**** BEGIN LOGGING AT Tue Oct 01 02:59:58 2013
Oct 01 13:55:23 <ayaka> about https://lwn.net/Articles/561462/ I wonder does dts need to be modify to work with new kernel when new kernel is released
Oct 01 14:02:49 <pabs3> ayaka: based on the article it seems that might be true but that Linux folks don't want it to be true
Oct 01 14:10:04 <ayaka> pabs3, so dts will like old board file you can't write once and forget
Oct 01 14:12:10 <pabs3> well board files are embedded in the kernel, so if you modify your kernel you might need to modify the board file. with dts you are more likely to update your kernel and so you might need to modify your dts when you do
Oct 01 14:14:09 <ayaka> about writing dts, I have a problem, how to find a full lists of compatible
Oct 01 14:17:04 <ayaka> I wanted to write a dts for a board whose SoC is s3c6410, but I don't know the compatible for each entry
Oct 01 14:20:04 <pabs3> I've no idea, but #kernelnewbies might be able to help
Oct 01 14:20:11 <ayaka> ok
Oct 01 14:20:42 <ayaka> as I don't have writed dts and don't use any board whose kernel with a dts
Oct 01 14:21:13 <ayaka> Could you give me a sample that when kernel updata that dts need to be modify
Oct 01 14:21:36 <ayaka> I don't know which part of dts shouldn't be modifed
Oct 01 14:25:37 <pabs3> I only know what I read in the article
Oct 01 14:28:53 <ayaka> pabs3, thank you
Oct 01 14:29:15 <pabs3> I wasn't very helpful :)
Oct 01 14:29:58 <ayaka> actually that article is a little difficult for me, as I only know the meaning of abi but I don't know what is it really for
Oct 01 14:30:20 <ayaka> pabs3, you have helped a lot, at least I know what I read it correct
Oct 01 14:31:33 <pabs3> ABI in this context I guess is the names the drivers accept from the dts.
Oct 01 20:21:13 <dos1> hmm
Oct 01 20:21:39 <dos1> DocScrutinizer05: do you remember if AUX LED on Freerunner was supposed to be dimmable?
Oct 01 20:22:15 <dos1> I can dim both power LEDs, but aux one shines brightly regardless of brightness value I set
Oct 01 20:22:41 <DocScrutinizer05> dos1: basically it's dimmable, yes
Oct 01 20:22:51 <DocScrutinizer05> ooh wait, aux
Oct 01 20:22:55 <DocScrutinizer05> I dunno
Oct 01 20:23:06 <DocScrutinizer05> for sure via bitbang
Oct 01 20:23:24 <DocScrutinizer05> but I can't recall if it had a PWM timer associated
Oct 01 20:25:02 <dos1> I came across this mail - http://lists.openmoko.org/pipermail/openmoko-kernel/2008-July/003842.html
Oct 01 20:25:32 <dos1> was it about AUX LED eating too much current, or something else?
Oct 01 20:25:39 <dos1> unfortunately link is bogus
Oct 01 20:26:35 <dos1> that's interesting too: http://lists.openmoko.org/pipermail/openmoko-kernel/2008-July/003841.html
Oct 01 20:26:35 <DocScrutinizer05> there been an issue with some LED driven without proper series resistor or sth like that
Oct 01 20:26:57 <dos1> on my device, when I turn on some power led, aux is turned off
Oct 01 20:27:18 <DocScrutinizer05> ooh wait, now I recall: the driver transistor *basis* pin had no series resistor, so GPIO had basically a short to ground
Oct 01 20:27:58 <dos1> and turning on aux led eats something like 50mA
Oct 01 20:28:01 <DocScrutinizer05> we replaced the driver transistor by a type with built-in basis resistor
Oct 01 20:28:37 <dos1> all this time I thought this is normal with a5 and it was fixed in a6
Oct 01 20:28:53 <dos1> but now I read there that all a5 devices should be reworked
Oct 01 20:29:31 <DocScrutinizer05> aiui yes
Oct 01 20:29:35 <dos1> either mine isn't or i don't understand
Oct 01 20:32:09 <dos1> now, which one is it? :)
Oct 01 20:41:55 <dos1> DocScrutinizer05: was the power leakage only about aux, or about all leds?
Oct 01 20:43:18 <dos1> if about all, then i guess the answer is "power leds were reworked, but aux was missed"
Oct 01 20:43:56 <dos1> if only about aux, then "i've got a device which wasn't intended to be sold"
Oct 01 20:44:50 <dos1> based on http://lists.openmoko.org/pipermail/community/2008-April/015843.html
Oct 01 21:37:00 <DocScrutinizer05> dos1: it was only one LED iirc
Oct 01 21:37:40 <DocScrutinizer05> ooh, maybe your first version might actually be right
Oct 01 21:38:08 <DocScrutinizer05> wait, I think I got a schematics of A4 or sth somewhere
Oct 01 21:40:08 <DocScrutinizer05> hmm no, GTA02_Schematic_MB_A5_1220.pdf
Oct 01 21:41:00 <dos1> well, A5 had this problem
Oct 01 21:41:32 <dos1> it was reworked after production and before shipping
Oct 01 21:42:01 <dos1> so I guess it should still be in the schematics?
Oct 01 21:45:54 <DocScrutinizer05> http://wstaw.org/m/2013/10/01/plasma-desktopdy3743.png
Oct 01 21:48:49 <DocScrutinizer05> and this is A7: http://wstaw.org/m/2013/10/01/plasma-desktopuj3743.png
Oct 01 21:49:21 <DocScrutinizer05> resp A5..A7
Oct 01 21:49:47 <DocScrutinizer05> R1564=0R for A5
Oct 01 21:50:25 <DocScrutinizer05> so yep, in A5 the AUX been borked
Oct 01 21:51:23 <DocScrutinizer05> God, this is almost lost in oblivion, years ago
Oct 01 21:52:14 <dos1> so, "reworked a5" should have aux led fixed, as there was no problem with power leds, right?
Oct 01 21:54:03 <dos1> hell, that mail is from april 2008
Oct 01 21:54:18 <dos1> my device has 20080621 on the sticker
Oct 01 21:57:34 <dos1> so I don't think this is some kind of RD device or sample Tony is mentioning in that mail
Oct 01 21:58:20 <dos1> interesting mystery :)
Oct 01 21:58:48 <DocScrutinizer05> dos1: the contrary, it seems A5 got reworked by replacing Q1501 which left AUX LED unfixed
Oct 01 21:59:34 <dos1> hmm, now i see
Oct 01 21:59:46 <dos1> ok then, mystery solved :D
Oct 01 21:59:49 <DocScrutinizer05> GTA02_Schematic_MB_A5_1220.pdf  in internal clearly shows that originally none of the 3 LED in A5 had a basis resistor
Oct 01 22:02:55 <dos1> DocScrutinizer05: thanks, now it's clear
Oct 01 22:03:02 <DocScrutinizer05> :-)
Oct 01 22:03:08 <dos1> so comments on Werner patches were wrong
Oct 01 22:03:34 <dos1> leds were fixed, except of AUX, which was the one patch was dealing with :)
Oct 01 22:09:37 <dos1> random unsolicitied fact: I've got a debubo
Oct 01 22:12:04 <dos1> so I'm looking forward for some random, clueless kernel and bootloader hacking after neo900 arrives :)
Oct 01 22:13:49 <dos1> but before that happens... I plan to work on some rock-solid SHR image for gta02 we can be proud of
Oct 01 22:15:57 <dos1> no one is active anymore... well, good, I'll be able to do some controversive stuff regarding our library infra ]:->
Oct 01 22:20:37 <jake42> dos1: 1+ ;)
Oct 01 22:20:49 <jake42> I'll try to help
Oct 01 22:21:11 <dos1> I consider my SHR installation pretty stable
Oct 01 22:21:19 <dos1> but it's customized
Oct 01 22:22:04 <dos1> I think I'll integrate some of the customizations I wouldn't otherwise, cause they don't fit in software model we were doing
Oct 01 22:22:23 <dos1> but they have one advantage - they work
Oct 01 22:22:48 <dos1> for instance, opimd-notifier should be integrated into libphone-ui infra
Oct 01 22:23:09 <dos1> but it's not - it's a standalone, optional python app
Oct 01 22:23:57 <DocScrutinizer05> dos1: freemangordon just asked for help from GTA04 kernel hackers
Oct 01 22:25:51 <DocScrutinizer05> dos1: I had to confess I don't know what's that
Oct 01 22:26:34 <dos1> DocScrutinizer05: opimd-notifier?
Oct 01 22:26:35 <DocScrutinizer05> s/what's/who's/
Oct 01 22:26:35 <apt> DocScrutinizer05 meant: dos1: I had to confess I don't know who's that
Oct 01 22:26:45 <dos1> hehe :)
Oct 01 22:26:49 <DocScrutinizer05> damn, my fingers type words I not even thought
Oct 01 22:27:23 <dos1> DocScrutinizer05: http://lists.goldelico.com/pipermail/gta04-owner/2013-September/005022.html
Oct 01 22:27:44 <DocScrutinizer05> but anyway such joint-venture aka help would be a first step of making friends between both communities
Oct 01 22:28:32 <DocScrutinizer05> mhm
Oct 01 22:28:43 <DocScrutinizer05> I'll forward that link
Oct 01 22:29:21 <DocScrutinizer05> I guess FMG knows better how to check commit sign-offs etc than I do, to find out who's those kernel hackers he wants to talk to
Oct 01 22:29:49 <dos1> I think gta04-owner maillist may be useful for him
Oct 01 22:30:06 <dos1> kernel topics are often discussed there
Oct 01 22:31:06 <dos1> Nikolaus might also be the right person to ask, he seems to be taking some care of the kernel by himself
**** ENDING LOGGING AT Wed Oct 02 02:59:58 2013