00:18.49 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-7456a3802b9eb642) |
00:24.18 | druidu | just a teaser |
00:24.20 | druidu | http://img80.imageshack.us/img80/2013/pictureds4.jpg |
00:24.37 | druidu | the framebuffer address was wrong, gotta rebase it |
00:50.17 | druidu | ok, I have color output and animated, so it means no crashes :) |
00:56.01 | *** join/#htc-linux datachaos (n=datachao@201.22.222.107.adsl.gvt.net.br) |
01:02.01 | BruteFrce | what device? |
01:06.01 | druidu | htc diamond |
01:09.43 | BruteFrce | you seen the jade yet? |
01:09.51 | BruteFrce | makes me all tingly |
01:18.14 | cr2 | is it possible to create a windows .exe file from a perl script ? |
01:23.17 | BruteFrce | you can make a .bat that calls a perl script |
01:23.32 | cr2 | i want a standalone .exe |
01:24.46 | ali1234 | http://www.indigostar.com/perl2exe.htm |
01:24.53 | cr2 | http://www.indigostar.com/perl2exe.htm |
01:24.56 | cr2 | yes :) |
01:24.59 | BruteFrce | you can use openpsx |
01:25.03 | BruteFrce | p2x |
01:25.50 | BruteFrce | openp2x its sole purpose it to make a perl script into an .exe |
01:27.19 | cr2 | This project has not yet created any file release packages. |
01:32.48 | BruteFrce | which one? |
01:33.51 | cr2 | openp2x |
01:42.17 | BruteFrce | seems like .par is the best method |
01:44.20 | BruteFrce | he already found that |
01:44.28 | BruteFrce | ooops |
01:49.36 | *** join/#htc-linux bash (n=spam@cpe-098-026-098-026.nc.res.rr.com) |
01:50.11 | BruteFrce | cr2 whats the script for? |
03:04.36 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-e37472d03e0831b5) |
03:10.33 | *** join/#htc-linux ALoGeNo (n=QUAKEIII@165.Red-88-27-115.staticIP.rima-tde.net) |
03:10.49 | ALoGeNo | lo all |
03:13.15 | *** join/#htc-linux EA2 (n=chipper@cpe-024-074-138-191.carolina.res.rr.com) |
03:57.57 | *** join/#htc-linux PoohbaLT (n=Poohba@c-98-235-52-97.hsd1.nj.comcast.net) |
04:12.42 | par | hello alogeno |
04:14.24 | ALoGeNo | hey par :) |
04:14.42 | ALoGeNo | how is your android? |
04:14.43 | par | how is your android running on your universal? |
04:14.46 | ALoGeNo | xD |
04:14.49 | par | haha same thing |
04:14.50 | ALoGeNo | lol |
04:14.53 | par | ;) |
04:14.58 | ALoGeNo | hehe |
04:15.02 | ALoGeNo | not now :) |
04:15.06 | par | i need to dig up more SD cards |
04:15.17 | par | i have them laying around my workbench area |
04:15.20 | ALoGeNo | oh you have a fault with that? |
04:15.32 | par | no no.. just lack of empty SD cards ;) |
04:15.40 | ALoGeNo | hehe okay |
04:15.51 | ALoGeNo | im using the old ones :) |
04:16.03 | ALoGeNo | 64MB and 128MB for that :) |
04:16.07 | par | haha |
04:16.15 | par | does your ts work yet? |
04:16.35 | ALoGeNo | and i have 256MB one that is RS-mmc and not work in nothing linux :P |
04:16.43 | par | or you aren't working on any kernel patches to bring up the ts |
04:17.02 | ALoGeNo | jesus christ! |
04:17.02 | ALoGeNo | xd |
04:17.18 | ALoGeNo | nopes im not developer or coder or that im only tester |
04:17.19 | ALoGeNo | XD |
04:17.24 | par | same here |
04:17.33 | ALoGeNo | :) nice |
04:17.39 | ALoGeNo | D: |
04:18.18 | par | on your universal you can navigate okay with the phone's keys right? |
04:18.36 | ALoGeNo | yes |
04:18.55 | ALoGeNo | the screen is dancing a bit like a rummble little rumble |
04:19.09 | ALoGeNo | but not allways.. |
04:19.30 | par | oh, can you functionally use android applications at all? |
04:19.40 | ALoGeNo | i was tested the new n8xx images of android but no results |
04:20.03 | druidu | almost finished the boot console for htc diamond :P |
04:20.14 | ALoGeNo | yes the old one you can open contacts and stuff but is bored becose i dont know the keys |
04:20.30 | ALoGeNo | congrats druidu |
04:20.32 | par | ah i see |
04:20.37 | druidu | thanks :) |
04:20.38 | par | yeah druidu, thats great |
04:21.56 | ALoGeNo | parim not much versated in mix kernels and stuff becose you need to know to patch many things i guess |
04:22.26 | par | yes, i have a good memory but i can't remember all the places |
04:22.40 | ALoGeNo | you remember the procces to run android no? |
04:22.42 | par | i haven't even looked at the project for a couple of years really |
04:23.02 | par | process? |
04:23.14 | ALoGeNo | is only copy the /bin in root directory, and later put the init in /sbin |
04:23.24 | par | well there was a kernel patch someone had out.. but i have the binaries like you |
04:23.28 | ALoGeNo | and the /bin dir is the busybox |
04:23.36 | par | oh, yup |
04:25.10 | ALoGeNo | i was do that proces with the new n8xx android images but nothin work.. (well that run and boot, but when after we see the red ball jumping , now is the logo of android becose i see him in the emulator, but in the uni we have blackscreen) |
04:26.04 | ALoGeNo | if you want i can paste the link of the images and you can try |
04:26.04 | ALoGeNo | xd |
04:26.24 | par | but these are for n8xx? |
04:26.27 | ALoGeNo | oh but you dont have empty cards.. np |
04:26.28 | ALoGeNo | :) |
04:26.33 | ALoGeNo | yep.. |
04:26.36 | par | yeah, no empty cards :) |
04:26.39 | ALoGeNo | you dont read the tuto? |
04:26.57 | par | which tut? |
04:27.32 | ALoGeNo | What i did: |
04:27.32 | ALoGeNo | - took the Kernel from handhelds.org (2.6.21-hh20) with |
04:27.32 | ALoGeNo | htcblueangel_defconfig |
04:27.32 | ALoGeNo | - patched it with http://android-on-n8xx.googlecode.com/files/linux-2.6.21_rx-34_androi... |
04:27.32 | ALoGeNo | - i didn't find a plain patch so i applied this one. Benno's |
04:27.32 | ALoGeNo | android.diff is for 2.6.23 and causes some trouble. |
04:27.34 | ALoGeNo | - activated Android specific settings and compiled it |
04:27.36 | ALoGeNo | - created the userland-img like in http://www.alextreme.org/drupal/?q=android_on_n810 |
04:27.38 | ALoGeNo | - copied the compiled busybox |
04:27.40 | ALoGeNo | - used haret to boot the kernel off the sd-card |
04:27.48 | par | yes, read that ;) |
04:28.36 | ALoGeNo | Altogether i've put up two tbz2: |
04:28.36 | ALoGeNo | - http://evil-packet.org/android/vfat-bootloader-kernel.tbz2 |
04:28.36 | ALoGeNo | Just copy this to the vfat partition (first partition) on your SD- |
04:28.36 | ALoGeNo | Card. |
04:28.36 | ALoGeNo | Android will autoboot as soon you plug the card into your BA. |
04:28.36 | ALoGeNo | - http://evil-packet.org/android/ext2-android-htc-blueangel-arm.tbz2 |
04:28.38 | ALoGeNo | Copy it to the ext2 partition (second partition) on your SD-Card. |
04:28.40 | ALoGeNo | I moved /init (android loader) to /sbin/android. |
04:29.02 | par | yes yes, i read it |
04:29.03 | ALoGeNo | is easy but i dunno if dont work for me for the universal or what.. |
04:29.39 | ALoGeNo | if i use that files works.. but if you get the new ones n8xx images of android and do that proces dont work |
04:29.44 | par | were those the images you are currently running on your uni? |
04:30.07 | par | oh i see |
04:30.10 | ALoGeNo | the ext2 one, is that i using and run.. |
04:30.19 | ALoGeNo | but letme search the new ones |
04:30.31 | par | well what are the versions of the new images for example |
04:32.57 | par | what exactly are the versions of the images.. and what patches are you trying to apply |
04:33.38 | par | what kernel did you compile for it etc. |
04:34.25 | ALoGeNo | http://androidboards.com/applications/ |
04:34.45 | ALoGeNo | im using the same kernel that im using in my debian |
04:35.13 | ALoGeNo | sumousou kernel and modules |
04:35.32 | ALoGeNo | sumoudou* |
04:36.15 | ALoGeNo | in this way work for me with the BA android image |
04:36.37 | ALoGeNo | but only the old one no the new one.. |
04:37.12 | ALoGeNo | the new one boot and say you are a gentellman or whatever.. |
04:38.26 | ALoGeNo | but instead of the red ball we get blackscreen and hang.. |
04:39.57 | par | maybe try with the trusty angstrom kernel at handhelds? |
04:40.15 | ALoGeNo | is the same kernel |
04:40.34 | ALoGeNo | 21hh20 |
04:41.27 | ALoGeNo | i was try all the f#cking krnels :P |
04:41.31 | par | yeah but patched with http://android-on-n8xx.googlecode. |
04:41.32 | par | <PROTECTED> |
04:41.40 | ALoGeNo | nopes |
04:41.55 | ALoGeNo | but the old one boots |
04:42.01 | ALoGeNo | without patch |
04:42.13 | par | oh, so you never pathced the old one |
04:42.18 | ALoGeNo | nope |
04:42.47 | ALoGeNo | and i dont have patch for uni |
04:42.52 | ALoGeNo | is BA patch |
04:43.31 | par | ah ok so you never modded the patch and not even used the patch at all |
04:43.40 | ALoGeNo | one more thing.. google.code n8xx was dissaper.. |
04:43.41 | ALoGeNo | Xd |
04:43.51 | ALoGeNo | dispear* |
04:43.51 | par | oh hahaha |
04:43.59 | ALoGeNo | or wahtever..xd |
04:44.00 | par | i dunno what to say then m8 |
04:44.05 | par | ;) |
04:44.08 | ALoGeNo | google is scared of nokia i think |
04:44.09 | ALoGeNo | xd |
04:44.21 | par | haha |
04:45.50 | par | i honestly thought those images were vogue images at first |
04:45.58 | par | because someone appeared to have it working with vogue images |
04:46.00 | *** join/#htc-linux goxboxlive (n=goxboxli@195.159.97.196) |
04:46.02 | par | (on the ba) |
04:46.22 | *** join/#htc-linux cr2_ (n=konversa@crpl6.physik.uni-wuppertal.de) |
04:46.49 | par | but when i saw they were nokia i got more interested |
04:47.45 | ALoGeNo | hell yeah, but i think that is all disapear vouge and n8xx in google.code fuck stuff.. |
04:50.30 | par | the n800 has a PXA320 |
04:50.40 | par | yeah |
04:51.07 | par | n800 pxa320 vs ba pxa263 |
04:51.22 | par | close enuff tho i guess |
04:56.51 | ALoGeNo | ooh.. i see |
04:57.27 | goxboxlive | hi paulproteus: Why havent you updated the git three with Qtopia 4.4? |
04:57.44 | ALoGeNo | hi goxboxlive |
04:58.01 | *** join/#htc-linux BruteFrce (n=IrcNet@m210e36d0.tmodns.net) |
04:58.51 | BruteFrce | does anyone know a way to startx with an option to start something like xterm when it starts? |
05:00.04 | par | ALoGeNo: i'll try with the new one on the ba when i get an empty card set up |
05:02.38 | par | ALoGeNo: but htc universal is pxa270 anyway :P |
05:03.20 | goxboxlive | hi ALoGeNo |
05:03.42 | par | ALoGeNo: if you can't make it work on the universal its 99% certain it won't work the way it is on the ba also. |
05:08.04 | ALoGeNo | roger par |
05:11.45 | BruteFrce | guess no one knows a way then? |
05:12.14 | ALoGeNo | i nope :P |
05:12.50 | BruteFrce | what a dumb lil thing to rebuild the whole userspace for |
05:13.02 | BruteFrce | one line in the xinirc |
05:13.11 | BruteFrce | xinitrc |
05:13.51 | ALoGeNo | BruteFrce serach the file that run the X sevrer and add & xterm |
05:14.29 | BruteFrce | explain that a little more please? |
05:15.14 | BruteFrce | i have tried startx && xterm no go |
05:15.27 | BruteFrce | if thats what ya mean |
05:15.47 | BruteFrce | alogeno |
05:16.12 | swetland | yawns |
05:18.13 | ALoGeNo | ley & only instead of && |
05:18.20 | ALoGeNo | let* |
05:18.28 | BruteFrce | let? |
05:18.39 | ALoGeNo | put* |
05:18.40 | ALoGeNo | xd |
05:18.55 | BruteFrce | what difference is there between 1 or 2? |
05:19.15 | ALoGeNo | i dunno but the other day it works for me |
05:19.53 | BruteFrce | with startx? |
05:20.03 | ALoGeNo | nope in one script |
05:20.10 | ALoGeNo | to run another thing |
05:20.26 | BruteFrce | because startx takes like 40 seconds to start so i think thats the problem |
05:20.41 | ALoGeNo | oh |
05:21.39 | ALoGeNo | i think that will be easy becose if you run safemode it run X and xterm ;) |
05:23.09 | BruteFrce | huh? |
05:23.38 | BruteFrce | im gunna try sleep 30 xterm |
05:23.58 | BruteFrce | then startx |
05:31.37 | ALoGeNo | hey i think that you need set the geometry |
05:32.04 | ALoGeNo | Below is a sample .xinitrc that starts several applications and leaves the window manager running as the ''last'' application. Assuming that the window manager has been configured properly, the user then chooses the ''Exit'' menu item to shut down X. |
05:32.05 | ALoGeNo | xrdb -load $HOME/.Xresources |
05:32.05 | ALoGeNo | xsetroot -solid gray & |
05:32.05 | ALoGeNo | xbiff -geometry -430+5 & |
05:32.05 | ALoGeNo | oclock -geometry 75x75-0-0 & |
05:32.07 | ALoGeNo | xload -geometry -80-0 & |
05:32.09 | ALoGeNo | xterm -geometry +0+60 -ls & |
05:32.11 | ALoGeNo | xterm -geometry +0-100 & |
05:32.13 | ALoGeNo | xconsole -geometry -0+0 -fn 5x7 & |
05:32.15 | ALoGeNo | exec twm |
05:32.37 | ALoGeNo | no? |
05:33.55 | ALoGeNo | http://www.xfree86.org/current/xinit.1.html |
05:34.52 | *** join/#htc-linux Magorium (i=Othello@gateway/tor/x-8172d10056cebb24) |
05:41.38 | *** join/#htc-linux ALoGeNoff (n=QUAKEIII@165.Red-88-27-115.staticIP.rima-tde.net) |
05:42.57 | *** join/#htc-linux datachaos (n=datachao@201.22.222.107.adsl.gvt.net.br) |
05:44.06 | tmzt | ALoGeNoff: if you want to start something in the background |
05:44.50 | tmzt | ALoGeNoff: (sleep 30; xterm&) |
05:44.50 | ALoGeNoff | oh nice is BruteFrce :) |
05:44.51 | ALoGeNoff | i try to help him :) |
05:45.04 | ALoGeNoff | lo tmzt :D |
05:46.10 | tmzt | BruteFrce: |
05:46.34 | tmzt | ALoGeNoff: missed that, just got back |
05:47.25 | ALoGeNoff | what you mean with that? |
05:48.25 | tmzt | ALoGeNoff: you guys get android on pxa? |
05:49.14 | ALoGeNoff | the old one yes, 2007 one, the new one not work for me |
05:53.21 | NetRipper | druidu, very nice work! :) |
05:59.43 | *** join/#htc-linux alo (n=QUAKEIII@165.Red-88-27-115.staticIP.rima-tde.net) |
06:01.18 | *** join/#htc-linux kiozen (n=oeichler@p5492A1DE.dip0.t-ipconnect.de) |
06:01.24 | par | tmzt: hi |
06:02.48 | tmzt | hi |
06:03.07 | tmzt | par: you trying to get android on uni? |
06:03.22 | tmzt | par: the new stuff shouldn't need anything special for ts at least |
06:03.30 | tmzt | par: don't konw about framebuffer |
06:04.44 | par | not on the uni.. i'd love to have it on the ba |
06:08.23 | par | very similar though.. |
06:16.03 | tmzt | haven't looked at the build system |
06:19.46 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-77-157-204.pppoe.mtu-net.ru) |
06:30.10 | tmzt | par: I'm a bit confused. You are trying same kenrel on ba as uni? |
06:49.08 | *** join/#htc-linux marex (n=marex@vasut.kolej.mff.cuni.cz) |
06:50.15 | *** join/#htc-linux rmoravcik (n=rmoravci@gtsgw.ttc.cz) |
06:53.28 | *** join/#htc-linux ginge (n=ginge@host81-149-155-58.in-addr.btopenworld.com) |
06:58.47 | *** join/#htc-linux timebomb (n=timebomb@p4FDE5885.dip.t-dialin.net) |
07:00.10 | par | tmzt: i'm not trying anything yet |
07:00.52 | par | ALoGeNo was using a ba kernel with the 2007 android |
07:01.02 | par | and apparently had it working on his uni |
07:02.46 | ALoGeNo | par ruun?? |
07:02.50 | ALoGeNo | how? |
07:03.00 | ALoGeNo | oh |
07:03.00 | ALoGeNo | xD |
07:03.07 | ALoGeNo | 2007 yes.. |
07:05.11 | ALoGeNo | par i was using universal kernel with BA tarball of android.. |
07:05.22 | ALoGeNo | not ba kernel |
07:08.15 | tmzt | ALoGeNo: hey, ok you have android for pxa from n800 and sumoudou kernel? |
07:08.36 | ALoGeNo | yes that is |
07:09.08 | ALoGeNo | but is the old android the new one using the same way dont work forme :P |
07:09.31 | ALoGeNo | you know wnat i mena? |
07:09.35 | ALoGeNo | wnat* |
07:09.39 | ALoGeNo | want* |
07:09.45 | ALoGeNo | mean* |
07:09.49 | ALoGeNo | X) |
07:10.01 | tmzt | touchscreen doesn't work? |
07:10.20 | ALoGeNo | nope, you must use the keyb |
07:10.52 | tmzt | I'm not sure its worth making old android work, new one can be compiled for pxa now |
07:11.00 | tmzt | it can be done though |
07:12.10 | ALoGeNo | i dont know how do that :( |
07:12.42 | tmzt | if you have the disk space, and I think python2.4, you can build according to google instructions |
07:12.43 | par | N800 is PXA320 |
07:12.58 | par | Universal is PXA270 and BA is PXA263 |
07:13.24 | tmzt | we have the source code for android now, we can build for basically anything (arm) |
07:13.58 | swetland | ARMv5t |
07:14.04 | ALoGeNo | buid appz but the whole os ? how? |
07:14.09 | tmzt | and we can make it work with native touchscreen coords, and see what it's doing with the framebuffer now |
07:14.12 | swetland | v4t will take a bit of work |
07:14.24 | tmzt | ok, pxa255/270 are armv5 |
07:14.38 | tmzt | only sa1100 and below (jornada) are armv4t |
07:14.54 | tmzt | and apparently s3cxxxx, one of those is |
07:14.59 | swetland | and the arm9 core in the neo freerunner |
07:15.18 | par | disk space? haha, isn't it like 2.2 gigs? |
07:15.19 | tmzt | would qemu-arm on device be possible, or way to slow? |
07:15.42 | swetland | ~1GB to download, 1GB to checkout, 2GB to build |
07:15.57 | swetland | tmzt: it would likely be pretty painful |
07:16.00 | ALoGeNo | wow |
07:16.03 | tmzt | and davlik?? is not using jazeel?? |
07:16.26 | swetland | the jazelle extensions are proprietary, require a sun scsl license for docs, and are outmoded |
07:16.35 | par | qemu..? what is it emulating to run android |
07:16.45 | swetland | arm is going to a more general model for supporting vms with the thumb2ee stuff |
07:16.54 | swetland | which is much nicer looking |
07:17.33 | tmzt | is there still qvga support, like old sdk emu? |
07:17.43 | tmzt | landscape |
07:17.58 | ALoGeNo | par i think that you do all the test with qemu or the emulator of anndroid (if they means that) |
07:18.04 | swetland | some of the newer UI is going to be a little squished in qvga |
07:18.16 | ALoGeNo | yes you have HVGA and QVGA tmzt |
07:18.52 | tmzt | I have hvga 640x240, I guess you mean, ALoGeNo |
07:18.53 | par | ALoGeNo ok, but i'm just curious wha they are emulating natively for Android anyway? |
07:19.20 | tmzt | but I mean Q (qvga landscape like old white skin was) |
07:19.51 | ALoGeNo | yes i think tmzt, i was try the emu yesterday, and go toooo slow forme in amd 3400+ and or my celeron 500Mhz |
07:19.51 | ALoGeNo | xd |
07:19.54 | tmzt | par: I mean running the qemu-arm emulator to make armv5 binaries work on armv4, not practical |
07:20.09 | par | tmzt: yep i agree with that |
07:20.52 | tmzt | the old ui looked better than wm6 on those type of devices, I was going to try it |
07:21.18 | tmzt | apps I am worried about would be maps/browser/phone |
07:22.13 | swetland | you can try arbitrary display sizes in the emulator |
07:22.18 | *** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
07:22.18 | swetland | with -skin 240x320 |
07:22.20 | swetland | or whatnot |
07:22.54 | tmzt | the old code was release as demo, or I remember wrong? |
07:23.16 | tmzt | home screen or whatever it is called |
07:24.56 | swetland | I think so, yeah |
07:26.47 | *** join/#htc-linux dante_ (n=dante@host176-48-static.15-79-b.business.telecomitalia.it) |
07:28.42 | ALoGeNo | hopes from something new to test becose is bored to try new android without results :P |
07:29.39 | tmzt | ALoGeNo: what kind of not results? (makes no sense in English either, sorry) |
07:29.58 | par | it makes most sense to get it built for ba, universal, (maybe even magician) those are closest to complete with htc-linux |
07:30.33 | ALoGeNo | whe i get the you are a gendelman have a blackscreen of the dead |
07:30.36 | tmzt | anyone can take magician.c and begin converting it for their device, with kernel26 source as reference |
07:30.56 | ALoGeNo | when* |
07:31.10 | par | tmzt: sure, thats great blanket statement :) |
07:31.49 | tmzt | the drivers are the hard part and they are mostly written. I don't know about egpio but cr2 has upstreamed driver ?? |
07:32.31 | par | you mean kevin o'connors htc-egpio? |
07:32.43 | par | it works with 5 devices so far i think? |
07:32.46 | tmzt | yeah, cr2 mentioned it and I phrased that wrong |
07:34.18 | par | i guess what i expect to see eventually is binaries floating around or in autobuild (under htc-pxa) for android ba, android-universal... etc. |
07:35.23 | tmzt | nothing should be android specific with android compiled from source, except changes to framebuffer which should not interfere with X (though they currently do) |
07:36.18 | tmzt | maybe there is something new, I haven't studied the source code yet. the is the hardware stuff at least to abstract it |
07:42.55 | tmzt | libhardware, for vibrator and leds and usb mass storage |
07:44.04 | *** join/#htc-linux lpotter (n=ljp@CPE-124-191-144-181.vic.bigpond.net.au) |
07:47.04 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
07:50.49 | par | how is bluetooth HID support |
07:51.11 | tmzt | on? |
07:51.26 | par | in general on htc-pxa for example |
07:51.58 | tmzt | depends on bluetooth/hci_uart/bluez version |
07:52.58 | tmzt | if you mean keyboard/mouse, you would pass the input device to X |
07:54.15 | par | 21-hh20-r16 |
07:54.58 | par | yeah, say a bluetooth 2.0 keyboard |
07:55.11 | par | does 21-hh20-r16 support that nowadays? |
07:55.19 | dcordes | are there bluetooth 2.0 keyboards? |
07:55.25 | tmzt | I guess I mean it's not specific to pxa or anything else, just how good the driver is and the other components |
07:55.42 | par | heh no i'm talking about kernel versions |
07:55.46 | tmzt | why would there be? maybe the new 2.1 pairing could be helpful |
07:56.13 | dcordes | yea that's why I ask. old bluetooth should sufice fore most input devices |
07:56.14 | par | just, i don't know what the codebase nowadays even supports :) |
07:56.28 | par | ok... |
07:56.42 | tmzt | that's why we need to get away from hh and to upstream 2.6.27, 2.6.26 at least |
07:56.46 | par | interestingly i have an old irda keyboard which i use under wince |
07:57.06 | tmzt | probably raw serial data, I have one too I haven't tried |
07:57.10 | par | cool |
07:57.12 | tmzt | as driver doesn't work on smartphone |
07:57.18 | par | yeah |
07:57.29 | par | probably a win2k3 driver right? |
07:57.49 | tmzt | and the weird circuit board that allows the transmitter to spin is all apart |
07:58.01 | tmzt | will probably put led on some wires now |
07:58.21 | tmzt | ppc only, I think |
07:58.44 | par | i meant win mobile |
07:59.03 | tmzt | actually thought it was just needed touchscreen, but wm tools?? (fake mouse) doesn't work either |
07:59.20 | tmzt | also tried adding to reg manually |
07:59.38 | tmzt | then will just get bt key |
07:59.40 | par | well, no biggie i don't have a prob with getting a usb keyboard |
07:59.57 | tmzt | for windows mobile? or linux |
07:59.58 | tmzt | ? |
08:00.06 | par | both |
08:00.21 | par | err oops meant bluetooth ;) |
08:00.33 | tmzt | linux should just work if protocol is simple enough and ficp works |
08:00.45 | tmzt | there was input method for gpe at least |
08:01.04 | par | so you are saying you want to get away from hh? |
08:01.09 | tmzt | could make userspace uart to uevent driver in simple .c |
08:02.00 | tmzt | I think it is neccessary to bring devices to supported status and gain from improvements |
08:02.56 | par | i thought the current work on the codebase was bringing it up to 2.6.26? |
08:02.57 | tmzt | msm is all on newer kernels (2.6.25 is oldest I think), omap850 is going there, jlime (jornadas and nec mobilepros) is |
08:03.53 | tmzt | it's even possible to build all mach variants in one zImage now, I think |
08:03.57 | par | i guess i would have to ask cr2.. but last time i peeked in here it was |
08:04.03 | dcordes | tmzt, oldest msm is 2.6.24 I think, but at 2.6.26 now |
08:04.05 | tmzt | we can have one pxa zImage |
08:04.28 | tmzt | dcordes: that could be right, wasn't sure |
08:04.40 | dcordes | where's that possible? in newer mainline kernels? |
08:04.59 | tmzt | to make testing easier, mostly a Kconfig change |
08:05.23 | tmzt | yeah, look on arm-kernel don't know how far back it's supported |
08:05.44 | par | yeah, has support been drummed up into mainline kernels these days? |
08:06.14 | par | what are these kernels that goog is using for android.. are they in mainline? |
08:06.40 | swetland | we're hoping to have android-2.6.27 and msm-2.6.27 in reasonable shape in a week or two |
08:06.41 | par | thats actually why i thought we were saved when i saw san here ;) |
08:06.57 | par | i was like oh joyous |
08:06.59 | tmzt | google pushed some early halibut stuff and basic msm support, and is doing some stuff for rmk now |
08:08.13 | tmzt | still not sure how to make the amss dependents stuff abstract though, really don't want to have to rip out stuff in upstream as general rule to patch for wince-based htc's |
08:08.37 | tmzt | which don't have rmnet and watchdog for internet, unfortunately |
08:09.51 | par | swetland: so just msm? i thought it was msm and something else |
08:11.14 | swetland | tmzt: we might have to just do a couple different versions of the smd/rpc/etc stuff |
08:11.30 | swetland | and make 'em a configure option based on amss version |
08:11.50 | tmzt | does xx25 have a table for the smd channels now? |
08:11.50 | swetland | from what I've seen of the patches for the wince amss builds things are pretty different in a number of ways |
08:12.08 | tmzt | that we could emulate in platform_data |
08:12.41 | swetland | well the actual structure of the channel data structures is different too, isn't it? or am I misremembering? |
08:13.07 | dcordes | swetland, that would be useful with the different amss configurations |
08:13.15 | tmzt | I think they just don't have a real structure and depend on numbers, not sure |
08:13.37 | swetland | we already use an amss version config to choose between dsp and rpc protocol versions |
08:14.24 | swetland | I'm working with the qualcomm folks to try to get stuff a bit more stable in the future, but of course that doesn't help with all the stuff you guys are doing to support older devices |
08:14.36 | dcordes | isn't smd channel setup also a hardware thing? |
08:14.43 | swetland | as you're stuck with whatever the older amss-for-wince builds do |
08:14.47 | swetland | no |
08:15.24 | tmzt | it's like saying a protocol over ethernet is hardware because win311 doesn't support it but the nic and the pc its running on do |
08:15.28 | swetland | it's just how amss uses shared memory to talk to the apps processor, which is purely a software thing |
08:15.30 | dcordes | so you could virtually select whatever memory areas by changing the amss? |
08:15.49 | swetland | there's not really special about the memory |
08:15.50 | tmzt | it just wonder if the version numbers overlap |
08:15.55 | swetland | it's just ram that both processors read/write |
08:16.15 | swetland | there are some data structures that they agree on and thus can pass commands/data/etc back and forth |
08:16.46 | tmzt | seeing how similar vogue and dia/raph are in proc_comm as I saw on here, makes sense |
08:17.21 | swetland | what you're probably seeing is the protocol evolving slowly as qualcomm changes things from version to version of amss |
08:18.00 | dcordes | swetland, basically I'm referring to the differences between vogue and kaiser and halibut smd setups which are the only ones I know |
08:18.02 | tmzt | exactly, but people where expecting more differences between non-a and a versions of the chip |
08:18.07 | swetland | a bunch of stuff changed during the 6.2.x amss development (the version g1 uses), some things due to qct making changes/adding features they were doing internally, some due to small changes we reqyuested, etc |
08:18.25 | swetland | the msm7xxx family are more similar than different |
08:18.28 | tmzt | in the way the two cores communicate when it is the software protocol on the other core that matters |
08:18.45 | swetland | and again, smd is purely a software issue. has nothing to do with the hardware differences |
08:18.50 | swetland | yup |
08:19.01 | tmzt | dcordes: halibut is just hardware, but it is running a specific amss (microkernel on arm9) |
08:19.28 | tmzt | dcordes: halibut used by the htc wince teams would be running their version and more like vogue |
08:20.00 | swetland | yeah. halibut is a eval/dev platform (big honkin' board). qct calls it "surf" |
08:20.39 | tmzt | the kind of cell phone you have to connect antennas to .. |
08:20.42 | swetland | and they develop amss on it -- the version that's supporting linux, the version that's supporting wince, etc |
08:21.00 | tmzt | not really carry in your pocket |
08:21.03 | dcordes | tmzt, that's all clear. I just didn't understand it is 100% software |
08:21.14 | swetland | http://cache.gizmodo.com/assets/resources/2008/02/big-qualcomm-prototype.jpg |
08:21.23 | swetland | that's a MSM7201A SURF |
08:21.24 | tmzt | it's like saying a tiny device based on omap3 versus your beagleboard |
08:22.12 | swetland | TI has similar huge ref boards for omap (H2 for 2420/30, etc) |
08:25.14 | tmzt | I guess we could hope that htc sees the beauty in ethernet packets instead of ppp chanels, the design in g1 looks really clean |
08:25.25 | dcordes | so if it is only software, why doesn't qualcomm use a more consistent setup of the smd? like on the msm7x00 devices one could use the same addresses for the at channels |
08:25.49 | swetland | yeah. qualcomm does some stuff that I think is strange, but I *love* the no-overhead rmnet/qmi stuff |
08:26.04 | swetland | having to hdlc encode, mux, and shovel ip data back and forth over ppp is barbaric |
08:26.07 | dcordes | tmzt, no data channels for the g1 used? how does rmnet operate? |
08:26.14 | par | hhahaa |
08:27.05 | tmzt | so your wakeup from network works better that way? |
08:27.13 | tmzt | I mean "push" |
08:27.21 | swetland | just less overhead |
08:27.29 | swetland | the linux network stack deals with ip packets |
08:27.35 | swetland | the radio network stack is packet based |
08:27.40 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
08:27.53 | swetland | having to encode/escape 'em as hdlc, stream to the other side, extract the ip data back out... just overhead |
08:28.31 | *** part/#htc-linux Kalemas (i=ejabberd@217.147.29.117) |
08:28.38 | tmzt | I got that, I would think that getting data from the network when only arm9 needs to be active and then notifying linux and arm11 would be better than polling |
08:29.14 | tmzt | since amss has to periodically wake up anyway |
08:29.18 | swetland | there's really no difference on that side |
08:29.23 | tmzt | ok |
08:29.38 | swetland | the network stack behaves the same. it's just how the packets are passed over to the apps core when there are packets for it |
08:30.49 | swetland | http://frotz.net/misc/powermonitor.png <- radio waking up to handle paging |
08:31.03 | *** join/#htc-linux NetRipper (n=netrippe@netripper.nl) |
08:33.00 | tmzt | I guess I want to keep arm11 powered down (sleep) until I get incoming mail or xmpp message or "tickle" and wake up, as well as for phone events. That would require more in the arm9 side though than android api appears to provide |
08:33.20 | par | wow, so yep.. much appreciated.. even if just msm only it still enlivens the effort as well as provides insight into what the whole project has been here |
08:33.38 | tmzt | par: what do you mean by msm only? |
08:33.45 | swetland | tmzt: um, that's exactly what we do |
08:33.53 | tmzt | cool |
08:34.01 | par | the goog going opensource |
08:34.03 | swetland | we enter power collapse (full power down of the a11) whenever possible in idle, and sit in that state in suspend |
08:34.47 | swetland | packet comes in, a9 wakes up, handles it, wakes up the a11, we do something then go right back to sleep (exact mechanics a bit more complex than that, but that's the basic state) |
08:35.25 | swetland | if you look at a trace of power consumption at the battery on g1, you'd see it basically flatline at 1.2mA and every ~1s (depending on network) spiking to ~140mA across a ~20ms window (network paging) |
08:35.46 | tmzt | so the a9 does see the ip packet, knows what port, etc. wasn't sure from your description. |
08:36.04 | swetland | a9 only knows an ip packet arrives. hands it to the a11 to process |
08:36.19 | *** join/#htc-linux lama (i=lama@gpi117.internetdsl.tpnet.pl) |
08:36.57 | *** part/#htc-linux kaze_ (n=kaze@ABordeaux-152-1-42-72.w82-125.abo.wanadoo.fr) |
08:36.57 | *** join/#htc-linux kaze_ (n=kaze@ABordeaux-152-1-42-72.w82-125.abo.wanadoo.fr) |
08:39.19 | tmzt | par: google is using msm in the g1 and has made source code available for kernel to work on that device. I just don't see what you mean by msm-only. before the android (user) code was released it was difficult to make it work on other devices and (I at least) was concentrated on the kernel side of that process. now we can write a libhardware for whatever device and see how android interacts with kernel |
08:39.59 | tmzt | so we can support whatever armv5t (pxa and newer) device with modifications if needed |
08:40.16 | par | tmzt: yes i agree thats why it much appreciated :) |
08:40.28 | swetland | hell, armv4t shouldn't be more than a couple days of work |
08:40.45 | par | like i said when i saw san was here i knew we were saved.. |
08:40.47 | swetland | get the compiler options tweaked, and hunt down where we have assembly that assumes v5t specifics |
08:41.01 | par | thank god its what they ended up going with. |
08:41.10 | swetland | ? |
08:41.20 | tmzt | I am saying now, short-term, we need to get ba/universal/magician on a newer kernel so that they can be brought forward (or even maintained upstream) to each new kernel released |
08:41.54 | swetland | getting stuff into upstream is an adventure ^^ |
08:42.05 | tmzt | then debian/angstrom/gizard/ and android can all be supported |
08:42.07 | swetland | but we're hoping to get there eventually |
08:42.16 | par | i've known san for a while.. since the 90's just as an industry acquaintanc |
08:44.06 | par | anyway it means no more limping so blindly forward after all the years |
08:44.29 | tmzt | the big difference is while before each machine needed its own way of doing gpio, gpio extenders (multi-function devices now), and whatever else, there is now gpio_ api, gpio devices, otg_transceiver for usb, usb gadgets, and all these other things that *should* work on all the arm variants |
08:44.53 | par | right ;) |
08:45.09 | swetland | really I think the biggest thing we need to do to ease porting is to continue into a (much needed) cleanup and overhaul of libhardware |
08:45.17 | par | versus.. hi this is a complete shot inthe dark hack we made up |
08:45.25 | swetland | to make platform specific support more pluggable and interfaces more well-defined |
08:46.06 | dcordes | swetland, what's libhardware? |
08:46.37 | swetland | it's intended to be the glue layer between the bulk of the android platform and architecture/platform specific stuff |
08:47.03 | swetland | for things that need to plug in at a higher level than the user/kernel boundary |
08:47.24 | tmzt | code running in davlik uses jni-like to talk to it? |
08:47.50 | swetland | usually uses jni to talk to lower level libraries |
08:48.10 | swetland | which then at the bottom use libhardware as an abstraction over architecture-specific driver interfaces |
08:48.24 | swetland | for say gps, sensors, 2d accel, etc etc |
08:48.34 | tmzt | the "int color" doesn't look like much of an abstraction, there is something above it? |
08:48.38 | dcordes | why don't you interface more directly? |
08:49.14 | swetland | because interfaces vary |
08:49.23 | swetland | on some hardware the camera may be a simple kernel driver |
08:49.41 | swetland | on msm7201a (right now), it's a proprietary library that talks to the dsp and video frontend |
08:50.00 | *** join/#htc-linux LanceHaig (i=c2d8ed19@foresight/member/lhaig) |
08:50.16 | swetland | on msm7201a, we use a lightweight raw audio interface. on some other platform somebody might want to use alsa to talk to the audio hardware |
08:50.23 | tmzt | and someday v4l might support compressed video -) |
08:50.29 | swetland | exactly |
08:50.31 | tmzt | in userspace |
08:50.59 | swetland | some things are kinda no-brainers -- we use /dev/input/* for general input event processing, because it's simple and broadly supported |
08:51.38 | tmzt | like, dcordes, you may get beagle to decode mp3 in hardware if it doesn't already (not hardware but dsp I mean) |
08:51.41 | swetland | other things are less clear-cut, and libhardware is intended (though the first cut is rough right now) to provide a bit of abstraction between the generic android platform and the specific hw it runs on |
08:52.20 | *** join/#htc-linux druidu (n=root@213.164.231.5) |
08:52.20 | tmzt | I guess freesmartphone is doing a more event-based api with specific behaviors configurable |
08:52.21 | swetland | totally -- on omap3 you might have some stuff to use the dspbridge to accelerate a/v |
08:52.29 | druidu | wohoooo |
08:52.34 | druidu | I have a beautiful console on my htc diamond |
08:52.41 | dcordes | :) |
08:52.46 | tmzt | fbcon or early? |
08:52.50 | swetland | congrats |
08:53.06 | druidu | I think it goes all the way, I get "proc_comm: TIMEOUT: modem has probably crashed, retrying" in an endless looop |
08:53.09 | druidu | goona disable it |
08:53.16 | druidu | my own hand-written console driver :P |
08:53.28 | druidu | works only with display powered on by wind0ze ce |
08:53.35 | druidu | yey, thanks! |
08:53.41 | tmzt | I saw that commit today when I was reading |
08:54.01 | druidu | swetland : any idea on that error? proc_comm interface not compatible? |
08:54.08 | druidu | tmzt : what commit? |
08:54.54 | tmzt | older than today, maybe september |
08:55.04 | swetland | might be shared memory interface difference between the wince supporting amss and the linux supporting one |
08:55.10 | swetland | or, the modem has stopped responding |
08:55.14 | druidu | any idea on the amss version? |
08:55.16 | druidu | the new one? |
08:55.29 | swetland | I don't know any details about the build they use for wince |
08:55.31 | druidu | 6.2.20 + new adsp |
08:55.40 | swetland | just that it's not identical to what we're using for linux |
08:56.27 | tmzt | druidu: I just remember it mentioning retrying if the modem didn't work, somewhere on git.android.com |
08:56.31 | dcordes | druidu, look at vogue-smd.c (htc-vogue branch) and kaiser-smd.c (htcmsm branch) in git.linxutogo.org |
08:56.50 | par | everything has a big impact, even a "green" impact.. as well as promises hope for people in developing countries. |
08:57.08 | par | anyway thanks :) |
08:57.30 | druidu | is it true that you're using a different flavour of linux for g1? I saw in that press release that it would be something from open kernel labs? |
08:57.38 | swetland | ahahhahahahahah |
08:57.40 | swetland | no. |
08:57.50 | druidu | there's a press release for that |
08:57.54 | druidu | http://www.tradingmarkets.com/.site/news/Stock%20News/1921360/ |
08:58.04 | druidu | maybe a different component? |
08:58.24 | swetland | no clue with the OKL folks are smoking, but there's no L4 on the apps/linux side of g1 |
08:58.33 | druidu | weird :) |
08:58.34 | swetland | http://android.git.kernel.org/?p=kernel/msm.git;a=shortlog;h=refs/heads/android-msm-htc-2.6.25 |
08:58.42 | swetland | that's what's running on g1 |
08:58.43 | tmzt | rlove's security looks interesting though, compared to "open" kernel |
08:58.46 | druidu | that's the branch I'm using now |
08:58.51 | druidu | cool B-) |
08:58.56 | tmzt | checking euids? |
08:59.00 | dcordes | druidu, I suggest it because diamond smd setup is more similar to kaiser/vogue than to g1. |
08:59.32 | *** join/#htc-linux miknix (n=miknix@gentoo/developer/miknix) |
08:59.43 | tmzt | druidu: oh I see, you get console and *now* you see that message, you just couldn't see it before? |
08:59.48 | dcordes | druidu, maybe you should first look at input/network ? |
08:59.48 | swetland | tmzt: we use extended groups to track and enforce various permissions. that was the simplest way to have the system be able to broadly grant (or deny) an app access to create ip/bt sockets |
09:00.20 | swetland | s/extended/supplementary/ |
09:00.39 | tmzt | easier than selinux or another pluggable api (if it was still possible) |
09:00.53 | druidu | swetland: and the board-trout? |
09:00.58 | druidu | is that the board on g1? |
09:00.59 | tmzt | yaffs2 supports XA? |
09:01.05 | swetland | yup |
09:01.13 | swetland | (board trout is g1) |
09:01.17 | druidu | tmzt : actually, now it magically works |
09:01.18 | druidu | no error |
09:01.23 | dcordes | druidu, does diamond have wifi? |
09:01.25 | druidu | I get all the way |
09:01.33 | druidu | including initializing net protocvols |
09:01.35 | druidu | yes |
09:01.53 | druidu | everything g1 has except memory card, compass and keyboard |
09:02.14 | dcordes | you could try the tiwlan |
09:02.16 | swetland | okay. must sleep. 'night folks. happy hacking |
09:02.24 | druidu | gonna have to debug the proc_comm thing / try code from vogue/kaiser |
09:02.25 | dcordes | good night |
09:02.37 | druidu | me too, @ school, got to go to some labs... |
09:02.40 | druidu | `night :) |
09:02.40 | ALoGeNo | nn swetland |
09:02.48 | par | nite |
09:05.38 | ALoGeNo | in that 10min i was learn more than in 2 weeks xD |
09:05.46 | *** join/#htc-linux Kalemas (i=ejabberd@jabbus.org) |
09:06.02 | ALoGeNo | reading here perhabs :) |
09:06.15 | *** part/#htc-linux Kalemas (i=ejabberd@jabbus.org) |
09:07.22 | par | thats why we at rikers G |
09:09.58 | kiozen | btw http://irclog.iclem.net/?chan=htc-linux is nicer that rikers, but less stable |
09:10.18 | kiozen | s/that/than/ |
09:12.57 | *** join/#htc-linux rakeem (n=rak@n058152149041.netvigator.com) |
09:15.06 | dcordes | bbl |
09:18.35 | *** join/#htc-linux ginge_ (n=ginge@host81-149-155-58.in-addr.btopenworld.com) |
09:24.17 | *** join/#htc-linux datachaos (n=datachao@201.22.222.107.adsl.gvt.net.br) |
09:28.27 | *** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
09:43.25 | cr2_ | hi |
09:45.17 | kiozen | cr2_: hi |
09:45.17 | kiozen | cr2_: did the fix work for you? |
09:45.19 | cr2_ | kiozen: no |
09:45.29 | kiozen | :( |
09:45.37 | cr2_ | kiozen: or i don't understand how to control the program. |
09:45.45 | cr2_ | which is your fault :) |
09:46.26 | kiozen | cr2_: shouldnt be much difference to the old version |
09:46.35 | kiozen | cr2_: you select the tdb |
09:46.46 | kiozen | cr2_: you select the basemap file |
09:46.54 | cr2_ | NetRipper: are you here ? |
09:46.55 | kiozen | that's it |
09:47.08 | cr2_ | kiozen: ok, i've done it |
09:47.30 | cr2_ | kiozen: the map is centered at 16N 46E |
09:48.49 | kiozen | cr2_: that is strange |
09:50.47 | cr2_ | swetland: the g1 code controls the clocks (and gpio setup) over proc_comm, but we need to do it by direct hardware access. |
09:51.58 | cr2_ | swetland: to idle the cpu, htc changes something in the SDRAM config area (timing?) |
09:55.53 | *** join/#htc-linux ltxda (n=anon@unaffiliated/ltxda) |
09:56.40 | *** join/#htc-linux BabelO_ (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
09:57.34 | BabelO_ | hi |
09:57.39 | BabelO_ | kiozen: small issue |
09:58.56 | kiozen | BabelO_: how big is it in real ;) |
09:59.08 | swetland | ms |
09:59.08 | BabelO_ | kiozen: |
09:59.10 | swetland | oops |
09:59.21 | BabelO_ | swetland: YOU ARE NOT SLEEPING , |
09:59.26 | BabelO_ | oups sorry caps |
09:59.45 | swetland | yes, I am failing at sleep |
10:00.49 | BabelO_ | swetland: ok, good night :) |
10:01.04 | BabelO_ | kiozen: multi selection of img is nto available on windows :( |
10:01.11 | kiozen | thinks: whereas swetland is falling asleep BabelO_ did not wake up yet :) |
10:01.42 | kiozen | BabelO_: multiselection of img? what is that? |
10:02.39 | BabelO_ | kiozen: selection of tdb file open a second selection file for img files |
10:03.03 | kiozen | BabelO_: yes, there can be only one basemap |
10:03.51 | kiozen | BabelO_: the usage of all other files is defined by the tdb file |
10:09.22 | BabelO_ | kiozen: i don't understand, if i select the tdb, i can select only one img, so only one is displayed |
10:09.39 | BabelO_ | kiozen: on linux, i ve selected the whole img file with multislect |
10:11.12 | cr2_ | BabelO_: i think that the file selection gui is actually inappropriate there. |
10:11.55 | cr2_ | BabelO_: because it is confusing, and implies to do what you are doing. |
10:12.41 | cr2_ | "file selection" == standard dialog |
10:15.04 | BabelO_ | cr2_: ok, so where is the problem ? img should be auto loaded from the location area |
10:15.17 | cr2_ | yes |
10:15.34 | cr2_ | it does not work for me now, but it must be another problem. |
10:16.05 | BabelO_ | cr2_: it was working for me when i try on linux with revision r814 |
10:16.13 | kiozen | cr2_: it says "Select base map for ..." |
10:16.43 | cr2_ | BabelO_: qlandkarte works for me as expected |
10:16.51 | cr2_ | BabelO_: !GT |
10:16.59 | BabelO_ | kiozen: (dx or dy) not compatible w<ith msvc :) |
10:17.02 | kiozen | cr2_: it's exactly the sam in qlandkarte |
10:17.29 | kiozen | you have to select the basemap, too |
10:17.31 | cr2_ | kiozen: but it also opens the big (empty) file selection dialog. i find it confusing. as an end-user :) |
10:18.20 | kiozen | GT or original? |
10:18.36 | cr2_ | original works, GT - not |
10:18.58 | cr2_ | must be something wrong with my setup. |
10:19.14 | kiozen | cr2_: try to remove |
10:19.28 | cr2_ | maybe it works, but it does not show anything :) |
10:19.32 | kiozen | ~/.config/QLandkarteGT/QLandkarteGT.conf |
10:19.37 | cr2_ | ok |
10:20.17 | BabelO_ | kiozen: ok i see what you mean by basemap, i select the wrong basemap :) can it be a automatic selection of basemap ? does tdb file describe it ? |
10:20.53 | kiozen | BabelO_: no, the real problem is that Garmin fucked it up |
10:21.18 | kiozen | BabelO_: usually a map is installed by an installer in win32 |
10:21.32 | BabelO_ | kiozen: ok, because i see the basemap at high level zoom and when i zoom in, it load relaetd map ... but i cant focus at the correct lat/long with that :) |
10:21.37 | cr2_ | kiozen: 'Select basemap for OSM-EU', and the gialog is empty |
10:21.37 | kiozen | BabelO_: this installer will tell the registry the basemap name |
10:21.41 | BabelO_ | bbl, food time |
10:21.54 | kiozen | cr2_: ? |
10:23.50 | kiozen | cr2_: the img files are in the same directory as the tdb? |
10:24.29 | kiozen | cr2_: and what are their extensions |
10:26.57 | cr2_ | kiozen: i think that it's extremely confusing for a non-mapsource person. |
10:27.07 | cr2_ | kiozen: yes. .img |
10:28.12 | cr2_ | kiozen: i've pressed 'enter', and they popped up. all of them |
10:29.11 | cr2_ | kiozen: now one needs to decide which one of the 4321 files is 'basemap' |
10:29.13 | kiozen | cr2_: I can't help it. There is no other way to know the basemap filename |
10:30.08 | kiozen | cr2_: it's just bad design by Garmin |
10:30.20 | cr2_ | kiozen: hmm. then GT should create an own .tdb from the .img collection |
10:30.23 | cr2_ | yeah. |
10:31.15 | cr2_ | i think garmin intentionally murks the water here. |
10:31.35 | kiozen | cr2_: no imho pure stupidity |
10:31.47 | cr2_ | dont believe it |
10:32.15 | cr2_ | it's only to lock you into mapsource+standard mapsets |
10:32.23 | cr2_ | "standard" |
10:32.30 | kiozen | cr2_: I can't see a real reason for it, but stupid osm uses a bad name for their basemap |
10:32.47 | cr2_ | ok, but then it's osm/mkgmap fault |
10:33.23 | kiozen | and that is the latest high quality basemap by osm: http://www.qlandkarte.org/shot6.png |
10:33.57 | cr2_ | hehe |
10:34.13 | kiozen | cr2_: this stuff is just badbadbad and goes worse |
10:34.44 | cr2_ | kiozen: well, let's write a mkgmap replacement in c++ |
10:35.12 | kiozen | nooooo not me! |
10:35.17 | cr2_ | but i'd like to dig into the routing graph setup first. |
10:35.26 | cr2_ | otherwise it's just not worth it. |
10:35.38 | cr2_ | bbl. food time |
10:35.41 | kiozen | bbl |
10:35.50 | NetRipper | cr2_, i am here, every now and then... i read druidu has fbcon working, will ask if he makes a .patch so we can work with it too ;) |
10:37.08 | *** join/#htc-linux ginge__ (n=ginge@host81-138-17-244.in-addr.btopenworld.com) |
10:38.53 | *** join/#htc-linux alo (n=QUAKEIII@165.Red-88-27-115.staticIP.rima-tde.net) |
10:41.43 | *** join/#htc-linux alo (n=QUAKEIII@195.Red-81-32-138.dynamicIP.rima-tde.net) |
10:41.52 | *** join/#htc-linux ginge__ (n=ginge@host81-138-17-244.in-addr.btopenworld.com) |
10:42.47 | *** join/#htc-linux ImCoKeMaN (n=imcokema@pool-96-249-151-72.hrbgpa.fios.verizon.net) |
10:43.23 | *** join/#htc-linux TheJoker (i=Othello@gateway/tor/x-fd6d257fea21fd0d) |
11:01.23 | NetRipper | cr2_, http://wiki.xda-developers.com/index.php?pagename=RaphaelLinux thought it'd be nice for the linux thread |
11:07.08 | *** join/#htc-linux Tinyboom (n=nahh@72.84-49-71.nextgentel.com) |
11:10.46 | *** join/#htc-linux tsdogs (n=tsdogs@net70-17.metalit.net) |
11:23.02 | *** join/#htc-linux datachaos (n=datachao@189.6.250.25) |
11:23.35 | *** join/#htc-linux ginge___ (n=ginge@host81-149-155-58.in-addr.btopenworld.com) |
11:24.25 | kaze_ | hello ginge any news on kaiser kernel? |
11:24.53 | ginge | um, no not really. I only just got it back from repair so not had chance to look at anything |
11:25.13 | ginge | someone else might have made progress in my absense |
11:25.15 | kaze_ | its all good since you got it back :) |
11:25.24 | kaze_ | ginge nope noone did |
11:25.34 | kaze_ | i talked with dcordes yesternight |
11:26.01 | ginge | pk. I guess I had better get back to writing code :) |
11:26.12 | kaze_ | that would be so nice |
11:26.42 | kaze_ | now that android source is out, it might help with wifi |
11:26.52 | kaze_ | or so i understood |
11:27.12 | ginge | possibly. I am more concerneed about SD/SDIO not working, which is what wifi runs through |
11:27.34 | kaze_ | sd is biggest concern it seems |
11:29.43 | *** join/#htc-linux team1 (n=team1@s559116c1.adsl.wanadoo.nl) |
11:53.51 | *** join/#htc-linux Xanthin (n=rnieva@salvia.nieva.com) |
11:57.15 | BabelO_ | kiozen: under windows i can't get this view :( http://www.qlandkarte.org/shot6.png |
12:04.51 | *** join/#htc-linux szsoftware (n=sven@e178219120.adsl.alicedsl.de) |
12:04.59 | kiozen | BabelO_: no map at all? |
12:07.27 | BabelO_ | kiozen: just the one i select, on linux i can select all and i see your view |
12:08.13 | kiozen | you picked the wrong basemap? |
12:17.42 | *** join/#htc-linux evildarknight (n=charles@41.207.132.103) |
12:27.39 | BabelO_ | kiozen: yes i picked the wrong |
12:27.56 | BabelO_ | kiozen: i see some bug in the render of maps, i send you one i have |
12:28.24 | BabelO_ | kiozen: bug is that the green area are on front of building |
12:29.24 | kiozen | BabelO_: try svn up first, I optimized draw order, everything else is a map bug |
12:30.09 | BabelO_ | kiozen: ok, but not a map bug, if i convert this img to russa render is good, i svn up.. look at my maps, density is high :) |
12:31.10 | BabelO_ | kiozen: bug is still here , i already svn up with my current build |
12:31.58 | kiozen | BabelO_: but russa will fail on other maps, polygon draw order is one of the bigges defiencies of the img format |
12:32.51 | BabelO_ | kiozen: can we define green area as transparent ? |
12:33.19 | kiozen | BabelO_: that will be a big performance hit |
12:33.49 | *** join/#htc-linux goxboxlive (n=goxboxli@139.84-48-212.nextgentel.com) |
12:53.20 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-78-243-191.pppoe.mtu-net.ru) |
13:19.19 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-78-243-191.pppoe.mtu-net.ru) |
13:20.58 | marajin | ginge: what'd you do to your poor kaiser this time? :P |
13:25.25 | ginge | battery wouldn't charge anhd it got hot enough to heat a small house |
13:27.03 | ezak | pray for ginge kaiser, so we can maybe get a workin android distro |
13:29.00 | ginge | we can only hope |
13:31.45 | *** join/#htc-linux Millow (n=emilio@kontoret.redpill.se) |
13:31.47 | Millow | hello |
13:31.54 | Millow | anything new about android + htc diamond? |
13:33.43 | dante_ | Millow: http://wiki.xda-developers.com/index.php?pagename=RaphaelLinux |
13:34.16 | Millow | dante_: yes, but now that android source is out I guessed maybe someone else had some more information :) |
13:36.01 | *** part/#htc-linux Millow (n=emilio@kontoret.redpill.se) |
13:45.51 | *** join/#htc-linux hlbot (n=adm@iclem.net) |
13:53.03 | *** join/#htc-linux timebomb (n=timebomb@p4FDE5885.dip.t-dialin.net) |
13:56.27 | *** join/#htc-linux Kalemas (i=ejabberd@jabbus.org) |
14:00.31 | *** join/#htc-linux szsoftwa1e (n=sven@e178207072.adsl.alicedsl.de) |
14:02.56 | szsoftware_ | swetland: kernel build trouble, after doing "make defconfig msm_defconfig && make zImage": http://pastebin.com/m4cf5d904 |
14:21.56 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-bea01fc56ea473f8) |
14:33.57 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-78-239-222.pppoe.mtu-net.ru) |
14:39.23 | *** join/#htc-linux FRITZ (n=FRITZ@189-19-168-237.dsl.telesp.net.br) |
15:07.34 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d87cc06.pool.einsundeins.de) |
15:07.39 | *** part/#htc-linux Kalemas (i=ejabberd@jabbus.org) |
15:09.42 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-78-237-23.pppoe.mtu-net.ru) |
15:09.53 | *** join/#htc-linux goxboxlive (n=goxboxli@139.84-48-212.nextgentel.com) |
15:14.58 | *** join/#htc-linux GPFerror (n=gpferror@cpe-76-187-41-132.tx.res.rr.com) |
15:16.47 | *** join/#htc-linux marmotta (n=skodde@unaffiliated/skodde) |
15:25.50 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-77-158-12.pppoe.mtu-net.ru) |
15:43.38 | *** join/#htc-linux imfloflo (i=58b44979@gateway/web/ajax/mibbit.com/x-94bb9e0f4de1e63b) |
15:43.59 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-77-159-67.pppoe.mtu-net.ru) |
15:47.28 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-9ed1ce0f5a0da5ed) |
15:55.41 | szsoftware | in the current android kernel source: what defines CONFIG_ANDROID_POWER ? can't compile with "make defconfig msm_defconfig" |
15:55.50 | szsoftware | cr2_: Do you know about this? |
15:58.01 | imfloflo | ask in #android channel |
15:59.37 | NetRipper | szsoftware, in your .config, check the CONFIG_ANDROID_* stuff |
16:00.11 | ali1234 | BabelO: could you pastebin a dmesg from artemis please? |
16:00.37 | *** join/#htc-linux lastik_ (n=eraser@80.92.96.29) |
16:00.38 | ali1234 | i just can't get these uarts going for some reason and i'm getting weird errors |
16:07.54 | cr2_ | NetRipper: have you tried to comment out the proc_comm calls in clk_* like it's done on vogue ? |
16:10.47 | *** join/#htc-linux scatterp_ (n=scatterp@82-35-152-46.cable.ubr05.enfi.blueyonder.co.uk) |
16:11.44 | *** join/#htc-linux Abracadabra (n=aaabraca@unaffiliated/abracadabra) |
16:12.05 | scatterp_ | hi i heard that the new google os has been released does any one know the effect it will have on drivers for htc linux yet |
16:12.12 | scatterp_ | or is it still being studyed ? |
16:13.13 | cr2_ | scatterp_: the core 7201A cpu support it there. |
16:13.51 | cr2_ | scatterp_: but there are differences in clock and gpio control. |
16:14.23 | cr2_ | the most important thing is that we have wifi driver now. |
16:15.05 | scatterp_ | will they be able to sniff ? |
16:15.29 | Abracadabra | Hi, how is Linux installed on HTC ? .... is it possible to boot it from a Micro SD card ? |
16:15.45 | scatterp_ | Abracadabra, if you look in the topic |
16:15.52 | scatterp_ | the links will explain how to do it |
16:16.00 | Abracadabra | or do we have to modify the firmware and/or install Linux over Windows Mobile ? |
16:16.22 | scatterp_ | its booted from inside windows no over writeing firmware yet |
16:16.31 | scatterp_ | cr2_, what about direct booting ? |
16:18.11 | cr2_ | scatterp_: g1 uses a different (?) bootlloader |
16:18.23 | scatterp_ | oh that sucks :( |
16:18.33 | cr2_ | scatterp_: there is no info about it (no dump) |
16:18.49 | scatterp_ | oh so maybe we will get a dump soon |
16:18.49 | imfloflo | bootloader as be release too |
16:18.54 | cr2_ | but there is certainly no source for it, and there will be no source for it. |
16:19.38 | cr2_ | imfloflo: hehe. this bootloader is no better than haret. |
16:20.05 | cr2_ | because it does not really initialize the hardware from scratch. |
16:20.07 | imfloflo | but with this bootloader you aren't oblige to boot in wm fisrt? |
16:20.48 | cr2_ | you can put linux in place of appsbl |
16:20.55 | cr2_ | in theory |
16:20.59 | imfloflo | so still waiting for a dump ? there are many people who already have g1 so why they don't dump it with the usb driver |
16:21.54 | szsoftware | NetRipper: there is no CONFIG_ANDROID_POWER in .config. So defconfig does not enable this, so it's a bug? |
16:22.23 | cr2_ | imfloflo: there are 6 mtd partitions afaik. "android (tm)" is only 1 of them |
16:22.49 | imfloflo | oki i didn't know |
16:22.59 | imfloflo | want to learn |
16:23.30 | cr2_ | imfloflo: and even the android itself is not 100% open source. i think many people miss this point. |
16:25.01 | scatterp_ | cr2_, i have the htc hermes is there a linux version with wifi avaialble yet ? |
16:25.07 | imfloflo | maybe one day |
16:25.52 | cr2_ | scatterp_: no, but it's a realistic project |
16:26.04 | *** join/#htc-linux LunohoD (n=alex@e180068214.adsl.alicedsl.de) |
16:26.22 | cr2_ | scatterp_: because you only need to apply power and setup the irq handler. all this data is known. |
16:27.04 | scatterp_ | i guess with the new gpios it leaves less options now for what the gpios could be right ? |
16:27.15 | scatterp_ | i mean the unknown gpios |
16:27.51 | cr2_ | http://wiki.xda-developers.com/index.php?pagename=HermesGPIO |
16:28.19 | cr2_ | hmm. it probably needs some better descriptions. |
16:28.42 | scatterp_ | also the wifi codes are not added there |
16:29.22 | *** join/#htc-linux Othello (i=Othello@gateway/tor/x-342e68463db32533) |
16:37.24 | szsoftware | NetRipper: cr2_ when I add CONFIG_ANDROID_POWER=y to .config, I get this error: .config:1147:warning: trying to assign nonexistent symbol ANDROID_POWER. Am I alone with that? |
16:37.57 | *** join/#htc-linux SanMehat (n=san@nat/google/x-260282151d9118de) |
16:38.15 | szsoftware | hello SanMehat :-) |
16:38.27 | SanMehat | hey sz :) |
16:38.29 | SanMehat | whats up? |
16:40.09 | szsoftware | it seems you and swetland reach a goal and now you're seen often here. :-) Look what the community does with your food? hehe |
16:40.25 | ali1234 | BabelO: i found the dmesg dump on xda-developers |
16:40.42 | szsoftware | SanMehat: seems, last month and weeks you where deep in work. |
16:40.59 | szsoftware | SanMehat: did you read the logs of the last 30min? |
16:41.27 | cr2_ | szsoftware: forget about power management for now |
16:41.45 | szsoftware | cr2_: but it's needed for compiling :-( |
16:42.21 | szsoftware | there is a ifdef CONFIG_ANDROID_POWER, and without that it doesn't build msm_fb.c |
16:42.44 | SanMehat | szsoftware: no i havent read logs.. sec |
16:43.54 | SanMehat | szsoftware: the logs calender doesnt have an entry for today yet |
16:44.26 | szsoftware | SanMehat: you don't need. It's the last written line by me. What in make menuconfig, sets CONFIG_ANDROID_POWER ? |
16:44.53 | szsoftware | SanMehat: you read on rikers? try iclem.net, its very fast |
16:45.41 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
16:49.25 | SanMehat | szsoftware: just define it in your defconfig |
16:49.37 | SanMehat | szsoftware: yeah we've been busy the last few months :) |
16:52.20 | szsoftware | SanMehat: so there's a bug in drivers/video/msm/msm_fb.c:513. This line needs a ifdef CONFIG_ANDROID_POWER too, it refers to power_on_panel(), which is in ifdef-section |
16:52.48 | SanMehat | szsoftware: please post a patch to our msm-linux list :) |
16:53.20 | SanMehat | i'll be happy to integrate it for you |
16:53.27 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
16:53.41 | szsoftware | SanMehat: ok :-) |
17:07.07 | NetRipper | szsoftware, do you make with ARCH=arm? |
17:07.53 | NetRipper | and if you change your .config, don't use 'make defconfig' anymore |
17:11.02 | szsoftware | NetRipper: yeah, I did. As I said to SanMehat, I think it's a lillte bug (missing ifdef), so it works without CONFIG_ANDROID_POWER |
17:13.05 | NetRipper | yes that too |
17:13.14 | NetRipper | i have CONFIG_ANDROID_POWER defined and it doesn't complain |
17:13.53 | NetRipper | man where's druidu |
17:13.57 | NetRipper | i want that console patch :p |
17:14.12 | szsoftware | ok, will check this, with editing msm_defconfig... |
17:15.48 | NetRipper | i normally just "cp arch/arm/configs/msm_defconfig .config" and then just "make ARCH=arm CROSS_COMPILE=arm-none-eabi-" |
17:15.48 | NetRipper | it'll ask for the missing pieces and continue building |
17:15.48 | NetRipper | :) |
17:15.56 | NetRipper | i wasn't even aware of 'make defconfig' ;) |
17:18.29 | SanMehat | make msm_defconfig is your friend :) |
17:22.05 | *** part/#htc-linux BabelO_ (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
17:29.32 | *** join/#htc-linux fnord (n=fnord@192.234.2.49) |
17:29.46 | *** join/#htc-linux druidu (n=root@213.164.231.5) |
17:35.33 | NetRipper | heya druidu |
17:36.04 | NetRipper | druidu, can you make a patch of your tree with the working console? :) |
17:43.35 | *** join/#htc-linux marex (n=marex@vasut.kolej.mff.cuni.cz) |
17:44.05 | BabelO | ali1234: yes my auto install image create a dmesg.txt on the FAT partition for debug :) |
17:46.28 | ali1234 | BabelO: i found the dmesg on the forum... but it didn't help. i think our kernel is broken with the UARTs |
17:46.57 | ali1234 | BabelO: this is what your dmesg says: http://pastebin.com/m2347352a this is what mine says: http://pastebin.com/m7551fda9 |
17:47.05 | BabelO | ali1234: your git repository.. maybe, because with linuxtogo git and 2.6.26, i ve no problem |
17:47.29 | ali1234 | BabelO: probably, we are looking at diffs now :) |
17:48.00 | BabelO | ali1234: look in htcartemis.c machine file there is clock settings for uart, wait i look at it too |
17:51.27 | BabelO | ali1234: |
17:51.31 | BabelO | <PROTECTED> |
17:51.31 | BabelO | <PROTECTED> |
17:51.31 | BabelO | <PROTECTED> |
17:51.31 | BabelO | <PROTECTED> |
17:51.55 | BabelO | and after that you call omap_serial_init() |
17:53.07 | ali1234 | BabelO: right right. i bet we dont do any of that in the wizard board file |
17:53.15 | ali1234 | BabelO: thanks, will try it |
17:54.06 | *** join/#htc-linux Guimli (n=guimli@ecu69-1-82-231-127-213.fbx.proxad.net) |
17:57.02 | druidu | yes |
17:57.16 | druidu | I was all day at school, after not sleeping last night... full day I can say :) |
17:57.28 | druidu | btw |
17:57.33 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-75-85-249-111.bak.res.rr.com) |
17:57.35 | ali1234 | BabelO: what about omap_board_config? is that actually used? |
17:58.06 | druidu | with the console fix and board-trout, kernel boots with no problems until proc_comm |
17:58.14 | druidu | NetRipper |
17:59.34 | *** join/#htc-linux zsircusr1 (n=zsirc@83.149.10.109) |
17:59.45 | *** join/#htc-linux LanceHaig (i=4e696aee@foresight/member/lhaig) |
18:00.59 | *** join/#htc-linux dcordes (n=dcordes_@f048070147.adsl.alicedsl.de) |
18:01.16 | druidu | NetRipper: let me make some more tests and I'll make you a patch |
18:01.51 | dcordes | hi |
18:06.22 | BabelO | ali1234: yes you need to add definition to board_config |
18:08.07 | *** join/#htc-linux LanceHaig (i=4e696aee@foresight/member/lhaig) |
18:08.19 | ali1234 | BabelO: it worked thanks |
18:09.37 | BabelO | ali1234: :) |
18:10.01 | ali1234 | BabelO: i added the htcartemis_bt to the devices array but that's not what i meant by board_config |
18:10.10 | dcordes | druidu, exclude proc_comm from your build? |
18:10.52 | dcordes | Funky Weasel is Jiggy wit it |
18:12.40 | szsoftware | dcordes: what is the plan concerning applying new stuff? android sources as new base and applying patches step by step? |
18:13.30 | dcordes | best would be to put ginge tree stuff on android-msm-htc-2.6.25 I guess |
18:14.19 | dcordes | since the former is based on the android-msm-2.6.25 branch and I assume android-msm-htc-2.6.25 is also, in a way. |
18:14.29 | szsoftware | dcordes: btw, Hi :-) |
18:14.38 | dcordes | hi |
18:14.53 | szsoftware | dcordes: yes, I have this in mind too |
18:16.22 | *** join/#htc-linux lastik__ (n=zsirc@83.149.10.109) |
18:19.07 | dcordes | szsoftware, ever got gprs working on the polaris with the tree from the tarball? |
18:25.20 | szsoftware | dcordes: no |
18:27.08 | diogene31 | dcordes: Hi there. I bought a kaiser today :) |
18:27.54 | kaze | :O |
18:27.56 | kaze | hooray |
18:28.08 | kaze | kaiser is a nice brick |
18:28.18 | kaze | you can even use hit to hurt people |
18:28.30 | kaze | but i still love mine :) |
18:28.59 | dcordes | diogene31, nice :) for 99€ with contract ? |
18:29.02 | cr2_ | diogene31: lol |
18:29.09 | kaze | btw diogene31 where did you find one for 99. ? |
18:29.12 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
18:29.18 | dcordes | hey kaze |
18:29.23 | kaze | hey dcordes |
18:29.28 | kaze | hoow are you |
18:29.34 | kaze | im just bacj from work :( |
18:29.35 | cr2_ | druidu: can you comment out the proc_comm calls (and add printks) to the clock api ? |
18:30.23 | diogene31 | kaze: FNAC shop, in Toulouse. The second less expensive contract :) |
18:30.32 | dcordes | kaze, I'm enjoying my last week of brutal unemployment :) |
18:30.38 | kaze | damn |
18:30.39 | *** join/#htc-linux bryanlet (n=b@206-248-163-182.dsl.teksavvy.com) |
18:30.47 | kaze | bought mine at easycash for 249 |
18:30.54 | kaze | contract lees but locked to orange |
18:31.00 | kaze | still its easy to unlock |
18:31.10 | dcordes | that's a nice offer |
18:31.38 | dcordes | vodafone germany sell the kaiser still for around 300 something with an expensive contract |
18:32.51 | kaze | my attempts at buildin kaiser kernel on freebsd always fail... |
18:33.08 | dcordes | kaze, what's the problem? no suitable toolchain? |
18:33.17 | kaze | anyone ever setup a workin cross compiling environnement on fbsd? |
18:33.23 | kaze | dcordes i think so |
18:33.35 | kaze | i tried to use linux toolchain with linux compat on |
18:33.37 | kaze | no go |
18:34.04 | kaze | and im a bit lazy to setup a linux box |
18:34.17 | kaze | aznd a bit busy too |
18:34.46 | kaze | will probably end up in vmware or something like that |
18:34.48 | *** join/#htc-linux Tinyboom (n=nahh@72.84-49-71.nextgentel.com) |
18:34.54 | dcordes | szsoftware, wow, I just applied http://headfuzz.co.uk/files/android/git-diff-2.6.25-170908.diff to the clean android-msm-htc-2.6.25 and it went ok. not too many fails |
18:35.41 | druidu | clock works I think |
18:35.53 | szsoftware | dcordes: cool, this patch is kinda old. but will do this also .-) |
18:35.54 | druidu | I tried adding printk delays but no luck, kernel won't boot |
18:36.25 | dcordes | szsoftware, well the stuff in the tarball I sent is basically that patch, plus polaris, plus garbage. |
18:36.37 | dcordes | so we have nothin much newer |
18:37.06 | dcordes | except you tried to rebase to the updated code (or whatever that's called) |
18:38.01 | szsoftware | dcordes: I think, we should add patches step by step to the new base. |
18:38.51 | diogene31 | dcordes: Where do you commit your work on the kaiser kernel ? |
18:38.54 | dcordes | why, only 10 hungs failed |
18:39.32 | kaze | diogene31 http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/htc-msm |
18:40.17 | kaze | (notice my eager to see someone new working on kaiser kernel too) |
18:40.25 | dcordes | diogene31, http://headfuzz.co.uk/files/android/git-diff-2.6.25-170908.diff this is the latest code on 2.6.25. the above branch has 2.6.24 |
18:40.36 | kaze | :X |
18:40.53 | szsoftware | dcordes: any progress with wifi? |
18:41.26 | cr2_ | druidu: printk delays ? |
18:42.00 | dcordes | diogene31, http://linuxtogo.org/~lgorris/index.php?path=kaiser-bootkit/ there's a tiny documentation. I hope it can answer some of your questions |
18:42.40 | druidu | yes |
18:42.51 | cr2_ | druidu: what for ? |
18:43.04 | druidu | to see everything, it scroll down really fast and I miss the first stuff |
18:43.24 | dcordes | szsoftware, no. I only checked it out and don't understand the structure. ginge said earlier he excepts the sdcc dma problem to interfer wifi as well |
18:43.28 | cr2_ | druidu: the proc_comm clock api does not exist on wince |
18:43.28 | diogene31 | dcordes: Thanks. I was just curious. In a couple of monthes, once mioa701 will be fully integrated in mainline, perhaps I'll push the kaiser upstream, if nobody did it before. |
18:43.57 | cr2_ | druidu: so it should be #ifdef'd ANDROID or semthing like that |
18:44.05 | druidu | actually, in board-trout proc_comm is used for camera/bluetooth stuff |
18:44.06 | dcordes | diogene31, ok. szsoftware and I will try and go 2.6.26 then |
18:44.33 | cr2_ | druidu: proc_comm will be used for -> |
18:44.37 | dcordes | diogene31, uhm sorry. 2.6.25 first. |
18:44.59 | cr2_ | druidu: http://wiki.xda-developers.com/index.php?pagename=RaphaelDEX# |
18:45.04 | dcordes | diogene31, just trying to get as close as possible to the google code. |
18:46.02 | cr2_ | druidu: the gpio setup (owner and alt), and the clocks are addressed directly |
18:46.17 | dcordes | druidu, are this the proc_comm functions? |
18:46.18 | cr2_ | druidu: without any arm9 help |
18:46.27 | szsoftware | haha, regarding that CONFIG_ANDROID_POWER issue: I was on the wrong branch ^^. |
18:47.38 | druidu | trout_gpio_write(NULL, TROUT_GPIO_USB_H2W_SW, cpld_usb_h2w_sw); |
18:47.42 | druidu | what does this do? |
18:48.07 | *** join/#htc-linux marex (n=marex@vasut.kolej.mff.cuni.cz) |
18:48.27 | szsoftware | dcordes: I'm curious, what dzo does with the new stuff. |
18:48.32 | tmzt | kaze: you try a chroot? |
18:48.51 | dcordes | szsoftware, same |
18:49.07 | szsoftware | dcordes: regarding dma, did'nt googles devs disable dma for now? can we do this too? |
18:49.16 | dcordes | szsoftware, did you try using his latest framebuffer code on the polaris? |
18:49.37 | dcordes | szsoftware, oh you're right. so maybe it's not relevant |
18:49.46 | imfloflo | polaris ?are you working on this? |
18:50.04 | szsoftware | dcordes: no, I'm not as proffessional hacker as I must be to fiddle this in. It's too complicated forme |
18:50.17 | szsoftware | imfloflo: I have polaris |
18:50.38 | dcordes | szsoftware, same. I did a diff and it's too much different from the current fb code in kaiser tree |
18:50.43 | *** join/#htc-linux sirox (n=sirox@91-67-142-232-dynip.superkabel.de) |
18:50.45 | sirox | hi |
18:50.50 | szsoftware | hi sirox |
18:50.52 | dcordes | szsoftware, we could ask dzo for help |
18:51.04 | imfloflo | ok so you work on polaris what are you trying to do ? |
18:51.10 | szsoftware | dcordes: we should! |
18:51.49 | sirox | how can i download the android-msm-htc-2.6.25 kernel from the android.git.kernel.org repository ? |
18:52.01 | sirox | and how can i compile the kernel ? |
18:52.05 | dcordes | szsoftware, we should present a usable diff of the .25 code with polaris for this purpose |
18:52.33 | szsoftware | imfloflo: at least I built fbi ipk. Or I'm compiling kernel, and try things. But do not really hack the kernel. |
18:53.17 | szsoftware | dcordes: a diff between what and what? |
18:53.25 | imfloflo | oki |
18:53.38 | dcordes | imfloflo, I will try and get that tiny 1.0 for ramdisk when the polaris/kaiser code is merged with the trout code |
18:53.50 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
18:54.30 | szsoftware | sirox: git-clone git://git.android.com/kernel.git |
18:54.37 | imfloflo | ok i want to try it too ^^ cause always 0.8 boring |
18:54.55 | szsoftware | sirox: then git-checkout -b android-msm-htc-2.6.25 origin/android-msm-htc-2.6.25 |
18:55.00 | dcordes | szsoftware, ina perfect world latest android-msm-htc-2.6.25 revision. |
18:55.13 | cr2_ | druidu: don't know. i'm looking at the h2w in more detail now, and there is no proc_comm here. only the already documented msm gpios |
18:55.33 | szsoftware | sirox: which device you're working on? |
18:55.38 | sirox | htc diamon |
18:55.39 | sirox | d |
18:56.00 | szsoftware | sirox: then make distclean && make msm_defconfig |
18:56.46 | sirox | ok thx, i will try |
18:56.48 | szsoftware | sirox: you know about druidu's progress? |
18:57.38 | dcordes | framebuffer on wince shipped msm7201A phones ftw!! |
18:57.55 | tmzt | does he have only early console or also fbcon? |
18:57.59 | sirox | szsoftware: no |
18:58.17 | dcordes | tmzt, I think he said an own fb driver iirc |
18:58.26 | dcordes | *fb console driver |
18:58.34 | sirox | szsoftware: who is druidu ? |
18:58.38 | szsoftware | I recommend reading the last days logs on iclem.net, he did a lot in a short time :-) |
18:59.17 | szsoftware | sirox: druidu came in this channel, and did a lot. Grep the logs for woohoo or so ;-) |
19:01.54 | sirox | szsoftware: do you know the day ? |
19:02.12 | szsoftware | sirox: last two |
19:03.00 | tmzt | has NetRipper got the patch yet? |
19:05.33 | *** join/#htc-linux Marnux (n=Marnux@83.101.56.192) |
19:05.59 | druidu | cr2_ : look at board-troat* from android kernel git |
19:06.37 | druidu | sirox: druidu is me, you can download latest kernel src with git |
19:07.11 | sirox | druidu: Oh hi druidu :D ... download like zsoftware said ? |
19:07.15 | cr2_ | druidu: disable all non-core drivers |
19:10.39 | dcordes | cr2_, I thought about how to disable proc comm. in the kernel config there is only SMD. proc_comm.c is built in any case. but it's unlikely that the rest of the code will built without because so much depends on it |
19:11.08 | *** join/#htc-linux ginge (n=baz@78-105-175-81.zone3.bethere.co.uk) |
19:11.48 | cr2_ | druidu: trout_gpio_write is the htc-egpio implementation. |
19:12.19 | cr2_ | dcordes: you need proc_comm for many things. |
19:12.39 | cr2_ | dcordes: but not for controlling the clocks, as g1 does. |
19:13.03 | cr2_ | dcordes: g1 just obfuscates some hardware access. |
19:13.43 | cr2_ | dcordes: pretending that it should go through arm9. |
19:13.48 | dcordes | cr2_, are you sure it's not needed for clock control? afaik vogue-pm.c is proc_comm functions only |
19:13.56 | dcordes | same on kaiser |
19:14.02 | cr2_ | which is not the case for wince amss |
19:14.31 | ginge | I'm with cr2_ |
19:14.41 | tmzt | was it only working because ce set up the clocks? |
19:14.52 | cr2_ | dcordes: here is the full proc_comm list -> http://wiki.xda-developers.com/index.php?pagename=RaphaelDEX |
19:14.56 | ginge | google had a lot of functions moved to amss for convenience |
19:15.00 | dcordes | ginge, how do you turn off the lcd? |
19:15.09 | ginge | magic |
19:15.42 | ginge | one mo, I'll find my notes |
19:16.31 | dcordes | not witchcraft? I see a lot of proc_comm in your kaiser-pm.c |
19:16.59 | dcordes | and iirc dzo said the power collapse code he tried also bases on the proc_comm |
19:17.14 | cr2_ | dcordes: 7 undocumented functions, btw. |
19:18.51 | dcordes | cr2_, undocumented proc_comm functions for the kaiser? |
19:18.53 | ginge | pm is not proc_comm if you look at my kaiser_pm.c it is 99% commented out :) |
19:19.17 | kaze | tmzt hi, chroot for building kernel on bsd? |
19:19.44 | tmzt | with linux syscalls it might work |
19:19.48 | cr2_ | dcordes: this list is for rapahel, but dzo says it's the same on vogue |
19:19.53 | kaze | mm good idea |
19:20.42 | dcordes | ginge, ok. do kaiser proc_comm functions match vogue ones? |
19:21.01 | ginge | function wise, yes. The addresses are different |
19:21.35 | cr2_ | ginge: the smem addresses don't match even on diam/raph |
19:21.38 | dcordes | ok I think we should put them in the wiki then |
19:22.01 | cr2_ | ginge: i wish there was a portable way to pick these offsets from wince |
19:22.02 | ginge | cr2_: they they seem to pick a number out of the air for each one |
19:22.13 | dcordes | cr2_, maybe we should make a unified proc_comm page with different rows for the device specific addresses |
19:22.45 | ginge | cr2_: I just wish they would at least amss common so it can be used more like a common API rather than the bag of crap it is at the momen |
19:22.49 | cr2_ | dcordes: i can look at the kaiser DEX |
19:23.23 | cr2_ | ginge: i think that the dex ioctl numbers are common |
19:23.42 | cr2_ | ginge: the -- commands come from nk.exe |
19:23.56 | ginge | ahh |
19:24.05 | ginge | dcordes: ok for turning off lcd... |
19:24.28 | ginge | we write to some registers in the right sequence |
19:25.12 | ginge | cant remember what most of the magic commands did, but they are just implemented to send the same old stuff that ce does |
19:25.48 | ginge | it uses a similar mechanism to the way we set clocks |
19:26.00 | cr2_ | ginge: something like that ? -> http://wiki.xda-developers.com/index.php?pagename=RaphaelLCD |
19:26.02 | ginge | iirc it also the clock memory space |
19:26.09 | dcordes | ok so they are not assigned to the proc comm function names and direct registers are used? |
19:26.17 | ginge | cr2_: yes exactly like that |
19:26.57 | cr2_ | ok |
19:27.16 | cr2_ | ginge: only the pwm registers are undocumented |
19:27.37 | cr2_ | but ymmv |
19:27.39 | cr2_ | http://wiki.xda-developers.com/index.php?pagename=RaphaelMDDI |
19:27.41 | ginge | yeah, couldn't figure that out myself |
19:28.59 | ginge | dcordes: no |
19:29.27 | ginge | dcordes: eg: |
19:29.27 | ginge | writel(0x0030000,MSM_BACKLIGHT_BRIGHTNESS); |
19:29.27 | ginge | writel(0,MSM_BACKLIGHT_BRIGHTNESS); |
19:30.35 | cr2_ | ginge: it's probably some pwm divisor |
19:30.38 | dcordes | does that write the value 0x0030000 to the address which is defined for the var MSM_BACKLIGHT_BRIGHTNESS ? |
19:31.00 | ginge | yes |
19:31.05 | cr2_ | is the MSM_BACKLIGHT_BRIGHTNESS in the MSM_CLK range ? |
19:31.17 | ginge | MSM_CLK_CTL_BASE+0x54; |
19:31.21 | cr2_ | ok |
19:31.34 | ginge | power at base+0x58 |
19:32.16 | cr2_ | ok, we need a replacement clk_* api for these |
19:32.27 | cr2_ | instead of the android's proc_comm |
19:32.30 | ginge | yeah, that would be a good idea |
19:38.43 | *** join/#htc-linux penguinz (n=opera@66-168-215-122.dhcp.athn.ga.charter.com) |
19:42.33 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-201287a76ba4cac3) |
19:42.42 | dcordes | szsoftware, lol did you see the msm_sdcc.c reject? |
19:44.30 | druidu | how do I turn off optimizations for a function? |
19:44.38 | szsoftware | dcordes: I patch again and specify reject-file manually, since I have still no *.rej |
19:44.53 | druidu | need to write hack a delay function and the compiler optimizez it to... NOTHING! |
19:45.13 | dcordes | szsoftware, patch in debian lenny does that automatically. I only do "patch -p1 < foo.diff" |
19:46.06 | cr2_ | druidu: hmm. we don't know how to control the RTC directly |
19:46.47 | cr2_ | druidu: read/write from some dummy memory location |
19:47.09 | szsoftware | dcordes: ok, now have all rejects in one file, moment.. |
19:47.24 | druidu | damn compiler is pretty smart |
19:47.29 | cr2_ | druidu: the space between the tags and L1 ptbl is not used |
19:47.30 | szsoftware | dcordes: oh, ubuntu doesn't |
19:47.33 | druidu | it's like it's guessing it's dummy code and strips it all out |
19:47.57 | tmzt | there's a pfm_delay in asm in the new usb isp116x driver for pxa255 |
19:48.20 | cr2_ | druidu: use volatile a=1; |
19:49.28 | dcordes | szsoftware, it worked the same for me in ubuntu as it does now in lenny |
19:49.32 | druidu | good idea, compiler should be afraid of touching volatile stuff :) |
19:49.45 | druidu | finally! |
19:51.08 | szsoftware | dcordes: As I said, I have no *.rej, and when I specify with patch -r rejects.txt, then in this file are only rejects, with no hint, what file is the hunk related to |
19:55.32 | dcordes | I have nfc what I am doing |
19:56.09 | NetRipper | druidu, when interrupts are on you can use (u|m)sleep() afaik |
19:56.43 | NetRipper | druidu, and i'm really looking forward to that patch :p |
19:57.53 | *** join/#htc-linux AstainHellbring (n=AstainHe@unaffiliated/astainhellbring) |
19:59.00 | NetRipper | druidu, oh and fyi, i added a RaphaelLinux page to wiki (also for diamond), in which we can keep other people informed.. you can also add to it if you feel like it when you make some progress :) http://wiki.xda-developers.com/index.php?pagename=RaphaelLinux |
19:59.28 | dcordes | cr2_, will trout battery driver work for us? |
19:59.57 | dcordes | assuming htc didn't reinvent the wheel for g1 it should, no? |
20:00.20 | ginge | dcordes: htc reinvent the wheel for each new phone... pretty much |
20:00.34 | cr2_ | dcordes: it uses proc_comm ? |
20:02.45 | dcordes | cr2_, doesn't look like it? but rpc. arch/arm/mach-msm/htc_battery.c |
20:03.21 | szsoftware | dcordes: ok, solved the rejects problem |
20:03.37 | druidu | NetRipper almost done :P |
20:03.41 | cr2_ | dcordes: ok, something to check |
20:04.05 | tmzt | rpc is another protocol over smd? |
20:04.36 | dcordes | szsoftware, ok I'm already done :) |
20:05.35 | szsoftware | dcordes: fine, all hunks? |
20:06.15 | dcordes | no left msm_sdcc.c out because it fails 100% due to rewrite |
20:06.45 | dcordes | doing a build now w/o MMC |
20:06.53 | szsoftware | dcordes: ahh, ok, you lol'ed before |
20:07.01 | NetRipper | can't we setup a git branch for raphael/diamond? |
20:07.36 | dcordes | NetRipper, why not? |
20:08.11 | szsoftware | dcordes: so msm_sdcc.c is a completly new file? Same name, 100% different content? |
20:08.33 | NetRipper | dcordes, is it easy to mirror branches and keep them up to date? i.e. if we have a clone of android-msm-htc-2.6.15 and we call it android-msm-raphael-2.6.15, would it be easy to merge any changes made later on in the htc tree to raphael tree? |
20:08.37 | NetRipper | :) |
20:08.44 | dcordes | ginge, do we need drivers/i2c/chips/mt9t013.o for kais? |
20:08.45 | NetRipper | dcordes, (i have almost no experience with git) |
20:09.20 | dcordes | szsoftware, it seems. guess it's due to the dma rewrite stuff |
20:09.33 | tmzt | is that the synaptics chip? |
20:09.53 | druidu | NetRipper : know about the wiki, read your post |
20:10.08 | NetRipper | k |
20:10.08 | NetRipper | ;) |
20:10.49 | dcordes | NetRipper, exactly what you explained (tracing a remote branch) is what was originally planned for the htc-msm (tracking git.android.com's android-msm) branch. we just never called in the new code ^^ |
20:11.32 | dcordes | I don't understand really how to set it up. pH5 prepaired that for us initially |
20:12.03 | *** part/#htc-linux penguinz (n=opera@66-168-215-122.dhcp.athn.ga.charter.com) |
20:12.07 | NetRipper | dcordes, when browsing the git over http, it shows some green tags with a 'merge'... but i dont know if its trivial to perform such a merge |
20:12.10 | *** join/#htc-linux penguinz (n=opera@66-168-215-122.dhcp.athn.ga.charter.com) |
20:12.31 | dcordes | ginge, oh that is the trout camera |
20:12.49 | dcordes | I wonder why make menuconfig automatically enables kconfigs which are not in the defconfig I chose? |
20:13.13 | tmzt | select? |
20:13.15 | ginge | dcordes: hmm? oh you are picking over the g1 source's carcass. wifi! |
20:13.20 | dcordes | tmzt, enable |
20:13.44 | tmzt | I mean select clauses in the Kconfig |
20:13.44 | dcordes | ginge, yep |
20:13.57 | NetRipper | dcordes, it takes the defaults for the values that are unknown to menuconfig, if you just make without menuconfig, it'll ask you the questions you want to see ;) |
20:14.48 | dcordes | NetRipper, ok thx will do that |
20:16.12 | dcordes | NetRipper, oops. that didn't really work. I removed .config and reran make htckaiser_defconfig . still unselected configs are built automagically. |
20:16.24 | NetRipper | eh |
20:16.25 | NetRipper | try this |
20:16.45 | NetRipper | cp arch/arm/configs/htckaiser_defconfig .config && make ARCH=arm |
20:16.59 | *** join/#htc-linux radem205 (n=aaa@dhcp-077-251-031-248.chello.nl) |
20:17.01 | dcordes | so the make *defconfig calls in the new config defaults? |
20:17.32 | NetRipper | i guess, "defconfig" is an acronym for default config anyway |
20:17.33 | dcordes | that works. |
20:17.54 | dcordes | yes |
20:17.57 | tmzt | same as copying arch/arm/config/msm_defconfig to .config ? |
20:18.06 | NetRipper | tmzt, yes |
20:18.27 | NetRipper | but make *defconfig probably automagically sets default values for unknowns |
20:19.51 | radem205 | |
20:19.51 | radem205 | is someone working to convert Android to the polaris? |
20:20.05 | dcordes | radem205, yes we will announce when there is 1.0 |
20:20.14 | dcordes | other priority now |
20:20.28 | radem205 | ok, i'm sorry for the question. I understand that! Good work! |
20:20.33 | dcordes | no problem |
20:20.37 | dcordes | you know the xda forum? |
20:20.46 | szsoftware | dcordes: I did not check msm_sdcc.c.rej completely, but the patch seems to be weird. There are almost no really differences, no? |
20:21.05 | dcordes | yes only context lines change |
20:21.15 | dcordes | but nothing important as I see it |
20:21.42 | radem205 | Yes I know, I've looked into the Kaiser Android topic so I was curious :) |
20:22.33 | NetRipper | cr2_, nice job on the MDDI wiki page |
20:24.56 | szsoftware | dcordes: so msm_sdcc.c should work like what we already have? |
20:26.52 | szsoftware | dcordes: can you send me your patch against android-msm-htc-2.6.25 ? |
20:32.57 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
20:33.37 | dcordes | szsoftware, yep wait |
20:33.51 | dcordes | let me try and boot first |
20:34.58 | *** join/#htc-linux chab7 (n=kvirc@183-17-175.ip.adsl.hu) |
20:39.19 | druidu | how can I make git-diff show new files? |
20:39.49 | dcordes | druidu, git-add file1/to/show.c file2.h file3 |
20:40.21 | dcordes | git-status brings up what you changed, and what is added or not |
20:41.03 | dcordes | you must also git-add changed files which are not new |
20:41.49 | NetRipper | druidu, before you make your patch, make this change please: http://pastebin.com/m47ca0e93 |
20:42.02 | NetRipper | druidu, the uart irqs can be enabled, just the debug timer interrupt is causing problems |
20:42.46 | NetRipper | druidu, and to make a patch: git diff -p |
20:42.51 | NetRipper | druidu, and to make a patch: git diff -p > mypatch.patch |
20:42.55 | druidu | well, I'm not used to this stuff so well: http://octav.oricum.ro/android/htc_fb_console.c http://octav.oricum.ro/android/htc_fb.diff http://octav.oricum.ro/android/linux-on-htc-diamond.jpg |
20:42.59 | *** join/#htc-linux |chab7| (n=kvirc@183-17-131.ip.adsl.hu) |
20:43.20 | druidu | the diff is really minor, just adding the file to the build tree and 2 kernel options |
20:43.23 | dcordes | druidu, delicious |
20:43.42 | kaze | i m in love with screen -d -r |
20:43.45 | druidu | I'm not making this a formal release, the code is pretty crappy anyway, but works perfectly and supports color too |
20:43.52 | kaze | so built a new kernel dcordes ? |
20:43.54 | NetRipper | kaze, you'll adore screen -x :p |
20:44.10 | NetRipper | druidu, very nice :) |
20:44.25 | kaze | netRipper ok i stay mouth shut :) |
20:44.26 | dcordes | kaze, almost :) hold on |
20:44.32 | druidu | the picture's also bad but well, had no batteries for the camera so used a damn phone |
20:44.52 | NetRipper | i'll make a pic if you want |
20:44.56 | dcordes | kaze, huh? |
20:45.05 | NetRipper | got a nice macro lens |
20:45.14 | ginge | use QList<QTreeWidgetItem *> |
20:45.16 | kaze | dcordes ? |
20:45.20 | ginge | wrong channel |
20:45.27 | kaze | about the screen? |
20:45.27 | druidu | basicly, with board-trout I can go until gpios are initialized which use proc_comm and make it go boom |
20:45.47 | druidu | but with board-halibut I still crash in kmem_cache_create |
20:46.08 | NetRipper | druidu, cr2 suggested to simply comment out the proc_comm for now, as it shouldn't be needed to get usb working |
20:46.25 | druidu | not sure about that flag I pasted ealier |
20:46.31 | kaze | screen is "a full-screen window manager that multiplexes a physical terminal between several processes" |
20:46.31 | druidu | usb_h2w |
20:46.58 | druidu | but we have a wonderful screen to print debug messages now :) |
20:47.14 | druidu | NetRipper: hope it works in Raphael, but I see no reason for it not to |
20:48.26 | druidu | also, .c should go in /drivers/android/ for the .diff I gave you, and you need to disable android ram console and enable framebuffer support + VGA FONT_8x16 (the mini font was SO tiny) |
20:48.37 | NetRipper | druidu, i'm going to try right away |
20:48.50 | NetRipper | druidu, ok |
20:49.34 | *** join/#htc-linux eXist (n=pimentel@yosemite.yosemite.edu) |
20:54.35 | druidu | btw, there's a param to add delay to each console write, it's configurable but it's not really accurate :) |
20:54.37 | NetRipper | druidu, it's working |
20:54.47 | druidu | good! :) |
20:54.49 | NetRipper | druidu, well lol yes i saw, i put it to 0 but it still has a huge delay |
20:54.50 | NetRipper | :p |
20:54.58 | NetRipper | like a second between each printf |
20:55.05 | druidu | msm_timer_init: setup_irq failed for dg_timer |
20:55.10 | NetRipper | druidu, but, best thing, it works |
20:55.10 | NetRipper | :) |
20:55.31 | NetRipper | hangs ont he proc_comm loop indeed |
20:55.33 | kaze | you post the fb pics on xda forum and i bet it goes crazy |
20:55.39 | druidu | still have no idea how I get stuff **before** I print the green message |
20:55.46 | NetRipper | i'll make a sharp pic, i'll let you post it druidu ;p |
20:55.47 | druidu | 'cause that is when I init the console |
20:55.59 | NetRipper | druidu, maybe it's in buffer already? |
20:56.10 | NetRipper | like for early console support |
20:56.16 | druidu | no, I clear the buffer |
20:56.28 | cr2_ | druidu: where does it come from : <druidu> msm_timer_init: setup_irq failed for dg_timer |
20:56.39 | druidu | somewhere in the boot log |
20:56.40 | cr2_ | the kernel messages are buffered |
20:56.59 | NetRipper | ah yes |
20:57.04 | druidu | cr2_ : I get that error |
20:57.07 | cr2_ | the only missing message is 'Decompressing kernel...............................' |
20:57.12 | NetRipper | that also explains why it shows 0.000000 on the first few messages |
20:57.14 | cr2_ | the famous 'dots' |
20:57.29 | cr2_ | druidu: git link ? |
20:57.38 | cr2_ | druidu: i don't have the tree locally |
20:57.40 | cr2_ | hm. wait |
20:57.40 | druidu | hehe, it's not buffered, the "decompressed" kernel never knows it was ever compressed :) |
20:57.57 | cr2_ | <PROTECTED> |
20:58.03 | cr2_ | it comes from putc() |
20:58.06 | cr2_ | in uncompress.h |
20:58.15 | cr2_ | which is disabled now. |
20:58.46 | cr2_ | it makes sense only if you have a serial/IR console |
20:59.27 | druidu | basicly, it's really easy to print anything, just write it raw to some buffer and start a dma by specifying physical address |
21:00.00 | *** join/#htc-linux szsoftwa1e (n=sven@e178207072.adsl.alicedsl.de) |
21:01.13 | druidu | NetRipper: anyway, post it on xda wiki if want, I don't have a lot of time to prepare it for a proper release and stuff :D |
21:01.36 | druidu | and maybe include the patch for irqs crash too |
21:01.54 | druidu | cr2_ : http://android.git.kernel.org/?p=kernel/msm.git;a=blob;f=arch/arm/mach-msm/timer.c;h=9e225da1ed0f5df53874a9884380e40c923a02ab;hb=android-msm-htc-2.6.25 |
21:02.13 | druidu | last lines |
21:02.18 | NetRipper | druidu, i'll make a patch against the android-msm-htc-2.6.25 |
21:02.48 | cr2_ | druidu: strange |
21:02.51 | NetRipper | druidu, what do i change to remove the delay? setting it to 0 doesn't make it go away :p |
21:02.55 | druidu | yeah, I use that one too |
21:03.20 | druidu | yeah, didn't get to test it enough, just look in the file and search for delay :P |
21:03.43 | druidu | the delay should delay by 1 milisecond, but I think it's off by a factor of 1000 or something... |
21:04.13 | druidu | I have a delay(1) after DMA, and one at the end of console_write |
21:04.59 | NetRipper | ok |
21:06.33 | cr2_ | druidu: the framebuffer is 18bpp ? |
21:06.57 | cr2_ | should be pita for X. |
21:07.04 | druidu | 16 |
21:07.05 | druidu | 565 |
21:07.10 | NetRipper | 2 bytes |
21:07.29 | cr2_ | /* 666 18BPP */ |
21:07.29 | cr2_ | dma2_cfg |= DMA_DSTC0G_6BITS | DMA_DSTC1B_6BITS | DMA_DSTC2R_6BITS; |
21:07.30 | druidu | 16 bits, 565 format... I think RGB (red being MSB) |
21:07.41 | druidu | no, that's internal format I think |
21:07.45 | druidu | I supply 16 bits |
21:07.49 | cr2_ | ok |
21:08.27 | druidu | anyway, framebuffer code from google is working 99% sure |
21:08.42 | druidu | only thing we need is panel power on/off and vsync |
21:08.42 | NetRipper | druidu, anyway why don't you use msleep() |
21:08.47 | NetRipper | for the delay() |
21:08.48 | druidu | I did try |
21:08.49 | druidu | it crashed |
21:08.51 | NetRipper | oh |
21:08.52 | druidu | dunno why... |
21:08.57 | NetRipper | well |
21:09.00 | druidu | you try it, maybe I did smth wrong |
21:09.01 | NetRipper | hm |
21:09.03 | cr2_ | mdelay ? |
21:09.03 | NetRipper | ok |
21:09.17 | cr2_ | msleep can be interrupted ? |
21:09.18 | *** join/#htc-linux BruteFrce (n=IrcNet@m120e36d0.tmodns.net) |
21:09.27 | druidu | but, probably jiffies wasn't initialized (calibrating delay loop is much further in the code) |
21:09.43 | BruteFrce | anyone looked at the android source yet? |
21:10.00 | cr2_ | druidu: what stage does the kernel reach now ? |
21:10.21 | cr2_ | NetRipper: fb penguin is enabled ? |
21:10.37 | NetRipper | haven't |
21:10.40 | kaze | bruteFrce even i wouldnt have asked such obvious thing |
21:10.42 | NetRipper | should i try? |
21:10.51 | NetRipper | ill try |
21:10.58 | NetRipper | the adventurous that i am |
21:11.50 | BruteFrce | lmao |
21:12.15 | *** join/#htc-linux sirox (n=sirox@91-67-142-232-dynip.superkabel.de) |
21:12.19 | BruteFrce | okay who has looked at it, nd how dificult does it look to adapt to oter devices? |
21:12.31 | ginge | BruteFrce: everyone, depends |
21:12.34 | cr2_ | NetRipper: it may crash :) |
21:12.52 | kaze | how difficult is it to adapt to correct syntax? |
21:12.53 | NetRipper | cr2_, yea wouldn't it need special support in driver? |
21:12.54 | NetRipper | :) |
21:13.01 | BruteFrce | does it just epend on the hardware drivers? |
21:13.13 | BruteFrce | deend |
21:13.22 | BruteFrce | blah im sick of this keypad |
21:13.26 | ginge | BruteFrce: depends on what part of android sources you are referring to for a start. Do you mean kernel or that java bobbins? |
21:13.48 | BruteFrce | the whole os |
21:14.19 | ginge | BruteFrce: not sure who here looked closely at it. We tend to work more with the kernel stuff in here. |
21:14.27 | BruteFrce | including its weird java xserver like thingy |
21:14.30 | ginge | #android has more userspace related chitchat |
21:15.35 | ginge | BruteFrce: so, the kernel stuff is nothing really amazing, and is quite specialised to the g1. Luckily we have been working quite close to the google source tree for a while, and we can get android booted on a few devices now. You can even boot on non msm chipsets |
21:16.30 | BruteFrce | really anything omap850? |
21:16.38 | szsoftware_ | ginge: what do you say about ti wlan driver from google? |
21:16.45 | ginge | sure |
21:16.52 | BruteFrce | i have ben using the linwizard kernel |
21:16.55 | druidu | cr2_: I see the tcp/ip stack init... |
21:16.59 | druidu | it's pretty much all the way |
21:17.03 | ginge | BruteFrce: http://forum.xda-developers.com/showthread.php?t=400900 |
21:17.15 | *** join/#htc-linux ssapphiree (n=sapphire@91.78.238.212) |
21:17.21 | ginge | szsoftware_: not looked at it yet. Shouldn't be anything too odd though, right? |
21:17.58 | szsoftware_ | ginge: would that dma issue prevent the driver from working? |
21:17.59 | cr2_ | druidu: all the way is crashing on mounting rootfs :) |
21:18.01 | ginge | BruteFrce: to boot android on most devices you only really need a supportting kernel. The patches are quite minimal |
21:18.04 | *** part/#htc-linux ssapphiree (n=sapphire@91.78.238.212) |
21:18.11 | ginge | szsoftware_: no, the wifi can use pio |
21:18.19 | druidu | cr2_: lol, I compiled it with yaffs2 support only |
21:18.24 | druidu | and gave it no initrd |
21:18.39 | druidu | probably no nand support anyway, it's all driver related |
21:18.40 | cr2_ | druidu: do you have cegcc ? |
21:18.40 | druidu | :) |
21:18.45 | szsoftware_ | ginge: did'nt google workaround that dma issue? |
21:18.50 | druidu | nope, why? |
21:18.54 | ginge | szsoftware_: not really |
21:18.54 | druidu | aaa, actually |
21:18.55 | druidu | I do |
21:18.59 | druidu | I did recompile haret on win |
21:18.59 | NetRipper | cr2_, no image, i enabled the images in kernel, should i assign one anywhere? |
21:19.01 | cr2_ | you can compile linload.exe |
21:19.09 | druidu | I know |
21:19.13 | cr2_ | ok |
21:19.16 | druidu | no need now I guess |
21:19.22 | druidu | it's far from package-ready |
21:19.23 | druidu | :) |
21:19.31 | cr2_ | NetRipper: complie the linloade.xe |
21:19.37 | ginge | szsoftware_: they didn't do any work around for us specifically. They (San) implemented the rest of the msm sdcc driver irq handlers and hardware status functions instead of relying on polling |
21:19.51 | ginge | szsoftware_: it doesn't solve the issue we have been seeing |
21:19.52 | *** join/#htc-linux ssapphiree (n=sapphire@91.78.238.212) |
21:19.57 | BruteFrce | ginge i cant view youtube videos on my phone |
21:19.58 | druidu | we could load up the dumped g1 image in a initrd with yaffs2 |
21:20.06 | ginge | BruteFrce: just read the title then |
21:20.08 | BruteFrce | well i can but its a huge hassle |
21:20.09 | druidu | and dunno, see if it boots with no drivers or something :) |
21:20.17 | ginge | BruteFrce: in a nutshell, yes ;) |
21:20.39 | BruteFrce | ginge so does it require some complicated toolchain to build? |
21:20.44 | *** part/#htc-linux ssapphiree (n=sapphire@91.78.238.212) |
21:20.54 | cr2_ | druidu: mtd driver needs offsets. |
21:21.20 | ginge | BruteFrce: couldn't tell you bud. It should just need gcc for the target. It pulls in most things you need anyway |
21:21.29 | NetRipper | druidu, http://www.netripper.nl/raphael/pics/ |
21:21.37 | NetRipper | also for the rest ;) |
21:21.46 | ginge | thats if you can get the repo script to connect to the repositories it killed o release :) |
21:21.52 | NetRipper | pretty neat :) |
21:21.57 | druidu | posted smth on xda forum |
21:22.04 | NetRipper | ok |
21:22.07 | cr2_ | NetRipper: add printk() to the clock api |
21:22.11 | ginge | druidu: awesome |
21:22.17 | BruteFrce | sweet, but you dont know anyone whos tryin to get the java gui workin? |
21:22.21 | cr2_ | so we will know where it crashes |
21:22.35 | ginge | BruteFrce: we can already boot the older android on our devices |
21:22.43 | BruteFrce | i hear its not xserver |
21:22.54 | ginge | BruteFrce: no, it uses its own stuff |
21:22.55 | BruteFrce | older android? |
21:22.59 | cr2_ | 64M |
21:23.07 | ginge | BruteFrce: beta releases |
21:23.32 | AstainHellbring | nice stuff there NetRipper |
21:23.46 | druidu | hehe, good quality pics :) |
21:24.53 | BruteFrce | and the radio stuff works outta the box? |
21:24.54 | ginge | BruteFrce: if you are interested in the user space apps, I recommend you talk to the guys over at #android |
21:25.05 | BruteFrce | no thanks |
21:25.09 | ginge | BruteFrce: assuming your kernel works properly, yeah |
21:25.17 | BruteFrce | this is where i always hang out |
21:25.25 | druidu | it should, firmware is almost the same for radio right? |
21:25.32 | szsoftware_ | dcordes: did you try to compile ti-wlan? |
21:25.36 | druidu | btw, there are 3 options for amms fw version |
21:25.42 | druidu | ls |
21:25.52 | ginge | BruteFrce: I will be less polite :) I know nothing at all about android other than what the kernel does ;) |
21:25.55 | *** join/#htc-linux datachaos (n=datachao@201.37.120.236) |
21:26.04 | druidu | cr2_: we gotta make proc_comm work, everything is based on that |
21:26.12 | dcordes | szsoftware_, no I'm having a slight look on smd then I will post it |
21:26.18 | druidu | don't really know what to disable so it's not used anymore |
21:26.28 | cr2_ | druidu: proc_comm itself works. look at vogue |
21:26.33 | ginge | druidu: just return at the top of proc_comm |
21:26.44 | druidu | tried, still endless loop I think |
21:26.53 | cr2_ | druidu: only in the clock api |
21:26.54 | dcordes | ginge, what fixed the smd1 again? was it only going back to the git entry-macro.S or also something else? |
21:26.59 | BruteFrce | ginge i have been adapting linwizard for my excalibur and have had some people help me wrote a few patches for my lcd and keymap, could i just apply those to the android kernel? |
21:27.01 | druidu | hhmm, let's see |
21:27.15 | ginge | dcordes: add the missing channel definition in the top of the function in smd_tty |
21:27.18 | cr2_ | druidu: check the vogue patch |
21:27.37 | ginge | BruteFrce: yeah. like I say, android kernel isnt too special |
21:27.47 | dcordes | ginge, only that? |
21:27.51 | ginge | yup |
21:27.52 | cr2_ | druidu: and you should return some dumb id, when the clock is queried |
21:28.12 | dcordes | ok so I will leave entry-macro |
21:28.14 | BruteFrce | except the radio stuff will work? cause it wnt with linwizard right now |
21:28.30 | ginge | BruteFrce: I think you answered your own question |
21:30.33 | *** join/#htc-linux Marnux (n=Marnux@83.101.56.192) |
21:30.38 | cr2_ | druidu: things like that -> |
21:30.42 | cr2_ | -clk_put(mddi->clk); |
21:30.43 | cr2_ | +//clk_put(mddi->clk); |
21:31.12 | BruteFrce | ginge no i dont get it.. |
21:31.25 | dcordes | if (n == 0) { |
21:31.26 | dcordes | name = "SMD_DS"; |
21:31.26 | dcordes | } else if (n == 27) { |
21:31.26 | dcordes | name = "SMD_GPSNMEA"; |
21:31.26 | dcordes | } else { |
21:31.26 | dcordes | return -ENODEV; |
21:31.29 | ginge | BruteFrce: if your linwizard kernel does not support radio, android app wont |
21:31.30 | dcordes | } |
21:31.32 | dcordes | ginge, where? :/ |
21:31.36 | ginge | dcordes: passsstebin |
21:31.47 | druidu | anyway, clock doesn't seem to stop the boot process |
21:31.52 | druidu | gives error and continues |
21:31.58 | druidu | proc_comm comes in later |
21:32.00 | dcordes | ginge, http://pastebin.ca/1234164 |
21:32.08 | kaze | bruteFrce if radio didnt work with linwizard it wont neither with android |
21:32.15 | ginge | okay, what channel number is the smd? 7 or something? |
21:32.37 | BruteFrce | ginge okay, but if i have a working kernel can i just switch with an android userspace? or do i need to be rnning there kernel? |
21:32.44 | BruteFrce | running |
21:32.46 | dcordes | ginge, AT=smd0, PPP=smd1, GPS=smd7 |
21:32.50 | dcordes | that's how it is in the git code |
21:33.11 | kaze | bruteFrce why dont you just try? |
21:33.17 | ginge | BruteFrce: their kernel is out kernel is everyones kernel. The android patches only do a few things, and android itself only needs a few extra kernel things. |
21:33.54 | dcordes | ginge, here I have smd0 smd1 and smd27 |
21:34.02 | ginge | so you can either add android patches to your kernel, or port your patche to android. Both are similar |
21:34.08 | dcordes | but trying to access smd1 I get no such device |
21:34.10 | BruteFrce | kaze becuase its alot of work in my situation or i would obviously |
21:34.29 | swetland | ginge: everone's kernel! |
21:34.39 | kaze | cant you just try your kernel with dcordes initrd for an example? |
21:35.10 | ginge | swetland: doh |
21:35.17 | kaze | you just need haret, zimage and android.bin |
21:35.23 | swetland | sounded good to me |
21:35.28 | dcordes | kaze, oh better use the linwizard one or so |
21:35.36 | ginge | dcordes: http://pastebin.ca/1234165 |
21:35.39 | swetland | we need to get some examples out there of how to create build targets for other devices for android. |
21:35.45 | swetland | I'll try to do that this afternoon |
21:35.45 | kaze | yup* |
21:35.56 | dcordes | thx |
21:35.57 | cr2_ | druidu: comment out goto again; |
21:35.59 | kaze | i meant linwizard kernel with your initrd |
21:36.02 | ginge | swetland: sounds like a good plan |
21:36.09 | druidu | ditto |
21:36.09 | druidu | :) |
21:36.11 | swetland | right now everyone is building the android system for "generic" |
21:36.15 | ginge | swetland: just a list of minimum deps would be good |
21:36.23 | druidu | I suspect parent function keeps asking |
21:36.25 | BruteFrce | android.bin ? is that a userspace? |
21:36.33 | druidu | lemme try some more |
21:36.36 | swetland | but you can drop in vendor/qct/surf/ with a handful of files to allow you to build for surf, for example |
21:37.59 | *** join/#htc-linux data|2 (n=datachao@201.21.247.155) |
21:38.13 | ginge | dcordes: I added channel 1 (smd1) to that if check. Thats all that was missing there. I gave it a dumb name, you might want to change that |
21:38.15 | cr2_ | druidu: return success here -> |
21:38.18 | cr2_ | again: |
21:38.18 | cr2_ | <PROTECTED> |
21:38.18 | cr2_ | <PROTECTED> |
21:39.03 | NetRipper | cr2_, as workaround you mean? |
21:39.21 | cr2_ | NetRipper: yes, we don't need working proc_comm anyway |
21:39.25 | NetRipper | ok |
21:39.27 | dcordes | druidu, sorry for not reading log but did you get further than proc_comm init? commented the proc comm init out or so? |
21:39.38 | dcordes | ginge, building |
21:39.39 | cr2_ | NetRipper: for booting |
21:39.41 | NetRipper | cr2_, it doesn't return at all anyway |
21:39.59 | cr2_ | NetRipper: to avoid endless return loops |
21:39.59 | NetRipper | cr2_, it loops within the proc_comm_wait_for |
21:40.33 | NetRipper | yes |
21:40.50 | cr2_ | NetRipper: in the main proc_comm routine |
21:41.03 | dcordes | BruteFrce, yes a very bad one. I ought to replace that with a working and more complete angstrom-console image |
21:41.05 | *** join/#htc-linux miknix (n=miknix@gentoo/developer/miknix) |
21:41.31 | druidu | didn't get to do it yet, gonna try it soon |
21:41.36 | druidu | dcordes ^^ |
21:41.46 | NetRipper | cr2_, it doesn't return to the main proc_comm routine |
21:41.58 | NetRipper | i hacked it in the subroutine now |
21:42.01 | NetRipper | i'll test |
21:42.04 | cr2_ | ok |
21:42.11 | dcordes | BruteFrce, you might try the linwizard initrd |
21:42.33 | druidu | wouldn't it be more fruitful if we tried to make it work, instead of disabling it? :) |
21:42.37 | BruteFrce | thats what i use |
21:42.46 | druidu | do we know what's wrong? maybe just some addr or smth |
21:43.03 | ginge | druidu: do you guys have your smem mapped? |
21:43.05 | dcordes | swetland, how do I add the tiwlan driver to my kernel build? |
21:43.09 | BruteFrce | i want to try the android gui with my kernel |
21:43.13 | dcordes | swetland, in trout branch |
21:43.15 | cr2_ | druidu: it it a clock proc_comm or some other ? |
21:43.37 | druidu | ginge : smem? as in the memory for framebuffer or what? |
21:44.00 | druidu | gpios I think |
21:44.01 | ginge | the shared memory between the arm chips |
21:44.10 | druidu | look in screenshot from NetRipper |
21:44.22 | druidu | probably |
21:44.31 | druidu | arm msm already did that |
21:44.35 | BabelO | cr2_: around ? |
21:44.53 | ginge | druidu: doesn't mean it is right |
21:45.08 | dcordes | NetRipper, wow the vga resolution looks mind blowing |
21:45.10 | druidu | 0x01F00000 => 0xE0100000 |
21:45.15 | NetRipper | boots quite abit |
21:45.16 | dcordes | I have 3x6 font on the kaiser qvga screen :( |
21:45.18 | druidu | it is I think |
21:45.26 | NetRipper | lol this is 8x16 |
21:45.31 | druidu | hahhaha, tried the 4x4 tiny font but wasn't able to read it |
21:46.26 | druidu | so I switched to 8x16 |
21:46.26 | dcordes | ginge, works. |
21:46.27 | dcordes | ginge, now preparing patch |
21:46.27 | druidu | it's a crime for the eyes to put a smaller one |
21:46.27 | ginge | dcordes: :) |
21:46.27 | druidu | what patch? |
21:46.27 | dcordes | druidu, rebase of kaiser code to trout branch |
21:46.29 | szsoftware_ | swetland: I see, you deliver a own build-system. Is this a derivation of one known or a totally new one? |
21:46.30 | NetRipper | ouch |
21:46.41 | NetRipper | it hits msmfb_probe() |
21:46.46 | NetRipper | should i disable that one? |
21:46.51 | NetRipper | it probably conflicts |
21:47.04 | druidu | nope, it has no panels, shouldn't |
21:47.13 | druidu | unless the trout-panel is added |
21:47.17 | dcordes | szsoftware_, google don't have the words openembedded openmoko angstrom etc in their dictionary file ^^ |
21:47.21 | NetRipper | msmfb_probe() installing 480x640 panel |
21:47.28 | NetRipper | oh |
21:47.30 | NetRipper | screen went black now |
21:47.38 | *** join/#htc-linux datachaos (n=datachao@201.21.247.155) |
21:47.38 | druidu | anyway, I say it again -- if we disable power on/off code for panel |
21:47.45 | druidu | we could use all the fb code |
21:47.50 | *** join/#htc-linux exco (n=exco@e181105133.adsl.alicedsl.de) |
21:47.55 | dcordes | szsoftware_, remember, it's the _first_ open mobile platform |
21:47.58 | druidu | the mddi/mdp works perfectly |
21:48.20 | NetRipper | there's garbage on the screen |
21:48.29 | druidu | when? |
21:48.33 | druidu | after fb enable? |
21:48.38 | NetRipper | yes |
21:49.00 | druidu | they use different addresses |
21:49.02 | NetRipper | device doesn't hang though |
21:49.07 | NetRipper | it updates _something_ |
21:49.11 | NetRipper | but just generates red and green pixels |
21:49.16 | NetRipper | on random locations |
21:49.18 | druidu | I used wince physical and mapped it to some random memory space like 0xE0... |
21:50.07 | szsoftware_ | dcordes: ok, these days we won't be worry about incompatibilities, no? :-) |
21:50.14 | NetRipper | oh i got screen back |
21:50.32 | szsoftware_ | dcordes: I just install their build-system, because ti-wlan seems to be delivered for that |
21:50.33 | NetRipper | nah the two drivers are definitely conflicting |
21:50.35 | NetRipper | somehow |
21:50.46 | dcordes | szsoftware_, ah yea I also noticed the android makefile thing |
21:50.57 | dcordes | in the tiwlan branch root |
21:51.02 | szsoftware_ | dcordes: and I am too lazy and too dumb, to strip it down from their metadata |
21:51.28 | szsoftware_ | dcordes: it's a damn weird directory structure ^^ |
21:51.37 | NetRipper | msmfb_start_dma 891.937 ms after vsync request |
21:51.38 | NetRipper | can't be good |
21:51.39 | NetRipper | :P |
21:51.40 | *** join/#htc-linux BruteFrce (n=IrcNet@m120e36d0.tmodns.net) |
21:51.58 | BruteFrce | ali1234 you alive? |
21:52.03 | druidu | NetRipper : basicly, my driver only interacts with lcd controller via the update function, in which it schedules a dma transfer |
21:52.05 | ali1234 | no i'm dead |
21:52.20 | druidu | probably because of my delay :) |
21:52.22 | NetRipper | druidu, what's msmfb supposed to do? |
21:52.34 | *** join/#htc-linux Zinbolic (i=Zinbolic@84.238.80.225) |
21:52.36 | NetRipper | druidu, i made your delay 100 times less |
21:52.37 | druidu | it's a framebuffer, like vesafb for pc |
21:52.39 | NetRipper | so its only about 10 ms now |
21:52.50 | druidu | it has hardware acceleration too |
21:52.56 | NetRipper | oi |
21:52.56 | druidu | scaling, paning etc |
21:53.04 | NetRipper | windows found usb hardware |
21:53.19 | NetRipper | unidentified |
21:53.20 | NetRipper | bah |
21:53.24 | NetRipper | not even usbnet |
21:53.30 | druidu | it's bulletproof, we only need to make the client side -> power on, power off, vsync, backlight |
21:53.44 | NetRipper | off -> on |
21:53.46 | NetRipper | i hope |
21:53.50 | BruteFrce | ali1234 could you build a linwizard kernel with the excalibur patches and the android patches, i want to try the android userspace |
21:54.28 | ali1234 | what android patches? |
21:54.39 | szsoftware_ | dcordes: so, maybe it's a occasion, building a current android-roofs or initrd for us.. |
21:55.12 | dcordes | szsoftware_, you will probably get a thousand dollars if you do |
21:55.23 | druidu | at some time we're gonna have to implement everything, especially good power management, otherwise it would be unusable |
21:55.36 | NetRipper | druidu, yes |
21:55.44 | ginge | ali1234: omap? |
21:55.47 | BruteFrce | someone said to run the android userspace i would need tp patch the android kernel with the excalibur patches or build my kernel with the android patches |
21:55.52 | NetRipper | still would be a nice gadget to show off |
21:55.52 | NetRipper | ;) |
21:56.25 | ginge | ali1234: BruteFrce: http://elinux.org/Android_on_OMAP |
21:57.14 | BruteFrce | is that where the patces are? |
21:57.23 | BruteFrce | patches |
21:57.30 | ginge | BruteFrce: ther is no one patch source afaik |
21:57.38 | ginge | BruteFrce: too many platforms |
21:58.05 | ginge | BruteFrce: that page is about the best resource for collated information on what you would need to do to get it running on an omap |
21:58.42 | BruteFrce | im just going by what someone in here said, they said to apply android patches to my kernel |
21:58.55 | dcordes | http://linuxtogo.org/~lgorris/git-diff-android-msm-htc-2.6.25_kaiser_20082210.diff |
21:59.02 | BruteFrce | why would they say that if there is no patches? |
21:59.23 | ginge | BruteFrce: short memory. I said it above! |
21:59.28 | kaze | there is not one source of paatch |
21:59.34 | szsoftware_ | dcordes: Thank you! |
21:59.35 | ginge | http://elinux.org/Android_on_OMAP#Patch_extraction |
21:59.36 | kaze | there are multiples |
21:59.45 | BruteFrce | what are you talking bout |
21:59.47 | ginge | ie you need to extract the patches, and rewrite for your platform |
22:00.29 | ginge | BruteFrce: "(22:34:37) ginge: so you can either add android patches to your kernel, or port your patche to android. Both are similar" |
22:00.30 | BruteFrce | if smeone has already got it working on omap850 then there is a patch |
22:00.43 | ginge | see above URL |
22:00.59 | NetRipper | <PROTECTED> |
22:01.05 | NetRipper | what .config option am i missing? |
22:01.08 | BruteFrce | no because i have to anually write ith a pen then type i into opera |
22:01.46 | ginge | BruteFrce: okay write that url down and look later. It basically has all you need on it. |
22:01.59 | BruteFrce | i have no pc |
22:02.09 | ginge | BruteFrce: how do you expect to compile up a kernel? |
22:02.10 | *** join/#htc-linux scatterp (n=scatterp@host81-153-90-18.range81-153.btcentralplus.com) |
22:02.22 | kaze | ginge, i have no word for how big your patience is |
22:02.33 | ginge | kaze: believe me it is failing |
22:02.35 | BruteFrce | i thought it was as simple as applying a patch to my kernel and replacing my usersapce because you said it was |
22:02.58 | dcordes | how do I find out current git revision of the branch I'm in? |
22:03.06 | BruteFrce | but i was misinformed so nevermind, thanks for your help |
22:03.27 | ginge | not sure about misinformed... more misunderstandings here |
22:03.41 | BruteFrce | no |
22:03.45 | kaze | haha |
22:03.56 | ginge | /gives up |
22:04.03 | kaze | rofl |
22:04.17 | BruteFrce | you didnt say build a patch then apply it, you just said apply the patch, leading me to elieve it already existed |
22:04.27 | BruteFrce | believe |
22:04.32 | kaze | gf is wandering at me because i laugh at the ccomp now i have to explain ... |
22:04.52 | ginge | BruteFrce: "(22:34:37) ginge: so you can either add android patches to your kernel, or port your patche to android. Both are similar" plural. also when referring to kernel code a patch is generally source |
22:05.00 | BruteFrce | tell her its because of your ignorance and moronity |
22:05.14 | ginge | hey no need for that |
22:05.36 | BruteFrce | theres no need for his trollingeithr |
22:05.44 | BruteFrce | trolling either |
22:06.07 | szsoftware_ | dcordes: haha I askes me that already, dunno if it's right or wrong: git-rev-parse --tags | tail -n1 ? |
22:07.06 | druidu | I get to msm_i2c_probe now |
22:07.07 | kaze | im not quite sure im the one trolling there ;\ |
22:07.13 | ginge | grins |
22:07.29 | dcordes | szsoftware_, thx just stated the date. |
22:07.52 | BruteFrce | i dont think anyone values your opion ;u |
22:08.34 | BruteFrce | anyways i apologize to the channel for the disruption |
22:08.38 | cr2_ | NetRipper: g1 panel is not vga |
22:08.54 | szsoftware_ | kaze: BruteFrce I like this sort of forum/chat because we are polite and patient. This is in contrast to many windows-, filesharing-, or mobilephone-usage-related sites |
22:09.02 | NetRipper | cr2_, that might explain yea, but it does detect 480x640 |
22:09.08 | dcordes | szsoftware_, so you gonna put the polaris framebuffer geometry hacks and on screen keyboard? |
22:09.17 | szsoftware_ | kaze: BruteFrce please calm down :-) |
22:09.44 | szsoftware_ | dcordes: how did you stated the date? |
22:09.47 | BruteFrce | shhh |
22:09.50 | kaze | szsoftware_ no problemeven tried to help him through pms |
22:09.52 | BruteFrce | just drop it |
22:10.01 | dcordes | szsoftware_, I mean I said which date the revision was pulled |
22:10.06 | cr2_ | NetRipper: ok, then you need to do this -> http://wiki.xda-developers.com/index.php?pagename=RaphaelLCD |
22:10.35 | BruteFrce | lol ya he helped me by insulting and cursing at me |
22:10.51 | druidu | ok, got to the msmfb init too, also garbage on the screen |
22:10.54 | szsoftware_ | dcordes: I dont' know ^^, this git-stuff is a jungle ... |
22:11.10 | druidu | there are lots of calls to proc_comm, with no proc_comm nothing will work |
22:11.26 | dcordes | szsoftware_, nevermind. I will look into adding polaris |
22:11.29 | szsoftware_ | dcordes: sorry about the fb hack and so on, but the code was too complicated for me |
22:11.33 | cr2_ | NetRipper: btw, you can add reading the toshiba gpios, and detect the panel id |
22:11.45 | dcordes | szsoftware_, nono I mean just the present code (which is also in the tar) |
22:12.07 | cr2_ | druidu: proc_comm is power, clocks and phone |
22:12.19 | *** part/#htc-linux BruteFrce (n=IrcNet@m120e36d0.tmodns.net) |
22:12.21 | dcordes | szsoftware_, the diff only has kaiser |
22:12.29 | cr2_ | druidu: unless you suspend it should not be a problem |
22:12.49 | NetRipper | cr2_, if i knew how.. |
22:12.54 | cr2_ | druidu: btw, i see the difference in the SMD channel offsets |
22:12.54 | dcordes | ginge, look at you. now you hurt his feelings |
22:12.56 | szsoftware_ | dcordes: ah ok. But the onscreen kb never worked in 2.6.25 |
22:13.17 | dcordes | szsoftware_, ok sorry for the confusion then |
22:13.19 | cr2_ | NetRipper: well, i guess it's a minor problem now ? |
22:13.25 | ginge | dcordes: don't make me feel bad. I already feel kind of bad because I couldn't really help him. |
22:14.25 | dcordes | ginge, nvm ^^ do you have the trout branch? |
22:14.31 | szsoftware_ | dcordes: I will have a try to change polaris stuff in your patch, according to ginges latest diff, right? |
22:14.41 | ginge | dcordes: yeah |
22:14.47 | ginge | just looking through that patch now |
22:15.04 | cr2_ | NetRipper: the raphaelLCD is not 100% perfect too. the vreg* stuff needs to be sorted out (it's proc_comm), and some hitachi code is missing too |
22:15.35 | *** join/#htc-linux ssapphiree (n=sapphire@ppp91-77-157-251.pppoe.mtu-net.ru) |
22:16.10 | dcordes | szsoftware_, yep it's almost 1:1, just with the trouts swimming everywhere |
22:16.14 | druidu | heeeeeeeeeeeeeey! |
22:16.21 | cr2_ | :) |
22:16.23 | druidu | check out board-trout-panel.c |
22:16.27 | druidu | they have sharp panels there! |
22:16.36 | druidu | static struct mddi_table mddi_sharp_init_table[] = { |
22:16.41 | cr2_ | druidu: they have all knd of panels there |
22:17.05 | druidu | maybe adjust the id? they have id 0 for sharp |
22:17.21 | druidu | maybe a different kind of id for same panel? |
22:17.22 | cr2_ | druidu: how do they detetc the id ? |
22:17.31 | druidu | <PROTECTED> |
22:17.49 | *** part/#htc-linux ssapphiree (n=sapphire@ppp91-77-157-251.pppoe.mtu-net.ru) |
22:17.54 | cr2_ | druidu: ok. is it the same as RapahelLCD ? |
22:18.09 | cr2_ | if bit 0x10 is not set -> id+=1 |
22:18.09 | cr2_ | if bit 0x04 is not set -> id+=2 |
22:18.13 | ginge | dcordes: that sdcc clocks a bit high, did that come from one of my patches? |
22:19.08 | tmzt | cr2_: what is this? < WyrM> you need DSS and EDSS values, at the time that wince sets SSE. |
22:19.32 | cr2_ | tmzt: it's SCCR0 register |
22:19.47 | cr2_ | tmzt: haret-w.exe crashed you you ? |
22:19.58 | druidu | trout_init_panel uses vreg which use proc_comm... |
22:20.05 | cr2_ | tmzt: then get gnu-haret and try 'dump pxa27xssp' |
22:20.37 | cr2_ | druidu: you can assume that the vreg is set by wince. for now |
22:20.42 | dcordes | ginge, sorry forgot to add that I reviewed what I did and sd works now. it is your exact cod from http://headfuzz.co.uk/files/android/git-diff-2.6.25-170908.diff yep |
22:20.45 | miknix | hey BabelO |
22:21.11 | cr2_ | druidu: so comment the call out |
22:21.23 | ginge | dcordes: ok. its more like to crash at those high speeds |
22:21.35 | BabelO | ola miknix |
22:21.41 | tmzt | he, left, theres good news too? |
22:21.46 | tmzt | BrtFrc |
22:21.46 | cr2_ | BabelO: lokking for me too ? |
22:22.26 | miknix | BabelO, I just found a nice resource for a tty driver example |
22:22.47 | miknix | BabelO, have a look at drivers/serial/jsm/jsm_driver.c |
22:22.47 | tmzt | cr2_: where is recent gnuharet, I only have koc's and haret-w |
22:23.01 | BabelO | cr2_: no, just to tell you that M works on wince now, just got a pb displaying the map... don't know if path is wrong or .. i m debugging |
22:23.01 | miknix | I totally miss it in the last time |
22:23.36 | BabelO | miknix: ok |
22:23.52 | dcordes | ginge, did you boot with the patch? |
22:23.52 | miknix | BabelO, can you have a look at drivers/serial/jsm/jsm_driver.c:450 ? |
22:24.26 | miknix | wrong file, sorry |
22:24.26 | ginge | dcordes: not yet. my phone is flat again |
22:24.30 | miknix | BabelO, drivers/serial/jsm/jsm_tty.c:450 |
22:24.39 | BabelO | miknix: on GIT it is end of file for me |
22:24.45 | ginge | dcordes: I need to buy a new battery. The faulty charging on the whole killed it |
22:24.56 | dcordes | ginge, misinterpreted the logs then. somehow thought you just got it back |
22:24.59 | BabelO | miknix: ok i see |
22:25.20 | miknix | brd->channels[i]->uart_port.irq = brd->irq; |
22:25.20 | miknix | brd->channels[i]->uart_port.uartclk = 14745600; |
22:25.22 | miknix | (..) ? |
22:25.45 | ginge | dcordes: I did just get it back. You keep your battery when it goes for repair. The battery lasts about 4 hours before discharging. It was related to the charging fault it was repaired for. |
22:26.14 | ginge | dcordes: I can't use the phone and charge at the same time, it just wont supply enough juice |
22:26.16 | BabelO | miknix: yes it seem to be that, like BA driver |
22:26.46 | dcordes | ginge, ah flat=platt ^^ |
22:26.52 | NetRipper | cr2_, boots quite far actually |
22:27.02 | NetRipper | cr2_, i temporarily disabled proc_comm completely |
22:27.04 | cr2_ | druidu: i check the bits, the panel id algorithm is different |
22:27.10 | cr2_ | NetRipper: ok. |
22:27.14 | miknix | BabelO, so, as far as I understood. we have to map the omap dsp fifo to a uart like in that example right? |
22:27.17 | ali1234 | ginge: i recommend shopblue.co.uk, cheap battery direct from china :) |
22:27.19 | cr2_ | NetRipper: so where does it fail now ? |
22:27.23 | dcordes | ginge, you sent it directly to htc? or got a branded? |
22:27.24 | NetRipper | cr2_, doesn't get to usb gadget |
22:27.30 | ginge | ali1234: cheers |
22:27.33 | kaze | dcordes by any chance you need a guinea pig on kaiser? |
22:27.33 | NetRipper | cr2_, last lines are PPP BSD compression module registered |
22:27.36 | ginge | dcordes: branded. Orange dealt with it |
22:27.43 | dcordes | kaze, yes |
22:27.46 | BabelO | miknix: yes like that |
22:27.48 | cr2_ | NetRipper: and then hangs ? |
22:27.53 | NetRipper | yes |
22:27.56 | kaze | im there then |
22:27.57 | NetRipper | cr2_, but im not sure if it hangs |
22:28.02 | NetRipper | cr2_, or in some endless loop |
22:28.13 | dcordes | kaze, http://linuxtogo.org/~lgorris/kaiser-bootkit/README.txt |
22:28.17 | miknix | BabelO, thanks, I was getting confused. cause my last approach was using a virtual tty which doesnt need a uart at all |
22:28.21 | cr2_ | NetRipper: but no oops ? |
22:28.25 | NetRipper | cr2_, no oops |
22:28.32 | cr2_ | NetRipper: the usb id does not change ? |
22:28.43 | NetRipper | cr2_, let me check in linux |
22:28.45 | NetRipper | i mean |
22:28.46 | NetRipper | windows sucks |
22:28.50 | cr2_ | LOL |
22:28.51 | ginge | lol |
22:28.54 | miknix | ali1234, I think that file is a good source of information.. |
22:29.28 | NetRipper | cr2_ |
22:29.28 | NetRipper | [4431415.333723] usb 5-6: new high speed USB device using ehci_hcd and address 120 |
22:29.30 | dcordes | kaze, would be nice if you could test booting, building, perhaps gprs |
22:29.32 | NetRipper | [4431415.618612] usb 5-6: configuration #1 chosen from 1 choice |
22:29.34 | kaze | dcordes i suppose you mean the new zimage in root folder not the readme? |
22:29.35 | NetRipper | but no id |
22:29.55 | dcordes | kaze, well check the BREAKING NEWS |
22:30.20 | szsoftware_ | dcordes: /mydroid$ make :-) will surely wait years for the whole build... |
22:30.28 | ali1234 | miknix: uart subsystems handles many different kind of uart/fifo through drivers... we have to either use one of those or define a new type. then the standard tty stuff will "just work" |
22:30.36 | dcordes | szsoftware_, dude we only need the acx |
22:30.38 | cr2_ | NetRipper: lsusb ? |
22:31.03 | cr2_ | druidu: MULTICABLE_3_5_EARPHONE ? |
22:31.18 | druidu | what's with that? usb? |
22:31.21 | dcordes | kaze, would be ace if you could do a manual build and then try to boot that |
22:31.22 | szsoftware_ | dcordes: what do you mean, we only need the acx? |
22:31.25 | NetRipper | cr2_, Bus 005 Device 121: ID 18d1:d00d |
22:31.27 | dcordes | szsoftware_, wifi |
22:31.29 | NetRipper | d00d? |
22:32.04 | druidu | can we get a shell running? :) |
22:32.06 | szsoftware_ | dcordes: ah, because the whole build? |
22:32.32 | cr2_ | NetRipper: 18d1 is qualcomm or google or what ? |
22:32.44 | NetRipper | no idea yet |
22:32.47 | druidu | 0BB4 is what windows shows me now |
22:32.50 | cr2_ | NetRipper: you may try lsusb -vvvvv |
22:32.51 | druidu | for mass storage device |
22:32.59 | NetRipper | google doesn't know |
22:33.00 | cr2_ | 0bb4 is htc |
22:33.08 | *** join/#htc-linux Zinbolic (i=Zy@84.238.80.225) |
22:33.22 | cr2_ | the first number is vendor, second is device |
22:33.28 | dcordes | szsoftware_, yes |
22:33.39 | cr2_ | <PROTECTED> |
22:33.53 | dcordes | NetRipper, lol raphael is the d00dphone? |
22:34.05 | cr2_ | yeah ;) |
22:34.13 | ginge | lol |
22:34.31 | NetRipper | cr2_, http://www.netripper.nl/raphael/lsusb.txt |
22:34.35 | kaze | dcordes kernel loads sees sd |
22:34.39 | cr2_ | what do androids set as usb vendor ? |
22:34.56 | dcordes | kaze, do you boot android.bin? |
22:34.59 | cr2_ | usb.ids has nothing for 18d1 |
22:35.04 | szsoftware_ | dcordes: ok, will learn how to compile parts of android only. Must have a closer look to android build system... |
22:35.06 | kaze | dcordes yup |
22:35.08 | ginge | dcordes: the patch against my orig tree looks quite small. good work |
22:35.10 | NetRipper | cr2_, is that id hardware or software? |
22:35.14 | kaze | booted |
22:35.16 | cr2_ | software |
22:35.17 | druidu | oard-halibut.c: .vendor_id = 0x18d1, |
22:35.18 | dcordes | kaze, cool. try calling |
22:35.19 | NetRipper | cr2_, if it's software we shoudl be able to see it in kernel code? |
22:35.24 | druidu | *board |
22:35.26 | cr2_ | yes |
22:35.30 | cr2_ | you can change it |
22:35.31 | NetRipper | druidu, ah you found it |
22:35.36 | cr2_ | to match pxa27x |
22:35.40 | druidu | you running trout or halibut? |
22:35.41 | szsoftware_ | dcordes: android-build system, don't they even have a cool name for that? ^^ |
22:35.43 | NetRipper | trout |
22:35.47 | druidu | mine doesn't boot with halibut |
22:35.53 | druidu | so how come you have halibut usb id? |
22:35.56 | kaze | calling works ! |
22:36.07 | cr2_ | druidu: it#s vendor id |
22:36.13 | druidu | yes |
22:36.15 | ginge | kaze: check data too |
22:36.21 | cr2_ | and you can set any junk you like |
22:36.37 | cr2_ | what does pxa27x_udc have ? |
22:36.38 | dcordes | ginge, it was easy. only context changed, not content. it's really 1:1 your diff, just fiddled it |
22:36.40 | dcordes | in |
22:36.58 | ginge | dcordes: better working off the trout branch though |
22:36.59 | kaze | dcordes i need to setup gprs link firt? |
22:36.59 | dcordes | kaze, nice |
22:37.01 | druidu | check out msm_hsusb_pdata in board-halibut |
22:37.01 | kaze | first |
22:37.17 | cr2_ | NetRipper: 0xff vendor specific |
22:37.34 | druidu | well, gonna go sleep a bit, good luck with the hacking... can't wait to get rid of wince for good :P |
22:37.37 | cr2_ | NetRipper: so your linux (desktop) does not know which driver to load |
22:37.39 | dcordes | kaze, don't really know about gprs in android on the kaiser. never tried |
22:37.48 | dcordes | kaze, only in the angstrom console image. there it works fine |
22:37.51 | NetRipper | cr2_, ok i think you can force id's on modprobe usbnet |
22:38.04 | kaze | incoming call works |
22:38.09 | ginge | kaze: just load the browser and see what happens |
22:38.15 | dcordes | szsoftware_, fancy handing your FBI image so I can construct a nice initrd and sd tar from it? |
22:38.16 | cr2_ | NetRipper: try to force the modprobe usb-serial vendor=foo dev=bar |
22:38.20 | kaze | voice dcordes unkown host |
22:38.29 | NetRipper | yes ok |
22:38.31 | NetRipper | let me get my laptop |
22:38.34 | dcordes | kaze, huh? |
22:38.39 | cr2_ | NetRipper: modinfo usb-serial |
22:38.42 | kaze | - voice |
22:38.49 | ginge | kaze: oh well. need to set up gprs apn |
22:38.56 | kaze | browser ways unknown host |
22:39.04 | kaze | oh yeah |
22:39.05 | dcordes | aah |
22:39.14 | kaze | i need to run script in angstrom? |
22:39.36 | dcordes | ginge, right, android.bin reads the apn user and pass from cmdline? |
22:39.36 | ginge | are you using the image I hacked to acceppt kernel apn info? |
22:39.36 | ginge | dcordes: |
22:39.38 | cr2_ | NetRipper: complie the linload.exe |
22:39.44 | dcordes | lol |
22:39.49 | cr2_ | NetRipper: and enable the ramdisk & friends |
22:40.01 | dcordes | ginge, I'm not sure but I think so. because some people in the kaiser thread report they go online with it |
22:40.11 | szsoftware_ | dcordes: you wanna have the ipk's of fbi ? |
22:40.15 | ginge | dcordes: yeah you have to pass the username password and apn in |
22:40.18 | dcordes | szsoftware_, nope. just your image |
22:40.21 | cr2_ | dcordes: does it support multitasking ? or it's like iephone ? |
22:40.27 | ginge | tries to remember |
22:40.28 | kaze | devtools hangs |
22:40.28 | dcordes | szsoftware_, with it installed. and/or other fancy stuff |
22:40.45 | NetRipper | cr2_, you want with or without long delays between each printf? |
22:40.46 | szsoftware_ | hmm ok, will email that |
22:40.51 | dcordes | cr2_, android browser? |
22:40.56 | szsoftware_ | dcordes: ok, will email that |
22:41.15 | kaze | soft reset |
22:41.17 | ginge | okay... ppp.apn=youapn ppp.user=gprsuser ppp.pass=password I think |
22:41.18 | dcordes | szsoftware_, thanks. it's only console, right? |
22:41.21 | cr2_ | 3 and 5 are GND says the xda-dev pic |
22:41.38 | cr2_ | NetRipper: without the delay, if it will not crash it |
22:41.41 | ginge | you need to get your apn address for your carrier |
22:41.55 | NetRipper | ginge, if you dont pass that in command line, will it hang on loading PPP module? |
22:42.03 | NetRipper | ginge, or is it only to get radio functional? |
22:42.34 | ginge | NetRipper: only to get radio functional. it just sends them to the baseband processor via the uart |
22:42.42 | ginge | username and password are used with ppp |
22:42.52 | NetRipper | ginge, ok, just asking as it currently hangs at PPP |
22:42.53 | NetRipper | :P |
22:43.00 | NetRipper | coincidental |
22:43.02 | ginge | heh :) |
22:43.18 | dcordes | ginge, excuse my bad memory. did you ever got X11 working? I remember vaguely that you once mentioned you did X11 in the gizard image. |
22:43.38 | ginge | dcordes: not as such, no. I got it to load, but it was pretty messed up |
22:43.51 | ginge | so I went back to the kernel where I belong :) |
22:44.08 | dcordes | ginge, ok can you detail messed up? |
22:44.11 | ginge | I treat X11 like a magical black box |
22:44.13 | szsoftware_ | dcordes: its only angstrom -console, yes |
22:44.41 | ginge | dcordes: top half was sort of working but had image colour corruption. Bottom half had horizontal lines running down it. |
22:44.59 | cr2_ | NetRipper: remove PPP & friends |
22:45.05 | ginge | framebuffer is screwed somewhere |
22:46.02 | kaze | i find the new kernel to be way faster |
22:46.21 | cr2_ | kaze: lol |
22:46.21 | miknix | acx? working on acx? |
22:46.21 | ginge | that makes no sense |
22:46.47 | kaze | cr2_ just an impression? |
22:46.53 | dcordes | ginge, I wonder why on earth kaiser displays android perfectly and on the raw fb console it can draw .jpg files and play videos with mplayer and X11 just refuses to work |
22:46.58 | cr2_ | miknix: where does tiwlan enable power/irq ? |
22:47.13 | NetRipper | cr2_, when i have ttyUSB0 can i just cat it? |
22:47.16 | ginge | dcordes: no idea bud. Dont know a lot about X :( |
22:47.19 | cr2_ | dcordes: some 18bit mess ? |
22:47.20 | dcordes | ginge, was the picture frozen or only messed up? |
22:47.24 | cr2_ | NetRipper: yes |
22:47.30 | ginge | dcordes: only messed up |
22:47.33 | cr2_ | NetRipper: cu/minicom |
22:47.35 | dcordes | cr2_, I have no clue. what 18 bits? |
22:47.44 | cr2_ | dcordes: 666 setup |
22:47.55 | dcordes | ginge, with the angstrom X11, it also seems to freeze in the very first frame |
22:48.18 | cr2_ | dcordes: it may suffer from the A register setup |
22:48.19 | dcordes | cr2_, ? :( |
22:48.41 | miknix | cr2_, I tried dumping that but I got nothing.. or either the gpios were masked or the power is controlled over something else |
22:48.45 | cr2_ | dcordes: only swetland can help here. |
22:49.07 | cr2_ | dcordes: or the direct register dumping |
22:49.12 | cr2_ | miknix: wiki ? |
22:49.15 | dcordes | cr2_, which registers? |
22:49.28 | cr2_ | dcordes: mddi/mdp |
22:49.43 | cr2_ | dcordes: you use msm_fb ? |
22:49.48 | szsoftware_ | dcordes: may I email the 12MB rootfs.tar.gz now? |
22:49.49 | dcordes | cr2_, yes |
22:50.06 | ali1234 | dcordes: maybe your X has the wrong colour depth or something? |
22:50.18 | miknix | cr2_, http://pastebin.com/d2c573bba <-- for wifi on |
22:50.27 | dcordes | szsoftware_, yes you could also use instant messenger or funny ad supported file hosting services |
22:50.57 | cr2_ | miknix: what does wiki say ? i think i've added at least the bt gpios. |
22:51.06 | ginge | dcordes: if you use MSN messenger they are the same |
22:51.10 | NetRipper | cr2_, what do you mean by friends? |
22:51.20 | dcordes | ginge, I have the full program ^^ |
22:51.26 | szsoftware_ | dcordes: its on it's way |
22:51.30 | cr2_ | NetRipper: ppp related modules, bsd_compress and such. |
22:51.33 | NetRipper | ah ok |
22:51.48 | ginge | dcordes: of what? |
22:51.58 | cr2_ | NetRipper: you may disable all phone related things. |
22:52.12 | cr2_ | NetRipper: maybe smd too. |
22:52.20 | miknix | cr2_, 121 |
22:52.23 | dcordes | ginge, the full assortment of nonfree IM protocols |
22:53.06 | ginge | dcordes: I see. if you use windows msn you end up buying random things you never knew you needed because it advertises to you all day long (grumble grumble) |
22:53.18 | szsoftware_ | swetland: SanMehat how to compile only ti-wlan within your build-system? make ti ? |
22:53.28 | cr2_ | miknix: but does omap-sd support sdio ? |
22:54.00 | *** join/#htc-linux scatterp_ (n=scatterp@82-35-152-46.cable.ubr05.enfi.blueyonder.co.uk) |
22:54.38 | miknix | cr2_, omap-sd? |
22:54.55 | cr2_ | miknix: omap-mmc or how it is called ? |
22:55.32 | NetRipper | cr2_, i can't disable smd, too many dependencies |
22:55.41 | NetRipper | it won't compile |
22:55.43 | dcordes | ginge, yes horrible. I have several Jack Lelane Power Juicers and Nicer Dicers. |
22:55.54 | ginge | LOL |
22:55.55 | miknix | cr2_, yes, I think it works over sdio |
22:55.58 | dcordes | ginge, did you try a build? |
22:56.04 | cr2_ | NetRipper: ok |
22:56.31 | ginge | no, gnome is in the process of bombing out. I could do with a reboot! |
22:56.33 | cr2_ | NetRipper: did you compile the cdc gadget ? |
22:56.36 | NetRipper | yes |
22:56.57 | NetRipper | cdc is just gadget + ethernet, right? |
22:56.58 | cr2_ | NetRipper: usb-serial is a different thing. |
22:57.03 | cr2_ | yes |
22:57.09 | NetRipper | yes i put both usb-serial and gadget |
22:57.22 | cr2_ | ok |
22:57.24 | NetRipper | or will that cause probs? |
22:57.35 | cr2_ | you may remove usr-serial |
22:57.47 | cr2_ | cdc_ether is better |
22:58.57 | dcordes | ginge, no need I can double check the patch locally |
23:00.08 | NetRipper | CONFIG_USB_GADGET_M66592=y |
23:00.10 | NetRipper | cr2_, that one right? |
23:00.12 | NetRipper | not pxa |
23:00.24 | szsoftware_ | dcordes: did you get mail? |
23:00.26 | cr2_ | hmm. |
23:00.33 | dcordes | checking |
23:00.45 | miknix | anyone owning a neo freerunner? |
23:00.46 | NetRipper | cr2_, can't even select the other two |
23:00.49 | NetRipper | cr2_, can't even select the others |
23:01.04 | cr2_ | ok, it's a difficult question :) |
23:01.09 | NetRipper | :P |
23:01.11 | NetRipper | just so you know :P |
23:01.17 | NetRipper | i guess pxa is for other arch's |
23:01.25 | cr2_ | NetRipper: what does android use ? usb-serial ? |
23:01.40 | NetRipper | i believe it has it disabled by default, let's see |
23:02.15 | NetRipper | cr2_, yes it seems to be usb-serial |
23:02.22 | NetRipper | cr2_, most certainly not gadget |
23:02.31 | cr2_ | ok |
23:02.37 | NetRipper | when using the "USB Function" you can select MSM7K high speed usb stuff |
23:02.43 | cr2_ | that's evil. |
23:02.56 | cr2_ | then you can't use rootfs.cgz |
23:03.07 | NetRipper | doesn't boot initrd yet |
23:03.22 | cr2_ | swetland: does g1 support cdc_ether ? |
23:03.55 | dcordes | szsoftware_, got it, thanks. I will add up the ppp script. do you have a text browser installed? |
23:04.02 | dcordes | I mean in that image |
23:06.09 | NetRipper | cr2_, that msmfb driver is interfering with console output |
23:06.31 | NetRipper | cr2_, so good chance it just doesnt update properly |
23:07.48 | dcordes | szsoftware_, think we should add lynx or so |
23:09.36 | cr2_ | NetRipper: can you disable the console output driver ? |
23:09.47 | szsoftware_ | dcordes: I think not, thought about links, its in the oe packages, dont know wether its in the rootfs or not |
23:09.50 | NetRipper | cr2_, msmfb or druidu's? |
23:09.51 | cr2_ | NetRipper: you don't really need it if msm_fb will work |
23:09.52 | NetRipper | i'd say msm |
23:10.01 | cr2_ | druidu's |
23:10.08 | NetRipper | msm_fb doesn't seem to properly work, else we would've seen it work before |
23:10.17 | szsoftware_ | dcordes: links is links2 and has graphics through libdirectfb |
23:10.19 | NetRipper | oh |
23:10.20 | NetRipper | or not |
23:10.20 | NetRipper | hm |
23:10.22 | NetRipper | sec |
23:10.56 | cr2_ | NetRipper: msm_fb misses the init and power, but wince has already done it. |
23:11.04 | dcordes | szsoftware_, no unforunately not. I only have my OE tree setup for the zaurus atm. would be better to directly install it to the image |
23:11.43 | dcordes | szsoftware_, did they push your fbi metadata so I can manually build a new image tomorrow? |
23:12.14 | szsoftware_ | dcordes: they said, yes, will look at the bug report.. |
23:12.52 | szsoftware_ | dcordes: how do I update my oe git to latest ? |
23:13.07 | dcordes | git-pull |
23:13.17 | szsoftware_ | dcordes: git-pull . |
23:13.21 | NetRipper | provided you checked out using --track |
23:13.21 | szsoftware_ | dcordes: ? |
23:13.42 | dcordes | NetRipper, I didn't do that. still git-pull works for me. |
23:14.13 | NetRipper | dcordes, ok |
23:14.23 | NetRipper | dcordes, didn't for me :p it just said it updated, while it didnt do anything |
23:14.39 | dcordes | koen updated the old mtn instructions to git http://www.angstrom-distribution.org/building-angstrom |
23:15.05 | dcordes | NetRipper, did it as described in the howto and it works |
23:15.14 | NetRipper | dcordes, which howto? |
23:15.23 | dcordes | NetRipper, in the link |
23:15.27 | NetRipper | oh |
23:15.28 | NetRipper | sorry |
23:15.35 | szsoftware_ | dcordes: git says everytimes "Already up to date" ^^ |
23:15.42 | NetRipper | ah ok but you do --rebase |
23:15.44 | NetRipper | when git pulling |
23:15.52 | dcordes | szsoftware_, I guess then you have the uptodate data |
23:16.08 | dcordes | NetRipper, no :) only git-pull in my OE dir |
23:16.14 | szsoftware_ | dcordes: but its some weeks ago, since pulling |
23:16.48 | NetRipper | cr2_, omg, now msm_fb works indeed |
23:16.54 | NetRipper | cr2_, even a blinking icon |
23:16.59 | NetRipper | cursor |
23:17.28 | dcordes | szsoftware_, maybe it's broken. put a cronjob that does rm-rf every few days and clones again ^^ |
23:17.29 | NetRipper | cr2_, fails at: mdp_dma_to_mddi: busy |
23:18.02 | dcordes | ginge, did you look at the "ti" driver? |
23:18.26 | NetRipper | omg, i feel sorry for druidu, although its thanks to his driver that we got to see what it hung |
23:19.04 | NetRipper | it'll be useful again when we go for ppp module |
23:19.11 | NetRipper | cr2_, think i had a kernel panic |
23:19.23 | NetRipper | cr2_, or something like a stack trace with registers and method names |
23:21.28 | NetRipper | cr2_, INFO: task rpcrouter:13 blocked for more than 120 seconds |
23:21.35 | NetRipper | and after that some kind of stacktrace |
23:21.45 | *** join/#htc-linux Spicypixel (i=Spicypix@cpc4-linc8-0-0-cust509.nott.cable.ntl.com) |
23:21.54 | NetRipper | cr2_, i'll make a haret thingy for you, then im off to bed |
23:22.09 | NetRipper | at least it's no hard-hang, as it repeats the message every 120 seconds |
23:22.09 | NetRipper | :p |
23:22.13 | NetRipper | (or so) |
23:22.15 | Spicypixel | hey netripper before you got, is it you whos working on the Diamond and Android |
23:22.22 | Spicypixel | go* |
23:23.22 | dcordes | bbiab |
23:23.46 | cr2_ | NetRipper: ok, it's the proc_comm fault obviously. |
23:24.07 | Spicypixel | you guys are awesome |
23:24.49 | cr2_ | NetRipper: the problem was/is the proc_comm (and clk_*) which blocked everything |
23:24.49 | Spicypixel | can the Touch/Pro be used without a stylus? |
23:24.54 | NetRipper | cr2_, i commented out with proc_comm |
23:25.17 | NetRipper | Spicypixel, among others, yes |
23:25.19 | cr2_ | ok. but at least we understand what to do now. |
23:25.34 | cr2_ | where are the problems, and how to solve them. |
23:25.36 | Spicypixel | have to say, you guys arent half taking on a challenge |
23:25.45 | cr2_ | at least i understand it :) |
23:25.55 | NetRipper | cr2_, glad you do, i sure don't |
23:25.55 | NetRipper | :D |
23:26.15 | NetRipper | cr2_, anyway let me make that haret exec so you can see the framebuffer magic yourself |
23:26.22 | cr2_ | NetRipper: i#ll leave the hardcore i2c tracing to you :) |
23:26.29 | cr2_ | ok, thanks. |
23:26.49 | NetRipper | uhhh i2c tracing? if you learn me how, i will |
23:26.49 | NetRipper | :P |
23:27.15 | cr2_ | i still need to document my i2c findings. includng the accelerometer commands/data |
23:27.20 | NetRipper | ok |
23:27.28 | NetRipper | there is accelerometer support in G1 code |
23:27.42 | cr2_ | NetRipper: just mmutrace for i2c_send and i2c_recv registers. |
23:27.50 | cr2_ | NetRipper: this is a different device. |
23:28.02 | NetRipper | cr2_, aww |
23:28.06 | cr2_ | diam/raph accelerometer chips is different |
23:28.11 | NetRipper | ok |
23:28.33 | cr2_ | Spicypixel: without the stylus ? |
23:28.56 | NetRipper | hm, somehow i screwed up some fonts on my windows mobile |
23:29.09 | cr2_ | hehe. |
23:29.22 | NetRipper | i'm getting monospace fonts on some applications |
23:29.28 | NetRipper | almost like linux is taking over windows mobile |
23:29.42 | NetRipper | omg, skynet has awakened! |
23:29.43 | NetRipper | ;) |
23:29.53 | cr2_ | lol |
23:30.12 | cr2_ | it's the arm9 rpc sent a message to big brother. |
23:30.51 | Spicypixel | yeah im saying is the touchscreen on the touch/pro ... capacative? |
23:30.54 | Spicypixel | i think thats the term |
23:30.57 | Spicypixel | iphone like |
23:31.11 | cr2_ | Spicypixel: no, the keypad is capacitive |
23:31.20 | cr2_ | the lower part |
23:31.22 | Spicypixel | righto |
23:31.34 | Spicypixel | what about the Touch HD? |
23:31.44 | cr2_ | it really tracks your finger movement |
23:32.00 | cr2_ | in realtime, even if you don't touch it |
23:32.07 | Spicypixel | awesome |
23:32.13 | cr2_ | don't know abut HD |
23:32.37 | Spicypixel | its just nice to see some progress on this |
23:32.50 | Spicypixel | i for one think "chrome mobile" deserves 640x480 |
23:33.00 | cr2_ | yeah, but there are still some minor challenges. |
23:33.12 | Spicypixel | calling it minor is amazing in itself |
23:33.13 | Spicypixel | lol |
23:33.20 | Spicypixel | followed linux on the uni for... months |
23:33.52 | Spicypixel | it killed my devices battery, from broken power drivers, but it was worth it lol |
23:33.54 | cr2_ | compared to the wifi driver, and msm7201A support without having any docs, they can be seen as minor. |
23:34.00 | NetRipper | cr2_, http://www.netripper.nl/raphael/raph_20081022_framebuffer_magic.exe |
23:34.14 | cr2_ | Spicypixel: broken power drivers ? |
23:34.24 | NetRipper | cr2_, after boot, wait approx 3 minutes before it shows error messages |
23:34.49 | cr2_ | Spicypixel: well, it#s true. i remember that wince uses 1 more gpio for something, and linux driver does not take it into account. |
23:34.58 | Spicypixel | inded |
23:35.06 | cr2_ | NetRipper: downloading |
23:35.17 | Spicypixel | question, is there any 3d acceleration in the MSM7201A chipset? |
23:35.25 | Spicypixel | cause the iphone has the power vr stuff |
23:35.31 | NetRipper | cr2_, lol and try opening/closing your keyboard for some fun |
23:35.37 | AstainHellbring | yes it has 3d acceleration |
23:35.39 | cr2_ | yes, but it's not a "real" opengl |
23:35.47 | Spicypixel | ive coded some opengl ES |
23:35.51 | Spicypixel | its enough :P |
23:35.54 | cr2_ | ok |
23:36.01 | NetRipper | Spicypixel, you wrote the opengl es c# wrapper? |
23:36.06 | Spicypixel | negative |
23:36.09 | NetRipper | ah ok |
23:36.18 | Spicypixel | i worked on opengl es on a virtual machine at uni |
23:36.21 | Spicypixel | not much |
23:36.22 | NetRipper | ah |
23:36.29 | Spicypixel | just enough to see it was useful for phones |
23:36.34 | *** join/#htc-linux dcordes_ (n=dcordes@unaffiliated/dcordes) |
23:36.49 | *** join/#htc-linux Zylogue (n=Zylogue@ip68-12-77-66.ok.ok.cox.net) |
23:37.04 | NetRipper | cr2_, and try pressing some hardware keys, some are already read :) |
23:37.38 | NetRipper | "read" being overrated, but still |
23:38.12 | Spicypixel | which phone are you trying this on? |
23:38.16 | NetRipper | raphael |
23:38.22 | NetRipper | it'll be same for diamond though |
23:38.39 | cr2_ | NetRipper: but it's the full boot |
23:38.43 | NetRipper | cr2_, yes |
23:38.49 | NetRipper | cr2_, but doesn't get to initrd |
23:38.55 | cr2_ | Freeing init memory, then comes the initrd. |
23:39.00 | cr2_ | ok |
23:39.03 | NetRipper | cr2_, hm |
23:39.13 | NetRipper | cr2_, i'll add some more output at the point where it boots initrd |
23:39.20 | NetRipper | cr2_, the initrd may simply need recompiling |
23:39.29 | cr2_ | mmc works ? |
23:39.41 | cr2_ | the card is not detected |
23:39.46 | NetRipper | it goes pretty fast |
23:39.51 | cr2_ | hmm. we need to fix the gpios |
23:39.57 | cr2_ | yeah, wince SUCKS |
23:40.03 | dcordes_ | NetRipper, which initrd do you use? |
23:40.20 | cr2_ | NetRipper: we need to track all hardcoded/set gpios |
23:40.20 | NetRipper | dcordes_, the rootfs.cfz that Kevin2 used on the hermes for usb gadget |
23:40.56 | dcordes_ | cr2_, did you already figure the smd chans? |
23:40.59 | cr2_ | ok, see the oops |
23:41.12 | cr2_ | dcordes: yes, but not really. |
23:41.28 | cr2_ | dcordes: there are more important things to fix first |
23:41.48 | cr2_ | NetRipper: rtc does not work ? |
23:41.57 | NetRipper | cr2_, not sure |
23:42.09 | NetRipper | cr2_, but i really got to end this for tonight :p |
23:42.13 | *** join/#htc-linux datachaos (n=datachao@189.27.166.201.adsl.gvt.net.br) |
23:42.16 | NetRipper | my boss will kill me if im sleeping @ work |
23:42.17 | NetRipper | ;) |
23:42.25 | cr2_ | i wish to get the full dmesg :) |
23:42.28 | NetRipper | yea |
23:42.29 | cr2_ | hehe. |
23:42.30 | NetRipper | indeed |
23:43.05 | cr2_ | ok. gr |
23:43.12 | cr2_ | it was a great day :) |
23:43.27 | NetRipper | i'll try to get a prompt tomorrow |
23:43.42 | cr2_ | but i also see a lot of things for me to solve . |
23:43.43 | NetRipper | just to make sure we can get initrd started |
23:43.50 | cr2_ | ok. |
23:44.08 | NetRipper | ok good night for now |
23:44.12 | NetRipper | :) |
23:44.18 | cr2_ | good night :) |
23:44.19 | NetRipper | been a good day indeed |
23:44.19 | NetRipper | ;) |
23:44.55 | cr2_ | create board-raph |
23:45.01 | cr2_ | so we can fix the gpios. |
23:45.26 | dcordes_ | night NetRipper |
23:45.27 | cr2_ | actually there are a lot of things to fix. |
23:45.56 | cr2_ | lol |
23:46.17 | cr2_ | even 3 buttons work :) |
23:46.28 | cr2_ | on the left side |
23:46.41 | cr2_ | the matrix gpios are a bit mixed |
23:46.44 | cr2_ | it seems |
23:47.00 | dcordes_ | did somebody trace the keys already? |
23:47.19 | AstainHellbring | oh raph? |
23:47.25 | cr2_ | and the softreset works |
23:47.37 | cr2_ | dcordes: the pins are in wiki |
23:47.42 | dcordes_ | AstainHellbring, yea |
23:48.01 | AstainHellbring | dont know as anyone has but I have a raph that I can do it on for you |
23:48.16 | cr2_ | the key<board is on i2c. it needs a separeate driver anyway |
23:49.48 | NetRipper | lol @ < cr2_> and the softreset works |
23:50.22 | cr2_ | NetRipper: it drops the junk on fb |
23:50.51 | NetRipper | pressing soft reset gives output?? |
23:51.00 | NetRipper | i know about the buttons |
23:51.07 | NetRipper | and when you open/close keyboard |
23:51.20 | cr2_ | open/clse is a gpio |
23:51.44 | cr2_ | try the softreset |
23:51.55 | cr2_ | it may be handled by arm9 btw |
23:52.19 | cr2_ | there are several arm9-owned gpios |
23:52.24 | cr2_ | so why not. |
23:52.54 | NetRipper | well i didn't notice anything being written to fb when soft resetting |
23:53.06 | cr2_ | NetRipper: what about penguin at boot now ? |
23:53.07 | dcordes_ | cr2_, on kaiser softreset is cpld gpio |
23:53.19 | cr2_ | dcordes_: ok |
23:53.20 | NetRipper | cr2_, lol, i was just enabling logo |
23:53.25 | cr2_ | :D |
23:53.30 | NetRipper | cr2_, but it refers to a /logo.rle |
23:53.39 | NetRipper | the msm logo option at least |
23:53.50 | cr2_ | NetRipper: will make a nice screenshot :) |
23:53.53 | NetRipper | yea it would |
23:54.14 | cr2_ | it think that the standard fb logo should work |
23:54.40 | cr2_ | hehe. you may use the small font |
23:54.46 | NetRipper | <PROTECTED> |
23:54.46 | NetRipper | <PROTECTED> |
23:54.47 | cr2_ | to see more on the screen. |
23:54.47 | NetRipper | <PROTECTED> |
23:55.08 | cr2_ | is it readable with gimp ? |
23:55.16 | NetRipper | i cant find a logo.rle |
23:55.18 | cr2_ | wtf is in this logo anyway ? |
23:55.27 | NetRipper | no idea |
23:55.37 | NetRipper | and the / in front suggests it finds it somewhere |
23:55.38 | cr2_ | in the dump ? |
23:55.49 | NetRipper | during boot or compile-time |
23:56.02 | dcordes_ | NetRipper, try creating it in the kernel root? |
23:56.16 | NetRipper | it doesnt complain while compiling |
23:56.20 | NetRipper | so no idea if it looks for it |
23:56.26 | cr2_ | NetRipper: try the standard logo first |
23:56.28 | dcordes_ | maybe ignored if not present |
23:56.56 | NetRipper | cr2_, the default logo's probably only work if the driver actively checks for it |
23:57.24 | szsoftware_ | NetRipper: cr2_ what about a modified http://linuxtogo.org/~lgorris/kaiser-bootkit/htc-tux.jpg ? :-) |
23:57.37 | NetRipper | #ifdef CONFIG_FB_MSM_LOGO |
23:57.37 | NetRipper | #define INIT_IMAGE_FILE "/logo.rle" |
23:57.37 | NetRipper | extern int load_565rle_image( char *filename ); |
23:57.37 | NetRipper | #endif |
23:58.16 | NetRipper | looks like it searches runtime |
23:58.18 | NetRipper | not compiletime |
23:58.27 | NetRipper | so it'll have to be in initrd or something |
23:59.05 | NetRipper | ah! |
23:59.11 | NetRipper | found the logo.rle of android |
23:59.14 | NetRipper | g1 dump |
23:59.14 | NetRipper | ;) |
23:59.29 | dcordes_ | can you open it in gimp? |
23:59.38 | NetRipper | man |
23:59.41 | NetRipper | im a windows user |
23:59.45 | NetRipper | let me check if i have gimp |