IRC log for #htc-linux on 20100514

00:00.41hamagcok goodnight guys.
00:02.25*** join/#htc-linux surge (surge@pool-98-118-154-21.bflony.fios.verizon.net)
00:12.37*** join/#htc-linux xawen (~Admin@pool-71-179-244-8.bltmmd.fios.verizon.net)
00:25.34*** join/#htc-linux Hoochster (~hooch@173.216.128.12)
00:26.52*** join/#htc-linux xawen (~Admin@pool-71-179-244-8.bltmmd.fios.verizon.net)
00:32.38*** join/#htc-linux xawen (~Admin@pool-71-179-244-8.bltmmd.fios.verizon.net)
00:39.38*** join/#htc-linux zfxp (~zfxp@adsl-179-125-253.gnv.bellsouth.net)
00:41.25zfxpAny status on raph500?
01:29.43*** join/#htc-linux mpannen (~matt@adsl-75-51-3-31.dsl.milwwi.sbcglobal.net)
01:42.16*** join/#htc-linux [[NEO]] (~Neo31@unaffiliated/neo31)
01:43.33*** join/#htc-linux enjalot (~enjalot@c-68-35-253-116.hsd1.fl.comcast.net)
01:46.55*** join/#htc-linux DarkMasterHalo (~sam@modemcable187.118-80-70.mc.videotron.ca)
01:56.54jonprymicrop iocontrol dumps from wavedev init http://pastebin.com/w0GriLYb
01:59.46jonprylots of pointers in there. due to totally weird linkages, 0xbe*** is stack, duh. 0xafe** is heap. most everything else is either a function pointer, or pointer to a data segment in the PE.
02:00.09jonpryPE's are loaded everywhere from 0x20000 to 0x4300000
02:12.49*** join/#htc-linux marex (~marex@vasut.kolej.mff.cuni.cz)
02:54.40*** join/#htc-linux damentz (~damentz@adsl-99-69-147-242.dsl.aus2tx.sbcglobal.net)
02:58.49*** part/#htc-linux damentz (~damentz@adsl-99-69-147-242.dsl.aus2tx.sbcglobal.net)
03:07.17*** join/#htc-linux thedicemaster2 (~thedicema@e200071.upc-e.chello.nl)
03:14.50*** join/#htc-linux jumoit (~samuel@121.77.87.145)
03:40.29jonpryi think i am back on the right track again. Code is doing lots of things. In particular it is IOControlling DSP1:
03:40.38*** part/#htc-linux jumoit (~samuel@121.77.87.145)
03:40.45jonprynot sure what driver handles that
03:43.36*** join/#htc-linux jumoit (~samuel@121.77.87.145)
03:50.55*** join/#htc-linux Forsaken|Desk (~Forsaken|@dslb-084-058-220-232.pools.arcor-ip.net)
03:57.04*** join/#htc-linux DaveWM (dave@h-96-13-75-173.ip.alltel.net)
04:12.50*** join/#htc-linux awatt (~watsky@ip70-171-107-157.no.no.cox.net)
04:15.51*** join/#htc-linux swc|666 (~infidel20@unaffiliated/swc666/x-4934821)
04:29.39*** join/#htc-linux BHSPitMonkey (~stephen@unaffiliated/bhspitmonkey)
05:10.54*** join/#htc-linux kiozen (~oeichler@rgnb-5d878729.pool.mediaWays.net)
05:24.42*** join/#htc-linux jamenlang (~jamenlang@68-170-60-12.mammothnetworks.com)
06:25.08*** join/#htc-linux onen|openBmap (~quassel@vbo91-1-89-87-201-85.dsl.club-internet.fr)
06:37.52*** join/#htc-linux dzo__ (~dzo@mail.marginz.co.nz)
06:45.57*** join/#htc-linux Wout (~wout@internet.office.tweakers.net)
07:11.37*** join/#htc-linux dekar_ (~dekar@port-92-193-47-168.dynamic.qsc.de)
07:15.02*** join/#htc-linux caffeinejunky (~4c5e9a76@gateway/web/freenode/x-nqgxxldnoskcpxed)
07:18.12*** join/#htc-linux Razzee (~Razzee@p4FCAB905.dip.t-dialin.net)
07:20.23caffeinejunkyanyone here working on the hd2?
07:20.40caffeinejunkyi think i might have some helpful info
07:20.54Razzeesomebody was looking for other devs
07:21.50caffeinejunkyso we're short on guys working on the kernel i assume?
07:23.14*** join/#htc-linux kiozen (~kiozen@p54921132.dip0.t-ipconnect.de)
07:23.43Razzeei heard there are some, but i don't know one
07:24.01Razzeeand this doesn't seem like the place they meet
07:25.10caffeinejunkyah i see. well i dont really get on irc much unless its important.
07:25.22caffeinejunkyi just thought that this may help developing the kernel.
07:25.28*** join/#htc-linux dzo__ (~dzo@mail.marginz.co.nz)
07:26.36caffeinejunkysharp released the 1GHz IS01 developer's edition mid.
07:26.59caffeinejunkyit uses a snapdragon and runs a stripped down version of donut.
07:28.37caffeinejunkyif maybe we can get the kernel sources for this mid, it might help.
07:34.00*** join/#htc-linux awatt_xda (~watsky@ip70-171-107-157.no.no.cox.net)
07:34.59caffeinejunkyit sports a capacitative display running at 960X480, has a 5 megapixel AF camera plus a 0.43 front camera for video calls, 802.11 b/g wifi, bluetooth 2.1 + edr, micro sd slot, 1seg mobile tv tuner, 4 gigs of internal storage and a qualcomm 3g cdma data
07:37.19caffeinejunkyfound specs here: http://www.umpcportal.com/products/Sharp/IS01/
07:42.10caffeinejunkyit may not be made by htc but i guess theres nothing to lose
07:43.56caffeinejunkyif anyone developing for the leo sees my msg, drop a line on where i can reach u guys. i'll just look at the logs in the morning.
07:44.01caffeinejunkyim out.
07:44.04*** part/#htc-linux caffeinejunky (~4c5e9a76@gateway/web/freenode/x-nqgxxldnoskcpxed)
07:47.05*** join/#htc-linux dzo__ (~dzo@mail.marginz.co.nz)
08:01.16*** join/#htc-linux kvaster (~kvaster@93.84.112.80)
08:08.36*** join/#htc-linux dzo__ (~dzo@mail.marginz.co.nz)
08:10.30*** join/#htc-linux Markinus (~Miranda@gtng-4db04703.pool.mediaWays.net)
08:25.50*** join/#htc-linux brandon (~brandon@96-28-73-187.dhcp.insightbb.com)
08:28.37*** join/#htc-linux dzo__ (~dzo@mail.marginz.co.nz)
08:30.34*** join/#htc-linux Captnoord (~Captnoord@dc5147a47b.adsl.wanadoo.nl)
08:35.51*** join/#htc-linux LeTama (~IceChat7@vau06-1-82-228-252-90.fbx.proxad.net)
08:40.45Woutoi Captnoord
08:44.33Captnoordoi
09:02.50*** join/#htc-linux waleed (~quassel@91.186.246.231)
09:03.21*** join/#htc-linux mpannen (~matt@75.51.3.31)
09:03.47*** join/#htc-linux mickey|office (~Mickey@dialbs-092-079-168-007.static.arcor-ip.net)
09:05.07*** join/#htc-linux Neo31 (~Neo31@unaffiliated/neo31)
09:08.21*** join/#htc-linux GNUtoo (~GNUtoo@host109-48-dynamic.21-79-r.retail.telecomitalia.it)
09:08.37brandonLeTama any futher breakthroughs I suggest toolchain problems to Markinus and he said that is wasn't that
09:10.24LeTamaHi brandon, well, nothing new on my side, it's clear for me that static linking is broken, at least on my toolchain
09:10.55LeTamafor what reason exactly, I don't know
09:12.05brandonhave you read this http://www.friendlyarm.net/forum/topic/246 it goes in to pretty through explanation
09:15.44brandonI have a US HD2 and the kernel had been panicing every time but the new update allow me to boot busy box from a second partition on my sd card
09:16.03LeTamawell, their cause is ftp transfer trashing the executable, that's not what we have here.
09:16.43LeTamawhat update ?
09:17.22brandonMarkinus posted update yesterday
09:27.59*** join/#htc-linux Vito_| (~Sysel@94.142.234.1)
09:35.16*** join/#htc-linux hax0r_ (~hax0r@2a01:e35:8bd4:8bb0:215:afff:feb4:dd78)
09:43.22*** join/#htc-linux luminoso (~lumos@av-217-129-131-108.netvisao.pt)
09:56.50*** join/#htc-linux MethoS- (~clemens@134.102.106.250)
10:48.29*** join/#htc-linux thedicemaster (~thedicema@e200071.upc-e.chello.nl)
10:56.56*** join/#htc-linux majster (~majster@cpc12-aztw25-2-0-cust68.aztw.cable.virginmedia.com)
11:02.28*** join/#htc-linux GeekLad (~GeekLad@adsl-82-5-166.jax.bellsouth.net)
11:07.44LeTamabrandon: thanks for the info, didn't see it. It doesn't seem to change the static link issue
11:11.07GNUtoomickey|office, hi
11:12.36GNUtooLeTama, glibc or uclibc?
11:13.09LeTamaglibc
11:13.38GNUtooah ok because in the past I had issues with parts that should be inlined and that weren't in ld.so of uclibc
11:15.52*** join/#htc-linux r0w3zz (~r0w3zz@p54A8FEDD.dip.t-dialin.net)
11:19.03*** join/#htc-linux Razzee (~Razzee@p4FCAB905.dip.t-dialin.net)
11:34.26mickey|officemorning GNUtoo
11:35.49*** join/#htc-linux GNUtoo|oeee (~GNUtoo@host109-48-dynamic.21-79-r.retail.telecomitalia.it)
11:43.04*** join/#htc-linux Vito_| (~Sysel@94.142.234.1)
11:45.22*** join/#htc-linux Heart| (markettalk@static.177.79.46.78.clients.your-server.de)
11:57.39*** join/#htc-linux Untouchab1e (~Untouchab@226.80-203-35.nextgentel.com)
12:08.20*** join/#htc-linux dcordes (~dccordes@unaffiliated/dcordes)
12:12.45dcordeshi all
12:13.05dcordesLeTama, did you ever try the google toolchain?
12:22.16*** join/#htc-linux mickey|zzZZzz (~mickey@80.81.242.146)
12:28.58*** join/#htc-linux thont (~littlemat@155.69.170.143)
12:35.56GNUtoomickey|office, I've huge issues with xorg
12:36.23GNUtoomickey|office, under xorg suspend doesn't work because of my changes to msmfb
12:36.33GNUtoobut even fake suspend
12:38.31*** join/#htc-linux DaveWM (~dave@173-19-136-12.client.mchsi.com)
12:46.39*** join/#htc-linux Vito_| (~Sysel@94.142.234.1)
12:46.41*** join/#htc-linux templis (~templis@cl-1285.ham-01.de.sixxs.net)
12:46.59templishi
12:47.34*** join/#htc-linux cmg29 (~cgoffredo@bflo.corp.synacor.com)
12:49.45*** join/#htc-linux jumoit (~samuel@121.77.87.145)
13:00.11WoutCaptnoord: any idea on when you'll push the rhod temp update? or isnt it any good without the tables and stuff?
13:05.27*** join/#htc-linux cmg29 (~cgoffredo@bflo.corp.synacor.com)
13:12.40*** join/#htc-linux Untouchab1e (~Untouchab@226.80-203-35.nextgentel.com)
13:27.07*** join/#htc-linux evildarknight (~charles@41.211.72.35)
13:32.22*** join/#htc-linux Kawaiius (~yadayada@pool-72-84-153-225.slsbmd.east.verizon.net)
13:39.27*** join/#htc-linux onen|openBmap (~quassel@vbo91-1-89-87-201-85.dsl.club-internet.fr)
13:58.40mickey|officeGNUtoo: which changes? you did a lot to msmfb
13:58.54mickey|officefrom my point of view you can revert anything but the FBIOBLANK
13:59.21GNUtooouch that's precisely what I want to revert
13:59.40GNUtooif I don't revert I really need help for fixing it
13:59.58GNUtoobecause we have an unusable phone if we use Xorg
14:00.48GNUtoobtw leviathan is working on rotation?
14:01.07GNUtooI bet he will do it with xf86-video-msm which isn't the thing to do
14:01.17GNUtooxf86-video-fbdev is better
14:01.26GNUtoobecause it has already some rotation code
14:07.18LeTamahi dcordes
14:07.54LeTamano, didn't try static build with it yet
14:09.01LeTamaI modified adb to do dynamic link on adbd with it, it works fine
14:10.04dcordesLeTama, sounds good :)
14:10.10dcordesnew hope!
14:10.43LeTamaeven though I'm not able to have network over usb. I probably miss something for this
14:11.28*** join/#htc-linux balans (~Gebruiker@82-170-215-5.ip.telfort.nl)
14:17.43dcordesLeTama, I never used adb so I fear I can't help with this.
14:18.53mickey|officeGNUtoo: i don't understand why FBIOBLANK would kill suspend
14:19.05GNUtoomickey|office, basically it does that:
14:19.16GNUtooyou ask for it to suspend
14:19.25GNUtooand the display goes black
14:19.30GNUtoothen it comes back just after
14:19.46LeTamadcordes: ok. it would be nice to have network to be able to do telnet.
14:23.01mickey|officeGNUtoo: and are you sure that removing .fbioblank is fixing it?
14:23.10GNUtooI tried it
14:23.15mickey|officejust removing this bit?
14:23.15GNUtooI can retry
14:23.27GNUtoothe correct solution would be to get help on this mather
14:23.40mickey|officeworth a try on the Xorg list
14:25.01GNUtoook
14:33.34dcordesLeTama, so no kernel patches were needed for this to work?
14:34.48LeTamadcordes:no, nothing. I pulled latests changes from Markinus this morning, but it was working before
14:37.42*** join/#htc-linux myn (cereal@cpe-71-64-8-167.insight.res.rr.com)
14:47.28*** join/#htc-linux [acl] (~abel@96.246.167.90)
14:47.50dcordesLeTama, so something is wrong about the linking stuff
14:47.56dcordesLeTama, how can we solve this?
14:48.55LeTamadcordes: no idea. I'm not even sure we have to solve it.
14:51.22dcordesLeTama, so maybe the kernel is fine and we just need to adjust the way we compile userspace binaries?
14:51.28LeTamayes
14:52.16LeTamawe need to have the right options to compile and link all dynamically.
14:52.26GNUtoomickey|office, I don't think it's xorg specific,I think the issue is in the driver,more precisely concurency
14:52.31GNUtooor something like that
14:53.39LeTamaI don't know if static is mandatory for anything except that it makes initrd a bit big with the lib*so
14:56.45dcordesGNUtoo, does angstrom distro make use of static linking?
14:56.59dcordesLeTama, do you think no qsd devices can run statically linked binaries?
14:57.15GNUtooin oe you can compile static things
14:57.19GNUtoolike busybox-static
14:57.26dcordesGNUtoo, but normally it's all dynamic?
14:57.31GNUtoobut an entire static distro....no
14:57.42GNUtoobut we build -static version of packages
14:57.46GNUtooor libraries
14:57.59GNUtooso you could compile a static binary
14:58.32GNUtoobut if you want all static there must be a way
14:58.49dcordesno I don't want that. I only want to know if my angstrom images include these -static packages
14:59.00GNUtooimages no
14:59.06dcordesok
14:59.06GNUtoopackages in feed yet
14:59.16mickey|officeGNUtoo: yes, i agree. (concurrency)
14:59.49dcordesLeTama, ok so I can assume in my angstrom initrds statical linking can not be the problem of segfaults. so it must be the compile options
14:59.52GNUtoomickey|office, ok
15:00.29GNUtooI'll look if there is a kernel mailing list for video
15:00.34GNUtoothere should be
15:01.39GNUtoomickey|office, btw is levithan working on rotation,if yes on which xorg driver? xf86-video-msm or xf86-video-fbdev?
15:01.56mickey|officesorry, i have no idea
15:02.11mickey|officei don't think he's actively working on anything dream, but rather studying
15:02.11GNUtoook
15:04.28dcordesLeTama, can you tell me what flags you use to make your successfully running test things?
15:04.50GNUtooLeTama, did you try different distro?
15:06.03LeTamadcordes: see my last email on mailing list, it has the options
15:06.43LeTamaGNUtoo: no, I didn't
15:06.45dcordesLeTama, ok thanks
15:07.04KensanGNUtoo: actually, there's a arm-msm mailinglist: http://vger.kernel.org/vger-lists.html#linux-arm-msm
15:07.19GNUtooKensan, wow thanks a lot
15:07.40KensanGNUtoo: and theres one for fbdev
15:07.44GNUtoook nice
15:07.46GNUtoothanks a lot
15:07.50KensanGNUtoo: just look through the list of mailinglists
15:07.55GNUtooI'll look in one hour
15:08.13KensanGNUtoo: there might be some "inofficial" ones not hosted on vger so googling might show some other lists.
15:08.22GNUtoook
15:23.05*** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring)
15:25.15dcordesLeTama, ok found them. I will try to reproduce that in openembedded and make an angstrom with these flags
15:25.22dcordesbbl
15:25.41LeTamadcordes: good
15:31.16HiddenOnehmmm
15:31.26HiddenOneis there a way to run maemo 5?
15:31.39HiddenOne( on something else other than those nokia tablets )
15:32.03HiddenOnesay...Rhodium....or Topaz....or something :P
15:33.34*** part/#htc-linux jumoit (~samuel@121.77.87.145)
15:35.06phhHiddenOne: mer (opensource fork of maemo or something like that) does
15:37.38PSPdemonahh
15:37.40PSPdemongotcha
15:38.56phhI didn't plaid a lot with it
15:39.10phhbut IIRC I could web browse with it
15:39.18phh(on wifi)
15:39.22PSPdemonheh
15:39.48phhbut the current user demand is more about ubuntu than mer :p
15:39.52phh(which I understand
15:39.53phh)
15:56.30Captnoordre
16:03.37Captnoordphh how's gps?
16:03.38Captnoord:P
16:06.27cmg29Yeah, someone should commit something. We haven't had to update our files today yet. :)
16:07.33Captnoordin a moment
16:08.03Heart|hi... i'm new to android world on windows devices ;-)... so what shall use for my blackstone? which "combi"
16:08.04Captnoordneed to test
16:08.28Captnoordblackstone is one of the least supported devices...
16:08.32Heart|the TheBrilliantMistake one?
16:08.42Captnoordmainly because there are not many testers
16:08.45Captnoordor dev's having one
16:08.51Heart|hmm thats not good to hear :-(
16:12.41KhaytsusI saw the forum poast saying a fix is on the way.  How punny :D
16:13.10Captnoordpunny?
16:13.38KhaytsusGPX..  Fix..
16:13.50Captnoordk
16:13.55Khaytsuserm GPS :)
16:13.56Captnoordphh is always working on fixes
16:17.44KhaytsusWhich is greatly appreciated, I was just trying to make a sily jokx about a GPS "fix" being on the way.  <joke></joke>
16:17.55Captnoordlol
16:18.14CaptnoordI pollished the copper pipes
16:18.17Captnoordthen cleaned it
16:18.22Captnoordand clear coated it
16:18.26Captnoordis kinda high
16:18.28Captnoord:P
16:20.20KhaytsusPlumbing or watercooling? :)
16:21.08Captnoordplumbing
16:21.35*** join/#htc-linux kvaster (~kvaster@vpn-e0.bas-net.by)
16:24.22Captnoordphh I pushed a possible fix for the diamond problems
16:24.23*** join/#htc-linux enjalot (~enjalot@c-68-35-253-116.hsd1.fl.comcast.net)
16:24.24Captnoordpanel problems
16:26.37KhaytsusAnyone have an idea why the _ key does a right-cursor or tab or something weird?
16:27.13Captnoordcmg29: commited shit
16:27.19Captnoordnah
16:27.28Captnoordpushed raph navi fix
16:27.31Captnoordand diamond panel fix
16:30.54jonpryi am having trouble using msgget() and family. these seem to have been removed from bionic for whatever reason, so i added them back in. but i get errno == "function not implemented"
16:31.17jonpryafaict it is just a kernel syscall. does it need to be compiled in the kernel or something?
16:40.06jonpryah, CONFIG_SYSVRPC
16:44.40*** join/#htc-linux DaveWM (~dave@173-19-136-12.client.mchsi.com)
16:47.04*** join/#htc-linux Wout (~a@s55927d21.adsl.wanadoo.nl)
16:53.51stinebdxdandroid twitter has been busy lately
16:53.57Woutthere is one?
16:54.09stinebdhttp://twitter.com/xdandroid
16:55.10Woutnice
16:55.25stinebdare you the person that requested it?
16:55.36Wouteeeehm, don't think so?
16:55.36stinebdsomeone here did
16:56.17WoutI'm happy with it nontheless :P
17:01.29*** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring)
17:02.18phhanyway
17:02.22phhno GPS should be almost usable.
17:02.34phhI just think that android behaviour is stupid.
17:02.59phh(the app needs to be foreground to activate GPS ....)
17:03.01phh(and keep it)
17:03.07stinebdthat's true
17:03.11phhand with fixes that takes a long time, or even forever....
17:03.14phhit's REALLY stupid.
17:04.25cmg29captnoord:  Yay, commits! lol
17:06.50*** join/#htc-linux kiozen (~oeichler@rgnb-5d878729.pool.mediaWays.net)
17:10.13*** join/#htc-linux brandon2004 (~brandon@96-28-73-187.dhcp.insightbb.com)
17:16.00jonpryi got more rpc's out of wavedev. still no interrupt though. http://pastebin.com/frBMLZQu
17:16.53phhjonpry: you could do cpu_to_be32 :p
17:18.22jonpryi like be better :-)
17:19.04phhjonpry: I guess I'm supposed to check those values compared to what we do ?
17:19.09phhor you already did ?
17:19.27jonpryi didn't check it
17:19.30jonpryjust got it
17:19.46jonpryi'm still looking for the interrupt :-)
17:19.50phhok
17:19.53phhfair enough.
17:20.24phhgroumpf, gps lose its previous data sometimes
17:20.29phhI wonder if it's because of pdsm_end_session
17:20.39brandon2004LeTama take a look at this http://honeypod.blogspot.com/2007/12/dynamically-linked-hello-world-for.html
17:21.15brandon2004it has options for dynamically linked binaries for arm running android using the codesourcery toolchain
17:23.41Captnoordphh i've commited a fix for the panel bug out on suspend / wake
17:23.48Captnoordhope that was the problem
17:25.01cmg29capt: how's your rhod batt stuff?
17:25.21Captnoordwip as hell
17:25.34Captnoordits complicated again
17:25.44Captnoordthe temperature stuff needs work / implementation
17:25.45cmg29yeah, i figger'd as much
17:26.01Captnoordthe battery level could use a hack or 2
17:26.03jonpryphh: you mean it loses the fix?
17:26.22Captnoordcmg29: I will take another peek at the asm later tonight
17:26.33Captnoordneed to relax a bit
17:26.36cmg29cool :)
17:26.51*** join/#htc-linux leviathan (~davidlan@nariwork-03.ee.ethz.ch)
17:26.54leviathanGNUtoo: hi
17:26.57GNUtoohi
17:26.58cmg29relax? aren't you working on laying floor and painting pipes when you're not building android stuff?
17:26.59cmg29hehe
17:27.03GNUtooso basically you're on wifi too
17:27.10GNUtooleobaillard, pandora has some patches
17:27.16GNUtoooops
17:27.17GNUtoosorry
17:27.24GNUtooleviathan, panrora has some patches
17:27.26Captnoordcmg29: hehe.... good point
17:27.31Captnoordnah.... just relax
17:27.32GNUtoomost of them were submited upstream
17:28.10GNUtooleviathan, the one who are left are:
17:28.10leviathanyes
17:28.13leviathanI saw it
17:28.21GNUtoo0010-wl1251-use-sdio_readb_ext-for-ELP-to-work.patch
17:28.26leviathanhmm, ok
17:28.27GNUtoo0011-wl1251-add-reset-handling.patch
17:28.31GNUtoo0013-wl1251-prevent-scan-when-connected.patch
17:28.32LeTamabrandon2004: thanks for the pointer, will take a look
17:28.33GNUtoothat's all
17:28.38leviathank
17:29.00GNUtooapart that It works for 20/30 min and after I've some issues
17:29.08GNUtooI'm using PSM
17:30.40GNUtooand currently 05-12 compat wireless
17:31.05brandon2004I just read that seg faults will happen if boot_thread-mt is listed before pthread during linking
17:31.16brandon2004@LeTama
17:31.35GNUtooleviathan, here the script I use http://pastebin.com/kibRHGxP
17:31.44GNUtooit always comes up the first time
17:31.52GNUtoowith compat-wireless
17:31.54*** join/#htc-linux hamagc (~hamagc@72-57-134-191.pools.spcsdns.net)
17:31.59GNUtooit doesn't with stock
17:33.46jonprystupid adsp interrupt. i don't understand where it is. my windowsce implementation is getting scaringly good.
17:34.56Captnoordlol
17:35.06Captnoordlike wine beeing able to run virusses
17:35.10Captnoord:P
17:35.28jonprylol
17:35.47GNUtooleviathan, so what did you do with wifi?
17:36.29jonpryi don't let it write any files yet, so virus's probably crash
17:37.25hamagcso any word on cdma gps yet?
17:37.34leviathanGNUtoo: try'n to make PSM working with stock :)
17:37.48GNUtooleviathan, ah ok why not using compat-wireless?
17:37.53leviathanuhm
17:37.54GNUtooit got tons of fixes
17:38.08leviathanhmmmmmmm
17:38.13leviathanits worth a shot
17:38.14phhjonpry: if you boot winmo, then take a fix, then boot linux, A9 remembers of the previous sat position
17:38.17phhso you can fix really quickyl
17:38.20phhquickly*
17:38.29leviathanGNUtoo: is compat wireless in OE now?
17:38.32*** join/#htc-linux bzo (~chatzilla@netblock-68-183-234-14.dslextreme.com)
17:38.40phhbut sometimes it seems to lose it
17:38.45GNUtooleviathan, no but eric bernard got a bad patch for it
17:38.50GNUtoowhich is more recent
17:38.55GNUtoobut your patch may be better
17:39.10GNUtoobecause eric bernard's patch had no RCONFLICTS and similar
17:39.30phhI receive that in pm: "Hey, I'm interested in helping out with the XDANDROID project. I've got some decent firmware experience and want to help."
17:39.32GNUtooleviathan, but you could refresh your patch with the idea of selecting only wl12xx
17:39.35phhwhat the hell should I do with that ?
17:39.46stinebdtell him if he gets camera working, he's in
17:40.01GNUtooalso tell him to define firmware experience
17:40.29phhGNUtoo: I prefer stinebd's version.
17:40.47GNUtoook that's why I told "also"
17:41.58GNUtooleviathan, also on fb side things aren't great:
17:42.05GNUtoo*broken suspend because of fb
17:42.10phhanyway I don't care about devs experience
17:42.10phhI think that most people here started without any experience.
17:42.10phh(in this field I mena.)
17:42.11phhmean*
17:42.11GNUtoo*no rotation
17:42.13GNUtooetc...
17:42.35jonpryphh: maybe the previous satellite position isn't real. like sometimes it is just a cell position
17:42.48phhjonpry: it is real
17:42.50GNUtooyes but knowing someone experience could help,if someone is experienced with sound let make him work on sound for instance
17:42.56phhjonpry: in NMEA you can see sat position
17:43.09bzophh: did you say you figured out the gps rpc struct already?
17:43.12phhand I've checked them, they are correct
17:43.15phhbzo: some stuff
17:43.30phhwait a second I'll find you the file.
17:43.48bzook, thanks
17:43.59phhright I lost it again, so it's not a second, will be some minutes.
17:44.08bzoI'm looking at my rpc log, and it is a bit difficult to figure out without data to match :)
17:44.16phhah got it
17:44.25phhname file is TTTTT
17:44.40tmztphh: mer has kind of become meego but they aren't really sure about non-gl devices (like msm7k)
17:44.48phhtmzt: ok
17:44.57phhbzo: http://husson.hd.free.fr/TTTTT
17:45.08tmztrhobuntu is probably the best option going forward until the current arm repos can be rebuilt for our cpu
17:45.24phhyeah rhobuntu has really gone a good way
17:45.34bzophh: got it thanks. Let's see if my pos is all 666 :/
17:45.49phhbzo: I don't have positions in it :p
17:45.54phhI've time and sat positions
17:46.05stinebdwhy can't i reproduce this bootanim loop?
17:46.14bzophh: didn't you say it was reporting your pos as all 666?
17:46.37phhbzo: ah no 666.6 was the error dilution when it fixes only 2 sats
17:47.01bzoah, that is good I guess
17:47.04phhbzo: do you have some gps test program to dump your own RPC stuff ?
17:47.17bzojust your test.c
17:47.23phhyeah I've no clue how it can find 4 unknown variables with only 2 infos :p
17:47.40phhbzo: yeah right that's is enough
17:47.49bzoI'm looking at the vogue structure right now, maybe it will give me a clue
17:48.12phhbzo: so when you get a fix, you should get a message with procid=1 (the second field after 310005b), and then evid=0x7
17:48.17phh(two fields before 0xda3)
17:48.29bzothe vogue code is kinda funny though, first they copy the data with cpu_to_be32, then turn in back with be32_to_cpu
17:49.14leviathanGNUtoo: but dri would be nice
17:50.21bzophh: good, I see my data has a fix. At least I know I should see good data
17:50.58*** join/#htc-linux ac (~opera@e178208066.adsl.alicedsl.de)
17:51.07phhbzo: cool
17:51.14phhbzo: try to see if date and time are readable as mine is
17:51.22*** join/#htc-linux thont (~littlemat@155.69.170.143)
17:52.56bzophh: date time looks good
17:53.01jonpryi tried building wine under rhobuntu the other day, lol
17:53.15phhjonpry: ?!?
17:53.17phhwine on arm ?
17:53.29Captnoordhmmm
17:53.38Captnoordrunning windows apps on rhobuntu
17:53.38jonpryyeah, it is known to work on powerpc and alpha
17:53.40Captnoorddamn
17:53.50Captnoordwonder how long building takes
17:54.05jonpryin theory you can use winelib to compile any win32 app for arm/linux
17:54.30phhCaptnoord: windows ARM apps ?
17:54.39phhjonpry: ah I forgot you have your own windows app
17:54.54Captnoordphh lol
17:54.56Captnoordhehe
17:54.57Captnoordidd
17:55.22jonprymy goal was to use their threading/mem management as i became aware that it requires a great deal of accuracy to work
17:55.45jonprybut there threading stuff is processor specific and not written for ARM
17:56.24jonprycompiling is reasonably fast. ./configure takes a couple of hours
17:56.41Captnoordlol
17:56.52phhCaptnoord: that's called "reasonably".
17:56.57Captnoordlolz
17:57.04phhjonpry: last century they created something really great
17:57.07phhthat's called cross building
17:57.19Captnoordphh did you test my commit btw...
17:57.21jonpryyeah right.
17:57.26Captnoordyou will see waking up
17:57.29Captnoordis a lot faster somehow
17:57.38Captnoordmainly because of the the panel code now actualy working
17:57.49phhleobaillard: hear that ? :p
17:58.02Captnoordseems that i2c fails the first time on wakeup
17:58.11Captnoordand we didn't had any checks for that
17:58.16phhok
17:58.20jonpryi think the actual "make" had it worked, would have been about 3 hours, which is not that far from what it takes on x86
17:58.28leobaillard:p
17:58.37Captnoordleobaillard: please test
17:58.49leobaillardI will, just give me 10 minutes
17:58.51Captnoordwould like some feedback if it changes anything for ya
17:59.01phhbzo: so in my dump you have RPC and NMEA so you can guess
17:59.10phhthere aren't many values that are big enough to have position
17:59.30phhunknown values.
18:00.00bzoyeah, hopefully they don't use some weird encoding
18:00.22phhbzo: it seems it's weird
18:00.53bzolooks liek the vogue has lat/long in 64 bit fields
18:01.06stinebdCaptnoord: that boot loop happens with babijoee's release?
18:01.25Captnoordgot the release from connect-utb
18:01.30stinebdoh right
18:01.38stinebdsorry
18:01.50Captnoordcan tell you the exact file name
18:01.51chamonixhey guys
18:02.02CaptnoordXDANDROID21AOSP030510.7z
18:02.38phhbzo: float ?
18:02.47phhI've no clue how to read float values :D
18:03.04Captnoord0x3f800000 = 1.0f
18:03.15stinebdCaptnoord: thanks
18:03.27Captnoordstinebd: your welcome
18:03.36bzophh: hmm, maybe. They use 2 int hi/lo values to store, but maybe they convert it upstream?
18:03.51phhbzo: don't know
18:04.05phhI haven't understood this part of code of vogue
18:04.09phh(I haven't reallt tried.)
18:13.12GNUtooleviathan, yes but let's first have the basics no? I'll eat
18:20.11Captnoordleobaillard: and?
18:20.27leobaillardCaptnoord: it's booting...
18:20.35Captnoordk
18:21.23*** join/#htc-linux friendofgermany (~fog@p5B26F803.dip.t-dialin.net)
18:23.12*** part/#htc-linux friendofgermany (~fog@p5B26F803.dip.t-dialin.net)
18:24.55chamonixCaptnoord: have u done any scope reading with kernel_20100512_1414 ?
18:25.34Captnoordscope readings
18:25.34Captnoordnope
18:25.55stinebdphh: i'm working on the libhardware_legacy build bot now
18:25.56Captnoordchamonix: why
18:26.01phhstinebd: cool
18:26.06chamonixok, just asking because my phone didn't even reach 18 hrs with that kernel
18:26.08CaptnoordI only enabled topaz battery code
18:26.12CaptnoordoOoOoOo
18:26.15Captnoordhow's that possible
18:26.20CaptnoordI didn't change shit
18:26.22chamonixdunno, I'll re-run
18:26.53Captnoordchamonix: I believe you
18:27.03Captnoordits just that its hard to do changes
18:27.07leobaillardCaptnoord: what should your patch correct exactly ?
18:27.08Captnoordwhen there are so many regressions
18:27.08chamonixthat's not the problem, I just wanna be sure
18:27.26Captnoordleobaillard: it should fix the diam's board not waking up correctly
18:27.28Captnoordfor some users
18:27.32Captnoordit will be reported as
18:27.48Captnoordslow panel wakeup on call
18:27.52Captnoordsomething like that
18:28.05Captnoordat least on my raph with diamonds panel code
18:28.08Captnoordits now perfect
18:28.26Captnoordbefore it didn't woke up properly
18:28.32Captnoordnow it does instantly
18:28.37leobaillardwell, when waking up with power button, it seams faster, but when a call arrives, I have to wait till the third beep at least before it wakes up
18:28.47Captnoordhmmmm
18:28.50CaptnoordI dono
18:28.56leobaillardI'll give it another shot
18:29.12Captnoordits possible that i2c is blocking
18:29.13CaptnoordI dono
18:29.25leobaillardit was quite faster this time :)
18:29.52Captnoordbetter than before?
18:30.26leobaillardyeah, definitely
18:30.32Captnoordgood
18:30.34stinebdignore the ril and sensors commits
18:30.36leobaillardstill not perfect, but much better
18:30.41stinebdit's a side effect of the updated build system
18:30.44Captnoordleobaillard: small steps
18:30.45Captnoordhehe
18:30.48leobaillard;)
18:30.51leobaillardyour right
18:30.54leobaillardthx btw
18:30.59chamonixplouf leobaillard
18:31.08leobaillardplop chamonix :)
18:31.13chamonix:)
18:32.07*** join/#htc-linux phh (~quassel@2a01:e35:2e4b:b2b0:250:8dff:fee1:c793)
18:32.20leobaillardchamonix: while you're here, I have a question for you : in the led effects, is it normal that the charging animation only displays once ?
18:32.58leviathanso
18:33.04leviathanI'll lie down a bit
18:33.08leviathanI'm tired
18:33.37chamonixleobaillard: no
18:33.43chamonixwhat version do u have?
18:34.03leobaillard0.7.0.0
18:34.44stinebdphh: all set
18:34.51phhcool
18:35.01stinebdi've set the ril, sensors and libhardware_legacy to 10 minute checks
18:35.02phhlet's commit some breaking stuff.
18:35.07*** join/#htc-linux MethoS- (~clemens@134.102.106.250)
18:37.41chamonixleobaillard: I noticed exactly the opposite: mine goes on breathing even if I disconnect the charger..... that may mean that the events broadcasted by the battery manager r broken
18:38.02chamonixlet me check logcat
18:40.55chamonixleobaillard: I get D/BatteryBroadcastHandler( 1395): Status is BATTERY_STATUS_CHARGING regularly when I connect charger till charge is complete. till that point it breathes, when I unplug the charger I get BATTERY_STATUS_NOT_CHARGING and the breathing stops so here everything seems fine
18:41.03[acl]sup fellas. too lazy to read some logs today. Any good news ?
18:41.41phh[acl]: I pushed a rootfs that should have almost everything working :p
18:41.51phhperhaps deep sleep is broken still.
18:41.58chamonixlol
18:41.59phhbut gps and wifi works.
18:44.39leobaillardchamonix: no luck here, I get one breathe eVerytime it starts charging and that's all
18:44.58chamonixleobaillard: can u check logcat?
18:45.31leobaillardI can
18:45.57phhyeah sounds really weird
18:46.09chamonixBATTERY_STATUS_* is what u r looking for
18:46.31phhlooping is handled by microp
18:48.35Captnoordhmmmm
18:48.37Captnoorddudes
18:48.40Captnoordits possible
18:48.43Captnoordit doesn't get handled
18:48.50Captnoordbecause of my suspend hack of the battery driver
18:49.10leobaillardchamonix: everything appears to be good
18:49.29leobaillardit stops the effect once unplug and activates it when plugged and charging
18:49.34chamonixmean u see battery charging events?
18:49.35leobaillardbut the effect is "played" once
18:49.36Captnoordhmmm
18:49.37Captnoordokey
18:49.40leobaillardyeah
18:49.49CaptnoordoO
18:49.50Captnoordwierd
18:49.55chamonixok, it's led effects then
18:50.16Captnoordpossible
18:50.24chamonixleobaillard: do u see "D/EffectManager.doEffect( 1395): was called with effect *" ?
18:50.31leobaillardyup
18:50.40chamonixwhat is * ?
18:50.42chamonix3?
18:50.45leobaillard4
18:51.00chamonixok, fine as well
18:51.42leobaillardinteresting... I just went in the preferences of ledeffects and hit the test button of the effect, which didn't work
18:51.57chamonixlol, looks u found a bug
18:52.04chamonixu can keep it is u want :P
18:52.10leobaillardproud
18:53.52leobaillardapparently the problem only happens with this effect
18:54.05chamonixwait a sec
18:54.10phhleobaillard: which effect is it ?
18:54.13phhit's the blinking one ?
18:54.17phhI mean just on/off ?
18:54.22leobaillardfade
18:54.36chamonixlol, it's the only one that does not repeat
18:54.39chamonixby design
18:54.41leobaillardrofl
18:54.42phhchamonix: yeah it's microp "bug"
18:54.49Captnoordphh its not a bug
18:54.54phhleobaillard: I thought you said it was breathing /o\
18:55.01Captnoordits not loaded by the firmware
18:55.07leobaillardsorry, wasn't clear enough apparently
18:55.07Captnoordwe could modify that
18:55.14phhCaptnoord: oh really ?
18:55.17phhthat'd be great
18:55.18Captnoordyea
18:55.24Captnoordbut I won't be doing that
18:55.30Captnoordnot that important
18:55.31chamonixleobaillard: i'm working on a sequencer to work that around, it's a known limitation
18:55.33phhcustom microp would be really great
18:55.40phhespecially on topaz/rhod with those two stupid leds.
18:55.43Captnoordhehe
18:55.49leobaillardchamonix: is it safe to put breathing both to sleeping and charging ? I assigned the fade effect to charging to avoid any problem
18:55.52Captnoordbug me about it another day
18:55.56Captnoordand i'll check it
18:55.57phhwe should be able to do a continous colors between orange and green
18:56.02chamonixI need a sequencer for default android led stuff and for nexus rooted one
18:56.18chamonixleobaillard: yes, that's no problem
18:56.26leobaillardok
18:56.33jonpryMicroP.dll in vial?
18:56.57Captnoordnah
18:56.59Captnoordits not needed
18:57.03Captnoordmicrop is easy to reverse
18:57.09Captnoordnot like the battery driver
18:57.52jonprycould always load the battery driver
18:58.07jonpryin that case do you know what those IOControl's I pasted are doing?
18:58.10Captnoordyea
18:58.17Captnoord?
18:58.23Captnoordwhich io's?
18:58.38Captnoordbattery driver has microp io
18:58.40Captnoordand dex
18:58.47Captnoordmicrop is todo the charge led effect
18:58.52Captnoordwe currently use as sleep effect
18:59.09jonpryhttp://pastebin.com/w0GriLYb
18:59.53bzophh: in the vogue code, latitude and longitude are stored as double floats * 1E8
19:00.03stinebdstill can't get the boot loop
19:00.03jonpryi put in some stuff that translates winmo DEX controls into linux proc_comm. it is not really complete, but easy to finish
19:00.04phhmeh ?
19:00.04stinebdweird
19:00.06Captnoordwhats the io id?
19:00.09phhwhat's the point of floats if it's *1E8 ?!?
19:00.18bzophh: unfortunately, our data does not have any numbers that translate sensibly
19:00.19Captnoord0x1e8?
19:00.21jonpryb2fbc
19:00.25Captnoord0x1e800000
19:00.36bzo1 * 10^8
19:00.56Captnoordlol
19:00.57CaptnoordI see
19:01.01Captnoordi'm not really awake
19:01.04Captnoordzombi mode
19:01.05*** join/#htc-linux swc|666 (~infidel20@unaffiliated/swc666/x-4934821)
19:01.48bzothe code is: location.latitude = (double)data.latitude.data / 1.0E8;
19:01.57phhah right so it's no double
19:01.59phhit's uint32
19:02.05phhthat's better.
19:02.16phhwell more likely sint32.
19:02.16stinebdgps should work without sim right?
19:02.19bzowell, it's stored in a int32+uint32, then cast to a double
19:02.49bzoand I assume it's encoded like this: http://en.wikipedia.org/wiki/Double-precision
19:03.17phhstinebd: just terribly slowly.
19:03.17phhbzo: ok
19:03.17phhbzo: well you dont care about double encoding
19:03.28phhthe RPC messages are in int
19:03.55stinebdk, i'll go try it
19:04.21bzophh: if you cast it as a double, then it will interpret it through the encoding, no?"
19:04.32phhbzo: no
19:04.40Captnoordjonpry: doesn't seem to excist
19:04.42phhit will do uint => double conversion
19:04.43Captnoord:S
19:04.46Captnoordwierd shit
19:05.05bzoyeah, I think you're right, makes more sense why it's stored *10^8
19:05.59jonprythat iocontrol data is wierd. very big structure full of pointers.
19:06.04Captnoordbleh
19:06.05Captnoordnvm
19:06.05tmztjonpry: using led triggers?
19:06.06Captnoordfound it
19:06.19Captnoord.text:0001729C                 MOV     R3, 0xB2EF8
19:06.19Captnoord.text:000172A4                 CMP     R4, R3
19:06.19Captnoord.text:000172A8                 BEQ     loc_172F8
19:06.27Captnoordits
19:06.27CaptnoordWriteLCMBrightnessManualControlToI2C
19:07.00jonpryits created by the sound driver. phh suggested it is i2c
19:07.08Captnoordlol
19:07.13Captnoordby the sound driver
19:07.13Captnoordlol
19:07.51phhCaptnoord: wow.
19:07.58phhjonpry: do you have calltrace ?
19:08.00phhthat sounds "kind of" weird.
19:08.17Captnoord"sounds" wierd
19:08.18Captnoordyes
19:08.19Captnoordhehe
19:08.26jonpryi'm guessing it doesn't do anything i care about. but sometimes a function call modified memory, or has some crazy return that effect the flow of execution
19:08.56jonpryi have traces, but they are so big at this point i can't pastebin them
19:13.45Khaytsusphh: GPS working stable today?  Just did a /last gps and saw encouraging discussion :)
19:14.05phhKhaytsus: stable in the sense it won't kill the system every minute.
19:14.57KhaytsusThat's improvement :D
19:15.55jonpryhttp://prymfg.com/files/trace5-14.txt
19:24.49bzophh: what is your approximate gps altitude?
19:24.58phhbzo: it's written in the nmea log
19:25.21phh232.3meter or 48.0meter
19:25.37phhI think the first one is sea level, the second one some geoid
19:26.05bzoI think I have a reasonable guess, though I can't make the numbers match up
19:26.12bzomaybe a conversion factor, or offset?
19:26.35phhthat's likely
19:27.14bzoso, the 2 fields before the hours/minutes, may be velocity/bearing
19:27.24bzothen there are 3 0 fields before that
19:27.31bzothen altitude before that
19:27.37bzo2 fields of longitude before that
19:27.40bzothen 2 fields latitude
19:28.04bzowhen I compare your location and mine, they very roughly correlate
19:28.27bzothis also matches the vogue gps struct
19:34.26*** join/#htc-linux hamagc (~hamagc@cpe-98-148-205-28.socal.res.rr.com)
19:35.06hamagcLatest rootfs and kernal seems to have broken my radio (cdma sprint)
19:35.25phhcool.
19:35.38Captnoordlol
19:35.39*** join/#htc-linux Wout (~a@s55927d21.adsl.wanadoo.nl)
19:35.40hamagcLol
19:36.02hamagcKnown?
19:36.09phhI don't think I've changed anything related to radio ?!?
19:36.16Captnoordme neighter
19:36.22phhI'm sure it's stinebd's fault.
19:36.28Captnoordyup
19:36.30phh:p
19:36.32Captnoordits his fault
19:36.50hamagcHrm. Let me reboot if that doesn't fix ill back up a release
19:37.38cmg29I'm using cdma sprint with latest both, no issues here...
19:38.33Captnoordbooting with the latest kernel
19:38.37bzophh: hmmm, now I remember that the vogue had a conversion ratio to convert the raw data to the real coordinates
19:38.41Captnoordeven later than yours
19:38.41Captnoord:P
19:39.17cmg29*latest publicly available both
19:39.20cmg29:)
19:40.25*** join/#htc-linux Orkspalter (~Orkspalte@p54B587FE.dip0.t-ipconnect.de)
19:40.46*** part/#htc-linux Orkspalter (~Orkspalte@p54B587FE.dip0.t-ipconnect.de)
19:41.29*** join/#htc-linux hamagc (~hamagc@cpe-98-148-205-28.socal.res.rr.com)
19:41.43hamagcnope still no radio
19:42.19hamagcNow is this the rootfs or kernal?
19:42.29Captnoordchange to older rootfs
19:42.30*** join/#htc-linux Neo31 (~Neo31@unaffiliated/neo31)
19:42.35hamagcK
19:42.46Woutoi
19:42.56Captnoordoi
19:43.16bzophh: haha, the magic vogue gps correction factor is 1.035631, doesn't seem to quite match our data though. Maybe ours is different
19:44.01phhbzo: funny value.
19:45.28Captnoordchamonix: sometimes it doesn't get into deep sleep
19:45.28Captnoord:S
19:45.28Captnoordmaybe thats why it didn't got good battery time
19:45.28chamonixI'll check sleeplog too confirm that
19:46.29bzophh: I think the fields I proposed are right, your data: 487653758 -> 4840.592255, 615193946 -> 0609.116367
19:46.49Captnoordhttp://mathworld.wolfram.com/GoatProblem.html
19:46.52Captnoordthe number
19:46.52bzoconversion factor is not the same, but maybe the nmea data fix was at slightly different time than your rpc trace
19:47.03phhbzo: no it's not
19:47.08phhI checked carefully it was the same
19:47.14phhand you can check you have time from both :p
19:47.31bzoahh, right, but the numbers are too close to be a coincidence
19:47.41bzoand mine match similarly
19:47.41stinebdphh: probably is my fault
19:47.54phhstinebd: lol, how ?
19:48.00phhstinebd: you rebuilt ril ?
19:48.00Captnoordcheck the link dudes
19:48.00bzoso, we just have to figure out the correct conversion factor(s)
19:48.03Captnoordits a known constant
19:48.27stinebdphh: it ended up being different somehow... i changed the build environment to the latest xdaosp
19:48.35stinebdwith proper buildspec.mk and all that jazz
19:48.50Captnoordstinebd: you fixed raph android 2.1?
19:48.51Captnoord:P
19:49.05stinebdthere's nothing wrong with 2.1 on raph :P
19:49.10CaptnoordoO
19:49.11Captnoordcrap
19:49.15stinebdi've tried 25 times at least now
19:49.17stinebdaccording to fsck
19:49.18Captnoordokey
19:49.20phhCaptnoord: this number is far from the one we have
19:49.42Captnoordwoeps
19:49.44Captnoordyour correct
19:49.47stinebdwith old data.img, new data.img, no data.img... i just can't get it to crash
19:49.48Captnoordshould sleep
19:49.49phhbzo: err, but there is a "little" something
19:49.49phh4840.592255 isn't 4,840.59225
19:49.55phhit's 48°40.59225'
19:49.58Captnoordstinebd: then gimme your stuff
19:50.03Captnoordhehe
19:50.05Captnoordthe latest
19:50.07hamagcOk downloading 2nd posted rootfs.be back
19:50.08phhso there are *60 stuff.
19:50.08Captnoordmost modern
19:50.16Captnoordk
19:50.17Captnoord:P
19:50.19stinebdCaptnoord: i took the system.sqsh from the package you're using and the latest kernel, rootfs, initramfs
19:50.29Captnoordokey
19:50.34phhperhaps it's something else than system.sqsh ?
19:50.38Captnoordi'll retry
19:50.39bzophh: conversion factor? :}
19:50.55phhbzo: 1°=60' ?
19:50.57stinebdsomeone posted a logcat
19:51.05stinebdbut it looks like a provisioning package problem in that
19:51.13stinebdhttp://razze.reigninblog.de/logcat.txt
19:51.14Captnoordhmmmm
19:51.59phhbzo: that's why I didn't really try to find it
19:52.06phhbzo: I don't know what's the format of the data
19:52.22stinebdand that's from a diam300
19:52.24bzophh: let me convert them to dec and see how they correlate
19:52.25phhseconds ? minutes ? degree ? in fraction of pi ?
19:52.29stinebdCaptnoord: which raph do you have?
19:52.43Captnoordraph300
19:52.53stinebdhmm
19:52.57CaptnoordT-Mobile Touch Pro
19:53.00Captnoordbut let me try again
19:53.05stinebd300 is gsm?
19:53.06CaptnoordT-Mobile Touch Pro 4
19:53.08Captnoordyup
19:53.18phhTP4 ?
19:53.41Captnoordrhod is tp5
19:53.46stinebdheh
19:53.48phhok.
19:55.19Captnoordphh
19:55.20Captnoordhttp://www.wolframalpha.com/input/?i=+1.035631
19:55.30CaptnoordI dono if its correct
19:55.46Captnoordbut if its only 1.035631
19:55.52Captnoordthere are more possibilities
19:58.34*** join/#htc-linux LTxda (~anon@unaffiliated/ltxda)
19:59.23jonprythat reminds me of the jonpry constant of error. in chemistry class when my experimental data was off, i would multiply or add 1.03 to various quantities until it was right :-)
19:59.36phh-_-'
19:59.45phhwhy 1.03 ?
19:59.51bzophh: 48°40.59225 =48.676537583333335 (register=4867653758)
20:00.03jonpryseemed to work. there was no experiment that couldn't be abritrarily correct
20:00.04stinebd1.03 shall henceforth be defined as 1 jonpry
20:00.10phhbzo: I feel stupid.
20:00.15phhI should have tried that at least.
20:00.23phhbzo: anyway that's great
20:00.38bzowell, the other non-obvious thing is that it is a 64bit field
20:00.46bzogood to see we don't need a weird correction factor though
20:02.43bzophh: so do you think it will be better to get everything from rpc, or are you going to keep using the nmea from smd?
20:02.53phhbzo: everything from RPC
20:02.56phhI hate nmea.
20:03.18bzophh: cool :)
20:03.21phhI mean, NMEA is great, but I just don't want to parse it.
20:03.30phhbzo: you still need to read sats infos.
20:03.44phhthe one I decoded is available only on fixes
20:03.45jonprygpsd?
20:03.49bzoyeah, parsing the nmea seems very inefficient and round about
20:03.50phh(which is really weird)
20:04.03phhjonpry: that was the plan, but if bzo found what's needed, that will be rpc :p
20:04.31phhna the plan was libgpsd
20:04.43phhI've still the problem of reporting bearing to sensors driver
20:05.14bzophh: I'm out of time today
20:05.19bzomore later
20:05.23phhbzo: ok
20:05.39phhbzo: I'll start coding from RPC
20:06.04bzosounds good, I'll check back in this weekend
20:07.03*** join/#htc-linux thedicemaster (~thedicema@e200071.upc-e.chello.nl)
20:11.34*** join/#htc-linux hamagc (~hamagc@173-126-41-254.pools.spcsdns.net)
20:12.46hamagcok so a6d rootfs gives me radio. The 2 most recent lilled my cdma radio
20:13.02hamagcKillled
20:13.16hamagc*
20:13.29stinebdlogcat?
20:14.36hamagcWhere is it or how do make?
20:15.10phhcommandcenter ?
20:15.14phh(or controlcenter ? I never know)
20:15.57hamagcIll upload when I get home to a computer
20:16.36stinebdlol there's a logcat live wallpaper, nice
20:17.27stinebdSendLog on the market can email the logs
20:17.38*** join/#htc-linux Wout (~woutf@s55927d21.adsl.wanadoo.nl)
20:18.27phhstinebd: last time i've seen it, it launches as many logcat instances as you read logcats.
20:18.37phhie you might end with 100s of logcat runnings after some tests.
20:18.42Captnoordstinebd: trying to boot 2.1
20:18.44Captnoordagain
20:18.44Captnoord:P
20:21.59stinebdmake sure you logcat it before it kills init
20:22.11hamagcEmail?
20:22.22Captnoordi'm not a hero with adb
20:23.06Captnoordstinebd: isn't there a way to make logcat log it to sd always
20:23.19stinebdnot that i know of
20:24.00hamagcEmail for logcat? I can do that on the road
20:24.35stinebdbryan@southcape.org
20:24.47Captnoordhmmm
20:24.49Captnoordin short
20:24.55Captnoordhow can I logcat it
20:24.55Captnoord?
20:24.56Captnoord:P
20:25.00Captnoordwindows user
20:25.00Captnoordbleh
20:25.13stinebdadb logcat
20:25.16stinebdoh windows
20:25.17stinebdwho knows
20:25.28stinebdboot into linux and adb logcat ;)
20:25.32Captnoordlol
20:25.44Captnoordhmmmm
20:25.45makkonenadb logcat should work in windows, if adb is working in windows
20:26.44*** join/#htc-linux dcordes (~dccordes@unaffiliated/dcordes)
20:27.35Captnoordhmmmm
20:27.39tmztwhere people from ARM vendors gave us informations about their work on getting support for products in mainline kernel. Looks like 2.6.35 will be used in Maverick release.
20:27.41Captnoordandroid stk downloading
20:27.42Captnoordbleh
20:27.47makkonenshould there be a new automated push of the RIL to the rootfs? I don't see that it's been updated.
20:27.48tmzthttp://marcin.juszkiewicz.com.pl/?p=1681
20:28.01hamagcSent
20:28.05tmztsince apparently this includes snapdragon it should be good news
20:28.55stinebdmakkonen: i updated the build bot's aosp tree and the hash changed for some reason
20:29.05hamagcOk be back later
20:29.54stinebdugh, sendlog doesn't do logcat -b radio
20:30.07stinebd1 star it on the market
20:30.10makkonenah. I see. sounds like that's what killed hamagc's radio, so... that's weird.
20:32.19KhaytsusWhy does the dosfsck get stuck sometimes, even on a clean card?  I've seen it several times, even when I *just* fsck'ed it.
20:32.38makkonenI don't think I've ever seen it get stuck.
20:33.05KhaytsusDoes about 30% of the time for me I think..  last line is dosfsck 3.0.1, 23 Nov 2008, FAT32, LFN
20:33.11KhaytsusAnd nothing else, ever....  until I reboot
20:33.13makkonenhuh.
20:33.15phhhuhu I've never had such problem either
20:35.36KhaytsusHmm, odd..  I wonder why.  I can almost always reboot and pick Android from gen.y to shorten the cycle time (rarely use gen.y otherwise really) and it'll boot right up
20:42.04Heart|is TheBrilliantMistake sometimes here in the channel (he's making the blackstone eclair port)
20:46.23stinebdyeah, ril is broken alright
20:47.37phhgood job stinebd ! :p
20:47.42phhHeart|: there is no "port" to do ?!?
20:48.36Heart|don't know the correct english word sorry ;-)... he's doing http://forum.xda-developers.com/showthread.php?t=621498
20:49.09phhwell it's just a thread ..
20:49.31phhwith a cab file right
20:49.39stinebdphh: see anything useful in here? http://maduin.southcape.org/~bryan/android/deadril.txt
20:49.51stinebdI/RILJ    ( 1290): Couldn't find 'rild' socket; retrying after timeout
20:50.21phhstinebd: sounds like ril is crashing
20:50.42stinebdwhy?
20:51.13phhD/AT      ( 1181): AT< OK
20:51.16phhmeans ril is alive
20:51.20phhI/RILJ    ( 1290): Couldn't find 'rild' socket; retrying after timeout
20:51.23phhmeans ril is unaccessible
20:51.27phheither dead or broken socket
20:51.38phhand ril seems reinited afterwards
20:51.40phhso I'd vote for dead
20:51.49stinebdProcess 1530 attached with 4 threads - interrupt to quit
20:51.49stinebd[pid  1530] read(12,  <unfinished ...>
20:51.49stinebd[pid  1529] recv(-2147433872,  <unfinished ...>
20:51.49stinebd[pid  1528] select(4, [3], NULL, NULL, NULL <unfinished ...>
20:51.49stinebd[pid  1527] setup(
20:51.58stinebdthats all i get out of strace
20:52.14*** join/#htc-linux Neo31 (~Neo31@unaffiliated/neo31)
20:52.40stinebdwe have too much stuff dependent on 2.0, eh?
20:52.44*** join/#htc-linux brandon2004 (~brandon@96-28-73-187.dhcp.insightbb.com)
20:52.44Captnoord-2147433872 == 0xffffffff
20:52.53Captnoordinvalid handle
20:53.18Captnoord8000C270
20:53.19CaptnoordI mean
20:53.25Captnoordstill invalid handle
20:57.39stinebdhttp://maduin.southcape.org/~bryan/android/strace-rild.log
20:58.33stinebdnobody look at my password
20:59.09*** join/#htc-linux balans (~Gebruiker@82-170-215-5.ip.telfort.nl)
20:59.58phhyup.
20:59.59phhno idea.
21:00.11phh1623  recv(1075875584,  <unfinished ...>
21:00.13phhthe problem starts here
21:00.17phhbut hum
21:00.22stinebdyeah
21:00.54Woutnighty night
21:00.59Wouthappy debuggin' :P
21:09.53[acl]acl
21:09.55[acl]is
21:09.57[acl]here
21:10.02[acl]what did i miss ?
21:10.57Captnoordokey
21:11.11Captnoordphh for adb anything I should do in startup.txt?
21:13.48jonpryargh, i get a kernel dump. any idea what this is from/means? http://pastebin.com/r4Fk3d56
21:14.01[acl]ahh fine .. ill read the logs :-)
21:15.03phhjonpry: you killed smd
21:15.04phhwell done
21:15.18phhjonpry: reproductible ?
21:15.28stinebdphh: i have rild build warnings if you're interested
21:15.33stinebds/rild/ril
21:16.21Captnoordstinebd: okey... do I need to change something to use adb
21:16.30Captnoordas it now screams at me that it can't find any devices
21:16.30Captnoordbleh
21:16.31Captnoord:P
21:16.36jonpryhappens everytime, lol. rpc seems to still work though
21:16.38stinebdCaptnoord: don't think so. it has to be plugged on boot, though.
21:16.43jonprynote the reply right after the crash
21:16.45Captnoorddid that
21:16.57stinebdi don't know about windows, though.
21:17.10stinebdit may need the device in a different usb mode
21:18.31*** join/#htc-linux mpannen (~matt@adsl-75-51-3-31.dsl.milwwi.sbcglobal.net)
21:19.42phhjonpry: perhaps it's nothing to worry about ? :p
21:21.08Captnoordhmmmm
21:21.17Captnoordstinebd: what adb module you use
21:21.24CaptnoordI assumed I needed the 2.1 stuff
21:21.26stinebdmodule?
21:21.35Captnoord[23:21] <Captnoord> I assumed I needed the 2.1 stuff
21:21.43Captnoordthe 1.7 api
21:22.08Captnoordhttp://groups.google.com/group/android-beginners/browse_thread/thread/ff713181959c48ee?pli=1
21:22.10Captnoordhmmmm
21:22.14jonprykernel crash doesn't sound like a good way to get anything working
21:22.17stinebdoh, i use some adb that phh gave me
21:22.32phhit's 1.6's
21:22.33Captnoordk
21:22.35Captnoordokey
21:22.39Captnoordgonna download 1.6
21:22.40phh~usb_function_switch
21:22.41aptrumour has it, usb_function_switch is /sys/devices/platform/msm_hsusb/usb_function_switch
21:22.44phhCaptnoord: write 2 to that file.
21:22.50phhmight help win's adb
21:22.56phhCaptnoord: na windows 1.6 adb has the same problem
21:23.04*** join/#htc-linux freakonlegs (~Craigtbal@host86-180-100-216.range86-180.btcentralplus.com)
21:24.04Captnoordphh [23:22] <apt> rumour has it, usb_function_switch is /sys/devices/platform/msm_hsusb/usb_function_switch
21:24.07Captnoordon android?
21:24.09Captnoordin*
21:24.16phhyes
21:24.21Captnoordas I can't get into android
21:24.23Captnoordas it loops
21:24.25Captnoordat boot
21:24.29phhhaha -_-'
21:24.32Captnoordhmmm
21:24.35phhCaptnoord: user.eclair.conf
21:24.36Captnoordmaybe I can hack it in
21:24.38phhin sdcard/conf
21:24.40Captnoordk
21:25.10Captnoordecho /sys/devices/platform/msm_hsusb/usb_function_switch 2
21:25.27Captnoord?
21:25.33Captnoordor just
21:25.34Captnoord=2
21:25.48phhecho 2 >  /sys/devices/platform/msm_hsusb/usb_function_switch
21:26.05Captnoordthanks
21:27.25Captnoordthanks
21:27.26Captnoordbingo
21:27.48Captnoordjust type logcat
21:29.12Captnoordhehe
21:29.13Captnoordk
21:30.38Captnoordits a some sort of java error
21:32.24stinebdhttp://pastebin.com/W1bzvk6t - ril warnings
21:32.54Captnoordxdadevelopers packages borkes it
21:33.09Captnoordis there a way to dump the adb
21:33.12Captnoordbleh
21:33.22Captnoordneed to find it out myself of course
21:33.23Captnoordbleh
21:33.33stinebdwindows should do output redirection
21:33.38Captnoordyea
21:33.40Captnoordhmmm
21:33.47stinebdhttp://www.microsoft.com/resources/documentation/windows/xp/all/proddocs/en-us/redirection.mspx?mfr=true
21:33.49Captnoordadb logcat >> log.txt
21:34.05stinebdoh wait
21:34.11stinebdi saw something on the market about the xda app
21:34.14stinebdit can't be used with JIT
21:34.35CaptnoordoO
21:34.37Captnoordwhy
21:34.38Captnoord:P
21:34.49stinebdno idea
21:35.05Captnoordanyway I want to check it anyway
21:35.08phhthey did it too cleanly
21:35.10stinebdhttp://twitter.com/xdadevelopers/status/13927007092
21:35.47Captnoordyup logcat dumps its content now nicely
21:36.02Captnoordwhat does the app do?
21:36.03Captnoord:P
21:36.11stinebdit's a client for the forums
21:36.16Captnoordlo
21:36.16Captnoordk
21:36.20Captnoordokey crash
21:36.22Captnoordnow I can wait
21:36.24Captnoordtill it reboots
21:36.27Captnoordand then its finished
21:36.27Captnoord:P
21:36.52CaptnoordI/vold    ( 1189): Android Volume Daemon version 2.0
21:36.52CaptnoordW/vold    ( 1189): Volume '/sdcard' already mounted at startup
21:37.09stinebdi hope so
21:37.13stinebdeverything runs off of there
21:37.46CaptnoordD/libEGL  ( 1208): loaded /system/lib/egl/libGLES_android.so
21:37.54CaptnoordI tought it was all software render
21:38.02stinebdthat's the software rendering library
21:38.10stinebdlibGLES_qcom is hardware 3d
21:38.14Captnoordk
21:39.17CaptnoordW/PackageParser( 1208): No actions in intent filter at /system/app/xdaforum.apk Binary XML file line #16
21:39.20Captnoordhere it goes wrong
21:39.30phhCaptnoord: it says blabla died just after that ?
21:39.40stinebdit should die in the status bar
21:39.46stinebdif it's anything like the other guy
21:39.54Captnoordphh there are a couple of more of those messages
21:39.57Captnoordthen I get
21:39.58CaptnoordD/dalvikvm( 1208): GC freed 3849 objects / 390232 bytes in 190ms
21:39.58CaptnoordD/PackageManager( 1208): Scanning app dir /data/app
21:39.58CaptnoordE/System  ( 1208): Failure starting core service
21:39.58CaptnoordE/System  ( 1208): java.lang.NullPointerException
21:39.58CaptnoordE/System  ( 1208): at com.android.server.PackageManagerService.scanDirLI(PackageManagerService.java:1971)
21:39.58CaptnoordE/System  ( 1208): at com.android.server.PackageManagerService.<init>(PackageManagerService.java:590)
21:39.59CaptnoordE/System  ( 1208): at com.android.server.PackageManagerService.main(PackageManagerService.java:335)
21:39.59CaptnoordE/System  ( 1208): at com.android.server.ServerThread.run(SystemServer.java:116)
21:40.00CaptnoordE/AndroidRuntime( 1208): Crash logging skipped, no checkin service
21:40.19phhok
21:40.24phhso xdaforum.apk is clearly the problem
21:40.24stinebdyeah
21:40.28phhI told you it's AndroidApps.
21:40.38stinebdlol
21:40.44phhI wouldn't have bet on xdaforum app though -_-'
21:40.44stinebdhow does that kill the whole system?
21:40.54phhstinebd: android's stupid.
21:41.09phh(it's one of my mottos along with DETU)
21:41.17stinebdpatches welcome
21:41.31phhyeah... no thanks I don't go in that stuff :p
21:41.44phh(you're right on the idea though.)
21:42.10stinebdso anyway, ril
21:42.22stinebdbig problem
21:42.23Captnoordsomeone remove the xda app from the rootfs
21:42.24Captnoord:P
21:42.29stinebdit's not in rootfs
21:42.32stinebdit's on your sd card
21:42.34phhCaptnoord: it's in AndroidApps
21:42.38phhwait
21:42.39phhno
21:42.40phhit's in system
21:42.45stinebd:o
21:42.45phh[23:39:17] <Captnoord> W/PackageParser( 1208): No actions in intent filter at /system/app/xdaforum.apk Binary XML file line #16
21:42.48Captnoordk
21:42.50stinebduggggh
21:42.52Captnoordtrue
21:43.00stinebdgotta stop putting market apps in system.sqsh
21:43.02phhwho the hell putted that in system ?!?
21:43.07phhstinebd: yeah ...
21:43.16phhI had hard time to delete all market apps from it
21:43.21phhand now babijoee reverse it.
21:43.30phh(hard time... well took me 10minutes with your kitchen :p)
21:43.35Captnoordhehe
21:43.47stinebd8 of those minutes in the uml image
21:43.55phh:p
21:44.05phhstinebd: not really, the list of apps is quite long
21:44.13phhand I had to check some many times
21:44.20stinebdyeah i'll be making that more friendly
21:44.22phhI even did a mistake
21:44.37phh(market needing google talk.. what ?)
21:44.50stinebdi got big plans for the kitchen
21:44.57stinebdbut it's gonna take me forever at this rate
21:45.06phhstinebd: webbuilder on xdandroid.com ? :p
21:45.16stinebdprobably
21:45.41Captnoordstinebd: written in what language?
21:45.48stinebdx86 asm
21:45.53Captnoordlol
21:46.05stinebdphp frontend, python backend
21:46.21stinebdi use fastcgi on the server so php goes fast fast fast
21:46.22Captnoordk
21:46.46stinebdand runs as a normal user and all that crap
21:48.32Captnoordmount /media/D_DRIVE/zimage/system.sqsh /mnt/android -t squashfs -o loop
21:48.33Captnoordhmmmmm
21:48.39Captnoordbleh
21:48.57stinebdheh, our squashfs's are read-only
21:49.34Captnoordwhy
21:49.35Captnoord:S
21:49.36Captnoordbleh
21:49.39Captnoordforces
21:49.43stinebdbecause they have to be
21:49.48stinebdthat version doesn't support write
21:49.54Captnoordk
21:49.56stinebdand the kernel doesn't support a newer version
21:50.31jonpryhow do the winmo irq's map to real irq's. like i have 3d, 3e, 3f, and 40?
21:51.06Captnoordjonpry: yup
21:51.22*** join/#htc-linux r0w3zz (~r0w3zz@p54A8F719.dip.t-dialin.net)
21:51.58jonpryyup what?
21:52.11Captnoordthey don't do anything fancy with the irq numbers
21:52.22Captnoordhmmmmm
21:52.31Captnoordwakes up
21:52.33jonprytoo many
21:52.43Captnoord3e time today
21:52.46CaptnoordI should hit the bed
21:52.47Captnoordbleh
21:53.48jonpryshould resolve map to adsp kind of stuff
21:54.28CaptnoordI could have tried to installed the squashfs tools
21:54.28Captnoordbleh
21:55.20jonprywonders if ADSP interrupt numbers are actually correct
21:56.38Captnoordremoves xdaforum.apk
21:56.50phhjonpry: hum
21:56.56phhjonpry: I can't see how it's possibly wrong.
21:57.24phhI mean haret has the correct values
21:57.39phhjonpry: 3d/3e/3f/40 can be the M2A interrupts
21:57.43phh(2-5)
21:58.40jonprymakes no sense though. like int-3b?
22:01.33Captnoordcan be gpio ints
22:01.42Captnoordhttp://www.htc-linux.org/wiki/index.php?title=Raphael_IRQ
22:02.08Captnoordhttp://www.htc-linux.org/wiki/index.php?title=Raphael_GPIO
22:02.12Captnoord3b is unknown
22:03.11jonpryit should be M2A stuff
22:03.37Captnoordhmmm
22:04.03jonpryif i could find out what module oeminterruptenable is in...
22:04.42Captnoordnk.exe
22:05.39*** join/#htc-linux r0w3zz (~r0w3zz@p54A8F719.dip.t-dialin.net)
22:05.40jonpryreally? they compile a custom nk.exe for each platform?
22:05.47Captnoordyup
22:06.03jonpryweird. well that will definately help
22:06.04Captnoordwe got
22:06.05CaptnoordOEMInterruptHandler
22:08.14jonpryi already know how that works
22:09.48Captnoordstinebd: how to re-mksquash the stuff?
22:13.48stinebdCaptnoord: mksquashfs-3.1, but you need to preserve permissions
22:13.57stinebdnot sure if it's possible in windows
22:14.13Captnoordi'm not doing that in windows
22:14.16Captnoordvm
22:14.19stinebdok
22:15.05stinebdeasiest way is to do it all as root then
22:15.37stinebdsudo -s, unsquashfs-3.1 system.sqsh, make your changes, mksquashfs-3.1 squashfs-root system.sqsh.new
22:15.59Captnoordhmmmmm
22:16.01Captnoordokey
22:16.04CaptnoordI think I got the point
22:16.20Captnoordthe sudo part is important
22:16.21*** join/#htc-linux jamenlang (~jamenlang@68-170-60-12.mammothnetworks.com)
22:16.21Captnoordbleh
22:19.36thontcan someone tell me how to collect modules and create modules-2.6.*.tar
22:19.40thont?
22:19.49Captnoordyou just have to rename it
22:20.03thonthow?
22:20.10thontrename the old one?
22:20.17Captnoordpossible
22:20.27Captnoordbut you have build your own kernel?
22:20.33thontyes
22:20.40Captnoordokey
22:20.43thontI created zImage 1 day ago
22:20.46thontbut when I boot up
22:20.47*** part/#htc-linux freakonlegs (~Craigtbal@host86-180-100-216.range86-180.btcentralplus.com)
22:20.48Captnoordcheck the logs
22:20.49thontwith old module
22:20.56Captnoordphh and I have logged it
22:20.57thontwifi not working
22:21.06Captnoordmodule isn't loaded
22:21.14Captnoordbecause it has the wrong name
22:21.15thontI no
22:21.19thontoh
22:21.24thonthow to correct the name then
22:21.29thontwhere is the log?
22:21.33thontin wiki?
22:21.51dcordesKensan, ping
22:21.53Captnoordits in the kernel
22:21.59Captnoordits beeing generated on the fly
22:22.11Captnoordcheck the module git repo
22:22.14stinebdril version mismatch?!
22:22.16Captnoordthere is some shit to be found
22:22.51thontsorry, I dun understand, but I'll try to figure it out
22:23.13stinebdlol
22:23.15Captnoordanother git repo
22:23.16stinebdphh: we need a new kitchen base :(
22:23.22phhstinebd: meh ?
22:23.36stinebdE/RILC    ( 1181): RIL_register: RIL_RadioFunctions * null or invalid version (expected 2)
22:23.41stinebd(2 = RIL_VERSION)
22:23.46thontlibhardware_legacy???
22:23.51stinebdaosp is up to RIL_VERSION=3 now
22:24.10stinebdso the libhtcgeneric-ril that the build bot built has RIL_VERSION=3
22:24.16stinebdand our current system has RIL_VERSION=2
22:25.01stinebdchanging the s_callbacks in RIL_register to give version=2 as opposed to version=RIL_VERSION (3) makes it work
22:25.31stinebdnot sure why they changed it
22:25.33stinebdgood ol' google?
22:25.48phhstinebd: argh
22:26.07stinebd:)
22:28.41*** join/#htc-linux dekar_ (~dekar@port-92-193-47-168.dynamic.qsc.de)
22:30.01Captnoordbleh
22:30.03brandon2004LeTama: I used buildroot to create a rootfs and it boot to sh just fine no errors but how do you issue commands over usb
22:30.07Captnoordremoving squashfs 4
22:30.08Captnoordbleh
22:30.08Captnoord:P
22:30.34Captnoordinstalling 3.1
22:30.46Captnoordbleh
22:30.49Captnoordtakes for ever this
22:30.52CaptnoordI should be in bed
22:30.53Captnoord:P
22:30.54stinebdlocal att data is down
22:30.55stinebd:(
22:31.22LeTamabrandon2004: well, you need network or adb I presume
22:31.26stinebdphh: in the mean time, i'm going to create a local branch to track master of the ril repo, with s_callbacks.version=2
22:31.41stinebdphh: after we do another release, remind me to change that back ;)
22:31.42phhstinebd: if that workaround it ok
22:32.05stinebdit works fine, since all the code in our ril hasn't changed
22:32.05brandon2004LeTama: how would you use adb without android
22:32.16stinebdbut i can't test data
22:32.25Captnoordstinebd: your sure its 3.1
22:32.25stinebdbecause it's not working at all for anybody around here
22:32.28Captnoordand not 3.6
22:32.42Captnoordcaptnoord@captnoord-laptop:~/dev/android$ sudo unsquashfs system.sqsh
22:32.42CaptnoordMajor/Minor mismatch, filesystem on system.sqsh is (3:1)
22:32.42CaptnoordI only support Squashfs 3.0 filesystems!  Later releases will support older Squashfs filesystems
22:32.44Captnoord:P
22:32.48Captnoordas I installed 3.1
22:32.49Captnoordbleh
22:32.51brandon2004LeTama: I used "init=/bin/sh"
22:32.53phhhehe
22:32.54phhmandriva rox.
22:33.01phhI have mksquashfs3 and mksquashfs
22:33.07phhboth does what they are supposed to do
22:33.08Captnoordk
22:34.44LeTamaBrandon2004: hum, ok, don't think you can without a script. I'm using an init script that tries to setup things
22:35.29Captnoordokey
22:35.33Captnoord3.3 works
22:35.37brandon2004LeTama: could I download ur script?
22:35.41LeTamaI'm not there yet too, I managed to launch adbd, but adb says device ID 0000, offline.
22:37.08LeTamaI guess I need more android stuff. I can give it to you, but you won't have much more
22:37.29brandon2004that's fine I just need a starting place
22:37.41*** join/#htc-linux aatreya (~ab438139@gateway/web/freenode/x-pddopmdtneslityu)
22:37.49aatreyaa2dp is stuttering on Android 2.1
22:37.49LeTamayou need to compile adbd as dynamic too
22:37.52aatreyaI'm on a RAPH800
22:37.56aatreyaany ideas why that might be happening?
22:37.56Captnoordhmmm works
22:38.01brandon2004yea I figured that
22:38.31aatreyait only happens on the AOSP build
22:38.49aatreyaon the HTC sense build, it works fine
22:39.39LeTamabrandon2004: here it is: http://pastebin.com/eLwLyjus
22:39.51LeTamalot's of stuff in it not needed
22:39.54phhaatreya: ATM, devs have a pin problem with A2DP.
22:40.10phhbefore even talking about stuttering.
22:40.18Captnoordsame error
22:40.21LeTamagot to run, good luck
22:40.25stinebdheh
22:40.27CaptnoordI think with another packages
22:40.29Captnoord-s
22:40.31brandon2004thanks
22:40.32stinebddamn bluetooth
22:40.36stinebdworks fine on my n1
22:41.05*** part/#htc-linux cj (~cjac@173-10-126-202-BusName-Washington.hfc.comcastbusiness.net)
22:44.46*** join/#htc-linux aatreya (~ab438139@gateway/web/freenode/x-bydtbnfryawuwsav)
22:45.07aatreyaphh: oh, interesting.  pairing works fine for me, if that's what you're referring to
22:45.57stinebddoes it pop up a pin entry screen?
22:46.24Captnoordstinebd: hmmm removed the app
22:46.27Captnoordstill crashing
22:47.58stinebdnice
22:48.13Captnoordhmmm
22:48.24Captnoordmaybe I should try adb with bugreport shit or something
22:52.28*** join/#htc-linux cr2_ (~cr2@ip-109-85-122-27.web.vodafone.de)
22:53.11Captnoordwelcomes cr2_
22:53.36cr2_hi Captnoord
22:53.39phhcr2_: \o/
22:53.42cr2_any good news ? :)
22:53.46phhcr2_: gps ! :p
22:53.53cr2_hi phh
22:53.56cr2_on leo ?
22:53.59stinebdok i'm done for the weekend
22:54.00phherr
22:54.03brandon2004can anyone tell me how to build adbd? like where is it located
22:54.07phhjust good old msm7k
22:54.15cr2_phh: ok
22:54.22cr2_phh: what was the problem ?
22:54.29phhcr2_: i'm still not sure
22:54.35phhI've rewritten everything in userspace
22:54.37PSPdemonbtw phh, what did you do to get maemo ( mer rather ) running on rhobuntu?
22:54.37phhusing some other dumps
22:54.50phhand it changed somehow, the crashed was only 10seconds after get_position
22:55.06phhand jonpry made me think we have to send ACKs to A9
22:55.21brandon2004+
22:55.34cr2_phh: acks to what ?
22:55.44phhcr2_: A9 sends callbacks messages
22:55.49phhon 3100005b
22:56.01cr2_phh: afair we did not get any A9 messages to our registered servers
22:56.55phhcr2_: and afair A9 crashed as soon as we send get_position
22:56.57phhnot 10seconds after it
22:57.05cr2_did you change the cb server registration code ?
22:57.13phh?
22:57.17phhcb server doesn't have to be registered
22:57.35cr2_why ?
22:57.44phhbecause they are callbacks
22:57.54phhcr2_: see android's librpc
22:58.01phhthey say that.
22:58.25cr2_hm. i don't understand it
22:58.43phhcr2_: when you say "register" you mean say to the A9 we have this server right ?
22:59.09cr2_"say" is done by the rpcrouter
22:59.30cr2_but you need to add the server to the local endpoint/server table
22:59.32phhperhaps that's where we were wrong then
22:59.41phh(librpc somehow bypasses rpcrouter)
23:00.02cr2_weird android mess ;)
23:00.09phhyup.
23:00.23phhnow we get nice NMEA sentences :p
23:00.31phhI'm still not sure about AGPS stuff
23:00.41cr2_time to get desire ;)
23:00.55phhpfff
23:01.00cr2_NMEA on smd27 ?
23:01.01phhi'm the only one getting an acer liquid ?
23:01.02phhcr2_: yup
23:01.34cr2_and what do you do on your librpc channel ?
23:01.44cr2_acer liquid, what's that ?
23:02.09phhsnapdragon 760MHz, 256MB memory
23:02.15phha cheap N1 basically.
23:02.24phh(and opensource gps driver ! :p)
23:02.41cr2_eh ?
23:03.08phh?
23:03.11cr2_but desire sucks compared to leo.
23:03.20cr2_<phh> (and opensource gps driver ! :p)
23:03.20phhlol ? just because of the screen ?
23:03.29thonthow to view boot log again?
23:03.33cr2_yeah, and the design.
23:03.52phhcr2_: lol who cares about design ?
23:04.02cr2_thont: dmesg ?
23:04.05thonttried
23:04.11phhcr2_: git://android.git.kernel.org/platform/hardware/qcom/gps.git
23:04.37phhcr2_: am I right thinking only XTRA is needed for AGPS ?
23:04.40thontbut it's not the same as I saw on the screen when the phone is booting
23:04.49cr2_phh: appel sued htc because of leo :) that means it is a really well designed device :)
23:04.55phhlol
23:05.01phhI thought it was android devices they sued
23:05.14brandon2004anyone here working on hd2 I think I just got somewhere with LeTama script he was unable to connect and I just got it too
23:05.53cr2_phh: they are morons anyway.
23:06.19phhcr2_: who ? htc or apple or both ?
23:06.21cr2_phh: XTRA is the data you pick from the qualcomm servers ?
23:06.25cr2_phh: appel
23:06.50phhcr2_: no clue it's a closed source part
23:06.54brandon2004dcordes: I just got connection over adb to hd2
23:07.18cr2_phh: i'm not interested in the spyware part :)
23:07.19tmztxtra.bin is the data from htc
23:07.29phhcr2_: the spyware part is open -_-'
23:07.32thont"version magic '2.6.27.46-01072-gfcfddc9 preempt mod_unload ARMv6 ' should be '2.6.27.46-01088-gef24e75"
23:07.38thontdo I need to rename the file?
23:07.41Captnoordyup
23:07.44phhcr2_: see the sources, the *_ni.cpp files
23:07.45thontthanks
23:07.46thontfinally
23:07.47phhit's the spyware part :p
23:08.02tmztyou not really providing more than mcc though I think
23:08.04tmztor mnc
23:08.12dcordesbrandon2004, very good. did you compile the binaries used yourself?
23:08.23tmztit's just a binary with the rest of almanac/ephemeris
23:08.26brandon2004dcordes: yes
23:08.28cr2_ok
23:08.29tmztand it really helps at least on tp2
23:08.35thontcan I know where is the version name come from?
23:08.44thontversion magic
23:09.00cr2_phh: so gps works on all 7201A now ?
23:09.00brandon2004dcordes: I used LeTama's init script though
23:09.08phhcr2_: not 6150
23:09.15phhcr2_: we don't have an NMEA channel on them
23:09.19phhso we need to read data from RPC
23:09.23phhbut it's almost done too
23:09.32dcordesbrandon2004, I'm working on fixing openembedded to use the compiler flags LeTama (the ones used by kernel) suggested on the mailing list.
23:10.23cr2_phh: it's 7501A
23:10.31dcordesphh, did you document the commands somewhere? I think I asked for it few days ago but I was on the go so I don't know if you replied.
23:10.35phhcr2_: right.
23:10.41dcordescr2_, hi
23:10.52phhdcordes: it's hard to document anything when I just use dumped values /o\
23:10.56cr2_phh: the same 1 and 4 fragmented CB messages like on kaiser ?
23:10.59brandon2004dcordes: thoses aren't the one's I used, I used the google toolchain to build the kernel, and then I used buildroot I can give you my makefile
23:10.59cr2_dcordes: hi
23:11.10phhcr2_: mmm ?
23:11.18phhI've seen nothing fragmented
23:11.27phhsome POSITION message
23:11.30phhstart pd_event
23:11.32phhend pd_event
23:11.34phhend pd_session
23:11.51phhsome other messages I haven't understood yet (satellite infos I guess)
23:12.16Captnoordstill doesn't work
23:12.17Captnoordthen
23:12.22Captnoorddisable jit
23:13.01cr2_phh: there are CB messages 1 and 4 (must be in wiki afair)
23:13.10phhcr2_: yes
23:13.19phhit's the 1 that contains fix infos
23:13.23tmztcr2_: looks like ubuntu might be working on a non-android kernel for snapdragon
23:13.36cr2_tmzt: for which amss ?
23:13.37dcordesbrandon2004, is buildroot some part of android?
23:13.49tmzt2.6.35 kernel
23:13.54phhcr2_: it contains info on fixed satellites too so it's a bit weird
23:13.54tmztah, probably a smartbook one
23:13.54cr2_phh: ok, maybe.
23:14.21cr2_phh: but these messages are longer than 512bytes, so they are fragmented. afair
23:14.41phhcr2_: ah fragmented from the rpcrouter point of view
23:14.42phhright
23:14.49cr2_but they hacked them on kaiser somehow
23:14.53brandon2004dcordes:http://buildroot.uclibc.org/
23:14.55dcordestmzt, if I were a big company selling qsd laptops I would also ship with ubuntu
23:14.56cr2_so it may be a good reference.
23:15.10phhcr2_: we have no fragmentation problem anyway
23:15.19cr2_ok
23:15.45phhonly remaning problems are AGPS and 6150
23:15.45dcordesbrandon2004, ok can you send me that makefile?
23:15.51phhand perhaps NI if someone wants it
23:15.57phhbut err... no thanks :p
23:16.11phh(that's the spyware part)
23:16.17tmzthttps://code.launchpad.net/~ubuntu-branches/ubuntu/lucid/linux-qcm-msm/lucid
23:16.20*** join/#htc-linux freakonlegs (~Craigtbal@host86-180-100-216.range86-180.btcentralplus.com)
23:16.31cr2_phh: what about tvout ?
23:16.43phhcr2_: I did some quick tries but it just crashes the kernel
23:16.52phhthere is a codeaurora driver for 2.6.29 or 2.6.32
23:16.55tmztphh: agps is really download only
23:16.56phhbut annoying to backport
23:17.02phhtmzt: ok
23:17.05tmztit doesn't send location to network and back
23:17.07phhtmzt: no position injection ?
23:17.23tmztanymore than what's already reported to network, but that's just mcc/mnc
23:17.42phhnot cellid ?
23:17.54tmztyou shouldn't have fine position reported except to emergency services
23:18.00tmztwell, yeah, I would think cellid
23:18.04tmztbut that's not in gps code anyway
23:18.06phhI know that
23:18.08phhthat part is NI
23:18.23tmztah ok
23:18.50tmztthe qualcomm agps (xtra/gpsone) just speeds up the fix
23:19.01tmztbut the data is downloaded and sent to the gps firmware
23:19.12brandon2004dcordes: makefile is at http://pastebin.com/Kw7iLexe .config is at http://pastebin.com/GmRKP22G
23:19.42phhtmzt: but approximate position should help a lot in a quick fix ?
23:19.47cr2_tmzt: the cellid may provide a good approximation for your location
23:19.59tmztyeah
23:20.02cr2_tmzt: so it makes searching for the satellites easier
23:20.08tmztbut I don't know if there's a local blob or not
23:20.13tmztor if there's just xtra.bin
23:20.18cr2_and downloading the ephemeris anyway
23:20.40phhwell there is a pdsm_inject_position function in android
23:20.42tmztI do know that not updating the gps on my tp2 makes fixes take forever
23:20.43cr2_xtra.bin depends only on time
23:20.48tmztwith it they're are instant
23:20.51phhand pdsm_use_this_position in wince's rpc
23:21.08tmztit injects the position into the gps firmware
23:21.09dcordesbrandon2004, thanks can you point me to where the compiler flags are? maybe it has some debug output so you can see them while building.
23:21.11phhI should try with AGPS off and xtra on
23:21.15tmztas well as the contents of xtra.bin
23:21.24cr2_if you have the ephemeris and your approximate location, you know where are the satellites
23:21.55cr2_phh: does it boot with "gps on" now ?
23:22.00phhcr2_: no
23:22.11phhcr2_: but agps and xtra data remains :p
23:22.26phhso we can get a quick fix in wimo, then quit the app then boot linux and we'll have quick fix in linux just after that
23:22.49dcordesbrandon2004, did you try to build the same stuff with a toolchain other than google one?
23:22.52dcordesbrandon2004, e.g. codesourcery
23:23.24brandon2004dcordes: yes and the code sourcery produced a kernel panic
23:23.58brandon2004dcordes, but cortex-a8 and above must use gcc-4.4.x
23:24.06cr2_phh: on tp2 ?
23:25.56tmztwell, the changelog on this ubuntu stuff makes little sense
23:26.04tmztit appears it's going to be the chrome amss though
23:26.21brandon2004dcordes, the compiler options are varibles in the makefile not sure what it used but if you build with that make and config it will show the compiler flags
23:29.33tmzthttp://bazaar.launchpad.net/~ubuntu-branches/ubuntu/lucid/linux-qcm-msm/lucid/revision/1/arch/arm/mach-msm/smd_nmea.c#arch/arm/mach-msm/smd_nmea.c
23:29.40tmztnot sure if this is relevant
23:30.13thontI renamed the modules***.tar.gz already, but there is still version magics mismatch error
23:30.17thontany suggestion?
23:30.30thontdo I nid to recompile the module?
23:31.07*** join/#htc-linux r0w3zz (~r0w3zz@p54A8F719.dip.t-dialin.net)
23:31.37dcordesbrandon2004, I've been building most of the leo kernels with gcc-4.3.2
23:32.20dcordesbrandon2004, well with arm-2008q3 toolchain from cs that has gcc-4.3.2
23:32.27brandon2004dcordes, yes and your kernels havent worked for me...I always have to use my own builds
23:32.42phhcr2_: TP/TD/blac/TP2/TD2.
23:33.02dcordesbrandon2004, I will put that bold into wiki
23:33.22brandon2004dcordes, I think that would be best
23:33.37leobaillardcr2_: but after testing, without launching an app in winmo before booting, I get a fix in ~3minutes, which is ok
23:34.09phhleobaillard: because you're lucky.
23:34.18leobaillardand that too :-°
23:34.23phhlast time I tried, it arn for 30minutes without getiing anything
23:34.47leobaillarddid you try on your toilet seat ? maybe that's your answer
23:35.04phhon my opened window.
23:35.48leobaillardmy window was closed and I'm in a dense urban area
23:36.14dcordesbrandon2004, do we a cpu in the qsd8250 that uses the exact cortex-a8 instruction set?
23:37.00brandon2004dcordes, no but it uses scorpion which is a slightly modified cortex-a8
23:37.09*** join/#htc-linux teknojunkey (~e@cpc5-broo7-2-0-cust239.know.cable.virginmedia.com)
23:37.43dcordesbrandon2004, is there any 'official' documentation on this?
23:37.44brandon2004dcordes, but all binaries I have compiled based on the a8 have worked
23:37.57tmztwhat is this???
23:37.59tmzthttp://bazaar.launchpad.net/~ubuntu-branches/ubuntu/lucid/linux-qcm-msm/lucid/revision/1/arch/arm/mach-msm/nohlt.c#arch/arm/mach-msm/nohlt.c
23:38.33dcordes<PROTECTED>
23:39.36Kensandcordes: pong
23:39.57brandon2004dcordes: http://en.wikipedia.org/wiki/Snapdragon_(processor) it's in the first paragraph under general info
23:40.26teknojunkeyhello, does anyone have any androids for the touch pro 2...  i have a fresh factory reset
23:40.44phhteknojunkey: google ?
23:40.47tmztwhich tp2?
23:41.05tmztI have a link to the forum thread but I haven't quite figured out what to download
23:41.07phhtmzt: both are in same thread anyway
23:41.25tmzthttp://forum.xda-developers.com/showthread.php?t=627997
23:41.41*** join/#htc-linux darkstar62 (~darkstar6@97-126-97-88.tukw.qwest.net)
23:41.46tmztsame zImage now too? just xdandroid with the cdma=1 and no sim boot
23:42.27phhtmzt: there is a huge "download now" link :p
23:42.27dcordesbrandon2004, ok. does that buildroot use any *neon* compile flags when you set ARM_CORTEXA8 ?
23:42.27teknojunkeyiv tried a couple (cant remember which) but they all fail a .net error .. strange cause its win6.5
23:42.46dcordesKensan, fixing to configure qsd correctly in openembedded
23:43.01phhand I don't think it's xdandroid,  I think it's the system.sqsh from stinebd's kitchen
23:43.12phh(ie the system.sqsh that isn't broken.)
23:43.49Kensanis trying to catch up by reading the backlog
23:43.56Kensandcordes: lot of activity here...
23:44.07brandon2004dcordes, not sure is that a problem
23:44.24dcordesbrandon2004, I don't know
23:44.29tmztthere's a download link in the first post? I don't see it
23:44.34Kensandcordes: why do we need gcc 4.4 at least?
23:44.53tmzthere's the rootfs with gps http://xdandroid.southcape.org/rootfs/rootfs-latest.zip
23:45.01tmztthat's why I thought it was xdandroid
23:45.01phhtmzt: "download now"
23:45.03phhimage
23:45.40phhat the end of the post
23:45.40tmztokay, thought that was an ad
23:45.40dcordesKensan, I don't know
23:45.40brandon2004Kensan, because it supports the newer arm processors
23:45.40Kensanwhat's the ubuntu linux-qcm-msm stuff?
23:45.40phhtmzt: lol
23:45.40phhthat happens.
23:45.41tmztstill trying to figure that out
23:46.05Kensanbrandon2004: so the qsd8250 is not properly supported in older compilers I see.
23:46.48thontplz help ... I already renamed the modules-2.6.27.46-01072-gfcfddc9.tar to modules-2.6.27.46-01088-gef24e75.tar already, but there is still version magics mismatch error: version magic '2.6.27.46-01072-gfcfddc9 preempt mod_unload ARMv6 ' should be '2.6.27.46-01088-gef24e75 preempt mod_unload ARMv6 '
23:46.49Kensanbrandon2004: but older compilers should still be able to generate compatible object code, no?
23:46.55brandon2004Kensan, I am almost positive that have been the source of all the problems, none of dcordes, kernels builds worked for me I always had to recompile
23:47.21brandon2004Kensan, yes its just the cross compile
23:47.41Kensanthont: your modules do not match the kernel so you need to recompile.
23:48.20Kensanbrandon2004: can you elaborate?
23:48.23thontwhere is the repo for modules source?
23:49.06thontKensan: libhardware_legacy?
23:50.20tmztkernel modules? no
23:50.26brandon2004Kensan, I honestly haven't created any object code using anything under 4.4 because that what came on my ubuntu distro but as long as you use 4.4.0 toolchain from google to cross_compile, I don't see why 4.3 wouldn't work
23:50.41thonttmzt: what do u mean?
23:50.42tmztmodules are part of the kernel in most cases, though the wireless might be separate
23:51.16thonttmzt: then how I can collect them and create modules*.tar?
23:51.50Kensanbrandon2004: hm... that's what I am thinking. Imho it's more likely that the compiler flags are the source for the issues...
23:52.12Kensanthont: where did you download the kernel from?
23:52.27thontKensan: I compiled it
23:52.37thontKensan: latest source
23:53.14Kensanthont: well when you compile your kernel you should also have compiled the modules.
23:53.29Kensanthont: that is unless you have disabled all modules and built everything in.
23:53.45thontKensan: sure, I compiled them
23:53.59thontbut then what modules***.tar is used for?
23:54.04brandon2004Kensan, that very well could be true, however if 4.4 works just fine with no problems why waste time fiddling with old compiler options
23:54.39Kensanbrandon2004: well it would be nice to know *why* the issue arises to understand the nature of the problem.
23:55.10Kensanbrandon2004: but obviously if 4.4 works fine then you can just use that and not worry about it any more ;)
23:55.28Kensanbrandon2004: can you also run statically linked binaries without problems?
23:56.08Kensanbrandon2004: if I recall correctly LeTama was able to run dynamically linked programs but not static stuff.
23:56.15brandon2004Kensan, I do know for sure that for cortex a8 you have to use 4.4, and according to wikipedia it says the scorpion core in the snapdragon is based of the a8 so...that is why I stuck
23:56.57Kensanbrandon2004: ok. well that's a valuable information.
23:57.11brandon2004Kensan, I though LeTama's problem was with dynamically linked files, and that he had to use statically linked files, which is a problem cause it results in much larger binaries
23:58.16dcordesbrandon2004, I thought it was the other way around
23:58.52thontKensan: does the latest kernel need the module**.gz anymore?
23:59.11thontKensan: I saw a line in the log that: module is already unpacked for this kernel version
23:59.46tmztcr2_: http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/lucid/linux-qcm-msm/lucid/revision/1/arch/arm/mach-msm/internal_power_rail.c#arch/arm/mach-msm/internal_power_rail.c

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.