**** BEGIN LOGGING AT Thu Nov 29 02:59:59 2012 Nov 29 16:37:32 Jay7: ping Nov 29 16:37:50 ant_work: pong Nov 29 16:38:10 hi there Nov 29 16:38:41 see, I promised to Koen I'll send the setup for the new Narcissus Nov 29 16:38:51 ah, nice Nov 29 16:38:52 (for Zaurus) Nov 29 16:38:59 now, it implies ubifs... Nov 29 16:39:15 ah Nov 29 16:39:31 I'd not add a patch to OE... Nov 29 16:39:51 when is deadline? :) Nov 29 16:41:13 no idea :) maybe during seasons holidays Nov 29 16:41:36 I think we have on emonth.. Nov 29 16:41:46 I hope we will setup hackaton on 8 Dec Nov 29 16:42:09 great Nov 29 17:04:54 btw we'll have to resync fstype.c Nov 29 17:05:10 I'll send you a patch Nov 29 17:06:20 ah, I've read kexwec may now need a --atags option Nov 29 17:06:38 default is device-tree now :/ Nov 29 17:07:21 I don't remember but I think we already discussed the need of more options for kexec Nov 29 17:24:51 seems we need to re-discuss some things Nov 29 17:26:51 and collect them before :) Nov 29 17:30:15 yes, I'd start with the 'old' patches, ubifs, menu Nov 29 17:30:23 then optimizations Nov 29 17:30:28 then new features Nov 29 17:31:02 so I can take a git tag and update OE release Nov 29 22:55:42 dromede: it looks like if we insist we got all patches merged in 3.8 Nov 29 22:56:35 i hope so Nov 29 22:56:56 i'll send two more in a few minutes Nov 29 22:57:02 the pxa25x patches are already in linux-next Nov 29 22:57:02 just some cleanups in spitz.c Nov 29 22:57:11 good news Nov 29 22:57:40 now the sharpsl_param ... Nov 29 22:58:01 btw, have you tried booting w/out kexec? Nov 29 22:58:14 my first idea was that we need a more recent kexec Nov 29 22:58:38 but it looks like you've found the reason Nov 29 22:59:16 if you want to boot without kexec, with the sharp bootloader, you need to get your kernel below ~1.2MiB Nov 29 22:59:39 i don't think this would be easy with todays kernels Nov 29 22:59:49 sure, I mean it would fail even booted by the bootloader Nov 29 23:00:01 of course Nov 29 23:00:05 I have to test with linux-kexecboot_3.4 Nov 29 23:00:12 i could use your help with this: Nov 29 23:00:14 http://lists.infradead.org/pipermail/linux-arm-kernel/2012-November/135380.html Nov 29 23:00:24 i'm pretty sure it's akita specific Nov 29 23:00:59 how do you trigger the issue on akita? Nov 29 23:01:23 so the problem is, every time i try to change the backlight, i get a big WARN_ON output Nov 29 23:01:31 just by writing to /sys/class/backlight/corgi_bl/brightness Nov 29 23:01:41 i'm pretty sure it's akita specific Nov 29 23:02:02 ah, ok, I can quickly test with 3.2 images Nov 29 23:02:11 exactly :) Nov 29 23:02:32 any 3.? kernel will do Nov 29 23:02:42 probably older ones Nov 29 23:03:03 i'm currently running 3.7-rc7 on my akita Nov 29 23:03:09 and i stumbled upon a weird bug Nov 29 23:03:37 when i try to but vanilla 3.7-rc7, i get no backlight Nov 29 23:04:11 and then i find out that the max7310 gpio expander is not even registered in 3.7-rc7 Nov 29 23:04:16 but there's no reason for that Nov 29 23:05:15 ok, I can tell you that locomo-bl reacts ok. corgi is charging... Nov 29 23:05:51 git bisect pointed me to this commit: Nov 29 23:05:52 http://comments.gmane.org/gmane.linux.kernel.commits.head/346901 Nov 29 23:06:31 and truly enought, if i revert that commit, the max7310 gets registered and works fine Nov 29 23:07:03 i simply can't see any correlation between this commit and this bug Nov 29 23:08:03 corgi mis ok, I went from 31 to 16. no warnings on screen Nov 29 23:08:10 *is ok Nov 29 23:08:16 ok, how about terrier? Nov 29 23:08:19 now spitz...one mom Nov 29 23:10:13 spitz (3200) reacts ok as well Nov 29 23:10:26 I dimmed from 31 to 16. no warnings Nov 29 23:11:06 ok then, as i suspected, it's only akita related Nov 29 23:11:15 i.e. with 255 I get write error: Invalid argument Nov 29 23:11:43 31 is maximum Nov 29 23:12:24 only akita has backlight gpio controls on it's max7310 expander Nov 29 23:12:51 hm.. it accepts values up to 47. 48 is invalid Nov 29 23:12:54 corgi and other spitz variants backlight gpio controls are on the scoop chip Nov 29 23:13:20 i never tried going above 31 Nov 29 23:13:29 but yes, 48 is defined as maximum in spitz.c Nov 29 23:13:46 I don't see much diff Nov 29 23:14:30 is it 5 bits ? Nov 29 23:15:53 ah, 0 is darl, 1 is actually acceptable as min Nov 29 23:15:57 *dark Nov 29 23:16:27 seems active range is indeed 0-31 Nov 29 23:16:47 must be logaritmic Nov 29 23:18:23 dromede: it's funny how akita subtly differs from corgi and spitz (not talking about tosa yet) Nov 29 23:20:20 akita has one less scoop chip and one less cf slot Nov 29 23:20:31 these scoop schips are amon other things, gpio expanders Nov 29 23:20:53 instead of the second scoop chip, it has the max7310 Nov 29 23:21:07 and tosa is a very different beast indeed Nov 29 23:21:37 and the best one of them all in terms of hardware Nov 29 23:21:48 I thought akita instead had that pca953x stuff Nov 29 23:21:57 as gpio expander Nov 29 23:22:05 pca 953x == max7310 Nov 29 23:22:08 same thing Nov 29 23:22:12 (note that I vaguely imagine what it is :) Nov 29 23:22:19 :) Nov 29 23:22:41 it just gives you a bunch of gpio lines over an i2c bus Nov 29 23:23:27 I see..I remember what happens if you forget to configure it... Nov 29 23:44:30 about that patch, I guess it's touching angel (bootloader) code in head.S Nov 29 23:54:43 falling asleep...gn **** ENDING LOGGING AT Fri Nov 30 02:59:58 2012