| 00:12.28 | *** join/#htc-linux PsWii60 (i=pswii60@has.r0oted.info) |
| 00:13.06 | cr2 | does not boot anymore :) |
| 00:13.09 | dcordes | cr2, did you try ppp.sh on raphael ? |
| 00:13.15 | dcordes | hm |
| 00:14.04 | cr2 | no |
| 00:14.50 | dcordes | http://www.youtube.com/watch?v=AkbeDCpKeAM |
| 00:16.19 | dcordes | msm8250 device |
| 00:16.24 | *** join/#htc-linux thedicemaster (n=thedicem@j89051.upc-j.chello.nl) |
| 00:17.25 | dcordes | quite a buggy kernel |
| 00:17.49 | dcordes | problems with hsusb ? |
| 00:19.15 | cr2 | no, changed the driver name |
| 00:19.39 | dcordes | uh oh poor rootfs in nand seems screwed |
| 00:23.23 | cr2 | fixed the boot |
| 00:25.12 | cr2 | [ 5.290771] mddi: registering panel TC358720XBG |
| 00:25.13 | cr2 | [ 5.290832] mddi: publish: |
| 00:25.14 | cr2 | [ 5.291473] mddi_toshiba_probe |
| 00:25.16 | cr2 | [ 5.291687] vsync on gpio 97 now 0 |
| 00:25.18 | cr2 | [ 5.292419] toshiba GPIODATA=0x000000c0 panel_id=0 at toshiba_mddi_enable |
| 00:25.19 | cr2 | [ 5.292510] found panel_id=0 at toshiba_mddi_enable, panel=Hitachi |
| 00:25.30 | *** join/#htc-linux Bry8Star (n=Bry8Star@70.239.151.41) |
| 00:25.40 | dcordes | cr2, you pull the log from usbnet? |
| 00:25.43 | cr2 | need somebody with diam (sharp) to check it |
| 00:25.57 | cr2 | ssh over usb |
| 00:27.14 | cr2 | [ 5.290283] XC=1 |
| 00:27.18 | dcordes | did you ever try ppp ? |
| 00:27.32 | dcordes | I dont't mean the script but in general on raph |
| 00:27.50 | cr2 | only manually |
| 00:27.56 | dcordes | NOTE: Angstrom DOES NOT support bluez-utils because bluez-utils 3.x has been replaced by bluez4 |
| 00:28.00 | cr2 | i need to rip the umts card from nc10 |
| 00:28.05 | dcordes | successfully? |
| 00:28.23 | cr2 | yes |
| 00:28.36 | cr2 | it was when i've fixed umts data channel |
| 00:28.56 | cr2 | what is bluez4 ? |
| 00:29.11 | cr2 | compile bluez4 |
| 00:29.18 | cr2 | i need hcitool |
| 00:34.46 | *** join/#htc-linux dcordes_ (n=luke@unaffiliated/dcordes) |
| 00:39.46 | dcordes_ | cr2, any patches before I take nap? |
| 00:41.32 | cr2 | wait 1 min :) |
| 00:43.10 | dcordes_ | ok regarding the images. OE had some hickup and I rebuild some now |
| 00:44.05 | dcordes_ | I will just add bluez4 and see if it provides hcitool. if not I will add it the normal way |
| 00:44.29 | cr2 | ok |
| 00:44.40 | cr2 | i'm recompiling to check if i broke something |
| 00:45.52 | cr2 | need to fix the lcd on/off and gps |
| 00:48.45 | |Jason8| | hey cr2 |
| 00:48.45 | cr2 | brf6100 |
| 00:48.46 | cr2 | brf6150 |
| 00:48.48 | cr2 | brf6300 |
| 00:48.51 | |Jason8| | great job with android and linux :D |
| 00:48.57 | cr2 | hm. where does this hciattach come from ? |
| 00:49.09 | cr2 | hi |Jason8| |
| 00:49.36 | cr2 | <PROTECTED> |
| 00:49.40 | cr2 | it is smart |
| 00:50.00 | dcordes_ | hm= |
| 00:50.07 | cr2 | the .bts version comes directly from BT firmware |
| 00:50.15 | *** join/#htc-linux stickboy (n=anonymou@cpe-74-68-145-150.nyc.res.rr.com) |
| 00:50.32 | dcordes_ | ok |
| 00:50.46 | cr2 | dcordes_: making a diff |
| 00:52.44 | cr2 | dcordes_: https://privatepaste.com/648xppegEP |
| 00:53.30 | cr2 | nothing substantial there. just a code cleanup |
| 00:56.47 | dcordes_ | cr2, ok good night |
| 01:00.47 | *** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a) |
| 01:01.51 | cr2 | hehe. broken the diamond. |
| 01:01.55 | cr2 | and maybe others |
| 01:03.05 | cr2 | dzo: i've changed the mddi client driver name from "mddi_c_d263_0000" to TC358720XBG |
| 01:03.49 | cr2 | dzo: it will break the diamond panel code. |
| 01:04.05 | cr2 | good night |
| 01:05.43 | *** join/#htc-linux AstainHellbring (n=AstainHe@unaffiliated/astainhellbring) |
| 01:39.08 | *** join/#htc-linux rashire (n=ed1112wa@pool-98-114-209-5.phlapa.fios.verizon.net) |
| 01:53.11 | *** join/#htc-linux BHSPitLappy (n=BHSPitLa@unaffiliated/bhspitmonkey) |
| 01:57.18 | tmzt | dcordes: can you push the thing for you on people? |
| 01:57.40 | tmzt | for dcordes |
| 01:58.25 | tmzt | I'm also working on a new db driver |
| 01:58.33 | *** join/#htc-linux rashire (n=ed1112wa@pool-98-114-209-5.phlapa.fios.verizon.net) |
| 01:58.57 | tmzt | starting by adding single buffer updates to the existing one |
| 01:59.17 | tmzt | but that causes tearing in my test |
| 01:59.51 | tmzt | but I think I figured out how to adapt via drm driver |
| 02:09.31 | *** join/#htc-linux PsWii60 (i=pswii60@has.r0oted.info) [NETSPLIT VICTIM] |
| 02:09.31 | *** join/#htc-linux root2 (n=root@rgnb-5d87473b.pool.mediaWays.net) [NETSPLIT VICTIM] |
| 02:09.31 | *** join/#htc-linux bertramt__ (n=chatzill@63.246.89.17) [NETSPLIT VICTIM] |
| 02:09.31 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-67-49-192-228.bak.res.rr.com) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux tmzt (n=tmzt@adsl-99-52-65-233.dsl.akrnoh.sbcglobal.net) |
| 02:09.32 | *** join/#htc-linux DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux Dinde (i=kayser@81-65-176-209.rev.numericable.fr) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux madCoder` (n=madcoder@c-68-42-30-56.hsd1.pa.comcast.net) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux bartman (n=bart@tau.jukie.net) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux dtrack (n=LOL@78-86-59-180.zone2.bethere.co.uk) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux astrolite (n=astrolit@c-d7c9e455.023-352-6c6b701.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux Perkka (n=perkka@ua-83-227-207-92.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux Kevin2 (n=Kevin2@207-237-194-161.c3-0.avec-ubr2.nyr-avec.ny.cable.rcn.com) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux KungFunk (n=cow@c-e1d571d5.014-46-73746f28.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux ecze (n=ecze@eczema.ecze.com) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux darkstar62 (n=darkstar@97-126-104-157.tukw.qwest.net) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux jmaurice (i=jmaurice@snickers.org) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux high-rez (n=gus@carrera.bourg.net) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux thinko (n=jdoe6alp@smaug.rackdragon.com) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux nozze (n=nozze@u193-11-162-41.studentnatet.se) [NETSPLIT VICTIM] |
| 02:09.32 | *** join/#htc-linux infernix (i=nix@unaffiliated/infernix) |
| 02:28.19 | tmzt | s/db/fb/ |
| 02:59.41 | *** part/#htc-linux PsWii60 (i=pswii60@has.r0oted.info) |
| 03:05.14 | *** join/#htc-linux root2 (n=root@rgnb-5d874278.pool.mediaWays.net) |
| 03:17.21 | *** join/#htc-linux MethoS- (n=clemens@134.102.106.250) |
| 03:31.43 | *** join/#htc-linux MrPippy_ (n=pip@adsl-75-36-57-196.dsl.sndg02.sbcglobal.net) |
| 03:44.47 | parmaster | yo tmzt |
| 04:30.04 | *** join/#htc-linux droid0011 (n=g1@p4FDCC5FE.dip.t-dialin.net) |
| 04:32.24 | *** join/#htc-linux sd00 (n=smac@82-39-194-231.cable.ubr02.jarr.blueyonder.co.uk) |
| 04:32.55 | *** join/#htc-linux leviathan (n=quassel@c-82-192-226-27.customer.ggaweb.ch) |
| 05:20.12 | *** join/#htc-linux rashire2 (n=ed1112wa@pool-98-114-209-5.phlapa.fios.verizon.net) |
| 06:59.59 | *** join/#htc-linux root2 (n=root@rgnb-5d87dfcd.pool.mediaWays.net) |
| 07:00.40 | *** join/#htc-linux Echo31 (n=olivier@mir31-4-82-240-194-54.fbx.proxad.net) |
| 07:19.44 | *** join/#htc-linux timebomb (n=tb@f054021101.adsl.alicedsl.de) |
| 07:27.16 | *** join/#htc-linux kiozen (n=kiozen@rgnb-5d87c591.pool.mediaWays.net) |
| 07:35.19 | *** join/#htc-linux kvaster (n=kvaster@live.bn.by) |
| 08:10.03 | *** join/#htc-linux Captnoord (n=Captnoor@dc5147a47b.adsl.wanadoo.nl) |
| 08:42.05 | *** join/#htc-linux MLM (n=mlvdmeid@5ED0BCBD.cable.ziggo.nl) |
| 08:49.49 | *** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl) |
| 08:56.16 | tmzt | parmaster: hey |
| 08:57.23 | townkat | hi |
| 08:57.32 | townkat | tmzt: u have a blackstone ? |
| 08:57.38 | tmzt | no |
| 08:57.46 | tmzt | cdma raph |
| 08:58.46 | tmzt | what works on blak now? |
| 08:59.03 | townkat | in linux or android ? |
| 08:59.16 | tmzt | ? |
| 08:59.34 | townkat | android is unchanged from 2009-03-08 |
| 08:59.42 | tmzt | kernel |
| 08:59.44 | townkat | in heard linux have sound |
| 08:59.57 | tmzt | should be in android too |
| 09:00.03 | tmzt | just build from git |
| 09:00.23 | townkat | i can try to test now |
| 09:00.38 | tmzt | can you build it? |
| 09:00.52 | townkat | i just build it last knight for the first time |
| 09:00.57 | townkat | i folowed the steps from wiki |
| 09:01.00 | townkat | is it ok ? |
| 09:01.04 | townkat | http://wiki.xda-developers.com/index.php?pagename=BlackstoneLinux |
| 09:01.24 | tmzt | to build from git? |
| 09:01.38 | townkat | OPTIONAL |
| 09:01.38 | townkat | If you want to compile your own kernel, follow these steps (on linux): |
| 09:01.49 | tmzt | ok |
| 09:02.16 | townkat | but i can try with your zimage too if u can give me |
| 09:02.32 | townkat | bu this only give zimage, i dont know where to get the rest of files |
| 09:02.44 | tmzt | I don't have one for blak |
| 09:02.50 | tmzt | use the ones you have |
| 09:03.00 | townkat | the old ones? from 03-03 ? |
| 09:03.41 | tmzt | from when? |
| 09:03.55 | townkat | <PROTECTED> |
| 09:04.09 | tmzt | march? |
| 09:04.26 | tmzt | see if you can find newer ones |
| 09:04.34 | tmzt | prorbaly dzo |
| 09:04.44 | tmzt | just use blak mtype |
| 09:04.49 | townkat | cr2 tol dme to get a diamond one |
| 09:04.55 | tmzt | yes |
| 09:05.08 | townkat | i dont know what mtype is |
| 09:05.11 | tmzt | but change default.txt to use blak mtype |
| 09:05.15 | townkat | ahh |
| 09:05.17 | tmzt | set mtype |
| 09:05.21 | townkat | just mtype should be changed ? |
| 09:05.25 | tmzt | yes |
| 09:05.28 | townkat | k |
| 09:10.04 | townkat | hmm, mtype is the same as diamond, 2030 |
| 09:10.31 | townkat | or not, i already changed it :) |
| 09:32.08 | townkat | it hang at Haret: Booting linux screen |
| 09:32.28 | townkat | i left it for 5 minutes |
| 09:34.12 | Captnoord | townkat last build? |
| 09:34.19 | Captnoord | cr2 broke the diamond build |
| 09:34.56 | Captnoord | 03:01cr2hehe. broken the diamond. |
| 09:34.57 | Captnoord | 03:01cr2and maybe others |
| 09:34.57 | Captnoord | 03:03cr2dzo: i've changed the mddi client driver name from "mddi_c_d263_0000" to TC358720XBG |
| 09:34.57 | Captnoord | 03:03cr2dzo: it will break the diamond panel code. |
| 09:34.57 | Captnoord | 03:04cr2good night |
| 09:35.17 | townkat | :) |
| 09:35.54 | townkat | no, i buit it for blackstone before that time |
| 09:35.59 | Captnoord | k |
| 09:41.37 | *** join/#htc-linux mrmoku|a` (n=mrmoku@ppp-93-104-80-241.dynamic.mnet-online.de) |
| 09:41.52 | townkat | diam startup.txt have set RAMSIZE 0x08000000 |
| 09:42.08 | townkat | mi old txt have set ramsize 0x6000000 |
| 09:42.09 | townkat | set ramaddr 0x10000000 |
| 09:42.29 | townkat | but it does not seem to start with either one |
| 10:04.33 | *** join/#htc-linux kvaster (n=kvaster@live.bn.by) |
| 10:08.38 | phh | where can i find dcordes initrd, cr2 seems to love ? |
| 10:21.43 | toer | http://linuxtogo.org/~lgorris/ |
| 10:22.16 | phh | thanks |
| 10:34.08 | *** join/#htc-linux MethoS- (n=clemens@134.102.106.250) |
| 11:15.02 | *** join/#htc-linux tsdogs (n=tsdogs@tsdogs.metalit.net) |
| 11:16.04 | *** join/#htc-linux GNUtoo (n=gnutoo@host134-152-dynamic.51-79-r.retail.telecomitalia.it) |
| 11:38.22 | *** join/#htc-linux pleemans (n=toi@d54C2A96D.access.telenet.be) |
| 11:41.54 | phh | hum, aceelerometer works on my diam100 with only uncommenting it |
| 11:56.31 | phh | (and angstrom is really nice.) |
| 12:16.51 | *** join/#htc-linux thedicemaster2 (n=thedicem@24.132.89.51) |
| 12:59.13 | *** join/#htc-linux Zinbolic (n=zinbolic@84.238.80.215) |
| 13:49.01 | *** join/#htc-linux GNUtoo (n=gnutoo@host134-152-dynamic.51-79-r.retail.telecomitalia.it) |
| 14:23.51 | *** join/#htc-linux cr2 (n=cr2@109.84.130.177) |
| 14:25.57 | cr2 | ./arch/arm/mach-msm/board-htcblackstone-panel.c: .name = "mddi_c_d263_0000", |
| 14:26.05 | cr2 | ./arch/arm/mach-msm/board-htcdiamond-panel.c: .name = "mddi_c_d263_0000", |
| 14:26.14 | cr2 | ./arch/arm/mach-msm/board-trout-panel.c: .name = "mddi_c_d263_0000", |
| 14:26.15 | cr2 | ./arch/arm/mach-msm/board-sapphire-panel.c: .name = "mddi_c_d263_0000", |
| 14:26.40 | cr2 | the last 2 are not important, but the first 2 are broken now |
| 14:27.20 | phh | cr2: have you some uncommited patch about bluetooth ? |
| 14:27.38 | cr2 | phh: for which phone ? |
| 14:27.43 | phh | raph |
| 14:27.55 | cr2 | it should work with the head |
| 14:28.01 | phh | ok |
| 14:28.14 | phh | so diam will need some other work :( |
| 14:28.30 | cr2 | change |
| 14:28.32 | cr2 | ./arch/arm/mach-msm/board-htcdiamond-panel.c: .name = "mddi_c_d263_0000", |
| 14:28.37 | cr2 | the .name to |
| 14:28.46 | cr2 | TC358720XBG |
| 14:28.53 | phh | yes i already did that |
| 14:29.02 | cr2 | ok |
| 14:29.03 | phh | the build work, everything work as expected |
| 14:29.08 | phh | but hciattach will timeout |
| 14:29.44 | cr2 | you need to apply power |
| 14:30.03 | cr2 | echo 1 > /sys/class/rfkill/rfkill0/state |
| 14:30.06 | cr2 | to enable |
| 14:30.08 | cr2 | or |
| 14:30.13 | cr2 | echo 0 > /sys/class/rfkill/rfkill0/state |
| 14:30.17 | cr2 | to disable |
| 14:31.19 | cr2 | for blac100 the |
| 14:31.22 | cr2 | <PROTECTED> |
| 14:31.28 | cr2 | is wrong anyway |
| 14:34.18 | phh | ok i have to rebuild with rfkill |
| 14:37.46 | phh | works :) |
| 14:39.31 | phh | (hciattach at least) |
| 14:40.25 | phh | now have to build some bluez tools... |
| 14:40.46 | cr2 | bluez4 says dcordes |
| 14:41.57 | ali1234 | bluez4 is tricky |
| 14:42.06 | phh | it needs dbus :/ |
| 14:42.15 | ali1234 | yep. you'll probably need the old bluez3 utils to actually do anything |
| 14:42.21 | ali1234 | ie pand |
| 14:42.44 | phh | i think things like rfcomm may work |
| 14:42.46 | ali1234 | you can check if its working with hciconfig hci0 and hcitool scan |
| 14:42.48 | cr2 | ali1234: i fear that it will pull a lot of crap with it |
| 14:43.01 | ali1234 | yeah it does if you consider dbus crap... |
| 14:43.05 | cr2 | like gtk dependency and such ;) |
| 14:43.10 | ali1234 | glib |
| 14:43.13 | ali1234 | not gtk |
| 14:43.34 | cr2 | it was gtk+x11 in the old days |
| 14:43.45 | phh | oO |
| 14:43.49 | phh | maybe not that much dependencies |
| 14:43.55 | ali1234 | you dont need bluez for android afaik |
| 14:44.02 | ali1234 | only hciattach |
| 14:44.12 | cr2 | ali1234: i need hcitool |
| 14:44.12 | ali1234 | bluez-utils i mean |
| 14:44.24 | ali1234 | hcitool should come with hciattach? |
| 14:44.26 | cr2 | i don't use android, as you know |
| 14:44.28 | phh | cr2: i built hcitool if you want |
| 14:44.36 | cr2 | at least it's not in the initramfs |
| 14:45.01 | ali1234 | i don't use android either |
| 14:45.03 | cr2 | phh: rfcomm ? |
| 14:45.11 | phh | cr2: too |
| 14:45.12 | ali1234 | you guys should look at gizard aka gentoo armel :) |
| 14:45.14 | phh | don't know if it works |
| 14:45.24 | phh | i have to find my bluetooth usb dongle ... |
| 14:46.00 | cr2 | ali1234: i like ncurses interfaces. they cut the crazy stuff down |
| 14:46.27 | cr2 | and for the insane gui stuff i prefer qt4 |
| 14:46.47 | ali1234 | we looked at most every phone type distro and they all suck :( |
| 14:46.48 | cr2 | but it's all mvho |
| 14:46.50 | phh | scanning works \o/ |
| 14:47.13 | phh | now rfcomm. hard one. |
| 14:47.22 | ali1234 | they are really only pda distros |
| 14:47.31 | ali1234 | eg gpe... has no dialer that we could find |
| 14:47.37 | ali1234 | or bluetooth manager |
| 14:47.46 | ali1234 | it has BT but its so old it doesnt work any more |
| 14:48.04 | phh | Can't create RFCOMM TTY: Operation not supported |
| 14:48.06 | phh | hum. |
| 14:48.13 | cr2 | ali1234: too bad trolls were too greedy to gpl qtopia2. it would have been a different world now. |
| 14:48.26 | ali1234 | phh: maybe you didnt include rfcomm in the kernel? |
| 14:48.43 | ali1234 | all that stuff comes down to kernel support |
| 14:48.47 | phh | i think i did mmm |
| 14:48.56 | phh | argh |
| 14:48.59 | phh | rfcomm but not rfcomm tty |
| 14:49.05 | ali1234 | that would be it... |
| 14:49.12 | cr2 | yeah |
| 14:49.33 | ali1234 | i got synergy-bluetooth running on wizard |
| 14:49.52 | ali1234 | for kb and mouse :) |
| 14:49.58 | ali1234 | works a bit slow though |
| 14:50.53 | cr2 | why ? |
| 14:51.06 | ali1234 | why i did or or why is it slow? |
| 14:51.12 | ali1234 | dunno why it is slow |
| 14:51.15 | cr2 | why is it slow |
| 14:51.21 | ali1234 | i guess it polls a something a bit too much |
| 14:51.38 | phh | maybe the bluetooth stack doesn't suits realtime |
| 14:51.49 | ali1234 | i can't fault the BT stack |
| 14:52.18 | ali1234 | never seen it fail... ever |
| 14:52.27 | ali1234 | unlike wifi |
| 14:52.41 | ali1234 | which i have never seen work properly |
| 14:53.14 | cr2 | there are picky usb bt interfaces |
| 14:53.18 | phh | speaking about wifi, any idea what means 'Failed to start config manager' ? |
| 14:53.30 | phh | said by tiwlan in dmesg. |
| 14:53.39 | ali1234 | no idea |
| 14:53.44 | ali1234 | we didn't start playing with that yet |
| 14:53.48 | phh | ok |
| 14:53.51 | ali1234 | although it should work for us on omap too |
| 14:53.59 | ali1234 | but we only just got a recent enough kernel going |
| 14:54.05 | cr2 | ali1234: do we have a eeprom/fw update tool ? |
| 14:54.09 | *** join/#htc-linux GlemSom (n=glemsom@93.163.75.202) |
| 14:54.09 | ali1234 | waiting for upstream to take our patches atm |
| 14:54.13 | phh | i'll try the proper driver instead of TI one |
| 14:54.21 | phh | but it seems a lot of build troubles. |
| 14:54.58 | ali1234 | cr2: for tiwlan? no idea, like i said, didn't investigate it yet. all i know is we have the same or very similar chip on wizard, and connected in the same way it seems |
| 14:54.59 | cr2 | phh: the problem is how to extract the eeprom, and how to load it. and the firmware. |
| 14:55.19 | cr2 | ali1234: pxa too |
| 14:55.27 | cr2 | ali1234: so think portable :) |
| 14:55.33 | phh | cr2: you mean calibration with eeprom ? |
| 14:55.38 | cr2 | ali1234: and s3c2442 too |
| 14:55.54 | ali1234 | i wish all hardware worked as easily as the omap gsm stuff |
| 14:55.59 | ali1234 | that couldn't be any easier, really |
| 14:56.08 | cr2 | phh: it has your ethernet id, and some othe r things. |
| 14:56.11 | phh | cause i got it by hand (firmware is in the \Windows\ and calibration can be found with dumping whole nand and find headers as stated in http://htc-linux.org/wiki/index.php?title=WL1251) |
| 14:56.26 | ali1234 | write firmware to specific physical addres -> get a AT interface on FIFO at another address |
| 14:56.34 | cr2 | ali1234: lol. omap gsm is easy ? |
| 14:56.51 | cr2 | ali1234: the AT, maybe. |
| 14:56.52 | ali1234 | compared to MSM it's simple |
| 14:56.58 | phh | argh |
| 14:57.02 | phh | scan no longer works :'( |
| 14:57.03 | cr2 | gsm audio ? |
| 14:57.13 | cr2 | phh: ? |
| 14:57.21 | cr2 | phh: powerdown, and up |
| 14:57.36 | cr2 | phh: and load the firmware. it may help |
| 14:57.55 | ali1234 | gsm audio is not done anything like MSM... on omap there is a sound card on applications core, and the gsm is connected onto one of it's inputs |
| 14:57.57 | cr2 | firmware patches in .bts, to be more accurate |
| 14:57.59 | ali1234 | so all you have to do is unmute it |
| 14:58.12 | *** join/#htc-linux ppman (n=mkern@pool-71-250-24-161.nwrknj.east.verizon.net) |
| 14:58.16 | phh | cr2: the hciattach in angstrom seems not to be able to read it properly :/ |
| 14:58.28 | ali1234 | now, making a driver for the app core sound is a different matter... |
| 14:58.29 | cr2 | ali1234: i2s over spi ? |
| 14:58.42 | ali1234 | i2s over McBSP i believe |
| 14:58.52 | ali1234 | certainly i2s |
| 14:58.56 | cr2 | phh: strange. i can decompile it without any problems |
| 14:59.05 | phh | hu ? i'll try again then |
| 14:59.14 | ppman | halp, apparently I need a data.img with premade dex files for the latest rogers build on my kaiser |
| 14:59.38 | ali1234 | ppman: would you like the moon on a stick with that? |
| 14:59.49 | ppman | yummy |
| 14:59.57 | cr2 | ppman: we have degraded to an msm7201A channel recently |
| 15:00.06 | ppman | but seriously, my device boots to a unresponsive black screen |
| 15:00.16 | GNUtoo | <PROTECTED> |
| 15:00.42 | cr2 | GNUtoo: missing pin ? |
| 15:00.45 | ppman | cr2: what's wrong with 7201? we still do have these phones, you know... |
| 15:00.56 | ali1234 | nah missing pin has it's own code |
| 15:00.57 | GNUtoo | ok I'll check |
| 15:01.04 | phh | cr2: if i put the firmware i get: |
| 15:01.04 | cr2 | ppman: that's why i say "degraded" |
| 15:01.05 | phh | Firmware file : /lib/firmware/TIInit_5.3.53.bts |
| 15:01.05 | phh | Loaded BTS script version 1 |
| 15:01.05 | phh | texas: changing baud rate to 4000000, flow control to 1 |
| 15:01.05 | phh | Initialization timed out. |
| 15:01.08 | ali1234 | CME 100 literally means "other error that doesn't have a code" |
| 15:01.35 | GNUtoo | ok thanks |
| 15:01.39 | ali1234 | GNUtoo: on wizard/omap gsm we get that error if we send the wrong command on the wrong AT interface (it has two, one for commands, and one for data) |
| 15:02.04 | cr2 | phh: hciattach ? |
| 15:02.11 | phh | cr2: yes |
| 15:02.19 | ali1234 | it sometimes also happens if you try to do a reset (ATZ) while connected on ppp |
| 15:02.29 | ali1234 | but it basically means the modem is confused and needs to be reset |
| 15:02.52 | GNUtoo | ali1234, mmm doesn't seem that because he can probe the network etc... |
| 15:03.18 | ali1234 | well, you're using MSM i guess like most people.... completely different modem, different firmware etc |
| 15:03.22 | ali1234 | so not much help |
| 15:03.25 | GNUtoo | ok |
| 15:03.31 | ali1234 | but best thing to do is reset and try again |
| 15:03.34 | cr2 | phh: is it so dumb to not change the line speed on the host side ? or maybe the host side doesn'T switch to 4MHz ? |
| 15:03.38 | GNUtoo | yes htc dream G1 but the person is using a dev phone |
| 15:03.43 | ali1234 | and try to figure out the sequence that causes it, and avoid it in future |
| 15:03.50 | ppman | so what do I do about this whole booting to blank issue? |
| 15:03.57 | ali1234 | lots of AT devices need crazy workarounds for stuff like this |
| 15:04.04 | ali1234 | just ask the n-m people :) |
| 15:04.33 | cr2 | phh: do you have dmesg after [17:01] <phh> Initialization timed out. |
| 15:04.59 | GNUtoo | ok |
| 15:05.06 | ali1234 | initialization timeout? i know what causes that :) |
| 15:05.22 | ali1234 | assuming you;re still talking about bluetooth |
| 15:05.31 | cr2 | ali1234: the fw changes baud rate to 4MHz |
| 15:05.40 | ali1234 | sure |
| 15:05.42 | cr2 | on the brf63x0 side |
| 15:05.50 | ali1234 | we had an issue on wizard |
| 15:05.55 | ali1234 | the bluetooth driver spams a lot of debug onto dmesg |
| 15:06.00 | ali1234 | that goes to the framebuffer |
| 15:06.06 | ali1234 | and the framebuffer is slow when scrolling |
| 15:06.07 | cr2 | phh: check the dmesg |
| 15:06.12 | phh | i'm |
| 15:06.16 | ali1234 | and so the dmesg output was causing a timeout |
| 15:06.22 | ali1234 | so we disabled debugging... and it worked |
| 15:06.27 | phh | ali1234: oO |
| 15:06.27 | GNUtoo | thanks a lot btw |
| 15:06.41 | phh | cr2: http://pastebin.com/m7a641ed1 |
| 15:06.47 | *** join/#htc-linux FR^2 (n=frquadra@frquadrat.de) |
| 15:06.52 | ali1234 | it took ages to track that one down |
| 15:06.53 | cr2 | i've set the BT clock to operate at 4x speed. maybe it |
| 15:06.57 | cr2 | 's too much |
| 15:07.18 | cr2 | for the M/N divider to operate at 64MHz |
| 15:07.54 | cr2 | [ 708.379638] clock-wince: set mdns: 35, 64000000; bitidx=26, offset=dc, ns=0 |
| 15:07.55 | cr2 | [ 708.396240] clock-wince: set_mdns_host_clock: 35, freq=32000000 |
| 15:08.02 | cr2 | err. i think it picks the SD clock here. |
| 15:08.30 | cr2 | we need to restructure the clock-wince.c, and add the switch() on clock id. |
| 15:08.51 | cr2 | since we know how to generate the (most important) clocks now |
| 15:09.22 | cr2 | phh: should it be 32MHz ? |
| 15:09.32 | phh | no idea |
| 15:09.35 | cr2 | i'm not sure, but thought it was 64 |
| 15:09.46 | cr2 | need to do the math again. |
| 15:10.00 | *** join/#htc-linux sdt555 (n=titus@147.145.40.44) |
| 15:10.06 | *** part/#htc-linux sdt555 (n=titus@147.145.40.44) |
| 15:10.13 | cr2 | <PROTECTED> |
| 15:10.28 | cr2 | yeah. asking for 32 will change the SD clock speed ;) |
| 15:10.50 | cr2 | a bit short-sighted implementation. |
| 15:11.12 | cr2 | <PROTECTED> |
| 15:11.31 | cr2 | phh: hmm. how does it come to 32 ? |
| 15:11.56 | cr2 | [17:07] <cr2> [ 708.379638] clock-wince: set mdns: 35, 64000000; bitidx=26, offset=dc, ns=0 |
| 15:12.08 | cr2 | ah, maybe i know it. |
| 15:12.22 | phh | reading the SD in the same time ? |
| 15:12.38 | phh | hum it should with hciattach |
| 15:12.43 | phh | shouldn't* |
| 15:12.50 | cr2 | no. the clock frequency is used as a key instead of the clock id |
| 15:12.58 | phh | oO |
| 15:13.11 | cr2 | so if two clock ids have the same frequency, it's a problem. |
| 15:13.25 | cr2 | well, back in the old time, it was ok |
| 15:13.29 | cr2 | to do that |
| 15:13.50 | cr2 | now we need to support much more clocks, so it became a problem |
| 15:14.24 | cr2 | <PROTECTED> |
| 15:14.33 | cr2 | this is the problem. |
| 15:18.27 | cr2 | need to add the id to struct mdns_clock_params |
| 15:19.57 | phh | will do an initramfs with bluez meanwhile |
| 15:20.30 | cr2 | phh: make this line the last in the list for now |
| 15:20.33 | cr2 | <PROTECTED> |
| 15:21.03 | phh | after SD ? |
| 15:21.08 | cr2 | yes |
| 15:21.23 | cr2 | will it break the SD ? hmm. |
| 15:21.38 | phh | i will be able to try anyway |
| 15:22.50 | cr2 | the list should be sorted on frequency , i guess |
| 15:22.59 | cr2 | for this code to work |
| 15:23.08 | GNUtoo | basically the call is like this: the person calls,then it hangs up immediately...is it because there are no carrier? |
| 15:23.11 | cr2 | we need to add the clock id asap |
| 15:24.34 | phh | hates waiting for Windows Mobile to boot. |
| 15:24.34 | cr2 | and the code should do a full match search. |
| 15:24.40 | cr2 | LOL |
| 15:25.47 | phh | first, SD isn't broken. |
| 15:25.55 | cr2 | ok |
| 15:27.14 | phh | hu |
| 15:27.15 | phh | changed |
| 15:27.18 | phh | texas: changing baud rate to 4000000, flow control to 1 |
| 15:27.18 | phh | Cannot send hci command to socket: Connection timed out |
| 15:27.18 | phh | Can't initialize device: Connection timed out |
| 15:27.29 | cr2 | dmesg ? |
| 15:27.39 | phh | http://pastebin.com/m35d05616 |
| 15:28.13 | cr2 | why did it return back to 115200 ? |
| 15:28.40 | phh | that's the default speed of hciattach |
| 15:28.58 | cr2 | hmm. |
| 15:29.11 | cr2 | i doubt that stty knows about 4000000 |
| 15:29.29 | cr2 | it was a problem already on blueangel |
| 15:29.38 | phh | ioctl(3, SNDCTL_TMR_START or TCSETS, {B4000000 -opost -isig -icanon -echo ...}) = 0 |
| 15:29.41 | phh | in strace hciattach |
| 15:29.43 | cr2 | that stty didn't know 460800 and 921600 |
| 15:29.57 | cr2 | ok, so it know it. |
| 15:30.39 | phh | does msm hs does too ? |
| 15:31.08 | cr2 | [ 115.638061] clock-wince: set mdns: 35, 64000000; bitidx=26, offset=dc, ns=0 |
| 15:31.10 | cr2 | [ 115.652954] clock-wince: set_mdns_host_clock: 35, freq=64000000 |
| 15:31.24 | cr2 | tcsetattr ioctl ? |
| 15:32.44 | phh | TCSETS according to strace, don't know if it is the same one |
| 15:33.04 | cr2 | ok |
| 15:33.04 | cr2 | need to think abou it. |
| 15:33.23 | cr2 | but you should get a coonect at 115200 |
| 15:33.24 | phh | i'd guess it happens on line 359 in msm_serial_hs |
| 15:34.09 | phh | i can't have both firmware and 115200 ... |
| 15:34.21 | cr2 | yes. |
| 15:34.30 | cr2 | at least now now |
| 15:34.58 | cr2 | to remove the frequency switch, we need a .bts compiler |
| 15:35.16 | phh | ok |
| 15:35.20 | cr2 | <PROTECTED> |
| 15:35.22 | cr2 | <PROTECTED> |
| 15:35.38 | cr2 | the comment is misleading. cut and paste |
| 15:35.51 | phh | now all i get without firmware is timeouts .... |
| 15:36.02 | phh | (leading to host is down and such things) |
| 15:36.34 | cr2 | and with firmware you can't connect at all ? |
| 15:36.57 | phh | cr2: hciattach exits with the timeout |
| 15:37.11 | cr2 | ok |
| 15:37.30 | cr2 | <PROTECTED> |
| 15:37.32 | cr2 | <PROTECTED> |
| 15:37.33 | cr2 | <PROTECTED> |
| 15:37.35 | cr2 | <PROTECTED> |
| 15:37.36 | cr2 | <PROTECTED> |
| 15:39.50 | cr2 | <PROTECTED> |
| 15:40.15 | cr2 | and > 460800 disable the baud divider... |
| 15:40.16 | phh | cr2: all the bluetooth related requests should go through hciattach ? |
| 15:40.21 | cr2 | it's ok |
| 15:40.53 | cr2 | do you have hci0 device ? |
| 15:40.58 | phh | yup |
| 15:41.16 | cr2 | then all goes through hci0 |
| 15:41.29 | phh | http://pastebin.com/m34582061 if that can help (don't think so.) |
| 15:41.56 | phh | i tried a distant rfcomm connect <mac of the device> and RX bytes hasn't changed |
| 15:42.03 | cr2 | hciconfig hci0 up |
| 15:42.28 | phh | i already tried, tried hciconfig hci0 reset too |
| 15:42.30 | cr2 | UP RUNNING |
| 15:42.50 | cr2 | does scan see the other side ? |
| 15:43.01 | phh | no |
| 15:43.21 | cr2 | compile the initramfs, i'll check with my bt gps. |
| 15:43.26 | cr2 | hehe. where is it... |
| 15:43.51 | phh | hum i can also try with other bluetooth devices |
| 15:43.53 | phh | but first initramfs ok :) |
| 15:44.06 | phh | cr2: hcitool/hciconfig/rfcomm is enough ? |
| 15:44.47 | *** join/#htc-linux dcordes (n=chatzill@f049131055.adsl.alicedsl.de) |
| 15:45.06 | dcordes | hi |
| 15:45.13 | dcordes | cr2: I have good and bad news |
| 15:45.39 | cr2 | hi dcordes |
| 15:45.48 | cr2 | yes. good news first |
| 15:46.30 | dcordes | the new fresh new rootfs is finished. I also built some extra images with nice X based gui and fso etc. |
| 15:46.39 | cr2 | ok |
| 15:46.48 | cr2 | phh: yes. |
| 15:46.50 | dcordes | and bad news is I extracted zubuntu in / haha |
| 15:47.04 | cr2 | on your computer ? |
| 15:47.17 | dcordes | yes on the desktop computer |
| 15:47.25 | cr2 | hehe. shit happens |
| 15:47.44 | dcordes | I just find it funny |
| 15:48.00 | dcordes | good opportunity to go back to debian. |
| 15:48.06 | cr2 | lol |
| 15:48.22 | dcordes | now I have a hybrid rootfs with x86 and arm binaries ^^ |
| 15:48.48 | cr2 | does 'file' work ? |
| 15:49.09 | dcordes | I think that was present in the zubuntu tarball so, most likely no |
| 15:49.21 | dcordes | I was just wondering... ls cannot execute binary ?!?!1 |
| 15:49.27 | cr2 | i've had such problem during aout->elf linux conversion |
| 15:49.38 | cr2 | yes, ls is probably broken now. |
| 15:49.48 | GNUtoo | dcordes, which phone? |
| 15:49.50 | cr2 | use 'echo *' instead |
| 15:50.07 | cr2 | because echo is builtin bash |
| 15:50.16 | dcordes | GNUtoo: I have the kovsky. but I'm talking about the funny adventures of extracting ARM rootfs into x86 / |
| 15:50.25 | cr2 | and bash is alive i guess :) |
| 15:50.27 | GNUtoo | dcordes, ah ok |
| 15:50.52 | GNUtoo | dcordes, you extracted your oe tarball to / in your x86...ok :( |
| 15:51.14 | dcordes | dcordes: no it wasn't OE but 'zubuntu' |
| 15:52.33 | GNUtoo | ah ok ubuntu for zaurus |
| 15:52.34 | dcordes | GNUtoo: did you see the XDA link cr2 posted in here yesterday? |
| 15:52.39 | GNUtoo | no |
| 15:52.40 | *** join/#htc-linux x29a (n=x29a@unaffiliated/x29a) |
| 15:52.42 | GNUtoo | what is it? |
| 15:53.03 | phh | cr2: http://husson.hd.free.fr/angstrom-bluez.cpio.gz |
| 15:53.07 | dcordes | GNUtoo: some folks who have similar aim as you. removing all the proprietary stuff from androed |
| 15:53.10 | phh | but i haven't checked it works yet |
| 15:53.16 | phh | (but no reason it wouldn't) |
| 15:53.23 | cr2 | phh: downloading |
| 15:53.25 | GNUtoo | dcordes, wow!!! |
| 15:53.56 | dcordes | GNUtoo: http://forum.xda-developers.com/showthread.php?t=564263 |
| 15:54.03 | GNUtoo | thanks a lot |
| 15:54.24 | cr2 | strange. 8GB sdhc is not seen on this usb dongle. but 4GB is ok. |
| 15:54.38 | phh | cr2: sure it's sdhc ? |
| 15:54.44 | phh | (some 4GB are not SDHC) |
| 15:55.03 | cr2 | has HC logo |
| 15:55.12 | phh | ok |
| 15:55.14 | phh | weird then yes. |
| 15:55.50 | phh | hum, i got wrong somewhere. |
| 15:55.57 | phh | original initramfs is 7.1MB, mine 13 |
| 15:56.13 | phh | haha, forget to strip. |
| 15:56.13 | *** join/#htc-linux balans (n=BRPOLDB@212-123-149-239.ip.telfort.nl) |
| 15:58.28 | phh | hum that's not it |
| 15:58.39 | phh | dcordes: how do you build your initramfs ? |
| 15:58.43 | cr2 | Warning: 256-byte inodes not usable on older systems |
| 15:58.48 | dcordes | phh: bitbake console-image |
| 15:59.34 | dcordes | hmm need to go buy CDR for installation |
| 16:00.50 | cr2 | phh: linuxrc link is ../../bin/busybox |
| 16:00.58 | cr2 | does not look right |
| 16:01.03 | phh | ergh |
| 16:01.14 | phh | i just extracted one and added the bluez files |
| 16:02.02 | cr2 | phh: you have zimage 2.6.25 inside |
| 16:02.05 | cr2 | =junk |
| 16:02.24 | cr2 | 1.3MB |
| 16:02.26 | phh | got it |
| 16:02.36 | phh | i extracted an old one |
| 16:04.04 | cr2 | no playwav2 ? |
| 16:04.43 | phh | it is angstrom 20081127 ... |
| 16:04.55 | phh | i can put whatever you want |
| 16:06.33 | cr2 | hmm. where did i download the debian image yesterday.. |
| 16:06.53 | phh | if you've a better base give it :) |
| 16:07.08 | cr2 | phh: the image of dcordes ? |
| 16:07.10 | dcordes | cr2: 'opkg install bluez-utils' (optionally extract initrd to card before). you have the usb net working, don't you ? |
| 16:07.43 | dcordes | but I will prepare the new one once debian is installed |
| 16:07.50 | dcordes | shopping now bbl |
| 16:10.05 | phh | cr2: http://husson.hd.free.fr/angstrom-bluez.2.cpio.gz |
| 16:10.23 | phh | and i tried it, it boots and bluetooth utilities works |
| 16:14.41 | phh | argh |
| 16:14.44 | phh | scan works |
| 16:15.36 | cr2 | can't find bt gps |
| 16:16.34 | phh | ok i know why it didn't work last time, i was too far from the dongle |
| 16:16.39 | phh | it seems it is very low powered ... |
| 16:19.11 | phh | i got a rfcomm connection working |
| 16:19.29 | phh | (i just had to remember than the MAC isn't just after "connect" but there is a ID before ....) |
| 16:19.58 | cr2 | phh: edit the wiki page |
| 16:20.45 | cr2 | i'm decompressing the 245MB debian image onto sd card |
| 16:20.59 | phh | http://wiki.xda-developers.com/index.php?pagename=RaphaelLinux you mean ? |
| 16:21.15 | cr2 | omg |
| 16:21.24 | cr2 | 2864376 22% /media/disk |
| 16:21.32 | cr2 | 2.8GB uncompressed :) |
| 16:21.35 | phh | haha |
| 16:22.44 | cr2 | will take some time to umount |
| 16:22.53 | phh | use mount -osync |
| 16:23.05 | cr2 | too late |
| 16:24.36 | phh | oops |
| 16:24.51 | phh | wanted to change the 'usb' part on the wiki and i failed |
| 16:32.58 | phh | no that wifi works, back to wifi ... i'll have to understand this "awesome" TI driver. |
| 16:34.24 | cr2 | phh: do you have your own nand dump ? |
| 16:34.28 | phh | yep |
| 16:34.42 | cr2 | do the 2 eeprom copies differ ? |
| 16:34.45 | phh | [ 1238.352752] TIWLAN: 1049.213224: Working on a 1251 PG 1.2 board. |
| 16:34.45 | phh | [ 1238.382385] TIWLAN: 1049.242887: Starting to process NVS... |
| 16:34.45 | phh | [ 1238.438385] TIWLAN: 1049.298887: NVS version 6.3.7 found. |
| 16:34.45 | phh | [ 1238.449951] TIWLAN: 1049.310453: Radio type is 0x1C. |
| 16:34.45 | phh | [ 1238.477142] TIWLAN: 1049.337644: Finished processing NVS. |
| 16:34.46 | phh | [ 1238.501800] TIWLAN: TIWLAN: Failed to start config manager |
| 16:34.50 | phh | here is where i get. |
| 16:35.19 | phh | cr2: 2 eeprom ? i got three time the 0x02 0x11 0x56 0x06 0x1C 0x06 header |
| 16:35.26 | phh | the two last ones are identical |
| 16:35.27 | phh | not the first one |
| 16:35.27 | cr2 | Radio type is eeprom afair |
| 16:35.40 | cr2 | phh: check wiki |
| 16:36.04 | phh | all i've found is http://htc-linux.org/wiki/index.php?title=WL1251 |
| 16:36.44 | cr2 | RaphaelNAND |
| 16:36.53 | phh | cr2: do you know if this driver does need some specific useland daemon maybe ? |
| 16:37.07 | cr2 | in android maybe |
| 16:37.19 | cr2 | there is some loader too |
| 16:37.25 | phh | well i built the module from the android git ... |
| 16:37.28 | cr2 | check how they use wifi on kaiser |
| 16:37.31 | phh | yes the loader is built |
| 16:37.37 | phh | and running |
| 16:37.52 | phh | well just the "start config managed failed" thing. |
| 16:37.52 | cr2 | check how to load eeprom |
| 16:38.08 | cr2 | then we need to disable the DMA 8 on wifi sd channel |
| 16:38.43 | cr2 | there is some ascii file with tiacx options |
| 16:38.45 | cr2 | let me check |
| 16:38.54 | *** join/#htc-linux root2 (n=root@rgnb-5d874749.pool.mediaWays.net) |
| 16:39.38 | cr2 | btw, android has its own bt firmware, and wifi fw |
| 16:39.57 | phh | 0x2500040 and 0x2500840 are where to find wifi eeprom |
| 16:40.01 | phh | and yes both are identical to me |
| 16:40.13 | cr2 | ok |
| 16:40.25 | cr2 | can we describe its structure ? |
| 16:40.42 | cr2 | maybe compare to acx100 |
| 16:41.01 | phh | ouch |
| 16:41.02 | cr2 | there is wpa_supplicant |
| 16:41.25 | cr2 | and tiwlan.ini |
| 16:41.35 | cr2 | do you have tiwlan.ini ? |
| 16:41.40 | phh | yes wlan_loader needs it |
| 16:42.57 | phh | cr2: do you think i should try android's fw ? |
| 16:44.25 | cr2 | maybe |
| 16:44.39 | cr2 | check how they deal with the eeprom first |
| 16:44.53 | cr2 | i've read something about eeprom-less operation |
| 16:45.59 | cr2 | tiwlan.ini, firmware.bin and nvs_map.bin |
| 16:46.45 | cr2 | i think eeprom is "sent" as an ATAG |
| 16:46.53 | cr2 | in g1 |
| 16:47.04 | cr2 | but we can just create a file with it. |
| 16:48.53 | phh | I think i'll try the alternative driver ... the ati one is unreadable ... |
| 16:49.11 | phh | -a |
| 16:49.53 | cr2 | SGH-M200 ? |
| 16:50.03 | cr2 | is it some neighbours phone ?? |
| 16:51.02 | cr2 | a780 sees some "Computer" ? |
| 16:51.31 | phh | no idea how you may see your phone mmm |
| 16:51.52 | tmzt | if it has pan the class will probably be computer |
| 16:52.11 | phh | tmzt: nothing launched |
| 16:52.15 | tmzt | that's a Samsung |
| 16:52.40 | cr2 | yes, a780 is visible when i let it |
| 16:52.51 | phh | cr2: and not your GPS ? |
| 16:52.53 | cr2 | A780 |
| 16:53.08 | cr2 | but if i search from it i see Computer |
| 16:53.18 | cr2 | phh: forgot gps at work |
| 16:53.25 | phh | ah |
| 16:53.44 | phh | cr2: well it's most likely, since bluez is mainly for desktops ... |
| 16:54.08 | phh | cr2: you can change it in /etc/bluetooth/hcid.conf or something like that if that matters |
| 16:55.40 | cr2 | pin_code_request: |
| 16:56.01 | phh | haha tough one |
| 16:56.17 | phh | you'll need the whole thing for that |
| 16:56.22 | phh | with glib dbus and friends |
| 16:56.52 | cr2 | ok. was just for fun |
| 16:57.14 | cr2 | will try to copy and load .bts |
| 16:59.17 | cr2 | phh: where is the hciattach source ? |
| 16:59.25 | cr2 | which we use in initramfs |
| 16:59.40 | phh | ask dcordes |
| 17:00.15 | phh | there is two hciattach for TI in bluez 4.54 |
| 17:00.28 | phh | ( http://www.kernel.org/pub/linux/bluetooth/bluez-4.54.tar.gz ) |
| 17:00.33 | phh | but it won't be the same version i guest |
| 17:00.35 | phh | guess* |
| 17:02.34 | cr2 | ok |
| 17:02.40 | cr2 | times out for me too |
| 17:03.02 | phh | and clocks seems fine ? |
| 17:03.14 | cr2 | err |
| 17:03.20 | cr2 | i forgot to patch the kernel |
| 17:04.33 | cr2 | i'll add clock id to the struct. |
| 17:04.43 | cr2 | to avoid such confusion in the future |
| 17:05.07 | cr2 | druidu wants to create something resembling clock-7x30.c |
| 17:09.24 | cr2 | eh ? |
| 17:09.38 | cr2 | wince exports linux partition too ? |
| 17:11.07 | phh | oO |
| 17:11.37 | cr2 | block device. |
| 17:11.44 | cr2 | but i will not risk to write there :) |
| 17:11.53 | phh | :) |
| 17:13.13 | cr2 | booting |
| 17:14.11 | cr2 | isthis blue rectangle some redundant feature ? |
| 17:14.36 | phh | blue rectangle ... ? |
| 17:14.50 | cr2 | near the penguin |
| 17:15.06 | cr2 | it was some virtual keyboard hack afair |
| 17:15.12 | cr2 | for diamond |
| 17:15.14 | phh | it is still normaly |
| 17:15.27 | cr2 | raph does not really need it |
| 17:15.29 | phh | i have to use it when usbnet doesn't work you know. |
| 17:15.42 | phh | cr2: just then blabal keyb=off |
| 17:15.54 | phh | msmvkeyb_toggle=off |
| 17:16.10 | phh | i think you have the "hide" option |
| 17:17.42 | cr2 | can't mount the second partition |
| 17:18.16 | cr2 | Disk /dev/mmcblk0: 4075 MB, 4075290624 bytes |
| 17:18.26 | cr2 | <PROTECTED> |
| 17:18.27 | cr2 | /dev/mmcblk0p1 1 65 253859 6 FAT16 |
| 17:18.29 | cr2 | /dev/mmcblk0p2 66 1018 3722418 83 Linux |
| 17:18.52 | cr2 | root@htcraphael:~# mount /dev/mmcblk0p2 /mnt/card/ |
| 17:18.52 | cr2 | mount: /dev/mmcblk0p2 already mounted or /mnt/card/ busy |
| 17:19.02 | phh | cat /proc/mounts ? |
| 17:19.18 | cr2 | root@htcraphael:~# cat /proc/mounts |
| 17:19.30 | cr2 | <PROTECTED> |
| 17:19.43 | cr2 | this initramfs is not very good |
| 17:19.43 | phh | hu ? mmcblk0 not p1 ? |
| 17:20.28 | cr2 | [ 394.111022] kjournald starting. Commit interval 5 seconds |
| 17:20.29 | cr2 | [ 394.222930] EXT3 FS on mmcblk0p2, internal journal |
| 17:20.31 | cr2 | [ 394.231475] EXT3-fs: mounted filesystem with ordered data mode. |
| 17:20.59 | cr2 | /dev/mmcblk0p2 3663888 799512 2864376 22% /media/card |
| 17:21.00 | cr2 | /dev/mmcblk0p1 253592 9104 244488 4% /media/cf |
| 17:21.26 | cr2 | ok. |
| 17:21.39 | cr2 | now may try to mount it as root instead of the initramfs |
| 17:22.49 | phh | mount -t tmpfs -o size=100K tmpfs /media/cf/dev |
| 17:22.49 | phh | cp -a /dev/* /media/cf/dev |
| 17:22.49 | phh | umount -l /proc |
| 17:22.49 | phh | umount -l /sys |
| 17:22.49 | phh | exec switch_root /media/cf /init |
| 17:22.51 | phh | i'd say. |
| 17:23.32 | cr2 | don't see hcitool in this 2.8GB distro ? |
| 17:23.36 | phh | haha |
| 17:25.40 | phh | cr2: what's the current status of GPS btw ? |
| 17:25.45 | phh | it still crashes the system ? |
| 17:27.29 | cr2 | the code needs a patch |
| 17:27.52 | cr2 | we need to register 2 CB rpc servers |
| 17:28.04 | cr2 | and they need to pick the messages sent by arm9 |
| 17:28.14 | cr2 | and probably just do nothhing |
| 17:28.21 | *** join/#htc-linux BHSPitLappy (n=BHSPitLa@unaffiliated/bhspitmonkey) |
| 17:28.32 | phh | hu ? and that's it ? |
| 17:28.37 | cr2 | but at least the message struct should be declared properly, and they are to be acked. |
| 17:28.42 | cr2 | yes |
| 17:29.12 | cr2 | maybe they should be shut down on gps down |
| 17:29.28 | cr2 | like it's done for snd/audmgr |
| 17:29.52 | cr2 | and if we are not interested in agps, things will be even more simple. |
| 17:30.26 | cr2 | root=/dev/mmcblk0p2 |
| 17:30.44 | cr2 | btw, if the wifi is enabled it will be mmcblk1 ? |
| 17:30.49 | phh | no |
| 17:30.58 | phh | [ 6.856842] mmc1: new high speed MMC card at address 0001 |
| 17:30.58 | phh | [ 6.891357] mmcblk0: mmc1:0001 000000 3932160KiB |
| 17:31.17 | phh | it's already on "mmc1" and it is still named mmcblk"0" |
| 17:31.18 | cr2 | ok, it' s not a block device |
| 17:31.29 | cr2 | i mean wifi |
| 17:31.30 | tmzt | what? |
| 17:32.02 | *** join/#htc-linux tsdogs (n=tsdogs@net203-187-146.mclink.it) |
| 17:32.05 | tmzt | MMC is not SDIO |
| 17:32.10 | tmzt | i linux |
| 17:32.12 | tmzt | in |
| 17:32.15 | phh | hu ? |
| 17:32.35 | cr2 | [ 7.049591] mmc0: new high speed SDHC card at address f34f |
| 17:32.36 | cr2 | [ 7.069671] mmcblk0: mmc0:f34f SU04G 3979776KiB |
| 17:32.38 | cr2 | [ 7.077728] mmcblk0: p1 p2 |
| 17:33.11 | cr2 | phh: you have microMMC card ? |
| 17:33.17 | tmzt | phh: switchroot |
| 17:33.19 | cr2 | does such thing exist ? |
| 17:33.24 | phh | cr2: no idea |
| 17:33.26 | tmzt | will do that for you |
| 17:33.29 | phh | it's just a diam100 that's all i know |
| 17:33.44 | cr2 | phh: it's reported by CIS |
| 17:33.55 | cr2 | ah |
| 17:34.04 | cr2 | it's not a "real" SD |
| 17:34.08 | phh | no |
| 17:34.11 | phh | it's a soldered one |
| 17:34.15 | cr2 | tmzt: lol, they used MMC :) |
| 17:34.24 | tmzt | yeah |
| 17:34.26 | phh | SD != MMC ? |
| 17:34.48 | cr2 | you don't need to pay the license fees |
| 17:34.56 | cr2 | phh: no |
| 17:35.11 | cr2 | phh: there is also mmcv4 |
| 17:35.44 | cr2 | i have such card, but htc phones do not have "real" mmcv4 readers |
| 17:35.53 | cr2 | so it works in the backwards mode |
| 17:36.17 | cr2 | need some food. |
| 17:36.18 | cr2 | bbl |
| 17:36.22 | phh | :) |
| 17:39.02 | *** join/#htc-linux expanderz (n=expander@bas3-montreal31-1279786760.dsl.bell.ca) |
| 17:40.01 | expanderz | Anyone on? |
| 17:40.07 | phh | mmmm |
| 17:40.12 | expanderz | ahhh |
| 17:40.44 | tmzt | but linux calls block devices mmc |
| 17:40.52 | expanderz | I wish I could ditch windows mobile... |
| 17:41.41 | *** join/#htc-linux MrPippy (n=pip@adsl-75-42-169-223.dsl.sndg02.sbcglobal.net) |
| 17:41.53 | expanderz | Hey guys, how is Android coming along for the Touch Diamond? |
| 17:42.17 | expanderz | Last build was in Sept, right? |
| 17:42.28 | phh | you know it's still september ? :D |
| 17:42.47 | expanderz | I know, lol |
| 17:43.11 | phh | well almost everything works except GPS, WiFi, bluetooth and camera ... |
| 17:43.21 | expanderz | good news |
| 17:43.45 | expanderz | thanks for the info... |
| 17:43.57 | *** join/#htc-linux darkstar62 (n=darkstar@97-126-104-157.tukw.qwest.net) |
| 17:44.00 | phh | but that's not new. |
| 17:44.48 | expanderz | I tried it a few weeks ago.... |
| 17:45.38 | expanderz | Wil google ever support the touch or will it always be a hack? |
| 17:46.02 | expanderz | but if the hack works...that's ok too |
| 17:47.40 | expanderz | ummmm |
| 17:47.58 | tmzt | which touch? |
| 17:48.21 | expanderz | Diamond touch HTC |
| 17:48.40 | expanderz | the first gen |
| 17:49.03 | townkat | may i build for blackstone now? or git still broken ? |
| 17:49.34 | townkat | not sure i used the terms correctly :) |
| 17:53.57 | *** join/#htc-linux cr2 (n=cr2@109.84.130.177) |
| 17:57.53 | *** join/#htc-linux timebomb (n=tb@f054021101.adsl.alicedsl.de) |
| 18:00.22 | townkat | cr2, is blackstone build broken ? |
| 18:01.16 | phh | townkat: yes |
| 18:01.36 | townkat | thnx |
| 18:01.47 | phh | [16:28:10] <cr2> ./arch/arm/mach-msm/board-htcblackstone-panel.c: .name = "mddi_c_d263_0000", |
| 18:01.47 | phh | [16:28:17] <cr2> ./arch/arm/mach-msm/board-htcdiamond-panel.c: .name = "mddi_c_d263_0000", |
| 18:01.47 | phh | [16:28:26] <cr2> ./arch/arm/mach-msm/board-trout-panel.c: .name = "mddi_c_d263_0000", |
| 18:01.47 | phh | [16:28:28] <cr2> ./arch/arm/mach-msm/board-sapphire-panel.c: .name = "mddi_c_d263_0000", |
| 18:01.50 | phh | [16:28:52] <cr2> the last 2 are not important, but the first 2 are broken now |
| 18:02.53 | townkat | thank you phh |
| 18:03.15 | phh | townkat: but for diamond the fix was really easy |
| 18:03.24 | phh | (but i don't know if it is a real one or not) |
| 18:03.39 | *** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
| 18:03.42 | phh | just remplace mddi_thing to TC358720XBG |
| 18:04.22 | townkat | nice, but we need for a fix for blac too :) |
| 18:04.43 | phh | townkat: it's maybe the same thing for blackstone |
| 18:05.31 | cr2 | back |
| 18:07.37 | cr2 | phh: for diamond the fix is real |
| 18:07.44 | phh | ok |
| 18:08.07 | cr2 | but blac100 should go int epson client code |
| 18:08.18 | cr2 | not toshiba |
| 18:15.01 | *** join/#htc-linux Captnoord (n=Captnoor@81.71.164.123) |
| 18:16.08 | phh | cr2: btw, can you commit accelerometer for diamond ? |
| 18:16.23 | phh | it actually works |
| 18:17.19 | phh | (drop comments in board-htcdiamond.c around line 160) |
| 18:17.26 | cr2 | tmzt: does X work for you ? |
| 18:17.34 | cr2 | phh: wait for dcordes |
| 18:17.40 | phh | ok |
| 18:17.56 | phh | hum i have another patch, but don't know why |
| 18:18.40 | phh | ho right, it's maybe a feature |
| 18:19.00 | phh | cr2: to be able to build msm_nand, i had to add "EXPORT_SYMBOL(msm_dmov_exec_cmd);" in dma.c |
| 18:19.02 | phh | is that wanted ? |
| 18:19.34 | cr2 | works for me as is |
| 18:19.54 | phh | oh, because i used it as a module |
| 18:19.56 | phh | not in kernel |
| 18:20.15 | phh | no big deal then. |
| 18:20.25 | cr2 | still a bug |
| 18:21.13 | cr2 | but if your rootfs is on nand, you'll never know :) |
| 18:21.29 | phh | i think my rootfs won't be on nand before months. |
| 18:21.52 | cr2 | why? |
| 18:22.23 | phh | i prefer having a proper ROM "just in case" |
| 18:22.26 | cr2 | we may do dualboot |
| 18:22.30 | phh | hum |
| 18:22.36 | phh | windows mobile takes a lot of space you know |
| 18:22.49 | phh | is nand quicker than mmc ? |
| 18:23.33 | cr2 | i think so |
| 18:23.51 | phh | let's try. |
| 18:24.19 | cr2 | need to write some imgfs tools first |
| 18:24.26 | phh | dd: /dev/mtdblock0: Input/output error |
| 18:24.28 | phh | oops ... |
| 18:24.44 | cr2 | bs |
| 18:24.59 | phh | 8192 |
| 18:25.05 | phh | [ 3480.301147] msm_nand_read_oob: unsupported ops->len, 512 |
| 18:25.05 | phh | in dmesg |
| 18:25.07 | phh | strange |
| 18:25.11 | cr2 | the driver does not support unaligned |
| 18:25.23 | phh | <PROTECTED> |
| 18:25.24 | cr2 | yes. 512 is too little |
| 18:25.35 | cr2 | 10k ? |
| 18:25.44 | cr2 | does not work ? |
| 18:25.55 | phh | i tried with 10000, same thing |
| 18:26.08 | cr2 | i used the block device |
| 18:26.16 | cr2 | and assigned the partition |
| 18:28.25 | phh | read(0, 0xac008, 8192) = -1 EIO (Input/output error) |
| 18:28.27 | phh | i really don't get it |
| 18:29.24 | cr2 | it's an incomplete driver |
| 18:29.39 | phh | i know, but it did work tuesday .. |
| 18:29.43 | cr2 | and some things need to be clarified |
| 18:30.15 | phh | anyway i successfully dumped mine once |
| 18:30.22 | phh | i have got my calibration thing so i don't need |
| 18:31.11 | cr2 | 1 question |
| 18:31.34 | cr2 | how are you going to create the partitions, if you want to preserve wince ? |
| 18:31.52 | phh | i can do it on the "internal storage" |
| 18:32.06 | phh | (meaning the soldered SD) |
| 18:32.33 | cr2 | how big are they ? |
| 18:32.55 | phh | the NAND is 256MB, the storage is 4GB |
| 18:33.11 | cr2 | wince itself |
| 18:33.17 | phh | oh |
| 18:33.37 | phh | don't know |
| 18:33.40 | phh | wince says it is 30MB |
| 18:33.47 | phh | but i don't believe it |
| 18:33.47 | cr2 | you need to know some things about it |
| 18:34.23 | cr2 | there are some linux questions too |
| 18:34.49 | cr2 | 1. should we implement byte access ? |
| 18:34.50 | phh | like reading "partition table" ... |
| 18:35.15 | phh | FS already knows howto handle bloc access |
| 18:36.39 | phh | byte access would be useless except for the wifi eeprom, but it would be cleaner to do it in this part |
| 18:37.12 | cr2 | suggest how to extract eeprom in haret |
| 18:37.42 | cr2 | there is special wince ioctl to read the eeprom data, actually. |
| 18:39.41 | phh | well that would be an "enduser" thing, for the moment wifi doesn't work even with these files |
| 18:39.53 | cr2 | SUPPORT_WRONG_ECC_CONFIG |
| 18:39.59 | cr2 | do we need it ? |
| 18:40.27 | cr2 | can you point me to the wlan_loader source in gitweb ? |
| 18:41.04 | phh | in android git it's android/system/wlan/ti/sta_dk_4_0_4_32/CUDK/tiwlan_loader |
| 18:41.41 | phh | don't know where is gitweb for android mmm |
| 18:42.23 | *** join/#htc-linux pleemans (n=toi@d54C2A96D.access.telenet.be) |
| 18:43.34 | phh | http://android.git.kernel.org/?p=platform/system/wlan/ti.git;a=blob;f=sta_dk_4_0_4_32/CUDK/tiwlan_loader/tiwlan_loader.c;h=f4c137748e9710244691dbd4fde8fe686c9e9fcc;hb=HEAD |
| 18:45.35 | phh | but it isn't standalone, it needs a lot of other files ... |
| 18:45.37 | cr2 | nvsPtr = (volatile unsigned long *)mmap(0,0x1000,PROT_READ|PROT_WRITE,MAP_SHARED,fd,0x13F13000); |
| 18:45.39 | cr2 | lol |
| 18:47.00 | phh | what ? |
| 18:47.44 | cr2 | maybe make it standalone ? like playwav2 |
| 18:49.09 | phh | anyway i think (almost) everything is done in kernel |
| 18:50.24 | phh | oops i forgot |
| 18:50.30 | cr2 | 449 #define WIFI_NVS_LEN_OFFSET 0x0C |
| 18:50.31 | cr2 | 450 #define WIFI_NVS_DATA_OFFSET 0x40 |
| 18:50.32 | phh | just after loading wlan.ko, i get: |
| 18:50.33 | cr2 | 451 #define WIFI_NVS_MAX_SIZE 0x800UL |
| 18:50.35 | phh | [ 40.561065] mmc0: Command expiry misfire |
| 18:50.51 | cr2 | do you have dma enabled ? |
| 18:50.51 | phh | that might be a problem |
| 18:50.58 | cr2 | how do we disable it ? |
| 18:51.06 | phh | no idea |
| 18:51.11 | cr2 | you may post the full dmesg with wifi |
| 18:51.47 | phh | http://pastebin.com/m24a88226 |
| 18:52.33 | cr2 | [ 36.220153] .... ERROR in size of beacon_probeRsp_t struct 0x00000057 should be 0x00000053..... ? |
| 18:53.07 | phh | yes, i guess android repo is somehow broken |
| 18:53.17 | cr2 | omg |
| 18:53.28 | cr2 | [ 39.337768] clock-wince: pc_clk_set_rate: id=19 rate=11000000 |
| 18:53.29 | cr2 | [ 39.359649] clock-wince: set mdns: 19, 11000000; bitidx=7, offset=a4, ns=0 |
| 18:53.31 | cr2 | [ 39.384643] clock-wince: set_mdns_host_clock: 19, freq=7372800 |
| 18:53.43 | cr2 | picks the bt clock ;) |
| 18:53.47 | phh | -_- |
| 18:54.09 | cr2 | it may be ok, but just weird |
| 18:54.29 | phh | it did the same before your "patch" but you might want the messages without it ? |
| 18:54.53 | cr2 | <PROTECTED> |
| 18:55.08 | phh | ok |
| 18:55.18 | cr2 | and the MN does not know how the output will be used |
| 18:55.46 | cr2 | if the register layout is the same (and it is more or less the same, it seems) then why not |
| 18:56.10 | cr2 | just there is some difference between 11MHz and 7.3 MHz |
| 18:56.25 | cr2 | and the 32 vs. 64MHZ was completely unacceptable |
| 18:57.04 | cr2 | hm. why do they request 11MHz ? |
| 18:58.15 | cr2 | http://www.htc-linux.org/wiki/index.php?title=MSM_SDIO |
| 18:58.49 | cr2 | wifi? 19000000 1 0x05 0x02 3 2 4 1 |
| 18:59.09 | cr2 | SRC=4 |
| 18:59.15 | phh | erf, 19Mhz against 7 |
| 18:59.16 | cr2 | it's mpll aka PLL0 |
| 18:59.26 | cr2 | phh: in wince |
| 18:59.34 | cr2 | PLL0 ~=245MHz |
| 18:59.36 | phh | oh |
| 19:00.22 | cr2 | 245*1/5/4 |
| 19:00.28 | cr2 | how much is that ? |
| 19:00.38 | phh | 12.25 |
| 19:01.26 | cr2 | i need to compare to the MSM_CLK page. it's more readable |
| 19:01.38 | cr2 | or with the kernel, which i've patched yesterday |
| 19:01.54 | cr2 | so TI requests 11MHz ? |
| 19:02.13 | phh | according to dmesg yes |
| 19:02.21 | phh | and i won't read the TI driver to check that. |
| 19:02.24 | cr2 | phh: maybe we should write some TODO list |
| 19:02.33 | cr2 | then it must be somewhere in the driver code |
| 19:03.03 | phh | mmm sdio part mustn't be big |
| 19:08.32 | cr2 | 245760000*1/5/4 |
| 19:08.52 | phh | almost same thing, 12.35 Mhz |
| 19:09.01 | phh | err 12.28 |
| 19:09.06 | cr2 | close to 11 |
| 19:09.16 | cr2 | although wince says "19" |
| 19:11.22 | cr2 | but for sdio the speed should not be an issue |
| 19:11.34 | phh | ok |
| 19:12.01 | cr2 | the enabled dma is workse |
| 19:13.18 | phh | hu ? |
| 19:13.22 | cr2 | <PROTECTED> |
| 19:15.22 | cr2 | don't see an easy way to disable it though |
| 19:16.08 | phh | the problem is that dma is shared ? |
| 19:16.28 | cr2 | 12.288MHz |
| 19:16.33 | cr2 | it may be a problem |
| 19:16.48 | cr2 | btw, can you add the line for 12.288MHz ? |
| 19:17.06 | phh | sure, just give me the line :) |
| 19:18.14 | phh | i'm just thinking: |
| 19:18.18 | cr2 | MSM_CLOCK_REG(12288000, 1, 0x5, 0x2, 3, 2, 4, 1), /* 12.288MHz */ |
| 19:18.36 | phh | [ 39.337768] clock-wince: pc_clk_set_rate: id=19 rate=11000000 <---- isn't 19 supposed to be 4MHz ? |
| 19:18.38 | cr2 | but if the driver wants 11, it will pick 12 |
| 19:19.04 | cr2 | phh: find this place in the wifi driver |
| 19:19.18 | phh | ok |
| 19:19.26 | *** join/#htc-linux MLM (n=mlvdmeid@5ED0BCBD.cable.ziggo.nl) |
| 19:19.29 | phh | i put this line between 12MHz and 19MHz i guess ? |
| 19:19.43 | cr2 | rate=11000000 e |
| 19:19.49 | *** join/#htc-linux luke-jr (n=luke-jr@2002:62b3:1d4c:0:20e:a6ff:fec4:4e5d) |
| 19:19.50 | cr2 | means it was requested |
| 19:19.57 | cr2 | yes |
| 19:20.35 | cr2 | awk '{print 921600*16}' |
| 19:20.36 | cr2 | 14745600 |
| 19:21.04 | cr2 | awk '{print 3686400*16}' |
| 19:21.05 | cr2 | 58982400 |
| 19:21.28 | cr2 | <PROTECTED> |
| 19:21.34 | luke-jr | so what is the potential for this project? :p |
| 19:21.40 | cr2 | i'm thinking if this is the best way to get 64MHz |
| 19:22.08 | cr2 | because wince uses these divisors for 1/4 clock , i.e. 16MHz |
| 19:22.21 | cr2 | and i've just disabled the prescaler |
| 19:22.26 | cr2 | =4 |
| 19:23.48 | cr2 | phh: but if they really want 11MHz, we can make such and entry close to it |
| 19:24.06 | phh | i'm not sure this timer is a problem |
| 19:24.08 | cr2 | phh: the spl uses 120000 bps instead of 115200 actually |
| 19:24.50 | cr2 | yeah, you just operate the bus at 7MHz |
| 19:25.07 | cr2 | makes transfers 2 times slower. |
| 19:26.15 | cr2 | err |
| 19:26.38 | cr2 | the divided baud rates are /4 |
| 19:26.53 | cr2 | the undivided should be the same ??? |
| 19:27.08 | cr2 | for bt. |
| 19:27.15 | cr2 | need to think about it. |
| 19:28.40 | luke-jr | any chance on a free Rhodium? :p |
| 19:29.49 | cr2 | luke-jr: compile the blac100 kernel (not the latest git revision), and change the framebuffer address in htc_fb_console.c |
| 19:30.00 | luke-jr | cr2: really? and it works? :o |
| 19:30.12 | cr2 | it will crash in the process. |
| 19:30.39 | luke-jr | -.- |
| 19:30.42 | cr2 | the topaz crash screenshot is at htc-linux.org |
| 19:31.10 | cr2 | but the crash data help to fix the differences. |
| 19:32.00 | cr2 | luke-jr: it's just a manpower problem :) |
| 19:32.03 | luke-jr | I see. |
| 19:32.15 | luke-jr | so in other words, by the time it's usable it will be obsolete? :p |
| 19:33.06 | cr2 | i doubt it |
| 19:33.48 | cr2 | i prefer to make raph/diam usable first |
| 19:33.58 | cr2 | maybe because i have one ;) |
| 19:34.33 | luke-jr | I don't have any. |
| 19:34.36 | luke-jr | in the market XD |
| 19:35.25 | cr2 | but since rhod100 has the same cpu, and many peripherals are the same, it is much easier to support it, than a completely unknown device. |
| 19:35.48 | cr2 | well, if i had one, i will still finish raph100 first. |
| 19:36.11 | cr2 | because it's a better phone (smaller) |
| 19:36.14 | luke-jr | so is the SoC fully supported? O.o |
| 19:36.25 | luke-jr | wants as big as is reasonable in his pocke |
| 19:36.26 | luke-jr | pocket |
| 19:36.47 | cr2 | i have athena for that |
| 19:37.01 | luke-jr | Athena? |
| 19:37.24 | luke-jr | too little RAM |
| 19:37.26 | cr2 | an older device |
| 19:37.32 | cr2 | yes. |
| 19:37.35 | *** join/#htc-linux root2 (n=root@rgnb-5d87df24.pool.mediaWays.net) |
| 19:37.38 | luke-jr | if it had more RAM I might consider it :p |
| 19:37.52 | luke-jr | but my N810 proves 128 MB is too little for even KDE without apps |
| 19:37.52 | cr2 | but a fast ide hdd, and real sirf3 gps |
| 19:38.03 | cr2 | not that shitty qualcomm gps |
| 19:38.29 | luke-jr | is the Clio an upgraded Athena? |
| 19:38.33 | luke-jr | I couldn't find specs on that |
| 19:38.42 | cr2 | i prefer apps to kde |
| 19:38.46 | luke-jr | I'm guessing from the pic it's too big for a pocket |
| 19:38.56 | cr2 | clio is an msm device ? |
| 19:39.01 | luke-jr | http://wiki.xda-developers.com/index.php?pagename=HTC_Clio |
| 19:39.05 | cr2 | 4" |
| 19:39.10 | luke-jr | the website categorized it with Athena |
| 19:39.23 | luke-jr | is bad at putting measurements into visuals |
| 19:39.24 | cr2 | no |
| 19:39.29 | luke-jr | does 4" fit in a pocket? XD |
| 19:39.32 | cr2 | it's a x86 device |
| 19:39.35 | luke-jr | oh :/ |
| 19:39.37 | cr2 | a real notebook |
| 19:39.42 | luke-jr | ... |
| 19:39.45 | luke-jr | I want a real notebook |
| 19:39.47 | luke-jr | but not x86 |
| 19:40.01 | luke-jr | and one that fits in my pocket :p |
| 19:41.56 | luke-jr | meh |
| 19:42.00 | luke-jr | found specs, 7" display |
| 19:42.01 | luke-jr | way too big |
| 19:42.18 | cr2 | notebook |
| 19:48.48 | phh | cr2: it seems that the wifi driver wants the 29 IRQ, but according to http://htc-linux.org/wiki/index.php?title=Raphael_IRQ it seems wrong |
| 19:49.35 | luke-jr | cr2: is the size the only difference between Rhod and Raph? |
| 19:49.41 | phh | hum, not that one, else it insmod will return -ENOMEM |
| 19:51.15 | cr2 | luke-jr: they have different phone software. i've documented some differences, but need a bit more research |
| 19:51.35 | cr2 | phh: wifi irq ? |
| 19:52.18 | phh | cr2: it seems TI driver asks for an irq in the init code |
| 19:52.27 | phh | but it depends on deeper TI functions so not totally sure ... |
| 19:52.28 | cr2 | yes, it's hardcoded |
| 19:52.36 | cr2 | change it there |
| 19:52.56 | cr2 | don't know a better way with the .27 kernel |
| 19:53.14 | phh | to 24? |
| 19:53.33 | cr2 | no |
| 19:53.34 | phh | called INT_SDC1_0 |
| 19:53.42 | cr2 | it's wrong wiki page |
| 19:53.45 | phh | ok |
| 19:53.55 | cr2 | raphael_gpio |
| 19:54.33 | cr2 | 29 0x1d 3,0 13 I wifi irq |
| 19:54.36 | cr2 | 29 |
| 19:54.43 | cr2 | so it's 29 |
| 19:54.52 | phh | so it's the good one. |
| 19:54.57 | cr2 | yes |
| 19:55.01 | luke-jr | cr2: hardware I mean. I have no interest in retaining their original software |
| 19:55.21 | luke-jr | are these phones really USB client only, or does the hardware support host/OTG mode |
| 19:55.48 | cr2 | luke-jr: amss is the phone firmware. you more or less must retain it |
| 19:56.02 | cr2 | raph100 supports usb host/otg |
| 19:56.10 | cr2 | don't know about rhod100 |
| 19:56.22 | luke-jr | cr2: firmware, sure, but that's independent from the software... |
| 19:56.33 | cr2 | ok |
| 19:56.39 | luke-jr | isn't it? |
| 19:57.02 | cr2 | i meant phone firmware (aka amss), not phone software (aka wince) |
| 19:57.26 | cr2 | wince can be replaced with linux |
| 19:57.35 | cr2 | amss is another story |
| 19:57.38 | luke-jr | but amss doesn't run on the CPU, correct? |
| 19:57.45 | cr2 | google phones use different amss |
| 19:57.53 | cr2 | there are 2 cpus |
| 19:58.05 | cr2 | one runs amss, the other wince/linux |
| 19:58.07 | luke-jr | well, amss is like a hardware modem except supports voice? |
| 19:58.14 | cr2 | yes |
| 19:58.20 | luke-jr | k |
| 19:58.31 | luke-jr | presumably the audio goes through Linux still? |
| 19:58.44 | cr2 | it's a shared DSP |
| 19:59.17 | luke-jr | but I could setup an Asterisk on the Linux side and have it use the modem to call people independent of me? |
| 19:59.38 | cr2 | i think yes |
| 19:59.45 | cr2 | but nobody tried it so far. |
| 19:59.54 | luke-jr | heh |
| 20:00.03 | luke-jr | well, I guess it would be like Bluetooth headset setup |
| 20:00.23 | cr2 | this should certainly work |
| 20:00.25 | luke-jr | Rhod300 is significantly different beast than Rhod100 or no? |
| 20:00.37 | cr2 | don't know. but it's unlikely |
| 20:00.56 | luke-jr | unlikely the same or unlikely different? XD |
| 20:00.59 | cr2 | if both are gsm/umts devices, they should be more or less identical |
| 20:01.20 | luke-jr | oh, no difference at all? O.o |
| 20:01.27 | luke-jr | wonders why they have different numbers |
| 20:01.31 | cr2 | raph100 and raph800 have some substantial differences, but one is CDMA and the other is gsm |
| 20:01.52 | cr2 | missing second camera, different keyboard. such things. |
| 20:02.06 | luke-jr | O.o |
| 20:02.19 | luke-jr | but more or less the software end shouldn't care? |
| 20:02.29 | cr2 | yes |
| 20:03.18 | luke-jr | so if I picked up a Rhodium, it is likely I could get Gentoo+KDE up and working on it? ;) |
| 20:03.27 | luke-jr | background on me: I've got Linux booting on a cable modem in the past |
| 20:03.34 | luke-jr | (though not usable) |
| 20:03.43 | cr2 | phh: can you do a haret test for me ? |
| 20:04.03 | cr2 | luke-jr: they say zubuntu works |
| 20:04.12 | cr2 | luke-jr: ok |
| 20:06.35 | luke-jr | would I need an "unlocked" phone, or does that matter once WinCE is wiped? |
| 20:07.24 | cr2 | it's on the phone side. |
| 20:07.35 | luke-jr | oh, that amss thing? :/ |
| 20:07.36 | cr2 | !linux issue |
| 20:07.42 | cr2 | yes |
| 20:08.04 | cr2 | they hack/reflash it, but it has nothing to do with linux. |
| 20:08.14 | luke-jr | can we modify the amss side and swap it with an unlocked one? XD |
| 20:08.30 | cr2 | check xda-dev |
| 20:08.36 | luke-jr | ponders if amss is ARM or MIPS and small enough to reverse engineer |
| 20:08.45 | cr2 | arm9 |
| 20:08.54 | cr2 | 15MB arm+thumb code |
| 20:08.58 | cr2 | elf image |
| 20:08.59 | luke-jr | ouch, pretty big |
| 20:09.33 | luke-jr | N810's battery charger is only 110 KB and I'm still quite far from anything usable |
| 20:09.34 | luke-jr | :( |
| 20:09.44 | cr2 | it may be interesting to remove some junk from there. but may be too difficult |
| 20:10.21 | cr2 | battery charger done in software ? |
| 20:10.34 | phh | cr2: sure |
| 20:10.52 | luke-jr | cr2: yeah |
| 20:10.57 | cr2 | phh: telnet ? |
| 20:11.08 | luke-jr | N810 has proprietary userland blobs for battery charger and GPS |
| 20:11.28 | phh | just wait a minute for WinMo to boot |
| 20:11.44 | cr2 | phh: ok |
| 20:11.56 | cr2 | luke-jr: the lcd was good. everything else, hmm... |
| 20:12.17 | luke-jr | cr2: ? |
| 20:12.37 | cr2 | luke-jr: n810 |
| 20:12.48 | luke-jr | oh |
| 20:12.49 | luke-jr | pretty much, yeah |
| 20:12.50 | cr2 | phh: looking where is the baud divisor registers |
| 20:13.13 | luke-jr | the best I can do with N810 is wardrive, if the GPS feels like working |
| 20:13.24 | phh | cr2: how ? |
| 20:13.25 | luke-jr | totally not worth the $200 I spent on it :( |
| 20:13.45 | cr2 | phh: pd 0xa86000d8 8 |
| 20:14.06 | cr2 | luke-jr: others like it, but i think there is better hardware |
| 20:14.26 | phh | a86000d8 | 0019ff9f ffb80b44 | ....D... |
| 20:14.44 | cr2 | if we had a bit more committed people, htc would have been the biggest linux bastion :) |
| 20:15.00 | cr2 | phh: bt on., 4MHz mode |
| 20:15.02 | luke-jr | I can't become committed unfortunately :/ |
| 20:15.37 | cr2 | luke-jr: imvho htc makes the best phone hardware |
| 20:15.44 | cr2 | aka best linux hardware :) |
| 20:15.52 | luke-jr | seems like it |
| 20:16.09 | luke-jr | I get labelled a troll in #Maemo for even looking at non-Nokia stuff :/ |
| 20:16.32 | luke-jr | as if Nokia somehow deserves support |
| 20:16.36 | cr2 | i wait for maemo being reqritten in qt :) |
| 20:16.58 | luke-jr | the GTK codebase was crappy enough even for GTK code, so I'm not too hopeful about the Qt port |
| 20:17.28 | luke-jr | the N810's soft-GPS driver requires GTK+ and is hard-coded to chown a file to user name "user" or error out |
| 20:17.43 | phh | cr2: anything else ? |
| 20:17.59 | cr2 | phh: looking for the register |
| 20:18.05 | phh | ok |
| 20:18.08 | luke-jr | to get it working in Gentoo at all, I had to write a LD_PRELOAD lib and zero out some bits |
| 20:18.19 | cr2 | hehe |
| 20:18.26 | luke-jr | (I refuse to install GTK :p) |
| 20:19.19 | cr2 | UARTDM_CSR_ADDR |
| 20:19.41 | cr2 | /* write only register */ |
| 20:19.43 | cr2 | #define UARTDM_CSR_ADDR 0x8 |
| 20:19.44 | cr2 | err |
| 20:19.48 | cr2 | and nontraceable. |
| 20:20.05 | cr2 | phh: bad. need to find another way ;) |
| 20:22.18 | cr2 | M=25 N=96 D=48 P=1 S=4 C=2 |
| 20:23.39 | cr2 | phh: add the PRE=3 to the 64MHz line, and try to load firmware |
| 20:24.03 | cr2 | it is |
| 20:24.03 | phh | hu ? |
| 20:24.06 | cr2 | <PROTECTED> |
| 20:24.08 | cr2 | no |
| 20:24.10 | cr2 | now |
| 20:24.14 | cr2 | change it to |
| 20:24.22 | cr2 | <PROTECTED> |
| 20:24.40 | phh | just 0 => 3 ? |
| 20:24.50 | cr2 | yes |
| 20:25.34 | cr2 | phh: the list should be sorted by frequency |
| 20:26.00 | phh | the one you pasted is 4MHz ? |
| 20:26.32 | phh | so just before the /* SD */ ? |
| 20:26.58 | tmzt | so, anything new? |
| 20:27.09 | tmzt | cr2: did you see what I typed last nigh? |
| 20:27.10 | cr2 | phh: move it to be the last entry in the table |
| 20:27.11 | tmzt | nigt |
| 20:27.20 | cr2 | tmzt: what ? |
| 20:28.29 | luke-jr | can't find any sign of someone who has run Linux on a Rhodium |
| 20:28.30 | tmzt | 17:31 < tmzt> dcordes: can you push the thing for you on people? |
| 20:28.30 | tmzt | 17:32 < tmzt> for dcordes |
| 20:28.30 | tmzt | 17:32 < tmzt> I'm also working on a new db driver |
| 20:28.30 | tmzt | 17:33 -!- rashire [n=ed1112wa@pool-98-114-209-5.phlapa.fios.verizon.net] has joined #htc-linux |
| 20:28.33 | tmzt | 17:33 < tmzt> starting by adding single buffer updates to the existing one |
| 20:28.36 | tmzt | 17:33 < tmzt> but that causes tearing in my test |
| 20:28.39 | tmzt | 17:34 < tmzt> but I think I figured out how to adapt via drm driver |
| 20:28.40 | tmzt | s/db/fb |
| 20:29.06 | tmzt | luke-jr: some people have asked about it and I think tried it, and the amss is closer to google's |
| 20:29.17 | tmzt | do you have a rhod? |
| 20:29.26 | luke-jr | tmzt: thinking of buying one |
| 20:29.49 | tmzt | it should be supportable after a cleanup of our tree |
| 20:29.57 | cr2 | tmzt: sounds good |
| 20:30.04 | tmzt | cr2: I'm going to try importing the aurora fb driver actaully |
| 20:30.50 | tmzt | then work on drm |
| 20:30.50 | cr2 | tmzt: btw, i didn't see at channel in the topa smem dump |
| 20:30.50 | phh | cr2: failed :p |
| 20:30.50 | luke-jr | hmm |
| 20:30.51 | luke-jr | how is 2D/3D accel btw? |
| 20:30.51 | cr2 | tmzt: maybe they use qmi there too |
| 20:30.51 | tmzt | since stefen_schmidt/mickeyl will need a working ddx for the Pre and I think they are very similar (mdp/omap2fb) |
| 20:30.52 | phh | cr2: http://pastebin.com/m2297603f |
| 20:30.52 | cr2 | luke-jr: not free software |
| 20:31.07 | luke-jr | :/ |
| 20:31.14 | luke-jr | cr2: it does exist as non-free? |
| 20:31.29 | tmzt | 3d? |
| 20:31.31 | cr2 | phh: ok. it's something for me to think about |
| 20:31.44 | cr2 | luke-jr: yes. and it works. inside android |
| 20:31.45 | tmzt | that requires the libhgl from g1/sapp/hero |
| 20:31.49 | phh | cr2: i hope you have a great todo list manager :p |
| 20:32.07 | tmzt | so you are working on bluetooth or wifi now? |
| 20:32.17 | cr2 | both |
| 20:32.19 | luke-jr | cr2: but not vanilla X.org? |
| 20:32.39 | tmzt | luke-jr: that's what I'm working on, I want to let X drive the dma to the panel |
| 20:32.44 | tmzt | instead of the msm_fb driver |
| 20:32.50 | luke-jr | O.o |
| 20:33.19 | cr2 | luke-jr: i've tried to boot a 2.8GB debian rootfs right now. it didn't work. but maybe i'm missing some kernel fb update patch |
| 20:33.30 | tmzt | as for libhgl, it should be possible to wrap it with mesa code for gles |
| 20:33.35 | cr2 | tmzt: why Xfbdev didn't work for me |
| 20:33.36 | tmzt | like google does with libagl |
| 20:33.55 | luke-jr | meh, I'd rather 100% free than 3D anyway |
| 20:33.59 | cr2 | tmzt: is it possible to ltrace libhgl ? |
| 20:34.00 | tmzt | cr2: with the update thread? there's a ioctl that needs to return success or disable checking the pixclock value |
| 20:34.03 | tmzt | I had a patch |
| 20:34.24 | tmzt | probably, but without a lot of gpu knowledge it would be difficult to do |
| 20:34.27 | cr2 | tmzt: how is wince doing it |
| 20:34.35 | cr2 | ok |
| 20:34.38 | tmzt | not sure it's worth it on pre 7x727 at hardware |
| 20:34.41 | tmzt | doing what? |
| 20:34.46 | tmzt | 3d? badly :) |
| 20:35.01 | tmzt | they seem to use pmem also |
| 20:35.03 | cr2 | tmzt: i'll try to cleanup the mddi setup for raph100 now, but some things are different from g1 |
| 20:35.15 | tmzt | if I get drm working that would be a starting place for a 2d driver using gpu |
| 20:35.19 | phh | (i confirm 3D on android works *really* great) |
| 20:35.24 | tmzt | since the memory management will be done |
| 20:35.40 | tmzt | sure, you splitting out the client chips? |
| 20:35.46 | phh | (neocore is only 20% slower than on G1 <3) |
| 20:35.50 | cr2 | tmzt: yes. i don't care about 3d. but 2d accel, bitblit and such are great |
| 20:35.55 | tmzt | I can't follow the existing code and I know it's not being used on raph500 |
| 20:36.20 | cr2 | tmzt: which mddi client do you have ? |
| 20:37.47 | cr2 | phh: the BT todo is only to find out the wince baud divisor value for 4MHz speed |
| 20:38.09 | phh | cr2: ok |
| 20:38.22 | phh | the wifi thing is a lot harder :/ |
| 20:38.28 | phh | ni actual idea of what is wrong... |
| 20:38.31 | phh | no* |
| 20:38.58 | cr2 | M=25 N=96 D=48 P=1 S=4 C=2 |
| 20:39.27 | cr2 | 0x19 0x60 0x30 prescaler=2 ?? pll0 |
| 20:39.38 | cr2 | then your speed should be |
| 20:43.09 | cr2 | 245760000*25/96/2 |
| 20:43.09 | phh | cr2: i looked at interrupts and to me it seems *really* weird: |
| 20:43.10 | phh | http://pastebin.com/m658332f9 |
| 20:43.10 | cr2 | 32MHz hmm |
| 20:43.38 | *** join/#htc-linux root2 (n=root@rgnb-5d87df24.pool.mediaWays.net) [NETSPLIT VICTIM] |
| 20:43.38 | *** join/#htc-linux luke-jr (n=luke-jr@2002:62b3:1d4c:0:20e:a6ff:fec4:4e5d) |
| 20:43.38 | *** join/#htc-linux MLM (n=mlvdmeid@5ED0BCBD.cable.ziggo.nl) [NETSPLIT VICTIM] |
| 20:43.38 | *** join/#htc-linux balans (n=BRPOLDB@212-123-149-239.ip.telfort.nl) [NETSPLIT VICTIM] |
| 20:43.38 | *** join/#htc-linux GNUtoo (n=gnutoo@host134-152-dynamic.51-79-r.retail.telecomitalia.it) |
| 20:43.38 | *** join/#htc-linux kiozen_away (n=kiozen@rgnb-5d87c591.pool.mediaWays.net) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux rashire (n=ed1112wa@pool-98-114-209-5.phlapa.fios.verizon.net) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux infernix (i=nix@unaffiliated/infernix) |
| 20:43.39 | *** join/#htc-linux nozze (n=nozze@u193-11-162-41.studentnatet.se) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux thinko (n=jdoe6alp@smaug.rackdragon.com) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux high-rez (n=gus@carrera.bourg.net) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux jmaurice (i=jmaurice@snickers.org) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux ecze (n=ecze@eczema.ecze.com) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux KungFunk (n=cow@c-e1d571d5.014-46-73746f28.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux Kevin2 (n=Kevin2@207-237-194-161.c3-0.avec-ubr2.nyr-avec.ny.cable.rcn.com) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux Perkka (n=perkka@ua-83-227-207-92.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux astrolite (n=astrolit@c-d7c9e455.023-352-6c6b701.cust.bredbandsbolaget.se) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux dtrack (n=LOL@78-86-59-180.zone2.bethere.co.uk) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux bartman (n=bart@tau.jukie.net) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux madCoder` (n=madcoder@c-68-42-30-56.hsd1.pa.comcast.net) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux Dinde (i=kayser@81-65-176-209.rev.numericable.fr) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) |
| 20:43.39 | *** join/#htc-linux tmzt (n=tmzt@adsl-99-52-65-233.dsl.akrnoh.sbcglobal.net) |
| 20:43.39 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-67-49-192-228.bak.res.rr.com) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) [NETSPLIT VICTIM] |
| 20:43.39 | *** join/#htc-linux bertramt__ (n=chatzill@63.246.89.17) [NETSPLIT VICTIM] |
| 20:43.43 | cr2 | phh: change the prescaler to 1 (from 3) |
| 20:43.44 | cr2 | it's the sdio side |
| 20:43.45 | tmzt | cr2: did you confirm if wifi data is over sd bus? |
| 20:43.45 | cr2 | <PROTECTED> |
| 20:43.45 | cr2 | <PROTECTED> |
| 20:43.45 | cr2 | <PROTECTED> |
| 20:43.45 | cr2 | <PROTECTED> |
| 20:43.45 | cr2 | it's sdc1, sdc2, sdc3, sdc4 |
| 20:43.46 | *** join/#htc-linux dzo (n=dzo@mail.marginz.co.nz) |
| 20:43.46 | cr2 | tmzt: it should be easily traceable |
| 20:44.05 | tmzt | there are instruction for tracing dma sg? |
| 20:44.07 | cr2 | tmzt: but the firmware/eeprom are loaded over sdio |
| 20:44.16 | tmzt | not the data itself I mean, but the commands/addresses |
| 20:44.39 | phh | cr2: and tiwlan0 on 93 ? |
| 20:44.47 | cr2 | tmzt: if you don't get SD irqs using wifi, wifi does not use it. 'wi' should be enough |
| 20:44.58 | tmzt | what are those numbers? |
| 20:45.31 | cr2 | phh: the tiacx irq is msmgpio, not msm |
| 20:45.39 | cr2 | it's an extrernal device |
| 20:45.53 | cr2 | tmzt: /proc/interrupts |
| 20:45.56 | phh | oh ok |
| 20:46.14 | cr2 | 87: 0 msmgpio msm-sdcc (slot) |
| 20:46.24 | cr2 | it's probably gpio 0x26 -> SD detect |
| 20:46.45 | tmzt | right, ok |
| 20:46.50 | cr2 | so 93 (-87)=6 |
| 20:46.51 | tmzt | why cmd? |
| 20:47.05 | cr2 | 0x26+6 |
| 20:47.18 | tmzt | if it's pio would there be irqs then? |
| 20:47.19 | cr2 | tmzt: androids called it this way |
| 20:47.32 | cr2 | pio means unused |
| 20:47.47 | cr2 | tmzt: is it possible to disable dma for sdc ? |
| 20:48.02 | tmzt | in pdata probably |
| 20:48.18 | tmzt | doesn't uboot do that? |
| 20:48.29 | cr2 | 23 0x17 3,0 7 I SD card inserted (and irq) |
| 20:48.37 | tmzt | right, ok |
| 20:48.37 | cr2 | 29 0x1d 3,0 13 I wifi irq |
| 20:48.51 | cr2 | tmzt: didn't look |
| 20:49.04 | tmzt | I think they just labeled it 'wifi' to distinguish it from the sd slot |
| 20:49.08 | cr2 | phh: so gpio23 is 87: 0 msmgpio msm-sdcc (slot) |
| 20:49.18 | tmzt | but they should be the six lines to the sd bus (wifi is 4bit right?) |
| 20:49.21 | cr2 | phh: you will get the same for gpio 29 |
| 20:49.40 | cr2 | tmzt: it's setup this way. |
| 20:49.46 | phh | i should get 93 you mean ? whici is the one i get |
| 20:49.57 | cr2 | yes |
| 20:50.01 | cr2 | it's wifi irq |
| 20:50.08 | cr2 | sd detect is 87: |
| 20:50.36 | cr2 | phh: yes. and even 1 irq :) |
| 20:51.18 | cr2 | phh: i'm looking for the posting about diam wifi |
| 20:51.27 | phh | cr2: i bet i'm gonna try with 2 for BT. |
| 20:51.39 | cr2 | 1 |
| 20:51.46 | tmzt | from druidu? |
| 20:51.51 | phh | http://pastebin.com/m592a1494 |
| 20:51.51 | cr2 | no |
| 20:52.18 | phh | (hciattach does the same thing as usual) |
| 20:52.18 | cr2 | phh: the dmesg will not change |
| 20:52.24 | cr2 | ok |
| 20:52.24 | phh | ok |
| 20:52.40 | cr2 | so i really need to check the baud divisor |
| 20:52.55 | cr2 | phh: who needs 4Mbit BT ? |
| 20:53.08 | tmzt | who needs 4Mbit wifi? |
| 20:53.14 | phh | who needs BT ? :D |
| 20:53.24 | cr2 | gps will be happy with 4800 :) |
| 20:53.40 | phh | cr2: currently the maximum is 500kbauds (if i read correctly), for good A2DP it's too low |
| 20:53.42 | cr2 | tmzt: i mean which application really needs this bandwidth |
| 20:53.50 | cr2 | it costs mA |
| 20:53.54 | phh | (yes SBC totally sucks.) |
| 20:54.01 | cr2 | ok |
| 20:54.22 | tmzt | did I say wifi? |
| 20:54.26 | tmzt | I mean IrDA |
| 20:54.29 | phh | tmzt: you said wifi. |
| 20:54.32 | phh | oh. |
| 20:54.46 | phh | IrDA and BT aren't exactly the same thing. |
| 20:55.35 | tmzt | yes, I know |
| 20:55.52 | tmzt | but there is 4Mbps IrDA |
| 20:56.09 | cr2 | phh: hciattach /dev/ttyHS1 texas 4000000 ? |
| 20:56.24 | phh | cr2: without firmware then ? |
| 20:56.31 | cr2 | with |
| 20:56.34 | phh | ok |
| 20:56.59 | cr2 | without checking the baud divisor, it's all guesswork |
| 20:57.09 | phh | [ 508.349761] clock-wince: pc_clk_set_rate: id=35 rate=7372800 |
| 20:57.13 | phh | hciattach doesn't even get it ... |
| 20:57.25 | phh | ioctl(3, SNDCTL_TMR_TIMEBASE or TCGETS, {B4000000 -opost -isig -icanon -echo ...}) = 0 |
| 20:57.25 | phh | ioctl(3, SNDCTL_TMR_START or TCSETS, {B4000000 -opost -isig -icanon -echo ...}) = 0 |
| 20:57.25 | phh | ioctl(3, SNDCTL_TMR_START or TCSETS, {B115200 -opost -isig -icanon -echo ...}) = 0 |
| 20:57.27 | phh | "WTF" |
| 20:57.40 | cr2 | it starts at 115200 |
| 20:57.58 | cr2 | sends the .bts command "switch to 4000000" |
| 20:58.09 | cr2 | and then should switch to 4000000 himself |
| 20:58.22 | cr2 | the response does not come |
| 20:58.37 | phh | well if i say hciattach 4000000 he shouldn't go at 115200 at all ... |
| 20:58.37 | cr2 | so he says "Timed out" and returns to 115200 |
| 20:58.52 | cr2 | there is -s flag for initial speed |
| 20:59.10 | cr2 | i should really look at the hciattach source |
| 20:59.48 | tmzt | that's what it is, a few gpios and set the ldisc |
| 21:00.49 | phh | cr2: i've just been thinking... the strace of hciattach might be useful ? |
| 21:01.05 | phh | since it actually reads something after TCSETS 4M |
| 21:01.29 | phh | http://pastebin.com/m75aaa0a2 |
| 21:04.22 | phh | (i'll let you understand the BTS to know what happens/should happens.) |
| 21:04.41 | cr2 | 0 4 7 |
| 21:04.51 | cr2 | i'll look at .bts |
| 21:05.10 | *** join/#htc-linux billyonthemounta (i=Miranda@81.13.236.29) |
| 21:05.30 | tmzt | 0 4 7? |
| 21:06.54 | phh | cr2: that's what hciattach can read from BT chip |
| 21:07.16 | cr2 | # HTC |
| 21:07.18 | cr2 | # Update the Baudrate to 4000000 bps |
| 21:07.19 | cr2 | #-------------------------------------------------------------------------------- |
| 21:07.21 | cr2 | SEND_COMMAND: ogf=0x3f ocf=0x336 4 -> |
| 21:07.23 | cr2 | ACTION_SEND_COMMAND: 0x01 0x36 0xff 0x04 0x00 0x09 0x3d 0x00 |
| 21:07.24 | cr2 | ACTION_WAIT_EVENT : 5000 msec, 7 size, data = 0x04 0x0e 0x04 0x00 0x36 0xff 0x00 |
| 21:07.25 | cr2 | ACTION_SERIAL_PORT : BTS changing baud rate to 4000000, flow control to 1 |
| 21:07.27 | cr2 | it's the first command in the .bts |
| 21:08.45 | cr2 | http://wiki.xda-developers.com/index.php?pagename=UniversalBluetooth |
| 21:09.00 | phh | there is even not the right number of bytes in send ... |
| 21:09.05 | phh | oops |
| 21:09.10 | cr2 | 0x336 4 |
| 21:09.12 | phh | it is. |
| 21:09.51 | cr2 | ACTION_SERIAL_PORT : BTS changing baud rate to 4000000, flow control to 1 |
| 21:10.01 | cr2 | should set the right port speed |
| 21:10.12 | phh | it does according to strace |
| 21:10.25 | cr2 | ACTION_WAIT_EVENT : 5000 msec |
| 21:10.30 | cr2 | does not time out ? |
| 21:11.03 | cr2 | the ext command is |
| 21:11.03 | cr2 | SEND_COMMAND: ogf=0x3f ocf=0x237 2 -> |
| 21:11.04 | cr2 | ACTION_SEND_COMMAND: 0x01 0x37 0xfe 0x02 0x03 0x35 |
| 21:11.05 | cr2 | ACTION_WAIT_EVENT : 5000 msec, 7 size, data = 0x04 0x0e 0x04 0x00 0x37 0xfe 0x00 |
| 21:11.22 | cr2 | 0x237 2 + |
| 21:11.26 | cr2 | unknown |
| 21:11.38 | cr2 | but used by brf6150 too |
| 21:11.44 | phh | we receive 4 bytes instead 7 and not the good ones /o\ |
| 21:12.06 | cr2 | 7=3 (header)+4 |
| 21:12.18 | cr2 | ACTION_SEND_COMMAND: 0x01 0x36 0xff 0x04 0x00 0x09 0x3d 0x00 |
| 21:12.24 | cr2 | 7 bytes |
| 21:12.47 | cr2 | + 0x00 (end) |
| 21:13.30 | cr2 | if we get junk, then the clock speed is wrong |
| 21:13.48 | cr2 | can be only the baud divisor |
| 21:13.48 | phh | yep. |
| 21:14.06 | cr2 | the clock divisors i see in wince |
| 21:14.52 | cr2 | the baud divisor in g1 code is 0xff (aka 4) |
| 21:15.18 | cr2 | need to find the value written into this register in wince. |
| 21:15.44 | cr2 | it may have been traced, but the device hangs on tracing thr UART2DM regs |
| 21:16.18 | cr2 | you may try, but i think it will be the same result |
| 21:16.34 | cr2 | dump mmu 2 0xa0300000 0x1000 |
| 21:17.05 | cr2 | i need to find what's written into 0xa0300008 |
| 21:17.57 | cr2 | # Enable deep sleep mode |
| 21:17.59 | cr2 | SEND_COMMAND: ogf=0x3f ocf=0x10c 9 -> |
| 21:19.35 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 21:24.48 | cr2 | phh: hciattach -n -s 115200 /dev/ttyHS0 texas 4000000 flow |
| 21:24.54 | cr2 | full command on g1 |
| 21:25.34 | cr2 | git://android.git.kernel.org/platform/external/bluez.git |
| 21:26.11 | cr2 | but they are not using .bts |
| 21:28.19 | phh | hciattach works but hciconfig hci0 up not |
| 21:28.57 | tmzt | cr2: what do I need to get the audio path for microphone? |
| 21:32.03 | phh | and s/4000000/115200/g works as usual |
| 21:32.44 | cr2 | phh: dmesg ? |
| 21:32.53 | cr2 | phh: which clock like are you using ? |
| 21:33.06 | cr2 | http://android.git.kernel.org/?p=platform/external/bluez.git;a=blob;f=utils/tools/hciattach_tialt.c;h=4df91f658dbe2c203f41fdb4e5cdd0f0293083bb;hb=36cb3b0a0fcde224b549e2d46275a7481320b3c3 |
| 21:33.12 | phh | <PROTECTED> |
| 21:33.17 | cr2 | this is how they load dump |
| 21:33.28 | cr2 | phh: try 0 for the prescaler |
| 21:35.23 | *** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821) |
| 21:39.25 | cr2 | phh: there was some simle patch for sdio loader, changing timeout. |
| 21:40.06 | phh | well i've seen somethings like this for kaiser, but i read the patch and... just didn't get what they were doing |
| 21:40.18 | *** part/#htc-linux balans (n=BRPOLDB@212-123-149-239.ip.telfort.nl) |
| 21:41.59 | phh | same, hciattach works, hciconfig hci0 up doesn't |
| 21:42.01 | phh | http://pastebin.com/m3ccbfc63 |
| 21:42.41 | cr2 | hciconfig -a ? |
| 21:43.29 | phh | hci0: Type: UART |
| 21:43.29 | phh | <PROTECTED> |
| 21:43.29 | phh | <PROTECTED> |
| 21:43.29 | phh | <PROTECTED> |
| 21:43.30 | phh | <PROTECTED> |
| 21:45.14 | cr2 | not found |
| 21:45.47 | tmzt | not found? |
| 21:45.51 | tmzt | what wasn't found? |
| 21:49.45 | *** join/#htc-linux dzo (n=dzo@mail.marginz.co.nz) |
| 21:50.23 | cr2 | 337 /* Print LMP version */ |
| 21:50.25 | cr2 | 338 fprintf(stderr, "Texas module LMP version : 0x%02x\n", resp[10] & 0xFF); |
| 21:50.26 | cr2 | 339 |
| 21:50.28 | cr2 | 340 /* Print LMP subversion */ |
| 21:50.29 | cr2 | 341 fprintf(stderr, "Texas module LMP sub-version : 0x%02x%02x\n", resp[14] & 0xFF, resp[13] & 0xFF); |
| 21:50.40 | cr2 | phh: do you get these messages ? |
| 21:50.53 | phh | no |
| 21:51.01 | phh | cr2: which hciattach do i have to build ? |
| 21:51.45 | phh | hciattach, hciattach_ti or hciattach_tialt ? |
| 21:52.08 | cr2 | alt loads binary fw |
| 21:52.12 | cr2 | not the .bts |
| 21:52.43 | cr2 | i actually don't see the bts loader here |
| 21:54.14 | cr2 | dcordes used OE to compile ? |
| 21:54.20 | cr2 | it must have .bts paches |
| 21:55.37 | luke-jr | cr2: btw, is your aim in here to port Android, or to port Linux? |
| 21:57.28 | cr2 | linux |
| 21:57.40 | cr2 | android is just an application |
| 21:58.00 | phh | err |
| 21:58.05 | cr2 | if googels have done it right, there would have been no need for android "porting" |
| 21:58.09 | phh | it's an application with really dirty hacks in the kernel |
| 21:58.13 | luke-jr | Android userland does not support Linux |
| 21:58.15 | cr2 | yeah |
| 21:58.17 | luke-jr | just its fork |
| 21:58.43 | ali1234 | android kernel isn't a fork, it's a patch... |
| 21:58.49 | luke-jr | in other words, can I get a vanilla kernel with this hardware support? :p |
| 21:58.54 | phh | ali1234: what's the difference ? |
| 21:58.57 | cr2 | phh: not only in the kernel, but also in the firmware |
| 21:59.05 | luke-jr | ali1234: patches are fork deltas |
| 21:59.07 | phh | cr2: oO |
| 21:59.24 | ali1234 | the difference is small but important |
| 22:00.06 | ali1234 | the android patches are relatively easy to apply to any kernel |
| 22:00.07 | cr2 | ali1234: such patches as pmem undermine the kernel dma api |
| 22:00.09 | luke-jr | the reason I ask is I see all these HTC git repos named "android" in there |
| 22:00.28 | cr2 | luke-jr: buzzword compliance |
| 22:00.29 | luke-jr | but I don't want android code on my Gentoo phone :p |
| 22:00.42 | ali1234 | luke-jr: then you want gizard |
| 22:00.48 | tmzt | luke-jr: what phone do you want? |
| 22:00.50 | ali1234 | aka gentoo armel |
| 22:00.52 | tmzt | or have |
| 22:00.56 | luke-jr | tmzt: I want to get a Touch Pro2 I think |
| 22:01.01 | tmzt | ok, right |
| 22:01.06 | luke-jr | unless there's something better |
| 22:01.07 | cr2 | ali1234: does it have omap hacks ? |
| 22:01.14 | luke-jr | 512 MB RAM would be a nice upgrade :p |
| 22:01.15 | ali1234 | no |
| 22:01.16 | tmzt | that would be a good gentoo phone (even if I would prefer debian) |
| 22:01.28 | luke-jr | ali1234: Gentoo has someone already working on HTC phones? |
| 22:01.34 | ali1234 | yep |
| 22:01.35 | cr2 | ali1234: does it mean i can use it on pxa too ? |
| 22:01.48 | tmzt | oh, I have a keymap I'm working on for raph |
| 22:01.51 | cr2 | not on s3c244x probably |
| 22:01.51 | ali1234 | cr2: it's gentoo... you can recompile it on any arch you want |
| 22:01.55 | tmzt | but I would like to see the u key work |
| 22:02.01 | cr2 | because it's armv4 |
| 22:02.07 | luke-jr | cr2: ... |
| 22:02.08 | ali1234 | gizard is just an overlay with phone-type stuff in it |
| 22:02.10 | tmzt | armv5 |
| 22:02.15 | tmzt | pxa right? |
| 22:02.18 | tmzt | oh |
| 22:02.18 | luke-jr | cr2: Gentoo will run on anything Linux supports, and more |
| 22:02.21 | ali1234 | http://tinderbox.dev.gentoo.org/embedded/ |
| 22:02.27 | tmzt | gentoo can be built for whatever you want |
| 22:02.50 | luke-jr | I had Gentoo on my Zaurus back in the day |
| 22:02.53 | cr2 | i want a portable binary distro |
| 22:02.54 | ali1234 | dont ask me how to use that stuff though |
| 22:03.02 | luke-jr | cr2: that is self-contradictatory |
| 22:03.08 | ali1234 | you need to ask solar, he's doing most of it |
| 22:03.13 | cr2 | luke-jr: not for arm |
| 22:03.29 | luke-jr | cr2: if it's only "for arm", then it's not portable. |
| 22:03.46 | cr2 | well, portable enough |
| 22:03.48 | tmzt | what? |
| 22:03.53 | luke-jr | http://distfiles.gentoo.org/releases/arm/autobuilds/20090922/ <-- generic Gentoo stages |
| 22:03.55 | cr2 | portable means i can use it on any htc phone |
| 22:04.06 | tmzt | luke-jr: you have to build for an isa |
| 22:04.07 | luke-jr | cr2: portable means any architecture/platform :p |
| 22:04.11 | tmzt | not really anyway around that |
| 22:04.18 | tmzt | it's portable from source |
| 22:04.41 | ali1234 | luke-jr: that's not quite the same... gizard builds up a whole rootfs for you, rather than having to compile stuff on your phone or in qemu |
| 22:04.55 | luke-jr | ali1234: I'm perfectly fine building on my phone. |
| 22:04.56 | ali1234 | or rather gizard is the result of a bunch of scripts that do that |
| 22:05.15 | ali1234 | well, we're not. wizard only has 64mb and a slow slow processor |
| 22:05.19 | luke-jr | :p |
| 22:05.31 | luke-jr | yeah, that would kinda screw it up |
| 22:05.43 | luke-jr | the only problem I couldn't overcome on my N810 was qt-webkit |
| 22:05.50 | luke-jr | it required more than the 128 MB RAM to link |
| 22:06.00 | ali1234 | yeah it is huge |
| 22:06.09 | luke-jr | KDE itself went ok |
| 22:06.14 | ali1234 | when android was released they killed the webkit repo :) |
| 22:06.24 | luke-jr | uh? |
| 22:06.50 | ali1234 | the original android source download script pulled from the webkit repo, and it went down for a while. google had to make a mirror |
| 22:06.51 | cr2 | luke-jr: why do you want to have kde on the phone ? i'm using mainly only the window manager from it. |
| 22:06.55 | ali1234 | because webkit code is huge |
| 22:07.13 | tmzt | it has backends for everything |
| 22:07.36 | luke-jr | cr2: because I don't want a phone. I want a handheld PC with always-on internet and "oh, I can do phone via my Bluetooth headset too" bonus |
| 22:07.42 | cr2 | luke-jr: on my desktop machine |
| 22:07.46 | ali1234 | yeah, i want the same |
| 22:07.53 | luke-jr | cr2: oh, I only use KDE/Qt apps |
| 22:07.57 | cr2 | luke-jr: a sane idea. |
| 22:08.14 | ali1234 | it would be nice if i could use it as a phone too though, and so far android seems like the only solution for that :( |
| 22:08.15 | cr2 | luke-jr: firefox/gimp ? |
| 22:08.18 | ali1234 | the only workable one anyway |
| 22:08.26 | cr2 | luke-jr: openoffice ? |
| 22:08.28 | luke-jr | cr2: no. they suck. |
| 22:08.32 | tmzt | luke-jr: well if bluetooth works |
| 22:08.38 | luke-jr | Konqueror and Krita win easily |
| 22:08.40 | tmzt | cr2: try abiword |
| 22:08.44 | luke-jr | KOffice is decent, too |
| 22:08.56 | cr2 | abiword segfaults much more often that oo |
| 22:09.07 | luke-jr | I rarely need even KOffice |
| 22:09.14 | luke-jr | KWrite is good enough |
| 22:09.29 | luke-jr | I can write up some HTML for Konq if I need formatting |
| 22:09.31 | cr2 | luke-jr: i need to read the .doc , .xls and .ppt attachments |
| 22:09.47 | luke-jr | cr2: KOffice then :p |
| 22:10.07 | cr2 | is not so good in reading them for some reason |
| 22:10.23 | luke-jr | just need to extract the real text content, right? :p |
| 22:10.34 | ali1234 | that's what strings is for |
| 22:10.37 | tmzt | antiword is good |
| 22:10.39 | cr2 | sometimes, edit and send back |
| 22:10.45 | tmzt | not .rtf or .docx support |
| 22:10.45 | luke-jr | ali1234: strings won't work with compressed data |
| 22:10.58 | luke-jr | cr2: oh, then that's easy |
| 22:11.00 | luke-jr | KHexEdit |
| 22:11.02 | luke-jr | :p |
| 22:11.07 | cr2 | had some funny situations because of that :) |
| 22:11.21 | luke-jr | shrug |
| 22:11.33 | luke-jr | if I have to send Office documents, I get my boss to convert from ODW |
| 22:11.38 | ali1234 | for that matter google docs solves the problem nicely too |
| 22:11.38 | luke-jr | or whatever the open doc fmt is |
| 22:11.50 | luke-jr | no |
| 22:11.54 | luke-jr | Google is worse than Microsoft |
| 22:12.04 | ali1234 | i don't care :) |
| 22:12.05 | cr2 | where i have screwed the names of editors, and attached internal comments. because of the -.doc support bugs :) |
| 22:12.20 | ali1234 | if you don't want me to open your doc in google docs, don't send it to my gmail :) |
| 22:12.45 | cr2 | gmail is spyware microsoft can only dream of |
| 22:12.57 | ali1234 | microsoft have never been in the spyware business |
| 22:13.17 | cr2 | indirectly |
| 22:13.17 | ali1234 | they're trapped in the 90s |
| 22:13.17 | luke-jr | anyhow |
| 22:13.17 | luke-jr | if you want OO.o on your phone, you'll need to find one with 1 GB RAM or cross-compile |
| 22:13.17 | luke-jr | :p |
| 22:13.17 | luke-jr | bbl |
| 22:13.22 | ali1234 | that's why spyware is such an issue on windows.... they just dont get it |
| 22:14.41 | cr2 | google gets it right. with google maps and earth |
| 22:14.50 | tmzt | ocx? |
| 22:15.08 | cr2 | i really hate them for not allowing tracing imagery for openstreetmap |
| 22:17.40 | *** join/#htc-linux BHSPitLappy_ (n=BHSPitLa@adsl-66-140-35-240.dsl.rcsntx.swbell.net) |
| 22:17.53 | cr2 | phh: compile the tiacx.ko and add it to initramfs |
| 22:18.11 | phh | cr2: tiacx.ko ... ? where does that one come from ? |
| 22:18.16 | cr2 | phh: and the loader. then it can be debugged independently |
| 22:18.23 | cr2 | phh: the wifi driver |
| 22:18.42 | phh | the TI official one ? |
| 22:18.56 | cr2 | phh: you have an unofficial ? |
| 22:19.21 | phh | there is one: http://bobcopeland.com/android_wifi.html |
| 22:19.25 | cr2 | actually i think that fixing wifi and gps is more important than 4000000 |
| 22:19.30 | phh | i think so. |
| 22:19.56 | phh | i can build this wifi driver, but it needs some things to loads in the board, but i can't see what it is .. |
| 22:20.17 | phh | cr2: you really need to have it in initramfs ? |
| 22:20.31 | phh | since it's most likely there will be "some" rebuilds |
| 22:20.38 | tmzt | of? |
| 22:20.40 | cr2 | wpa can wait |
| 22:20.56 | cr2 | phh: then we need to merge some working source into git |
| 22:21.13 | cr2 | right now i have no idea which driver to use, and how to build it |
| 22:23.16 | cr2 | if we can't fix it in .27 , let's move to .29 |
| 22:23.45 | cr2 | i just want to finish with gps and raph panel. |
| 22:24.02 | phh | cr2: there is some backports to .27, i'll try it |
| 22:25.05 | cr2 | ali1234: is there some rootfs that i can just untar to a /dev/sda2 on SD, and mount ? |
| 22:25.14 | ali1234 | yeah |
| 22:25.30 | ali1234 | anything here will work http://tinderbox.dev.gentoo.org/embedded/linwizard/ |
| 22:25.45 | ali1234 | there might be a MSM specific one, tmzt, didn't solar make you one? |
| 22:26.09 | tmzt | yes |
| 22:26.18 | tmzt | and he also made me a full init version |
| 22:26.19 | cr2 | tmzt: do i need a patch for X |
| 22:26.25 | cr2 | kernel patch |
| 22:26.30 | tmzt | tmzt: look on people, I have a small one |
| 22:26.52 | cr2 | then i'm wondering why it's not in the kernel yet. |
| 22:27.03 | tmzt | http://people.openezx.org/tmzt/htcraphael-msmts-and-vres-1.patch |
| 22:27.09 | tmzt | because it breaks android |
| 22:27.33 | tmzt | I like that msmts, msm_fb and vkeyb are split with a hook |
| 22:27.38 | tmzt | much cleaner than the old code |
| 22:27.46 | ali1234 | ah: http://tinderbox.dev.gentoo.org/embedded/images/ |
| 22:28.23 | cr2 | msm-20090623.tar.bz2 ? |
| 22:28.33 | ali1234 | yep |
| 22:28.38 | cr2 | ok |
| 22:28.44 | ali1234 | and the -dev- one probably contains a toolchain too :) |
| 22:29.11 | tmzt | yes |
| 22:29.15 | tmzt | but it boots slowly |
| 22:29.39 | ali1234 | you can also install packages with qmerge - like emerge without the compiling |
| 22:29.45 | *** join/#htc-linux Weiss (n=taw27@zazu.bitwiz.org.uk) |
| 22:29.58 | ali1234 | at least you can on the wizard images |
| 22:30.49 | tmzt | hey Weiss |
| 22:31.12 | tmzt | Weiss: the framebuffer the lcm updates from is in a remote chip |
| 22:32.04 | tmzt | over the mddi bus |
| 22:32.05 | cr2 | tmzt: if android does non-standard things we should separate them, and mark appropriately |
| 22:32.24 | tmzt | cr2: that patch makes X work with the update thread in the kernel |
| 22:32.27 | cr2 | tmzt: with its own video ram |
| 22:32.31 | tmzt | right |
| 22:32.48 | cr2 | tmzt: what about raw framebuffer ? |
| 22:32.59 | tmzt | what raw framebuffer? |
| 22:33.11 | cr2 | i'd like to check the qlandkarteM from kiozen |
| 22:33.18 | tmzt | without update? X will start but it won't render anything |
| 22:33.26 | tmzt | with update thread that should work |
| 22:33.27 | cr2 | used by qtembedded |
| 22:33.30 | tmzt | right |
| 22:33.40 | cr2 | ok |
| 22:34.02 | cr2 | i still can't believe that the toshiba/epson does not support autorefrsh |
| 22:34.36 | tmzt | cr2: can you explain what's going on to Weiss? he's the one who built the drm driver for freerunner/glamo chip |
| 22:35.23 | cr2 | tmzt: i'm not 100% sure about the role of MDP too |
| 22:36.04 | tmzt | I want to replace msm_fb.c and use mdp.c with a new msmmdp_drv drm1 driver |
| 22:36.10 | Weiss | i'm not necessarily the best person to advise on this, but: |
| 22:37.11 | cr2 | tmzt: i see that mdp can do scaling, rotating and alpha blending. |
| 22:37.18 | tmzt | yes |
| 22:37.36 | Weiss | - "DRI2-style" memory management (usually, but not necessarily, using a GEM interface, and abstracting details of where things are resident in memory away to the kernel), is a good idea (some of the memory management things done in DRI1 are really horrible - statically dividing the memory between various uses, rather than having a single unified memory management) |
| 22:37.42 | tmzt | raster says that those are probably faster in software though |
| 22:38.27 | tmzt | right, that's the goal :) I think I can get it to work with a fixed framebuffer size though until we can reallocate parts of smi |
| 22:38.38 | tmzt | gem is certainly a goal though |
| 22:38.49 | tmzt | particularly for the offscreen buffers |
| 22:39.34 | cr2 | tmzt: what is the kmalloc upper limit on arm ? |
| 22:39.59 | tmzt | largest allocation? |
| 22:40.05 | tmzt | page size is 4k |
| 22:40.27 | cr2 | i have 128k in mind |
| 22:40.32 | tmzt | don't know mm layer well though, just what I've read about it |
| 22:40.37 | cr2 | but it may be more |
| 22:40.47 | tmzt | it's usually 2^n |
| 22:40.51 | tmzt | related to page size |
| 22:40.56 | tmzt | not sure how |
| 22:40.59 | ali1234 | doesn't pmem kmalloc huge regions of memory? like 8mb |
| 22:41.08 | Weiss | I'm trying to think of a nice way to fit the LCM updates into the DRM/KMS framework.. perhaps have a "notify update" ioctl that tells the kernel to update a region or the whole screen. I'd suspect it's best to keep that high-level, though, rather than having userspace talk to the kernel in terms of DMA transfers |
| 22:41.27 | tmzt | how would you avoid that? |
| 22:41.37 | Weiss | you know about "vmalloc"? |
| 22:41.39 | tmzt | mmap the control registers directly? |
| 22:41.50 | cr2 | Weiss: no, kmalloc |
| 22:42.20 | tmzt | the via driver uses it's own allocator and an array of pointers instead of a linked-list |
| 22:42.20 | ali1234 | maybe i'm confused and it just uses ioremap or something |
| 22:42.26 | tmzt | one of the things I like about it |
| 22:42.29 | ali1234 | don't listen to me :) |
| 22:42.36 | Weiss | tmzt: it's a good strategy to keep userspace away from registers as far as possible.. |
| 22:43.16 | cr2 | Weiss: the sound/video DSP needs about 8MB, and can't do scatter-gather. afaik |
| 22:43.41 | tmzt | I only see pmem.c allocing small data structures |
| 22:43.54 | tmzt | can't do sg? |
| 22:43.56 | tmzt | oh |
| 22:44.10 | tmzt | <PROTECTED> |
| 22:44.45 | tmzt | if the address space isn't used for ram by linux we can ioremap it |
| 22:45.12 | tmzt | I think drm lets us use hw addresses from userspace, like pmem does |
| 22:45.15 | tmzt | maybe not in drm2 |
| 22:45.17 | tmzt | dri2 |
| 22:48.17 | tmzt | cr2: this isn't the dsp, is that true of mdp as well? |
| 22:50.39 | *** join/#htc-linux dzo (n=dzo@mail.marginz.co.nz) |
| 22:59.03 | *** join/#htc-linux riddlebox (n=james@75-132-225-75.dhcp.stls.mo.charter.com) |
| 22:59.53 | riddlebox | anyone trying the latest kernel on a raph800? It just powers off for me after it starts to boot |
| 23:00.00 | tmzt | in android? |
| 23:00.05 | tmzt | yes it does that on raph500 also |
| 23:00.09 | tmzt | not sure why |
| 23:00.21 | tmzt | disabling the PM features (except CONFIG_PM) doesn't help |
| 23:01.02 | tmzt | so my assumption that it was related to power collapse seems to have been wrong |
| 23:01.03 | riddlebox | tmzt, yeah |
| 23:01.17 | tmzt | it could also be smd related but I don't think so |
| 23:02.03 | riddlebox | I havent had a kernel work good for a long time |
| 23:03.14 | riddlebox | I try the autobuild site everyday hoping things will work but no luck |
| 23:03.54 | tmzt | they won't work until they are fixed |
| 23:04.01 | tmzt | but we don't know what causes that |
| 23:04.15 | luke-jr | riddlebox: tmzt: bisect? |
| 23:04.18 | riddlebox | i wish there was some way to run a trace |
| 23:04.18 | tmzt | if you can get an adb logcat or ssh logcat |
| 23:04.35 | tmzt | luke-jr: yeah, but the recent patches won't work very well with bisect |
| 23:04.49 | riddlebox | luke-jr, bisect? |
| 23:05.27 | luke-jr | tmzt: yeah, Git needs a bisect-rebase function |
| 23:06.26 | riddlebox | it just seemed like all of a sudden the power stuff went south, I think it was working half way decent like a month ago |
| 23:13.14 | riddlebox | I am going to try the diamond version |
| 23:14.15 | tmzt | if you use mtype from raph800 it will probably do the same thing |
| 23:14.24 | tmzt | if you don't I'm not sure if it will work or how well |
| 23:14.29 | tmzt | as the diam version is for diam100 |
| 23:14.43 | *** join/#htc-linux dzo_ (n=dzo@mail.marginz.co.nz) |
| 23:15.13 | riddlebox | tmzt, so far it hasnt powered off, it is taking forever to boot |
| 23:15.26 | tmzt | ok |
| 23:15.28 | tmzt | wait |
| 23:16.39 | riddlebox | it booted but I get the power off screen right away |
| 23:16.52 | tmzt | power off? but it doesn't actually turn off? |
| 23:17.11 | tmzt | dzo: I don't understand this PM stuff |
| 23:18.25 | *** join/#htc-linux Midnigh2ker (n=AndroidI@unaffiliated/midnightoker) |
| 23:18.34 | tmzt | hey |
| 23:18.55 | Midnigh2ker | hey |
| 23:19.10 | Midnigh2ker | i might dissapier at any moment. playing taxi in the car |
| 23:19.20 | tmzt | ok |
| 23:19.37 | riddlebox | tmzt, it actually rebooted the phone |
| 23:19.49 | Midnigh2ker | im just curious what you were talking about in #android |
| 23:20.12 | tmzt | sure, he wanted to use the serial port on the G1 |
| 23:20.23 | Midnigh2ker | oh right |
| 23:20.26 | Midnigh2ker | ah time to go |
| 23:20.28 | tmzt | riddlebox: that's probably because you are using the diam100 mtype then |
| 23:21.32 | riddlebox | tmzt, how do I use the diam100 mtype? |
| 23:21.41 | tmzt | change default.txt |
| 23:21.44 | tmzt | I think you are though |
| 23:21.57 | riddlebox | i have 2039 |
| 23:22.33 | *** join/#htc-linux Orbital_sFear (n=ldupin@bbb72-0-189-133.bendbroadband.com) |
| 23:22.35 | tmzt | is that raph100? |
| 23:22.38 | tmzt | hello Orbital_sFear |
| 23:22.41 | Orbital_sFear | :) |
| 23:22.45 | Orbital_sFear | hey |
| 23:22.55 | riddlebox | tmzt, raph800 i think |
| 23:22.58 | tmzt | ok |
| 23:23.17 | tmzt | your're right |
| 23:23.35 | tmzt | s/ur/u/ |
| 23:26.31 | *** join/#htc-linux Renfield (n=Renfield@c-24-218-184-34.hsd1.nh.comcast.net) |
| 23:32.21 | Orbital_sFear | tmzt: I think I got it, I can write a small c program that writes messages to the kernel using syslog.h |
| 23:32.27 | Orbital_sFear | thanks for the help |
| 23:32.31 | tmzt | to the kernel? |
| 23:32.36 | Orbital_sFear | well |
| 23:32.38 | Orbital_sFear | no |
| 23:32.40 | tmzt | syslogd runs in userspace |
| 23:32.42 | Orbital_sFear | but tot he system logger |
| 23:35.18 | Orbital_sFear | damn, ya you're right, its not going to klog |
| 23:40.22 | *** join/#htc-linux dzo (n=dzo@121.98.128.127) |