00:04.23 | cr2 | somebody still around ? back to hacking :) |
00:12.58 | rob_w | yep |
00:13.14 | rob_w | 1:16 here ..cu all |
00:22.06 | lkcl | a |
00:22.22 | lkcl | already so soon, cr2? oh - he's gone hacking |
00:25.37 | iclem | can't get this himalaya to charge under linux |
00:28.37 | cr2 | can't link with gx.lib |
00:30.24 | cr2 | found this software http://www.geekshop.be/rien/lcab |
00:39.01 | lkcl | lessavaquicklook... |
00:43.33 | cr2 | have only three problems left: |
00:43.46 | cr2 | out/video.obj:video.cpp:(.text+0x68): undefined reference to `GXBeginDraw()' |
00:43.56 | cr2 | out/output.obj:output.cpp:(.text+0x6c0): undefined reference to `CreateDialogParamW' |
00:44.05 | cr2 | out/script.obj:script.cpp:(.text+0x4c): undefined reference to `isspace' |
00:44.50 | cr2 | the last two are the bugs in includes and can be worked around. |
00:46.29 | cr2 | gx.lib is found. if i write foobar.lib, then i get an error. hmm... |
00:48.06 | cr2 | there was something fishy about underscores, but i don't see them with nm. |
00:49.09 | cr2 | it is probably easier simply to get rid of this GAPI stuff ;) |
00:58.53 | cr2 | are these really equivalent : ?GXBeginDraw@@YAPAXXZ void * GXBeginDraw(); |
00:59.11 | cr2 | lkcl: do you know how to use c++filt ? |
00:59.12 | lkcl | oh yeh - i think i cut those out, too :) |
00:59.24 | lkcl | naah. |
00:59.36 | lkcl | i cut out the draw stuff entirely. |
00:59.47 | lkcl | couldn't be bothered with the penguin crap. |
00:59.51 | lkcl | or the __try stuff. |
00:59.57 | lkcl | if it doesn't work, so what? :) |
01:00.01 | cr2 | very wise, indeed. |
01:06.36 | cr2 | i don 't get it. why arm-wince-pe-c++filt can't demangle the m$ symbols??? |
01:07.08 | lkcl | probably because nobody's tried it before. |
01:07.29 | lkcl | you have to be _very_ specific in your symbol naming conventions on c++ linking. |
01:08.05 | lkcl | dude you are _much_ better off just running the wince MS arm cross-compiler under wine. i did that in about 30 mins. |
01:11.43 | cr2 | but it also has linking problems... |
01:13.42 | lkcl | i encountered no such problems - it just worked. |
01:14.05 | lkcl | ii libwine 0.9-1 Windows API Implementation (Library) |
01:14.05 | lkcl | ii libwine-dev 0.9-1 Windows API Implementation (Development file |
01:14.05 | lkcl | ii wine 0.9-1 Windows API Implementation (Binary Loader) |
01:14.05 | lkcl | ii wine-doc 0.0.20040121-1 Windows Emulator (Documentation) |
01:14.09 | lkcl | ii wine-utils 0.9-1 Windows API Implementation (Utilities) |
01:14.11 | lkcl | ii winesetuptk 0.7-1 Windows Emulator (Configuration and Setup To |
01:14.16 | lkcl | that's under debian. |
01:14.18 | lkcl | which version are you using? |
01:14.47 | lkcl | the handhelds.org wiki HARET page does warn you that if you're not using a recent version of wine, to expect problems, and to upgrade to avoid the problems. |
01:17.47 | cr2 | heh. it's just not true. i had no problems with the old versions of wine. i 'm tried 0.9 and 0.9.4 on x86 and x86_64 and got the same error. |
01:18.20 | cr2 | maybe i should look for the really old version and compile it. |
01:21.50 | lkcl | hmmm... |
01:21.56 | lkcl | how odd. |
01:22.29 | cr2 | winedump sym "?GXSuspend@@YAHXZ" |
01:22.39 | cr2 | int __cdecl _global_GXSuspend_1(void) |
01:22.57 | lkcl | okay: i'm going to upload the wince thingy i hacked together to http://hands.com/~lkcl/haret (in a minute). |
01:23.11 | lkcl | it'll include a typescript file |
01:23.52 | lkcl | it definitely works - but just takes forever heeeeave, heeeave oh _drat_ wrong one, i edited it and left it in an uncompileable state :) |
01:25.18 | lkcl | yes - wine is known for being able to run specific programs but not others. |
01:25.37 | lkcl | the reason is simple: there are bugs in applications that were developed for specific versions of win32. |
01:25.45 | lkcl | and they worked at the time. |
01:25.53 | lkcl | _but_ of course, things get upgraded... |
01:25.57 | lkcl | OH - yes: i remember now. |
01:26.09 | cr2 | ok. let me know when the file is there. |
01:26.12 | lkcl | you need to shorten the number of things on the linker command-line |
01:26.14 | lkcl | :) |
01:26.31 | lkcl | if there are too many on the wine command-line, it cuts them off. |
01:26.34 | lkcl | that was it. |
01:27.14 | cr2 | burn them these developers |
01:27.31 | lkcl | okay, it's there. |
01:29.03 | cr2 | thanks. |
01:30.29 | cr2 | i think it is intentional. you can't link with another library if you can 't demangle the names. wine knows how to do it (lgpl software) and binutils - not. |
01:35.05 | lkcl | okay. we got visitors, they're being thingy. cranky cos there's not enough places for everyone to sleep. oops :) |
01:35.15 | lkcl | better go sort them all out... |
01:36.19 | cr2 | hey, you are compiling with evc3! |
01:44.30 | iclem | night |
01:53.06 | lkcl | yes? |
01:53.25 | lkcl | what it is is that you have to remove some of the additional (redundant) compile options from the linker line. |
01:53.45 | lkcl | that makes the whole line short enough for it to fit into whatever command-line buffers are too bloody short. |
01:53.51 | lkcl | under wine. |
01:54.01 | lkcl | and then it works. |
01:54.51 | lkcl | hey, the article on linuxdevices is going up in a few minutes... |
01:59.22 | cr2 | linuxdevices ? |
01:59.33 | cr2 | MontaVista's Mobilinux 4.0 gains SD/SDIO/MMC support |
02:05.55 | lkcl | http://linuxdevices.com/articles/AT7937511405.html |
02:05.57 | lkcl | YAAY! |
02:06.12 | lkcl | yeh. the article describing the legal minefield is very useful. |
02:06.19 | lkcl | basically: we're stuck with reverse-engineering. |
02:06.35 | lkcl | although, m-sys.com's web site now contains a _lot_ of downloadable appnotes |
02:06.53 | lkcl | including one for the diskonchip 32Mb and 64Mb millenium plus... |
02:23.28 | cr2 | you are a good journalist :) some technical inaccuracies here and there, but the text is well understandable by a "normal" reader. |
02:26.09 | lkcl | ta. |
02:26.20 | lkcl | any suggestions, because we have an opportunity to correct them. |
02:26.42 | lkcl | yes, it's a skill i deliberately cultivated. i really meant it about it being my vocation. |
02:27.53 | cr2 | easy ones: UTMC -> UMTS |
02:29.53 | lkcl | duh :) |
02:30.24 | lkcl | any more? |
02:30.38 | cr2 | rereading. |
02:34.45 | cr2 | the european model Motorola A780 is a great phone, it has GPS and is bundled with the NATIVE linux routeplanner from ALK (copilot) with maps of all western Europe/Scandinavia. |
02:35.17 | cr2 | you mention the really outdated model. |
02:36.40 | lkcl | oh yeh. |
02:38.16 | cr2 | i don't think we will need a couple of years to run Linux on the universal. partly thanks to the A780 and its source code availability. |
02:41.49 | lkcl | oh yeh??? _excellent_! where, where? it has similar hardware?? |
02:42.20 | lkcl | well... i'm being pessimistic. another success, another article. hurrah :) |
02:42.24 | cr2 | there is also a "security" observation. in spite of all whining about signed drivers, trusted API and such, we can run haret without doing anything. just install and run. ultimate virus dreamworld. |
02:43.58 | lkcl | que? i no unnerstand. oh. yeh. haret. "do you want to run haret at least once, yes/no?" |
02:44.12 | lkcl | hack it - hack the registry, bypass the question. |
02:44.46 | lkcl | i mean... the tools on forum.xda-developers.com _provide_ all the information necessary for viruses to unlock the ROM... |
02:45.51 | cr2 | for 2.4.20 look at http://sf.net/projects/e680 . For the 2.6.14 hacker kernel i will write Harald Welte from the openezx.org fame. I talked with him in Berlin about it. |
02:46.58 | lkcl | _excellent - is that on the Universal wiki pages? |
02:47.32 | cr2 | haret calles the functions that are "trusted" since wince 3.0 without having any certificate. |
02:49.04 | cr2 | A780 kernel has the camera drivers (not available anywhere else), and communicates with the GSM over USB OTG, like universal. |
02:49.55 | cr2 | it has pxa270 processor, so the drivers controlling the CPU voltage, speed, SRAM, etc. are also there. |
02:50.32 | lkcl | hey cr2, can you do a slashdot article-post, too? |
02:50.34 | cr2 | wiki ? not really. i have written some word about it on the xanadux wiki page. |
02:50.46 | lkcl | slashdot hate me :) |
02:51.22 | lkcl | rick has done some updates from suggestions from you - thanks. |
02:54.17 | cr2 | slashdot is a US site, and the A780 with GPS is not sold there. i have read that Bruce Perens has bought this phone. Look at the http://openezx.org site and wiki |
02:56.09 | lkcl | oooo |
02:57.51 | cr2 | i think that the really hot thing is this Qt version of copilot from ALK. it' s not some java crap, but a fully ported and good working program. the windows version costs really arm and leg, and here it is bundled and the maps are available. |
03:03.21 | lkcl | wow. |
03:03.46 | lkcl | worth trashing gpe and going with the other thing. the qt-based one. i always hated GTK, anyway :) |
03:05.07 | cr2 | qomunicator :) |
03:07.27 | cr2 | the phone app on A780 looks like gomunicator. the program interface on A780 is rough and simple. i really like it. but it is not free software. |
03:08.13 | lkcl | yeh, damn qomummymunicator |
03:08.39 | lkcl | i mentioned e680 links on a couple of universal pages, so as not to forget them. UniversalResearch. |
03:09.22 | lkcl | 3am - urk. |
03:09.43 | cr2 | 4am :) i will go to bed. good night! |
03:10.07 | lkcl | night cr2. |
03:22.13 | lkcl | http://wiki.xda-developers.com/index.php?pagename=Himalaya103BootloaderRescue |
03:22.29 | lkcl | i _thought_ there was a command to write to the SD card in the bootloader. |
03:22.32 | lkcl | d2s it's called. |
04:17.40 | lkcl | oh _maaan_. xilinx provided GPL code for the IIc device, to motorola. |
04:18.17 | lkcl | on the E680 and A780, it runs the touchscreen, keyboard, and some of the GPIO. |
11:36.42 | *** join/#htc-blueangel rob_w (n=bob@p85.212.184.202.tisdip.tiscali.de) |
11:36.43 | *** mode/#htc-blueangel [+o rob_w] by ChanServ |
12:01.06 | *** join/#htc-blueangel jurop88 (n=kvirc@81-208-36-84.ip.fastwebnet.it) |
12:04.50 | *** join/#htc-blueangel jurop88 (n=kvirc@81-208-36-84.ip.fastwebnet.it) |
12:17.43 | *** join/#htc-blueangel iclem (n=clement@iclem.net) |
13:05.40 | *** join/#htc-blueangel Olli_G (n=bbird@port-195-158-178-2.dynamic.qsc.de) |
14:52.08 | lkcl | heya folks |
14:57.28 | iclem | hi lkcl |
15:05.00 | *** join/#htc-blueangel petaflot (n=dave@lsn-boi-catv-c114-p094.vtx.ch) |
15:05.39 | petaflot | hi. happy new year to you all |
15:14.55 | *** join/#htc-blueangel JuNuKN (n=JuNuKN@Mc623.m.pppool.de) |
15:15.38 | JuNuKN | ~botmail read |
15:22.40 | iclem | hi JuNuKN |
15:23.23 | jurop88 | Hi everybody and happy new year to all! |
15:23.50 | JuNuKN | Hi iclem ... Happy new year ... something new ? ... Logs are again incompletet ! |
15:23.56 | iclem | happy new year to you all |
15:24.20 | iclem | nothing new for me, i'm trying to figure out why battery is not charging, and how to get battery status |
15:24.37 | jurop88 | may I put some questions? |
15:24.50 | iclem | of course |
15:24.58 | jurop88 | so... |
15:25.23 | jurop88 | I'm looking to buy a new palm/phone |
15:25.33 | jurop88 | and was looking for a Linux one |
15:26.00 | jurop88 | I followed the whole discussion about the porting... |
15:26.05 | jurop88 | for the XDAII |
15:26.20 | jurop88 | I am not a developer but have some practice with programming |
15:26.39 | jurop88 | and have a nice occasion to have a qtek 9090 for a good price |
15:26.55 | jurop88 | do you think I will be able to use it with LInux? |
15:28.33 | jurop88 | perhaps a more recent jasjar? |
15:28.33 | JuNuKN | 9090 is the qtek 9090 MDA3, which is ported with most of the features |
15:29.13 | JuNuKN | GSM and SMS with it is portet, Sound Keyboard USB and Seriell. |
15:29.40 | JuNuKN | But till now WLAN and BVluetooth is not complete (as I know) |
15:30.49 | jurop88 | ...so it means that I can make/receive phone calls the same way as a normal cell phone? |
15:32.00 | JuNuKN | Yes ... I am also not shure about the current status of standby-handling. Maybe it is still incomplete, but I would expect, that this will be the next, which is ready. |
15:33.14 | jurop88 | another question... |
15:33.39 | JuNuKN | Old way (till for a week or so) was a workarround , which leads to a batery live of standby 18 hours ... # |
15:33.56 | jurop88 | to hack the phone, I will need something particular or the USB connection is sufficient? |
15:35.22 | JuNuKN | You need HaRET and somehow a seriell or WLan-TCP/IP telnet connection to the device ... still running WinCE |
15:35.53 | iclem | yes usb is sufficent |
15:36.08 | JuNuKN | USB will worg for this TCP/IP connection .. |
15:36.51 | jurop88 | sorry if my questions seem stupid, but I will be sure BEFORE spending my money ;-) |
15:37.17 | iclem | have a look at http://www.handhelds.org/moin/moin.cgi/BlueAngel for current status |
15:37.40 | jurop88 | in case of problems, what are the chanches to revert to WinCE original status? |
15:37.55 | jurop88 | @iclem: thanks, I already eat all of it!!! |
15:37.57 | iclem | 100% ;) |
15:39.26 | iclem | linux do not erase WinCE |
15:39.27 | JuNuKN | At the moment the new OS is not flashed into the device it is only stored on a SD card, and without the SD-Card plugged and a hardreset WinCE comes back ;o) |
15:39.34 | iclem | (one day, I hope... ) |
15:39.41 | jurop88 | who is following the project at the moment? I am not so skilled to go on my foot, but I think to be sufficient enough to be of some help in case somebody else is on the same road |
15:40.43 | iclem | AFAIK there are at least three active people on blueangel port |
15:41.37 | jurop88 | @JuNuKN: I'll start to understand the mechanism! So there's a software emulation layer over WinCE? Let's say a form of soft-reset after initialising all services not driven by Linux via WinCE? |
15:42.23 | iclem | if fact it's not emulation, but a wince bootloader which is booting linux kernel (Haret) |
15:43.38 | jurop88 | ah, a dual boot machine! As i read, Haret is an acronym for the reverse-engineering layer, is it? |
15:43.39 | JuNuKN | No, it replaces WinCE ... but only in the RAM-Memory of the device, where the normal location of WinCE is. It will be loaded by a hartreset from falsh into RAM. And Same is now done by inserting the SD-card via a bootloader, running under WinCE |
15:44.09 | JuNuKN | HaRET has also a feature known from loadlin |
15:44.44 | JuNuKN | #And now my knowledge goes to an end ... iclem is the guy form more details |
15:47.03 | JuNuKN | iclem: Any internal news on Himalaya Port : sound chip / gomunicator? |
15:52.25 | jurop88 | another question: is it difficult to build the environment for cross-compiling? I have 2xi386, one (old) with Mandrake and the other with Debian, both updated with a 2.6 kernel but one with a gcc 3 and the other with gcc4 |
15:56.00 | JuNuKN | No it should be easy, because it is available as OpenEmbedded Project, which creates the crosscompilation chain and so on by itself. Don't know it exactly, because I didn't set it up until now .... I am a linux newby ... which is following the discussions since weeks |
15:59.07 | JuNuKN | By the way, this chat is most active at nights ... maybe its better to ask than !? |
15:59.08 | iclem | jurop88: there are precompiled crosstoolchains, easy to use |
16:00.06 | iclem | sufficient if you plan just to do kernel hack |
16:03.40 | iclem | but compilation will be slow on i386 |
16:06.51 | JuNuKN | There we are at the theme .... can you try to give me instructions to do a first compile with my toolchain |
16:06.55 | JuNuKN | ? |
16:07.26 | jurop88 | iclem: the working machine is a Celeron 2.4 Mhz, not properly a 386 ;-) |
16:08.49 | iclem | IIRC you have downloaded and installed hh.org crosstoolchain ? |
16:09.49 | JuNuKN | My current structure to the toolchain is at the moment /home/jn/xanadux/usr/local/arm/3.4.1/bin which I have to add to my Path variable ... Correct? |
16:10.25 | JuNuKN | Yes the handhels toolchain 3.4.1 |
16:10.29 | iclem | no I think it must installed in /usr/local/arm/3.4.1/bin |
16:10.57 | JuNuKN | So I have to move it as root? |
16:11.35 | iclem | don't know of it matters, I've done this way |
16:11.55 | iclem | with you setup try to go in another directory and type "arm-linux-gcc" |
16:13.20 | iclem | any error message ? |
16:13.40 | JuNuKN | OK, I let to old location exist as it was !? Set the path to it and try the command... |
16:13.57 | JuNuKN | The move as user said "no right" |
16:14.17 | iclem | yes try with your current setup before moving |
16:14.58 | JuNuKN | ... Pleas again a dummy Winuser help ... How to setup path ? PATH seems not to be defined? |
16:15.52 | iclem | echo $PATH |
16:16.10 | iclem | export PATH=$PATH: /home/jn/xanadux/usr/local/arm/3.4.1/bin |
16:16.23 | iclem | without space after : |
16:16.50 | JuNuKN | Ahhh thanks echo delivers the path now ... you see I am still not familar with it! |
16:17.40 | iclem | if it works you will have to put it into ~/.bash_profile |
16:18.43 | JuNuKN | returns arm-linux-gcc: no input file :O) great ! |
16:19.00 | iclem | lets go in you kernel tree |
16:19.05 | iclem | make himalaya_defconfig |
16:19.06 | iclem | make |
16:19.23 | JuNuKN | ... one moment ... first at it in my bash ... |
16:20.19 | iclem | INSTALL_MOD_PATH=<diretory you want the modules to be installed in> make modules_install |
16:20.54 | JuNuKN | Linux want not to show me that file .... one moment |
16:22.03 | JuNuKN | I only see a .bashrc and a .bash_history |
16:22.10 | JuNuKN | with mc |
16:22.12 | iclem | create it |
16:22.29 | JuNuKN | sudo mc |
16:22.43 | iclem | create it as user |
16:22.45 | JuNuKN | sorry wrong keyboard ;o) |
16:22.50 | JuNuKN | sudo mc |
16:23.02 | iclem | echo "export PATH=$PATH: /home/jn/xanadux/usr/local/arm/3.4.1/bin" > ~/.bash_profile |
16:23.30 | iclem | again without space ;) dumb cut&paste |
16:26.18 | JuNuKN | ok make is running .... |
16:27.05 | JuNuKN | But was iut handled correct ? it calls cc and maybe not the CrossChain compiler !? |
16:28.27 | cr2 | hi guys |
16:28.43 | cr2 | JuNuKN: put the cross-path before $PATH |
16:28.47 | JuNuKN | Hello cr2 and a happy new year |
16:28.53 | iclem | iirc kernel is choosing the gcc that correspond to the arch |
16:29.37 | JuNuKN | ? So it is still correct only to call the himalaya_defconfig and than the make |
16:29.55 | iclem | that what i do and it works |
16:30.09 | iclem | try ;) |
16:30.45 | JuNuKN | OK ... I never thought, that it is that easy .... It is still running ... no harddisk problem till now ;o) |
16:30.55 | cr2 | i use: PATH=/usr/local/arm/3.4.1/bin:$PATH make |
16:31.39 | cr2 | iclem: have you tried gomunicator ? |
16:31.45 | JuNuKN | It proceed without errors and now ? |
16:32.19 | iclem | yes, with a symlink /dev/gsm0 > /dev/ttyS0 |
16:32.32 | JuNuKN | I will write the stuff into wiki afterwards to make it reproducable for other beginners ... |
16:33.23 | JuNuKN | In the root there is now a vmlinux |
16:33.23 | iclem | error message i do not remind about message channel cannot be open and /dev/gsm0 not a directory |
16:33.29 | cr2 | iclem: /dev/ttyS2 http://wiki.xda-developers.com/index.php?pagename=HimalayaSerial |
16:33.58 | iclem | ok will give it a new try |
16:34.24 | iclem | JuNuKN: use arche/arm/boot/zImage |
16:34.38 | iclem | arch/ |
16:34.47 | iclem | that's the kernel |
16:35.13 | iclem | rename it or modify haret's script to reflect the name |
16:36.27 | JuNuKN | OK, I see ... that was easy :o)) .... So this kernel may start and may give some serial output ... but where is the initrd and how can I add stuff there ? |
16:37.59 | iclem | use another initRD, it works |
16:38.07 | iclem | this kernel is mostly builtin |
16:38.09 | JuNuKN | Espoecially the mmc-driver |
16:38.16 | iclem | ha |
16:38.33 | iclem | copy the modules in your initRD |
16:38.58 | JuNuKN | So there is no other kewrnel in the initRD ? |
16:39.30 | JuNuKN | I thought, that the first kernel only loads the ramdisk and than sitches the root and the kernel!? |
16:39.31 | iclem | no initRD contains a root file system |
16:39.55 | iclem | the first kernel is the only one |
16:40.08 | JuNuKN | I understand,- and have to read this night!! |
16:40.27 | iclem | it loads the initRD in ram and execute linuxrc |
16:41.17 | JuNuKN | So did I also compile the two mmc-modules with this command? And how to insert them into the initRD maybe permananent. Maybe modify a sh script? |
16:41.49 | iclem | have you make modules_install |
16:42.17 | JuNuKN | No not till now? Where must I do it under the device itself? |
16:42.27 | iclem | INSTALL_MOD_PATH=<diretory you want the modules to be installed in> make modules_install |
16:42.57 | JuNuKN | echo INSTALL_MOD_PATH |
16:42.58 | iclem | not this extracts the modules from the kernel tree and copy it in the /lib/modules dir |
16:43.06 | JuNuKN | uups |
16:43.37 | iclem | but as user you should install then elsewhere |
16:44.26 | iclem | to mount the initrd follow my recent "GPE without SD" wiki |
16:45.04 | JuNuKN | No I am completely disturbed |
16:45.37 | iclem | why ? |
16:45.55 | JuNuKN | I thought the Kernel loads the initRD ... initialized by harets command-line arguments and envirenment |
16:46.50 | iclem | it's right |
16:47.19 | iclem | the modules files are in initRD |
16:47.21 | JuNuKN | OK, than again the reason for the INSTALL_MOD_PATH ... module_install |
16:47.34 | JuNuKN | ... I am to fast |
16:47.47 | iclem | to get the modules files to be copied in you initRD ;) |
16:48.40 | JuNuKN | Where can I find the current, preconfigured initRD file, which I just compiled ? |
16:49.40 | JuNuKN | ... or better,- hopefully created with my make command |
16:50.47 | iclem | cr2: gomunicator says "could not get security status, Is SM card missing" |
16:51.06 | rob_w | iclem, try again |
16:51.11 | iclem | that's pretty faire, as there's not SIM card in the himalaya ;) |
16:51.25 | rob_w | duh |
16:51.34 | cr2 | great. is there any log feature in gomunicator ? |
16:52.50 | cr2 | rob_w: is there any userspace tool for the gsm data channel ? |
16:53.00 | iclem | i reboot with sim inserted, i really like to do the wame things twice |
16:53.26 | rob_w | cr2, no |
16:53.45 | rob_w | i was thinking of a AT input box inside gomunicator |
16:54.33 | rob_w | cr2, would u like to see something ? |
16:54.34 | cr2 | cp15 said something about modifying pppd. we have a common access model for all phones: two serial ports, so it will work everywhere. |
16:55.04 | JuNuKN | OK I found one ... under .../boot/bootp initRD.s but only with 110 byte ... |
16:55.19 | cr2 | rob_w: what about creg=2 ? |
16:55.38 | rob_w | cr2, hehe thats where i wanted to try that .. |
16:55.44 | iclem | JuNuKN: initRD is NOT compiled with kernel |
16:56.05 | iclem | you have to created one or modify existent |
16:56.07 | rob_w | sorry i amy so frustrated as my ba does not resume anymore .. and i ve not idea why |
16:56.29 | iclem | just add you modules in initRD's /lib/modules |
16:57.14 | JuNuKN | OK, please give me detailed informations .... for a windows user ;o) |
16:57.40 | iclem | ok, take the initRD from llange website |
16:57.58 | JuNuKN | ok, I try one moment .... |
16:58.41 | rob_w | ah it seems i did not use logread correctly ,, thats why i saw now messages |
17:01.19 | JuNuKN | OK I have doiwnloaded it into my /home/jn/xanadux directory name is initrd-2.6.12.gz |
17:01.51 | JuNuKN | size 5.4 MByte |
17:02.08 | iclem | gunzip |
17:02.24 | JuNuKN | cd .. |
17:02.30 | iclem | mkdir tmpinitrd |
17:02.34 | JuNuKN | cd ..# |
17:02.45 | iclem | mount -o loop initrd-2.6.12 tmpinitrd |
17:03.06 | iclem | the go to kernel tree |
17:03.34 | iclem | then : INSTALL_MOD_PATH=<path/to/tmpinitrd> make modules_install |
17:03.51 | iclem | ha may need to be root to mount |
17:04.23 | iclem | rob_w: gomunicator do not want my pin |
17:04.53 | iclem | no reaction, no error |
17:05.03 | iclem | just "enter pin:" |
17:06.11 | JuNuKN | I just unpacked the file ... I am not that fast ;o) |
17:06.17 | iclem | db_call_update_duration(-1) Segmentation fault |
17:06.48 | iclem | hopefully you've got the log ;) |
17:07.16 | JuNuKN | sudo mount ... doesn't work sais no sudoers file |
17:07.31 | iclem | su |
17:09.03 | JuNuKN | OK, now the mount worked |
17:10.35 | iclem | do the modules_install as root too |
17:11.12 | JuNuKN | Thanks for the tip ... ;o) I now was abvle to switch back .... with su jn |
17:11.18 | JuNuKN | su |
17:11.53 | iclem | type "exit " to return as user |
17:11.59 | JuNuKN | A forecast of mine ... Path is not set here !? |
17:13.03 | iclem | you must use "su - <username>" to load all environment , including $PATH |
17:13.34 | iclem | gomunicator do not dial |
17:14.08 | iclem | I'm using v0.1.3, is there a most recent compiled version i can try ? |
17:15.37 | iclem | got to go, bbl |
17:17.38 | JuNuKN | OK, it seems to work ... No error message ....nasic3_mmc.ko and others are processed ... |
17:18.12 | JuNuKN | Now umount <something>? |
17:24.36 | JuNuKN | iclem ? |
17:26.33 | iclem | yes |
17:26.47 | iclem | umount tmpinitrd |
17:26.56 | JuNuKN | OK, I have unmounted it like a normal mount ... |
17:27.08 | iclem | gzip initrd-2.6.12 |
17:28.00 | JuNuKN | Now the files are in the intrd !? Thanks !!! I will write a short doku tonight for the wiki! How to add other things like gomunicator? |
17:28.07 | JuNuKN | ls# |
17:29.26 | iclem | gomunicator needs GPE which is not included in this initRD |
17:30.30 | JuNuKN | Ahh ok .... |
17:31.32 | JuNuKN | And all neccessary kernel things are automatically copyed by the modules_install command? So I simply have to take any gpe image than!? |
17:31.37 | iclem | next step for you is to try your just compiled kernel to check if everything worked |
17:32.09 | iclem | sort of |
17:32.19 | JuNuKN | OK ... I will try later ... backup takes a lot of time under WinCE therefore I first have to update the kernel tree from cvs ... |
17:33.01 | JuNuKN | and do it again and write the doku,- finally also into wiki .... |
17:33.18 | JuNuKN | How to update the cvs tree from the commandline ? |
17:33.33 | JuNuKN | Do I need a clean before the make ? |
17:33.37 | iclem | tonight if you want, i can explain you how to use NFS |
17:34.46 | iclem | i do "make distclean" before cvs update |
17:34.50 | JuNuKN | Which are the advantages for it? if it is that simple to make changes on the initrd!? |
17:35.51 | JuNuKN | OK I did the make distclean |
17:35.55 | iclem | you can modify modules without rebooting ( 15 minutes economy) |
17:37.39 | JuNuKN | OK ... good Idea .... but maybe to much for my first start .... lets use the current learned . Thanks for the moment! What is the cvs update for my tree ? |
17:38.16 | JuNuKN | I have to go to eat in a few minutes and bring the kids to bed ... so I will met again in a few hours |
17:40.15 | iclem | be back later too |
17:41.11 | jurop88 | btw, thanks for all the info and see you later |
17:51.36 | *** join/#htc-blueangel eme (n=bosty@BSN-165-121-219.dial-up.dsl.siol.net) |
17:51.51 | eme | happy new year girls and boys! |
18:46.40 | cr2 | hi eme |
18:49.05 | cr2 | i have compiled 'gnuharet' with gcc, but it doesn't start. must be my fault. vlc compiled with the same tools works. |
20:11.49 | JuNuKN | Hey iclem ... Thanks again .... http://wiki.xda-developers.com/index.php?pagename=HowToCompile4Himalaya MY compile notices are online now ... be so kind and crosscheck it and correct it if needed. Thanks |
20:23.13 | JuNuKN | I see, noone is online ,,,, My PC linux machine has restarted itself without user action ! Don't know where to look for problems in the logs ... Maybe that was the initial problem of my FS charsh |
20:24.11 | JuNuKN | Maybe I have to switch of (somehow) acpi handling !? |
20:27.05 | lkcl | hello i'm not here, JuNuKN |
20:27.31 | JuNuKN | Sad, so no help for now ;o) |
20:31.17 | JuNuKN | .... or maybe you are really online ? |
20:31.51 | cr2 | wince programming is evil. i don't reach WinMain for some reason. hello world works. |
20:36.51 | lkcl | what is this "online" of which you speak? |
20:37.02 | JuNuKN | not away .... |
20:37.29 | JuNuKN | May you give me a tip how to update my cvs tree with commandline cvs ? |
20:37.41 | lkcl | cvs update -d -P |
20:37.54 | lkcl | in the linux-2.6-xda directory |
20:38.06 | lkcl | fairly easy, really. |
20:38.17 | lkcl | i'm trying to build a working initrd for the himalaya. |
20:39.45 | JuNuKN | doesn't work for me reports ... failed to open /home/jn/.cvspass I checked out as anonymous |
20:41.50 | JuNuKN | No ideas ? |
20:44.57 | JuNuKN | Now,- with the second try it seems to do something ... with the fail message ... !? We will see |
20:45.07 | lkcl | do touch ~/.cvspass |
20:45.10 | lkcl | then do it again. |
20:46.28 | JuNuKN | lkcl: OK I will do afterwards ... thanks. Do you create a completely new initrd? Or a image for the SD card as real filesystem? |
20:46.37 | JuNuKN | What is the difference? |
20:47.37 | lkcl | just adding stuff to the existing one |
20:47.49 | lkcl | hurrah - GPE 2.7 booting on himalaya, hurrah. |
20:47.54 | lkcl | ooo, entering runlevel 5... |
20:48.06 | lkcl | HEY COOL! GPE! |
20:48.09 | lkcl | with a touchscreen! |
20:48.25 | lkcl | dang. |
20:48.26 | JuNuKN | sounds great!!! And what else? SD-Card support ? |
20:48.29 | lkcl | wow this is wild. |
20:48.32 | lkcl | yep, sd-card. |
20:48.47 | lkcl | good god. sound module loaded. |
20:49.09 | JuNuKN | And speaker working ??? |
20:49.26 | lkcl | dunno yet... |
20:51.52 | lkcl | hmm... /dev/ttyS0 isn't recognised. |
20:52.03 | lkcl | the initialisation sequence for the himalaya gsm is VERY specific. |
20:52.08 | lkcl | and a complete. |
20:52.09 | lkcl | utter. |
20:52.11 | lkcl | friggin. |
20:52.14 | lkcl | pain. |
20:52.42 | cr2 | lkcl: what is ttyS0 ? |
20:52.44 | lkcl | it took me something like ... FOUR weeks just to get the right sequence to start the GSM phone up. |
20:52.52 | lkcl | uhhh.. dunno, cr2. |
20:53.01 | cr2 | gsm is /dev/ttyS2! |
20:53.03 | lkcl | how am i gonna test sound... *thinks*... |
20:53.21 | lkcl | well... it ain't there. |
20:53.36 | lkcl | and it _won't_ be there if someone's removed the initialisation code that powers up the GSM radio rom. |
20:54.10 | lkcl | it really _is_ a complete bastard. you have to power it up for a few ms, then power it down. flip some other GPIO then power back on. |
20:54.11 | JuNuKN | shit the ttys1, 2 and 0 ... |
20:54.22 | lkcl | i think you even have to send some AT commands in between. |
20:54.26 | lkcl | it was awful :) |
20:54.43 | cr2 | for the sound ? |
20:55.51 | lkcl | no, to power up the GSM radio module. |
20:56.03 | JuNuKN | Test the sound with some beeps or a mp3 or something else ? |
20:56.05 | lkcl | the sound module is loaded, recognised. |
20:56.18 | lkcl | yeh there's a test as part of the sound settings - no volume. oh well. |
20:56.25 | lkcl | can't have everything, i suppose. |
20:56.33 | cr2 | aah. do you have gomunicator inside gpe ? |
20:56.36 | lkcl | okay. i'm gonna hunt for the code. |
20:56.42 | lkcl | yep, cr2. |
20:57.13 | cr2 | link /dev/ttyS2 to /dev/gsm0 and run it. look in the /tmp/gomunicator.log |
20:57.28 | lkcl | the initialisation sequence is guaranteed to fail if the GSM radio module hasn't been properly initialised. |
20:57.37 | lkcl | it's empty. |
20:57.44 | cr2 | ok. |
20:58.22 | lkcl | why is himalaya_rs232_irq commented out? |
20:59.29 | lkcl | okay - it's this code, ser_ffuart_gpio_config |
21:00.22 | lkcl | ah, that's what those silly gsm reset gpio on and off messages are all about... hmmm. |
21:00.26 | cr2 | split hima with the BA code. you have different serial setup. |
21:02.00 | cr2 | http://wiki.xda-developers.com/index.php?pagename=HTC_Serial |
21:02.04 | lkcl | ta |
21:03.17 | lkcl | hmmm... |
21:03.45 | lkcl | why doe ser_ffuart_gpio_config keep getting called every couple of mins? |
21:04.24 | cr2 | no idea. where ? |
21:05.16 | lkcl | ah, it's cos /etc/inittab has an entry for tts/2 to start up a getty! |
21:05.28 | cr2 | fucking wince. you can't even fprintf with creating a window and a message queue... |
21:05.31 | cr2 | lol. |
21:06.09 | lkcl | damn. it's gone suspend bye-byes. and i was getting on _so_ well :) |
21:06.44 | cr2 | s/with/without/ |
21:06.56 | cr2 | suspend. so fast ? |
21:07.20 | lkcl | about 10 mins. of course i've not tried blueangel_suspend yet so it wouldn't wake up. |
21:14.05 | JuNuKN | Maybe ... if you make a GSM-call to the device |
21:14.57 | lkcl | oh. |
21:14.59 | lkcl | :) |
21:15.07 | lkcl | it helps to have a SIM card in it... |
21:15.11 | JuNuKN | ;o) |
21:15.53 | JuNuKN | yeahr I read it also ..... thet really also a gsm sim card is needed ... also for linux ;o) |
21:17.30 | cr2 | wow. got gnuharet gui running :) the bug must be somewhere deep in the code. |
21:17.55 | lkcl | excellent, cr2. |
21:19.16 | JuNuKN | .... whats the differnce ? |
21:22.03 | cr2 | JuNuKN: you are not compiling free software with some shitty microsoft compiler ;) it's gnu/haret :lol: |
21:23.40 | JuNuKN | OKOK ... but it is only the way ... not the end goal :o))) |
21:23.54 | JuNuKN | But this are principles ... |
21:24.39 | JuNuKN | I have to go soon, hopefully the logs are tomorrow completer as the last days !!! |
21:31.52 | lkcl | _right_. what am i buggering-up today? |
21:35.23 | JuNuKN | Overwrites from BA ...? |
21:40.30 | lkcl | come oooonnn |
21:40.37 | lkcl | boot DAMN |
21:50.08 | lkcl | oops i used cramfs instead of ext2 to make the initrd |
21:50.53 | lkcl | boot yer buffer YAAY |
21:51.04 | lkcl | wheeeee |
21:53.45 | cr2 | Detected CPU family: Unknown |
21:53.51 | cr2 | that pisses me off. |
21:54.41 | lkcl | huh? |
21:54.58 | lkcl | you finally got the irda going on the thingy - universal - then? |
21:58.42 | cr2 | not yet. but found a gpio. |
21:59.19 | cr2 | YES. it works now. HaRET Version 0.3.6gnu |
22:00.03 | cr2 | 100% microsoft-free solution. |
22:00.15 | cr2 | now i need to fix the known haret bugs. |
22:12.58 | *** part/#htc-blueangel jurop88 (n=kvirc@81-208-36-84.ip.fastwebnet.it) |
22:13.37 | *** join/#htc-blueangel jurop88 (n=kvirc@81-208-36-84.ip.fastwebnet.it) |
22:16.49 | lkcl | ah. |
22:16.51 | lkcl | got it. |
22:17.11 | lkcl | some idiot modified the serial initialisation sequence without understanding what's going on. |
22:17.28 | lkcl | because there was a complex function - pxa_gpio_op() - that doesn't now exist |
22:17.45 | lkcl | they replaced the call with two oversimplified (and incorrect) pxa_gpio_mode() calls. |
22:17.50 | lkcl | which is totally wrong. |
22:21.26 | cr2 | well. we probably need a hi0 branch. |
22:21.34 | lkcl | naah. |
22:21.42 | lkcl | the ba0 one is good enough. |
22:21.56 | lkcl | i found the missing function that whoever-it-was couldn't be bothered to cut across. |
22:22.28 | lkcl | they _sort_-of made part of it... yeh... there is much similarity. |
22:22.44 | cr2 | i'm tired but happy. |
22:22.48 | lkcl | but it's no good: pxa_gpio_mode only has two parameters: i need 3. |
22:22.50 | lkcl | ah, good :) |
22:29.07 | cr2 | haret now takes 3 seconds to compile here :) |
22:29.11 | lkcl | _excellent_ |
22:29.16 | lkcl | that's more friggin like it. |
22:35.10 | cr2 | i have WADDR, WADDR16, KMALLOC and VMALLOC commands now. and completely messed up pxa27x code... |
22:39.22 | lkcl | _excellent_ |
22:47.22 | cr2 | before testing IR i must fix the pxa27x gpio support. |
23:07.08 | cr2 | i get interesting results from the framebuffer code. without this silly gapi stuff. |
23:12.03 | lkcl | on the universal? |
23:13.49 | cr2 | yes. look. |
23:14.20 | lkcl | where,where! |
23:14.57 | cr2 | B1: form=1 bpp=16 fbaddr=A87AA000/A3FAA000 x=480/2 y=640/960 |
23:15.08 | cr2 | FB0: form=160 bpp=16 fbaddr=B0094020/A0094020 x=240/480 y=320/480 |
23:15.41 | cr2 | two different methods of calling the video driver give two different results. |
23:19.39 | *** join/#htc-blueangel apt (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
23:19.39 | *** topic/#htc-blueangel is HTC BlueAngel is still under development ! Feel free to ask and check also -> http://www.handhelds.org/moin/moin.cgi/BlueAngel | http://wiki.xda-developers.com/index.php?pagename=BlueangelResearch | http://gnulinux.biz/files/blueangel/sd/linux/ channel chat logs at http://ibot.rikers.org/%23htc-blueangel/ **upcoming gpe release is in org.openembedded.oz354fam083 dev tree !!!!*** HAPPY NEW YEAR 2006 *** |
23:23.51 | cr2 | it looks like two Visuals (form=1,160) simultaneously mapped into main memory. double buffering for SRAM ? hacky setup. |
23:26.39 | cr2 | #define FORMAT_565 1 |
23:28.25 | iclem | I'm unable to make battery charge |
23:29.21 | iclem | can't get what is wrong or lacking in the code |
23:30.03 | lkcl | 0000010 002 002 002 022 G S M V e r s i o n |
23:30.03 | lkcl | 0000020 : 1 . 0 6 . 0 1 G 2 3 V e |
23:30.03 | lkcl | 0000030 r s i o n : 1 3 3 7 . 1 6 002 |
23:30.03 | lkcl | 0000040 002 002 026 A T - C o m m a n d I n |
23:30.03 | lkcl | 0000050 t e r p r e t e r r e a d y \r |
23:30.22 | lkcl | oops :) |
23:30.45 | iclem | I've put some gpio/asic dump at http://iclem.net/himalaya if it can help |
23:31.40 | lkcl | oh - you wanna play with a new kernel and zImage, iclem? |
23:32.33 | iclem | oh yes |
23:33.00 | cr2 | lkcl: can you send at+creg=2 to the gsm ? is it supported on hima ? |
23:33.09 | lkcl | http://hands/com/~lkcl/himalaya |
23:33.24 | lkcl | will try... i'm doing a hexdump -c /dev/ttyS2 |
23:33.35 | lkcl | i'm not sure if i echo something to it if it will try to reinitialise it - again... |
23:34.11 | lkcl | nothing... no rection... |
23:34.13 | cr2 | just the echo is not enough. |
23:34.47 | cr2 | you must wrap it in 0x16 0x2 at+... |
23:34.58 | lkcl | hmm, goes gpe have a "terminal... oh bollocks." |
23:35.19 | cr2 | install uucp :) |
23:35.26 | lkcl | ah joy. |
23:35.58 | lkcl | no, i think i'll create a file and then copy that over... |
23:36.14 | iclem | must go, maybe see you later |
23:36.21 | lkcl | k iclem. |
23:36.33 | cr2 | echo should be capable to send oct and hex. |
23:36.59 | lkcl | what's 0x16 in octal? |
23:37.07 | lkcl | <PROTECTED> |
23:37.19 | lkcl | 26 |
23:37.21 | lkcl | \026 |
23:38.30 | cr2 | hmm. how will you get the response ? echo ...; cat ? |
23:38.31 | lkcl | what's the "ending" character?. ... oh right i get it. |
23:38.40 | cr2 | \r |
23:39.40 | lkcl | hmmm.. |
23:40.22 | cr2 | i used this with gps, but it's ugly. |
23:40.29 | lkcl | i think it's getting switched off again. |
23:40.37 | lkcl | must check the sequence properly... *sigh*... |
23:41.38 | cr2 | look into the gomunicator source. |
23:41.50 | lkcl | wow! the lcd brightness works on the himalaya! |
23:41.52 | lkcl | cool! |
23:42.18 | cr2 | const char prefix[2] = { 0x02, 0x16 }; |
23:42.28 | cr2 | sendCommand('%s') |
23:42.37 | lkcl | ack. |
23:42.37 | cr2 | const char postfix[2] = { 0x16, 0x02 }; |
23:42.49 | lkcl | okay. |
23:44.13 | lkcl | it's not the init stuff i'm concerned about, it's the GPIO initialisation stuff. |
23:44.17 | lkcl | in himalaya.c |
23:44.51 | lkcl | that's gotta be absolutely perfect. |
23:45.15 | cr2 | the command ends with CRLF |
23:45.42 | lkcl | that's \r\n isn't it. |
23:45.48 | lkcl | echo \002\026at+cpan=2\r\n\026\002 > /dev/ttyS2 |
23:45.49 | cr2 | yes. |
23:45.52 | lkcl | is what i'm doing. |
23:46.06 | cr2 | ; cat /dev/ttyS2 ? |
23:46.52 | cr2 | echo -e ? |
23:46.56 | lkcl | well if you do "cat /dev/ttyS2" then it tries to reinitialise the GSM modem... |
23:47.03 | lkcl | DUH. |
23:47.37 | lkcl | nope.... |
23:47.37 | cr2 | stty -crtscts -F /dev/ttyS2 |
23:47.54 | cr2 | or something like that. |
23:48.12 | lkcl | nope... |
23:48.41 | lkcl | * [-]crtscts enable RTS/CTS handshaking |
23:49.19 | cr2 | i think universal uses DTR for reset... |
23:49.50 | lkcl | oodear... |