**** BEGIN LOGGING AT Thu Jun 21 02:59:58 2012 Jun 21 05:59:32 SHR: 03Martin.Jansa 07meta-smartphone * re2926e17e9f8 10/meta-fso/recipes-freesmartphone/cornucopia/cornucopia-from-git.inc: meta-fso: cornucopia: bump SRCREV Jun 21 05:59:42 SHR: 03Martin.Jansa 07meta-smartphone * r32b602d8e1f8 10/meta-fso/recipes-freesmartphone/cornucopia/ (2 files in 2 dirs): meta-fso: libfsobasics: backport one patch from master to fix build of mterm2/mkdump with vala-0.16 and without fso-autorev Jun 21 06:35:21 SHR: 03shr-devel 07buildhistory * ra97ef8d50c9b 10/packages/armv4t-oe-linux-gnueabi/ (35 files in 35 dirs): packages: Build 201206210816 of shr 20120621 for machine om-gta02 on opmbuild Jun 21 06:50:01 SHR: 03shr-devel 07buildhistory * r0d1090106bae 10/packages/armv7a-vfp-neon-oe-linux-gnueabi/ (35 files in 35 dirs): packages: Build 201206210837 of shr 20120621 for machine nokia900 on opmbuild Jun 21 07:13:44 SHR: 03shr-devel 07buildhistory * rc3e18994c847 10/images/crespo/eglibc/chroot-image/ (build-id installed-package-sizes.txt): images: Build 201206210906 of shr 20120621 for machine crespo on opmbuild Jun 21 07:16:10 Hi all. Someone from SHR project ? Jun 21 07:18:01 nachognu: best ask your question rather than asking to ask Jun 21 07:19:29 Sorry :) Long time I'm not using IRC. I'm trying to install new SHR on my Freerunner (I've been using old for long time) and I'm not able to install the UBI image using neotool Jun 21 07:20:30 which image are you trying to install? what error are you getting when flashing it? Jun 21 07:20:55 And what bootloader are you using. Jun 21 07:21:13 Qi-ubi Jun 21 07:21:19 the old image was ubi also Jun 21 07:22:07 I tried some months ago and I do not susceed, so I recovered my backup image. I wanted to try now again (because I'm not able to mount my SD card) and same issue Jun 21 07:22:20 I', trying to install now jffs2 to try Jun 21 07:23:08 please paste a link to the image and the error log to pastebin.com Jun 21 07:23:08 I have no error flashing, but when start I see "Kernel Panic- Not syncing: VFS unable to mount root fs on unknonw-block(0,0) Jun 21 07:23:35 sounds like you are booting from SD card rather than NAND? Jun 21 07:23:44 nop. no SD card on the system Jun 21 07:24:03 and I used the normal procedure to flash (neotool + image...) Jun 21 07:24:14 hmm, I only ever got that error with SD Jun 21 07:29:23 nachognu: you're flashing .ubi, not .ubifs, right? Jun 21 07:29:28 jep Jun 21 07:29:37 Copying the error message.... Jun 21 07:29:42 Image I'm using: Jun 21 07:29:43 http://build.shr-project.org/shr-2012.07/images/om-gta02/shr-image-om-gta02-20120611074304.rootfs.ubi Jun 21 07:29:43 Bootloader : Jun 21 07:29:43 http://build.shr-project.org/shr-2012.07/images/om-gta02/qi-jffs2-s3c2442.udfu Jun 21 07:29:43 Kernel : Jun 21 07:29:43 http://build.shr-project.org/shr-2012.07/images/om-gta02/uImage-2.6.39-r12-om-gta02-20120525082154.bin Jun 21 07:30:16 nachognu: qi-jffs2 won't work with ubi Jun 21 07:31:57 Let me try again with the QI-UBI (I do not know now if I messed up the copy or the flash...) Jun 21 07:32:20 but I tryied first with the current bootloader I had from the old UBI version and also failed Jun 21 08:48:17 SHR: 03shr-devel 07buildhistory * r1bde51d32152 10/packages/armv4t-oe-linux-gnueabi/ (134 files in 134 dirs): packages: Build 201206210942 of shr 20120621 for machine om-gta02 on opmbuild Jun 21 09:00:29 The error: http://pastebin.com/ma4eUpFt Jun 21 09:01:04 I've copied manually the error message, I was not able with "cu -l" to retrieve the error on the console Jun 21 09:01:34 If you want I can try with the jffs2 image. Jun 21 09:09:17 nachognu: hm, looks like that ubi image is broken, or probably some major incompatibility between the ubi version used to create an image and the kernel ubi implementation. Jun 21 09:38:46 SHR: 03shr-devel 07buildhistory * rd93fa4a821df 10/packages/armv7a-vfp-neon-oe-linux-gnueabi/ (134 files in 134 dirs): packages: Build 201206211050 of shr 20120621 for machine nokia900 on opmbuild Jun 21 09:55:49 I though the same, but I've tried with other UBI images and are not working. I'm wondering if someone is using UBI format. Jun 21 09:57:17 using same computer / USB wire / NeoTool, I've been able to install my old UBI system. Trying to install JFFS2 format Jun 21 10:00:35 The MD5 for the UBI image is the same from the one on the server, so just could be the flashing method is failing or the image is not working Jun 21 10:11:04 SHR: 03shr-devel 07buildhistory * rdd377245614a 10/images/crespo/eglibc/chroot-image/ (build-id installed-package-sizes.txt): images: Build 201206211156 of shr 20120621 for machine crespo on opmbuild Jun 21 10:15:43 Hi, I'm looking for the FLUID source code, it was at svn.internal.openmoko.org/trunk/;module=fluid but it doesn't seem to be here any more.. Jun 21 10:16:11 Anyone knows where could I get it from? Jun 21 10:18:12 svn.internal? that doesn't sound like something that was ever public Jun 21 10:19:02 pabs3: actually atm http://svn.internal.openmoko.org/track is available so.. maybe it was someway public.. Jun 21 10:24:08 I would be very surprised if the fluid source was ever released. anyway, if it was and now isn't, that is probably a side effect of the recent server changes so you would have to contact Harald Welte: http://lists.openmoko.org/pipermail/community/2012-May/066975.html Jun 21 10:24:52 PaulFertser, DocScrutinizer: did you ever hear of the source for fluid being released?^ Jun 21 10:30:52 pabs3: I need it mainly to understand in which way the m0 input file is read (since it isn't the firmware binary file). With the fluid src we can understand how to flash a firmware into the calypso, we are trying to do this for the osmocom project.. Jun 21 10:35:10 bertani: thats part is easy, file says 'Motorola S-Record; binary data in text format' and you should use the srecord program to decode it to a binary file Jun 21 10:37:43 pabs3: you are right! But we still need the code to understand how the flashing process works Jun 21 10:38:16 bertani: but I wonder if flashing is needed at all since osmocom can just be loaded into ram? Jun 21 10:40:13 pabs3: actually it could also be flashed by using osmocon (http://bb.osmocom.org/trac/wiki/flashing) Jun 21 10:41:18 I guess osmocon doesn't yet know how to flash gta02? Jun 21 10:43:16 so my suggestions are: contact Harald Welte, if he isn't able to release it then you would need to reverse engineer how fluid is doing it Jun 21 10:44:53 pabs3: afaik osmocom can be flashed to FR Jun 21 10:45:03 osmocombb Jun 21 10:46:00 lindi-: not by using osmocon Jun 21 10:46:21 btw, is there some mail or web page that suggested the FLUID source code was public? Jun 21 10:47:43 pabs3: no, I just looked at the control file of the ipk and there was the reference to svn.internals.openmoko.org/trunk... but probably it wasn't public Jun 21 10:48:03 ah Jun 21 11:03:13 bertani: Harald just replied on the osmocombb mailing list Jun 21 11:03:50 pabs3: I know =) Jun 21 11:29:03 pabs3: what's FLUID now? Jun 21 11:29:26 the stupid flasher for calypso? Jun 21 11:29:30 the gta02 GSM firmware flasher, yeah Jun 21 11:29:55 it's obsolete since we ported it to ARM Jun 21 11:30:13 see MOKO11 GSM image Jun 21 11:30:35 and yes, it actually been semi-public once Jun 21 11:31:38 DocScrutinizer51: and you still have the src, right? :) Jun 21 11:31:54 I miht Jun 21 11:32:32 should be somewhere on people(.internal).openmoko.org/* Jun 21 11:33:21 joerg Jun 21 11:33:35 (maybe?) Jun 21 11:33:43 but actually I wonder where from I took the flashing tool (whatever it been) for that uSD MOKO11 image Jun 21 11:34:03 for sure not on mine, no Jun 21 11:34:34 bertani: ping me in 6h again Jun 21 11:34:38 maybe 7 Jun 21 11:34:48 not @ home now Jun 21 11:34:56 DocScrutinizer51: ok, thank you! Jun 21 11:35:05 yw Jun 21 11:42:41 bertani: http://www.google.com/search?q=FLUID+openmoko Jun 21 11:42:56 might yield some helpful links Jun 21 11:43:57 DocScrutinizerWK: I have already looked around on google but.. no luck Jun 21 11:45:16 ""Install http://people.openmoko.org/joerg/calypso_moko_FW/fluid_0.0+svn20070817-r2_armv4t_eabi.ipk on your device: "" Jun 21 11:48:50 that's the obsolete version? Jun 21 11:49:04 I might come to think we never had allowance from TI to pusblish src of FLUID Jun 21 11:49:32 only binary Jun 21 11:49:33 DocScrutinizerWK: well, this is the ipk but the src is not there.. Jun 21 11:49:41 DocScrutinizerWK: :( Jun 21 11:54:58 http://www.mail-archive.com/hardware@lists.openmoko.org/msg00289.html Jun 21 11:58:28 DocScrutinizerWK: I see, so probably you aren't allowed to share the src of fluid Jun 21 11:59:07 probably not, but I guess it might have "escaped" meanwhile Jun 21 11:59:26 together with a lot of other calypso stuff Jun 21 12:07:08 http://bb.osmocom.org/trac/wiki/osmocon Jun 21 12:07:22 might know more ;-D Jun 21 12:08:11 DocScrutinizerWK: got it :) Jun 21 12:10:06 otherwise the binary is available for RE anytime Jun 21 12:10:33 gott awork ;-D Jun 21 12:10:36 o/ Jun 21 12:12:19 thank you, bye! o/ Jun 21 12:37:35 mickey_office: ping Jun 21 12:39:35 mickey_office: about your question: I reverted the delegate class thing as it introduces a bug for the GTA02 somewhere Jun 21 12:40:16 I have not found it in a reasonable time so I reverted the changes Jun 21 12:48:05 SHR: 03shr-devel 07buildhistory * rbe8cc837bd4a 10/images/om_gta02/eglibc/shr-image/ (11 files): images: Build 201206211421 of shr 20120621 for machine om-gta02 on opmbuild Jun 21 12:51:33 mickey_office: if you have some free time I have a lot of API related things to talk about Jun 21 13:16:15 SHR: 03shr-devel 07buildhistory * rce10f6778854 10/images/nokia900/eglibc/shr-image/ (11 files): images: Build 201206211451 of shr 20120621 for machine nokia900 on opmbuild Jun 21 13:16:50 freesmartphone.org: 03morphis 07morphis/device-status * r675acad2f151 10specs/ (3 files in 3 dirs): org.freesmartphone.GSM: add enumeration for the device functionality level Jun 21 13:42:46 morphis: i try to free some time soon. one question now though… do you think it's good to get rid of the fine granular ALIVE_ states? how do you cope with command sequences now that have to be sent in different states? Jun 21 13:43:45 SHR: 03shr-devel 07buildhistory * rf6f8618f805e 10/images/om_gta04/eglibc/shr-image/ (11 files): images: Build 201206211518 of shr 20120621 for machine om-gta04 on opmbuild Jun 21 14:10:33 SHR: 03shr-devel 07buildhistory * r7ed3049f8470 10/images/crespo/eglibc/shr-image/ (11 files): images: Build 201206211545 of shr 20120621 for machine crespo on opmbuild Jun 21 14:36:04 bertani: maybe wpwrak knows details about FLUID, where from src came, and if it ever been public Jun 21 14:38:26 meh, gone Jun 21 14:39:56 * DocScrutinizer51 idly wonders what for bertani needed FLUID source Jun 21 14:48:46 DocScrutinizer51: for flashing osmocombb to the calypso baseband flash on gta02: http://lists.osmocom.org/pipermail/baseband-devel/2012-June/003095.html Jun 21 14:58:58 DocScrutinizer06: afaik, we only released the binary Jun 21 14:59:13 source belongs to TI Jun 21 14:59:52 here's an ARM ipk, if that helps: http://people.openmoko.org/joerg/calypso_moko_FW/fluid_0.0+svn20070817-r2_armv4t_eabi.ipk Jun 21 15:10:56 Apparently, Harald has already answered the question, or at least provided a patch to try. Jun 21 16:24:10 mhm Jun 21 16:56:42 Damn, tried SHR's 2.6.39. It doesn't boot even because it can't read the partition table from my uSD (even though it detects it). Jun 21 16:57:23 JaMa: hey :) is it you now whom i to report a bug? Jun 21 17:02:05 PaulFertser: did you try the usual uSD workarounds? Jun 21 17:02:28 lindi-: no, as my card used to work fine without them. Jun 21 17:03:25 lindi-: thanks for the hint. It should have been obvious for me, heh Jun 21 17:03:40 PaulFertser: different kernel, different issues unfortunately Jun 21 17:03:58 PaulFertser: there is this kernel option for some SD clock Jun 21 17:04:06 can't remember the exact name right now Jun 21 17:05:12 lindi-: oh, i do :) Jun 21 17:05:45 glamo_mci.sd_max_clk Jun 21 17:05:48 PaulFertser: also http://lindi.iki.fi/lindi/openmoko/sd_cards.txt Jun 21 17:06:54 lindi-: funny thing, tweaking that option is something i would suggest myself to anyone having problems with sd card. But what confused me is that it always worked for me fine on 2.6.29 Jun 21 17:07:05 :) Jun 21 17:10:29 lindi-: setting it to 10MHz made 2.6.39 boot Jun 21 17:10:40 lindi-: thanks! Jun 21 17:11:01 It seems i started to loose my head because of emotions and you got me back on track Jun 21 17:15:34 And it feels like it works really faster than 2.6.34 from qtmoko. Jun 21 17:17:05 hah Jun 21 17:18:28 PaulFertser, hi Jun 21 17:18:50 PaulFertser, since you run on microsd you can put a file that append the cmdline Jun 21 17:18:51 GNUtoo-hplaptop: hi! Jun 21 17:18:58 which contains stuff like for instance: Jun 21 17:19:01 GNUtoo-hplaptop: :D Jun 21 17:19:23 GNUtoo-hplaptop: my openmoko adventures several years ago started with hacking on Qi code :) Jun 21 17:19:31 ok Jun 21 17:19:43 glamo_mci.sd_max_clk=2000000 Jun 21 17:19:44 try that Jun 21 17:19:55 for me it makes partition detection work on my 32G card Jun 21 17:20:27 basically we may have an upgraded kernel at some point of time Jun 21 17:20:33 I just need a debug cable back Jun 21 17:21:14 I said may because it depend on how much bugs I get and on my ability to fix them fast enough Jun 21 17:21:37 GNUtoo-hplaptop: 10MHz works for me Jun 21 17:22:00 ok Jun 21 17:22:03 then try 10Mhz Jun 21 17:22:06 GNUtoo-hplaptop: in fact one can connect ~any uart-usb converter by soldering to the testpoints. Jun 21 17:22:17 Yes, i tried and booted 2.6.39 Jun 21 17:22:32 hmmm but then I need to use gta02 as main phone Jun 21 17:22:38 *to be able to Jun 21 17:52:02 Hm, soundcard is not detected on 2.6.39 though the modules were loaded automatically. Is this to be expected? ;) Jun 21 17:53:26 In fact it's not that funny for somebody willing to use his phone :) Jun 21 18:01:54 PaulFertser: yes need for glamo_mci.sd_max_clk started for some uSD with .31 already and mine was fine until .39 iirc and now is again fine with 3.2 Jun 21 18:02:34 JaMa: funny... Jun 21 18:02:37 PaulFertser: with shr rootfs or qtmoko? Jun 21 18:02:54 JaMa: with shr modules Jun 21 18:03:06 JaMa: plain Debian here, as always :) Jun 21 18:03:21 some modules were renamed, are you sure you have right set loaded? :) Jun 21 18:03:38 in 3.2 you need to load them in right order too :/ Jun 21 18:03:40 JaMa: they're auto-loaded based on modaliases Jun 21 18:04:03 SHR root@gjama ~ $ cat /etc/modules-load.d/snd-soc-neo1973-wm8753.conf Jun 21 18:04:03 snd-soc-s3c24xx Jun 21 18:04:03 snd_soc_s3c24xx_i2s Jun 21 18:04:03 snd-soc-dfbmcs320 Jun 21 18:04:03 snd-soc-wm8753 Jun 21 18:04:05 snd-soc-neo1973-wm8753 Jun 21 18:04:37 auto-loading didn't work for me.. Jun 21 18:05:40 That's damn not funny Jun 21 18:05:52 loading snd-soc-dfbmcs320 in random order hanged the device Jun 21 18:09:23 Ok, "modprobe snd-soc-dfbmcs320; modprobe snd-soc-neo1973-wm8753" brought sound back to me. Jun 21 18:11:06 JaMa: big thanks for the hint, really important to not let me use nokia1100 again. Jun 21 18:26:59 yw Jun 21 18:35:23 jeepingben: hi, can you confirm that 065 issue was only because of missing depmod -a call? Jun 21 18:36:02 BTW, the modules i untarred were not depmod Jun 21 18:41:30 PaulFertser: yup opkg install kernel-module-foo is suposed to f that Jun 21 18:41:46 and people untarring them are on their own Jun 21 18:41:57 JaMa: hehe Jun 21 18:42:54 that's what README.html says :) http://build.shr-project.org/shr-kernels/README.html Jun 21 19:07:28 PaulFertser: did you say something about Qi? :) src/cpu/s3c2442/start_qi.c:39:4: error: 'NULL' undeclared here (not in a function) with latest gcc-4.7.1 Jun 21 19:15:44 SHR: 03Martin.Jansa 07shr-chroot * r5cb45de42d80 10/ (79 files in 20 dirs): system upgrade Jun 21 19:17:28 JaMa: #include Jun 21 19:18:39 will do when other builds are completed just to let you know.. Jun 21 19:20:52 JaMa: thanks :) Jun 21 19:49:17 SHR: 03Martin.Jansa 07meta-smartphone * r1d38f1b483d2 10/meta-shr/recipes-connectivity/connman/ (connman/connman connman_0.79.bbappend): connman: drop .bbappend and custom initscript Jun 21 19:49:17 SHR: 03Martin.Jansa 07meta-smartphone * r1e01cb073fcb 10/meta-openmoko/recipes-bsp/qi/ (2 files in 2 dirs): qi: fix build with gcc-4.7.1 Jun 21 20:38:10 yoyoyo Jun 21 20:38:21 Receiving objects: 86% (781/904), 92.54 MiB | 88 KiB/s ects: 86% (781/904), 91.97 MiB | 96 KiB/s iB/s Jun 21 20:38:45 *BURRP* Jun 21 20:40:37 * DocScrutinizer05 wonders if those android devels ever heard of ncurses Jun 21 20:40:58 or plain vt100 Jun 21 20:41:18 even ANSII ESC sequences Jun 21 20:42:06 repo install asked if I want to use colors and shite Jun 21 20:42:38 repo sync clutters my xterm with rubbish Jun 21 21:40:33 JaMa, hi Jun 21 21:40:39 could you take care of http://www.shr-project.org/trac/ticket/2019 ? Jun 21 21:40:57 it would fix the cases where you have to ifdown usb0;ifup usb0 to get networking work Jun 21 22:08:25 GNUtoo: hi Jun 21 22:08:41 JaMa, depmod -a didn't work for me. From what I could tell, /boot/uImage was still pointed at the old kernel Jun 21 22:08:42 hi Jun 21 22:10:39 GNUtoo: do I need anything except cloning some repos and making a manifest to set up a custom android tree on a git server? and how to keep track of the master if I, say, fork cm9? Jun 21 22:11:06 Alex[sp3dev], hi, do you require a free version of cm9? Jun 21 22:12:01 GNUtoo: no, we're just planning to set up a custom private tree for our project Jun 21 22:12:52 ok Jun 21 22:13:06 on another note, I'm writing galaxy s2/nexus firmware loader and RIL support for libsamsung-ipc. I'm disassembling the blob and using the unfinished code from grindars. Hope to finish it by monday Jun 21 22:13:14 basically you need to setup a manifest Jun 21 22:13:14 http://pastebin.com/aeCFnNj4 at least, I got some reply from the modem Jun 21 22:13:19 in a git server Jun 21 22:13:30 ok Jun 21 22:13:58 mirage335, ^^^ Jun 21 22:14:16 ah sorry Jun 21 22:14:19 sgs2 Jun 21 22:14:24 not galaxy nexus Jun 21 22:15:27 GNUtoo: well, I'm planning to write support for both. they're 90% the same, except for fw location. and well, I will have to add nexus support anyway for our project @ work Jun 21 22:15:58 ok Jun 21 22:16:03 mirage335, ^^^ Jun 21 22:16:30 * Alex[sp3dev] wishes someone gave me hex rays Jun 21 22:16:36 lol Jun 21 22:17:54 jama, just happened again when upgrading from 067 to 068. UImage still points to old kernel Jun 21 22:18:17 sometimes I'm thinking of writing a decompiler. as in asm to C translator. but then I'm thinking that reverse engineering is silly and I should write something useful myself instead of digging others' crap Jun 21 22:29:45 Alex[sp3dev]: reverse engineering is not silly Jun 21 22:30:36 Alex[sp3dev]: also such a translator is not really the way to go, you want an interactive tool Jun 21 22:30:58 like integrated to gdb? Jun 21 22:31:01 or gdb scripts Jun 21 22:32:00 GNUtoo: like an IDA plugin Jun 21 22:32:21 there is something integrated to metasploit tough Jun 21 22:32:22 metasm Jun 21 22:32:59 lindi-: I feel pretty silly when people ask me what I'm doing at my free time or at work and I'm like "hacking drivers" Jun 21 22:33:08 it's a separate project tough Jun 21 22:33:19 and the metasploit version is out of date it seems Jun 21 22:33:24 but I've not tried theses yet Jun 21 22:33:40 http://metasm.cr0.org/ Jun 21 22:35:40 GNUtoo: gdb is imho quite weak, it doesn't do anything symbolic really Jun 21 22:36:08 it can do source debugging on unknown binaries if you have the right symbols I think Jun 21 22:36:11 GNUtoo: probably building on top of llvm would be a better idea nowadays Jun 21 22:36:28 GNUtoo: yeah but it can't even tell you where are function prologues Jun 21 22:36:36 like for instance like the person reversing the qualcomm baseband did Jun 21 22:36:38 GNUtoo: since gdb does not understand the code at all Jun 21 22:36:53 ok Jun 21 22:37:01 it can disassemble it for the user as text but it doesn't understand what the code does Jun 21 22:37:11 ok Jun 21 22:37:14 what IDA does good is recognizing function start/end, symbol xrefs and common structure names. and you can define custom symbols on the fly Jun 22 02:28:42 * pabs3 wonders how the Boomerang decompiler is going http://boomerang.sourceforge.net/ Jun 22 02:29:07 last news 2006, not so well I guess **** ENDING LOGGING AT Fri Jun 22 02:59:59 2012