**** BEGIN LOGGING AT Mon Jul 21 02:59:56 2008 Jul 21 09:46:07 Jay7: hey Jul 21 09:46:43 ant|work731: i have some trouble with u-boot patching Jul 21 09:47:06 for akita? Jul 21 09:47:22 yes, for change partition sizes Jul 21 09:48:55 this would be 'standard' : +define MTDPARTS_DEFAULT "mtdparts=sharpsl-nand:7168k(Logical),54272k(root),69632k(home)" Jul 21 09:49:01 file "include/configs/akita.h" should be patched? Jul 21 09:49:42 but if I want also increase 'logical' partition? Jul 21 09:50:18 and what size is now better suitable for 'logical'? Jul 21 09:50:26 and for testing :) Jul 21 09:50:27 I tried with 10m Jul 21 09:50:47 let me find a link...one sec Jul 21 09:51:30 ok, I'll try 10*1024(Logical),100*1024(root),*(home) Jul 21 09:54:33 #define NAND_LOGICAL_SIZE 0x700000 Jul 21 09:54:57 this define should be changed also? Jul 21 09:56:35 Jay7: http://www.h5.dion.ne.jp/~rimemoon/zaurus/memo_006.htm Jul 21 09:57:56 ant|work731: ok, I'll read it Jul 21 09:58:42 Jay7: I cannot find it but I have a ready-exemple for you somewhere... Jul 21 09:59:24 ant|work731: I know one person who already patched u-boot and changed partitions on corgi Jul 21 09:59:36 but seems that her is in vacation now :) Jul 21 10:01:26 ant|work731: btw, what is better translator from japanese to english? Jul 21 10:01:36 yahoo/google/other? Jul 21 10:01:41 uh, here you have to ask Laibsch Jul 21 10:01:46 :) Jul 21 10:01:55 he can Japanese Jul 21 10:08:59 Jay7: found something Jul 21 10:08:59 http://host.xzshare.com/?d=0AD57C191 Jul 21 10:10:38 10x, I'm reading previous link now :) Jul 21 10:11:18 I think, I should translate it to russian for our zaurus community :) Jul 21 10:11:47 I put that addresses on the xls file Jul 21 10:12:04 and some OE docs need to be translated.. Jul 21 10:13:06 Jay7: there was a developer for akita called CP57 or something similar Jul 21 10:13:20 he is russian :-) Jul 21 10:13:27 svs57? Jul 21 10:13:33 the best Zaurus-kackers are russians!!! Jul 21 10:13:41 yes, google for it Jul 21 10:13:43 he is that person in vacation now :)) Jul 21 10:13:53 holy sh*t Jul 21 10:14:09 I meet with some zaurus dev's at next week in moscow Jul 21 10:14:39 Jay7: have you seen in oebug 3738 there is nandlogical.bb now ;-) Jul 21 10:14:45 but they mostly going to nokia IT Jul 21 10:15:07 ant|work731: yes, it's good news :) Jul 21 10:15:09 Jay7: I'm thinking we could put together a new emergency image Jul 21 10:15:31 now is possible to reflash without rebooting :) Jul 21 10:15:57 svs57 thinking about new emergence image too :) Jul 21 10:16:05 but he is debian user now :) Jul 21 10:16:15 there are these initramfs-images ready now Jul 21 10:16:21 these come from debian Jul 21 10:16:35 I have talk with Civil (gentoo user) about boot loaders Jul 21 10:17:07 he says that altboot is better alternative and u-boot patches is unsupportable Jul 21 10:17:16 forget altboot Jul 21 10:17:24 probably kexec Jul 21 10:17:55 today's siyuation is: you can flash initramfs-kexec image Jul 21 10:17:58 what amount of overhead is in kexec-way? Jul 21 10:18:09 zero Jul 21 10:18:18 (is not like altboot) Jul 21 10:18:18 only at boot time? Jul 21 10:18:29 no, it's a command Jul 21 10:18:41 executed after init Jul 21 10:19:05 but again, you have to hardcode the cmdline Jul 21 10:19:18 if you use the Sharp's bootloader Jul 21 10:19:21 hm.. but it solves only problem with small 'logical' Jul 21 10:19:43 can root be increased with initramfs? Jul 21 10:19:48 -images Jul 21 10:29:48 re Jul 21 10:29:54 Jay7: sure Jul 21 10:31:25 ant|work731: just by hardcoding cmdline and reflashing? Jul 21 10:34:05 sharpsl.c Jul 21 10:34:25 needs love Jul 21 10:35:02 Jay7: BTW the new kernels >2.6.23 now supports ATAG list Jul 21 10:35:30 but the Sharp bootloader (angel) doesn't provide a sane list Jul 21 10:35:42 that's one more reason for u-boot Jul 21 10:35:54 kernel cmdline would be 'empty' Jul 21 10:36:04 and all hardcoded in u-boot Jul 21 10:36:36 both ways have pro's and con's :) Jul 21 10:37:07 zaurus hw should be supported by upstream Jul 21 10:37:12 (u-boot) Jul 21 10:37:14 hi Jul 21 10:37:21 yeah Jul 21 10:37:44 can somebody help me with Xfbdev and tslib touchscreen raw input device? Jul 21 10:37:46 or we should reinvent the wheel.. Jul 21 10:37:52 he he Jul 21 10:38:12 and support it by self :) Jul 21 10:38:31 in ts_test it works just great but Xfbdev doesn't like it Jul 21 10:38:45 is here any story of u-boot patches pushing to upstream? Jul 21 10:39:36 dcordes: what is hw/sw versions? Jul 21 10:40:24 Jay7: hw is machine htckaiser with modified ts2003.c, software is tslib 1.0 Jul 21 10:40:36 mostly what angstrom-x11 builds Jul 21 10:40:51 dcordes: and Kdrive ver.? Jul 21 10:41:06 I wonder what I need to pass for "Xfbdev" to use the stuff correctly Jul 21 10:41:11 Jay7: not sure Jul 21 10:41:23 does angstrom-x11 use kdrive? Jul 21 10:41:33 ipkg list_installed | grep xorg-xserver Jul 21 10:41:39 yep Jul 21 10:41:51 on zaurus at least :) Jul 21 10:42:01 let me boot. need ext2fsck Jul 21 10:42:11 Xfbdev is Kdrive part Jul 21 10:42:40 dcordes: check TSLIB_* env variables Jul 21 10:42:53 and /dev/input/* devices Jul 21 10:44:29 Jay7: TSLIB_* env variables is a good tip. we didn't get it to work at all in ts_* because of var problem Jul 21 10:44:39 I pass them in /etc/profile Jul 21 10:45:06 Xfbdev looks for some.. I forgot what exactly Jul 21 10:45:30 and my zaurus is now in emergency-only mode :) Jul 21 10:45:33 Jay7: pls have a look at http://host.xzshare.com/?d=0AD57C191 Jul 21 10:45:45 it's the nand map Jul 21 10:45:47 I have the vars like that in profile "export TSLIB_TSDEVICE=/dev/input/event0" Jul 21 10:45:50 is that correct? Jul 21 10:45:53 ant|work731: I looking into now :) Jul 21 10:46:55 dcordes: your /dev/ts* are created, aren't ? (I got udev issues recently...no ts btw) Jul 21 10:47:25 ant|work731: yep they are. and I can use the touchscreen in ts_test perfectly Jul 21 10:47:26 well Jul 21 10:47:32 except of the huge offsets :D Jul 21 10:48:54 Jay7: sash started creating a new partitioning schema but then reverted back... (U-oot, U-BootEnv..) Jul 21 10:49:26 ant|work731: did you now reason of reverting? Jul 21 10:50:03 exactly no, now all is in mt2 Jul 21 10:50:12 I think we should provide some semi-automated way for non-standard partitioning Jul 21 10:50:20 but how? Jul 21 10:51:25 Jay7: these partitions are hardcoded in the Zaurus bootloader Jul 21 10:52:27 hm.. and how Cacko changes partitions? Jul 21 10:53:00 there was menu 'NAND flash utils' with option 'resize root partition' Jul 21 10:54:29 well, beside nandlogical there is a file fsro_resize.c in survive-1.1.0.tar.bz2 Jul 21 10:56:07 Jay7: are the TSLIB_* vars used by Xfbdev, if I place them in the way like above? Jul 21 10:56:39 Jay7: C:\Documents and Settings\andrea\Desktop\search[1].html++zaurus++mtdparts&hl=it&ct=clnk&cd=17&gl=it.htm Jul 21 10:56:49 oops Jul 21 10:57:11 http://www.oesf.org/forum/index.php?showtopic=3361 Jul 21 10:57:45 dcordes: TSLIB_TSDEVICE is used imho Jul 21 10:58:17 dcordes: try to start TSLIB_TSDEVICE=/dev/.... Xfbdev .. Jul 21 10:58:48 ok Jul 21 10:58:49 ant|work731: fsro_resize looking interesting Jul 21 10:59:39 just 4 writes Jul 21 11:00:22 but I fear is only c7x0 Jul 21 11:03:04 ant|work731: may be just try it? Jul 21 11:03:18 I have default u-boot now on zaurus Jul 21 11:03:43 compiled from you or the one from pdaxrom Jul 21 11:03:47 ? Jul 21 11:03:53 from pdaxrom now Jul 21 11:04:08 so, 'wrong' cmdline Jul 21 11:04:20 ah.. Jul 21 11:04:34 no good :) Jul 21 11:05:52 but being that the sharp bootloader passes bogus value, linux-rp ignores it ;- Jul 21 11:05:55 ) Jul 21 11:06:11 so you can try only patching sharpsl.c and cmdline Jul 21 11:08:18 looks like fsro_resize rewrites somewhat in emergency image.. offset is 0x60000 Jul 21 11:08:54 emergency is not used in normal boot as i know.. Jul 21 11:09:10 is paraminf.bin Jul 21 11:09:39 ah, yes Jul 21 11:09:46 not emergency (zimage2.bin) Jul 21 11:10:23 but if you have u-boot, it doesn't make sense Jul 21 11:11:00 ok, I should go to customer right now Jul 21 11:11:03 because yes, in that case emergency starts there Jul 21 11:11:18 I try to patch u-boot and kernel later Jul 21 11:11:18 cu, thx for following my efforts Jul 21 11:12:15 how much time would you be here? Jul 21 11:12:45 grr.. I need some english learning :) Jul 21 11:13:02 what's your timezone Jul 21 11:13:04 ? Jul 21 11:13:15 MSD (GMT+4) at summer Jul 21 11:13:23 i have now 15:13 Jul 21 11:13:53 now I'm in office for 6 more hours, then break and connect as ant@home (here CEST 13.13) Jul 21 11:15:15 ok, I'll be back in ~3 hours Jul 21 11:15:32 then I want to hack my akita :) Jul 21 11:15:41 *g* Jul 21 11:16:49 akita! best choice. Jul 21 11:19:04 * ant|work731 -> food Jul 21 11:19:08 bbl Jul 21 11:20:08 dcordes: yes, i like tosa too.. but i don't have tosa :) Jul 21 11:20:30 SL-6000W is very interesting and rare device :) Jul 21 11:21:13 I'm away for 3 hours :) Jul 21 11:21:51 bye Jul 21 12:35:44 Jay7: looking again at http://tomoyo.sourceforge.jp/cgi-bin/lxr/source/drivers/mtd/nand/sharpsl.c?a=arm Jul 21 12:36:51 I did I mistake: mtd2 appears to be 58mb and mtd3 63mb on akita (standard) Jul 21 12:37:13 7+58+63=128 Jul 21 12:37:57 for husky/boxer it is 7+53+68 Jul 21 12:40:02 http://www.oesf.org/forum/index.php?showtopic=25055&st=0&p=171224&#entry171224 Jul 21 12:42:13 you'll have to check akita-angstrom.patch... Jul 21 12:42:17 sorry but I don't have akita ... Jul 21 12:42:30 bbl Jul 21 12:55:27 yo :) Jul 21 14:48:10 re Jul 21 14:48:30 Jay7: hrw could help us Jul 21 14:49:16 ant|work234: good :) Jul 21 14:49:51 hm.. I have read #oe talk Jul 21 14:50:02 single kernel - it's idea :) Jul 21 14:50:12 hrw says he won't support u-boot... Jul 21 14:50:24 koen agreed about that single kernel Jul 21 14:50:30 =) Jul 21 14:50:42 sory - phone call Jul 21 14:52:08 i understand that hrw won't support u-boot :) Jul 21 14:52:14 I won't too.. Jul 21 15:13:41 hm.. do we have kernel which does not pass CONFIG_CMDLINE? Jul 21 15:14:00 or it's already fixed? Jul 21 15:14:03 do I need to patch sharpsl.c? Jul 21 15:15:04 svs57 wrotes "You can change partitions size either by recompiling mtd/nand/sharpsl.c module _or_ adding bootarg" Jul 21 15:15:06 but not in 2.6.23 Jul 21 15:23:26 back Jul 21 15:24:16 in 2.6.23 ATAG patch was applied Jul 21 15:24:55 and supposedly they skipped cmdline Jul 21 15:24:59 but now we have 2.6.24 in angstrom 2007.12? Jul 21 15:25:07 exactly, need test! Jul 21 15:25:13 ok Jul 21 15:25:14 we even have 2.6.26 Jul 21 15:25:30 I have patched u-boot's akita.h Jul 21 15:25:35 great Jul 21 15:25:36 one question left Jul 21 15:25:58 #define NAND_LOGICAL_SIZE 0x700000 Jul 21 15:26:01 should I set here 10Mb? Jul 21 15:26:18 yes, you should test Jul 21 15:26:20 #define MTDPARTS_DEFAULT "mtdparts=sharpsl-nand:10240k(Logical),102400k(root),18432k(home)" Jul 21 15:26:48 http://www.hpc.ru/pda/board/index.php?t=121323&start=330 Jul 21 15:26:58 but only two partitions here... Jul 21 15:27:04 hm.. Jul 21 15:27:25 it's older, just to confirm you the 10240 Jul 21 15:27:29 ah.. I've read this but forget already :) Jul 21 15:27:56 typo?: else if (machine_is_corgi() || machine_is_shepherd()) { Jul 21 15:28:03 sharpsl_partition_info.size = 54 * 1024 * 1024; Jul 21 15:28:07 ?? Jul 21 15:28:14 ok, it's old Jul 21 15:29:02 'shepherd - 25Mb' http://www.oesf.org/forum/index.php?s=decb47fe0773c0e0eb7312dba8b3880d&showtopic=25055&view=findpost&p=171233 Jul 21 15:29:09 don't know yet :) Jul 21 15:29:24 you see how messed up it is ! Jul 21 15:29:51 i'm downloading svs57's u-boot :) Jul 21 15:30:03 at 128Kbps it's not so fast :) Jul 21 15:30:10 it's 128k Jul 21 15:30:13 ;-) Jul 21 15:35:16 standard for akita should be +define MTDPARTS_DEFAULT "mtdparts=sharpsl-nand:7168k(Logical),59392k(root),64512k(home)" Jul 21 15:40:36 ok, downloaded Jul 21 15:40:46 let's look in :) Jul 21 15:42:02 Jay7: I'm against having only 2 partitions, but yes, home could be mounted elsewhere Jul 21 15:43:41 ant|work234: I'm too Jul 21 15:44:02 having small /home is better than having no /home at all :) Jul 21 15:44:25 /home on NAND is used only for config's Jul 21 15:44:51 all 'real big data' is stored on SD Jul 21 15:45:17 but is to big as it is Jul 21 15:45:59 10M-12M would be enough Jul 21 15:46:25 I left 18432k Jul 21 15:46:45 btw, '-(home) should work? Jul 21 15:47:08 it works always? or better specify number Jul 21 15:48:04 should work in both cases, the last takes all the rest Jul 21 15:48:12 the last mtd Jul 21 15:48:56 ok Jul 21 15:49:11 http://www.kazanfoto.ru/index.php?pg=52&i=34234 Jul 21 15:49:19 rains in Kazan :) Jul 21 15:49:41 ~350km from me Jul 21 15:49:57 well, that's rain! Jul 21 15:50:13 (or a river ?) Jul 21 15:50:40 no, it was a street :) Jul 21 15:50:57 hm, svs don't change that define Jul 21 15:51:06 #define NAND_LOGICAL_SIZE 0x700000 Jul 21 15:51:21 m Jul 21 15:51:28 #define MTDPARTS_DEFAULT "mtdparts=laze-0:7168k(Logical),-(JFFS2)" Jul 21 15:51:46 but it's her parts.. Jul 21 15:51:54 what is 'laze-0'? Jul 21 15:52:05 is the 'name' Jul 21 15:52:11 and why not 'sharpsl-nand'? Jul 21 15:52:13 ah Jul 21 15:52:21 ok Jul 21 15:52:22 Jay7 would be ok as well Jul 21 15:52:31 hmhm.. Jul 21 15:53:07 but he's not changed define NAND_LOGICAL_SIZE because he's not changed 'logical' size Jul 21 15:53:20 i'll do grep :) Jul 21 15:53:59 oops Jul 21 15:54:11 he is having corgi, not akita :) Jul 21 15:55:13 in corgi.h NAND_LOGICAL_SIZE is not changed too, but logical is set to 10Mb Jul 21 15:57:42 this define is used in common/cmd_nand.c Jul 21 15:58:10 ant|work234: can you look into 'common/cmd_nand.c' for NAND_LOGICAL_SIZE? Jul 21 15:58:43 I think it should be set to real value.. Jul 21 15:59:04 I'm sorry, here at work I don't have OE Jul 21 15:59:19 ok Jul 21 15:59:31 I just set it to 10Mb Jul 21 15:59:43 then build u-boot and flash it up :) Jul 21 16:02:21 do you have a serial cable? Jul 21 16:03:35 not yet :) Jul 21 16:03:49 I'll get one at next week :) Jul 21 16:03:54 ok, you can always boot recovery console Jul 21 16:04:12 I can always do nand restore :) Jul 21 16:04:53 but serial cable is necessary for such things.. Jul 21 16:06:17 I dunno why sash abandoned the old (proper) partitioning http://www.pdaxrom.org/?q=node/89 Jul 21 16:06:25 of mtd1 Jul 21 16:07:26 th epoint is, if we get a new unified bootmenu+emergency kernel we can save place Jul 21 16:08:56 btw, you have mistype in akita-angstrom.patch Jul 21 16:09:15 define MTDPARTS_DEFAULT is without '#' in beginning Jul 21 16:13:35 compiled :) Jul 21 16:18:17 ant|work234: I place autoboot.sh, u-boot.bin, updater.pro, updater.sh on SD Jul 21 16:18:30 can I flash without emerg.img? Jul 21 16:23:12 no, you need nandlogical Jul 21 16:23:25 and at the moment is only in emergency Jul 21 16:23:42 actually you could compile it Jul 21 16:23:52 and update u-boot from shell Jul 21 16:24:00 u-boot.bin Jul 21 16:24:19 hm.. Jul 21 16:24:40 yes, to write on mtd1 you need nandlogical Jul 21 16:25:20 nandlogical /dev/mtd1 WRITE 0x0 0x20000 $file Jul 21 16:25:21 just nandlogical /dev/mtd1 WRITE 0x0 0x20000 u-boot.bin Jul 21 16:25:26 ok Jul 21 16:25:56 lets burn it :) Jul 21 16:26:06 no way :-] Jul 21 16:28:14 hm.. it does not started :)) before flashing :) Jul 21 16:28:20 it fears :) Jul 21 16:31:13 ok, loaded emergency :) Jul 21 16:31:28 now nandlogical'ing :) Jul 21 16:32:56 seems written Jul 21 16:34:59 nothing going on :) Jul 21 16:35:12 ? Jul 21 16:35:30 can't boot it into emergency Jul 21 16:36:30 hm.. serial cable would help, but I have no serial cable :) Jul 21 16:39:18 the emergency image shoul be untouched... Jul 21 16:40:54 testkey 101 -> nand read.logical 0xa0800000 0x00060000 0x00540000 Jul 21 16:42:35 ok, it's booted into emergency after reflashing with pdaxrom files (via C+D) and with new u-boot Jul 21 16:43:26 2 mtd partitions Jul 21 16:43:44 logical & JFFS.. Jul 21 16:43:48 hm.. Jul 21 16:44:13 seems pdax's standard u-boot Jul 21 16:44:33 wait, you are in emergency Jul 21 16:44:52 this is a second kernel Jul 21 16:45:05 2.6.18 with own cmdline Jul 21 16:45:20 ah.. Jul 21 16:45:29 so I should write angstrom kernel Jul 21 16:45:37 yes, and use mkimage Jul 21 16:45:46 and boot as normal Jul 21 16:45:58 yep Jul 21 16:46:06 ok, one moment :) Jul 21 16:46:59 for testing you could grab a 2.6.24 kernel from /unstable Jul 21 16:47:55 but ideally you should test with a kernel without mtdparts... in cmdline Jul 21 16:48:13 and it should work.... Jul 21 16:48:28 (I tested in december-january with older kernels...sorry...) Jul 21 16:49:05 and how I can disable mtdparts in cmdline? Jul 21 16:49:16 self-compile Jul 21 16:49:45 I just want use kernel from angstrom-2007.12 Jul 21 16:49:59 compiled on my desktop Jul 21 16:50:01 see linux-rp.inc Jul 21 16:52:43 and defconfig-akita Jul 21 16:57:11 default kernel is 2.6.24-r9 btw Jul 21 16:58:04 i found 3 entry in defconfig Jul 21 16:58:12 CONFIG_MTD* Jul 21 16:58:26 shoul I unset CONFIG_MTD_CMDLINE_PARTS? Jul 21 16:59:35 yes, that one Jul 21 16:59:52 somewhat else? Jul 21 17:00:02 unset or or you can insert it in cmdline Jul 21 17:01:53 I just comment this out in defconfig-akita Jul 21 17:01:57 have you found mkimage? Jul 21 17:02:10 yes, in gentoo portages :) Jul 21 17:02:26 but it's compiling now Jul 21 17:02:30 there is a mkeimage-native in OE too Jul 21 17:02:34 btw, 1.3.1 is ok? Jul 21 17:02:41 ah.. Jul 21 17:02:53 dunno, used the one in /tmp/u-boot Jul 21 17:03:06 get compiled with u-boot IIRC Jul 21 17:03:17 ok, I'll try 1.3.1 from portages, then -native from OE Jul 21 17:03:28 no, first look in your tmp Jul 21 17:03:36 where u-boot.bin was Jul 21 17:04:01 tools ? extra ? Jul 21 17:04:22 wow Jul 21 17:04:25 tools/mkimage Jul 21 17:04:29 10x :) Jul 21 17:04:53 there was discussion earlier on #oe about mkimage-native he he Jul 21 17:05:04 I think that is compiled for arm but it's for host Jul 21 17:05:31 now I have two mkimage :)) Jul 21 17:05:38 ok, I'll use OE's Jul 21 17:05:42 you lucky guy! Jul 21 17:05:53 btw, initrd should be stripped yet? Jul 21 17:06:02 yes Jul 21 17:06:13 then reflash both Jul 21 17:06:46 ok, going on :) Jul 21 17:16:44 BTW in OE mkimage-native is in u-boot-utils-native Jul 21 17:20:10 no, 10x, I have already two :) Jul 21 17:20:47 you immediatly see if it works from the stdout Jul 21 17:22:08 linux-rp is compiling.. Jul 21 17:22:49 initrd.bin is provided by linux-rp? Jul 21 17:23:01 no, this is rootfs Jul 21 17:23:24 well, the modules somehow Jul 21 17:23:41 ok, then waiting for linux-rp and console-image :) Jul 21 17:24:13 it would be interesting if it works without mtdparts= in cmdline Jul 21 17:24:18 as it should Jul 21 17:25:01 we will know after some time :) Jul 21 17:25:20 going home now, bbl in ~3 hours Jul 21 17:25:34 good work Jul 21 17:25:42 thx again for your time Jul 21 17:25:51 bye Jul 21 18:14:44 Jay7: can you help me again with tslib on Xfbdev? Jul 21 18:15:19 dcordes: I can try :) Jul 21 18:17:00 Jay7: uses Xfbdev the variables in /etc/profile? Jul 21 18:18:04 dcordes: cant say.. Xfbdev takes variables from current env Jul 21 18:18:34 check /etc/X11/Xinit.d/* Jul 21 18:18:46 may be here is right place Jul 21 18:18:51 I don't use Xinit Jul 21 18:18:55 ah.. Jul 21 18:19:02 only start with Xfbdev for test Jul 21 18:19:16 do you have success anyway with TSLIB_TSDEVICE? Jul 21 18:19:27 no. I tried ito Jul 21 18:20:02 maybe the vars are ok, because: In the background of Xfbdev, there is an error when ts event occurs Jul 21 18:20:12 do Xfbdev complains about TSLIB_TSDEVICE on stderr when you start it? Jul 21 18:20:40 not about the variable TSLIB_TSDEVICE Jul 21 18:20:51 about something like? Jul 21 18:20:55 I think it says something like "Unknown event: 0 0 240 320 " Jul 21 18:21:08 hm.. Jul 21 18:21:29 do you have ssh access or multiple consoles on your device? Jul 21 18:21:40 yes but no money for gprs Jul 21 18:22:04 I can switch between vt Jul 21 18:22:17 and device has keyboard Jul 21 18:23:24 try to start Xfbdev in foreground and check stderr for errors Jul 21 18:23:58 Jay7: can I rediret stderr to a file, in console I start it? Jul 21 18:24:05 when I tries to start Xfbdev in qemu-image, it complains about TSLIB_ variable Jul 21 18:24:19 yes, I think Jul 21 18:24:27 how? Jul 21 18:25:19 Xfbdev 2>stderr_file Jul 21 18:25:26 after exit look into stderr_file Jul 21 18:26:00 btw, don't specify -mouse option in command line of Xfbdev Jul 21 18:26:19 I tried -mouse tslib. why do not specify it? Jul 21 18:27:18 when -mouse specified, Xfbdev swithes to mouse, not to tslib Jul 21 18:27:25 afaik Jul 21 18:27:58 when using tslib, there is sould be no -mouse switch Jul 21 18:28:13 in /etc/X11/Xserver there is settings like #blballablablaa if device with tslib\ ARGS= $ARGS '-mouse tslib' Jul 21 18:28:17 or similar Jul 21 18:28:46 hm.. Jul 21 18:31:08 I found this info in OE bugzilla Jul 21 18:33:17 Jay7: which? Jul 21 18:33:27 can't find.. :) Jul 21 18:33:57 ok will look there. http://www.kazanfoto.ru/index.php?pg=52&i=34234 man is that your city? Jul 21 18:34:59 dcordes: no, it city in ~350-400km from me :) Jul 21 18:35:12 s/it/this/ :) Jul 21 18:35:59 lucky you Jul 21 18:36:34 that's climate change. we all need to stop using linux pdas. Jul 21 18:37:32 :)) Jul 21 18:42:47 dcordes: sorry, cant find this bug.. Jul 21 18:43:01 try set up Xinit and check it Jul 21 18:43:18 ok Jul 21 18:44:02 or just check xinit scripts in packages Jul 21 18:44:12 grep -ri xinit packages/* Jul 21 18:48:27 dcordes I already had to stop drinking fizzy drink.. I'm going to use my linux PDA come hell or high water.. Jul 21 19:01:55 can somebody hint me on where to find Xfbdev --help output? (can't scroll it on my device) Jul 21 19:02:54 dcordes: Xfbdev 2>output.txt Jul 21 19:03:09 Xfbdev --help Jul 21 19:03:41 or Xfbdev --help 2>&1 >output.txt Jul 21 19:05:16 thanks Jul 21 19:19:51 Jay7: the error I get is "Invalid update: 0 0 320 240" (many times, only when I touch) Jul 21 19:19:54 and also Jul 21 19:20:07 0 0 320 240 (like 6 times)\ 0 156 320 84\ 0 15 320 90\ 0 144 320 96\ 0 139 320 102\ Jul 21 19:20:35 sorry.. here is Kdrive/Xfbdev guru needed :) Jul 21 19:20:45 check some on #oe :) Jul 21 19:21:18 good idea Jul 21 19:25:17 Hi, I know this is not the best place to ask but I have a problem with Xfbdev, I can't find any help with elsewhere. in the tslib testing apps, my touchscreen works perfectly, but in Xfbdev, I get errors like "Invalid update: 0 0 320 240" "Invalid update 0 156 320 84" "Invalid update: 0 15 320 90" "Invalid update 0 144 320 96", plenty in a sequence, when I touch the screen in Xfbdev. Anybody seen something like that? Jul 21 19:25:33 this was for #OE :S Jul 21 19:25:39 :) Jul 21 20:37:41 ant__: re Jul 21 20:38:10 ant__: I have some success Jul 21 20:39:10 when I left logical partition as is (7Mb) then I can load without mtdparts in kernel cmd_line Jul 21 20:39:36 when I change size of logical then kernel can't mount rootfs Jul 21 20:39:44 can't find Jul 21 20:39:53 hi Jay7 Jul 21 20:40:18 did you hack sharpsl.c ? Jul 21 20:40:25 yes and no :) Jul 21 20:40:39 with 10Mb logical - no matters Jul 21 20:41:02 with standard logical I have now 100Mb root Jul 21 20:41:23 with patched sharpsl and without mtdparts in kernel cmd_line Jul 21 20:41:42 sharpsl.c: thats where the mtdpart string comes from Jul 21 20:41:49 I want try to test unpatched sharpsl.c Jul 21 20:42:12 this has standard partition layout Jul 21 20:42:23 so mtd1 = 7mb Jul 21 20:43:06 but with 10Mb logical I tries patched and unpatched sharpsl.c with and without mtdparts in cmd_line Jul 21 20:43:46 and u_boot ? Jul 21 20:44:22 now I return u-boot to 7Mb logical Jul 21 20:44:28 and set 100Mb root Jul 21 20:44:36 this works Jul 21 20:44:40 yes Jul 21 20:44:44 I tested it (on c7x0) Jul 21 20:44:50 I even tested 10mb Jul 21 20:44:59 hm.. Jul 21 20:45:13 then it was abandoned... Jul 21 20:45:18 kexec hype Jul 21 20:45:59 I'll try once more at morning :) Jul 21 20:46:16 unfortunately pfalcon (author of most of initramfs image) and thesing (the klibc guy) somehow abandoned the kexec project too Jul 21 20:46:28 hehe.. Jul 21 20:46:31 Jay7: the idea is: Jul 21 20:46:35 bad luck.. Jul 21 20:46:41 1) u_boot (install once and forget) Jul 21 20:46:59 2) unified kernel for flashing, emergency and bootmenu Jul 21 20:47:11 (all above fitting in mtd1) Jul 21 20:47:32 3) the above kernel kexec another kernel in /boot Jul 21 20:47:44 this is the actual kernel you use and replace Jul 21 20:47:55 and could resiode on mtd2 Jul 21 20:48:08 or could even reside on mtd3 Jul 21 20:48:23 or if we hack more, on mtdX Jul 21 20:49:07 u_boot would not need a console Jul 21 20:49:29 standard boot being the bootmenu-image, from which you chose which media to boot Jul 21 20:49:39 here is needed special crafted initramfs Jul 21 20:49:40 (defauklting to /boot) Jul 21 20:49:48 we have it Jul 21 20:49:51 alredy Jul 21 20:50:02 working with uclibc (small size) Jul 21 20:50:13 semi-working with klibc (minimal) Jul 21 20:50:29 Jay7: botwer thesing for the latter ;-) Jul 21 20:50:39 *bother Jul 21 20:50:42 :) Jul 21 20:51:23 but how we should operate with unsupported u-boot? Jul 21 20:51:29 it could nicely feet mtd1 (7mb or even 10) Jul 21 20:51:33 stay with 1.1.4? Jul 21 20:51:59 here you need the russian-connection: bother sashz! Jul 21 20:52:19 or find some u-boot guru to reach mainstream Jul 21 20:52:40 (we have already some on #oe) Jul 21 20:53:25 grr.. konversation has krashed :) Jul 21 20:53:31 np Jul 21 20:53:34 here you need the russian-connection: bother sashz! Jul 21 20:53:39 (we have already some on #oe) Jul 21 20:53:57 ok, I'll try to ask sashz Jul 21 20:54:03 uh, sorry, Jul 21 20:54:06 or find some u-boot guru to reach mainstream Jul 21 20:54:09 (we have already some on #oe) Jul 21 20:54:15 ;-) Jul 21 20:54:53 Jay7: as you read on hrw blog, it would be easy for an u-boot guy Jul 21 20:55:15 we don't even need console and environment Jul 21 20:56:05 at morning I have test once more configuration with 10Mb, then make some "freeze".. Jul 21 20:56:52 because I need working zaurus at 24 Jul 21 20:57:12 24.07.08 :) Jul 21 20:57:37 I go to Moscow at this date Jul 21 20:57:56 ok Jul 21 20:58:07 and after 03.08 I can test some things again Jul 21 20:58:21 Jay7: please tell me one more thing: which version of udev are you using? Jul 21 20:58:35 I'm actually debugging it on c7x0 Jul 21 20:58:43 ant__: on Z? from .stable branch Jul 21 20:58:43 (raCE CONDITIONS ?) Jul 21 20:58:56 latest stable is .092 Jul 21 20:59:00 and is ok Jul 21 20:59:22 the newer init (/etc/init.d/udev) is the prob Jul 21 20:59:30 oebug 4118 Jul 21 20:59:42 yes, I tries to build console-image from .dev Jul 21 20:59:47 let me solve this and I come back to u-boot Jul 21 21:00:04 and have random bugs Jul 21 21:00:12 e.g. getty does not start after booting Jul 21 21:00:29 please try udev_124 Jul 21 21:00:38 and chech the a.m. bug Jul 21 21:01:03 (I had missing /dev, thus no tty*) Jul 21 21:01:34 actually whole /dev/shm was not mounted :-( Jul 21 21:01:37 yes, I have same issue as in 4118 Jul 21 21:01:47 comment udevsettle Jul 21 21:02:03 I can't check right now Jul 21 21:02:08 the zaurus is *slow* enough for udev to finish the tasks ;-) Jul 21 21:02:13 when I build image from .dev - it seems like that Jul 21 21:03:34 I use .dev (stable is outdated) Jul 21 21:03:36 btw, I have same issue when started image for spitz in qemu-system-arm Jul 21 21:03:44 he he Jul 21 21:03:57 was reported only on tosa and c7x0... Jul 21 21:04:09 I use self created branch based on stable with some new packages from .dev :) Jul 21 21:04:18 ok, fine Jul 21 21:04:28 but I needed navit and thus... Jul 21 21:04:40 I can show you example how to start it up in qemu :) Jul 21 21:05:06 Jay7: I have the real device, no time to fiddle with qemu ;-) Jul 21 21:05:15 ok :) Jul 21 21:05:24 actually I have two SL-C860 Jul 21 21:05:33 one for me and one for my wife Jul 21 21:05:40 good machines :) Jul 21 21:05:41 the latter will run cacko Jul 21 21:05:55 until I get some mahjong in OE:-) Jul 21 21:06:16 killer app, you know... Jul 21 21:06:23 I've used cacko some time ago, but it is very outdated.. Jul 21 21:06:33 yee! majong :) Jul 21 21:06:41 mahjongg ;-] Jul 21 21:06:58 first image from angstrom I have used was opie-image :) Jul 21 21:07:04 he uses zaurus 99% for it Jul 21 21:07:18 *she Jul 21 21:07:25 very usable but don't have software :) Jul 21 21:07:37 too old Jul 21 21:07:46 no browser, no good mailer, no office apps Jul 21 21:07:51 interesting was the tslib pressure patch Jul 21 21:07:56 but PIM is very good Jul 21 21:08:19 so I switched to x11-image ;) Jul 21 21:08:36 what was it...petite-peinture using stylus-pressure? Jul 21 21:08:38 and now to console image with self-build env Jul 21 21:08:53 dunno :) Jul 21 21:09:06 don't remember exactly too... Jul 21 21:09:40 ok.. I want to sleep now :) Jul 21 21:09:47 good night Jul 21 21:09:49 so.. bis morgen :) Jul 21 21:09:57 tschuess Jul 21 21:10:16 yo :) **** BEGIN LOGGING AT Mon Jul 21 22:24:26 2008 **** ENDING LOGGING AT Tue Jul 22 02:59:57 2008