00:23.02 | *** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821) |
00:34.34 | *** join/#htc-linux root2 (n=root@rgnb-5d87dd8d.pool.einsundeins.de) |
00:36.04 | *** join/#htc-linux StarLite (n=nnscript@84-245-8-228.dsl.cambrium.nl) |
00:38.59 | *** join/#htc-linux PSquid (n=psquid@77-100-16-57.cable.ubr06.harb.blueyonder.co.uk) |
00:41.42 | *** join/#htc-linux Moku (n=John@f049036115.adsl.alicedsl.de) |
00:42.47 | *** part/#htc-linux PSquid (n=psquid@77-100-16-57.cable.ubr06.harb.blueyonder.co.uk) |
00:44.40 | DrHouse|HP | humm wonder why the keyboard on the Hermes works intermittently on Linux |
00:45.35 | tmzt | is it spi? |
00:46.09 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
00:47.47 | DrHouse|HP | i think i have no idea offhand |
00:48.07 | DrHouse|HP | is there a way i can find out? |
00:48.18 | tmzt | look at the source |
00:48.34 | tmzt | you might try porting upstream to it if the drivers are complete enough |
00:48.37 | tmzt | it might work better |
00:48.46 | tmzt | I believe 2400 support is now upstream |
00:52.01 | DrHouse|HP | yes it is SPI |
01:13.36 | *** join/#htc-linux root2 (n=root@rgnb-5d8740e5.pool.einsundeins.de) |
01:38.36 | *** join/#htc-linux root2 (n=root@rgnb-5d87ccc3.pool.einsundeins.de) |
01:42.26 | *** join/#htc-linux miknix (n=miknix@gentoo/developer/miknix) |
01:44.15 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
01:45.45 | miknix | arghh.. I'm going crazy |
01:46.15 | miknix | anyone with a HTC Jade 100? |
02:00.24 | *** part/#htc-linux darweth (n=cryptic@ool-ad03bcc0.dyn.optonline.net) |
02:34.53 | *** join/#htc-linux root2 (n=root@rgnb-5d87dcdc.pool.einsundeins.de) |
03:01.37 | *** join/#htc-linux root2 (n=root@rgnb-5d87ce06.pool.einsundeins.de) |
03:09.00 | *** join/#htc-linux mrmoku` (n=mrmoku@ppp-93-104-118-10.dynamic.mnet-online.de) |
03:24.37 | *** join/#htc-linux root2 (n=root@rgnb-5d87cb01.pool.einsundeins.de) |
03:48.40 | *** join/#htc-linux root2 (n=root@rgnb-5d87d72a.pool.einsundeins.de) |
03:52.21 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
04:21.56 | *** join/#htc-linux |Jason8| (i=JayAte@server1.tiltshellz.org) |
04:27.39 | *** join/#htc-linux root2 (n=root@rgnb-5d874ad8.pool.einsundeins.de) |
04:46.26 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
04:51.09 | *** join/#htc-linux droid0011 (n=mc@p4FDCDE1D.dip.t-dialin.net) |
05:00.32 | *** join/#htc-linux Amaranth__ (n=travis@74-221-38-55.longlines.com) |
05:01.26 | *** join/#htc-linux root2 (n=root@rgnb-5d87db53.pool.einsundeins.de) |
05:09.59 | *** join/#htc-linux Echo31 (n=olivier@mir31-4-82-240-194-54.fbx.proxad.net) |
06:09.37 | *** join/#htc-linux root2 (n=root@rgnb-5d87d6c5.pool.einsundeins.de) |
06:45.08 | *** join/#htc-linux marex (n=marex@thor.hackndev.com) |
06:54.28 | *** join/#htc-linux root2 (n=root@rgnb-5d874cfa.pool.einsundeins.de) |
07:13.09 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
07:17.25 | *** join/#htc-linux root2 (n=root@rgnb-5d87cfc1.pool.einsundeins.de) |
07:48.42 | *** join/#htc-linux root2 (n=root@rgnb-5d874da8.pool.einsundeins.de) |
07:51.46 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
08:09.03 | *** join/#htc-linux root2 (n=root@rgnb-5d87dcf2.pool.einsundeins.de) |
08:21.42 | *** join/#htc-linux StarLite` (n=nnscript@84-245-8-228.dsl.cambrium.nl) |
08:25.45 | *** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
08:41.32 | *** join/#htc-linux root2 (n=root@rgnb-5d874e42.pool.einsundeins.de) |
09:08.24 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
09:10.36 | *** join/#htc-linux MethoS- (n=clemens@host-091-097-243-131.ewe-ip-backbone.de) |
09:12.45 | *** join/#htc-linux root2 (n=root@rgnb-5d87c339.pool.einsundeins.de) |
09:28.36 | *** join/#htc-linux Amaranth (n=travis@ubuntu/member/Amaranth) |
09:52.33 | *** join/#htc-linux root2 (n=root@rgnb-5d87c24a.pool.einsundeins.de) |
10:01.30 | *** join/#htc-linux dcordes (n=luke@unaffiliated/dcordes) |
10:09.24 | *** join/#htc-linux leviathan (n=kvirc@c-82-192-226-27.customer.ggaweb.ch) |
10:10.04 | leviathan | msmfb_pan_display timeout waiting for frame start |
10:10.18 | leviathan | does anyone have an idea how to fix this? |
10:10.36 | leviathan | I bootet succesfully openmoko |
10:10.39 | leviathan | ok htc-dream |
10:10.56 | leviathan | but the msmfb claims about failing frames |
10:11.36 | leviathan | please stop ignoring questions about non-android topics and just look at this problem as htc-specific question |
10:11.53 | leviathan | http://wiki.openmoko.org/wiki/OpenMoko_on_HTC-Dream |
10:12.17 | leviathan | I started porting OpenMoko onto the HTC-Dream |
10:12.25 | leviathan | access to the chip works so far |
10:12.30 | leviathan | only sound is a problem |
10:12.35 | leviathan | an graphics |
10:12.44 | leviathan | and wifi has some bugs |
10:12.52 | leviathan | but this all will work soon |
10:13.02 | leviathan | its not soooo much work to do |
10:13.09 | leviathan | but now i need you help |
10:13.42 | tmzt | hello |
10:14.02 | tmzt | we aren't ignoreing questions |
10:14.06 | tmzt | we are asleep |
10:14.35 | tmzt | this channel is logged and people read the logs but sometimes it takes time to get an answer |
10:14.45 | tmzt | which chip? |
10:17.53 | leviathan | msmfb, the standard chip of htc-dream |
10:18.20 | leviathan | im using the kernel modul taken from the android kernel git repo |
10:19.45 | leviathan | root@om-gta02:~# Xfbdev |
10:19.45 | leviathan | Using **pending_return in dbus_connection_send_with_reply_setup() without pending_setup is deprecated and strongly discouraged |
10:19.48 | leviathan | Using **pending_return in dbus_connection_send_with_reply_setup() without pending_setup is deprecated and strongly discouraged |
10:19.51 | leviathan | Using **pending_return in dbus_connection_send_with_reply_setup() without pending_setup is deprecated and strongly discouraged |
10:19.57 | leviathan | Using **pending_return in dbus_connection_send_with_reply_setup() without pending_setup is deprecated and strongly discouraged |
10:20.06 | leviathan | unrecognised device identifier! |
10:20.08 | leviathan | Using **pending_return in dbus_connecti |
10:20.31 | leviathan | dmesg says: |
10:20.33 | leviathan | msmfb_pan_display timeout waiting for frame start |
10:20.54 | leviathan | sry for posting so much |
10:21.04 | tmzt | that's okay |
10:21.51 | leviathan | do you have an idea, how this comes? |
10:21.57 | *** join/#htc-linux raster (n=raster@enlightenment/developer/raster) |
10:22.03 | tmzt | yeah, there's a few reasons |
10:22.10 | raster | boo |
10:22.10 | leviathan | could the timeout in the kernel driver be to low? |
10:22.11 | tmzt | not sure which one it is in this case |
10:22.14 | tmzt | hello |
10:22.20 | leviathan | hi raster |
10:22.26 | tmzt | what does dbus have to do with this? |
10:22.44 | leviathan | tmzt: its an openmoko specific thing =) |
10:22.49 | tmzt | ok |
10:23.13 | tmzt | I believe the update thread (CONFIG_FB_MSM_UPDATE) needs to be enabled |
10:23.27 | tmzt | and the vyres needs to be reduced to one screen height |
10:23.46 | leviathan | ok, ok |
10:24.03 | tmzt | 14:01 < tmzt> CONFIG_FB_MSM_REFRESH=y |
10:24.06 | tmzt | I mean |
10:24.17 | tmzt | what kernel are you using? |
10:24.44 | raster | leviathan: yo yo :) |
10:24.47 | raster | tmzt: bing |
10:25.03 | tmzt | ? |
10:25.15 | raster | just acking :) |
10:25.21 | leviathan | 2.6.29 |
10:25.44 | raster | the fb driver looks wierd in that tis continually dmaing from one fb to another |
10:25.46 | leviathan | I took android-msm-2.6.29 and merged drivers from wireless-testing into it |
10:25.49 | raster | (eg from backing to front store) |
10:27.54 | dcordes | mhm |
10:28.00 | tmzt | dos that work? |
10:28.43 | leviathan | ONFIG_FB_MSM_REFRESH-Option does not exist in this kernel |
10:28.47 | leviathan | dont know why |
10:28.52 | dcordes | leviathan, what tmzt says. CONFIG_FB_MSM_UPDATE is necessary for Xfbdev, at least on the other msm7xxx devices. |
10:29.04 | dcordes | leviathan, because it is only in our own msm kernel |
10:29.14 | tmzt | right, that's what I mean |
10:29.29 | leviathan | android-msm-2.6.29 |
10:29.30 | tmzt | and you really shouldn't take our vkeyb into that kernel |
10:29.31 | leviathan | ? |
10:29.35 | dcordes | s/msm7xxx/msm7xxxx/ |
10:29.43 | tmzt | actually, this should be done in ddx were it belongs :) |
10:30.25 | raster | probably |
10:30.37 | raster | even then u'd be needing to do the whole stack of accel/hw supporty |
10:30.37 | dcordes | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=commitdiff;h=fcf0dc8dc232b237dc8d3098a5de09fab0ad9ba3 |
10:31.05 | raster | it's kind of a - put one toe in and it wont work until you jump in body and all |
10:31.12 | dcordes | raster, can you look at this? without the above hack Xfbdev doesn't refresh the screen. but android works fine without it. why? |
10:31.40 | raster | because xfbdev is drawing to the backbuffer |
10:31.54 | raster | nothng u draw gets to the front/real buffer without a copy |
10:32.06 | raster | i dont actually know if the frontbuffer is exposed in memory for writes |
10:32.11 | raster | (and reads) |
10:32.11 | tmzt | oh |
10:32.18 | raster | but if it is - thats where the fb should be |
10:32.21 | tmzt | this is not related to mdp at all? |
10:32.27 | tmzt | it's just the flipping? |
10:32.31 | raster | ot looks after a quick read like the fb is in thwe wrong spot |
10:32.33 | tmzt | we just had them backwords |
10:32.39 | raster | now that might be because it CANT be in th real spot |
10:32.44 | tmzt | we should be using the second framebuffer |
10:32.45 | raster | thus the copies |
10:32.56 | raster | thats what it looks like to me |
10:32.56 | tmzt | (yres offset=480) |
10:32.58 | leviathan | do you have wl12xx-driver in your kernel? |
10:32.59 | raster | not 100% sure |
10:33.04 | tmzt | leviathan: no |
10:33.29 | raster | but it looks like u're dmaing from one to the other continually to get refresh - thus. the logical thing is that the other is where the lcd scanout is going on |
10:33.38 | tmzt | I think if raster is right we can just fix this |
10:33.39 | raster | and thus - thats where a normal fb would live |
10:33.39 | dcordes | leviathan, not in git but it applies cleanly. |
10:33.42 | leviathan | hmm, if I could get access to the git-repo I could try to migrate it |
10:33.45 | tmzt | make the framebuffer default to the second offset |
10:33.51 | raster | trya quick fix - change fb to be the other one :) |
10:33.54 | tmzt | and change the exposed framebuffer to the frontbuffer |
10:33.58 | dcordes | leviathan, register at linuxtogo if you want the git accesss. |
10:34.03 | raster | quick test :) |
10:34.11 | tmzt | how can we do that? |
10:34.20 | tmzt | both of the are exposed as part of the fbdev |
10:34.30 | tmzt | s/the/them/ |
10:34.44 | tmzt | that's why virtual_yres is doubled |
10:35.49 | raster | fbdev just needs adjusting to put the kernel fb at an offset |
10:35.57 | raster | its perfectly capabhle of using another address |
10:36.13 | raster | expose 2 fb devices, not one |
10:36.17 | raster | fb0 will be the 2nd one |
10:36.32 | raster | likely there is a register to flip and tell it whihc one to scanout and it happens to be set to the 2nd by default from boot |
10:36.49 | raster | you will need to find that register/control for pageflipping later anyway. |
10:37.42 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
10:39.00 | dcordes | leviathan, if you want to get started with Xfbdev just integrate the above patch in your dream kernel. it's ugly but works well |
10:39.30 | dcordes | raster, is it difficult to realise what you said? I don't really understand it, else I would try |
10:39.46 | tmzt | well, it shouldn't be that hard to modify Xorg with a oe patch to use a differeent offset into the mmaped buffer right? |
10:40.07 | raster | dcordes: depends how much info u have on the msm and the amount of sample code |
10:40.14 | raster | (and how easy it is to read) |
10:40.18 | dcordes | ok so this is what we were talking about once, it's not the msm_fb but X problem |
10:40.26 | raster | tmzt: it'd be trivial |
10:40.37 | raster | but as such the kernels own console should work anyway |
10:40.41 | raster | without the copy hack |
10:41.02 | raster | does the kernels own console work right? |
10:41.05 | dcordes | without the hack, the framebuffer console works yes |
10:41.14 | raster | ie it draws to the right fb - or it actually set up the correct scanout fb? |
10:41.19 | raster | ok |
10:41.20 | tmzt | yeah |
10:41.25 | tmzt | we don't know how though |
10:41.29 | raster | so the kernel actually puts the console at the offset? |
10:41.32 | tmzt | fbcon works fine |
10:41.37 | raster | thats odd. |
10:41.44 | tmzt | maybe it does |
10:42.03 | raster | i dont have any basis to start with on my g1 - so i cant do more than guess atm |
10:42.10 | dcordes | also on a side note, Xfbdev does pretty fine with the refresh hack. you can even watch movies with mplayer at decent framerates. |
10:42.17 | raster | is there an oe tree for building - or is it in oe-dev? |
10:42.17 | dcordes | don't know if thats's relevant |
10:42.32 | raster | hmm |
10:42.37 | raster | that hack really looks bad to me |
10:42.42 | raster | i suspect u are giving up performance |
10:43.02 | raster | xfbdev will always be at a handicap as its a dumb fb software xserver |
10:43.12 | raster | all operatiosn are done with the cpu to a dumb fb |
10:43.25 | raster | any pointers to docs on how to get started? |
10:43.40 | raster | i have cupcake adp1.5 on my g1 |
10:43.51 | raster | i'm using it as my phone.. so non-destroying would be nice :) |
10:43.51 | dcordes | ouff.. wait a minute |
10:44.10 | leviathan | i was ennoyed about this big amount of binary drivers and the license of the android userspace so I forked the android kernel (2.6.29) and begun to integrate free drivers for wlan e.g. (wl12xx) |
10:44.19 | leviathan | and now comes another fbdriver |
10:44.29 | leviathan | openmoko-msm-2.6.29 |
10:44.38 | dcordes | leviathan, do you follow the lkml discussion about upstreaming dream? |
10:44.52 | dcordes | yet another userspace dependant dream kernel :P ? |
10:45.00 | dcordes | the "openmoko" makes it sound like that |
10:45.05 | leviathan | :) |
10:45.15 | dcordes | jk |
10:46.01 | raster | if there's somethnig to work off i can poke around my g1 a bit and see what i kab do kernel/xorg wise |
10:46.12 | raster | it'd be nice to get something going off sd |
10:46.31 | raster | i'd like to see how this performs relative to other platforms anyway... and i have a benchmark for that :) |
10:47.08 | raster | right now my "real linux" devices are uber-high res doobies with wimax instead of wifi :) |
10:47.13 | dcordes | raster, http://android.git.kernel.org/?p=kernel/msm.git;a=tree;f=drivers/video/msm there was a sort of readme in that folder |
10:47.58 | dcordes | it had some comments on the msm fb |
10:49.58 | tmzt | raster: JesusFreke recovery? |
10:50.06 | tmzt | you can get a shell with alt-x |
10:50.06 | raster | oooh no no - i was wondering if there is somehting i can actually begin with that runs |
10:50.22 | tmzt | what do you have now? |
10:50.29 | raster | tmzt: ummmm... i think i have.. or jad the jf recovery |
10:50.36 | raster | adp may have upgraded it... :/ |
10:50.51 | mickeyl | morning folks. any news wrt. audio? |
10:51.07 | dcordes | raster, here http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=drivers/video/msm/msm_fb.txt |
10:51.29 | tmzt | mickeyl: hey, leviathan is working on fso/shr for g1 |
10:51.38 | *** join/#htc-linux xperia (n=chatzill@77-56-95-46.dclient.hispeed.ch) |
10:51.47 | *** join/#htc-linux Amaranth__ (n=travis@74-221-38-55.longlines.com) |
10:51.52 | mickeyl | tmzt: great news. |
10:51.53 | tmzt | nothing without audio on raph, that's waiting on cr2 to come back from vacation mostly |
10:51.58 | mickeyl | hmm, k |
10:52.08 | mickeyl | backlight perhaps?` |
10:52.48 | mickeyl | waves to mrmoku |
10:53.13 | dcordes | leviathan, you should hook up with Gnutoo he has the same aim as you. |
10:53.15 | tmzt | I can upload what I have |
10:53.16 | mrmoku | waves back to mickeyl :) |
10:53.24 | tmzt | I didn't get the rpc working though |
10:53.27 | raster | dcordes: msm_fb.txt has gone |
10:53.31 | tmzt | can't really test anything on my phone right now |
10:53.49 | dcordes | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=drivers/video/msm/msm_fb.txt;h=f90e0995a721d506bcbb10c33f63753a5f313ed3;hb=htc-msm |
10:54.05 | raster | tmzt: which jf recovery do you use? |
10:54.21 | dcordes | heh |
10:54.23 | mickeyl | hmm, i try to understand why that is so difficult. isn't it just using an existing function and moving it into another class driver? |
10:55.59 | dcordes | raster, he has no dream but cdma raphael |
10:56.15 | raster | hmm ok |
10:57.03 | xperia | hello to all just a litlle question. can somebody tell me the range of the i2c adresses. the microp-klt isnt working on 0x66 and on 0x67 is allready microp-ksc working. now i need to test the other i2c adresses but i dont know the min and max range . |
10:57.48 | tmzt | mickeyl: from what driver? |
10:57.58 | tmzt | I mean what existing function |
10:58.11 | tmzt | dcordes: do you know anything about backlight driver? |
10:58.19 | dcordes | xperia, http://www.htc-linux.org/wiki/index.php?title=RaphaelMemoryMapPg2 |
10:58.20 | mickeyl | tmzt: led class driver -> backlight class driver |
10:58.49 | raster | dcordes: the jf recovery.img has an option to boot off sd? does it kexec the kernel from there? |
10:59.01 | tmzt | ah |
10:59.09 | tmzt | is backlight on klt? |
10:59.11 | tmzt | yeah |
10:59.29 | mickeyl | it's already working fine as led class driver. the only annoying thing is that it's wrong |
10:59.30 | tmzt | so the problem was I didn't know the proper way to export a function from maejrep's driver |
10:59.30 | mickeyl | :) |
10:59.53 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
10:59.59 | tmzt | http://people.openezx.org/tmzt/msm_klt_backlight.c |
11:00.03 | tmzt | maejrep: you here? |
11:03.15 | dcordes | xperia, do you have the aiv git? |
11:04.48 | leviathan | yeahhh |
11:04.51 | leviathan | hot |
11:04.58 | leviathan | I got a GUI! |
11:05.09 | tmzt | what did you do? |
11:05.10 | xperia | dcordes thanks a lot ofr the link. yeah i have it but i am not shure if i can give you his patch. my patch for the fbcon screen rotation on the keyboard slider event i can upload to a place. ps do you know if +0x50 DISP_WAKE |
11:05.46 | dcordes | xperia, I think I will push all of the aiv patches. |
11:06.53 | xperia | ah you also have aiv kernel patches in this case. hmmm didnt know that. well it will be good to have a central online repo so we dont need to work on decentralized repos |
11:07.15 | dcordes | xperia, right. |
11:07.29 | dcordes | do you know how to clone the aiv git? |
11:07.48 | dcordes | I have the web url but what must be used for git clone |
11:08.37 | xperia | i know how to klone a git repo but how i can klone aiv online repo i dont know. i have just downloaded a snapshot and added it to a new git |
11:09.13 | xperia | ahhh okay wait i will tell you the needed steps |
11:09.22 | *** join/#htc-linux miknix (n=miknix@gentoo/developer/miknix) |
11:11.54 | xperia | dcordes are you sure you have the right git url and not the http git web adress. normally you do this here. git clone git://domain/subfolders/kernel.git |
11:11.56 | xperia | git checkout |
11:12.03 | raster | aaah ok |
11:12.14 | raster | totally sure. dma is copying from first fb to 2nd |
11:12.19 | raster | lcd scanout is at second only |
11:12.33 | raster | x is askign fbdev for the whole fb - and of course is using the lower/first address |
11:12.48 | raster | the extra is "spare video ram" it wont use as its fbdev it doesnt use it for accel (pixmaps) |
11:13.03 | *** join/#htc-linux DJWillis (i=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) |
11:13.05 | dcordes | raster, you are a complete genious. finally somebody shed light on the X use on msm |
11:13.17 | raster | the question is.. how to change the location of the fb for the hw |
11:13.36 | raster | i'm looking around |
11:14.17 | xperia | dcordes you need a url with kernel.git at the end. if you dont have this the best thing is if you download the snapshot.gz from his site and added it to a new git repo |
11:14.49 | dcordes | doing this manually sucks big time this is why I asked you for the clone url. you don't know it so it's ok |
11:15.18 | leviathan | whats the native resolution of the htc-dream screen? |
11:15.25 | dcordes | xperia, option is you provide your git-diff . |
11:15.30 | leviathan | tmzt: whats the native resolution of the htc-dream screen? |
11:15.32 | dcordes | then we can add your stuff too., |
11:15.39 | tmzt | xperia: what does the url to clone online look like |
11:15.43 | tmzt | leviathan: hvga |
11:15.57 | tmzt | 480x320 I think |
11:16.06 | leviathan | ok, thx! |
11:16.13 | dcordes | http://dev.aiv-dev.info/git/?a=summary&p=X1%20kernel |
11:17.35 | raster | if i can get things set up to run kernels and have userspace enough to test (build and patch) i could see |
11:17.58 | tmzt | you have to flash kernel and initrd together as recovery.img |
11:18.05 | raster | but there must be something that controls which fb is scanned out |
11:18.15 | tmzt | the unpacking and repacking root page should have instructions |
11:18.22 | *** join/#htc-linux dzo (n=dzo@mail.marginz.co.nz) |
11:18.22 | raster | tmzt: oh kernel too |
11:18.29 | dcordes | dzo, hi |
11:18.32 | raster | revoery wont kexec? |
11:18.36 | raster | recovery |
11:18.39 | dzo | hi |
11:18.48 | tmzt | dzo: cmonex has some questions about raph700 |
11:18.55 | tmzt | which is msm7500 non-A based |
11:19.20 | dzo | ok, didn't know such a thing existed. |
11:19.35 | tmzt | yeah, it's a cdma device with r-uim |
11:19.42 | tmzt | but similar to the touch pro |
11:19.58 | dzo | Are you sure it's non-A, why would they do that? |
11:20.08 | tmzt | raster: it might if kernel supports |
11:20.18 | tmzt | dzo: no I'm not, but no2chem said it was |
11:20.37 | tmzt | they also said the spl is similar to raph800 (Sprint version) |
11:20.42 | tmzt | at least the usb detect is the same |
11:20.49 | tmzt | which makes no sense to me |
11:20.49 | raster | tmzt: hmm that'd be my preferred way right now - i'm leery of futzing with it as i'm using it as a phone |
11:20.57 | tmzt | it's 400mHz as well |
11:21.11 | tmzt | raster: it should be safe to do whatever you want with recovery.img |
11:21.23 | tmzt | and leave boot.img alone to boot your phone |
11:21.26 | raster | shouldnt the msm7 on the g1 be 500 somthing mhz? (but its underclocked to save power) ? |
11:21.41 | tmzt | if you have hardspl or engboot installed you can still recover with fastboot |
11:21.51 | tmzt | you can also boot with fastboot by the way |
11:21.56 | tmzt | but not change cmdline |
11:22.08 | raster | i swear i tried that a while back and it failed miserably |
11:22.12 | tmzt | 526 I think |
11:22.28 | tmzt | tried what? |
11:22.33 | raster | i just had pain on de-locking my g1 |
11:22.39 | raster | tried fastboot |
11:23.32 | xperia | dcordes: no i dont know his right git adress. just the web adress that you have posted. i can remember that he told me that he hasnt a git server running so git cloning could from his webserver could be difficult. i can give you my whole git kernel but maybe it would be better to wait a week or max two weeks till aiv came back from the hollidays so we setup the repo with him together. P.S... |
11:23.34 | xperia | ...are the i2c adresses that you have posted really the like on the xperia. i cant see any values like 0x67 and 0x66 for the microp-ksc and micop-klt |
11:23.37 | tmzt | it won't work on stock spl |
11:24.04 | tmzt | you have to flash rc29/rc7 first, gain root, then flash |
11:24.15 | raster | did that |
11:24.24 | raster | and i eventually got my way to adp1.5 (cupcake) |
11:24.56 | dcordes | xperia, I would like to add them already. there are some people who like to add xperia patches too and it's not so nice to see everybody do the double work. don't see problem about adding it now. |
11:25.30 | dcordes | xperia, we should use the kovsky name, not venus. kovs is used for codename and all other machines use codename too. |
11:25.33 | raster | oh bollocks |
11:25.44 | raster | adp updates wiped hardspl - back to a adp1 dev spl |
11:25.50 | raster | ptoohey! |
11:25.53 | raster | aaarghhh |
11:25.55 | tmzt | ah |
11:26.03 | tmzt | but dev spl is okay too |
11:26.10 | tmzt | it's not locked |
11:26.34 | raster | fastboot works? i swear i treid it after i had hardspl on (needed it to get further) |
11:26.49 | raster | actually i had to do all sorts of magic goldcard poo |
11:26.52 | raster | that was not fun |
11:27.18 | tmzt | yeah, you had de model |
11:27.21 | tmzt | eu |
11:27.26 | xperia | okay tell me then the best way how to do this. should i email you the git patches. about venus i have the same meaning. everytime i extract a original rom from htc kovsky is displayed and not venus. i am asking if venus could be the code name for the cdma version i am using the gsm version. |
11:27.54 | tmzt | do you have cdma version? |
11:27.58 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
11:28.21 | xperia | no the gsm version and kovsky is displayed after extracting a original htc rom |
11:28.30 | raster | actually i have .sg model |
11:28.36 | raster | same as .au model (same telco) |
11:29.29 | tmzt | sg? |
11:29.44 | raster | so i guess... is there a ready-made recover.img that i can flash_image that will kexec a kernel from sd (and pass boot params etc. to boot root off sd) ? |
11:29.52 | raster | .sg - singapore |
11:30.01 | tmzt | ah, ok |
11:30.09 | dcordes | xperia, htc uses the same codename for cdma and gsm. the number after the codename identifies cdma/gsm . like raph100 raph800 |
11:30.14 | tmzt | well, I guess we can build one |
11:30.19 | raster | i'd like not to have to so a flash_image every time i futz with a kernel change |
11:30.23 | dcordes | I think there is no cdma kovsky |
11:30.30 | raster | major discouragement :) |
11:30.35 | dcordes | is there one? |
11:30.38 | tmzt | getting kexec to work on arm is not easy |
11:30.48 | tmzt | it shoudl work but I have yet to see it |
11:31.09 | raster | it'd be damned useful :) |
11:31.26 | tmzt | what kind of kernel changes do you need to make repeatedly? |
11:31.51 | raster | probably msm_fb fiddling |
11:31.57 | raster | adding icotls for accel ops |
11:32.04 | tmzt | not easy to avoid sadly |
11:32.18 | tmzt | we don't have that because of haret, but we still have to boot ce everytime |
11:32.34 | tmzt | I guess porting uboot is possible but that's a lot more work |
11:32.46 | xperia | decorde: well in usa there is a cdma network. south america has also cdma networks. i cant belive that xperia is only sold in europe. we need also to change a line in haret as there also stay venus other wise we cant boot the kernel. but this shouldn be a problem |
11:32.54 | raster | oh uboot will be a big one |
11:33.10 | raster | at least for android - a kexec capable recovery img is probably easier |
11:33.46 | tmzt | there's some source for the new chips here: |
11:33.48 | tmzt | https://www.codeaurora.org/patches/quic/qsd/ |
11:33.50 | tmzt | and |
11:37.39 | dcordes | xperia, do you have my email? |
11:37.47 | xperia | no |
11:38.10 | xperia | you can post it over private mail if you want on xda |
11:38.16 | raster | well one thing - fiddling with a 2nd os boot is going to be fiddly with recovery needing a reflash from android every kernel change |
11:38.20 | dcordes | what's your xda nickname? |
11:38.28 | raster | what are you building for userspace? |
11:38.28 | xperia | ARAN |
11:39.06 | tmzt | https://www.codeaurora.org/patches/quic/le/ |
11:40.30 | dcordes | xperia, thanks. pm sent |
11:42.22 | dcordes | 6,2455 |
11:42.47 | xperia | great. i will now start with the diffs and sent you over email in the next 30 to 45 mins |
11:44.31 | tmzt | raster: do you think if you flashed back the hspl it would work from fastboot? |
11:45.01 | tmzt | you just need an update.zip with hspl in it I believe |
11:45.26 | raster | dcordes: ooh wait. the mdp code has the actual stuff that does work |
11:46.31 | raster | tmzt: i'll try maybe tomorrow night |
11:46.38 | raster | it's time for food here |
11:46.41 | raster | maybe later tonight |
11:46.48 | raster | just looking at the mdp stuff |
11:47.07 | raster | MDP_DISPLAY_STATUS |
11:47.15 | raster | MDP_DISPLAY0_START |
11:47.18 | raster | MDP_DISPLAY1_START |
11:48.17 | raster | hmm what are these used for |
11:48.26 | raster | but theres definite stuff there for configuring up multiepl displays |
11:48.29 | raster | vsync and more |
11:50.50 | raster | ewww |
11:50.55 | raster | non-enumed addresses |
11:52.11 | tmzt | non-enumed? |
11:52.26 | raster | well ok - msm_fb.c and mdp.[ch] imply there is some decent set of 2d hw controls there |
11:52.30 | *** join/#htc-linux StarLite (n=nnscript@84-245-8-228.dsl.cambrium.nl) |
11:52.34 | raster | multi-colorspace video overlay it seems |
11:52.35 | tmzt | the mddi bus is like a serial bus connecting the remote framebuffer chip to msm |
11:52.36 | raster | dma/blit |
11:52.39 | raster | m mmm. |
11:52.42 | tmzt | I think that's what start is about |
11:52.52 | raster | <PROTECTED> |
11:52.56 | raster | <PROTECTED> |
11:52.58 | raster | etc. |
11:52.59 | tmzt | one of the mddi ports is used for camera actually |
11:52.59 | raster | for example |
11:53.14 | raster | smells of control registers to me |
11:53.25 | tmzt | yeah, we don't have details on those exactly |
11:53.34 | raster | thats going to be the problem |
11:53.59 | tmzt | did you see the uboot patches, they have much more detail for some reason |
11:54.02 | tmzt | than the android kernel |
11:54.09 | raster | one of those registers offset from MDP_CMD_DEBUG_ACCESS_BASE will be a control reg for moving the lcd out to the first fb |
11:54.12 | raster | seems by default its the 2nd |
11:54.20 | raster | oh havent looked |
11:54.25 | raster | need to make dins |
11:54.42 | raster | i'm just getting a quiz at the info for now to see whats up - but need to find info on those offsets |
11:54.54 | raster | you can spend a lot of time guessing at them and poking them to see what they do |
11:55.06 | raster | (thus the need to hack and try different kernels) |
11:55.10 | tmzt | we know the driver can switch scanout buffer |
11:55.21 | raster | aaah but how |
11:55.22 | raster | :) |
11:56.14 | raster | at this point i'd be asking "so.. qhere's the SoC programming manual docs - that'll have the register maps and info on what they do" :) |
11:57.23 | raster | on my s3c6410... i just consult the docs. :) |
12:01.09 | tmzt | well yes |
12:01.19 | tmzt | but we don't have those |
12:01.37 | tmzt | except the defines in the uboot patches, which is for a newer version of the chip |
12:01.55 | tmzt | there's some pdf somewhere also, but I don't know where |
12:02.25 | tmzt | you have the sgx docs also on that chip? |
12:03.37 | tmzt | <PROTECTED> |
12:03.37 | tmzt | <PROTECTED> |
12:03.37 | tmzt | <PROTECTED> |
12:03.37 | tmzt | <PROTECTED> |
12:03.37 | tmzt | <PROTECTED> |
12:03.39 | tmzt | <PROTECTED> |
12:03.57 | tmzt | static int send_blit(const struct mdp_info *mdp, struct mdp_blit_req *req, |
12:03.57 | tmzt | <PROTECTED> |
12:03.57 | tmzt | { |
12:03.57 | tmzt | <PROTECTED> |
12:03.57 | tmzt | <PROTECTED> |
12:03.59 | tmzt | <PROTECTED> |
12:04.02 | tmzt | <PROTECTED> |
12:04.05 | tmzt | <PROTECTED> |
12:04.07 | tmzt | <PROTECTED> |
12:04.10 | tmzt | <PROTECTED> |
12:04.12 | tmzt | is that what you wanted? |
12:05.26 | *** join/#htc-linux LanceHaig (n=lanceh@foresight/member/lhaig) |
12:06.27 | LanceHaig | I want to get an htc hero as my replacement to the tytn II is it possible to tether your laptop to the hero so you can use the internet? |
12:06.57 | tmzt | I believe so |
12:07.07 | tmzt | #android might help more with that |
12:08.12 | raster | ppp is the overlay i think |
12:08.37 | tmzt | I think it's used for everyting, at least we had to set DST to fix the color issue |
12:08.40 | raster | anyway |
12:08.41 | raster | foot |
12:08.43 | raster | food |
12:09.00 | raster | and that could be |
12:09.08 | raster | using an rgb "overlay" |
12:09.10 | tmzt | it's possible they aren't changing the address, but the coordinates |
12:09.12 | raster | likely it supports multipel layers |
12:09.18 | tmzt | within the framebuffer |
12:09.22 | tmzt | (which is double) |
12:10.41 | tmzt | <PROTECTED> |
12:10.41 | tmzt | <PROTECTED> |
12:10.41 | tmzt | <PROTECTED> |
12:10.41 | tmzt | <PROTECTED> |
12:10.41 | tmzt | <PROTECTED> |
12:10.43 | tmzt | <PROTECTED> |
12:10.46 | tmzt | <PROTECTED> |
12:10.49 | tmzt | <PROTECTED> |
12:10.51 | tmzt | <PROTECTED> |
12:11.53 | tmzt | <PROTECTED> |
12:11.53 | tmzt | <PROTECTED> |
12:11.53 | tmzt | <PROTECTED> |
12:11.53 | tmzt | <PROTECTED> |
12:11.53 | tmzt | <PROTECTED> |
12:11.56 | tmzt | <PROTECTED> |
12:12.06 | tmzt | guess not then, yres is not doubled, just vyres |
12:12.23 | tmzt | but: |
12:12.25 | tmzt | <PROTECTED> |
12:12.25 | tmzt | <PROTECTED> |
12:12.25 | tmzt | <PROTECTED> |
12:12.25 | tmzt | <PROTECTED> |
12:12.27 | tmzt | <PROTECTED> |
12:15.12 | tmzt | this should really be pdata: |
12:15.21 | tmzt | <PROTECTED> |
12:15.24 | tmzt | <PROTECTED> |
12:15.27 | tmzt | <PROTECTED> |
12:15.30 | tmzt | <PROTECTED> |
12:15.32 | tmzt | <PROTECTED> |
12:15.35 | tmzt | <PROTECTED> |
12:16.05 | *** join/#htc-linux Echo31 (n=olivier@mir31-4-82-240-194-54.fbx.proxad.net) |
12:22.38 | LanceHaig | tmzt: thanks I will shoot over to #android |
12:26.54 | *** join/#htc-linux raster_ (n=raster@124-170-251-164.dyn.iinet.net.au) |
12:29.42 | dcordes | xperia, ok thanks |
12:37.47 | xperia | dcordes: i will send you in the next 15 minutes a download link for my kernel working directory with the newest and latest changes. i have in this directory a script for building the kernel called "buildnewimage.sh" that do everything what is needed. in this directory there are also files for a animated bootsplash. |
12:37.49 | xperia | if you want to see my diffs do "git diff" in the extracted dir. |
12:37.51 | xperia | For uplading the repo to the git server do only "git push". |
12:37.54 | xperia | that is the most easy way for testing the build and uploading the git repo. |
12:38.56 | dcordes | xperia, what happened to git diff ? |
12:40.06 | xperia | well for building and testing the kernel thats the most easy way. you can still do git diff in the dir for getting the needed patches |
12:40.42 | dcordes | I think it's a bit more convenient than sending the entire tree. |
12:41.47 | xperia | so you want only the git diff.gz |
12:42.32 | dcordes | yes with the aiv + your changes |
12:43.02 | xperia | okay then i will stop the uplad of the dir |
12:43.05 | dcordes | optionally your diffs against aiv tree |
12:46.12 | dcordes | xperia, are you online in the CET evening? |
12:52.04 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
12:52.31 | xperia | yeah do you want later to come back |
13:11.29 | dcordes | xperia, yes so we can discuss kosvky patches here |
13:11.33 | dcordes | I have to go now see you |
13:16.06 | *** part/#htc-linux LanceHaig (n=lanceh@foresight/member/lhaig) |
13:16.35 | *** join/#htc-linux MethoS-- (n=clemens@host-091-097-246-050.ewe-ip-backbone.de) |
13:22.02 | xperia | tmzt are you still here or are you on the way to sleep :-) |
13:26.38 | *** join/#htc-linux gmv (n=georg@84.114.50.242) |
13:28.37 | *** join/#htc-linux UnknownArtists (n=unknown@84.19.198.64) |
13:29.40 | *** part/#htc-linux gmv (n=georg@84.114.50.242) |
13:36.59 | *** join/#htc-linux StarLite` (n=nnscript@188.90.203.253) |
13:43.04 | *** join/#htc-linux StarLite (n=nnscript@84-245-8-228.dsl.cambrium.nl) |
13:46.24 | *** join/#htc-linux StarLite` (n=nnscript@84-245-8-228.dsl.cambrium.nl) |
14:02.06 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
14:06.04 | xperia | dcordes: i have sent you the needed infos and patches |
14:40.15 | xperia | can somebody tell me the range of the i2c adresses |
14:45.21 | xperia | decordes: will be come back as i see that you are online here |
14:54.43 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
15:58.52 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
16:13.11 | *** join/#htc-linux Alex[sp3dev] (n=alex_dfr@p26-shn1.elcomnet.ru) |
16:25.56 | *** join/#htc-linux marex (n=marex@thor.hackndev.com) |
16:26.04 | *** join/#htc-linux ali1234 (n=al@robotfuzz.co.uk) |
17:00.58 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
17:13.44 | *** join/#htc-linux stickboy (n=anonymou@cpe-74-64-103-242.nyc.res.rr.com) |
17:40.48 | *** join/#htc-linux BabelO (n=fcr@unaffiliated/babelo) |
18:09.04 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
18:32.15 | *** join/#htc-linux PoohbaLT (n=BlkPoohb@c-69-141-71-249.hsd1.nj.comcast.net) |
18:41.33 | *** join/#htc-linux q_tpro (n=q_tpro@host-106-145-66-217.spbmts.ru) |
18:41.41 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
18:41.41 | q_tpro | hi |
18:50.36 | *** join/#htc-linux xperia (n=chatzill@77-56-95-46.dclient.hispeed.ch) |
18:51.19 | xperia | hello q_tpro. how are you. did you fix your internet problems |
18:51.55 | *** join/#htc-linux SuperJMN (n=jmn@233.Red-217-126-241.staticIP.rima-tde.net) |
18:54.00 | *** part/#htc-linux SuperJMN (n=jmn@233.Red-217-126-241.staticIP.rima-tde.net) |
19:10.08 | *** join/#htc-linux q_tpro_ (n=q_tpro@host-106-145-66-217.spbmts.ru) |
19:16.29 | xperia | still the same logout problem q_tpro :-) |
19:25.46 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
19:30.45 | q_tpro_ | xperia: it's better now |
19:31.28 | q_tpro_ | xperia: can you trace the i2c read/write registers ? then you will know the i2c addresses |
19:36.55 | xperia | hey qtpro nice that you are back. look i have tested inside the driver all adresses from 0x00 to 0xff |
19:37.54 | xperia | the variable "data->version" had allways one of this value here in the file microp-klt.c |
19:39.01 | xperia | 0x0000, 0xcfcf, 0x00cf, 0xffff, 0x00ff |
19:39.20 | xperia | there where no other values |
19:40.03 | xperia | i am asking myself really if the keyboard backlight and the other leds are really controlled by microp-klt |
19:41.24 | xperia | you must know one interessting thing about keyboard backlight on the xperia. if you slide out the keyboard the backlight is only going on if the light sensor that is positioned right to the front video camera |
19:41.55 | xperia | is covered up till the phone think it is night other wise ot wont go on |
19:42.29 | xperia | so keyboard backlight and light sensor of the front camera act together |
19:43.21 | xperia | to answer your question i dont know how to trace with the haretconsole the i2c adresses. if you could give me a tip this would help sure a lot |
19:49.54 | xperia | q_tpro you should try getting a bidirectional broadband satelit connection. this gprs connection is nothing. p.s another possibility is to communicate with me over mail my adress is admin (dot) gnucash (at) gmail (dot) com |
19:51.20 | xperia | it would be great if i can have your mail adress for exchanging information |
19:51.45 | xperia | you can post it also over private message to my xda user account ARAN |
19:52.23 | *** join/#htc-linux Amaranth_ (n=travis@74-221-38-55.longlines.com) |
20:02.29 | *** join/#htc-linux Amaranth__ (n=travis@74-221-38-55.longlines.com) |
20:08.34 | *** join/#htc-linux Amaranth (n=travis@74-221-38-55.longlines.com) |
20:12.34 | *** join/#htc-linux PoohbaLT1 (n=BlkPoohb@c-69-141-71-249.hsd1.nj.comcast.net) |
20:13.59 | *** join/#htc-linux chab7 (n=chab7@85.67.206.132) |
20:36.56 | *** join/#htc-linux Zoolooc (n=fredsiba@p549527F2.dip0.t-ipconnect.de) |
20:39.40 | ali1234 | can anybody help me with usb host? i've got it almost working on the wizard... the host controller registers etc. but wheni plug something in to it, it all goes crazy |
20:41.23 | xperia | i cant as i ddint had the time to study he code |
21:40.35 | *** join/#htc-linux Zoolooc_ (n=fredsiba@p549521E7.dip0.t-ipconnect.de) |
21:53.01 | *** join/#htc-linux leaigor (n=laigor@188.134.36.14) |
22:00.03 | *** join/#htc-linux lvh (n=lvh@83.101.76.209) |
22:00.06 | lvh | Hi :-) |
22:49.49 | tmzt | host? https://www.codeaurora.org/patches/quic/le/PATCH_u-boot-2009.03_1181_usb_keyboard_20090617.tar.gz |
22:51.03 | *** join/#htc-linux xfiles (n=xfiles@c-98-193-151-150.hsd1.ky.comcast.net) |
22:59.58 | *** join/#htc-linux darkstar62 (n=darkstar@97-126-107-190.tukw.qwest.net) |
23:16.10 | *** join/#htc-linux BabelO (n=fcr@unaffiliated/babelo) |