**** BEGIN LOGGING AT Wed Sep 12 02:59:59 2012 Sep 12 08:08:42 dromede: any other fixes apart slob/slab and re-enabling cpufreq in your kernels? Sep 12 08:09:19 I'll work on defconfig fragments soon, for 3.4 Sep 12 08:10:48 btw: [PATCH 0/7] Add device-tree support to kexec-tools for ARM Sep 12 08:12:39 and OFC how-to "Booting DT kernel on a non-DT bootloader" Sep 12 09:07:51 dromede: any time to help writing similar kernel driver for Zaurus? Sep 12 09:07:53 http://lists.infradead.org/pipermail/linux-mtd/2012-August/043694.html Sep 12 09:07:58 ^? Sep 12 12:13:55 wow, I've got in touch with Zajec, the author, meybe he'll help us very soon :) Sep 12 13:02:09 bluelightning: we have to revert ant_work> JaMa: I think we should just revert http://cgit.openembedded.org/meta-handheld/commit/conf/machine/include/zaurus.inc?id=2ed328bd842c96496fb5bbeecd3f6c7ed2df4c65 Sep 12 13:05:51 ant_work: it's not really a revert, more of a reverse - i.e. the original change wasn't wrong at the time Sep 12 13:06:15 yes, the effect is the same. Sep 12 13:06:28 what I mean is, the commit message needs to explain why we're making the change back and does not need to refer to the original commit Sep 12 13:07:10 before doingthat, out of curiosity I'd like to see what is providing xserver-xorg-lite Sep 12 13:07:14 ... Sep 12 13:07:37 iirc RP said onc we should avoid to set XSERVER in the machine.conf Sep 12 13:08:18 and talking with Martin, Sep 12 13:08:19 JaMa> but mesa-xlib should be now changed too Sep 12 13:08:19 I have it set by distro config.. Sep 12 13:09:12 if the defaults are now for xserver-xorg and for mesa-dri I'd remove those settings from here alltogether Sep 12 13:09:54 well, or just repat the exact defaults here as documentation :) Sep 12 14:24:34 bluelightning: seems http://cgit.openembedded.org/openembedded-core/tree/meta/conf/distro/include/default-providers.inc?id=8b0a1ce417feea5f58fc08f54025343c7c7ff892 Sep 12 14:24:37 will fix the defaults.I'll send a patch commenting out Sep 12 14:25:31 why comment out? just delete... Sep 12 14:26:52 ok. btw I've read pb postinst runscript patch will not be merged in freeze-time Sep 12 14:27:22 if you look at the cpio of yesterday, there is still extra stuff Sep 12 14:28:08 I'm waiting to know that kernel boots on collie before touching that :) Sep 12 14:29:55 ouch.. Sep 12 14:30:10 it appears the patch has been merged only few mins ago... Sep 12 14:30:14 [oe-commits] Khem Raj : gcc-4.7: Fix build for armv4/EABI and ppc/Os Sep 12 14:30:18 this patch :/ Sep 12 14:32:23 why ouch? Sep 12 14:37:14 because the kernel you downloaded is 'old' Sep 12 14:37:51 somehow I thought gcc patch was already committed Sep 12 14:38:17 oh ok Sep 12 14:38:23 I haven't tested it yet so I'll wait Sep 12 14:38:35 see, time saved :] Sep 12 14:39:02 one less 'no, doesn't boot' :) Sep 12 14:39:08 heh Sep 12 15:06:50 bbl Sep 12 20:26:19 bluelightning: wow, that was a malformed header what I've sent Sep 12 20:27:03 ant__: heh... that's why I always add --confirm=always to git send-email ;) Sep 12 20:28:49 I'm rebuilding the collie k-kernel rigth now from scratch, will take 1 hour I guess Sep 12 20:30:51 ok cool Sep 12 20:31:07 I'm resending the patch now. remember the lzma one... Sep 12 20:31:16 yep Sep 12 20:31:31 #34869 Sep 12 20:31:57 so this kernel will use the lzma compression right? Sep 12 20:38:58 yes Sep 12 20:39:17 as the kexecboot one Sep 12 20:39:25 patch seems now in better shape :) Sep 12 20:41:05 #36437 Sep 12 20:44:06 bluelightning: once xz is well tested I'd move all Zaurus to that. Initial boot is a bit slower but we gain 500KiB Sep 12 20:44:31 (I've seen patches for 3.6 around, better wait :) Sep 12 20:52:38 bluelightning: pls check you can traverse the /oe dir on my server Sep 12 20:52:47 ...and reach the images... Sep 12 21:01:56 ant__: /oe says forbidden Sep 12 21:02:25 hm.. ok, I'll try to do it 'properly' Sep 12 21:09:33 can you pls try with /ipk now? Sep 12 21:35:20 ant__: nope sorry Sep 12 21:35:31 I mean still forbidden Sep 12 21:44:30 np, grab this: /collie/zImage-yocto-tiny-kexecboot-3.2.18+git1+486f7aec824b4127e91ef53228823e996b3696f0_1+da6291c46f37e2fb799d964c807d8d57ea89c375-r4.0-collie.bin Sep 12 21:52:29 ok got it Sep 12 22:03:16 would you try one lst time with /images? I added apache to my users temprarly.. Sep 12 22:13:00 ant__: nope sorry Sep 12 22:13:08 ant__: not selinux by any chance? Sep 12 22:15:56 no, just normal permission issue. I have an 'unconfigured' apache here...and /oe partition has andrea:users permissions Sep 12 22:17:40 hmm Sep 12 22:17:47 well, the file therein, /oe is root:root Sep 12 22:18:01 well unfortunately I can report this kexecboot image is no better on collie :( Sep 12 22:18:07 same behaviour Sep 12 22:18:43 ok, I have pulled Khem patches this time :/ Sep 12 22:19:50 can you disable the kernel boot logo as well? Sep 12 22:20:01 yes, sure Sep 12 22:20:33 sadly is kernel debug time :/ Sep 12 22:39:49 hm.. CONFIG_DEBUG_USER ? Sep 12 22:44:25 looking at the help text I think that's for userspace crashes Sep 12 22:44:32 later maybe. let'stry Sep 12 22:45:08 building now. if this fails I'll add init=/bin/sh Sep 12 22:45:33 no logo, no jffs2-ext4-ubi, printk, bug Sep 12 22:45:44 and debug in cmdline Sep 12 22:45:50 kexecboot text-ui only Sep 12 22:46:15 let's give one more chance to kexecboot :) Sep 12 22:50:19 bluelightning: pls get /collie/zImage-dev_kexecboot Sep 12 22:53:32 ant__: ok, I get text but it's unreadable Sep 12 22:53:45 ant__: it vaguely looks like a kernel panic Sep 12 22:54:06 no serial I suppose Sep 12 22:54:25 I mean, kernel could not open a serial Sep 12 22:54:38 it would be great if we could fix the framebuffer initialisation, that is also broken aside from whatever is stopping it booting Sep 12 22:55:00 I don't know, I don't have a serial cable to test with in any case Sep 12 22:57:58 ok, let's try to jump on /bin/sh Sep 12 22:58:02 mom Sep 12 23:01:16 MAGE_INSTALL = "kexec-klibc kexecboot-klibc ubiattach-klibc klibc-static-utils-sh klibc-static-utils-ls klibc-static-utils-mount" Sep 12 23:05:08 bluelightning: /collie/zImage-dev_shell Sep 12 23:08:27 you've the older, file renamed Sep 12 23:09:23 pls try /collie/zImage-dev_shell Sep 12 23:09:30 ah whoops, redownloading Sep 12 23:09:40 you're monitoring logs I see ;) Sep 12 23:10:02 yes, just to see how long it takes for some bot :) Sep 12 23:10:12 no bites yet? Sep 12 23:10:44 heh, no ^_^ Sep 12 23:11:50 I ought to swap my router with the one with usb-stick and share Sep 12 23:12:09 now, whaere is that one... Sep 12 23:13:15 no difference sorry :( Sep 12 23:14:07 Jay7: I'll pm you the link if you can d/l and post the serial logs.. Sep 12 23:15:34 hm, probably I should swap the master console to tty for you, we have it on serial Sep 12 23:15:43 ^ for bluelightning Sep 12 23:16:36 won't do me much good if it's not readable though... Sep 12 23:16:42 bluelightning: for fun, pls download kernelreplacement-2.6.31.12-r10-rep3.tgz Sep 12 23:17:05 ouch, let's rename it Sep 12 23:17:42 /collie/replacement-collie.tgz Sep 12 23:19:31 tell me this one boots... Sep 12 23:20:20 ah, it's for older kexecboot judging by the size Sep 12 23:21:15 bluelightning: tomorrow I'll try with 3.4 Sep 12 23:21:40 maybe we'll have better luck Sep 12 23:21:53 good night Sep 12 23:23:25 night **** ENDING LOGGING AT Thu Sep 13 03:00:00 2012