**** BEGIN LOGGING AT Tue Jul 22 02:59:57 2008 Jul 22 06:54:39 morning Jul 22 06:56:20 methril: hi Jul 22 07:40:34 morning Jul 22 07:53:00 morning, XorA! Jul 22 08:25:01 bonjour Jul 22 08:40:07 * * OE Bug 4449 has been created by ospite(AT)studenti.unina.it Jul 22 08:40:09 * * gpe-image fails configuring libsvg-cairo_20050601.bb Jul 22 08:40:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4449 Jul 22 08:52:44 g'mornin' Jul 22 08:53:20 03koen 07org.oe.dev * r25f2070b... 10/ (13 files in 3 dirs): linux omap2 git: update TWL4030 patches for beagle, add mrus clock patches Jul 22 09:14:45 ant|work: ping Jul 22 09:16:51 Jay7: pong Jul 22 09:16:56 morning Jul 22 09:17:10 hi hrw Jul 22 09:20:02 03koen 07org.oe.dev * r719f01c7... 10/ (6 files in 3 dirs): mythtv: start beating it into a state where it will crosscompile Jul 22 09:32:21 03koen 07org.oe.dev * r18e6c1bc... 10/ (1 packages/dsplink/dsplink_1.51.00.08.bb): dsplink 1.51: adjust default preference Jul 22 09:47:51 RP: ping Jul 22 09:56:12 mwester: ping Jul 22 09:56:26 * ant|work needs badly an ATAG master Jul 22 10:11:44 ant|work: what's your issue? Jul 22 10:13:25 hi pb Jul 22 10:13:42 the issue is the cmdline_mtdparts seems ignored Jul 22 10:14:10 (the one passed by u-boot) Jul 22 10:14:57 Which mechanism uboot passing the commandline via? Jul 22 10:14:59 and I see this comments http://www.rpsys.net/openzaurus/patches/export_atags-r2.patch Jul 22 10:15:22 hi RP, we think the kernel strips the mtdpart Jul 22 10:15:34 but dunno where to start looking for Jul 22 10:15:48 we = me and Jay7 Jul 22 10:16:21 The kernel doesn't strip specific bits Jul 22 10:16:22 ant|work: do you see the right stuff in /proc/cmdline? if so, it seems unlikely that your problem is with atags. Jul 22 10:16:57 I remember about bogus params skipped Jul 22 10:17:32 aerlier in Jan/Feb when kexec got momentum Jul 22 10:19:10 ant|work: as pb__ said, whats in /proc/cmdline ? Jul 22 10:19:31 sorry cannot check here now Jul 22 10:19:36 (no hw) Jul 22 10:19:56 was trying to help Jay7 on #angstrom since yesterday Jul 22 10:20:08 he is resizing mtd1 Jul 22 10:20:13 and cannot boot Jul 22 10:22:04 he's using uboot? Jul 22 10:22:11 RP: yes Jul 22 10:22:19 btw utx fixed nandlogical Jul 22 10:22:32 and now compiles for 2.6.2x Jul 22 10:23:02 there is a .bb attached (oebug 3738) Jul 22 10:23:56 RP: we are getting old waiting initramfs-bootmenu image .... Jul 22 10:24:35 thesing did great work but it is 98% Jul 22 10:24:41 ant|work: It sounds like some issue with your uboot and the kernel, perhaps disagreeing about how they pass the options Jul 22 10:25:30 Ignoring unrecognised tag 0x00000000 ? Jul 22 10:25:31 ant|work: I don't know what to advise though, apart from priting the commandline uboot wants to pass and then looking at the one the kernel gets Jul 22 10:26:03 uboot is corrupting the atags? You can see where that error comes from in the kernel source Jul 22 10:26:43 IIRC zaurus is treated differently as other machines (sharpsl.c) Jul 22 10:27:49 ant|work: You mean sharpsl_param.c ? Jul 22 10:28:23 don't have tree here...one sec Jul 22 10:29:26 yes, this one of course Jul 22 10:31:06 That doesn't have any affect on atags or the commandline Jul 22 10:31:29 There is no special case handling for the zaurus Jul 22 10:31:35 ok, it is done by cmdlinepart.c Jul 22 10:32:24 yes, as with any kernel Jul 22 10:32:45 so, in theory one should just adjust sharpsl_param.c Jul 22 10:32:55 why? Jul 22 10:33:02 and does not need a cmdline_mtdparts Jul 22 10:33:03 sharpsl_param has nothing to do with this Jul 22 10:33:18 RP: for resizing mtd1 Jul 22 10:33:45 sharpsl_param has nothing to do with the mtd sizing Jul 22 10:34:08 if you want to hack it, edit drivers/mtd/nand/sharpsl.c Jul 22 10:34:29 oops.. I meant this one, sorry Jul 22 10:34:47 as firstly stated Jul 22 10:34:47 That tries to parse any commandline and then falls back to some defaults Jul 22 10:35:26 ant|work: You then told me it was something else and lead me down the wrong path Jul 22 10:35:47 RP: there are two issues and I got confused Jul 22 10:36:05 first: resize mtd1 Jul 22 10:36:17 second: boot from there Jul 22 10:36:20 ;-) Jul 22 10:36:41 from there = after this change = from that point Jul 22 10:37:03 change the commandline to change the root device then Jul 22 10:37:50 it looks like with old linux-rp.2.6.23 (before atag's) it was possible Jul 22 10:37:57 not now Jul 22 10:38:29 nothing has changes that would affect that Jul 22 10:39:09 ok, I'll retest Jul 22 10:39:18 Jay7 is testing now on akita Jul 22 10:39:26 with 2.6.24 Jul 22 10:41:46 actually the situation is: Jul 22 10:41:47 ant|work321: kernel can't find init when logical size is increased Jul 22 10:42:08 now compiled with mtdparts in cmdline and with patched sharpsl Jul 22 10:43:59 03koen 07org.oe.dev * r87d8e0fc... 10/ (4 files in 3 dirs): mythtv: the configure does link tests first (which succeed) and then goes to check if the lib exists in LDPATH. This is bogus, so rig the path check. Jul 22 10:44:02 ant|work: There are too many variables - did the kernel boot from the right device? did it correctly see the new partition layout/size? Was the image correctly flashed to the flash partition? Jul 22 10:45:09 he could boot using standard partition size Jul 22 10:45:25 but yes, too many variables Jul 22 10:46:24 so you're saying he's trying to increase the size of the rootfs mtd partition, nothing else? Jul 22 10:46:50 as first step, yes Jul 22 10:46:58 if adding the command line option to set mtd size causes the kernel to not be able to find init, that would suggest that (a) the kernel is indeed interpreting the mtd commandline options, since otherwise adding the option would have no effect at all; Jul 22 10:47:07 (b) that being the case, atags are probably irrelevant Jul 22 10:47:28 so atags are only relevant using kexec ? Jul 22 10:47:32 atm? Jul 22 10:48:47 (c) most likely, the "extra" space that he is trying to expand the rootfs into contains garbage which is upsetting jffs2. Jul 22 10:48:47 pb__: unfortunately the guy does not have a serial cable... Jul 22 10:49:21 who is "the guy" I think there are still some OE serial cables about Jul 22 10:50:40 there are some serial cables left for Europe devs Jul 22 10:50:40 pb__: hmmm Jul 22 10:50:44 ant|work: I don't know what the situation is with atags and the zaurus bootloader specifically. But the point is that the only purpose atags serve is to get information into the kernel. If changing the commandline has any effect on the kernel's behaviour, even if it is a bad effect, then the kernel must be receiving the new commandline and hence the communication mechanism (atags or not) is clearly working correctly. Jul 22 10:51:33 I was misinterpreting the comment in export_atags-r2.patch Jul 22 10:51:34 so, in that case, it is almost certainly futile to obsess over the details of the parameter passing mechanism. I think you should be focussing your attention elsewhere. Jul 22 10:51:52 pb__: Presumably at the same time the kernel is changed, they're reflashing the rootfs, probably using a different method to normal to account for the larger size so its not as clear cut as that Jul 22 10:52:10 if "the guy" doesn't have a serial cable then I guess he should either get one, or redirect console output to the screen. Jul 22 10:52:33 RP: oh, right. Jul 22 10:52:49 Jay7 is the name Jul 22 10:52:59 well, in any case, this all seems like speculation. I guess what we need is hard facts :-) Jul 22 10:53:09 http://bugs.openembedded.net/attachment.cgi?id=5190 Jul 22 10:53:25 add some sleeps into printk so they can see whats going on. The kernel does print the commandline early in the boot process so that should help start to narrow the problem down Jul 22 10:53:41 but yes, more hard facts would be good Jul 22 10:54:58 pb__: flash_eraseall /dev/mtd2...mmm, me could use nandlogical for mtd1 Jul 22 10:55:12 the idea of garbage could be the point... Jul 22 10:55:23 You don't want to use nandlogical Jul 22 10:55:30 not for the rootfs Jul 22 10:56:10 only the 2.4 kernels can read the nandlogical junk Jul 22 11:07:40 RP: so this is not needed for other models ? http://bugs.openembedded.org/attachment.cgi?id=2204 Jul 22 11:07:58 I see it upstream for poodle :-) Jul 22 11:09:42 It went upstream? Jul 22 11:09:50 This is just totally wrong :( Jul 22 11:09:54 I have it on my Gentoo... Jul 22 11:10:25 it was oebug 2969 Jul 22 11:11:05 RP: forget , it is not my day...ù Jul 22 11:11:27 (need more sleep overnight =) Jul 22 11:12:10 ant|work: Its not in mainline, thankfully Jul 22 11:12:32 8-} Jul 22 11:13:27 hmm. what is a simple serial port terminal program that is in OE? Jul 22 11:13:41 picocom? Jul 22 11:15:18 czr_: microcom is part of busybox Jul 22 11:15:55 I think cu is also packaged somewhere, maybe as part of uucp Jul 22 11:16:12 and of course if all else fails you can always use cat :-) Jul 22 11:16:31 oh, plus minicom itself which has its own bb file Jul 22 11:19:55 thanks, trying picocom Jul 22 11:28:53 pb__: you miss screen Jul 22 11:40:37 zecke: I do Jul 22 11:47:35 03koen 07org.oe.dev * r50d1126a... 10/ (1 packages/mythtv/mythtv/configure.patch): mythtv: don't use 'uname' when cross compiling Jul 22 11:47:41 03koen 07org.oe.dev * r6f809a4f... 10/ (1 packages/mythtv/mythtv_svn.bb): mythtv: pass --arch as well as --cpu Jul 22 11:47:45 03koen 07org.oe.dev * r1049eaea... 10/ (1 conf/machine/neuros-osd2.conf): neuros-osd2: recommend dsplink modules Jul 22 11:47:51 03koen 07org.oe.dev * r45869292... 10/ (1 packages/tasks/task-base.bb packages/tasks/task-boot.bb): task-boot,base: bump PR Jul 22 11:51:10 03koen 07org.oe.dev * r05c935a7... 10/ (1 packages/linux/linux-neuros_git.bb): linux-neuros: bump SRCREV Jul 22 12:02:35 my tmp/work took 52GB.. Jul 22 12:03:08 with rm_work ? Jul 22 12:10:58 no Jul 22 12:11:58 I got only 32 Gb at max (2 machines) ... then I inherited rm_work ;-) Jul 22 12:48:17 03mickeyl 07org.oe.dev * rb663738c... 10/ (11 files in 4 dirs): linux-openmoko[-devel]: add squashfs to both builds, refactor more common stuff into the .inc file Jul 22 12:48:22 03mickeyl 07org.oe.dev * r66afa730... 10/ (1 packages/linux/linux-openmoko.inc): linux-openmoko.inc: add audio modules for both kernels, lets ses whether the right thing happens @ runtime Jul 22 12:58:23 03koen 07org.oe.dev * r21552d6d... 10/ (1 packages/vlc/vlc-davinci_0.8.6h.bb): vlc davinci: more configure params Jul 22 13:15:02 anyone using initramfs-module-bootmenu ? Jul 22 13:15:19 http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob;f=packages/initrdscripts/files/30-bootmenu.sh;h=c77bd662afa1145bbd2023477816d59fe19f3ffa;hb=d1e90988539db9107d344171d67b7eeec111856e Jul 22 13:15:35 yhis script fails because 'exp' is not in klibc Jul 22 13:15:59 other distros are enabling awk Jul 22 13:17:11 *expr of course Jul 22 13:23:39 * Jay7 is here Jul 22 13:23:49 uff.. too long day.. Jul 22 13:29:01 XorA: 10x, but I'll get one serial cable at next month in Moscow :) Jul 22 13:29:35 Jay7: ok. Jul 22 13:29:55 Jay7: we got those cables from amazon when they were 2usd each Jul 22 13:32:07 hrw: maslovsky got this from amazon too :) Jul 22 13:32:34 Jay7: if you meet sash/maslowsky in Moscow, give them free drink (we pay) Jul 22 13:32:43 (but make them finish u-boot !) Jul 22 13:33:12 ant|work: maslovsky should be on out meeting :) Jul 22 13:33:38 *our Jul 22 13:33:40 I like this russian-connection he he Jul 22 13:34:09 ah, tell them about OE ! Jul 22 13:34:15 :) Jul 22 13:34:17 insist !!! Jul 22 13:35:02 btw, thanks for all the terminal emu suggestions. found out that I didn't enable the relevant GPIOs in order to get the serial signals out of the board, hence the problems of getting anything done with serial test programs ;-) Jul 22 13:39:58 hrw: did you have tosa yet? Jul 22 13:40:28 Jay7: I still have to send it to Dmitry Jul 22 13:40:58 and it looks like it have to wait till September now as I have some parts here, some in other city Jul 22 13:41:02 I want to package an exe for system 2 in an ipk for system 1. The reason for this is sytem 1 downloads the exe to system 2. The problem I'm having is I get "Architecture did not match" from OE when building the ipk. Is there a way to bypass the Achitecture QA test for a specific file? Jul 22 13:41:20 hrw: ok :) Jul 22 13:41:41 system 2 is a very simple system that can't run a OS so having a ipk to install on it is not an option. Jul 22 13:44:51 pb__, RP: what facts exactly you needed? Jul 22 13:45:16 I can provide some.. but without serial cable now :) Jul 22 13:45:41 Jay7: btw - you will get original sharp cable or home-made one? Jul 22 13:46:04 hrw: seems like original Jul 22 13:46:05 Jay7: as many non-sharp ones do only one-way work on cxx00 Jul 22 13:46:37 Sharp-Zaurus-SL-5500-CE-170TS-RS-232C Jul 22 13:46:49 ok - thats original one Jul 22 13:46:50 from amazon Jul 22 13:46:55 03mickeyl 07org.oe.dev * radc849af... 10/ (1 packages/busybox/busybox_1.11.1.bb): busybox 1.11.1 apply first two hotfixes Jul 22 13:46:59 03mickeyl 07org.oe.dev * rc04c9a16... 10/ (1 contrib/python/generate-manifest-2.5.py): generate-manifest-2.5.py restore alphabetical order and update version Jul 22 13:47:21 and stylus pack instead of shipped plastic Jul 22 13:56:50 mickeyl, any chance you could post a strace from a successful nfs mount? Jul 22 13:57:58 sure, wait a second Jul 22 13:58:04 thanks Jul 22 14:00:53 http://rafb.net/p/wAHcIz80.html Jul 22 14:03:11 thanks Jul 22 14:03:32 do you use mount from util-linux-ng? Jul 22 14:07:44 root@om-gta01:/local/pkg/fso/framework/tools# ipkg search /bin/mount Jul 22 14:07:45 busybox - 1.9.2-r0 - /bin/mount Jul 22 14:08:32 ah, wait Jul 22 14:08:34 there's Jul 22 14:08:39 root@om-gta01:/local/pkg/fso/framework/tools# ls -l /bin/mount Jul 22 14:08:39 lrwxrwxrwx 1 root root 16 Jan 1 1970 /bin/mount -> mount.util-linux Jul 22 14:08:46 so, yes. it's from util-linux Jul 22 14:08:57 2.12r Jul 22 14:10:33 ah, I have util-linux-ng Jul 22 14:10:48 thanks Jul 22 14:13:01 Crofton: Startrek told us Next Gen is never as good :-D Jul 22 14:13:31 Next Gen >> star trek classic ^^ Jul 22 14:13:41 heh Jul 22 14:13:46 SHATNER! Jul 22 14:13:49 (flame is approching!) Jul 22 14:14:04 whatever Jul 22 14:14:05 (rise the shields!) Jul 22 14:14:22 Bringing up that topic is a kick-bannable offense Jul 22 14:14:28 * XorA throws anti matter at diego_at_work Jul 22 14:14:39 diego_at_work: Oh hi Jul 22 14:14:40 * Crofton cackles Jul 22 14:15:47 Lol. I thought I was in #freebsd, and was surprised to see diego_at_work here :P Jul 22 14:17:02 * mwester is frightened of Trekkies :( too strange.... Jul 22 14:18:07 * XorA is scared when a linux geek calls someone too strange Jul 22 14:18:25 touche! :D Jul 22 14:26:44 XorA: When you get a chance could you reply to the time for action email please Jul 22 14:29:18 RP: I have replied Jul 22 14:30:16 RP: I have some time for testing u-boot & kernel 2.6.24 on akita today Jul 22 14:30:38 RP: what kind on facts you want to see? Jul 22 14:30:55 XorA: Thanks :) Jul 22 14:31:30 Jay7: I don't really have time to talk you through it. Its just a question of debugging your problem step by step Jul 22 14:31:46 RP: ah, ok Jul 22 14:32:31 Jay7: Is your kernel getting the expected commandline? Is that commandline changing the mtd partitions as expected? Can you boot from the filesystem if you don't resize it but do write it using whichever util you use for that Jul 22 14:33:22 1 - can't say now, 2 - looks like (in dmesg), 3 - yes Jul 22 14:34:04 Jay7: If 2 and 3 are happening, it should just work... Jul 22 14:34:15 hm.. Jul 22 14:39:39 RP: what's the order for searching the "mtdparts=" string ?: 1)from atag's, 2)from cmdline, 3)from hardcoded partitioning Jul 22 14:46:56 is the order 3-2-1 ? Jul 22 14:50:17 ant|work: The kernel picks a commandline and then uses it. Where that commandline comes from is the thing that varies Jul 22 14:50:36 The hardcoded paritition layout should only apply if nothing was set in the commandline Jul 22 14:51:49 and atag_cmdline ? has the preference overall? Jul 22 14:52:45 tag should beat the one hardcoded into the kernel Jul 22 14:52:58 but you can tell which its using from the boot messages Jul 22 14:53:44 ->Ignoring unrecognised tag 0x00000000 ?? this ones ?? Jul 22 14:53:53 he Jul 22 14:55:40 so if I read "Using static partision definition" it is because the kernel is defaulting to this Jul 22 14:56:38 btw, can I safely add sleep() to printk()? Jul 22 14:56:55 or here should be some other sleep-function? Jul 22 15:01:33 Jay7: Use mdelay(1000) Jul 22 15:01:42 RP: 10x Jul 22 15:01:53 Its not a nice thing to do in general but should be fine for debugging purposes Jul 22 15:02:17 Hi RP Jul 22 15:02:18 ant|work: yes, I think so Jul 22 15:02:22 hi khem Jul 22 15:02:51 RP: Did you find the 25th hour :) Jul 22 15:03:49 khem: sorry, not yet :( Jul 22 15:04:04 I'm trying though, I did read your updated patch Jul 22 15:04:05 RP: all tools combinations I have tried on gcc 4.x and glibc 2.4+ work Jul 22 15:04:11 and the hint about it getting longer :) Jul 22 15:04:22 Yeah the last one I sent yersterday Jul 22 15:04:52 yeah because I am fixing more fixes unrelated to the patch Jul 22 15:05:40 khem: Would you be able to separate them out? Jul 22 15:06:13 RP: actually so far they are toolchain related Jul 22 15:06:21 so with some care I can Jul 22 15:06:49 khem: That would be good and we can perhaps get them in sooner Jul 22 15:07:09 but I think I should commit the gcc 4.3.1 sdk fix Jul 22 15:07:39 Should I add CONFIG_MTD_CMDLINE_PARTS to kernel in ideal? Jul 22 15:07:42 I didn't see that patch specifically... Jul 22 15:07:54 Jay7: probably, yes Jul 22 15:07:58 its in the bug too Jul 22 15:08:04 ok Jul 22 15:08:15 !oebug 4428 Jul 22 15:08:17 * * Bug 4428, Status: NEW, Created: 2008-07-14 18:41 Jul 22 15:08:18 * * : gcc-cross-sdk 4.3.1 fails to compile Jul 22 15:08:19 * * http://bugs.openembedded.net/show_bug.cgi?id=4428 Jul 22 15:10:30 khem: looks great, please commit :) Jul 22 15:10:39 RP: ok will do Jul 22 15:10:47 khem: cool Jul 22 15:10:58 * khem only have an hour before morning Jul 22 15:11:04 office Jul 22 15:12:42 Jay7: http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-10/msg10762.html Jul 22 15:12:59 RP: ^^ Jul 22 15:14:39 hm.. svs57 wrotes some about this Jul 22 15:18:04 ant|work: yes, and? Jul 22 15:21:51 RP: RK confirms the parameters passed by the bootloader have precedence Jul 22 15:23:09 over CONFIG_CMDLINE? yes, that has always been the case. Jul 22 15:23:36 CONFIG_CMDLINE exists solely to give you some way to provide parameters to the kernel if your bootloader doesn't have a means to pass them in. Jul 22 15:23:37 unless it is zaurus with sharp bootloader... Jul 22 15:23:49 so if we use u-boot we would not need any cmdline nor static partitioning ? Jul 22 15:24:06 hrw: That doesn't matter, the zaurus just ignores the zauus bootloader params Jul 22 15:24:17 it will look at any commandline passed via atags though Jul 22 15:24:19 this is the obscure point...ù Jul 22 15:24:20 ok Jul 22 15:24:38 who ignores what? Jul 22 15:25:06 The kernel ignores any commandline passes via the "old" position in memory method Jul 22 15:25:25 That is zaurus specific to stop the hardcoded 2.4 bootloader breaking things Jul 22 15:25:33 It will listen to any atags based commandline however Jul 22 15:25:44 ah, me shuts up Jul 22 15:25:51 hrw :-) Jul 22 15:26:12 03khem 07org.oe.dev * re4aab546... 10/ (5 files in 3 dirs): Jul 22 15:26:12 Fix ARCH_FLAGS_FOR_TARGET and --with-build-sysroot on gcc 4.3.1. Jul 22 15:26:12 We use ARCH_FLAGS_FOR_TARGET to overcome a bug in --with-build-sysroot implementation where it does not pass LDFLAGS_FOR_TARGET and CPPFLAGS_FOR_TARGET to libstdc++ during build. Jul 22 15:26:52 uboot is vaugely modern and should be passing any commandline through atags Jul 22 15:27:19 florian: you can test the .dev now and see if cross-sdk build with gcc 4.3.1 Jul 22 15:28:06 khem: will do so - many thanks! Jul 22 15:29:27 RP: yes http://www.nabble.com/Passing-mtdparts-to-the-kernel-to2930591.html Jul 22 15:30:04 * Jay7 now compiling kernel with default sharpsl.c, enabled CONFIG_MTD_CMDLINE_PARTS and mdelay() in printk() Jul 22 15:45:26 "vaguely modern" :-D Jul 22 15:46:24 mwester: he he, please defend u-boot ! Jul 22 15:47:15 * Jay7 should today make akita work with root > 64Mb :) Jul 22 15:47:18 anyway :) Jul 22 15:47:40 and anyhow :) Jul 22 15:48:58 03jluebbe 07org.oe.dev * r832167cf... 10/ (1 packages/freesmartphone/zhone_git.bb): zhone: depend on python-cairo for graphs Jul 22 15:49:04 Jay7: you can rate limit printk without mdelay Jul 22 15:49:22 khem: how to? :) Jul 22 15:50:01 Jay7: use jiffies Jul 22 15:50:05 03koen 07org.oe.dev * r252a60e4... 10/ (14 files in 3 dirs): qt4 x11 free: add 4.4.0 Jul 22 15:50:24 khem: sorry, I'm not a linux kernel guru :) Jul 22 15:51:07 I just want to slow down kernel output on boot Jul 22 15:51:11 Jay7: BOOT_PRINTK_DELAY Jul 22 15:51:26 │ Prompt: Delay each boot printk message by N milliseconds │ Jul 22 15:51:33 wow Jul 22 15:52:27 Jay7: http://rafb.net/p/OVrCNh67.html Jul 22 15:52:56 Jay7: hmmm then my solution wont work for you Jul 22 15:53:06 khem: :) Jul 22 15:53:50 mine solution is to debug a part on system which is called very frequently and you want to use printk debugging in their Jul 22 15:54:37 * khem leaves for work that pays bills Jul 22 15:57:08 03koen 07org.oe.dev * r8ae83242... 10/ (1 packages/dsplink/dsplink_1.51.00.08.bb): dsplink: fix preference typo Jul 22 15:58:50 fyi ~ we (bug) announced moving to Poky/OE here: http://bugblogger.com/bug-adopts-poky-linux-155/ Jul 22 16:01:06 03koen 07org.oe.dev * r344b8612... 10/ (1 packages/vlc/vlc-davinci_0.8.6h.bb): vlc: add more configure flags, bump SRCREV Jul 22 16:08:12 is mtn still current or is it git, yet? Jul 22 16:10:18 still mtn Jul 22 16:10:34 kgilmer: I want shot of xeyes on bug Jul 22 16:10:54 kgilmer: btw - does bug works when there are two display modules pluged in? Jul 22 16:13:34 hrw: let me try and get that :) and yes the bug supports two independent displays. Jul 22 16:18:18 kgilmer: what if 3 or 4 will get plugged? :) Jul 22 16:21:21 the SoC we use, mx31 only supports 2 displays. Jul 22 16:21:39 the other two (I think...) will mirror the two independent diplays hrw. Jul 22 16:21:54 at least that's the intended behavior. Jul 22 16:22:36 ok Jul 22 16:22:42 fb0 and fb1 they are? Jul 22 16:26:18 fb0 is the base lcd (on the base unit) Jul 22 16:26:19 fb1 and fb2 are the LCD modules Jul 22 16:26:28 eventually we will have a larger lcd that spans the entire device. Jul 22 16:26:57 hrw, here is the screenshot you requested: http://bugcommunity.com/wiki/images/0/03/LCDModule_screenshot.png Jul 22 16:27:04 pretty sweet, right? Jul 22 16:27:07 have a nice evening guys Jul 22 16:27:30 lovely TWM :) Jul 22 16:46:04 03koen 07org.oe.dev * ra1e27e3d... 10/ (4 files in 3 dirs): mythtv: disable opengl since qt4 doesn't build QtOpenGL Jul 22 16:46:06 * * OE Bug 4450 has been created by trini(AT)embeddedalley.com Jul 22 16:46:08 * * KERNEL_OUTPUT cannot be overridden and isn't always used Jul 22 16:46:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4450 Jul 22 16:51:23 There's a patch on that bug, btw, if anyone wants to take care of it Jul 22 17:10:34 now I can store on my zaurus files bigger than NAND :) Jul 22 17:55:28 /join #git Jul 22 17:55:57 fail Jul 22 18:04:16 lol Jul 22 18:10:56 don't worry Spyro, i fail all the time :) Jul 22 18:12:48 hehe :-) Jul 22 18:12:54 03koen 07org.oe.dev * r874f0942... 10/ (1 packages/mythtv/mythtv_svn.bb): mythtv svn: it now completes packaging for armv5te! Jul 22 18:13:04 piroko: its FAIL not fail :-) Jul 22 18:13:33 Ohh my bad Jul 22 18:19:31 hrw, bug for ya Jul 22 18:19:33 !oebug 4450 Jul 22 18:19:34 * * Bug 4450, Status: UNCONFIRMED, Created: 2008-07-22 18:51 Jul 22 18:19:35 * * trini(AT)embeddedalley.com: KERNEL_OUTPUT cannot be overridden and isn't always used Jul 22 18:19:36 * * http://bugs.openembedded.net/show_bug.cgi?id=4450 Jul 22 18:19:42 patch attached Jul 22 18:21:22 ehh.. Jul 22 18:21:42 du -sx / => 279675k Jul 22 18:21:49 on akita with 128Mb nand.. Jul 22 18:21:54 interesting... Jul 22 18:22:08 Jay7: remember that jffs2 has compression.... Jul 22 18:22:13 hrw: Jul 22 18:22:28 hrw: do not have any words... Jul 22 18:22:43 try "dd if=/dev/zero of=/test" and how much it will take after some time Jul 22 18:22:45 I made zero-filled test file :)) Jul 22 18:23:19 I hate cups Jul 22 18:23:51 "Unable to open parallel port device file: No such file or directory" when printer is on USB... Jul 22 18:23:59 hrw: thank you very much.. i started thinking that I create small black hole in nand :)) Jul 22 18:24:27 hrw: you save my brains :) Jul 22 18:27:22 why I feel that if something can be fucked then it will happen in linux land... Jul 22 18:27:30 btw, who can say me where is used zeroconf in real life? Jul 22 18:27:48 do I really needed avahi-daemon on my PDA? Jul 22 18:28:01 Jay7: Probably not :-/ Jul 22 18:28:15 Jay7: "ssh root@c7x0.local" without caring which IP c7x0 has is useful Jul 22 18:29:29 hrw: it's useful but I have dhcp on my home server :) Jul 22 18:29:59 and every wi-fi spot already have some.. Jul 22 18:30:29 I have not seen yet zeroconf in real life.. only on windows by default :) Jul 22 18:37:20 I'm using OpenEmbedded in combination with the Wii. (This is all based on hacks; not sanctioned by Nintendo.) I figured I'd let it be known here in case someone is half-interested :) - http://stephan.kochen.nl/proj/wii-oe/ Jul 22 18:37:40 stephank: You have my attention Jul 22 18:41:19 ARGHHH Jul 22 18:41:43 here's my oe build http://pastebin.com/m24d745a6 what could be the reason of matchbox-keyboard not beeing present? what should i do in order to have it in my image? Jul 22 18:42:02 oops hello Jul 22 18:43:24 Gnutoo: Use IMAGE_INSTALL += Jul 22 18:49:30 very cool stephank! Jul 22 18:58:46 :) Jul 22 18:59:02 03koen 07org.oe.dev * r91179464... 10/ (1 packages/dsplink/codec-engine_2.10.bb): codec-engine: enhance do_stage and install the OE version of the neuros CE helper script Jul 22 19:05:59 03koen 07org.oe.dev * r87df2840... 10/ (1 packages/dsplink/files/ticel-config): codec-engine: add neuros helper script Jul 22 19:07:04 that's all Jul 22 19:07:22 akita with non-standard NAND partitioning is reality :) Jul 22 19:16:44 piroko, ok thanks a lot Jul 22 19:17:18 Gnutoo: np Jul 22 19:20:41 btw, what is better mediaplayer for zaurus akita? Jul 22 19:21:20 from OE feeds Jul 22 19:21:26 *packages Jul 22 19:49:40 Jay7: mplayer Jul 22 19:51:01 03koen 07org.oe.dev * r54250edb... 10/ (1 packages/dsplink/codec-engine_2.10.bb): codec-engine: prepare do_stage and do_install for more libraries Jul 22 20:02:25 hello, i've a compilation faillure(wesnoth) http://www.pastebin.ca/1079855 Jul 22 20:02:42 what should i do? Jul 22 20:03:01 could someone help me or should i bugreport? Jul 22 20:06:06 piroko,=i have ERROR: Unable to parse conf/bitbake.conf (conf/local.conf:170: unparsed line: 'IMAGE_INSTALL -= "multitap-pad"')...and i need to remove the old keyboard to remplace it... Jul 22 20:06:23 s/=i/I Jul 22 20:07:50 Gnutoo: -= is not a valid bitbake directive. Jul 22 20:08:14 piroko, ok, but what should i do to remove the package? Jul 22 20:09:47 That I'm not sure :( Jul 22 20:10:10 util-linux-ng mount does not work with nfs .... Jul 22 20:10:29 piroko, ok...by the way do have any idea on what i should do for that compilation failure: http://www.pastebin.ca/1079855 ? Jul 22 20:12:28 Yeah Jul 22 20:12:45 Edit the Makefile Jul 22 20:13:06 Change INCLUDEPATH to the staging include path Jul 22 20:13:11 Instead of /usr/include Jul 22 20:14:07 Crud I gotta go. If you need further help, ask. It's a very common problem Jul 22 20:14:14 ok thanks Jul 22 20:14:40 i'll grep the bb files in order to see how they dealt with that... Jul 22 20:16:42 re Jul 22 20:44:55 03koen 07org.oe.dev * raadc5470... 10/ (1 packages/mythtv/mythtv_svn.bb): mythtv: package libmytdb as well Jul 22 20:46:25 khem: Good work - builds again! Jul 22 20:58:06 hello, i patched wesnoth's Makefile.am and it built it until the time where it enters in src...there there is a new Makefile.am and there is no occurences of /usr/include in this Makefile.am but the generated Makefile has some.... Jul 22 20:59:02 here's the grepped Makefile: http://pastebin.com/m77f8aa41 Jul 22 20:59:33 what should i do now? Jul 22 20:59:46 is there a way to patch the generated Makefile? Jul 22 21:04:34 normally the autogen.sh produces the Makefile.in from Makefile.am and configure...which i don't find produce the Makefile Jul 22 21:04:42 from the Makefile.in Jul 22 21:06:17 ah it's in the trunk directory Jul 22 21:44:17 Saving to: `/opt/sources/www.angstrom-distribution.org/unstable/sources/m4-1.4.11.tar.gz' Jul 22 21:44:17 100%[=====================================================================================>] 1,168,840 346K/s in 3.3s Jul 22 21:44:17 2008-07-22 14:25:47 (346 KB/s) - `/opt/sources/www.angstrom-distribution.org/unstable/sources/m4-1.4.11.tar.gz' saved [1168840/1168840] Jul 22 21:44:17 FINISHED --2008-07-22 14:25:47-- Jul 22 21:44:17 Downloaded: 1 files, 1.1M in 3.3s (346 KB/s) Jul 22 21:44:19 NOTE: Task failed: Fetch failed: ftp://ftp.gnu.org/gnu/m4/m4-1.4.11.tar.gz Jul 22 21:55:01 check the md5sum? Jul 22 22:12:39 03mickeyl 07org.oe.dev * rd1634d53... 10/ (1 packages/images/fso-image.bb): fso-image: ship proper illume theme Jul 22 22:12:43 03mickeyl 07org.oe.dev * recab97c7... 10/ (5 files in 3 dirs): illume: rename default theme to illume-theme-asu, depend on any theme Jul 22 22:12:48 03mickeyl 07org.oe.dev * r8709d29f... 10/ (1 packages/freesmartphone/illume-theme-freesmartphone_git.bb): add illume-theme-freesmartphone Jul 22 22:12:54 03mickeyl 07org.oe.dev * rd2cf665b... 10/ (1 conf/distro/include/fso-autorev.inc): fso-autorev.inc: add entry for illume-theme-freesmartphone Jul 22 22:15:51 RP: thanks for the inputs today Jul 22 22:15:58 pb__: you too Jul 22 22:25:59 morning all Jul 22 22:26:23 good morning! Jul 22 22:26:29 you're the first today Jul 22 22:26:33 ;-) Jul 22 22:26:38 How are you doing? Jul 22 22:26:46 How is life and university? Jul 22 22:30:06 both is great. Jul 22 22:31:03 Although its a pity that my time at university is over soon. Jul 22 22:31:24 Btw. did you notice that I put my git repository at l2go? Jul 22 22:45:56 no Jul 22 22:46:03 I don't even know about l2go Jul 22 22:46:10 How is the collie? Jul 22 22:46:18 Is it doing well? Jul 22 22:46:38 Oh man, it has to be late ;-) Jul 22 22:49:40 hi thesing Jul 22 22:50:05 I didn't do much in the past month. I'm catching up with mainstream now. Jul 22 22:50:14 I found new forces for finishing initramfs-bootmenu-image Jul 22 22:50:19 hopefully Jul 22 22:50:39 and will push my changes to my repository at linuxtogo. Jul 22 22:51:08 ant_: great. who is it? and what does he/she want to implement? Jul 22 22:51:11 good Jul 22 22:51:17 he is Jay7 Jul 22 22:51:27 :) Jul 22 22:51:28 look of th logs on #angstrom Jul 22 22:51:45 urgh, two typos in one line Jul 22 22:51:50 is very late... Jul 22 22:52:52 thesing: now I have akita with u-boot :) Jul 22 22:53:15 could you summarize? I don't really want to read logs. Jul 22 22:53:22 thesing: I like idea of having single kernel and initramfs with menu Jul 22 22:54:00 do you plan to implement the menu? Jul 22 22:54:16 so I can at least testing images Jul 22 22:54:41 yes, I can try implement menu Jul 22 22:55:08 but later.. after 4 august :) Jul 22 22:55:23 * thesing noticed that the last time that I started my development machine is one month ago. Jul 22 22:55:30 * thesing watches to much TV ;) Jul 22 22:55:46 TV is pure evil ;) Jul 22 22:56:18 I think the best solution would be a new implementation in C + klibc. Jul 22 22:56:41 you could just redo what the current shell scripts do. Jul 22 22:57:46 may be.. I cant say anything now.. I have not seen this image yet :) Jul 22 22:58:17 I need some investigating before Jul 22 22:58:33 may be shell is enough Jul 22 22:59:00 you need regular expressions which klibc doesn't have. Jul 22 23:00:11 so you can't use expr thats why the existing bootmenu doesn't work with klibc atm. Jul 22 23:01:59 thesing: other distros eare enabling 'awk' for initramfs. could help? Jul 22 23:02:08 I have look at 30-bootmenu.sh once more.. Jul 22 23:02:22 yes, it can be implemented in C Jul 22 23:02:42 thesing: we have 6mb of space... Jul 22 23:03:41 lua is 200Kb :) Jul 22 23:03:49 I have idea of lua-init :) Jul 22 23:03:53 ant_: my goal is minimal kernel + bootmenu = 1 MB Jul 22 23:04:06 ok, that's the regular way Jul 22 23:04:14 1,2 mb limit Jul 22 23:04:35 1,2Mb would be great.. Jul 22 23:05:05 then you would not need u-boot Jul 22 23:05:24 I personally need :) Jul 22 23:05:38 but by other reason :) Jul 22 23:07:12 its doable. currently klibc ash + several tools are small enough. So one C programm that does all the stuff can be written. Jul 22 23:07:52 but of course with working u-boot your motivation is different. Jul 22 23:08:18 thesing: but where is emergency image should be located? Jul 22 23:08:44 Jay7: the idea is to respect sharp tools Jul 22 23:09:04 with 6Mb we can make bootmenu initrams and emergency image in one Jul 22 23:09:04 and standards Jul 22 23:09:24 respecting standarts is true.. Jul 22 23:09:29 and a video too... Jul 22 23:09:37 ant_: not for me. The thing is that there is no u-boot for collie. Jul 22 23:09:48 because noone bothers Jul 22 23:10:19 there are here in OE several skilled people Jul 22 23:10:32 but zero interest for zaurus Jul 22 23:10:47 anymore Jul 22 23:11:12 there are better machines now. Jul 22 23:11:25 that's not a point Jul 22 23:11:33 btw, may be we should check other bootloaders? Jul 22 23:12:38 I think it is too much work to port a bootloader. and you don't win that much. Jul 22 23:12:45 with bootmenu we needed just 'dumb' loader Jul 22 23:13:39 thesing: from this point sharp loader is better alternative Jul 22 23:13:50 it works as is :) Jul 22 23:14:30 on collie I have 16 MB flash. 1 MB of this is for kernel. I want a kernel with initramfs+bootmenu in there that checks what images and kernels are available (sd, cf, flash) displays them and starts the selected. Jul 22 23:14:56 as a bonus it could remember decisions but this is secondary. Jul 22 23:15:37 grub :) Jul 22 23:16:51 ok, let's make it in next month :) Jul 22 23:33:06 * * OE Bug 4451 has been created by bluelightning(AT)bluelightning.org Jul 22 23:33:08 * * h2200-bootloader broken causing suspend failures Jul 22 23:33:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4451 Jul 23 00:05:47 * Jay7 -> sleep() Jul 23 00:06:07 sleep well Jul 23 00:59:49 hello, i've a scons problem: http://pastebin.com/m6eb358ae Jul 23 01:16:38 hello...is there someone there? **** ENDING LOGGING AT Wed Jul 23 02:59:56 2008