| 00:00.22 | psokolovsky | Kevin2, Ah, if go that far old, no, unfortunately ;-( wm2003 is minimum |
| 00:00.25 | cr2 | Kevin2: i have a working code, even wgpio (!). and 'dump gpio'. the only part which is broken is AllocPhysMem ;) |
| 00:00.34 | psokolovsky | I porbbaly hsould look for ppc2002 image for h3900 |
| 00:01.55 | cr2 | Kevin2: but some useful functions are missing on 2.11. SleepTillTick() for example. |
| 00:02.06 | *** join/#htc-linux JTRipper (n=jtripper@p54B1CA53.dip.t-dialin.net) |
| 00:02.32 | Kevin2 | cr2: I think SleepTillTick can be replaced with Sleep(1) or Sleep(0). |
| 00:02.35 | WizMaui | cr2: can you tell something from my dump? |
| 00:02.37 | JTRipper | !@=#+... my X just simply restarted... |
| 00:02.51 | cr2 | JTRipper: samsung seems to be more close to pxa than omap850. |
| 00:02.55 | Kevin2 | psokolovsky: Is there really a need to support old versions of ce? |
| 00:03.01 | cr2 | WizMaui: not really. |
| 00:03.17 | JTRipper | yeah, I know... omap850 is crap... |
| 00:03.28 | psokolovsky | Kevin2, we told about that already, I think yes ;-) |
| 00:03.32 | JTRipper | but it was the only thing i could afford |
| 00:03.54 | psokolovsky | Kevin2, but I can bother with that, only thing I ask is that you didn't remove old code. |
| 00:03.57 | cr2 | Kevin2: maybe. i have not disassembled the code, so don't know how it is implemented. but it lowers the CPU load significantly |
| 00:04.35 | psokolovsky | Kevin2, if you don't feel like working on it, pls just separate to some file, and I'll pick it up from there |
| 00:04.38 | cr2 | Kevin2: isn't wince tick=HZ=10ms ? |
| 00:05.10 | Kevin2 | cr2: I don't know, but the timer interrupt seems to fire 1000 times a second. |
| 00:05.37 | cr2 | yes, gettickcount() |
| 00:06.50 | cr2 | Kevin2: the code is here, if you are interested http://jornada820.cvs.sourceforge.net/jornada820/bootimage/haret-j820gnu |
| 00:07.09 | Kevin2 | psokolovsky: The main issue is that I can't test any changes I make. I was going to add in some code to do late binding of dll's. But again, if I can't test it, I don't know what use it will be. |
| 00:07.44 | cr2 | i must spend some time and commit the jornada820 code to hh.org CVS ;) |
| 00:07.48 | psokolovsky | Kevin2, if you use LoadLibrary(), it will work. Or do you do something more funky? |
| 00:08.22 | Kevin2 | psokolovsky: Right. I'm pretty sure it will work, but as a matter of course, I like to test everything before committing.. |
| 00:08.43 | cr2 | psokolovsky: i wanted to do RIL things in this way, but got lost in evil c++ casts ;-) |
| 00:08.47 | psokolovsky | Kevin2, all in all, I don't want to dictate rules and make your work harder. just if you kill code, fishing for it in old revs is harder then see at least something in the current rev. |
| 00:08.53 | Kevin2 | cr2: Is there a way I can get a tar of the code? (CVS checkout from jornada820... doesn't seem to work for me). |
| 00:09.13 | cr2 | Kevin2: i'll try. |
| 00:09.41 | psokolovsky | Kevin2, so, after all, feel free to decide what is threshold of old versions support. |
| 00:10.02 | Kevin2 | Haret wiki says -- cvs -d :pserver:anoncvs@jornada820.cvs.sourceforge.net:/cvs login -- but the password fails for me. |
| 00:12.19 | cr2 | Kevin2: i've checked out everything with -- cvs -z3 -d:pserver:anonymous@jornada820.cvs.sourceforge.net:/cvsroot/jornada820 co -P bootimage |
| 00:13.00 | cr2 | only haret-j820gnu is relevant |
| 00:13.20 | Kevin2 | cr2: What login password did you use? |
| 00:13.48 | Kevin2 | Hrmm. I guess I don't need to login. |
| 00:14.20 | psokolovsky | Kevin2, yep, try empty at first ;-) |
| 00:15.58 | cr2 | hmm. small patch is needed. |
| 00:16.02 | cr2 | --- src/script.cpp 28 Apr 2006 12:55:43 -0000 1.1 |
| 00:16.06 | cr2 | +++ src/script.cpp 16 Nov 2006 00:15:21 -0000 |
| 00:16.10 | cr2 | @@ -22,7 +22,7 @@ |
| 00:16.14 | cr2 | <PROTECTED> |
| 00:16.18 | cr2 | <PROTECTED> |
| 00:16.22 | cr2 | <PROTECTED> |
| 00:16.26 | cr2 | -#include "bench.h" |
| 00:16.30 | cr2 | +//#include "bench.h" |
| 00:16.34 | cr2 | <PROTECTED> |
| 00:16.42 | cr2 | --- src/wince/haret-res.rc 28 Apr 2006 12:56:38 -0000 1.1 |
| 00:16.46 | cr2 | +++ src/wince/haret-res.rc 16 Nov 2006 00:15:21 -0000 |
| 00:16.50 | cr2 | @@ -2,10 +2,10 @@ |
| 00:16.54 | cr2 | -IDB_LOGO BITMAP DISCARDABLE "img/penguin.bmp" |
| 00:16.58 | cr2 | -IDB_THERMORED BITMAP DISCARDABLE "img/thermored.bmp" |
| 00:17.02 | cr2 | -IDB_THERMOBLUE BITMAP DISCARDABLE "img/thermoblue.bmp" |
| 00:17.06 | cr2 | -IDB_EYES BITMAP DISCARDABLE "img/eyes.bmp" |
| 00:17.10 | cr2 | +//IDB_LOGO BITMAP DISCARDABLE "img/penguin.bmp" |
| 00:17.10 | cr2 | +//IDB_THERMORED BITMAP DISCARDABLE "img/thermored.bmp" |
| 00:17.11 | cr2 | +//IDB_THERMOBLUE BITMAP DISCARDABLE "img/thermoblue.bmp" |
| 00:17.11 | cr2 | +//IDB_EYES BITMAP DISCARDABLE "img/eyes.bmp" |
| 00:17.11 | cr2 | then it compiles. |
| 00:19.07 | cr2 | and runs. |
| 00:21.35 | *** join/#htc-linux scatterp- (n=deminish@212.32.67.196) |
| 00:23.52 | WizMaui | gn8 everyone |
| 00:24.35 | Kevin2 | cr2: I've looked through the diffs, and nothing pops out as being hard to support. Was there something in particular that made the sa1100 hard to do? |
| 00:28.15 | cr2 | Kevin2: no. not really. |
| 00:28.57 | Kevin2 | cr2: Okay cool. |
| 00:29.38 | cr2 | hmm. WS2.dll vs WINSOCK.dll comes to mind. and 0xc0000000 vs. 0xa000000, but it should not be a problem for you. |
| 00:29.49 | Kevin2 | What I've done with mainline haret, is added a couple new features - the first is a mechanism to add a machine class. This should allow sa1100 to be just another class that gets detected at startup. |
| 00:30.10 | cr2 | yes, that was really good. |
| 00:30.37 | cr2 | sa1100 and sa1110 actually. but they are the same cpu. |
| 00:30.38 | Kevin2 | The second, is I've added a way to add new commands and variables without having to modify script.cpp. This should mean we can add all the cool features of gnu-haret without making script.cpp too big. |
| 00:31.00 | cr2 | ok. |
| 00:31.22 | Kevin2 | The next thing I was going to add, was the ability to do late binding of functions. (So we don't need to require a dll be present.) |
| 00:31.29 | cr2 | then you'll be able to support samsung and omap too. |
| 00:31.37 | cr2 | ok. |
| 00:31.59 | Kevin2 | REG_CMD(0, "BOOT|LINUX", cmd_bootlinux, |
| 00:31.59 | Kevin2 | <PROTECTED> |
| 00:31.59 | Kevin2 | <PROTECTED> |
| 00:32.05 | Kevin2 | The above is how a new command is added. |
| 00:32.31 | cr2 | ok, so the help text is included automagically :D |
| 00:33.00 | Kevin2 | Yeah, it is all a compile time magic - similar to how the kernel can declare module parameters. |
| 00:33.11 | cr2 | ok. |
| 00:33.44 | cr2 | hmm. but the asm code may be different for different cpus ? |
| 00:34.11 | Kevin2 | I've removed most of the assembler from haret. |
| 00:34.27 | Kevin2 | The remaining stuff should be portable. (eg, enable/disable interrupts.) |
| 00:34.33 | cr2 | ok, but you can't get rid of inline. |
| 00:34.51 | cr2 | machine-dependent too ? |
| 00:34.54 | Kevin2 | Yeah, that stuff is still there, but that should be portable. |
| 00:35.45 | cr2 | i mean, these functions are just machine-dependent pointers ? |
| 00:36.01 | Kevin2 | Which functions? |
| 00:36.12 | cr2 | enable_interrupts |
| 00:36.23 | Kevin2 | enable interrupts is an arm standard. |
| 00:36.37 | cr2 | hmm. icmr2 on pxa27x ? |
| 00:37.21 | Kevin2 | One enables/disables interrupts via the cpsr register. |
| 00:38.00 | Kevin2 | Some code will be platform specific. My plan is to just have those commands disabled on non-conforming platforms. |
| 00:38.10 | cr2 | ok, we may change it later. |
| 00:38.12 | cr2 | ok. |
| 00:38.29 | Kevin2 | For example, "wi" is definitely pxa specific right now. So, I was just not going to list it is as available on non-pxa machines. |
| 00:38.44 | Kevin2 | REG_CMD(testPXA, "WI|RQ", cmd_wirq, |
| 00:38.44 | Kevin2 | <PROTECTED> |
| 00:38.44 | Kevin2 | <PROTECTED> |
| 00:38.56 | Kevin2 | That is what the "testPXA" is supposed to do -- it is just stubbed out right now. |
| 00:39.29 | cr2 | ok. |
| 00:41.06 | cr2 | falling asleep ;) good night. |
| 00:41.12 | Kevin2 | good night. |
| 01:05.53 | Kevin2 | Ughh. LoadLibrary and GetProcAddress are both CE3.0 functions! This kinda rules out ce2.1 users. |
| 01:06.49 | Kevin2 | Oops - they are in ce1.0.. Nevermind. |
| 01:26.22 | Kevin2 | Okay. I successfully added a late library loading capability. Syntax is: |
| 01:26.33 | Kevin2 | LATE_LOAD(AllocPhysMem, "coredll") |
| 01:27.14 | Kevin2 | The above will create a function pointer "ptr_AllocPhysMem" that can be used in place of the real function call. It will be NULL if the library doesn't successfully load. |
| 01:57.05 | lkcl | damnit. i need to read 1 byte from the serial port. i'm tempted to do it as a direct address read... YEP! |
| 02:54.25 | Kevin2 | cr2 / psokolovsky: Can you grab the latest haret from mainline cvs and see if it works for you? (Both new and old platforms?) |
| 02:55.32 | psokolovsky | Kevin2, will try to do soon, mingw32ce gcc seem to have been built for me |
| 02:55.36 | psokolovsky | thanks! |
| 02:55.59 | Kevin2 | You can also get it from: svn://www.shadowmite.com/home/svn/ppc6700/haret as haret-0.4.3.exe |
| 02:56.31 | psokolovsky | ok |
| 02:56.49 | Kevin2 | This version has the "wi" command for pxa systems. On old system (non pxa or pre-ce-4.0) the command wont exist and haret should still load. On new systems, however, all the functionality should be present. |
| 02:57.28 | psokolovsky | nice! |
| 02:57.40 | psokolovsky | pxa you mean 25x too? |
| 02:58.08 | Kevin2 | pxa 25x should still work. (We've had people with blueangel use wi correctly.) |
| 02:58.11 | *** join/#htc-linux dullard (n=jim@adsl-static-1-30.uklinux.net) |
| 02:59.23 | psokolovsky | ok |
| 03:16.23 | Kevin2 | psokolovsky: I've reenabled your gx.dll stuff now. It is using the LATE_LOAD implementation. Note, that I can't really test this, so you might want to give it a whirl. |
| 03:16.33 | Kevin2 | (You'll need to do a cvs up, as I just checked it in.) |
| 03:16.55 | psokolovsky | Kevin2, ok, going to have a bite and building it now ;-) |
| 03:41.39 | Kevin2 | Hrmm. Existing haret calls CeSetThreadQuantum which is a ce-3.0 and later function. (Though, it looks like the call can just be removed.) |
| 03:44.03 | psokolovsky | Kevin2, do I need to apply any patches to cegcc svn trunk? |
| 03:45.27 | psokolovsky | <PROTECTED> |
| 03:45.27 | psokolovsky | <PROTECTED> |
| 03:45.27 | psokolovsky | <PROTECTED> |
| 03:45.27 | psokolovsky | <PROTECTED> |
| 03:45.29 | psokolovsky | ;-( |
| 03:47.10 | psokolovsky | I read on the list your patches should have been applied? |
| 03:48.11 | psokolovsky | oops, almost the same is in your binary... |
| 03:48.42 | Kevin2 | I last built cegcc with rev 777 and that patch in haret cvs. |
| 03:48.57 | Kevin2 | What problem are you seeing? |
| 03:50.41 | psokolovsky | Kevin2, well, just booted using your 0.4.3.exe |
| 03:50.58 | psokolovsky | Kevin2, bit I'm getting "not a valid wince app" with my build ;-( |
| 03:51.51 | Kevin2 | Did you use src/build-mingw32ce.sh to build cegcc? |
| 03:52.49 | psokolovsky | yep |
| 03:54.37 | Kevin2 | Well, you might want to try the steps in the INSTALL file - that is use rev 777 with the patch. |
| 03:55.15 | Kevin2 | Good night. |
| 03:55.41 | psokolovsky | gnight! |
| 04:48.45 | *** join/#htc-linux Funklord (n=cow@213.114.213.203) |
| 04:48.46 | *** join/#htc-linux mndctrl (n=mind@2.81-167-1.customer.lyse.net) |
| 04:48.49 | *** join/#htc-linux scatterp- (n=deminish@212.32.67.196) |
| 04:48.49 | *** join/#htc-linux JTRipper|Prophet (n=jtripper@p54B1CA53.dip.t-dialin.net) |
| 04:48.49 | *** join/#htc-linux g3gg0__ (n=g3gg0@ppp-82-135-67-6.dynamic.mnet-online.de) [NETSPLIT VICTIM] |
| 04:48.49 | *** join/#htc-linux lamikr (n=chatzill@aragorn.kortex.jyu.fi) |
| 04:48.49 | *** join/#htc-linux psokolovsky (n=psokolov@239.usernat.ip.net.ua) |
| 04:48.49 | *** join/#htc-linux LunohoD_ (n=alex@e180071154.adsl.alicedsl.de) [NETSPLIT VICTIM] |
| 04:48.49 | *** join/#htc-linux rejon (n=rejon@64-121-195-22.c3-0.sfpo-ubr4.sfrn-sfpo.ca.cable.rcn.com) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux pierrox|away (n=pierrot@82.240.21.171) |
| 04:48.50 | *** join/#htc-linux cr2 (n=konversa@crpl22.physik.uni-wuppertal.de) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux baud123 (n=baud@ver78-1-82-240-29-152.fbx.proxad.net) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux iggy (n=iggy@gentoo/developer/iggy) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux hlbot (n=adm@iclem.net) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux pof (n=pof@62.57.1.173) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux anYc (i=mario@hadince17.hadiko.uni-karlsruhe.de) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux toi (n=pleemans@d5152D3B4.access.telenet.be) |
| 04:48.50 | *** join/#htc-linux eldu (n=damajor@nysa.e-geek.org) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux parmaster (i=par@dipole.idlepattern.com) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux derekS (n=DerekS@unaffiliated/dereks) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux Kevin2 (n=Kevin@207-237-52-30.c3-0.avec-ubr12.nyr-avec.ny.cable.rcn.com) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux Kmarc (i=kari@markos.biz) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux TeringTuby (n=maarten@189-66-dsl.ipact.nl) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux gw280 (i=authdeni@81.91.110.54) [NETSPLIT VICTIM] |
| 04:48.50 | *** join/#htc-linux ljp (n=lpotter@203.94.178.46) [NETSPLIT VICTIM] |
| 06:38.21 | *** join/#htc-linux iggy (n=iggy@12.45.184.235) |
| 07:11.24 | *** join/#htc-linux LunohoD (n=alex@e180118180.adsl.alicedsl.de) |
| 08:01.18 | *** join/#htc-linux dullard (n=jim@adsl-static-1-30.uklinux.net) |
| 08:07.39 | *** join/#htc-linux Fossi (n=cmewes@195.177.48.115) |
| 09:49.13 | *** join/#htc-linux rejon (n=rejon@64-121-195-22.c3-0.sfpo-ubr4.sfrn-sfpo.ca.cable.rcn.com) |
| 10:14.23 | cr2 | Kevin2: this is FUD about LoadLibrary. http://msdn2.microsoft.com/en-US/library/ms886736.aspx |
| 10:14.37 | cr2 | OS Versions: Windows CE 1.0 and later. |
| 10:14.49 | *** join/#htc-linux booba (n=booba@AAnnecy-256-1-113-18.w90-9.abo.wanadoo.fr) |
| 10:14.55 | cr2 | Link Library: Coredll.lib |
| 10:16.01 | cr2 | GetProcAddress - http://msdn2.microsoft.com/en-US/library/ms885634.aspx |
| 10:16.21 | cr2 | OS Versions: Windows CE 1.0 and later |
| 10:16.41 | cr2 | Link Library: Coredll.lib |
| 10:18.51 | cr2 | For Windows CE 3.0 and later, the ASCII version of this function, GetProcAddressA, is supported. With this version, the lpProcName parameter is of the type LPCSTR. |
| 10:28.53 | *** join/#htc-linux TripleDES (n=sergio@84.77.161.251) |
| 11:24.08 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.179) |
| 12:01.41 | *** join/#htc-linux psokolovsky (n=psokolov@239.usernat.ip.net.ua) |
| 12:53.27 | Kevin2 | cr2: Yeah. I realized I misread the page on LoadLibrary. |
| 14:16.02 | *** join/#htc-linux JTRipper (n=jtripper@p54B1CAAF.dip.t-dialin.net) |
| 14:26.37 | lkcl | well my pxa ioctls are working but not yet doing the job of initialising the phone |
| 14:29.17 | *** join/#htc-linux Speedy2 (n=Javier_6@cpe-70-95-222-168.san.res.rr.com) |
| 14:29.19 | Speedy2 | Hi all |
| 14:30.05 | lkcl | hi speedy |
| 14:30.43 | Speedy2 | Just learned about the effort to get Linux on the htc units, way cool! I used to hack on Linux for SA1110... |
| 14:31.35 | lkcl | ahh |
| 14:31.58 | lkcl | sa1110 - wasn't that the one with the cirrus logic arm processor? |
| 14:32.04 | Speedy2 | No. That's the StrongARM |
| 14:32.11 | Speedy2 | SA == StrongARM |
| 14:32.26 | Speedy2 | The Cirrus Logic ARMs IIRC start with 7's in their model names |
| 14:32.29 | Speedy2 | 7111, etc |
| 14:32.40 | lkcl | been a while |
| 14:33.19 | lkcl | i helped try to rescue the skyminder and i think that i ended up copying a lot of sa1100 code as example drivers |
| 14:34.03 | Speedy2 | What is the skyminder? |
| 14:34.31 | lkcl | a security device with GPRS and GPS in a PDA-sized unit |
| 14:34.44 | Speedy2 | Oh, cool. |
| 14:34.55 | Speedy2 | Who designed it? |
| 14:34.59 | Speedy2 | Garmin? |
| 14:35.03 | Speedy2 | Trimble? |
| 14:35.19 | lkcl | yeh it would have been if they hadn't had a manager who must have cost the company about GBP 400,000 of their 500,000 funding |
| 14:35.23 | lkcl | richard lightman |
| 14:36.18 | lkcl | well - richard did his best, against the total mistrust and stupid decisions of his manager, whom he eventually managed to get through to the directors that the guy was a total nightmare, but the damage was already done |
| 14:36.53 | lkcl | i was brought in at the last minute to try to finish off the device drivers, which they should have done a year before, but it was too late |
| 14:37.05 | Speedy2 | Did they mass produce any hardware? |
| 14:38.03 | lkcl | they produced about 5 revision 0.4 units and 20 revision 0.5 units, and had orders for thousands of units - _if_ they got the 0.4 hardware working. |
| 14:38.45 | Speedy2 | Did they? |
| 14:39.00 | lkcl | unfortunately, what happened was that i got enough hardware working on 0.4 to demonstrate that there was RF feedback from GPRS to the microphone circuit, and they had rushed through the 0.5 hardware a week beforehand... |
| 14:39.40 | lkcl | which was pretty much the story of the entire development cycle - rush,rush,rush - not listening to richard that he needed more time to fix things. |
| 14:39.59 | Speedy2 | sigh |
| 14:40.02 | Speedy2 | I think I know where this is going. |
| 14:40.21 | lkcl | they wouldn't delay by 3 days on the 0.1 to 0.2 revision and that cost them being able to boot linux, so it delayed the entire project by nine months. three days; nine months... |
| 14:40.26 | lkcl | well it's gone now :) |
| 14:41.03 | lkcl | they were one of the first people to actually try to put GSM and GPS on a single device. |
| 14:41.15 | lkcl | they got funding back in 1991 or thereabouts. |
| 14:41.27 | lkcl | got a DTI smartaward and everything. |
| 14:41.35 | lkcl | and it was a 'wow' idea. |
| 14:41.47 | lkcl | but now, every phone and his mum has GPS :) |
| 14:42.27 | lkcl | so they could now actually use commodity hardware like the a780 but the point of having a security device is that it _doesn't_ look like a phone. |
| 14:42.39 | lkcl | what's the first thing that a kidnapper is going to ask you to do? |
| 14:43.13 | lkcl | anyway. hello Speedy2 :) |
| 14:43.24 | lkcl | have you come back to haunt us, then? |
| 14:45.19 | Speedy2 | hehe |
| 14:45.26 | Speedy2 | one sec, phone |
| 14:45.54 | lkcl | la. |
| 14:46.25 | Speedy2 | So how well does Linux work on these HTC devices? |
| 14:46.46 | Speedy2 | I have a Zaurus 5000D and no cell phone...so it would be perfect to hear I could get one and run OPIE on it |
| 14:48.35 | cr2 | Speedy2: 5000D needs an spi driver to make the SD card work ;) |
| 14:49.55 | cr2 | psokolovsky: it seems there is a bug in asic3_pcmcia driver, and wrong info in the ipaq-asic3.h file. |
| 14:49.58 | Speedy2 | I meant status of the HTCs |
| 14:50.09 | Speedy2 | cr2: My 5000D has a battery life of like one day |
| 14:50.17 | lkcl | it depends on the device, dude. i nearly had a useable blueangel device with 2.6.12 except for lcd initialisation, which has since been fixed for 2.6.17, but everything else is a fuck-up because whoever ported it over to hh.org fucked it up |
| 14:50.19 | Speedy2 | So I basically use it as a glorified address book |
| 14:50.36 | lkcl | see http://hands.com/~lkcl/hp6915/linux.phones.txt |
| 14:50.39 | Speedy2 | Will things be "un-fucked" anytime? |
| 14:50.54 | lkcl | i've been writing an article which is still wip |
| 14:50.56 | lkcl | hehe |
| 14:51.14 | cr2 | Speedy2: i don't care, it can run on AC, but the binary SD driver just pisses me off. i can't use my MMC4 cards. |
| 14:51.27 | psokolovsky | hi! |
| 14:51.54 | lkcl | this is reverse-engineering - there are about eight to ten devices, and four to five of us working only _really_ part-time, with cr2 doing most of it |
| 14:52.08 | Speedy2 | cr2: I'm kind of opposite, I don't care about SD, but :/ |
| 14:52.12 | psokolovsky | cr2, pretty possible. I fixed bug in asic3_base today, regarding CPU-independence of irq handling for example ;-) |
| 14:52.45 | cr2 | Speedy2: i'm working really hard on the universal, and it seems the device will have full hw support. don't know about umts. |
| 14:52.59 | Speedy2 | UMTS is the old cell network, right? |
| 14:53.13 | cr2 | newest |
| 14:53.15 | lkcl | all of the devices are 'reaaaally close'. what's holding up universal (and hx4700 and hw6915) is the ak4641 sound isn't in the asoc patch and doesn't work in 2.6.17 |
| 14:53.15 | lkcl | no UMTS is 3G |
| 14:53.43 | Speedy2 | I guess if it supports GSM that would be great |
| 14:54.10 | cr2 | psokolovsky: look at the bottom http://wiki.xda-developers.com/index.php?pagename=ASIC3 |
| 14:54.26 | cr2 | Speedy2: it works on blueangel and universal |
| 14:54.30 | lkcl | blueangel and himalaya _were_ really close in xanadux 2.6.12 and have gone a major step backwards for 2.6.17 hh.org - uda1380 sound was holding things up for himalaya, and uda1380+lcd init was holding things up for blueangel |
| 14:55.06 | lkcl | but again, uda1380 is part of the SoC alsa sound rewrite so it'd be a bit dodgy. |
| 14:55.17 | cr2 | psokolovsky: this entry CF0_SLEEP_MODE seems bogus, but i need to look a bit more on the hx4700 driver. |
| 14:55.22 | lkcl | i only have the one device here with me in the uk - the hw6915 so am working on that. |
| 14:55.53 | lkcl | otherwise i'd definitely have given up on the hw6915 for a bit and tackled my blueangel or himalaya :) |
| 14:55.59 | lkcl | or my universal |
| 14:56.40 | cr2 | lkcl: current himalaya code is awful... |
| 14:56.51 | Speedy2 | The xda site is not coming up for me. |
| 14:56.53 | cr2 | but we should fix blueangel first. |
| 14:57.57 | cr2 | Speedy2: works for me. |
| 14:58.09 | Speedy2 | bizarre. |
| 14:58.20 | Speedy2 | Let me try resolving it elsewhere |
| 14:58.21 | lkcl | cr2: probably. that's because i wrote a lot of it :) |
| 14:59.15 | cr2 | lkcl: this was the first. if i have not seen it, i would not buy blueangel. |
| 14:59.45 | Speedy2 | hrm, site still not coming up |
| 14:59.58 | Speedy2 | So basically is the word on the street do NOT get an HTC for a little while ? |
| 15:00.22 | lkcl | well if you're a kernel hacker, then i would say definitely get one. |
| 15:00.35 | lkcl | but if you're not - then yes, defintely don't get one. |
| 15:00.37 | Speedy2 | hrm |
| 15:00.54 | cr2 | Speedy2: this is the best "linux" hardware. EU moto a780 does not count anymore. |
| 15:01.00 | Speedy2 | Why is that? |
| 15:01.31 | cr2 | what ? |
| 15:01.32 | lkcl | the universal is the sexiest one. 640x480 screen, two cameras, 3G, half-sized _complete_ keyboard (1234567890 _and_ QWERTY) |
| 15:01.54 | lkcl | 802.11, bluetooth, irda |
| 15:02.15 | lkcl | it's basically a mini-laptop. oh. and a 520mhz arm processor. |
| 15:02.42 | Speedy2 | Who's ARM? A Samsung ARM9? |
| 15:02.47 | cr2 | which goxboxlive once pushed to 624MHz |
| 15:02.49 | lkcl | intel pxa |
| 15:02.52 | lkcl | eek! :) |
| 15:02.56 | Speedy2 | Is is the EU a780 no longer count as sexy? |
| 15:02.58 | Speedy2 | wee |
| 15:02.59 | Speedy2 | err |
| 15:03.03 | Speedy2 | Why does the EU a780 no longer count. |
| 15:03.07 | Speedy2 | <cr2> Speedy2: this is the best "linux" hardware. EU moto a780 does not count anymore |
| 15:03.09 | *** part/#htc-linux WizMaui (n=WizMaui@62.112.90.179) |
| 15:03.31 | lkcl | well... it's only a phone :) |
| 15:03.43 | cr2 | it's not produced anymore, and the development seems to be dead after this openmoko stuff. |
| 15:04.03 | lkcl | the only thing missing from the universal is a hard drive and usb-host (rather than usb-client) |
| 15:04.19 | Speedy2 | How hard is it to acquire any of this HTC hardware? |
| 15:04.22 | cr2 | Speedy2: i have a780, it has gps, route planner and maps for the whole europe. |
| 15:04.37 | lkcl | ebay or a decent phone shop. |
| 15:04.43 | Speedy2 | Woah, can you extract the program it runs (GPS) and run it on another PDA, etc? |
| 15:05.14 | lkcl | on the a780? hmm... actually... *thinks*... |
| 15:05.16 | cr2 | Speedy2: and it's a "real" linux phone from the beginning. |
| 15:05.26 | lkcl | depends if it's qtopia-based or gpe-based... |
| 15:06.22 | lkcl | http://cgi.ebay.co.uk/XDA-Executive-Fully-Boxed-on-O2_W0QQitemZ280047694431QQihZ018QQcategoryZ3312QQrdZ1QQcmdZViewItem |
| 15:06.58 | lkcl | http://cgi.ebay.co.uk/NEW-XDA-EXECUTIVE_W0QQitemZ130046430636QQihZ003QQcategoryZ3312QQrdZ1QQcmdZViewItem |
| 15:07.21 | Speedy2 | So people have stopped work on the a780? That's silly, isn't the phone pretty new? |
| 15:07.40 | lkcl | http://cgi.ebay.co.uk/UNLOCKED-PDA-O2-XDA-2s-XDA-IIs-MDA-SPV_W0QQitemZ150055835752QQihZ005QQcategoryZ3312QQrdZ1QQcmdZViewItem |
| 15:07.57 | lkcl | no, they haven't stopped (talk to stefan on #openezx) |
| 15:08.26 | lkcl | and it's not really new - it's about two to three years old now. |
| 15:09.26 | lkcl | http://cgi.ebay.co.uk/O2-XDA-II-with-Pocket-Streets-Destinator_W0QQitemZ120052858142QQihZ002QQcategoryZ3312QQrdZ1QQcmdZViewItem |
| 15:11.20 | Speedy2 | I'm interested to know if the GPS mapping program for the a780 can be used on other platforms |
| 15:11.29 | Speedy2 | Did the a780 come with the GPS program cr2? |
| 15:18.51 | lkcl | speedy2: quite probably, if you compile up the right distro. |
| 15:18.59 | lkcl | and yes it does. |
| 15:19.09 | lkcl | the best people to ask are on #openezx |
| 15:19.16 | cr2 | Speedy2: it's from alk. but it is compiled by a modified gcc |
| 15:19.22 | lkcl | and then there's always opensource.motorola.org |
| 15:19.30 | lkcl | o god you're kidding, cr2? haha |
| 15:19.42 | lkcl | y |
| 15:19.52 | cr2 | and it is locked to the device, so you can run it elsewhere only as warez ;) |
| 15:20.07 | lkcl | dumb idiots... *sigh* |
| 15:20.11 | cr2 | lkcl: they have a different toolchain. |
| 15:20.15 | Speedy2 | alk |
| 15:20.16 | lkcl | oh good grief. |
| 15:20.18 | Speedy2 | CoPilot |
| 15:20.21 | cr2 | yes |
| 15:20.27 | Speedy2 | Wow, they ported it |
| 15:20.36 | lkcl | btw cr2 you wanna help me sort out this phone stuff? |
| 15:20.39 | Speedy2 | If they didn't have their heads up their ass they would make a version available for Linu |
| 15:20.40 | Speedy2 | Linux |
| 15:21.33 | cr2 | lkcl: we need a good dump of nk.exe and rilgsm.dll |
| 15:21.39 | cr2 | Speedy2: they did it only for motorola. |
| 15:22.02 | lkcl | o i was thinking more along the lines of comparing a wi / gpio dump to the kernel and userspace code i created |
| 15:22.19 | cr2 | it leaks memory sometimes, pushes you to the highway at a first opportunity, but is pretty good. |
| 15:22.42 | cr2 | if i not a free software zealot, i'd like it ;) |
| 15:22.59 | cr2 | s/i not/i was not/ |
| 15:23.23 | Speedy2 | No one has a GPS mapping/navigation program native Linux |
| 15:23.26 | Speedy2 | If these guys are smart... |
| 15:23.32 | lkcl | there's openstreetmap.org |
| 15:23.51 | cr2 | Speedy2: commercial. but there is zroadmap. |
| 15:24.14 | Speedy2 | What? |
| 15:24.17 | lkcl | and phil hands is considering buying a license of the converter-program which some guy wrote to turn the openstreetmap.org raw data into commercial-compatible format |
| 15:24.17 | cr2 | if you _have_ good maps, then zroadmap will be able to use them. |
| 15:24.53 | Speedy2 | Realistically, we need something that can read the Navteq databases |
| 15:25.27 | cr2 | lkcl: openstreetmap is nowhere for the street navigation. i tell you that as a major contributer there ;) |
| 15:25.35 | lkcl | ha ha |
| 15:25.55 | cr2 | Speedy2: if they are in shapfile/mapinfo format, then it's not a problem. |
| 15:26.32 | lkcl | why the hell isn't roadmap.sf.net in debian?? |
| 15:27.07 | cr2 | Speedy2: btw, look at navit. http://navit.sf.net . these are good maps :D |
| 15:27.12 | Speedy2 | cr2: I bet it's in a proprietary format. if you want some samples... |
| 15:27.37 | cr2 | lkcl: it is in OE. |
| 15:27.38 | Speedy2 | Hrm, any screen shots of this program? |
| 15:27.53 | lkcl | http://roadmap.digitalomaha.net/screenshots.html |
| 15:28.00 | lkcl | yeh but _that's_ no good :) |
| 15:28.29 | cr2 | Speedy2: proprietary does not help. |
| 15:29.03 | cr2 | the navit maps for germany are very good. |
| 15:29.05 | Speedy2 | I don't know, I don't know much about the GIS data formats. |
| 15:29.20 | cr2 | Speedy2: how are your sample named ? |
| 15:29.54 | Speedy2 | .dat, .psi, .mad , .its , .vlf, .dct, .ndx |
| 15:30.01 | cr2 | lkcl: these are not good screenshots. there is a fork called roadmap_editor, it supports turn restrictions and route planning. |
| 15:30.18 | Speedy2 | it seems that .mad and .its are the largest files |
| 15:30.30 | cr2 | Speedy2: imho, that does not help. |
| 15:30.37 | Speedy2 | I know :/ |
| 15:31.14 | Speedy2 | I see this header in one file |
| 15:31.14 | Speedy2 | ITOLITL |
| 15:31.17 | Speedy2 | ITOLITLS |
| 15:31.18 | cr2 | the alk CDs i got with a780 also have whole europe in some weird format. |
| 15:31.33 | Speedy2 | ALK has their own data map service |
| 15:31.36 | Speedy2 | (apparently) |
| 15:33.55 | cr2 | Speedy2: read the navit site, if you are interested in the european maps. |
| 15:38.45 | lkcl | cr2: can you read wi traces of serial data, directly, including the other 16550uart data? |
| 15:39.18 | lkcl | http://hands.com/~lkcl/hp6915/btuart.wi.gpio.trace2.txt |
| 15:39.25 | cr2 | with the python script in haretconsole. |
| 15:41.10 | lkcl | it didn't work i tried. |
| 15:41.57 | lkcl | it only does data in and data out, not control-commands DTR etc. etc. |
| 15:42.21 | Speedy2 | lkcl: Do you have any programs that can simulate GPS data? I'm trying to some testing with a program and I'd like to convince the PC it's in a car following a route |
| 15:42.25 | lkcl | i'm missing something - and i don't know what |
| 15:42.40 | lkcl | gpsd has an emulation mode |
| 15:43.25 | lkcl | you _should_ be able to grab any NMEA data from anywhere even from a garmin gps |
| 15:43.37 | Speedy2 | and just send it back to the serial port, eh? |
| 15:43.40 | lkcl | and then feed that back in. |
| 15:43.46 | lkcl | well, not quite, speedy2 :) |
| 15:44.08 | lkcl | but you get your program to read from a file instead from serial, just like gpsd does |
| 15:44.24 | Speedy2 | Well, I meant log the data, take a second PC and have it spew the data out via it's serial port |
| 15:44.43 | lkcl | oh yeh that'd likely do it, too. |
| 15:44.56 | lkcl | its!!! ITS!!! not 'via it is serial port'!!! |
| 15:45.05 | cr2 | lkcl: works for me. |
| 15:45.23 | Speedy2 | It's improper grammar but it seems appropriate |
| 15:45.29 | lkcl | ok will try it... but it doesn't do DTR etc. |
| 15:45.59 | lkcl | yes that's because nobody can frickin learn that's why. |
| 15:46.05 | cr2 | well, not DTR. |
| 15:46.37 | lkcl | yeh and CTS / RTS and DTR is what i suspect is kicking the eriksson radio-rom on to the next stage of initialisation |
| 15:47.09 | cr2 | Speedy2: there is a "gpsfake" program in the gpsd |
| 15:48.05 | cr2 | lkcl: isnt't RTS/CTS hardware flow control ? can you really follow there changes in software ? |
| 15:48.35 | lkcl | well i just set the tracemask to 0x1f and ... oh yeh good point |
| 15:48.48 | lkcl | ok - maybe what i said was misleading |
| 15:48.50 | cr2 | btuart does not have DTR/DSR afaik. |
| 15:49.00 | *** join/#htc-linux pH5 (n=ph5@e178195031.adsl.alicedsl.de) |
| 15:49.01 | cr2 | it is simulated by a gpio. |
| 15:49.18 | lkcl | GPIO 42 |
| 15:49.28 | cr2 | ok. |
| 15:50.04 | lkcl | ok what's _this_? 0124ef97: debug 096792b0: e5812000(str) 0000005d ae200004 |
| 15:50.21 | cr2 | ? 42 is RXD. |
| 15:50.51 | lkcl | oh ok |
| 15:51.14 | cr2 | i don't rememember the 16550 register layout. |
| 15:51.32 | lkcl | 0x5d is being written to the oh it's the interrupt enable register |
| 15:52.15 | cr2 | +4 is IRQ enable. |
| 15:52.18 | lkcl | 15823000(str) 000000c5 ae200008 |
| 15:52.26 | lkcl | 15823000(str) 000000c3 ae200008 |
| 15:52.33 | cr2 | +8 is IRQ id |
| 15:52.34 | lkcl | e5813000(str) 0000002b ae200010 |
| 15:52.50 | cr2 | modem control. |
| 15:53.25 | lkcl | ok. |
| 15:53.50 | lkcl | you think any of those would have an impact on what's going on? |
| 15:54.02 | lkcl | changing the irq id twice seems a bit weird |
| 15:54.41 | lkcl | oh wait when you write to it it's the FIFO control register |
| 15:54.55 | cr2 | MCR -- it's RTS (readonly) |
| 15:54.55 | lkcl | ah _ha_. |
| 15:55.12 | lkcl | +8 is FIFO control register on write. |
| 15:56.17 | lkcl | 0xc5 is UART_FCR_ENABLE_FIFO | FCR_CLEAR_XMIT | UART_FCR_T_TRIG_00 |
| 15:56.54 | lkcl | 0xc3 is UART_FCR_ENABLE_FIFO | FCR_CLEAR_RCVR | UART_FCR_T_TRIG_00 |
| 15:57.04 | lkcl | i think that's it... |
| 15:57.11 | lkcl | it's gotta be |
| 15:57.37 | lkcl | oh for goodness sake. |
| 15:57.59 | *** join/#htc-linux lamikr (n=chatzill@aragorn.kortex.jyu.fi) |
| 16:01.11 | lkcl | ok i think i got it. there's a pre-init stage and a post-init stage |
| 16:02.29 | lkcl | in the serial code |
| 16:20.24 | lkcl | oops i accidentally got opie running :) |
| 16:21.22 | cr2 | psokolovsky: yes, there is a bug in ipaq-asic3.h header, but only hx4700_pcmcia is affected. |
| 16:21.59 | cr2 | lkcl: we need a generic tsc2046 driver. |
| 16:22.02 | lkcl | ack |
| 16:27.02 | lkcl | yeh i hacked a copy of the htcuniversal_ts2.c code |
| 16:28.06 | pH5 | ack |
| 16:28.21 | pH5 | good evening :) |
| 16:29.00 | lkcl | evening pH5 |
| 16:30.25 | *** join/#htc-linux pleemans (n=peter@d51A5E421.access.telenet.be) |
| 16:31.14 | psokolovsky | cr2, could you post to k-d on it? |
| 16:38.40 | cr2 | psokolovsky: i think only the owners of asic3 documentation have something to say ;) |
| 16:39.16 | lkcl | i'm getting a couple of interrupts from the eriksson-gsm-module. |
| 16:40.23 | cr2 | lkcl: on what line and under which conditions ? |
| 16:40.30 | cr2 | #define ASIC3_EXTCF_OWM_SMB (1 << 5) /* OWM bus selection */ |
| 16:40.30 | cr2 | #define ASIC3_EXTCF_CF0_SLEEP_MODE (1 << 6) /* CF0 sleep state control */ |
| 16:40.30 | cr2 | #define ASIC3_EXTCF_CF1_SLEEP_MODE (1 << 8) /* CF1 sleep state control */ |
| 16:40.44 | lkcl | BTUART. after sending the sequence of 5 asic3 gpio inits |
| 16:40.46 | psokolovsky | cr2, do you want to change something? let's have it reviewed. or even if not, pls make it noted somewhere, as irc log searching is not very scalable ;-) |
| 16:41.53 | cr2 | should be |
| 16:41.55 | lkcl | now i have opie starting i must switch off irda :) |
| 16:41.57 | cr2 | #define ASIC3_EXTCF_OWM_SMB (1 << 5) /* OWM bus selection */ |
| 16:41.57 | cr2 | #define ASIC3_EXTCF_UNDOC_BIT6 (1 << 6) /* used by OWM and PCMCIA */ |
| 16:41.58 | cr2 | #define ASIC3_EXTCF_CF0_SLEEP_MODE (1 << 7) /* CF0 sleep state control */ |
| 16:41.58 | cr2 | #define ASIC3_EXTCF_CF1_SLEEP_MODE (1 << 8) /* CF1 sleep state control */ |
| 16:42.37 | lkcl | ok i gonna need something else to do, soon, cos i gonna go mad. |
| 16:42.53 | lkcl | like try to find acx. |
| 16:44.16 | cr2 | psokolovsky: i'll post it to the ml. |
| 16:44.43 | psokolovsky | thanks! |
| 16:45.19 | psokolovsky | just noone can process all the issues in realtime, and it's worth to ponder about many of them ;-) |
| 16:46.49 | cr2 | the hx4700_pcmcia setup code is also not quite kosher, but since it works... |
| 16:53.41 | lkcl | btw - what's the 'clockrate' on the wi timings? |
| 16:53.51 | cr2 | 1msec |
| 16:54.24 | lkcl | ok so 109107 - 109091 is 16 ms right? |
| 16:54.36 | lkcl | GPIOD4 S 1 ? 109091 |
| 16:54.36 | lkcl | GPIOD8 S 1 ? 109107 |
| 16:54.53 | lkcl | the delay between asic3 gpio d8 and d4 is appx 16 ms right? |
| 16:55.15 | cr2 | yes. |
| 16:55.15 | lkcl | ok. |
| 16:55.15 | lkcl | then why the xxxx isn't this working? :) |
| 16:55.29 | cr2 | but there is a SleepTillTick() call there, so... |
| 16:55.38 | cr2 | how do you test it ? |
| 16:55.51 | cr2 | cu -l /dev/ttyS1 -s 921600 ? |
| 16:56.30 | cr2 | busybox can't 921600 last time i have looked there. |
| 16:56.51 | cr2 | stty in busybox, to be more accurate. |
| 16:56.59 | lkcl | http://hands.com/~lkcl/hp6915/serial.c |
| 16:57.30 | lkcl | _really_ simple - opens /dev/ttyS1 and writes a binary blob to it then reads the response |
| 16:57.34 | cr2 | hmm. *cfspeed ? |
| 16:57.47 | cr2 | use 'chat' ? |
| 16:58.04 | lkcl | but the init sequence - reading GPIOB0 - is still 0 instead of 1 |
| 16:58.16 | cr2 | have you done 'stty' before writing there ? |
| 16:58.25 | cr2 | hmm. |
| 16:58.26 | lkcl | no - but it's even before that stage |
| 16:58.34 | cr2 | ok, i see. |
| 16:58.51 | lkcl | am gonna just try 20ms delays all round |
| 16:59.18 | lkcl | see what that does :) |
| 17:00.35 | lkcl | are these wi timings actually accurate, to within +/- 1ms? |
| 17:01.35 | lkcl | htcsable configure phone: 0 |
| 17:01.35 | lkcl | GSM_DCD (?) tries=4 of 3 |
| 17:01.53 | lkcl | which means: first stage init done (5 gpios) followed by B0 still being 0 instead of 1 |
| 17:06.13 | *** join/#htc-linux goxboxlive (n=goxboxli@9.80-202-160.nextgentel.com) |
| 17:07.04 | cr2 | lkcl: that's a hardware timer. did you just copy the DCD code from universal ? |
| 17:07.12 | lkcl | DCD code? :) |
| 17:07.16 | lkcl | what DCD code? |
| 17:08.32 | cr2 | GSM_DCD (?) tries=4 of 3 |
| 17:08.49 | cr2 | it was my quick and dirty hack to mimick wince. |
| 17:08.58 | lkcl | ah yes that yes i did |
| 17:09.14 | lkcl | _that_ dcd code :) |
| 17:10.10 | *** join/#htc-linux booba (n=booba@AAnnecy-256-1-113-18.w90-9.abo.wanadoo.fr) |
| 17:10.11 | cr2 | btw, what is IPM ? |
| 17:10.52 | lkcl | IPM? interrupts per minute? no idea. |
| 17:10.58 | cr2 | intel power manager, probably. |
| 17:11.05 | lkcl | ahh. |
| 17:13.24 | cr2 | lkcl: DCD is 0x80 bit. can't say on which ABCD. because the data seg is fux0red. |
| 17:14.03 | lkcl | that's 1<<7 |
| 17:14.53 | cr2 | AC is 0x100 |
| 17:15.36 | lkcl | AC is 0x100 - that's D8 |
| 17:16.12 | lkcl | none of the bits i'm looking at here from the wasic3gpio trace are 0x80 |
| 17:16.19 | cr2 | so 68 is D. |
| 17:16.53 | lkcl | B0, B2 are input-checkers (which i thought were the dcd and bb things) |
| 17:17.34 | lkcl | the power-up sequence is D4, D8, C12, C13, D5 |
| 17:18.01 | lkcl | then 100ms wait checking for B0 |
| 17:18.41 | cr2 | 0x200 is the hp irq. |
| 17:19.00 | lkcl | then a BTUART irq and some UART initialisation, and GPIO42 goes 1, 0 then a BTUART irq and GPIO42 goes 1 |
| 17:19.31 | lkcl | 0x200 => 1<<9 |
| 17:20.12 | lkcl | D4: 0x10, D8: 0x100, C12: 0x1000, C13: 0x2000, D5: 0x20 |
| 17:21.11 | lkcl | then we have the BTUART initialisation, including resetting the FIO just like in pxa.c |
| 17:21.17 | cr2 | both are missing in your list. |
| 17:21.26 | lkcl | then B2 is checked for being 1 |
| 17:21.40 | lkcl | what are missing, dude? |
| 17:22.28 | lkcl | 0x200 - the hp irq? |
| 17:22.57 | cr2 | headphone |
| 17:23.17 | cr2 | 0x2 - miniSD ? |
| 17:23.41 | lkcl | oh headphone right. |
| 17:24.52 | cr2 | 0x400 :Ext, whatever that means. had problems on other phones already ;) |
| 17:25.02 | lkcl | oo i got a little blue light to go on! |
| 17:26.47 | cr2 | 0x4 - U |
| 17:26.57 | cr2 | i need the data seg... |
| 17:28.31 | lkcl | doesn't help that stty reports that the baud rate is 9600... |
| 17:29.31 | Speedy2 | For the HTCs, is the phone part (GSM) sufficiently independant of the PDA? |
| 17:29.32 | cr2 | stty -F /dev/ttyS0 speed 921600 ? |
| 17:29.48 | Speedy2 | i.e. does the GSM chipset just continually grab data from the mic and "cell phone it" out? |
| 17:29.54 | cr2 | then your stty is bad. |
| 17:29.57 | cr2 | Speedy2: yes. |
| 17:30.09 | cr2 | on most phones. |
| 17:30.16 | lkcl | invalid arg 921600 but 115200 worked |
| 17:30.51 | cr2 | there are phones that store GSM calibration data in main RAM. try to run haret on them ;) |
| 17:30.59 | lkcl | yep btuart is 921600 baud according to dump pxa2xxserial |
| 17:31.01 | cr2 | lkcl: your stty is fscked. |
| 17:31.04 | lkcl | hooray! |
| 17:31.08 | cr2 | try 'cu' |
| 17:31.18 | lkcl | o rats i just rebooted |
| 17:31.31 | cr2 | but i'm not sure here too. |
| 17:31.41 | cr2 | are you using busybox ? |
| 17:32.13 | cr2 | then submit a patch. probably to uclibc, or how it is called. |
| 17:33.13 | lkcl | busybox, yep. |
| 17:33.52 | cr2 | you are the first who has 921600 uart. |
| 17:34.28 | lkcl | hooray! |
| 17:34.30 | cr2 | i had this problem with BT on universal, but it is solved by hciattach. |
| 17:34.42 | cr2 | no need for stty. |
| 17:37.25 | lkcl | hm cu not on here |
| 17:37.48 | lkcl | hm i might just code it up into my little program - do the ioctl myself. |
| 17:38.03 | cr2 | ttycfsetspeed ? |
| 17:38.43 | cr2 | no. what's the posix name ? |
| 17:39.01 | lkcl | i downloaded the source of minicom earlier... |
| 17:39.17 | lkcl | tcsetattr |
| 17:39.22 | lkcl | cfsetispeed |
| 17:39.50 | cr2 | man termios :) |
| 17:40.49 | cr2 | #define B921600 0010007 |
| 17:40.53 | lkcl | dats it |
| 17:40.58 | lkcl | ok gimme 1sec to write code... |
| 17:41.03 | cr2 | :) |
| 17:43.04 | lkcl | any other suggestions on params? |
| 17:43.10 | lkcl | while i'm at it? :) |
| 17:44.24 | cr2 | rts/cts |
| 17:44.27 | cr2 | 8bit |
| 17:44.31 | cr2 | 1stop |
| 17:46.34 | goxboxlive | cr2: Have you got Opie running? |
| 17:46.56 | goxboxlive | I have noticed something we might could use as a phone. |
| 17:48.20 | Speedy2 | It would be awesome get the HTC with a SIP client as well |
| 17:48.44 | cr2 | goxboxlive: no. i'm looking at the charging code. |
| 17:48.54 | goxboxlive | Speedy2 there is a SIP client for Opie it's named Kphone |
| 17:48.58 | Speedy2 | Is charging under software control on the HTCs? |
| 17:49.08 | cr2 | Speedy2: it has nothing to do with HTC. |
| 17:49.10 | Speedy2 | (that's so offensive, battery charging under main software control) |
| 17:49.23 | goxboxlive | cr2: Then i wouldnt disturb you, we can take it later. I'll try yo play with it. |
| 18:00.36 | lkcl | cr2 - why are both htcuniversal_ts.c _and_ htcuniversal_ts2.c included? |
| 18:01.05 | lkcl | in arch/arm/mach-pxa/htcuniversal/Makefile ? |
| 18:02.14 | cr2 | lkcl: because the _ts.c is a port of hx4700 .16 driver, and _ts2.c of .12 driver. .12 works better. |
| 18:02.24 | lkcl | oh. |
| 18:02.26 | lkcl | err |
| 18:02.34 | cr2 | the same with the sound ;) |
| 18:03.07 | lkcl | the ts2.c one i can't get it to read any interrupts or anything |
| 18:03.09 | cr2 | if we have ported the .12 driver we'd have working sound. |
| 18:03.31 | cr2 | ts2.c is for asic3 irq. |
| 18:03.48 | cr2 | take the irq code from .12 hx4700 driver. |
| 18:03.52 | lkcl | i changed that (asic3 irq to gpio) in both |
| 18:03.56 | lkcl | oh right good point |
| 18:04.12 | goxboxlive | cr2: just port it then just until you have the .16 working |
| 18:04.17 | lkcl | yeh i should use hx4700 stuff |
| 18:04.29 | cr2 | goxboxlive: too much work ;) |
| 18:04.44 | lkcl | ok that light happily goes on each time i access /dev/ttyS1 |
| 18:05.30 | cr2 | lkcl: there are 3 params for the tsc2046 driver: SSP1 or 2, irq and the evil powersaving bits. HTC made the life difficult. |
| 18:05.40 | cr2 | which light ??? |
| 18:05.42 | lkcl | hooray |
| 18:05.42 | goxboxlive | ok |
| 18:05.51 | lkcl | the blue light next to the power button |
| 18:06.07 | cr2 | bluetooth ? |
| 18:06.16 | lkcl | doesn't sound quite right, does it... |
| 18:06.28 | cr2 | then you must be pulling too much gpios ;) |
| 18:06.39 | lkcl | o dear. |
| 18:07.01 | cr2 | can't you check it with 'btctl 10' and 'btctl 11' which one is for the bluetooh ? |
| 18:07.19 | lkcl | *sigh* |
| 18:07.45 | lkcl | ok. BTUART is used for gsm. |
| 18:07.55 | cr2 | BTUART is only the name. |
| 18:07.58 | lkcl | FFUART is used for phone. |
| 18:08.08 | lkcl | yeh |
| 18:08.14 | cr2 | has nothing to do with bluetooth. |
| 18:08.36 | cr2 | ok, on universal it's like you've written. |
| 18:09.05 | lkcl | so /dev/ttyS1 is the one (inappropriately) named BTUART |
| 18:09.38 | lkcl | and /dev/ttyS1 is definitely linked to /dev/tts/1 |
| 18:09.52 | lkcl | and my debug messages are definitely appearing then |
| 18:10.28 | lkcl | and why is there a getty on /dev/ttyS0??? |
| 18:10.56 | cr2 | vi /etc/inittab ? |
| 18:11.17 | cr2 | because on some fscking ipaq it is (was) so. |
| 18:11.26 | lkcl | nothing in there |
| 18:11.46 | cr2 | getty is called from /etc/inittab |
| 18:12.01 | cr2 | otherwise you can just kill it. |
| 18:12.20 | lkcl | which explains why bluetooth initialisation is happening :) |
| 18:12.51 | cr2 | LOL. |
| 18:13.32 | cr2 | it was better on universal (and apache). the zimage decompression dots went directly to the phone :) |
| 18:13.46 | lkcl | ha ha i was editing my /etc/inittab |
| 18:16.31 | lkcl | so... about the ts in universal - it still doesn't quite explain why both are in there. are they under different entries in /dev/input or something? |
| 18:17.25 | goxboxlive | cr2: this is strange |
| 18:17.32 | goxboxlive | I called my self |
| 18:17.51 | cr2 | so ? |
| 18:17.52 | goxboxlive | I can hear in one of the speakes |
| 18:18.05 | cr2 | LOL!!!!!! |
| 18:18.06 | goxboxlive | i think it is only the clamshell speaker |
| 18:18.18 | goxboxlive | the mic doesnt work |
| 18:18.36 | goxboxlive | There where some echo and dealy, but we are realy on our way :-) |
| 18:18.41 | cr2 | are there 3 speakers ? |
| 18:19.02 | goxboxlive | yes there are one in the back of the lcs |
| 18:19.11 | goxboxlive | s/lcs/lcd/ |
| 18:19.15 | cr2 | 2 in the box and 1 in the cover ? |
| 18:19.27 | goxboxlive | yes |
| 18:19.36 | cr2 | that explains gpioa1 and a6. |
| 18:19.44 | goxboxlive | havent you noticed???????????????? |
| 18:19.47 | cr2 | or is a possible explanation. |
| 18:19.58 | cr2 | never thought about it :) |
| 18:20.32 | goxboxlive | I also tried headphones, but there where no sound in them, the sound was still out of the lcd speaker |
| 18:20.34 | cr2 | hmm. or are they switched in parallel ? |
| 18:20.48 | goxboxlive | no i dont think so |
| 18:21.06 | cr2 | there is an AT command to switch between BT,HP and speaker. |
| 18:21.26 | cr2 | universal is a complex device. |
| 18:21.46 | cr2 | analog and digital audio routing is not easy to guess. |
| 18:21.54 | goxboxlive | When i play a song in WinCE there are only sound out of the two ones, not the lcd one |
| 18:22.14 | goxboxlive | I guess the LCD one is only activated when the clam is locked |
| 18:22.21 | cr2 | hmm. then it must be the AT command. |
| 18:22.46 | goxboxlive | Actually cr2, there are two speakers in the lcd |
| 18:22.53 | lkcl | there may be AT commands to increase volume |
| 18:22.57 | cr2 | so it is probably connected to the GSM |
| 18:23.02 | goxboxlive | I forgot the one at top of the screen |
| 18:23.28 | cr2 | lkcl: yes, but also the switching. universal is more complex than sable. |
| 18:23.28 | goxboxlive | There are one over the lcd, and one at back of the lcd screen. You can see the holes |
| 18:23.33 | lkcl | good grief. |
| 18:24.03 | cr2 | these are certainly parallel ? |
| 18:24.04 | lkcl | yeh, i can tell - by you having two touchscreen drivers, dude |
| 18:24.15 | cr2 | i did not open the lcd cover. |
| 18:24.28 | goxboxlive | hold on i'll try to call my self |
| 18:24.40 | cr2 | lkcl: we have 2 cams, so what's with 4 speakers ? :) |
| 18:25.29 | cr2 | why didn't they solder the gps chip in universal ... |
| 18:25.32 | cr2 | would be a real monster device. |
| 18:25.41 | goxboxlive | No, they are not parallell |
| 18:25.51 | cr2 | stereo ??? |
| 18:25.53 | lkcl | only an extra $30 retail price... |
| 18:26.09 | lkcl | yes of course they're stereo speakers, cr2!!!! |
| 18:26.15 | lkcl | couldn't you guess?? |
| 18:26.15 | goxboxlive | The one in back of lcd screen is only active when the clam shell is clsed ( the lcd screen not visable) |
| 18:26.28 | cr2 | but the MSM chip can do accel 3D, so i'm not surprised. |
| 18:27.11 | cr2 | goxboxlive: then we should do more AT logging. i have never tried it in this config. |
| 18:27.14 | goxboxlive | The one at the lcd screen side is active when the clamshell is open |
| 18:27.20 | lkcl | argh there goes that little blue light again |
| 18:27.21 | goxboxlive | ok |
| 18:27.37 | cr2 | aah. PDA mode and GSM mode in wavedev.dll |
| 18:28.07 | goxboxlive | cr2: Universal? |
| 18:28.20 | cr2 | one head is good, many heads is better. |
| 18:28.21 | cr2 | yes. |
| 18:29.08 | pH5 | cr2: are the AT commands for audio switching issued from wavedev.dll on the universal? |
| 18:29.19 | cr2 | lol. so it's not the 2 speakers where you hear the call sound :))) |
| 18:29.42 | goxboxlive | so, actually, we have sound working now. Only needs the right gpio's to route it the right way. Am i right? |
| 18:30.27 | cr2 | pH5: http://wiki.xda-developers.com/index.php?pagename=HTC_GSM |
| 18:30.57 | cr2 | AT@AUDIOSET =%u,0,0,%u U+ $1=0,2,3,7 ; $2=2 |
| 18:30.58 | goxboxlive | But one thing is for sure, we have only one mic. |
| 18:31.43 | cr2 | goxboxlive: i'm not quite sure about the speakers, but the mic is certainly switched in parallel. |
| 18:32.03 | cr2 | you can test it in wince. call a phone, and crash wince :) |
| 18:32.24 | cr2 | bt continues working. also on ba. |
| 18:32.28 | goxboxlive | cr2: There is only one mic. Not two in parallell. |
| 18:32.53 | cr2 | mic is connected to the ak4641. that's for sure. |
| 18:33.16 | cr2 | don't know about the MSM chip. but it's easier that way. |
| 18:34.10 | cr2 | otherwise you need to ADC the data and transfer over SSP. it is doable only on ba/hima/magician where they split I2S and SSP. |
| 18:35.12 | cr2 | pH5: such things are DevSpecific. add your data here -- http://wiki.xda-developers.com/index.php?pagename=rilgsm.dll |
| 19:26.46 | pH5 | cr2: done |
| 19:28.54 | cr2 | pH5: great. they are rather consistent. does your phone work ? |
| 19:32.02 | pH5 | no. I would have bragged about it here. |
| 19:34.28 | cr2 | :) |
| 19:37.54 | *** join/#htc-linux Speedy2 (n=Javier_6@cpe-70-95-210-2.san.res.rr.com) |
| 19:42.16 | lkcl | wb speedy2 |
| 19:58.40 | lkcl | HOORAY i got a read 0 bytes |
| 19:59.40 | lkcl | i think i done it... |
| 20:00.33 | lkcl | ok i managed it once |
| 20:03.28 | lkcl | argh it was a simple bug: i wasn't calling the asic3 init functions correctly. ha ha |
| 20:03.32 | lkcl | very funny. |
| 20:03.32 | lkcl | not |
| 20:03.43 | Speedy2 | re |
| 20:12.59 | *** join/#htc-linux BabelOued (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 20:18.09 | lkcl | AH HAAHAHAHAHAHAAAA |
| 20:18.11 | lkcl | i got it. |
| 20:21.22 | cr2 | what ? |
| 20:22.25 | lkcl | gsm initialisation. |
| 20:22.39 | lkcl | it's working |
| 20:22.56 | lkcl | i made a stupid mistake using the asic3 set function. |
| 20:23.30 | lkcl | which leads me to believe that there's a bug in the asic3 set routine. |
| 20:23.37 | lkcl | or at least a safety feature missing |
| 20:24.23 | cr2 | ok. can you send the data ? |
| 20:25.07 | lkcl | i'm just copying what i'm seeing in http://hands.com/~lkcl/hp6915/btuart.trace.wi.decoded.txt |
| 20:27.14 | cr2 | does it answer ? |
| 20:35.47 | lkcl | yep! |
| 20:36.01 | lkcl | i'm up to the 3rd exchange, copying that trace |
| 20:36.59 | lkcl | oops - 2nd exchange fails, with a timeout. |
| 20:37.21 | lkcl | hmmm... |
| 20:40.35 | cr2 | lkcl: can you try a new haret to dump nk.exe and rilgsm.dll. ida barfs, but the file looks more or less ok. |
| 20:40.41 | lkcl | sure |
| 20:40.51 | cr2 | i'll upload it. |
| 20:41.18 | lkcl | ack |
| 20:43.30 | cr2 | done. |
| 20:46.49 | *** join/#htc-linux Scott_Waye (n=scott@213-152-36-33.dsl.eclipse.net.uk) |
| 20:48.07 | Scott_Waye | Anyone know what the status of the linwizard project on sourceforge is? |
| 20:48.11 | lkcl | awww i have to boot out of linux... |
| 20:48.22 | lkcl | linwizard? is that an htc wizard project? |
| 20:48.26 | Scott_Waye | yes |
| 20:49.01 | cr2 | hmm. there is something broken in this code with hte later wince5 versions ... |
| 20:49.31 | lkcl | y'know - i so don't know why people don't do their research _before_ creating a new project. |
| 20:49.38 | lkcl | thanks Scott |
| 20:49.54 | Scott_Waye | I guess you mean that this work is already handled somewhere else? |
| 20:50.34 | lkcl | well there are people working on it - probably independently so it would appear - who drop in occasionally here. |
| 20:52.39 | Scott_Waye | ok, thanks. |
| 20:53.22 | lkcl | ok. they do know. |
| 20:53.23 | lkcl | http://wiki.xda-developers.com/index.php?pagename=WizardResearch |
| 20:53.29 | lkcl | or it seems so, anyway. |
| 20:53.50 | lkcl | http://www.handhelds.org/moin/moin.cgi/HTCWizard |
| 20:54.55 | lkcl | o god it's an OMAP 850 device. don't bother with it, Scott_Waye. |
| 20:55.01 | Scott_Waye | And theres a readme here |
| 20:55.02 | Scott_Waye | http://omap850.om.funpic.de/downloads/wizard/readme |
| 20:55.05 | lkcl | i bought one. great device... |
| 20:55.35 | Scott_Waye | Whats wrong with it? |
| 20:55.48 | Scott_Waye | apart from underpowered |
| 20:56.00 | lkcl | the TI omap processor and the fact that you have to upload binary-blob firmware to the DSP |
| 20:56.25 | lkcl | and that the GSM radio parameters are memory-mapped into the omap's memory space |
| 20:56.44 | Scott_Waye | meaning its next to impossible to reverse engineer? |
| 20:56.53 | lkcl | nothing is impossible |
| 20:57.06 | lkcl | it's just ... just... |
| 20:57.21 | lkcl | well, i have no idea how it got FCC approval, _especially_ with wince running on it! |
| 20:57.45 | lkcl | the htc apache (i think) aka the orange spv c550 is the same. |
| 20:58.11 | lkcl | basically there's a whole range of devices (underpowered as you say) which use the OMAP series of processors - and they're all dodgy as hell |
| 20:58.39 | Scott_Waye | Oh, so its easy to reconfigure the radio to break the regulations - at least in theory? |
| 20:58.41 | lkcl | they come bottom of my list of projects to support, and i love htc phones |
| 20:58.43 | lkcl | CORRECT! |
| 20:58.44 | lkcl | :) |
| 20:59.03 | lkcl | in theory??? all you have to do is write to the correct memory location and you've changed your phone's IMEA address! |
| 20:59.25 | lkcl | or, how about putting the GSM power-output to 200% the FCC-approved limit?? |
| 20:59.32 | lkcl | and burning your ear off??? |
| 20:59.50 | lkcl | or setting fire to the device?? |
| 20:59.59 | lkcl | really _great_ design |
| 21:00.08 | Scott_Waye | Cheap I suppose |
| 21:00.16 | lkcl | and this is in wince-unprotected memory so a virus could be written to blow up your phone |
| 21:00.26 | cr2 | lkcl: not apache, tornado. |
| 21:00.35 | lkcl | ahh yes that's the one. got one of them, too :) |
| 21:01.25 | cr2 | http://vivien.chappelier.free.fr/typhoon/index.html |
| 21:03.27 | cr2 | lkcl: do you have windows ? :) |
| 21:05.23 | lkcl | cr2: nope! |
| 21:05.32 | lkcl | yes i know about that one, cr2 |
| 21:05.47 | lkcl | well, i have a couple of disks kicking around and i've even tried reactos |
| 21:07.18 | lkcl | Scott_Waye: you might want to talk to vivien to see how he's getting on (pretty good it seems) cos they're OMAP as well. 730 in this case |
| 21:08.15 | Scott_Waye | ok, I'll do that. |
| 21:20.11 | *** part/#htc-linux Scott_Waye (n=scott@213-152-36-33.dsl.eclipse.net.uk) |
| 21:32.43 | lkcl | what command again cr2? |
| 21:35.29 | *** join/#htc-linux goxboxlive (n=goxboxli@9.80-202-160.nextgentel.com) |
| 21:36.19 | cr2 | lswin |
| 21:38.48 | goxboxlive | lkcl: Do you have a working phone now? |
| 21:40.24 | cr2 | goxboxlive: tell me about the phone/sound again. |
| 21:40.32 | goxboxlive | cxr2: ok |
| 21:40.56 | goxboxlive | What do you want to know? |
| 21:41.02 | goxboxlive | There are 4 speakers. |
| 21:41.35 | goxboxlive | two on lcd (both sides) and two on the keyboard unit (stereo) |
| 21:43.51 | goxboxlive | The one at the on the outside (when clam is closed) is ringing when reciveing a call, and it work as a lure when when the clam is closed (( you know what i mean with closed "lcd down") |
| 21:44.59 | goxboxlive | When i am in a call and open the lcd and flip it to portrait mode the speaker on front stops working and the one over the lcd takes over. |
| 21:45.38 | goxboxlive | So, i guess, the two speakrs i have told you about now is workimg with the phone part of the phone. |
| 21:46.36 | goxboxlive | The others on the keyboard part (stereo) is the ones that is a part of the PDA ( like video, music warningsan so on. |
| 21:46.46 | goxboxlive | cr2: Was this clear enough? |
| 21:49.48 | goxboxlive | When i used linux and made a call, i could hear the voice on the universal in the speaker above the lcd, no mather if it where closed or opened. |
| 21:51.22 | cr2 | the lcd speakers are mono, and are switched depending on the clamshell rotation. |
| 21:51.57 | goxboxlive | cr2 that's right |
| 21:52.15 | cr2 | switched probably in hardware. |
| 21:52.39 | cr2 | there is an option in the phone app, where you can switch it to "loud" ? |
| 21:52.57 | goxboxlive | I dont think it has something to do with the rotation, it seems like it has a open lid button or something |
| 21:53.29 | goxboxlive | yes, there is ( You mean handsfree right ) |
| 21:54.19 | cr2 | i can see a button "sound off" |
| 21:54.31 | goxboxlive | ok |
| 21:54.50 | goxboxlive | hardware or gpio? |
| 21:57.20 | cr2 | with handsfree the sound goes to the stereo speakers ? |
| 21:58.03 | cr2 | "handsfree" "sound off" buttons on the screen. |
| 21:59.12 | goxboxlive | Yes i tried that out now. When sppeakers on during a phone call (handsfree) the sound goes to the stereo speakers |
| 22:00.16 | cr2 | and if i use the headphone ? |
| 22:00.28 | goxboxlive | And, when the phoine is reciving a call (phone rings) the ringing is also in the stereo speakers |
| 22:01.05 | goxboxlive | No mather if the clamshell is closed or open |
| 22:01.18 | cr2 | there is one gpio with unclear function related to the headphone. |
| 22:01.22 | goxboxlive | Headphone i havnet tried in wince |
| 22:01.47 | cr2 | where does it ring with the bt headset ? |
| 22:02.52 | goxboxlive | I dont have BT handsfree now, but as far as i remember it rang as normal, but it also rang in the BT headset |
| 22:05.33 | goxboxlive | Mybe the unknown gpio is the "voice dial" function usning the handsfree headset. |
| 22:06.10 | cr2 | i don't remember right now, but when you call another phone with bt headset, the ring is in the bt headset. |
| 22:06.37 | cr2 | but it takes some time, before the sound is switched in the bt headset. |
| 22:07.00 | goxboxlive | Yes the ring is in the BT headset |
| 22:07.24 | goxboxlive | i havent noticed that |
| 22:07.40 | goxboxlive | there is a delay in sound |
| 22:09.59 | cr2 | yes. |
| 22:11.05 | cr2 | EyeOfTheStorm has a stereo bt headset, i have asked him if he can play music to it. |
| 22:11.34 | cr2 | if the answer will be yes, then it is necessary to dump the bt datastream. |
| 22:13.10 | goxboxlive | Yes, that works cr2. It has been made an new AKU3.2 something. And it supports BT stereo handset. |
| 22:13.35 | goxboxlive | But that is kind of new stuff, i guess it is only included in beta roms and unofficial roms |
| 22:14.12 | cr2 | i want to know if the data goes through the btuart. |
| 22:14.54 | cr2 | otherwise, there is a PCM data path to the headset. |
| 22:15.28 | lkcl | ok dump on its way. over 300 memory pages |
| 22:15.36 | *** join/#htc-linux florian (n=fuchs@84-245-181-178.bln.bpool.celox.de) |
| 22:15.42 | lkcl | SPLAT when it got down to 5 pages |
| 22:16.07 | cr2 | the question is: where (and when) to switch what on/off. |
| 22:16.08 | lkcl | goboxlive: i have been able to talk to the gsm module for a couple of data exchanges. |
| 22:17.04 | cr2 | lkcl: mine crashes at page=4. something is broken there,but i don't understand what. |
| 22:17.10 | lkcl | goboxlive, was it you who got initrd working with 2.6.18-hh0? |
| 22:17.31 | lkcl | sounds like an off-by-one-word bug to me |
| 22:17.46 | cr2 | if i just memcpy() the data, then there is no crash. so the crash is because of the WriteFile() function. |
| 22:17.46 | lkcl | but what do i know? :) |
| 22:18.01 | lkcl | ah |
| 22:18.02 | lkcl | oops. |
| 22:18.33 | lkcl | ok how about copying into a bigger memory buffer or just into a buffer and _then_ using WriteFile? |
| 22:18.38 | cr2 | i have even called LockPages on this virtual memory, and it succeeded. |
| 22:18.45 | goxboxlive | lkcl: Not initrd but iniramfs, i compiled a kernel with a etx2 bootstrap image, but ithad restricted functions |
| 22:18.57 | lkcl | ah right |
| 22:19.51 | cr2 | lkcl: hmm. maybe to call allocphysmem() ... the binaries are all <4MB, so it should work. |
| 22:20.19 | cr2 | another option is just to hexdump then to stdout. |
| 22:21.01 | cr2 | i think the second option is better. |
| 22:21.16 | cr2 | fot the first one i still need to call fwrite(). |
| 22:29.16 | lkcl | hmmm.... 02 0c 00 81 00 41 54 45 30 3b 56 30 0d bc 9d - the last two bytes are looking suspiciously like a crc16 of some kind |
| 22:29.38 | florian | re |
| 22:48.30 | cr2 | lkcl: calculate crc16. gather all such strings to cover the full linar space :) |
| 22:48.42 | cr2 | s/linar/linear/ |
| 22:48.49 | cr2 | if it's not crc16 |
| 22:50.20 | lkcl | nk.exe: GetFrame() - CRC Error |
| 22:50.54 | lkcl | ah ha! |
| 22:51.01 | lkcl | ok so it's crc |
| 22:51.05 | lkcl | of some kind. |
| 22:51.56 | *** join/#htc-linux FossiFoo (n=Fossi@e176100247.adsl.alicedsl.de) |
| 22:52.36 | cr2 | crc16 is used by mmc afaik, so it must be in the kernel. |
| 22:52.44 | cr2 | and crc8 |
| 22:53.10 | lkcl | ah joy. |
| 22:59.27 | lkcl | hm of course not a good idea to include _their_ crc value in the calculations... |
| 22:59.49 | cr2 | :) |
| 22:59.59 | cr2 | hexdumping... |
| 23:02.00 | lkcl | hmmm... |
| 23:03.40 | cr2 | crashed. |
| 23:03.56 | lkcl | oops |
| 23:05.38 | cr2 | with zeros at the end. the same problem like with your file. |
| 23:09.18 | lkcl | hmm |
| 23:09.33 | cr2 | lib/crc16.c ? |
| 23:09.41 | lkcl | yeh saw it |
| 23:11.01 | cr2 | does it work ? |
| 23:11.42 | lkcl | ok it contains the right 'code' - 0xd201 so there's a chance it's the right one |
| 23:11.49 | lkcl | for a 2-byte bit of data... |
| 23:22.07 | lkcl | ARGH! :) |
| 23:24.28 | cr2 | does not match ? |
| 23:25.27 | cr2 | lkcl: we need somebody with windows, who can properly extract nk.exe and rilgsm.dll |
| 23:32.24 | cr2 | or to write the perl bindings for the lzx decompressor ;) |
| 23:38.38 | FossiFoo | nk.exe from your phone? |
| 23:38.53 | FossiFoo | i mean phonetype |
| 23:39.02 | FossiFoo | or is it a general mw5 thing? |
| 23:39.25 | *** join/#htc-linux florian (n=fuchs@84-245-181-178.bln.bpool.celox.de) |
| 23:40.09 | lkcl | no it dont damn match |
| 23:49.15 | cr2 | FossiFoo: yes. |
| 23:50.37 | lkcl | hmh. |
| 23:50.40 | *** join/#htc-linux rejon (n=rejon@64-121-195-22.c3-0.sfpo-ubr4.sfrn-sfpo.ca.cable.rcn.com) |
| 23:50.48 | lkcl | you reckon there's enough in that decoded nk.exe, cr2? |
| 23:50.52 | lkcl | even with the last 4 pages missing? |
| 23:59.14 | cr2 | it's not only 4 pages, there are a lot of zeros at the end. |