00:04.22 | awelux | I should upload what I have soon |
00:04.51 | awelux | It might help you with debuging the 2.6.15 kernel |
00:12.04 | cr2 | that will be nice. |
00:15.38 | awelux | do we diable both, normal and fast irq's ? |
00:15.47 | awelux | *disable |
00:22.39 | cr2 | all. |
07:49.48 | *** join/#htc-blueangel rob_w|mis (n=rob_w@p549BBB82.dip0.t-ipconnect.de) |
08:23.01 | *** join/#htc-blueangel awelux_ (n=awelux__@dslb-084-058-134-223.pools.arcor-ip.net) |
09:58.18 | *** join/#htc-blueangel dullard (n=jim@adsl-static-1-30.uklinux.net) |
10:47.46 | *** join/#htc-blueangel hlbot (n=adm@82.67.76.179) |
11:01.05 | *** join/#htc-blueangel hlbot (n=adm@82.67.76.179) |
11:41.15 | *** join/#htc-blueangel hlbot (n=adm@iclem.net) |
12:22.14 | *** join/#htc-blueangel geers_ (n=DevzMan@c51008F31.inet.catch.no) |
16:36.35 | *** join/#htc-blueangel cr2 (n=konversa@crpl22.physik.uni-wuppertal.de) |
16:53.14 | *** part/#htc-blueangel rob_w|mis (n=rob_w@p549BBB82.dip0.t-ipconnect.de) |
20:26.24 | awelux_ | cr2, the kernel is linked at address 0xC0008000 |
20:26.32 | awelux_ | and not 0xa0008000 |
20:26.43 | awelux_ | which seams to be the main reasin it fails |
20:26.46 | awelux_ | *reason |
20:27.11 | awelux_ | do you have an idea why the kernel is linked at the address 0xc0008000? |
20:27.40 | awelux_ | It's slightly possible that this is intentional because the MMU is switchted on at that point, |
20:28.08 | awelux_ | But I think they have a 1:1 mapping for the start of the kernel |
20:28.49 | cr2 | hi awelux |
20:29.14 | cr2 | the 0xC0008000 is the address where all linux kernels are linked. |
20:29.21 | cr2 | it is virtual. |
20:29.49 | cr2 | also on x86. |
20:29.58 | cr2 | is it 3GB ? |
20:32.29 | awelux_ | ok. then I've to double check the mmu table that is build |
20:33.16 | awelux_ | shouldn't we pass the 0xc0008000 address somehow to the kernel ? |
20:33.43 | cr2 | no. |
20:34.01 | cr2 | the kernel will setup it itself. |
20:34.17 | cr2 | can you do a |
20:34.25 | cr2 | pd 0xa0032000 100 |
20:34.36 | awelux_ | ok |
20:34.47 | cr2 | 2 times |
20:34.51 | awelux_ | ok ;) |
20:34.54 | cr2 | and look at the 0x80 offset |
20:39.37 | cr2 | pd 0xa0032000 0x100 |
20:39.39 | cr2 | of course |
20:40.00 | awelux_ | some connection problems here with haret. will take a while to reinitialize |
20:57.10 | awelux_ | HaRET(2)# pd 0xa0032000 0x100 |
20:57.10 | awelux_ | a0032000 | a0038210 a0038220 a0038230 a0038240 | .... ...0...@... |
20:57.11 | awelux_ | a0032010 | a0038250 a0038260 a0038270 a0038280 | P...`...p....... |
20:57.11 | awelux_ | a |
20:57.13 | awelux_ | ... |
20:57.24 | awelux_ | a0032060 | a0038390 a00383a0 a00383b0 a00383c0 | ................ |
20:57.24 | awelux_ | a0032070 | a00383d0 a00383e0 a00383f0 a0038400 | ................ |
20:57.24 | awelux_ | a0032080 | 00003b70 00000000 00000000 00000000 | p;.............. |
20:57.25 | awelux_ | a0032090 | 00000000 00000000 00000000 00000000 | ................ |
20:57.25 | awelux_ | a |
20:57.26 | awelux_ | ... |
20:57.49 | awelux_ | a0032070 | a00383d0 a00383e0 a00383f0 a0038400 | ................ |
20:57.49 | awelux_ | a0032080 | 00002630 00000000 00000000 00000000 | 0&.............. |
20:57.49 | awelux_ | a0032090 | 00000000 00000000 00000000 00000000 | ................ |
20:57.59 | awelux_ | so just a0032080 changes |
20:58.32 | awelux_ | switched on flight mode |
20:58.37 | awelux_ | nmno difference |
20:59.27 | cr2 | ok. that means your usb host is stopped. |
21:00.05 | cr2 | this 0x80 counter corrupts memory on the universal is the radio is off. |
21:02.34 | awelux_ | the usb host shouldn'T be off |
21:02.54 | awelux_ | I mean in flight mode the counter is still running. |
21:10.15 | cr2 | yes. |
21:43.23 | awelux_ | I think I've trouble to access the vram after the mmu was switched on |
21:43.51 | awelux_ | what happens if I access an unmapped address? |
21:44.03 | awelux_ | Do I get the physical address or an exception? |
21:46.46 | cr2 | an exception |
21:46.57 | awelux_ | ok. |
21:47.04 | awelux_ | that explains a lot i think |
21:47.07 | cr2 | you need to map the fb. |
21:47.19 | cr2 | add the pagetable entries. |
21:47.37 | awelux_ | easier said then done. But I will try. |
21:47.38 | cr2 | there was some VIDEO macro in the kernel, but imho, never used. |
21:48.04 | cr2 | are you crazy ? :) |
21:48.49 | awelux_ | lol. Did you mean that as a joke earlier? |
21:59.52 | cr2 | it can be done, but is a really crazy stuff. |
22:00.06 | cr2 | btw, i have working sound on the universal :) |
22:02.10 | cr2 | bluetooth is not working. probably i still need to pull the reset asic3 gpio. |
22:17.10 | awelux_ | I've seen some code special for debugging over serial |
22:17.16 | *** join/#htc-blueangel JuNuKN (n=JuNuKN@Mfb35.m.pppool.de) |
22:18.41 | awelux_ | hi JuNuKN |
22:18.53 | JuNuKN | Hello awelux_ |
22:19.04 | cr2 | use IR instead. |
22:19.12 | cr2 | hi JuNuKN |
22:19.23 | JuNuKN | Hello cr2 .... gratulation |
22:20.18 | cr2 | thanks. but the touchscreen is very annoying, making everything unusable... |
22:20.58 | JuNuKN | Yes I saw it ! But I am shure you will find out the trick! |
22:21.13 | cr2 | eventually. |
22:21.18 | JuNuKN | But sound!!! on universal ... ! |
22:21.45 | awelux_ | is there a chance that with a deifferent setup of the ts chip everything will start to be linear ? |
22:22.01 | awelux_ | yep congrat cr2 |
22:22.08 | awelux_ | *congrats |
22:22.22 | cr2 | it is actually a more simple sound setup than ba/hima/magician. |
22:22.44 | cr2 | i don't even know which ts chip is used. |
22:23.01 | cr2 | you can test this driver, because magician has this chip. |
22:23.15 | JuNuKN | maybe some combination between the intensity of stylus pressure which will define some kind of multiplier or what ever. |
22:23.58 | cr2 | it is probably easier to disassemble the wince dll and look for the setup values. |
22:24.21 | JuNuKN | 192.168.0.55192.168.0.55 |
22:24.36 | JuNuKN | ... oh wrong keyboard ... sorry |
22:25.06 | cr2 | i will solve the nonlinearity problem. there is also another one with some random jumps. |
22:25.18 | cr2 | which are not really random. |
22:25.28 | cr2 | so averaging does not help very much. |
22:30.56 | JuNuKN | Maybe some to slow clock speed, which does not allow to messure the coordinates correct for the hardware ? |
22:33.16 | cr2 | don't know. i will wait a bit. there are two other drivers for the compatible hardware. |
22:33.44 | cr2 | the ultimate solution is of course to look what the wince is doing. |
22:35.35 | cr2 | you can test it on the magician anyway, a i have said. |
22:36.55 | cr2 | awelux, why don't you want to use the IR port for debugging ? |
22:37.15 | JuNuKN | Interesting! I will test also later, if magician has a state, where I can test with it. |
22:38.04 | cr2 | once the kernel will boot, the framebuffer and touchscreen are the same as on the universal. |
22:38.15 | cr2 | sound is the same as on blueangel. |
22:38.27 | cr2 | no asic3 - no pain. |
22:38.37 | cr2 | everything is on pxa272. |
22:38.52 | cr2 | a dream machine :) for porting.. |
22:39.35 | JuNuKN | :) but to less ram for a liunux machine!? Or not? |
22:40.02 | cr2 | the same as on the unuversal. with smaller framebuffer, so it will be faster. |
22:41.53 | JuNuKN | I have only arround 20MB free memory of my 64MB (under CE). That is the biggest disadvantage against himalaya or blueangel. And the screen was better on Hima and on BA. Especially in sunlight! |
22:43.25 | cr2 | but the dpi. |
22:43.42 | JuNuKN | Nevertheless, the form factor is great! It is still wearable ;o) |
22:43.42 | cr2 | and you have a smaller battery now. |
22:44.05 | cr2 | yeah. i prefer a780 as a real phone. |
22:44.06 | JuNuKN | The battery holds longer as on hima! |
22:44.26 | cr2 | it's obvious. |
22:44.34 | cr2 | less RAM, smaller screen. |
22:44.48 | cr2 | less hardware |
22:44.52 | cr2 | no 2D accel. |
22:44.55 | JuNuKN | But there are optional bigger batteries for Hima and Blueangel ... Maybe later again an BA |
22:45.19 | cr2 | ba has a bigger battery |
22:45.35 | cr2 | which does not fit into the cradle :) |
22:45.43 | cr2 | and ba is very fast |
22:45.47 | JuNuKN | The screen has same resolution :o) and I don't want to watch movies on it ;o) |
22:45.49 | cr2 | compared to the universal. |
22:46.02 | cr2 | ba is the best for movies. |
22:46.20 | cr2 | because of the hardware IDCT. |
22:47.09 | cr2 | as i have said, i wish they have put the universal LCD into ba. a dream machine. |
22:48.55 | cr2 | it is a pity that nobody is doing ba development now. |
22:48.59 | JuNuKN | I would prefer the hima .... the hardware seems more robust to me against BA. Especially the keyboard I would dislike on BA, because I would evertime expect, that it got contact problems when pulling out and in every day. But one big advantage on BA is Wifi .... which is yet not usable under linux ... why I bought a hima last time. |
22:49.50 | cr2 | no contact problems on ba. |
22:50.13 | cr2 | it is only the LCD that moves. |
22:50.34 | cr2 | and it will loosen over time. |
22:51.00 | JuNuKN | But for the moment I like my magician. Do you think it will work over years? I read a lot on ebay, from people, which send there ba to repair because of keyboard problems. |
22:51.57 | JuNuKN | And the "flat-band-cable", wouldn't it break with the time? |
22:52.18 | cr2 | in principle, it is a solid design. spi keyboard. |
22:52.35 | cr2 | the flat LCD cable ? |
22:52.51 | JuNuKN | Yes, that was the thing I ment. |
22:53.01 | JuNuKN | Or the cable to the keyboard |
22:53.11 | cr2 | don't think so. it is very long, so the mechanical stress is minimal. |
22:53.33 | cr2 | the keyboard cable is fixed and does not move. |
22:54.19 | cr2 | it's like the flatband on the scanner. |
22:54.30 | JuNuKN | Maybe I will buy such a device next time again, when my display on magician breaks. It would not be the first display, which break on my devices :o) |
22:54.31 | cr2 | also very long one. |
22:54.48 | cr2 | heh. |
22:55.02 | cr2 | that's why i like a780 and universal. |
22:55.04 | JuNuKN | First on PALM, Next on MDA1 |
22:57.04 | JuNuKN | But now I will stop to interrupt you and will see, what comes to run this night ;o) |
22:57.58 | cr2 | :) i am tired for today. |
22:58.12 | cr2 | and i need to write the asic3 dumper some time. |
22:58.35 | cr2 | and i need a commandline tool to talk to the htc modem. |
22:58.48 | cr2 | and the stupid openembedded does not work. |
22:59.05 | cr2 | and why did they included 'tslib' in the ba ??? |
22:59.19 | cr2 | tsdev, not tslib. |
22:59.37 | cr2 | an unnecessary kernel module. |
23:02.17 | cr2 | the channel header here says "org.openembedded.oz354fam083", but there is no BA config there!!! |
23:02.52 | cr2 | only in org.openembedded.dev and dreambox, which do not want to compile at all. |
23:04.30 | awelux_ | strange world ;) |
23:05.24 | awelux_ | cr2, I don't want to use the ir because I don't have an ir port on my pc |
23:05.34 | awelux_ | Only on the notebook |
23:05.48 | awelux_ | but that's so noisy that I don't like it |
23:06.10 | cr2 | ok. |
23:07.40 | cr2 | i can think about only one reason it does not boot. memory corruption by some device. |
23:08.04 | JuNuKN | awelux_ : I do my sync meanwhile only via IR or cable, because BT will pair again and again :o) |
23:09.31 | cr2 | i have found the LCD contrast control on the universal. |
23:09.55 | JuNuKN | cr2 ... or one interrupt vector for a interrupt method, which is not disabled and not initialized? |
23:11.00 | cr2 | the kernel requirements to the bootloader are clear: stop the dma, stop the interrupts, switch mmu off. that's all. |
23:11.16 | cr2 | the kernel will take care of the rest. |
23:12.00 | cr2 | haret should power down and switch off all the devices. |
23:12.07 | cr2 | including LCD actually. |
23:12.30 | cr2 | it is difficult to debug |
23:12.59 | cr2 | that's why the serial port is the best option. |
23:13.09 | cr2 | or the IR in out case. |
23:13.48 | awelux_ | I dug out an old IR reciever for the ir remote control project |
23:13.54 | awelux_ | was it itda? |
23:14.00 | awelux_ | *irda ? |
23:14.00 | cr2 | since you don't know the purpose of all the gpios, it can be difficult to power up everything again. |
23:14.07 | JuNuKN | hopefully nothing is forgotten!? Better to initialize whole memory with nops and a final jump to catch possible illegal method calls ...?, |
23:14.24 | cr2 | that's why i am writing the asic3 dumper for the universal. |
23:14.50 | cr2 | SIR at 9600, if you don't change anything. |
23:14.58 | cr2 | that's how wince works. |
23:15.13 | awelux_ | ah lirc |
23:15.24 | awelux_ | I've an reciever for lirc |
23:18.16 | awelux_ | but it doiesn't work |
23:19.07 | cr2 | hmm. |
23:19.40 | cr2 | can you compile the kernel with the simple fb ? |
23:20.11 | cr2 | we can then modify haret to switch everything off. |
23:20.38 | cr2 | btw, what is the fb memory dma descriptor ? |
23:20.54 | cr2 | pd 0xSRAM 0x100 |
23:22.21 | cr2 | oh. #error Machine byte order unknown. |
23:22.25 | cr2 | in glibc. |
23:22.32 | cr2 | again. |
23:23.04 | cr2 | fcking OE. |
23:24.12 | cr2 | it was working for OZ. hmm. but it uses different gcc. |
23:25.22 | awelux_ | what is OE ? |
23:25.25 | awelux_ | and OZ? |
23:26.10 | JuNuKN | open embedded ... some kind of gentoo ;o) and Open Zaurus ... |
23:29.11 | awelux_ | ah. thanks |
23:29.46 | awelux_ | HaRET(1)# pd 0xSRAM 0x100 |
23:29.47 | awelux_ | Error: line 1: Expected a number, got 0xSRAM |
23:29.47 | awelux_ | Error: line 1: Expected <vaddr> <size> |
23:30.21 | JuNuKN | open embedded will manage to download all sourcecodes, compile a cross compilation chain, kernel and everything ... If you have prepared a script for it. |
23:33.07 | cr2 | the SRAM address is needed :) |
23:33.18 | awelux_ | HaRET(4)# pd 0xa0000000 100 |
23:33.19 | awelux_ | a0000000 | 80080bb4 00003a7f a0660000 00000001 | .....:....f..... |
23:33.19 | awelux_ | a0000010 | ffffc764 6000001f 4000003f 5ffff9cf | d......`?..@..._ |
23:33.19 | awelux_ | a0000020 | feffbbef 7f77fffd 6fff7ffd f7577ff7 | ......w....o..W. |
23:33.21 | awelux_ | like that ? |
23:33.39 | cr2 | 0x5c000000 |
23:33.45 | awelux_ | oh |
23:33.58 | cr2 | 0xa0000000 is DRAM. |
23:34.35 | cr2 | JuNuKN: OE is a complete mess now. i get different error messages on different computers |
23:34.36 | awelux_ | sig. I think I will never understand this asm |
23:34.55 | cr2 | and the error messages change in the last days. |
23:35.00 | awelux_ | seamingly harmless code causes everything to fail |
23:35.05 | cr2 | they are making some massive chanhes there. |
23:35.57 | JuNuKN | Yes I have had same problems as I tested it. And at the end my machine crashes the filesystem ... but that seems more a hardware problem on my desktop. |
23:37.09 | JuNuKN | You should use a special tagged version to start with OE. |
23:38.03 | cr2 | does not work also. |
23:38.07 | cr2 | i have tried. |
23:38.40 | JuNuKN | Wasn't it immolo, which prepared OE for BA!? |
23:38.55 | JuNuKN | Maybe he can be from help here. |
23:41.41 | cr2 | lkcl has written a lot of patches and actually compiled gpe and opie. |
23:41.49 | cr2 | some time ago. |
23:42.03 | cr2 | even on an amd64 cpu. |
23:44.42 | cr2 | good night. |
23:48.14 | JuNuKN | good night cr2 |
23:48.22 | awelux_ | gn8 |
23:49.33 | JuNuKN | awelux_ : what is the actual status of magician port ? |
23:49.55 | awelux_ | the asm seams instable |
23:50.04 | awelux_ | the objdump is ok though |
23:51.25 | JuNuKN | So no running kernel at the moment? |
23:51.36 | awelux_ | guess I've to buy a IR serial port. |
23:51.43 | awelux_ | no. no kernel yet |
23:53.13 | JuNuKN | You do it the first time? |
23:54.03 | awelux_ | yes |
23:54.30 | awelux_ | This arm asm is confusing the heck out of me |
23:54.36 | JuNuKN | Great! I am a linux beginner :o) |
23:54.49 | awelux_ | it's stopping at odd places |
23:55.05 | awelux_ | for no obvious reason |
23:56.58 | JuNuKN | I have to less time to "learn" it .... I have familie and job. I may expect, that you are student? |
23:57.32 | JuNuKN | That where good days! A lot of time for such researches! |
23:57.51 | awelux_ | working |
23:58.02 | awelux_ | but I started with linux while I was still a student |
23:58.07 | JuNuKN | That time windows programming was hip, because win3.0 came out. |
23:58.34 | awelux_ | Started in the 5th semester with linux. |
23:58.56 | JuNuKN | I also started with linux, but only on user level and a view small escapes in details ... |
23:59.08 | awelux_ | and at the end of the 8th I decided that I should spend more time with my studies to finish at all. |
23:59.31 | JuNuKN | I started to use linux 2 years ago. ... and tested it several times/years before |