**** BEGIN LOGGING AT Tue May 12 02:59:59 2015
May 12 11:31:31 <lumag> ant_work, hello.
May 12 11:31:47 * lumag is back from tons of public holidays here. So back to patches.
May 12 11:32:16 <lumag> hello bluelightning Jay7
May 12 12:27:43 <bluelightning> hi lumag
May 12 12:27:47 <bluelightning> lumag: saw those, thx
May 12 14:45:10 <ant_work> hello lumag, bluelightning
May 12 14:45:30 * ant_work is still wasting money/time repairing old vintage hi-fi
May 12 16:19:41 <ant_work> bbl
May 12 20:49:03 <ant_home> lumag, should I just jump in the kernel discussion and ask obvious things?
May 12 20:49:29 <ant_home> i.e. why no SysRQ for locomokbd while collie is the unique user?
May 12 20:49:42 <ant_home> (and poodleofc)
May 12 20:50:17 <ant_home> the kernel keymap is unsatisfactory
May 12 21:03:17 <lumag> ant_home, try to :)
May 12 21:04:11 <lumag> I mean if you have arguments other than mine, please do so.
May 12 21:04:46 <ant_home> pls correct me: only collie and poodle are using locomokbd
May 12 21:04:55 <lumag> yep
May 12 21:05:10 <ant_home> who/why decided that one was the right keymap back then (2.4/2.6)?
May 12 21:05:49 <ant_home> the old locomokbdtweak is around since 10yrs
May 12 21:06:02 <lumag> 1) It supported only poodle.
May 12 21:06:12 <ant_home> now that you are rewriting the driver I'd rather improve the keymap
May 12 21:06:17 <lumag> 2) It correctly maps all keys to corresponding keys.
May 12 21:06:31 <lumag> Remapping them to Alt and SysRQ are not an improvements, but rather a hack.
May 12 21:06:51 <ant_home> hm.. corgi has SysRQ iirc
May 12 21:07:10 <ant_home> or is it done in userspace..checking
May 12 21:07:43 <ant_home> my point i these devices have a fixed kb (yes, bluetooth..blah)
May 12 21:07:48 <lumag> In kernel space.
May 12 21:08:46 <lumag> That's actually a problem. Ideally (note this fine word) we should have few more entries in include/uapi/linux/input.h (for example, it misses the Contacts key IIRC).
May 12 21:08:54 <lumag> And then the userspace should cope with that.
May 12 21:08:58 <ant_home> http://lxr.free-electrons.com/ident?i=KEY_SYSRQ
May 12 21:09:37 <ant_home> just spitz and corgi
May 12 21:09:37 <lumag> So that all keys return proper input events (instead of cooked F18-like things).
May 12 21:09:58 <lumag> However opie will break -- it can not handle 'extended' keys.
May 12 21:10:00 <ant_home> ideally
May 12 21:10:02 <ant_home> :)
May 12 21:10:25 <bluelightning> opie can always be fixed ;)
May 12 21:11:03 <lumag> For tosa I ended up with a Kconfig entry that switches between 'compatible' and 'extended' keys.
May 12 21:11:29 <lumag> It's not just opie. The whole qte will need to be changed. I don't think bluelightning will bless such thing.
May 12 21:11:46 <lumag> qt4e doesn't have such problem.
May 12 21:12:07 <lumag> (IIRC)
May 12 21:12:14 <lumag> Neither does xorg
May 12 21:13:04 <ant_home> 637 #define KEY_ADDRESSBOOK         0x1ad   /* AL Contacts/Address Book */
May 12 21:13:10 <ant_home> ^input.h
May 12 21:14:11 <bluelightning> lumag: hmm...
May 12 21:20:11 <lumag> bluelightning, ?
May 12 21:42:00 <bluelightning> lumag: that was in reply to you suggesting we'd need substantial changes to qte
May 12 21:44:26 <lumag> Yeah :)
May 12 21:45:07 <lumag> ant_home, Yeah. Unfortunately that is > 0xff (or even 0x80 - I don't remember ATM), so it won't work with plain console IIRC.
May 12 21:45:28 <lumag> And qte uses console to get input events.
May 12 21:46:12 <ant_home> aboutthat..touchscreen is prolly still broken on collie :/
May 12 21:46:28 <ant_home> I didn't even test anymore...
May 12 21:46:44 <lumag> That is in queue.
May 12 21:47:02 <ant_home> poodle's one is/was surpsisingly ok
May 12 21:47:17 * ant_home switching lights on
May 12 21:47:18 * lumag wants to finish with locomo & sa1100 irqs before going further
May 12 21:47:29 <ant_home> darn thinkpad black kbd
May 12 21:47:31 <lumag> poodle uses completely different chip to handle ts
May 12 21:48:09 <ant_home> yes but iirc there is an issue with qt rotation..both have rotated fb
May 12 21:48:31 <ant_home> but maybe the problem is in the uc*** driver
May 12 21:50:31 <lumag> Hmm.
May 12 21:51:04 <lumag> Especially mentioning that collie code is handled by a special case in nearly all functions in the touchscreen part of that driver.
May 12 21:51:58 <ant_home> I have some old logs and iirc the readings were off
May 12 21:52:08 <ant_home> I was testing it with anarsoul
May 12 21:52:33 <lumag> I don't want to think about that ATM.
May 12 21:52:38 <ant_home> he got his ipaq working but iirc with opie extra patch
May 12 21:53:21 <ant_home> yes, the ucb driver has beed deeply changed..must be the culprit:)
May 12 21:53:44 <ant_home> kep it in queue
May 12 21:53:56 <ant_home> argh...I need an old AT keyboard :p
May 12 21:54:16 <lumag> Today I was hit by another 'feature' . The backlight will no go off on first fb0 blank.
May 12 21:54:30 <lumag> It will go on second and subsequent blanks.
May 12 21:54:39 <lumag> Fine Freescale code
May 12 21:54:52 <lumag> a55944ca82d287ca099ca90413af857af9086773
May 12 21:55:07 <ant_home> hmm..maybe I've seen that on gcw0..smthg similar
May 12 21:55:20 <lumag> Yep, that is a generic bug.
May 12 21:55:29 <ant_home> ah :)
May 12 21:56:00 <lumag> colleagues, gtg now. It's already 1:00 a.m and I'd like to eat something.
May 12 21:56:16 <ant_home> eek
May 12 21:56:20 <lumag> (next todo: review SPI patch review, fix shortcomings in keyboard driver).
May 12 21:56:25 <lumag> and resumbit the serie.
May 12 21:56:45 <lumag> Oh, And wait for response from MFD maintainer.
May 12 21:56:51 <lumag> bye!
May 12 21:56:56 <ant_home> just let me knowif/how should I break-in with dumb msgs on lakml
May 12 21:56:58 <ant_home> bye
May 12 21:57:20 <lumag> I don't think that makes sense. Let's keep that as a separate patch in meta-hh.
May 12 21:57:44 <ant_home> k
**** ENDING LOGGING AT Wed May 13 02:59:58 2015