**** BEGIN LOGGING AT Thu Nov 27 02:59:59 2014 Nov 27 08:52:06 ant_work, hello. Nov 27 08:52:12 Not all of them. Nov 27 08:54:58 ant_work, patches 1-4 should be enough. Nov 27 08:55:32 It should be sufficient for fixing the IRQ0 issue. Nov 27 09:32:22 gm Nov 27 09:32:53 ah, ok, I have used the first patch you sent to lakml Nov 27 10:46:06 ant_work, this serie is better, because it contains fixes for irq0 Nov 27 10:49:24 so we have pending 6 clock + 8 irq + 16 locomo + 1 gpio-charger patches Nov 27 10:49:26 :/ Nov 27 10:50:19 to be fair I shouldn't take in account the 16 locomo patches Nov 27 11:26:26 ant_work, that is why I have that asana board. Nov 27 11:52:50 ant_work, what should be the deault trigger for green led? I'm thinking about switching from NAND to mmc0. Nov 27 11:53:06 Especially because collie doesn't have nand Nov 27 11:54:25 Hmm. Something broke spi. Nov 27 12:01:24 fixed Nov 27 12:28:31 regression in 3.18-rc6? Seems there isn't much new concerning collie Nov 27 12:28:54 no, in my patches. Nov 27 12:29:02 I tried to optimize spi driver a bit. Nov 27 12:29:43 I'm going to send you another locomo patch serie. Nov 27 12:29:46 Now 18 patches. Nov 27 12:30:28 It will need IIO, m62332 DAC, regulators and fixed regulator Nov 27 12:30:33 - Kconfig changes Nov 27 12:36:34 Hmm. got oom on poodle while booting the kernel. Nov 27 12:38:46 with the images I uploaded? Nov 27 12:38:58 these run fine on real hw Nov 27 12:40:50 no. my testing kernel Nov 27 12:41:08 what is the memory size on poodle? Nov 27 12:42:15 32m Nov 27 12:42:36 a.k.a. be carefull with compression options Nov 27 12:43:02 especially with lzma/xz Nov 27 12:49:14 maybe that's the issue Nov 27 12:49:52 Hmm. even collie has 64M, has it not? Nov 27 12:56:54 unfortunately not...c700, 5000d and 5600 have just 32mb Nov 27 12:57:13 for long time c700 and poodle where misterioulsy not booting ;) Nov 27 12:57:46 but it was failing at initial decompression, not OOM Nov 27 13:01:19 ok. Nov 27 13:01:43 So 5500 is better than 5600. Nov 27 13:01:48 thank you, sharp :) Nov 27 13:03:53 5600 is really an hybrid Nov 27 13:05:42 Display is better, but mem is worse. Nov 27 13:09:21 I don't remember the specs but maybe the dram of pxa are faster? Nov 27 13:10:06 as for nor vs. nand it's not so evident for that limited sizes Nov 27 13:10:38 nand is much more prone to ecc eroors Nov 27 13:10:50 (modern nands at least) Nov 27 13:12:09 btw, next to-do in the list is debugging the touchscreen calibration Nov 27 13:12:15 in userspace Nov 27 13:12:44 after kernel driver changes I'm under the impression some values are swapped or maybe the scale changed Nov 27 13:12:59 ucb1x00 (a touch screen + ADC +...) on collie is old, semi-unmaintained, strange, etc. Nov 27 13:13:02 fact is, only poodle can succerssfully run OPIE Nov 27 13:13:12 and calibrate Nov 27 13:14:34 I'm really puzzled....me and anarsoul tried to debug it but did never finish Nov 27 13:14:49 my guess is rotation is the culprit Nov 27 13:15:10 ...but calibration is ok on poodle which has rotated fb so... Nov 27 13:15:55 same issues with xinput-calibrator fwiw Nov 27 13:16:12 ok, next in to-do list;) Nov 27 13:16:33 poodle IIRC has a different touchscreen chip. Nov 27 13:16:43 ah...I still wonder if we should write a little kernel module for reading nand partitioning for pxa zaurus Nov 27 13:17:05 That would be nice. Nov 27 13:17:16 or use the BLKPG ioctl to repartition on the fly in kexecboot Nov 27 13:17:44 in kernel would be surely better... Nov 27 13:54:06 next "big" topic for me would probably be a pm/charger/etc. on poodle. Nov 27 14:03:41 heh, yes Nov 27 14:04:14 we must check the embedix sources and verify if there was smthg special on poweroff Nov 27 14:04:21 like for corgi Nov 27 14:08:00 charger. Nov 27 14:08:35 after several experiments, my poodle boots into kexecboot, scans nand (led is blinking) and then powers off. Nov 27 14:08:43 Probably due to low battery. Nov 27 14:08:49 It worked an hour ago. Nov 27 14:09:07 It's a pity that they have so small amount of memory. Nov 27 14:09:15 s/they/it/ Nov 27 14:10:17 Hmm. http://www.rpsys.net/openzaurus/patches/poodle_serial_vcc-r0.patch Nov 27 14:10:27 gtg now, sorry Nov 27 14:10:45 bye Nov 27 15:31:36 ge ant_work bluelightning Nov 27 15:31:49 wb Nov 27 15:31:56 hi lumag_, ant_work Nov 27 15:32:32 hi bluelightning Nov 27 15:33:53 what's up? you seem really busy since some months..are you maybe planning to marry ? Nov 27 15:33:56 :p Nov 27 15:37:08 ant_work: I did already get married a few months ago as it happens :) Nov 27 15:37:38 I guess I mostly stopped doing OE-related stuff at home, which is maybe a bit sad Nov 27 15:38:06 it wasn't a conscious decision Nov 27 15:38:27 I spend way too much time watching youtube videos at home if I'm not busy doing something else... Nov 27 15:38:28 bingo ! Nov 27 15:39:09 enjoy Nov 27 15:40:49 you'll have a little time-window for your hobbies by having little kids around ;) Nov 27 15:41:11 bluelightning, congratulations Nov 27 15:42:41 ..but once kids grow forget about TV/tablets...all will be monopolized by them Nov 27 15:43:06 be prepared to watch a lot of cartoons Nov 27 15:47:42 thanks :) Nov 27 15:47:52 ant_work: heh... well, no kids yet Nov 27 16:52:47 ok, gpio-charger for collie got accepted. Nov 27 16:53:26 by rmk Nov 27 17:04:58 nice, one less Nov 27 17:11:06 I really never cease to be amazed by the work you guys are doing with this... thanks, FWIW :) Nov 27 17:16:41 no need to thank, is just an hobby Nov 27 17:16:56 I find that more interesting than tv or crosswords Nov 27 17:17:00 ;) Nov 27 17:18:00 I guess I just never expected all of this gnarly kernel stuff to get resolved, I'm pleased that it is Nov 27 17:18:50 mostly thx to lumag :) Nov 27 17:21:42 bluelightning, it's just my current project at work gave me some spare time. Nov 27 17:22:00 And quite significant, unfortunately. Nov 27 17:22:15 And hacking a kernel is fun. Nov 27 17:28:12 what is the difference between sl-5600 and b600? Nov 27 17:28:25 I thought that those are the same, but 2.4 kernel differentiates them. Nov 27 17:31:00 I thought they were pretty identical Nov 27 17:31:47 *b500 Nov 27 17:33:30 Because sharpsl_battery driver in one of 2.4 versions has the following: Nov 27 17:33:43 #if POODLE && !SHARP_SL_J Nov 27 17:33:43 ... Nov 27 17:33:48 #elif CORGI Nov 27 17:33:49 ... Nov 27 17:33:53 #else Nov 27 17:33:58 ... Nov 27 17:33:59 #endif Nov 27 17:34:02 Strange Nov 27 17:42:18 next item to get through Russell: irqs and no-cpufreq. Nov 27 17:59:31 bbl **** ENDING LOGGING AT Fri Nov 28 02:59:58 2014