00:22.35 | *** join/#htc-linux g3gg0__ (n=g3gg0@ppp-82-135-4-70.dynamic.mnet-online.de) |
00:42.25 | *** join/#htc-linux g3gg0_ (n=g3gg0@ppp-62-245-211-203.dynamic.mnet-online.de) |
00:44.25 | drno | cr2: still awake? |
02:03.08 | *** join/#htc-linux LunohoD (n=alex@e180111136.adsl.alicedsl.de) |
02:06.13 | lkcl | Babel0: typecast iobase to a char* or preferably an int |
02:07.19 | lkcl | u16 *arr=(u16*) ( ((int)iobase) + 4 ); |
02:07.25 | lkcl | just to make sure. |
02:09.38 | lkcl | psokolovsky: yes, it was 0.12.4 - i also made a couple of minor changes: don't worry about them, trash them if necessary: i'll soon find them again. |
02:24.26 | *** join/#htc-linux bd2 (n=ufo@89.113.84.151) |
04:40.37 | *** join/#htc-linux RoEn_PC (n=roen@p54A65EFD.dip.t-dialin.net) |
05:35.10 | *** join/#htc-linux psokolovsky_ (n=psokolov@237.usernat.ip.net.ua) |
07:49.03 | *** join/#htc-linux goxboxlive (n=goxboxli@9.80-202-160.nextgentel.com) |
08:01.07 | *** join/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk) |
08:01.13 | *** join/#htc-linux Fossi (n=cmewes@195.177.48.115) |
08:05.55 | *** join/#htc-linux TripleDES (n=sergio@89.129.175.52) |
08:11.14 | *** part/#htc-linux teslamike (n=teslamik@CPE0011d864891e-CM0011e6c7aaa5.cpe.net.cable.rogers.com) |
08:31.08 | *** join/#htc-linux booba (n=booba@AAnnecy-256-1-94-95.w90-10.abo.wanadoo.fr) |
09:25.13 | *** join/#htc-linux psokolovsky_ (n=psokolov@237.usernat.ip.net.ua) |
10:37.12 | *** join/#htc-linux babel_work (n=Fabrice@faim1.itesoft.com) |
10:39.55 | babel_work | hi |
10:40.32 | babel_work | lkcl : thanks for the tips :) but nothing new with these lines |
13:16.56 | cr2 | babel_work: you don't read the output after writing to the registers ? i'll look for my old c code for universal. |
13:19.47 | lkcl | allo cr2 |
13:20.41 | cr2 | hi lkcl |
13:20.55 | lkcl | am still struggling with uda1380 sound... |
13:20.55 | cr2 | the ak4641 is a really fine chip. |
13:20.58 | lkcl | cool! |
13:21.00 | cr2 | :) |
13:21.22 | lkcl | so no surprise that it's in most of the latest htc phones, then |
13:21.27 | cr2 | i've looked at the block diagramm, and you can switch anything to everywhere. |
13:21.37 | lkcl | o good grief... |
13:21.54 | cr2 | 4 analog inputs, 3 analog outputs, 2 digital i/o. |
13:22.00 | lkcl | howling feedback from speaker to microphone, internally? |
13:22.01 | cr2 | all switchable. |
13:22.06 | cr2 | yes. |
13:22.13 | lkcl | good grief |
13:22.21 | cr2 | with attenuator controls everywhere. |
13:22.34 | lkcl | ak - is that Akai ? |
13:22.35 | cr2 | we can do a better job than wince |
13:22.37 | cr2 | yes. |
13:22.55 | cr2 | you can do very clever things with it. |
13:23.04 | lkcl | ahh, i think i had a sound module from them - very good one - so it wouldn't surprise me |
13:23.17 | cr2 | but i need to test my thoughts first. |
13:23.21 | lkcl | ack. |
13:23.27 | lkcl | btw you still got a blueangel? |
13:23.39 | cr2 | it's as if i was at htc, and designed the module. |
13:23.40 | cr2 | yes. |
13:23.48 | lkcl | he he |
13:23.54 | babel_work | hi cr2 |
13:24.02 | babel_work | hi lkcl |
13:24.06 | lkcl | allo babel |
13:24.07 | cr2 | what we need now is the real i2c mixer code for ak4641. |
13:24.30 | cr2 | the available one is only for hx4700, and is _very_ rudimentary. |
13:24.38 | cr2 | hi babel_work |
13:25.05 | lkcl | would you have some spare time to help me bash this uda1380 at some point? the interpolator isn't running because we haven't worked out how to clock it from the ASIC3 yet |
13:25.16 | lkcl | oh. quick question. |
13:25.25 | cr2 | maybe i should plot a switching diagramm. |
13:25.32 | lkcl | in asic3 dump, where is the clock speed for CX - of 32khz - coming from? |
13:25.43 | cr2 | lkcl: i'll print the uda1380 code and datasheet today. |
13:26.05 | cr2 | xtal on asic3 |
13:26.28 | cr2 | is it kHz ? |
13:27.35 | lkcl | kHz - yep. |
13:28.22 | cr2 | 32768 Hz ? |
13:29.08 | cr2 | imho, we can calculate the "external clock" frequency |
13:29.12 | lkcl | <PROTECTED> |
13:29.12 | lkcl | <PROTECTED> |
13:29.12 | lkcl | <PROTECTED> |
13:29.12 | lkcl | <PROTECTED> |
13:29.23 | lkcl | <PROTECTED> |
13:29.30 | lkcl | <PROTECTED> |
13:29.31 | lkcl | <PROTECTED> |
13:29.31 | lkcl | <PROTECTED> |
13:29.37 | cr2 | yeah, the dot (.) |
13:29.40 | cr2 | ok. |
13:29.41 | lkcl | where do those come from? |
13:29.42 | babel_work | cr2 : i ve add your code (without read register because value are output correctly from acx.ko), but nothing more |
13:31.57 | lkcl | ok i can look at gnu-haret/src/ |
13:31.57 | cr2 | babel_work: but if you have the right numbers, then the memory location is ok. |
13:31.57 | cr2 | lkcl: the only source of asic3 info is the ipaq-asic3.h header by CPQ. |
13:31.58 | cr2 | everything else is RE |
13:31.58 | lkcl | ok. |
13:31.58 | cr2 | well, i've added some RE data to ipaq-asic3.h |
13:32.08 | *** join/#htc-linux TripleDES (n=sergio@89.129.175.52) |
13:34.52 | lkcl | it's tricky. SACR bit2 is set (BCLK), SADIV is 0xd (01101). but the i2s is OFF. |
13:37.58 | lkcl | the uda1380 interpolator only responds when its sysclk or wsi or something are correctly set. |
13:42.05 | lkcl | cr2: you got wavedev.dll or the bootloader for blueangel somewhere? |
13:42.41 | lkcl | and if so would you mind doing a quick .lst for me so i can take a look at it? |
13:43.19 | babel_work | cr2 after car /proc/drivers/acx_...._diag, i saw rx register FULL. SO i think my next test is activate full acx-mem debug to see irq handler |
13:43.37 | babel_work | lkcl : i have this dll, you want it ? |
13:43.53 | lkcl | babel: sure. i can copy it to my server. |
13:44.15 | babel_work | i send it by mail :) |
13:44.43 | cr2 | lkcl: i'll do it on you server |
13:44.52 | lkcl | ta to you both. |
13:45.19 | cr2 | babel_work: i think it's some pcmcia setup bug. |
13:46.34 | babel_work | cr2 i remove all pcmcia and keep only power and reset pin and gpio17 for irq |
13:50.56 | lkcl | ok i gotta go shopping folks |
14:15.34 | cr2 | http://lkml.org/lkml/2006/12/13/342 |
14:15.55 | cr2 | i don't think we have any binary modules on arm. |
14:16.15 | cr2 | hmm. maybe g3, but we will take care of it. |
14:17.11 | cr2 | and this fcking sl5500 sd driver, but i'm mosty done with it. |
14:34.28 | cr2 | lkcl: what is 0xac44 ? |
14:35.16 | babel_work | 44100 khz ? non ? |
14:39.21 | cr2 | yes :) ok, now i get it. |
14:40.11 | cr2 | 0x30 is the i2c_addr >> |
14:40.27 | cr2 | 0x18<<1 |
14:52.59 | cr2 | blueangel_audio needs 3 clocks: http://wiki.xda-developers.com/index.php?pagename=ASIC3 |
14:53.28 | cr2 | CDEX_Source1, CDEX_CONTROL_CX and CDEX_EX1 |
14:53.49 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.231) |
14:56.02 | cr2 | #define CLOCK_CDEX_EX1 (1 << 14) /* R/W: 24.576 MHz crystal */ |
15:08.38 | *** join/#htc-linux apt (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
15:08.38 | *** topic/#htc-linux is HTC Linux Channel: Find logs at http://apt.rikers.org/%23htc-linux/ | please check http://handhelds.org/moin/moin.cgi/HTC_2dPhones | http://wiki.xda-developers.com/index.php?pagename=Xanadux |
15:16.55 | cr2 | lkcl: ~/ba/wavedev.lst |
15:34.00 | *** join/#htc-linux psokolovsky (n=psokolov@237.usernat.ip.net.ua) |
15:41.06 | *** join/#htc-linux oOZe (n=oOze@p54960B89.dip0.t-ipconnect.de) |
15:45.34 | *** join/#htc-linux asylumed (n=insanity@196.211.28.91) |
15:52.59 | oOZe | hello, i´ve got a question .. my Linux is not booting on my BlueAngel "kernel-panic connot mount rootfs" anyone kwows what that means? |
15:53.40 | Fossi | that the kernel can't mount the rootfs |
15:53.53 | asylumed | oOze: what are you booting off of? |
15:53.54 | oOZe | what can i do to fix it ? |
15:54.05 | Fossi | if you don't know what that means, you should prolly read up on linux before trying to run it on a cellphone ;p |
15:54.06 | oOZe | im booting from an 1 GB SD card |
15:54.10 | oOZe | ;) |
15:54.25 | asylumed | oOZe: you compiling from cvs? |
15:54.51 | oOZe | no, i downloaded the image from handhelds.org |
15:55.07 | oOZe | and followed the instructions there |
15:55.11 | *** join/#htc-linux Ralith (n=ralith@66.150.12.86) |
15:55.30 | asylumed | oOZe: kernel version? |
15:56.17 | asylumed | oOZe: http://www.gnulinux.biz/files/blueangel/people/rob_w/sd/ |
15:56.28 | oOZe | ups sorry |
15:56.57 | oOZe | thank u i will try this one |
15:59.34 | *** join/#htc-linux bd2 (n=ufo@89.113.84.151) |
16:01.34 | *** join/#htc-linux psokolovsky_ (n=psokolov@237.usernat.ip.net.ua) |
16:32.58 | *** join/#htc-linux JTRipper_ (n=jtripper@p54B1DEE8.dip.t-dialin.net) |
16:40.20 | cr2 | http://sourceforge.net/projects/qgnokii |
16:42.11 | *** join/#htc-linux pH5 (n=ph5@e178251178.adsl.alicedsl.de) |
16:44.46 | Hymie | cr2 / dudes: do you think hermes will boot successfully soon? If we can get the bastard booting, I might be able to help with some debugging of things.. once I have a linux environment to work from |
16:45.12 | Hymie | cr2: btw, how can you tell why it isn't booting? does garet fully support a serial console in the kernel as it tries to boot? |
16:46.14 | cr2 | similar ipaq ? |
16:46.21 | Hymie | well, it had a lot of the same hardware |
16:46.25 | Hymie | just a hope ;) |
16:46.43 | cr2 | to see something, you need either the right LCD parameters, or ... |
16:46.44 | Hymie | it was on some russian page, and it seemed to have the same CPU and a few other things |
16:46.55 | cr2 | wanted to write serial port, but there is nothing on hermes. |
16:47.22 | Hymie | ah, so we need full USB support to even start a kernel serial console.. and by then, the kernel has booted a lot I'm sure |
16:47.36 | Hymie | might as well use telnet at that point, heh |
16:47.50 | cr2 | we can live with other things, what we certainly need an LCD. |
16:47.59 | Hymie | yes, I guess so |
16:48.05 | cr2 | we can't support an ATI fb now |
16:48.13 | cr2 | so we need to adapt the vsfb |
16:48.21 | cr2 | and reuse the wince setup. |
16:48.26 | Hymie | hum |
16:48.35 | cr2 | the framebuffer address space is already known. |
16:48.36 | Hymie | too bad vesafb didn't just work out of the box |
16:48.50 | cr2 | it's not vesafb |
16:48.58 | Hymie | oh, I thought with vsfb it would be |
16:49.27 | Hymie | sorry |
16:49.27 | cr2 | vsfb != vesafb |
16:49.38 | cr2 | have you compiled the rx3715 ? |
16:50.01 | Hymie | no... so, first step for me is to get 2.6.19.x, and cross compile to rx3715? |
16:50.35 | cr2 | we need to start with some working s3c24xx port. |
16:51.01 | Hymie | ok, so many s3c24xx patches are not in the main kernel |
16:51.55 | cr2 | does not matter now. |
16:52.23 | cr2 | we need the minimal cpu setup, and to bolt the vsfb on it to see the penguin on the LCD. |
16:52.27 | Hymie | I see some talk on google on handhelds for this |
16:53.08 | cr2 | let's start with rx3715 |
16:53.34 | Hymie | http://www.handhelds.org/hypermail/kernel-discuss/19/1976.html <-- I see some talk here.. so people are working right now on some problems |
16:54.40 | Hymie | looks like there are ipaqs with the same cpu... I think that russian guy was trying on an ipaq with the same cpu... has anyone played with the sources from that page? |
16:55.00 | cr2 | we don't care about asic3 issues. |
16:55.37 | Hymie | just reading, learning |
16:55.49 | cr2 | ok. |
16:55.51 | Hymie | can you point me at someplace, then push me? |
16:56.22 | Hymie | http://handhelds.org/moin/moin.cgi/HpIpaqRx3715Kernel <-- you think these kernels are a good start? |
16:56.24 | Hymie | or any suggestions? |
16:57.00 | cr2 | forget about precompiled. |
16:57.15 | cr2 | i'll try the defconfig |
16:57.59 | cr2 | heh. there is no defconfig |
16:58.21 | Hymie | doh |
16:58.34 | Hymie | Kernel configuration file |
16:58.34 | Hymie | Put it in your kernel source dir, rename it to .config |
16:58.34 | Hymie | <PROTECTED> |
16:58.34 | Hymie | <PROTECTED> |
16:58.42 | Hymie | that's not def, but.... |
16:58.47 | Hymie | at least shows what they're using |
16:59.23 | cr2 | ok. |
17:00.05 | Hymie | just different --target? |
17:00.06 | Hymie | that's all |
17:00.07 | Hymie | ? |
17:00.12 | cr2 | arch/arm/mach-pxa/hx4700/Kconfig:70: unknown option "cool" |
17:00.17 | Hymie | eheh |
17:00.20 | Hymie | "cool" ;) |
17:00.26 | cr2 | somebody did not check his patches ;) |
17:00.47 | babel_work | cr2 nice startt qgnokii :) |
17:02.10 | cr2 | corrected |
17:02.43 | cr2 | babel_work: we can even remove the windows compatibility junk from there :) |
17:02.46 | cr2 | qpegnokii |
17:03.48 | cr2 | drivers/usb/gadget/Kconfig:474:warning: choice values currently only support a single prompt |
17:03.48 | cr2 | drivers/usb/gadget/Kconfig:455:warning: choice values currently only support a single prompt |
17:04.02 | cr2 | why is the poor pxausb is so fcked up... |
17:04.55 | *** join/#htc-linux oOZe_ (n=oOze@p549610FD.dip0.t-ipconnect.de) |
17:05.07 | babel_work | cr2 : yes, it is a good idea... but difficult to build opie/qtopia from scratch without OE :( .. |
17:05.27 | cr2 | without OE ? |
17:06.05 | cr2 | you only need to avoid the qt widgets not available in qpe |
17:06.17 | cr2 | and there are not that many of them. |
17:06.26 | babel_work | cr2 : yes not using openembedded, only with tgz files |
17:06.39 | cr2 | expect a big revamp with opie2 anyway. |
17:06.46 | babel_work | yes, i have made some dev with QT for mythtv |
17:06.56 | cr2 | we can write our own .bb file |
17:07.04 | cr2 | look at zroadmap package. |
17:08.21 | cr2 | arch/arm/mach-s3c2410/built-in.o: In function `s3c2410_pm_drvinit': |
17:08.21 | cr2 | mach-rx3715.c:(.init.text+0xb3c): undefined reference to `s3c2410_sysclass' |
17:08.40 | cr2 | pm is not that necessary. |
17:09.52 | *** join/#htc-linux psokolovsky_ (n=psokolov@237.usernat.ip.net.ua) |
17:09.53 | Hymie | cr2: looks like I need to compile binutils and gcc for the target? |
17:10.31 | babel_work | hymie : get this http://www.codesourcery.com/gnu_toolchains/arm/download.html |
17:10.37 | babel_work | add it to your PATH |
17:10.57 | babel_work | and use make CROSS_COMPILE=arm-none-linux-gnueabi- |
17:11.42 | rmoravcik | cr2: http://pastebin.ca/279040 |
17:12.07 | Hymie | babel_work: ok.. I'm using 64bit os, but I'll just chroot for this 32bit stuff |
17:12.11 | Hymie | not sure what it depends upon |
17:13.18 | babel_work | Hymie no depends i think, i use suse 10.1, no problem :) |
17:13.25 | cr2 | Hymie: i use the toolchain compiled in OE for angstrom. |
17:13.58 | cr2 | on dual amd64/8GB ram. but it is a multimegabyte setup. |
17:14.04 | Hymie | and what angstrom is ;) |
17:14.12 | cr2 | http://openembedded.org |
17:14.29 | cr2 | http://www.openembedded.org/wiki/GettingStarted |
17:14.50 | cr2 | http://angstrom-distribution.org |
17:15.26 | Hymie | ok, I go to read |
17:15.43 | cr2 | it's everything you'll ever need :) |
17:16.49 | babel_work | Hymie : if you have enougth disk space, it is the best solution |
17:25.58 | cr2 | lol. 14804480 ~/stuff |
17:26.09 | cr2 | a lot of junk there. |
17:26.40 | babel_work | cr2 : gpe + opie + bootstrap ? |
17:34.37 | cr2 | no gpe, no boostrap, but a lot of old opie. |
17:35.10 | cr2 | babel_work: have you already tried qgnokii ? |
17:35.45 | cr2 | hmm. how do i config vsfb ? |
17:36.18 | babel_work | cr2 : no, i ve started build on my home computer... and now going to my house to test it :) |
17:37.25 | cr2 | ok. |
17:43.14 | babel_work | cr2 : oh oh |
17:43.14 | babel_work | there are people very tired here:) |
17:43.37 | babel_work | i found this in some code today StrToInt('-'+IntToStr(nPrecision)) |
17:46.28 | cr2 | LOL |
17:46.42 | cr2 | c++ is evil :) |
17:47.10 | babel_work | it is borland delphi code :) |
17:51.00 | *** part/#htc-linux babel_work (n=Fabrice@faim1.itesoft.com) |
18:03.22 | cr2 | qgnokii gui is too simple :) |
18:04.31 | lkcl | allo darlins |
18:09.01 | *** join/#htc-linux RoEn_PC (n=roen@p54A65EFD.dip.t-dialin.net) |
18:11.14 | cr2 | lkcl: the samsung stuff in hh.org CVS is not in a great shape. |
18:17.52 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
18:18.53 | lkcl | cr2: oops... |
18:19.28 | pH5 | hey |
18:22.14 | *** join/#htc-linux goxboxlive (n=goxboxli@9.80-202-160.nextgentel.com) |
18:22.55 | BabelO | cr2: qgnokii... it is a very very basic interface :( |
18:34.42 | cr2 | BabelO: yes, i've seen. |
18:35.51 | cr2 | pH5: any progress with the uda1380 ? |
18:37.08 | cr2 | i'm thinking about writing an awk ak4641 command parser. |
18:37.29 | goxboxlive | hi cr2 |
18:37.49 | lkcl | ah ha. ok. i have something (from wavedev). |
18:37.54 | lkcl | i think it goes something like this: |
18:38.00 | goxboxlive | Gnokii wouldnt build. I guess we have to get support for the HTC phones before we can build it. |
18:38.06 | goxboxlive | hi lkcl |
18:38.13 | pH5 | hi goxboxlive |
18:38.14 | cr2 | hi goxboxlive |
18:38.15 | lkcl | allo goboxlive. |
18:38.20 | cr2 | 00e5e9c9: addr=0x24 command= MIC value= 0x1d |
18:38.20 | cr2 | 00e5f47c: addr=0x24 command= MIC value= 0x1c |
18:38.20 | cr2 | 00e5ff1c: addr=0x24 command= PGA value= 0x2c |
18:38.20 | cr2 | 00e62056: addr=0x24 command= BTIF value= 0x00 |
18:38.21 | cr2 | 00e62ad4: addr=0x24 command= MIC value= 0x1c |
18:38.21 | cr2 | 00e71d89: addr=0x24 command= PWR1 value= 0x93 |
18:38.24 | pH5 | cr2: not yet |
18:38.30 | cr2 | loook like that now. |
18:38.32 | pH5 | I'm waiting for lrg to reshuffle his soc code |
18:38.40 | cr2 | but i'd like to extract the bits. |
18:38.57 | cr2 | it seems that universal uses the ak4641 to its full extent. |
18:39.26 | cr2 | and you can record simply everything |
18:39.47 | cr2 | mic_ext, mic_int, aux_in, BT_in. |
18:40.26 | cr2 | goxboxlive: right now the ak4641.c driver has only mic_int hardcoded. |
18:40.29 | lkcl | set alt fn asic3-gpioA15 to 1 (very weird) |
18:40.40 | cr2 | lkcl: it is already ? |
18:40.56 | lkcl | gimme 1sec, cr2... |
18:41.12 | cr2 | it's in the same code block where the CDEX clocks are enabled. |
18:42.05 | cr2 | goxboxlive: i guess the working phone mic is a question of switching from mic_int to aux_in |
18:42.24 | *** join/#htc-linux pleemans (n=peter@d51A5E612.access.telenet.be) |
18:42.42 | cr2 | in the same way you can switch to mic_ext (headphone) and BT_in (BT headset). |
18:43.30 | cr2 | you only need to be careful what you switch/mix and setup the appropriate attenuator bits. |
18:44.04 | lkcl | set bit 14 of CDEX (aka EX1) |
18:44.35 | cr2 | yes, i've already patched blueangel_audio some time ago. |
18:45.20 | lkcl | sleep for 1ms |
18:45.35 | lkcl | i gotta go through it, cr2, because we're missing something. |
18:46.03 | lkcl | so far it looks like they're setting alt fn asic3-gpioA15 to 1 - not set asic3-gpio to 1. |
18:46.11 | goxboxlive | cr2: Ok, seems like you have full control now with the sounddriver now :-) |
18:46.15 | lkcl | it could be that the switching is relevant |
18:47.21 | cr2 | goxboxlive: i'm only wondering what is the purpose of this at@audioset stuff. |
18:47.46 | cr2 | if everything goes through aux_in / mono_out |
18:48.03 | goxboxlive | Maybe it has something to do with the "button" on the headset? |
18:48.24 | cr2 | don't know. |
18:48.40 | cr2 | at@talk is not very useful |
18:48.45 | lkcl | then, ASIC3 clock clear bit 0 and set bit 1 |
18:48.58 | cr2 | it reports the in/out call times. |
18:49.07 | cr2 | lkcl: where ? |
18:49.24 | lkcl | .text:015B99C8 |
18:50.04 | goxboxlive | cr2: The button has two operations AFAIK. Answer a call (push the button), and activating a voice speed call (hold the button for 2-3 sec.). |
18:50.11 | lkcl | _then_ set CONTROL_CX |
18:50.18 | lkcl | i mean - it's just plain ... _weird_! |
18:50.27 | cr2 | lkcl: ? |
18:50.30 | lkcl | yes cr2? |
18:50.38 | lkcl | what dude? |
18:50.39 | goxboxlive | cr2: Well, maybe AT£TALK is the voice speed dialing command. |
18:51.08 | cr2 | goxboxlive: at@talk is time. voice speed call, i don't know what it is and how to use it :) |
18:52.01 | cr2 | lkcl: it's the setting of the CDEX_Source1 bit |
18:52.13 | lkcl | yep. |
18:52.22 | cr2 | clear two lower bits (0xfffc) and then | 0x2 |
18:52.27 | lkcl | and importantly, clearing Source 0. |
18:52.32 | cr2 | yes. |
18:52.33 | lkcl | *sigh* |
18:52.50 | cr2 | which brings me to the following |
18:53.05 | lkcl | but the timing is important. that 1ms delay corresponds with the stuff in the uda1380 spec about doing a reset whilst the clock is enabled. |
18:53.24 | cr2 | ex0 is 32768 Hz, ex1 is 24576000 Hz |
18:53.38 | lkcl | yerrrs? |
18:53.45 | cr2 | if source1 is ex1, and the divisor is 768 |
18:53.57 | cr2 | then you get 32000 Hz for your sound clock. |
18:54.00 | cr2 | weird. |
18:54.16 | lkcl | that's ... 32000 ... yep. oh well. |
18:54.19 | lkcl | 32000/8. |
18:54.29 | lkcl | oops. meant to type that @ python prompt :) |
18:54.35 | cr2 | what does the .12 code has for the external clock frequency ? |
18:54.47 | lkcl | i have _no_ idea. |
18:54.50 | cr2 | 32/8=4 |
18:54.55 | lkcl | yeh. |
18:55.15 | goxboxlive | cr2: Voice speed dial = Lets say i have your name in my contact book on phone linking to your phone number. I record a voice (let's say 'cr2') and link it to your name and phone number in my phonebook. When i call you all i have to do is, hold the button in for 2-3 sec. and after a short beep i say your name. It will recognize "cr2" and automaticly call your number. |
18:55.49 | lkcl | The DAC part of the UDA1380 can operate from 8 to 100 kHz sampling frequency (fs) |
18:55.54 | lkcl | that's from the uda1380 spec. |
18:56.08 | cr2 | it should produce the 44100 bitclk with default settings. |
18:56.27 | cr2 | goxboxlive: where do you "say" the name ? |
18:56.51 | cr2 | and any mic ? (int/ext/bt) ? |
18:57.11 | goxboxlive | cr2: It's great to have when you are driving a car, you dont have to touch the phone. Just hit the button on the headset/Bluetooth headset / or the vioce dail buttoin on the side of the phone. |
18:57.23 | pH5 | with the uda1380's divider settings that lkcl posted, sysclk should be 768*44100 |
18:57.48 | cr2 | pH5: the ssp on ba is set to 16bit MSB, like the i2s |
18:57.54 | goxboxlive | If you use headset, then the headset mic is used, if no headset the mic in the device is used. And with BT headset the mic on the BT headset is used. |
18:59.16 | cr2 | goxboxlive: that means the voice is sampled, transfered over I2S to pxa27x, recognized somehow, and the ATDT12345678 is send to the phone ? |
19:00.05 | cr2 | all can be done in userspace. the key is the voice recognition. |
19:01.20 | cr2 | answering the call with the button is another issue. i don't know how it is handled . maybe i should try to call, run wasic3gpio , and press the button. |
19:02.12 | lkcl | cr2: any chance you could double-check this weirdness for me: .text:015B9990 |
19:02.38 | lkcl | looks to me like it's setting ASIC3 A alt-function |
19:02.55 | goxboxlive | cr2: Yes i guess that is the routing . |
19:03.03 | cr2 | <PROTECTED> |
19:03.03 | cr2 | <PROTECTED> |
19:03.11 | cr2 | pH5: val =44100 |
19:03.11 | lkcl | bit 15 |
19:03.17 | cr2 | lkcl: looking. |
19:03.21 | lkcl | ta |
19:03.54 | cr2 | pH5: SSPEXTCLK/(2*(div+1)) is the ssp clock |
19:05.38 | cr2 | lkcl: 1. set gpioa15 to alt |
19:05.57 | cr2 | CDEX_EX1 on |
19:06.06 | lkcl | yeh, that's what i thought. ok now need to find where it's cleared, just to confirm. |
19:06.10 | cr2 | msleep(1) |
19:06.28 | cr2 | clear CDEX source1 and source0 |
19:06.35 | cr2 | set CDEX_source1 |
19:07.19 | lkcl | oh. it's clear them _both_ first, then set source 1. |
19:07.21 | lkcl | argh |
19:07.47 | cr2 | set CONTROL_CX |
19:08.03 | cr2 | that's all |
19:08.20 | lkcl | no, it's clear source0 and set source1 simultaneously |
19:08.25 | lkcl | then set control_cx. |
19:08.50 | psokolovsky_ | Hi! |
19:08.52 | lkcl | dang. |
19:08.55 | lkcl | allo psoko. |
19:09.01 | cr2 | source1 is not cleared. |
19:09.09 | psokolovsky_ | Anybody worked already on cleaning up tsc2046 dirt around? |
19:09.10 | cr2 | only EX1 and CONTROL_CX |
19:09.36 | lkcl | we're working out the uda1380 initialisation/reset sequence for blueangel. |
19:09.45 | cr2 | psokolovsky_: i've tried, but it is still messing with the pxa SSP, so it's pxa-specific. |
19:09.48 | lkcl | cr2 gave it a start, psoko. |
19:09.52 | cr2 | htcuniversal_ts2.c |
19:10.07 | cr2 | but we will also need it on hermes = s3c24xx |
19:11.34 | psokolovsky_ | cr2, sure. but at least it seems to be the same for all machines, so it's good first step to factor out SSP handling out of TS drivers. So, you or someone else don't have any code? |
19:11.36 | cr2 | psokolovsky_: the main problem is the 24bit spi. i have set it to 16bit, but it was really unusable, then 32bit, it was better but there was still some bug the the decoding of the samples. |
19:11.53 | cr2 | look at htcuniversal_ts2.c |
19:12.08 | psokolovsky_ | cr2, so, univ != hx4700? |
19:12.25 | cr2 | irq on asic3, different SPI port. |
19:12.41 | psokolovsky_ | ok ;-\ next time then ;-) |
19:12.44 | cr2 | the rest is the same. and i have cleaned up the bits. |
19:13.17 | psokolovsky_ | I thought it's just exactly the same code in ~4 ports now... |
19:13.48 | cr2 | psokolovsky_: if you'll set your irq, your ssp port and the pen_down() then it will work. |
19:13.48 | psokolovsky_ | ok |
19:13.58 | cr2 | look at the code first, and then jump to the conclusions :) |
19:14.38 | cr2 | struct tsc2046_mach_info { |
19:14.42 | cr2 | <PROTECTED> |
19:14.43 | psokolovsky_ | I just grepped around, and so crap like (1 << 20) in bunch of places ;-) |
19:14.46 | cr2 | <PROTECTED> |
19:14.50 | cr2 | <PROTECTED> |
19:14.54 | cr2 | <PROTECTED> |
19:14.58 | cr2 | <PROTECTED> |
19:14.58 | cr2 | <PROTECTED> |
19:14.59 | cr2 | }; |
19:14.59 | cr2 | #define TSC2046_SAMPLE_X 0xd0 |
19:14.59 | cr2 | #define TSC2046_SAMPLE_Y 0x90 |
19:15.01 | psokolovsky_ | cr2, thanks for hint, will review later |
19:15.03 | cr2 | these are the parameters. |
19:16.02 | psokolovsky_ | cr2, by "few places" I mean nor htcuniversal/, but mach-pxa/ |
19:16.10 | cr2 | yes. |
19:16.18 | cr2 | and it's evil. |
19:16.39 | cr2 | you only need to fill the tsc2046_mach_info |
19:17.57 | psokolovsky_ | ok |
19:25.13 | cr2 | pH5: if it is stereo, then the SSPEXTCLK is 2822400 Hz ? |
19:28.04 | lkcl | ok gotta go shopping: back later |
19:31.15 | pH5 | that would be 64*sample_rate. in slave mode the uda1380 accepts bitclocks from 32 to 128*sample_rate |
19:38.55 | *** join/#htc-linux Fossi (n=Fossi@e176112082.adsl.alicedsl.de) |
19:43.51 | BabelO | cr2: i ve activated full debug for acx |
19:51.17 | goxboxlive | How do i get the keycodes ? Isent it dumpkey? I just want the keycode for the buttons i press, not the wholes list |
19:52.52 | cr2 | showkey |
19:53.16 | goxboxlive | ahhh thx |
19:53.47 | goxboxlive | btw. DId you know that gpe is showing the right time? Opie doesnt. |
19:54.00 | cr2 | btw, why do you have only english in the opie ? |
19:54.04 | *** join/#htc-linux ellisway (n=ellis@host-87-74-241-174.bulldogdsl.com) |
19:54.15 | cr2 | time needs the rtc support ? |
19:54.16 | goxboxlive | cr2: Yes |
19:54.41 | cr2 | i have no idea where opie gets the time :) |
19:55.09 | goxboxlive | I know where Opie gets the time |
19:55.27 | cr2 | where ? |
19:55.52 | goxboxlive | It shows the time form when it was build, also the date. But GPE is showing the right time. (in the right corner) |
19:58.16 | cr2 | oh. i've already missed 140 oe revisions :) |
19:58.24 | goxboxlive | :-) |
20:02.20 | lkcl | allo am back |
20:32.50 | goxboxlive | psokolovsky_ Are you around? |
20:51.30 | goxboxlive | cr2: Why is camera, light and "red hangup" button on the side of lcd screen keycode 68 ? |
20:55.05 | *** join/#htc-linux rob_w (n=bob@p85.212.174.68.tisdip.tiscali.de) |
20:55.56 | cr2 | static struct pxa27x_keyboard_platform_data htcuniversal_kbd |
20:56.19 | cr2 | light ? |
20:57.30 | goxboxlive | maybe |
20:57.39 | goxboxlive | the light button |
20:57.55 | goxboxlive | beside the volume slider |
20:58.37 | cr2 | hmm. |
20:59.33 | cr2 | { KEY_F10, GPIOA_BUTTON_BACKLIGHT_N,1, "backlight_button"}, |
20:59.33 | cr2 | { KEY_F24, GPIOA_BUTTON_RECORD_N,1, "record_button"}, |
20:59.33 | cr2 | { KEY_F10, GPIOA_BUTTON_CAMERA_N,1, "camera_button"}, |
20:59.42 | cr2 | KEY_F10 ;) |
20:59.50 | goxboxlive | :-) |
20:59.53 | cr2 | i've just copied the old data. |
21:00.13 | goxboxlive | i am trying to make a xmodmap |
21:00.24 | cr2 | what _should_ the camera send ? |
21:00.37 | cr2 | and the phone buttons ? |
21:00.48 | goxboxlive | Camera = keycode 68 |
21:01.15 | cr2 | what "qtopia phone whatever" sends ? |
21:01.25 | goxboxlive | yes |
21:04.11 | cr2 | i'll try to add suspend/resume to pxa27x_keyboard. |
21:05.21 | goxboxlive | good |
21:07.05 | cr2 | compiling |
21:16.26 | goxboxlive | cr2: have you tried it yet?? |
21:26.03 | *** join/#htc-linux g3gg0_ (n=g3gg0@ppp-82-135-7-126.dynamic.mnet-online.de) |
21:27.07 | BabelO | cr2 : i want to see register mcmem0 and mcmem1 at adress 0x48000028 |
21:27.24 | BabelO | is this exact |
21:27.26 | BabelO | HaRET(4)# pd 0x48000028 0x8 |
21:27.26 | BabelO | 48000028 | 0001c416 00000000 | ........ |
21:27.26 | BabelO | H |
21:28.16 | BabelO | for register mcmem0 ? |
21:28.43 | cr2 | goxboxlive: i've put it in CVS |
21:28.49 | goxboxlive | did it work? |
21:29.05 | *** join/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk) |
21:29.07 | cr2 | don't know :) but it compiles without any warnings. |
21:29.51 | cr2 | rmoravcik: why can't i compile the rx3715 code in hh.org CVS ? can you point me to a working .config ? |
21:29.56 | goxboxlive | ok, the i am compiling and try it out. :-) |
21:30.20 | rmoravcik | cr2: do you mean problem with pm? |
21:30.36 | cr2 | BabelOued: what are these registers ? |
21:30.52 | cr2 | rmoravcik: if i disable the pm, i get other linking problems. |
21:31.06 | goxboxlive | <PROTECTED> |
21:31.06 | goxboxlive | <PROTECTED> |
21:31.06 | goxboxlive | kernel/module.c: In function ‘module_add_driver’: |
21:31.06 | goxboxlive | kernel/module.c:2282: warning: ignoring return value of ‘sysfs_create_link’, declared with attribute warn_unused_result |
21:31.18 | rmoravcik | cr2: i solve this problem with http://pastebin.ca/279040 |
21:31.31 | cr2 | goxboxlive: it's not my code :) |
21:32.16 | cr2 | rmoravcik: ok, i'll try. |
21:32.20 | rmoravcik | cr2: http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2006-November/036929.html |
21:33.03 | cr2 | ok. |
21:33.21 | cr2 | but i think you should add rx3715_defconfig to CVS. |
21:33.33 | rmoravcik | cr2: http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2006-November/036992.html |
21:34.41 | cr2 | rmoravcik: post it to the k-d list ? |
21:34.51 | rmoravcik | cr2: i want add it, but first i want add udc, touchscreen driver |
21:35.11 | *** join/#htc-linux rob_w (n=bob@p85.212.169.250.tisdip.tiscali.de) |
21:35.44 | rmoravcik | cr2: if you want |
21:37.27 | goxboxlive | cr2: No it still doesnt resume. |
21:38.07 | cr2 | i'm not so well informed about the samsung support, so it's better to hear what others think about it. |
21:38.32 | cr2 | goxboxlive: i have added the code that was available on other devices. |
21:38.58 | cr2 | goxboxlive: does the irq counter increase ? |
21:39.11 | goxboxlive | hmm, well i still doesnt resume thow. Why is that. The driver is resuming, but the keyboadr is just dead. |
21:39.15 | goxboxlive | I'll check |
21:39.30 | rob_w | ahh my new dsl router arrived ,, hopefully that is the end of broken down connections .. |
21:39.46 | cr2 | :) |
21:40.13 | cr2 | goxboxlive: why does bitbake-image compiles the .18-hh0 kernel ?? |
21:41.18 | goxboxlive | cr2: because of the ochi-hcd is set as a |
21:41.24 | goxboxlive | is not set as a module |
21:41.46 | goxboxlive | I forget to change the htcuniversal_defconfig before i made it. |
21:42.02 | goxboxlive | I have 19-hh5 now |
21:43.11 | cr2 | ok, tell me if the keyboard irqs are coming through after resume. |
21:44.12 | goxboxlive | root@htcuniversal:~$ cat /proc/interrupts |
21:44.12 | goxboxlive | <PROTECTED> |
21:44.12 | goxboxlive | <PROTECTED> |
21:44.22 | goxboxlive | o they dont |
21:44.47 | goxboxlive | But could it be related to the usb0 is not resuming either? |
21:45.00 | cr2 | unlikely. |
21:45.10 | goxboxlive | ok |
21:45.53 | *** join/#htc-linux pH5_ (n=ph5@e178251178.adsl.alicedsl.de) |
21:46.08 | cr2 | goxboxlive: have you tried to boot the kernel with Kevin's 'boot by resume' ? |
21:46.53 | *** join/#htc-linux rob__w (n=bob@p85.212.157.245.tisdip.tiscali.de) |
21:49.55 | goxboxlive | No, not for a while, Has a meade a new Haret version? |
21:52.45 | rob__w | hey we broke the 40 users , hurray ! |
21:53.27 | cr2 | goxboxlive: no, this feature was available for some time. |
21:54.04 | BabelO | cr2: these register are for pcmcia Socket 0 timing config |
21:54.07 | goxboxlive | yes, it didnt wake up. It suspended, but it didnt resume |
21:56.17 | *** join/#htc-linux FossiFoo (n=Fossi@e176112082.adsl.alicedsl.de) |
21:56.50 | cr2 | ok, then we need to debug it. |
22:04.18 | cr2 | good night |
22:07.18 | goxboxlive | good night |
22:09.35 | pH5_ | good night |
22:18.53 | *** join/#htc-linux rmoravci1 (n=rmoravci@pc-nbhcnceqpo8wov6n0m468h4u6ax6am4.users.student.utc.sk) |
22:31.46 | *** join/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk) |
22:37.44 | *** join/#htc-linux Ralith (n=ralith@66.150.12.86) |
22:55.52 | *** join/#htc-linux ralith_ (n=ralith@66.150.12.86) |
23:36.04 | *** join/#htc-linux Ralith (n=ralith@soggy202.drizzle.com) |