| 00:08.25 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 00:13.46 | ali1234 | linus is talking about ARM at linuxcon right now |
| 00:13.59 | ali1234 | this is very relevant to anyone working on kernels here :) |
| 00:14.10 | ali1234 | http://events.linuxfoundation.org/events/linuxcon/live-video-streaming (needs signup) |
| 00:15.46 | ali1234 | "i accidentally merged three drivers for the same hardware and i didn't realise" - greg kh |
| 00:16.17 | ali1234 | here's the no-signup link : http://www.ustream.tv/channel/linux-foundation-live |
| 00:24.33 | Cass | hey .. nice one .. cheers |
| 00:25.09 | Cass | linus sounds very american now |
| 00:25.20 | Cass | affected accent .. much ! |
| 00:25.24 | arrrghhh | heh |
| 00:34.25 | Cass | heh slate gnome3 .. way to go |
| 00:41.21 | arrrghhh | lol he just runs out |
| 00:41.30 | Cass | yeah beer on ice |
| 00:41.53 | arrrghhh | that was spiffy, thx ali1234 |
| 00:42.00 | Cass | yeah man cheers .. |
| 00:42.07 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-jdshcdqirestffec) |
| 00:52.21 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 00:55.20 | *** join/#htc-linux WyrM (~WyrM@unaffiliated/wyrm) |
| 01:14.22 | *** join/#htc-linux hardwalker (~hardwalke@122-117-115-146.HINET-IP.hinet.net) |
| 01:36.33 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 02:03.57 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 02:10.56 | *** join/#htc-linux dobrin (~dobrin@85.91.150.26) |
| 02:15.06 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 02:20.34 | *** join/#htc-linux programmer8922 (~Evan@97-119-137-120.albq.qwest.net) |
| 02:22.11 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 02:41.27 | *** join/#htc-linux leviathan (~quassel@2001:470:26:484:6ef0:49ff:fee6:8dca) |
| 02:42.36 | *** join/#htc-linux raymonddull (45f57d6f@gateway/web/freenode/ip.69.245.125.111) |
| 03:05.49 | *** join/#htc-linux rajkosto (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 03:15.08 | *** join/#htc-linux mitsutaka (~mitsutaka@pw126226162224.32.tss.panda-world.ne.jp) |
| 03:24.43 | *** join/#htc-linux raymonddull_ (45f57d6f@gateway/web/freenode/ip.69.245.125.111) |
| 03:25.02 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 03:30.20 | *** join/#htc-linux Rob2223 (~Miranda@p4FFF25FB.dip.t-dialin.net) |
| 03:50.38 | *** join/#htc-linux raymonddull (45f57d6f@gateway/web/freenode/ip.69.245.125.111) |
| 03:51.38 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 04:34.08 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@94.52.236.39) |
| 05:02.40 | *** join/#htc-linux Berger- (~noreply@90.149.127.254) |
| 05:21.23 | *** join/#htc-linux mitsutaka (~mitsutaka@rt.miraclelinux.com) |
| 05:25.24 | *** join/#htc-linux avinashhm (~avinash-h@192.91.66.189) |
| 06:31.53 | *** join/#htc-linux NeoMatrixJR (~NeoMatrix@173-20-63-62.client.mchsi.com) |
| 06:31.54 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@94.52.236.39) |
| 06:45.03 | *** join/#htc-linux Rajko (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 07:13.21 | *** join/#htc-linux turbolapin (~quassel@188.4.197.98.dsl.dyn.forthnet.gr) |
| 07:25.05 | *** join/#htc-linux dekar_ (~dekar@95.143.165.210) |
| 07:26.26 | *** join/#htc-linux Cotulla (~w3489sdf@nat100-255-205-109.tvoe.tv) |
| 07:40.36 | *** join/#htc-linux arene (b40f3caf@gateway/web/freenode/ip.180.15.60.175) |
| 07:46.55 | *** join/#htc-linux Rajko (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 07:47.48 | *** join/#htc-linux Rajko (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 07:48.18 | *** join/#htc-linux bitrot (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 08:00.28 | *** join/#htc-linux rajkosto (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 08:02.35 | *** join/#htc-linux NeoMatrixJR2 (~NeoMatrix@173-20-63-62.client.mchsi.com) |
| 08:19.42 | *** join/#htc-linux paulk (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 08:21.31 | *** join/#htc-linux Segnale007 (~Segnale00@ppp-241-32.26-151.libero.it) |
| 08:22.20 | *** join/#htc-linux gauner1986 (~Miranda@ip-178-202-172-84.unitymediagroup.de) |
| 08:26.39 | *** join/#htc-linux rzk_ (rzk@95-25-219-22.broadband.corbina.ru) |
| 08:47.14 | *** join/#htc-linux arrrghhh (~arrrghhh@unaffiliated/arrrghhh) |
| 10:21.46 | *** join/#htc-linux Segnale007 (~Segnale00@ppp-241-32.26-151.libero.it) |
| 10:31.16 | *** join/#htc-linux Berger- (~noreply@90.149.127.254) |
| 11:02.44 | *** join/#htc-linux helicopter88 (~helicopte@host94-112-dynamic.46-79-r.retail.telecomitalia.it) |
| 11:03.21 | *** join/#htc-linux CptAJ (CptAJ@186.90.219.5) |
| 11:06.36 | *** join/#htc-linux avinashhm (~avinash-h@192.94.92.11) |
| 11:08.41 | *** join/#htc-linux MethoS- (~clemens@134.102.106.250) |
| 11:16.21 | *** join/#htc-linux Segnale007 (~Segnale00@ppp-241-32.26-151.libero.it) |
| 11:28.55 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@94.52.236.39) |
| 11:33.18 | *** join/#htc-linux turbolapin (~quassel@188.4.197.98) |
| 11:43.32 | *** join/#htc-linux turbolapin_ (~quassel@178.128.217.22) |
| 12:22.26 | *** join/#htc-linux turbolapin (~quassel@178.128.217.22) |
| 12:30.35 | *** join/#htc-linux avinashhm (~avinash-h@192.91.66.186) |
| 12:47.38 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 13:16.12 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-164-254.netcologne.de) |
| 13:18.17 | *** join/#htc-linux programmer8922 (~Evan@97-119-137-120.albq.qwest.net) |
| 13:26.11 | *** join/#htc-linux mitsutaka (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
| 13:31.37 | *** join/#htc-linux fpu (~nonperson@178-36-146-86.adsl.inetia.pl) |
| 13:32.09 | fpu | hey |
| 13:47.51 | Cotulla | hi |
| 13:50.05 | Cotulla | who wanted to know |
| 13:50.14 | Cotulla | yes LEO can support more than 1 USB host device |
| 13:51.18 | fpu | Cotulla: what is the meaning of it if there's only one usb port? |
| 13:51.27 | arrrghhh | lol |
| 13:51.30 | arrrghhh | never heard of a hub? |
| 13:51.31 | arrrghhh | :P |
| 13:51.32 | fpu | that you can theoretically solder on more sockets? |
| 13:51.36 | fpu | oh hub! |
| 13:51.40 | Cotulla | yeah |
| 13:51.43 | arrrghhh | solder on more sockets bwhahahaha |
| 13:51.44 | Cotulla | hub |
| 13:51.46 | arrrghhh | fpu, i want to see that |
| 13:51.49 | fpu | silly |
| 13:51.53 | Cotulla | there 5 USBH ports btw |
| 13:51.55 | fpu | hehe |
| 13:51.55 | Cotulla | in CPU |
| 13:52.04 | Cotulla | 2 HS and 3 FS |
| 13:52.19 | fpu | I never thought there must be some support for usb hubs, always thought they are just regular devices |
| 13:53.01 | Cotulla | lol |
| 13:53.11 | Cotulla | I connected one leo to another |
| 13:53.30 | arrrghhh | i connected 100 together and made a movie |
| 13:53.32 | Cotulla | and see SD card in MAGLDR USB MS |
| 13:53.44 | Cotulla | :D |
| 13:53.48 | arrrghhh | wait, that was PS2's... nvm :P |
| 13:54.40 | Cotulla | what else I can connect .. . |
| 13:55.19 | Cotulla | keybaord |
| 13:55.23 | fpu | keyboard, aye |
| 13:55.31 | fpu | Cotulla: what kernel? |
| 13:55.34 | fpu | linux_on_wince_htc? |
| 13:55.40 | Cotulla | no... |
| 13:56.06 | Cotulla | keybaord works |
| 13:56.13 | fpu | linuxtogo? |
| 13:56.41 | Cotulla | no. it's my WM65 driver ;) |
| 13:56.52 | fpu | aa ok :-) |
| 13:57.06 | Cotulla | maybe add host support in magldr |
| 14:05.57 | fpu | hcitool scan gives empty output on android :-( |
| 14:07.19 | fpu | uhh nevermind, me silly again |
| 14:20.14 | Cotulla | ? |
| 14:22.07 | *** join/#htc-linux gauner1986 (~Miranda@ip-109-91-117-8.unitymediagroup.de) |
| 14:28.58 | fpu | well, still no luck |
| 14:29.12 | fpu | maybe any of you has any idea about bluetooth dun on android? android phone being client |
| 14:32.29 | *** join/#htc-linux rajkosto (~rajkosto@cable-188-2-156-181.dynamic.sbb.rs) |
| 14:52.33 | *** join/#htc-linux ray|yar (~raymonddu@c-69-245-125-111.hsd1.mi.comcast.net) |
| 14:52.33 | *** join/#htc-linux raymonddull (~raymonddu@c-69-245-125-111.hsd1.mi.comcast.net) |
| 14:56.12 | *** join/#htc-linux Markinus (~Adium@212.255.47.70) |
| 14:58.50 | *** join/#htc-linux emwe (~emwe@cable-86-56-10-158.cust.telecolumbus.net) |
| 15:24.33 | *** join/#htc-linux [acl] (~abel@96.246.167.90) |
| 15:25.06 | *** join/#htc-linux rob_w (~bob@ppp-88-217-79-73.dynamic.mnet-online.de) |
| 15:25.08 | *** join/#htc-linux toastcfh (~toastcfh@41-32.200-68.tampabay.res.rr.com) |
| 15:28.17 | [acl] | emwe: bro quick question |
| 15:29.01 | [acl] | emwe: is the liblights on the current rootfs try to access the sysfs for the autobl ? or was that only gingerbread ? |
| 15:29.10 | Cotulla | sup dupper |
| 15:29.49 | *** join/#htc-linux avinashhm (~avinash-h@192.91.66.189) |
| 15:35.17 | [acl] | Cotulla: hello my good man |
| 15:36.19 | Cotulla | hey |
| 15:36.30 | Cotulla | what's up? |
| 15:36.39 | [acl] | Cotulla: tired.. very very tired |
| 15:36.46 | Cotulla | me too |
| 15:37.01 | Cotulla | hate java coding |
| 15:37.12 | gauner1986 | java is cool |
| 15:37.13 | gauner1986 | :P |
| 15:37.20 | Cotulla | gsodder with us :) |
| 16:03.52 | *** join/#htc-linux Alex[sp3dev] (~alexander@86.110.163.19) |
| 16:04.43 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 16:36.18 | *** join/#htc-linux GlemSom (~glemsom@0x5da34bca.cpe.ge-1-1-0-1105.sdnqu1.customer.tele.dk) |
| 16:58.07 | *** join/#htc-linux turbolapin_ (~quassel@46.12.134.243.dsl.dyn.forthnet.gr) |
| 17:00.22 | *** join/#htc-linux gauner1986 (~Miranda@ip-109-91-117-8.unitymediagroup.de) |
| 17:08.26 | *** join/#htc-linux turbolapin (~quassel@46.12.134.243.dsl.dyn.forthnet.gr) |
| 17:08.31 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@94.52.236.39) |
| 17:09.41 | *** join/#htc-linux CptAJ2 (CptAJ@186.90.219.5) |
| 17:10.55 | *** join/#htc-linux D3tul3 (~Oliver@pool-96-234-135-148.bltmmd.east.verizon.net) |
| 17:19.39 | *** join/#htc-linux dekar_ (~dekar@drms-590c45bc.pool.mediaWays.net) |
| 18:05.59 | *** join/#htc-linux ali1234 (~ajbuxton@robotfuzz.co.uk) |
| 18:21.34 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 18:24.13 | *** join/#htc-linux |Jeroen| (~jeroen@d5152B25B.access.telenet.be) |
| 18:30.54 | *** join/#htc-linux helicopter88 (~helicopte@host94-112-dynamic.46-79-r.retail.telecomitalia.it) |
| 18:34.04 | *** join/#htc-linux MN-- (~yaaic@host86-134-36-116.range86-134.btcentralplus.com) |
| 18:43.25 | *** join/#htc-linux Segnale007 (~Segnale00@ppp-241-32.26-151.libero.it) |
| 18:49.03 | *** join/#htc-linux GlemSom (~glemsom@0x5da34bca.cpe.ge-1-1-0-1105.sdnqu1.customer.tele.dk) |
| 18:54.34 | *** join/#htc-linux toastcfh (~toastcfh@41-32.200-68.tampabay.res.rr.com) |
| 19:04.31 | *** join/#htc-linux onen|openBmap (~quassel@vbo91-5-88-186-17-64.fbx.proxad.net) |
| 19:10.55 | *** join/#htc-linux fpu (~nonperson@178-36-3-163.adsl.inetia.pl) |
| 19:11.19 | fpu | good evening |
| 19:14.54 | MN-- | evening fpu |
| 19:15.31 | *** join/#htc-linux ali1234 (~ajbuxton@robotfuzz.co.uk) |
| 19:29.01 | emwe | Alex[sp3dev]: hey. i am in the midth of porting (copying) your dynamic amss/adsp setup. does it guarantee that all rpc devices are published prior to adsp probe chiming in? asking because in 2.6.35.14 it's good luck if rpc devices are registered in time or not during boot for adsp probe to trigger or not. |
| 19:30.22 | Cotulla | emwe, are u TPZ man? |
| 19:30.25 | Cotulla | hi |
| 19:30.38 | emwe | Cotulla: amongst others yes. topa was my first device. |
| 19:31.02 | Cotulla | TPZ has nand android already? |
| 19:31.10 | Alex[sp3dev] | emwe: yes, i have added a conditional variable and a wait event specially for that and during NAND testing have fixed the bug in the first implementation |
| 19:31.12 | emwe | no. nobody dared to yet. |
| 19:31.25 | Cotulla | are u working under TPZ? |
| 19:31.31 | emwe | Alex[sp3dev]: ah cool. then i am just going along. thanks! |
| 19:31.40 | Cotulla | hey Alexx |
| 19:31.45 | Alex[sp3dev] | hey Cotulla |
| 19:32.06 | emwe | Cotulla: it's become secondary device for testing. using the donated RHOD400 as main device. doing dev for both. but they are both quite similar anyway. |
| 19:32.25 | Cotulla | okay... |
| 19:32.33 | Alex[sp3dev] | emwe: how about migrating rhod to microp-ng? i think we've already fixed the keyboard, right? |
| 19:32.35 | emwe | does anybody of you know how the arm machine registry or rather its maintainer thinks about renaming machine types? |
| 19:32.53 | emwe | Alex[sp3dev]: ehm, it is. |
| 19:32.56 | Alex[sp3dev] | emwe: i think he doesn't feel good about unmainlined things anyway |
| 19:33.29 | emwe | Alex[sp3dev]: because some "dumbass" registered "whitestone" not "htcwhitestone" like all others of our machtypes are namend. |
| 19:33.39 | Cotulla | heh |
| 19:33.46 | Cotulla | I always hate this list honestly ;) |
| 19:33.56 | Alex[sp3dev] | emwe: does it really matter unless you want some ugly macros?) |
| 19:34.09 | Cotulla | and u always can do define |
| 19:34.16 | emwe | emwe: somehow i like things being in line with the other crap :) |
| 19:34.34 | emwe | lol. Alex[sp3dev] ^^ |
| 19:34.39 | Markinus | emwe_ hi |
| 19:34.43 | Markinus | here the list |
| 19:34.44 | Markinus | http://www.arm.linux.org.uk/developer/machines/index.php |
| 19:34.54 | Alex[sp3dev] | emwe: somehow i would prefer if we had a working port first.. and yeah.. i'm thinking of getting a tegra smartphone maybe.. or maybe will wait for tegra3 |
| 19:34.56 | emwe | oh a machines list. goodi |
| 19:35.15 | emwe | Alex[sp3dev]: i can still resist my inner wish for a faster device. :) |
| 19:35.31 | emwe | Markinus: don't you miss your old TOPA sometimes? haha |
| 19:35.46 | Cotulla | lol |
| 19:35.54 | Markinus | hehe, sometimes ;) But HD2 is a great devcie, there you can forott other stuff fast ;) |
| 19:35.59 | Markinus | forgott |
| 19:36.00 | Cotulla | :D |
| 19:36.15 | Alex[sp3dev] | emwe: yeah. honestly we've spent too much time reversing msm so i don't want to give up.. and we can easily get SHR running. the only problem is that msm is VERY slow |
| 19:36.21 | *** join/#htc-linux paulk (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 19:36.37 | Cotulla | ur MSM7k is slow yes |
| 19:36.44 | emwe | Markinus: ah heh. i thought that was a link to a machine mailling list i had overlooked. i know the list. dunno if it's worth registering there and if it's possible to rename mach type defs. |
| 19:36.47 | Cotulla | our LEO's much faster |
| 19:37.08 | Markinus | the mach types is comming from here |
| 19:37.19 | Markinus | its not here then you should add it |
| 19:37.24 | Markinus | *if |
| 19:37.59 | emwe | Markinus: it's already there but named inconsistently to all our 6 other machtypes. we have "htcfoo" not just "foo". |
| 19:38.18 | Alex[sp3dev] | Cotulla: huh. i didn't even mention newer MSMs. and honestly, despite tegra's lacking NEON, i'd go for it since it has a working GLES/OpenMAX drivers and a v4l2 driver and most stuff is already in mainline.. much less efforts than with msm.. |
| 19:38.40 | Cotulla | okay |
| 19:38.59 | emwe | Alex[sp3dev]: yeah, we spent a shitload of time already. wouldn't even know what to buy at this time. |
| 19:41.33 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 19:42.46 | emwe | Alex[sp3dev]: i haven't looked at this at all and am still a bit ignorant about the lower level stuff, but is it possible to obtain two i2c_client's on driver probe if that is going through microp-ng? |
| 19:47.35 | Alex[sp3dev] | emwe: no, because it is designed to handle one i2c device. why would you need that? if you need keyboard leds, i'd suggest using led triggers as done for kovsky. this way you can enable both i2c and gpio leds when a key is pressed |
| 19:48.20 | emwe | rhods buttons are ksc controlled, not klt. that's why i am asking. |
| 19:49.03 | emwe | i'd like to consolidate the whole led handling into one managing file anytime in the future when i'v gathered enough issues to know what it could be made like. |
| 19:49.09 | Alex[sp3dev] | and what prohibits you from writing two microp-ng clients for one i2c device? you pass an array of platform_device and each one gets i2c_client |
| 19:49.57 | emwe | it's rather one client for two i2c devices. |
| 19:50.20 | Alex[sp3dev] | okay, i see what you want to do. if you don't want to keep led code in the board file, just make a separate file and export the devices via a header so that the board can include them. |
| 19:51.12 | Alex[sp3dev] | emwe: a slight hint: you don't really need to have the single handler for all leds even if you want them to turn on simultaneosly. that's what triggers are for |
| 19:52.04 | emwe | Alex[sp3dev]: somehow the triggers didn't work for me with userland. they always got reset. at leat that's what i saw for the keyboard led device. |
| 19:52.17 | emwe | as soon as userland chimed in the trigger was [none] again |
| 19:53.15 | Alex[sp3dev] | you know what i always say.. "works for me". look into how it is implemented on kovsky. and yes, if the userland is broken, it must be fixed instead of breaking the kernel |
| 19:53.49 | emwe | well, the trigger is set until userland chimes in. anyway, that piece works just fine without triggers :) |
| 19:54.33 | emwe | no idea what resets the trigger. looked like as soon as the brightness hook was written to, it got reset. |
| 19:54.42 | emwe | don't remember in detail anymore. |
| 19:56.26 | Alex[sp3dev] | triggers are good because 1. you don't need to add hooks to the input (keypad driver) and 2. we can do it in the unified way for all devices. anyway you can try triggers right now with color leds, for example. |
| 19:56.33 | Alex[sp3dev] | i'll grep android source for triggers ;) |
| 19:57.09 | *** join/#htc-linux balans21 (~user@53543A69.cm-6-5a.dynamic.ziggo.nl) |
| 19:57.24 | emwe | i've found nothing in there. and if so, it should be only liblights. |
| 19:57.45 | Cotulla | Alex, u too carry about non important things :) |
| 19:58.06 | emwe | lol |
| 19:58.25 | Alex[sp3dev] | Cotulla: i care for the source code being easy to update and i don't care at all for 'user experience' and crap like that |
| 19:59.18 | Cotulla | easy to update - it's one code block with universal functions-thunks to interact with kernel and etc |
| 19:59.45 | Cotulla | today u using platform data, tomorrow they will be absolute part of linux kernel |
| 20:00.34 | Alex[sp3dev] | if you want user experience, you can go and buy any commercial device. and figure out that official sources are crap. vendors even hack gpio_keys and mmc core instead of using platform data. and that you have to rewrite everything to make it working with a newer version. there are very few native linux devices integrated upstream. i can't remember any except n900 and freerunner |
| 20:01.12 | Cotulla | yes they hack |
| 20:01.31 | Cotulla | because it's utopian idea to make universal code |
| 20:02.28 | Alex[sp3dev] | no it is not. it just takes much time. and whatever the excuse for bad code is (for example, "we're not paid for that") does not prohibit me to call it all crap. |
| 20:03.29 | Cotulla | anyway by doing such code it become large, it become hard to support it, then it will start self destructing |
| 20:05.36 | Cotulla | users of this code need to fill long structures with more than 20 different fields which extremely bad documented. then u need search reference for each field, analyze it. |
| 20:05.38 | Alex[sp3dev] | blah-blah-blah. there is no valid reason for the devices on the same SoC not to be able to have the unified zImage. btw, PXA devices can do that. if the SoC vendor integrates to mainline, things 'just work' (like on pxa, omap and tegra). |
| 20:06.09 | Cotulla | if all parameters like defines, code usually very simple and it's easy to understand basic ideas |
| 20:06.13 | Cotulla | and modify code for them |
| 20:06.37 | Cotulla | unified zImage? also utopian idea] |
| 20:06.46 | Cotulla | one device - one zImage - one source tree |
| 20:06.50 | Alex[sp3dev] | fuck you |
| 20:07.50 | Cotulla | lol |
| 20:08.00 | *** join/#htc-linux gauner1986 (~Miranda@tmo-097-162.customers.d1-online.com) |
| 20:08.03 | Cotulla | why I need copy x86 alpha and etc each time? |
| 20:08.09 | Cotulla | wtf I need them for arm? |
| 20:08.10 | D3tul3 | hm maybe it's not worth adding work queues to klt if you guys are on your way to bringing ng to the masses |
| 20:08.15 | Alex[sp3dev] | seriously, the why keep different trees? the whole purpose of open-source is to keep code centralized so that you can commit to upstream once and then it is fixed by other people (btw, that's the reason corporations try to integrate Hyper-V and stuff) |
| 20:08.55 | Cotulla | upstream ? but there are we and where upstream ? |
| 20:08.57 | Alex[sp3dev] | Cotulla: because you're not the only person using linux. and it is easier to keep all stuff in one place. it is not compiled into kernel, so why would you care? |
| 20:09.13 | Cotulla | why easy to keep unused stuffs? |
| 20:09.24 | Cotulla | it's easy to split them |
| 20:09.31 | ali1234 | Cotulla: you should make a fork, and delete all other arch then |
| 20:09.47 | Cotulla | remove unused data to save storage and increase search speed |
| 20:09.53 | ali1234 | nobody is going to try to stop you |
| 20:09.54 | Alex[sp3dev] | ali1234: it's because some people still do not use git and are copying each file manually |
| 20:10.21 | Cotulla | I am sure it's like legend - somebody said that it must be so and all become to believe to this ;) |
| 20:10.48 | ali1234 | nope |
| 20:10.51 | Alex[sp3dev] | Cotulla: no, it is just previous experience with pxa270 where everything was perfect |
| 20:11.08 | Cotulla | u had one device there to develop, right? |
| 20:11.20 | ali1234 | it's evolution where all the people that tried to use linux the way you want to do it failed hard |
| 20:11.34 | Alex[sp3dev] | i still want to integrate p525 and loox upstream but i cannot find time to do that working on msm |
| 20:11.47 | Cotulla | u sure they will grab them? |
| 20:12.07 | ali1234 | it's not hard to get code in mainline |
| 20:12.12 | Alex[sp3dev] | Cotulla: yes i am |
| 20:12.17 | ali1234 | as long as you test it against a recent kernel |
| 20:12.26 | ali1234 | not some random tree from 18 months ago |
| 20:12.34 | Cotulla | yeah |
| 20:13.12 | Cotulla | wait, but how they will update it? |
| 20:13.17 | Alex[sp3dev] | anyway. bottom line is that we have problems only because qualcomm have not integrated their code. and you all know why they did not |
| 20:13.21 | Cotulla | if nobody have there p525? |
| 20:13.33 | ali1234 | update it? |
| 20:13.37 | ali1234 | why would they update it? |
| 20:13.42 | Cotulla | update and test |
| 20:13.47 | ali1234 | why would anyone even touch it? |
| 20:13.50 | Cotulla | if it will be included into mainline |
| 20:13.54 | Alex[sp3dev] | Cotulla: in most cases it is enough to verify it compiles when someone makes changes. then if some user reports a regression, he's welcome to send patches |
| 20:13.54 | Cotulla | by Alex |
| 20:14.06 | Cotulla | yeah, I know this "enough to verify" |
| 20:14.11 | Cotulla | so it's same like it not there |
| 20:14.15 | ali1234 | this is why we have -rc kernels |
| 20:14.21 | ali1234 | if someone broke your driver, report it |
| 20:14.28 | ali1234 | patch won't be accepted until it is fixed |
| 20:14.30 | Alex[sp3dev] | cmon, you're given everything for free and under GPL, why would you want more? |
| 20:15.28 | Cotulla | as well pdata is very awful to search references to it |
| 20:15.30 | Cotulla | :( |
| 20:16.19 | Alex[sp3dev] | use kdevelop if you want to find it like in eclipse or VS. does anyone prohibit you? |
| 20:16.34 | Cotulla | are u using? |
| 20:16.46 | Alex[sp3dev] | i'm using vim and clang for that |
| 20:16.53 | Cotulla | and eclipse and vs never can find anything complex |
| 20:16.57 | Cotulla | it's nature of C/C++ |
| 20:19.19 | Alex[sp3dev] | so the main problem with android devices is that android kernel is too hacked. when mainline refused to accept pmem and wakelocks, instead of discussing possible solutions the google developers deciced to make a fork. and it is quite sad to see that, for example, for tegra, they're writing everything twice - once for android, once for mainline and chromium os. and that's the morale. if you don't follow the rules, you have to do twice the effort |
| 20:19.55 | Cotulla | why they setup rules? |
| 20:20.11 | Cotulla | and it's GPL it means that google can change it how they want |
| 20:20.36 | Cotulla | so call it "hacked" also wrong - it just own Google's version ;) |
| 20:20.48 | ali1234 | i agree there, and so does linus |
| 20:20.58 | ali1234 | you should really watch the talk he did yesterday at linuxcon |
| 20:20.59 | Alex[sp3dev] | by 'hacked' i mean ugly solutions, not forking itself |
| 20:21.10 | ali1234 | he addresses directly all this shit |
| 20:21.13 | Cotulla | ugly for u |
| 20:21.43 | Alex[sp3dev] | Cotulla: ok, we've wasted like half an hour and it is like the tenth time we talk about that. stop trolling me, please |
| 20:21.51 | Cotulla | I am not trolling |
| 20:22.13 | Cotulla | and it's not for to decide if solutions are ugly or not |
| 20:22.31 | Cotulla | Google had some target and they implement stuffs for it |
| 20:23.20 | Cotulla | also possible that goolge interfaces are more simple and more suitable for slow cpus |
| 20:24.43 | Alex[sp3dev] | yeah, inventing a custom libc with java on top is a suitable solution. why would they need java when games are still compiled to native code? they're just deliberately breaking ABI and using obfuscated drivers. and also hardcode for specific hardware instead of writing generic code. that sucks. everything else is ok. |
| 20:26.44 | Cotulla | java - good for UI development. they even have plugin for Eclipse. of course it complete sucks if we compare with WP7 tools. |
| 20:27.06 | Cotulla | Java just more easy. |
| 20:27.15 | ali1234 | everything sucks compared to the WP7 tools |
| 20:27.18 | Cotulla | yes |
| 20:27.29 | Cotulla | fast emulator |
| 20:27.33 | Cotulla | good UI designer |
| 20:27.34 | Alex[sp3dev] | ali1234: lemme guess. because only WP7 tools were written by M$ |
| 20:27.35 | Cotulla | and etc |
| 20:27.40 | Cotulla | yeah |
| 20:27.42 | Cotulla | and not only |
| 20:27.45 | Cotulla | compiler and etc |
| 20:27.48 | ali1234 | Alex[sp3dev]: no, just because they are actually good |
| 20:27.49 | Cotulla | google just stole a lot of |
| 20:27.51 | Cotulla | free software |
| 20:27.52 | Cotulla | :D |
| 20:27.57 | Cotulla | stolen Java |
| 20:27.58 | Alex[sp3dev] | Cotulla: oh, finally |
| 20:28.01 | Cotulla | reuse free GCC |
| 20:28.06 | Cotulla | reuse free Linux kernel |
| 20:28.08 | Cotulla | andetc |
| 20:28.20 | Cotulla | even can't make something normal from it |
| 20:28.35 | *** join/#htc-linux turbolapin (~quassel@79.103.89.29) |
| 20:28.35 | ali1234 | Cotulla: have you ever tried developing for blackberry? that is the worst... |
| 20:28.35 | Alex[sp3dev] | really, that's a problem - they _reuse_ everything but break ABI and API and don't commit to upstream. in other words, raping the community |
| 20:28.46 | ali1234 | lol |
| 20:28.46 | Cotulla | ali1234, no |
| 20:29.10 | Cotulla | why u think that upstream is something great? for me it's some kind of people who wants to control other how others should code... |
| 20:29.23 | ali1234 | they don't want that at all |
| 20:29.29 | ali1234 | linus doewsn't care if you fork the kernel |
| 20:29.46 | ali1234 | he doesn't care if google does it either |
| 20:29.48 | Cotulla | yeah, but most people like Alex, will call my kernel shit. it's hidden thing. |
| 20:29.57 | ali1234 | all this "google is raping the community" stuff is BS |
| 20:30.14 | ali1234 | i only care because you are just making it harder on yourself :) |
| 20:30.21 | Cotulla | hidden thing that only "upstream is good. other - sucks." |
| 20:30.24 | Cotulla | no? |
| 20:30.55 | Alex[sp3dev] | if there were no upstream, then we could not have such stuff as Mesa, Firefox, Xorg and even libc. because there is upstream linux kernel, you have the same ip stack, i2c drivers, alsa and stuff for *all* platforms and you don't have to rewrite half the kernel to make it work on a new hardware |
| 20:31.12 | Cotulla | right |
| 20:31.24 | Cotulla | but I think better restrict via license |
| 20:31.32 | Cotulla | ability to make a lot of useless clones |
| 20:32.02 | Cotulla | like Bolgen OS |
| 20:32.06 | Cotulla | :D :D |
| 20:32.42 | ali1234 | Cotulla: if your model works so much better, why isn't *BSD used more for smartphones? |
| 20:32.44 | Alex[sp3dev] | well.. but then again, you cannot call it 'free' software if it is restricted. moreover, if it was restricted, corporations would not use linux and it would be nowhere. somewhere even farther than BSD systems |
| 20:33.12 | Cotulla | anyway we forgot that most things are done for money :) so all our talks are abstract] |
| 20:33.30 | Cotulla | Google wanted a lot of apps - they choose Java and not compatible libc |
| 20:34.27 | Cotulla | I am worry about developing |
| 20:34.37 | Cotulla | that in open source a lot of project clones |
| 20:34.47 | Cotulla | which never finished |
| 20:35.16 | ali1234 | looks at a htc wizard... |
| 20:35.21 | Cotulla | ali1234, it's used. Darwin ^^ |
| 20:36.12 | Alex[sp3dev] | most projects dedicated to porting software to phones cannot be finished because hardware goes out of use in like 2 years and most projects take 3-4 years unless all hardware specs are available |
| 20:36.28 | Cotulla | slide view to leo512, gold, leo1024 and to bee :o |
| 20:37.16 | Cotulla | Alex[sp3dev], or maybe u too much think about useless things like how to pass data to driver? pdata and etc. just used global variable and job done. nobody will touch it after u, 90%. |
| 20:37.17 | Alex[sp3dev] | that's not quite fair. leo is very close to nexus and most stuff like camera 'just worked'. besides, much stuff like clocks was reversed for msm7200A |
| 20:38.18 | Alex[sp3dev] | Cotulla: i doubt your theory is good. when i know how hardware is supposed to work, it takes less than a day to write the driver. when we don't know how it works, it can take months just to figure out |
| 20:38.59 | ali1234 | knowing how it works spoils the fun |
| 20:39.10 | Cotulla | sometimes u need look at this as black box |
| 20:39.59 | Cotulla | deep looking good ofcourse |
| 20:40.05 | Cotulla | I like it ;) |
| 20:40.25 | Alex[sp3dev] | dunno. honesly there's not much fun in writing drivers because most of them are very identical. there are a lot of more interesting stuffs like opengl, dsp and video. but to play with them you need a good working kernel first |
| 20:40.52 | Cotulla | yeah |
| 20:41.09 | Alex[sp3dev] | okay, enough of that. that was a microsoft and qualcomm sponsored provocation to distract me from writing camera driver |
| 20:41.10 | Cotulla | but u spend time to thinking "how cool to pass parameters to driver" |
| 20:41.31 | Cotulla | I won't steal ur time tomorrow |
| 20:43.13 | Cotulla | SDCC <3 too |
| 20:43.14 | Cotulla | :P |
| 20:50.54 | Alex[sp3dev] | good news is that in future we'll only have to write the drivers for the peripherals and board files will mostly go away. FDT will be used like it is used for tegra and i heard that for tegra new boards are accepted only in fdt-compatible format |
| 20:51.11 | Cotulla | what is DFT? |
| 20:51.13 | Cotulla | lol |
| 20:51.15 | Cotulla | FDT |
| 20:51.21 | Cotulla | is it something like UEFI? |
| 20:51.24 | Alex[sp3dev] | DFT is a bunch of geeks ;) |
| 20:51.29 | Cotulla | :D |
| 20:52.17 | Alex[sp3dev] | FDT is a declarative format for describing driver settings (i.e., pdata). you just write one file, then u-boot and linux parse it and you don't have to write board file. i'll find an example for you |
| 20:52.37 | Cotulla | some kind of meta data? |
| 20:52.43 | Alex[sp3dev] | yes |
| 20:52.45 | Cotulla | OMG |
| 20:52.57 | Alex[sp3dev] | well, it is actually compiled to binaries |
| 20:53.01 | Cotulla | megabytes of XML parsers |
| 20:53.04 | Cotulla | oh well |
| 20:53.06 | Cotulla | binaries |
| 20:53.07 | Alex[sp3dev] | not XML ;) |
| 20:53.16 | Cotulla | binaries not bad |
| 20:53.20 | Cotulla | but u need tools for them and etc |
| 20:53.24 | Cotulla | config build process right |
| 20:53.38 | *** join/#htc-linux MN-- (~yaaic@host86-134-36-116.range86-134.btcentralplus.com) |
| 20:54.11 | Alex[sp3dev] | like this http://git.chromium.org/gitweb/?p=chromiumos/third_party/u-boot.git;a=blob;f=board/nvidia/seaboard/tegra2-seaboard.dts;h=c1b93d2319cda8b7eab5082981e75c645af51215;hb=chromeos-v2011.03 |
| 20:55.35 | Alex[sp3dev] | actually this is not a new idea. it was used on PowerPC computers for ages but was ported to arm recently |
| 20:56.31 | Cotulla | NVRAM? |
| 20:56.47 | Alex[sp3dev] | what about nvram? |
| 20:56.59 | Cotulla | is it used |
| 20:57.04 | Cotulla | for settings |
| 20:57.05 | Cotulla | on ppc? |
| 20:57.43 | Alex[sp3dev] | dunno.. afaik, ppc uses openfirmware interface and all low-level stuff is done by the firmware in the motherboard |
| 20:58.04 | *** join/#htc-linux gauner1986 (~gauner198@ip-109-91-117-8.unitymediagroup.de) |
| 21:02.46 | Alex[sp3dev] | btw, on recent kernels (around 35 or later) patches for runtime detection of zreladdr and phys_offset were introduced. that, for example, allows to boot the single kernel on PXA with ram at 0xa0000000 and at 0xa8000000. I think that with accurate coding it may be likely that when new SoCs are integrated, it will be possible to support multiple CPUs in one kernel. or at least different variations of the same one via FDT. if some corporation decid |
| 21:02.46 | Alex[sp3dev] | es to ship arm laptops with linux preinstalled, that will happen for sure. you know, money invested into development decides how stuff is done in the end |
| 21:03.07 | Cotulla | is it still |
| 21:03.09 | Cotulla | 2MB align? |
| 21:03.10 | Cotulla | :P |
| 21:04.03 | *** join/#htc-linux eugenb (~Adium@dslb-188-100-030-044.pools.arcor-ip.net) |
| 21:04.06 | Alex[sp3dev] | dunno. wasn't that a msm-specific thingie? |
| 21:04.40 | Cotulla | no |
| 21:04.42 | Cotulla | generic |
| 21:06.30 | Alex[sp3dev] | i didn't look into it. maybe that was done to group pages in larger blocks to make accessing them faster or boot code simpler. |
| 21:06.31 | *** part/#htc-linux eugenb (~Adium@dslb-188-100-030-044.pools.arcor-ip.net) |
| 21:06.45 | Cotulla | lol |
| 21:06.59 | Cotulla | it's just stupid |
| 21:07.03 | Cotulla | arm have sections 1 mb |
| 21:07.09 | Cotulla | so 1 mb align is normal |
| 21:07.15 | Alex[sp3dev] | if it is stupid, go ahead and fix it |
| 21:07.16 | Cotulla | or ever without align |
| 21:07.24 | *** join/#htc-linux Segnale007 (~Segnale00@ppp-241-32.26-151.libero.it) |
| 21:07.46 | Cotulla | FLPT needs to be 64k aligned ofcourse |
| 21:07.53 | Cotulla | it fucked down me once |
| 21:09.47 | *** join/#htc-linux toastcfh (~toastcfh@unaffiliated/toastcfh) |
| 21:25.00 | *** part/#htc-linux Alex[sp3dev] (~alexander@86.110.163.19) |
| 21:28.30 | *** join/#htc-linux toastcfh (~toastcfh@41-32.200-68.tampabay.res.rr.com) |
| 21:32.00 | *** join/#htc-linux helicopter88 (~helicopte@host94-112-dynamic.46-79-r.retail.telecomitalia.it) |
| 21:32.08 | *** join/#htc-linux toastcfh (~toastcfh@unaffiliated/toastcfh) |
| 21:33.45 | *** join/#htc-linux WyrM (~WyrM@unaffiliated/wyrm) |
| 21:41.14 | WyrM | DT sucks. |
| 21:41.51 | WyrM | manufacturers will just use it to release less info when they have to release the kernel source as required by the gpl. |
| 21:42.21 | WyrM | like motorola did with magx. |
| 21:42.45 | WyrM | IIRC the original DT implementation for ARM was sent by them. |
| 21:52.04 | *** join/#htc-linux xawen1 (~xawen@pool-98-117-40-230.bltmmd.fios.verizon.net) |
| 21:56.06 | *** part/#htc-linux Cotulla (~w3489sdf@nat100-255-205-109.tvoe.tv) |
| 22:28.40 | *** join/#htc-linux bukington (~bukington@fac34-2-82-228-151-145.fbx.proxad.net) |
| 22:30.15 | *** join/#htc-linux mitsutaka (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
| 23:06.55 | *** join/#htc-linux arif-ali_ (~arif-ali@88-111-132-115.dynamic.dsl.as9105.com) |
| 23:12.40 | *** join/#htc-linux programmer8922 (~Evan@97-119-137-120.albq.qwest.net) |