01:06.44 | *** join/#htc-linux asylumed_ (n=insanity@196.211.28.90) |
02:03.10 | *** join/#htc-linux meshuga (i=meshuga@c-71-231-139-142.hsd1.or.comcast.net) |
02:08.18 | *** join/#htc-linux meshuga_ (i=meshuga@c-71-231-141-145.hsd1.or.comcast.net) |
05:23.36 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
07:08.18 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-151-198.pools.arcor-ip.net) |
08:54.55 | *** join/#htc-linux rob_w (n=bob@p213.54.219.224.tisdip.tiscali.de) |
09:06.22 | *** join/#htc-linux Nalim (n=Miranda@213.151.92.182) |
10:01.02 | rob_w | hey all - anyone workiong on htc stuff ? |
10:02.15 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
10:30.42 | *** join/#htc-linux skodde (n=skodde@213-140-16-178.fastres.net) |
11:11.08 | *** join/#htc-linux rob_w (n=bob@p213.54.221.211.tisdip.tiscali.de) |
13:00.41 | *** join/#htc-linux rob_w (n=bob@p213.54.218.173.tisdip.tiscali.de) |
13:49.06 | *** join/#htc-linux rob_w (n=bob@p213.54.218.173.tisdip.tiscali.de) |
14:18.34 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-153-015.pools.arcor-ip.net) |
15:04.53 | *** join/#htc-linux alge (n=alge@chello080109231226.4.uni-klu.teleweb.at) |
15:05.11 | alge | hi |
15:08.10 | awelux | hi alge |
15:08.33 | alge | is there a way, when using openembedded, to get the needed kernel-module-* packages (dependencies of other kernel-modules) added automatically into to rootf image? |
15:11.33 | alge | I am currently using someting like "tmp/cross/bin/arm-linux-modinfo tmp/rootfs/lib/modules/2.6.16-hh5/kernel/drivers/*" to get the dependencies done step by step ... |
15:54.52 | alge | another question: i'm trying to boot the hh5 (2.6.16) kernel with rob_w's patches and a familiar-unstable gpe-image. it boots, I can ping it via usb0 but no ssh login, no Xfbdev starting ... I modified the 2.6.12 initrd a little bit, changed lib/modules to .16, removed its Xfbdev and dropbear- was that wrong? but shouldn't Xfbdev and dropbear come from the filesystem on the sd card part2? |
15:57.44 | alge | another problem is that the kernel warns about missing its initial console, so I cant see meesages from the initrd ... |
16:28.10 | alge | is it possible to generate the initrd as some openembedded *-image? has someone a *.bb file to do this? |
16:38.45 | *** join/#htc-linux Nalim (n=Miranda@213.151.92.182) |
16:44.38 | awelux | alge, You could ask the openembedded question in the openembedded channel #oe |
16:45.54 | awelux | About dropbear not starting. I've seen scripts where dopbear was started from the initrd image and only later the real image from sd card was mounted as root. |
16:46.20 | awelux | If you remove dropbear in this case it won't get started automatically from sd card later. |
16:48.07 | awelux | I'm not sure what to make from the initial console warning though. Do you have the option console on framebuffer activated? |
16:48.37 | alge | I think so |
16:49.14 | alge | I see kernel messages on the console, but i added some echo to the linuxrc in the intrd and this daos not appear |
16:51.21 | alge | console on fb is set: CONFIG_FRAMEBUFFER_CONSOLE=y |
16:52.23 | alge | Can it be that the binaries in the initrd (which come from the 2.6.12 initrd) are to old/incompatible with my familiar-unstable rootfs? |
16:55.44 | awelux | hmm. let me think. I had this problem too. |
16:56.51 | awelux | I think I had a buggy haret when I had the problem and wasn't able to pass kernel parameters. |
16:57.03 | awelux | In the kernel parameters I specify the console. |
16:57.55 | awelux | This migh also explain why you have problems with the starup scripts. |
16:58.16 | rob_w | hey dudes |
16:58.29 | awelux | I've root=/dev/mmc/blk0/part2 init=/linurc console=tty0 in my kernel parameters |
16:58.33 | awelux | nhi rob_w |
16:59.38 | alge | I 'll try two things: use the old unmodified initrd (but with 2.6.16 modules) and build a new one as oe image in initramfs format (can be doen without loop mounting and so without root access, i think) |
17:00.43 | rob_w | i also starting a new rootfs as mine is having alignment traps all over |
17:01.09 | alge | ah, I have root=/dev/ram0 ... I tried /dev/mmcblk0p2 but this didnt work, will try /dev/mmc/blk0/part2 first |
17:03.11 | awelux | <-reboot |
17:05.14 | rob_w | alge still will need /dev/mmcblk0p2 |
17:05.43 | rob_w | brb |
17:08.17 | *** join/#htc-linux rob_w (n=bob@p213.54.218.173.tisdip.tiscali.de) |
17:09.44 | alge | but then cannot open root device .,.. |
17:10.04 | rob_w | ? but you where booting 2.6.16 ? |
17:10.12 | alge | yes |
17:10.20 | rob_w | are u really sure ? |
17:10.42 | alge | there i only a zimage-2.6.16-hh5 on my sd card :-) |
17:11.14 | alge | the kernel sees the two partitions on mmc0: |
17:11.19 | alge | mmcblk0: p1 p2 |
17:11.57 | alge | but then ist says cannot open root device, maybe the /dev/ entriy is missing? |
17:11.59 | rob_w | wonders me .. probably ur rootfs is old as it then doesnt use udev .. as udev has it on /dev/mmcblk0pX |
17:12.55 | alge | what do you men with rootfs, the initrd or the part2? |
17:13.11 | rob_w | ah |
17:13.24 | rob_w | u are using initrd |
17:13.48 | alge | yes, i enabled it in the htcbluangel_defconfig |
17:13.58 | rob_w | without initrd u will be able to boot on mmcblk0 |
17:14.16 | alge | ok , i ll try |
17:14.56 | alge | will it work with a rootfs generated with recent openembedded as gpe-image? |
17:21.48 | rob_w | yes |
17:22.11 | rob_w | depending on your definition of "recent" |
17:37.00 | alge | yesterday :-) |
17:39.40 | rob_w | well ... that might work :-O |
17:47.49 | rob_w | alge, can you boot now ? |
17:50.42 | alge | still connat open root device mmcblk0p2 |
17:50.51 | alge | s/connat/cannot/ |
17:51.25 | rob_w | u use my defconfig ? |
17:51.34 | alge | I'll check what the differences between my .config and yours are |
17:52.10 | alge | I took yours but did some modifications, just for "training purposes" :-) |
17:53.24 | rob_w | lemme update it and show u the one i currently use |
17:54.45 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-097-037.pools.arcor-ip.net) |
17:55.15 | rob_w | updated now |
17:57.01 | alge | ok, now there are more differences, I 'll recompile with yours, thanks |
18:03.38 | alge | I guess you also have to update your patch, cause you are now using CONFIG_MACH_BLUEANGEL_LEDS=m but with current htcblueangel.diff (07-19) compilation fails at "arch/arm/mach-pxa/htcblueangel/htcblueangel_leds.c:234: error: unknown field ..." |
18:37.20 | *** join/#htc-linux rob_w (n=bob@p213.54.218.173.tisdip.tiscali.de) |
18:40.19 | alge | hi rob_w, could you please check if your new defconfig file also needs newer patches? |
18:40.58 | rob_w | alge, is it not working ? |
18:41.12 | alge | not compiling |
18:41.16 | rob_w | what ? |
18:41.23 | rob_w | where does it break |
18:41.35 | alge | htcblueangel_leds |
18:41.53 | alge | guess your patch is now older than the config file |
18:42.13 | rob_w | what does htcblueangel_leds say ? |
18:42.40 | alge | arch/arm/mach-pxa/htcblueangel/htcblueangel_leds.c:234: error: unknown field `hw_duty_get' specified in initializer |
18:42.46 | rob_w | but u finished a zImage , right |
18:42.46 | alge | ... and so on |
18:42.56 | alge | no |
18:43.28 | rob_w | did u implement htcblueangel_leds into the kernel ? |
18:43.50 | rob_w | make it a module .. check if led support in general is also selected |
18:44.27 | alge | CONFIG_CLASS_LEDS=y ist set |
18:44.41 | alge | and CONFIG_MACH_BLUEANGEL_LEDS=m |
18:44.57 | alge | but that comes from your config file |
18:45.08 | alge | if i di not mix up things |
18:45.18 | alge | s/di/did/ |
18:46.00 | rob_w | a bit strange .. as u say BA_LEDS is a module .. you should be able to compile the kernel itself |
18:46.40 | rob_w | ah now i know |
18:47.16 | rob_w | deselct htcbluangel_leds for now .,.i am able to compile it .. so i would need to update some patches .. |
18:47.54 | alge | ok, i understand now, of course zImage has been built, but oe stops working, so I have to disable BA_LEDS. |
18:48.45 | alge | I just thought you had been working on the led stuff because it was in your config file. |
18:49.12 | rob_w | yeah . iwas |
18:49.27 | rob_w | you will need to change include/linux/leds |
18:49.44 | alge | how |
18:50.16 | rob_w | get the .12 tree and compare that leds.h to the .16 include/linux/leds.h |
18:50.27 | rob_w | we altered the led interface a bit |
18:51.34 | alge | I recognized already that hw_duty_get is now blink_frequency_get, but there was also another function name that changed ... |
18:53.52 | rob_w | yeah get the xanadux .12 and compare |
18:54.15 | rob_w | my wife needs cookies .. i brb from hte gasstation ;-) |
18:54.22 | alge | i'm currently diffing it |
19:10.58 | *** join/#htc-linux ChanServ (ChanServ@services.) |
19:10.58 | *** mode/#htc-linux [+o ChanServ] by irc.freenode.net |
20:09.31 | *** join/#htc-linux Nalim (n=Miranda@213.151.92.182) |
20:28.00 | rob_w | any news alge |
20:30.18 | alge | i disabled ba_LEDS, it compiles, but I still get no boot device ... I've to double check if I copied the right files to sd ... |
20:31.37 | alge | can pleas show your kernel commandline? maybe I'm stillmissing something |
20:32.32 | alge | I have: set CMDLINE "root=/dev/mmcblk0p2 console=tty0 " |
20:33.13 | alge | and the curious thong is, that the partitions p1 and p2 are reported from the kernel |
20:33.22 | alge | s/thong/thing/ |
20:33.35 | rob_w | set CMDLINE "root=/dev/mmcblk0p2 console=tty0" |
20:34.07 | alge | strange |
20:34.13 | rob_w | indeed |
20:34.52 | alge | it cannot be related to the rootfs we are using at all, because at this point it doesnt natter |
20:35.00 | rob_w | i even boot opie on /dev/mmcblk0p3 |
20:35.34 | rob_w | try syncing your fat sd partition once more |
20:35.43 | rob_w | maybe fat plays tricks on you |
20:35.58 | alge | maybe, I'll try now |
20:39.21 | alge | still the same |
20:40.08 | rob_w | what mkfs.xx did you use to make the rootfs partition ? |
20:40.21 | alge | ext2 |
20:41.46 | rob_w | really strange |
20:41.52 | alge | partition type is 83 |
20:42.10 | rob_w | thats cool |
20:42.35 | alge | ...? |
20:43.50 | rob_w | it is either something really stupid .. or someone plays a trick on u |
20:43.56 | alge | your patch is against that handhelds.org cvs kernel? maybe we differ here? |
20:44.35 | alge | but it applies whithout problems, so what |
20:45.02 | rob_w | right .. and i didnt touch mmc or asic support in my patches |
20:45.57 | alge | I got my kernel with something like |
20:46.00 | alge | cvs://anoncvs@anoncvs.handhelds.org:/cvs/linux-2.6.16-hh2;method=pserver;module= linux/kernel26 |
20:46.02 | rob_w | hmm i am unsure how to proceed |
20:46.51 | rob_w | i was using hh4 |
20:47.16 | alge | despite using hh2 I got hh5 |
20:47.16 | rob_w | cat CVS/Tag |
20:47.17 | rob_w | NK2-6-16-hh4 |
20:47.53 | rob_w | well i wasnt compiling it inside oe .. |
20:48.54 | alge | I have no Tag files in my CVS dirs, maybe I checkout a new tree with an explicit tag |
20:50.10 | rob_w | ?? Houston , we didnt get hte last message .. |
20:51.10 | rob_w | my glibc generates locals .. i be back in 15 |
20:51.21 | alge | ok |
20:56.52 | alge | now I get it once more with |
20:56.56 | alge | cvs -d :pserver:anoncvs@anoncvs.handhelds.org:/cvs co -r K2-6-16-hh5 linux/kernel26 |
20:57.52 | alge | and I'll try -r K2-6-16-hh4 too |
21:00.09 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-133-190.pools.arcor-ip.net) |
21:12.00 | alge | BTW, I also apply tsc2200.h.diff and tsc2200soc.diff , is that ok? |
21:15.14 | *** join/#htc-linux bosjo (n=bosjo@130.80-203-23.nextgentel.com) |
21:48.56 | alge | things get stranger: with K2-6-16-hh4 I get rejects from the htcblueangel.diff: |
21:49.10 | alge | | can't find file to patch at input line 8162 |
21:49.10 | alge | | Perhaps you used the wrong -p or --strip option? |
21:49.10 | alge | | The text leading up to this was: |
21:49.10 | alge | | -------------------------- |
21:49.10 | alge | | |--- kernel26/arch/arm/mach-pxa/htcblueangel/Kconfig 2006-06-16 01:01:57.000000000 +0200 |
21:49.10 | alge | | |+++ kernel26/arch/arm/mach-pxa/htcblueangel/Kconfig 2006-07-19 19:09:55.000000000 +0200 |
21:49.12 | alge | | -------------------------- |
21:49.14 | alge | | No file to patch. Skipping patch. |
21:49.16 | alge | | 1 out of 1 hunk ignored |
21:49.18 | alge | | can't find file to patch at input line 8176 |
21:49.20 | alge | | Perhaps you used the wrong -p or --strip option? |
21:49.22 | alge | | The text leading up to this was: |
21:49.24 | alge | | -------------------------- |
21:49.26 | alge | | |--- kernel26/arch/arm/mach-pxa/htcblueangel/Makefile 2006-06-16 00:57:45.000000000 +0200 |
21:49.28 | alge | | |+++ kernel26/arch/arm/mach-pxa/htcblueangel/Makefile 2006-07-18 12:44:26.000000000 +0200 |
21:49.30 | alge | | -------------------------- |
21:49.32 | alge | | No file to patch. Skipping patch. |
21:49.34 | alge | | 1 out of 1 hunk ignored |
21:50.27 | alge | but the large parts of the patch apply correctly |
21:52.58 | alge | same for TK2-6-16-hh5 |
21:53.43 | alge | I think I'll stop now, maybe tomorrow is a better day to try again ... :-) |
21:55.28 | alge | ok, I mistypep thw WORKDIR in my bb file, of course the patch can only fail then ... |
22:28.16 | alge | now with a TK2-6-16-hh4 kernel + patches: I dont see any console output, the haret boot logo sits there, nothing happens |