apply weiss patches too) Jun 28 20:47:02 JaMa, I didn't clone andy-tracking, it was some other head Jun 28 20:47:17 *checkout Jun 28 20:47:36 I don't have the commit string anymore though.. Jun 28 20:47:37 cz_jc: http://lists.shr-project.org/pipermail/shr-devel/2010-June/002785.html Jun 28 20:48:53 cz_jc: drm-traking is old 2.6.29 with weiss's patches.. but that was before WSOD on every resume fix.. so you would notice using it :) Jun 28 20:49:21 JaMa, well.. that's a new one.. Jun 28 20:49:30 (I mean the error) Jun 28 20:49:49 JaMa, yes, it was drm-tracking Jun 28 20:51:46 cz_jc: and do you have MFD_GLAMO_DRM enabled? (instead of MFD_GLAMO_FB) Jun 28 20:52:11 JaMa, nope, CONFIG_MFD_GLAMO_FB=y Jun 28 20:52:42 JaMa, so looks like we're about to confirm your error then :) Jun 28 20:52:49 I'm at configure now Jun 28 20:53:32 checking whether to use KMS... no Jun 28 20:54:11 JaMa, it didn't check for DRI, that means it's not dependant on libdrm when without kms ? Jun 28 20:54:35 what xserver version are you using? 1.8.1.902 Jun 28 20:55:45 yup you shouldn't need libdrm (with glamo patches for glamo DDX and mesa patches for mesa-dri Jun 28 20:55:55 JaMa, 1.7.6 Jun 28 20:56:03 perfect Jun 28 20:56:14 cz_jc: this happen for first time with 1.9 RC1-RC3 Jun 28 20:56:14 JaMa, the KMS symbols aren't in nm anymore Jun 28 20:57:09 cz_jc: but I didn't want to downgrade to 1.8 xserver, because of different ABI between those 2 (rebuilding all drivers again..) Jun 28 20:57:20 JaMa, btw. does it mean that when --disable-kms, it doesn't use direct rendering ? Jun 28 20:58:33 JaMa, well, xorg started just fine Jun 28 20:58:43 I'll try to load fluxbox and play with it Jun 28 20:58:57 I think no, but Weiss knows all details (http://www.bitwiz.org.uk/s/how-dri-and-drm-work.html can give you some hints) Jun 28 21:00:38 nice blog :) Jun 28 21:00:55 I'll try also xrandr (that got broken pretty often) Jun 28 21:01:46 JaMa, well it works nicely :) Jun 28 21:01:52 no errors so far Jun 28 21:03:40 anybody able to fetch eds-dbus_svn.bb ? Jun 28 21:03:48 wanna make sure problem is not on my side Jun 28 21:08:40 dcordes: svn: Could not open the requested SVN filesystem Jun 28 21:08:58 dcordes: and I've already removed it from my downloads dir Jun 28 21:12:25 hmm sucks Jun 28 21:14:45 thanks for trying it Jun 28 21:31:10 SHR: 03zub 07libphone-ui-shr * r8e8aa6f1ddd1 10/src/view/quick-settings-view.c: quick-settings: try harder to retrieve the list of profiles Jun 28 21:31:18 SHR: 03zub 07libphone-ui-shr * rcb8423d5a255 10/src/ (9 files in 2 dirs): Fix code to follow libphone-ui-shr conventions (suggested by TAsn). Jun 28 21:31:44 sorry for the huge delay ;( Jun 28 21:33:41 leviathan, /serial_cable_status Jun 28 21:34:24 <[Rui]> When are you guys considering to do the next efl bump? Jun 28 21:34:27 dcordes, JaMa I remember not beeing able to fetch it either Jun 28 21:36:02 [Rui]: as soon as there is something interesting :) Jun 28 21:37:06 <[Rui]> JaMa: hehe, well... nothing much on EFL, I'd say, but as you've pinned elmdentica do E17 I was just wondering when elmdentica users could take benefit from the recently implemented gag lsits Jun 28 21:37:12 <[Rui]> s/lsits/lists Jun 28 21:38:23 [Rui]: I can bump only elmdentica if you want Jun 28 21:40:59 <[Rui]> Well, if it isn't too much trouble, then could you bump it to 49930 please? Jun 28 21:42:04 hm, does anyone know what is s3c-sdi used for on the FR? Jun 28 21:42:06 [Rui]: do I need to test it first (on device) or can I belive you that our older efl and your newer elmdentica will be ok? Jun 28 21:43:23 <[Rui]> JaMa: nothing new efl-wise in elmdentica, nothing but the usual elementary functions (and not even the most recent fancy objects), so it should be a-ok Jun 28 21:43:37 okey Jun 28 21:44:05 <[Rui]> JaMa: I mean, my local copy is from 201005142132 and it works on my laptop Jun 28 21:45:01 <[Rui]> gag lists are cool if one wants to avoid trolls or trendy topics (like vuvu.*) Jun 28 21:45:11 [Rui]: ok, please ping me when you notice next efl bump without upgrading elmdentica (in case I forgot to remove it or someone else does the EFL bump) Jun 28 21:45:28 * [Rui] will :) Jun 28 21:46:04 <[Rui]> JaMa: oh, and thanks for using libjson_svn in shr-u, the author seems unresponsive Jun 28 21:46:31 yw Jun 28 21:47:52 <[Rui]> weird... Jun 28 21:47:55 <[Rui]> NOTE: Running task 14 of 7742 (ID: 1057, /media/devdisk/openmoko/shr-unstable/openembedded/recipes/stage-manager/stagemanager-native_0.0.1.bb, do_package) Jun 28 21:47:55 <[Rui]> cp: cannot stat `/media/devdisk/openmoko/shr-unstable/tmp/work/x86_64-linux/stagemanager-native-0.0.1-r12/sysroot-destdir//media/devdisk/openmoko/shr-unstable/tmp/*': No such file or directory Jun 28 22:21:44 [Rui]: newer elmdentica will be built tonight (after gcc/eglibc combo), but I see there are new snapshots in efl.. so maybe I'll test and bump sooner Jun 28 22:24:20 <[Rui]> JaMa|ZZz: just as well, good night! :) **** ENDING LOGGING AT Tue Jun 29 02:59:57 2010