| 00:02.59 | cr2 | the directory tree structure changed substantially, and is being fixed now. |
| 00:03.08 | pancake | oh ok :/ |
| 00:07.49 | *** join/#htc-linux g3gg0_ (n=g3gg0@ppp-82-135-0-47.dynamic.mnet-online.de) |
| 05:11.38 | *** join/#htc-linux ndiscreet (n=chatzill@h46088f98.area4.spcsdns.net) |
| 06:04.36 | *** join/#htc-linux RoEn_PC (n=roen@p54a64791.dip.t-dialin.net) |
| 06:15.25 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250) |
| 06:21.37 | *** join/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 07:04.58 | *** join/#htc-linux rmoravcik (n=rmoravci@adsl-d181.84-47-56.t-com.sk) |
| 08:26.08 | *** join/#htc-linux TripleDE1 (n=sergio@226.pool85-50-149.dynamic.orange.es) |
| 08:36.40 | *** join/#htc-linux rob_w (n=bob@p85.212.61.201.tisdip.tiscali.de) |
| 08:42.09 | *** join/#htc-linux psokolovsky (n=psokolov@empty-crossing.volia.net) |
| 08:52.01 | *** join/#htc-linux lilliput (n=Touriste@82-47-146-8.cable.ubr11.brad.blueyonder.co.uk) |
| 09:33.35 | *** join/#htc-linux hugo_nz (n=hugo@132.181.52.15) |
| 09:53.23 | *** join/#htc-linux TripleDES (n=sergio@194.pool85-50-204.dynamic.orange.es) |
| 09:58.20 | *** join/#htc-linux tsdogs (n=twostupi@84.18.144.39) |
| 10:08.49 | *** join/#htc-linux TripleDE1 (n=sergio@91.pool85-50-128.dynamic.orange.es) |
| 10:12.21 | *** join/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 11:07.27 | *** join/#htc-linux g3gg0_ (n=g3gg0@ppp-82-135-0-47.dynamic.mnet-online.de) |
| 11:15.13 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 11:22.15 | *** join/#htc-linux TripleDES (n=sergio@40.pool85-50-158.dynamic.orange.es) |
| 11:26.46 | *** join/#htc-linux JuNuKN (n=chatzill@p549abc63.dip0.t-ipconnect.de) |
| 11:30.02 | *** join/#htc-linux pH5 (n=ph5@e178222049.adsl.alicedsl.de) |
| 12:36.55 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 12:43.32 | *** join/#htc-linux tsdogs (n=twostupi@62.123.180.130) |
| 13:03.47 | *** join/#htc-linux psokolovsky_ (n=psokolov@82.193.98.2) |
| 13:31.57 | *** join/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 13:41.36 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250) |
| 13:57.10 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 14:35.42 | *** part/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 14:58.08 | BabelO | hi |
| 15:04.11 | tsdogs | BabelO: hi |
| 15:04.27 | tsdogs | new build of opieII seems to be running :) |
| 15:04.42 | tsdogs | though it's using 95% of cpu and is very slow |
| 15:04.44 | BabelO | oh oh, what did you change ? |
| 15:05.02 | tsdogs | The base of OE I think |
| 15:05.17 | BabelO | % of cpu ? which process ? |
| 15:05.21 | tsdogs | The dbus and bluetooth was not included |
| 15:05.23 | tsdogs | qpr |
| 15:05.30 | tsdogs | s/spr/qpe/ |
| 15:05.45 | tsdogs | qpe |
| 15:06.02 | tsdogs | I'm rebuilding now with dbus and bluetooth |
| 15:06.06 | BabelO | maybe it is some autodetect process |
| 15:06.41 | tsdogs | I have no idea, but themes seem to do take a lot of processing (?) |
| 15:06.59 | BabelO | yes try crisp or gel theme |
| 15:07.15 | tsdogs | I have a battry issue too. |
| 15:07.37 | tsdogs | It does not seem to be reading battery status, and says that battery is in critacal state |
| 15:08.00 | BabelO | oh.. remove line system("apm -s) in custom.cpp |
| 15:08.18 | tsdogs | ok I'll try. |
| 15:08.26 | BabelO | normaly it read /sys/class/battery/main-battery |
| 15:08.42 | tsdogs | yes and that's there... |
| 15:09.27 | tsdogs | I have to go now, baby time :) |
| 15:10.04 | tsdogs | If you want to try my new version you can find it here: http://84.18.144.44/tsdogs/ |
| 15:11.41 | BabelO | i m waiting you build a bluetooth with alsa version... mine always compile with oss |
| 15:12.26 | tsdogs | I haven't checked about alsa. I'll do it tonight. :) cya later. |
| 15:12.58 | BabelO | bye |
| 15:44.15 | *** join/#htc-linux Sti_0239 (n=Where@did75-14-82-236-19-158.fbx.proxad.net) |
| 15:47.47 | Sti_0239 | hi |
| 15:48.11 | BabelO | hi |
| 16:19.53 | *** join/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 16:23.02 | *** part/#htc-linux Sti_0239 (n=Where@did75-14-82-236-19-158.fbx.proxad.net) |
| 16:23.34 | *** join/#htc-linux Guest5272 (n=test@84.235.5.232) |
| 16:23.42 | Guest5272 | hello |
| 16:57.37 | *** join/#htc-linux jeanseb (n=jeanseb@gazypan.dyndns.org) |
| 17:06.39 | *** join/#htc-linux psokolovsky_ (n=psokolov@82.193.98.2) |
| 17:36.10 | *** join/#htc-linux psokolovsky (n=psokolov@82.193.98.2) |
| 17:42.34 | Kevin2 | Hi |
| 17:43.17 | Kevin2 | ph5: Yes, apache cpld uses half the register to show the status of the input pins, and the other half as the irq trigger bitmap. |
| 17:44.17 | BabelO | hi Kevin2 |
| 17:44.40 | Kevin2 | BTW, it is possible to trace through a suspend/resume in haret. Haret loads a new set of irq handlers which log to an in memory buffer - it runs regardless of the state of haret.exe. When you resume (and wifi reestablishes the telnet session) the buffer will be flushed to your screen. |
| 17:44.56 | Kevin2 | Hi BabelO |
| 17:44.59 | BabelO | where can i get your latest haret ? the one which can monitor ATI ;) |
| 17:45.17 | Kevin2 | http://www.handhelds.org/~koconnor/haret/haret-20070510.exe |
| 17:46.05 | BabelO | thx ;) i ll try it now |
| 17:49.23 | BabelO | Kevin2: another question, on BA and h4000 acx is on pcmcia pxa port. Do you think htc studie memory map to got an memory map image fit in pcmcia port range of pxa, but the real adress mapping is somewhere else ? |
| 17:49.39 | BabelO | so that why we can't trace it ? |
| 17:50.35 | Kevin2 | I'm not sure what you mean - "studie memory map"? |
| 17:50.46 | pH5 | hi Kevin2, thanks. |
| 17:51.22 | pH5 | does htcapache have wakeup irqs on there, too, like power cable or bt irq? |
| 17:52.08 | Kevin2 | There are a few things which we can't trace - writes to different virtual addresses, writes via dma, and potentially writes when the cpu is in "abort" mode. It is always possible that some code could disable tracing during certain regions, but that seems quite unlikely. |
| 17:53.14 | BabelO | ok, thx, i m trying now |
| 17:53.30 | BabelO | looking at log for your command line ;) |
| 17:53.39 | Kevin2 | pH5: Yes. The power plug irq is on cpld. |
| 17:54.20 | Kevin2 | pH5: BTW, pastebin.ca seems to be down, so I couldn't see what you pasted yesterday. |
| 17:54.50 | Kevin2 | BabelO: Your memory addresses will likely be different. You need to run "dump mmu" and then find the memory areas to trace. |
| 17:56.09 | BabelO | Kevin2: i ve my mmu dump here http://linuxbox2.free.fr/blueangel/mmu-blueangel.txt |
| 17:58.16 | pH5 | Kevin2: great. include/linux/soc/htc_cpld.h is in cvs now and example platform data for apache is at http://rafb.net/p/k1y6HY32.html |
| 17:58.44 | pH5 | Kevin2: this is a proposal, I haven't written anything besides the magician_cpld code that is in cvs yet, so improvements are welcome. |
| 17:59.18 | BabelO | Kevin2: hown can i start your new haret ? |
| 17:59.37 | pH5 | I want to combine htcapache-egpio.c and magician_cpld.c into a common htc-cpld driver for at least apache,athena and magician. |
| 18:00.08 | Kevin2 | pH5: Okay - I'll review it. |
| 18:00.23 | Kevin2 | BabelO - which physical address range did you want to watch? |
| 18:00.48 | BabelO | Kevin2 :it is adress 2c000000 |
| 18:01.26 | BabelO | but when i start your haret i see a nice splash screen booting linux ... |
| 18:01.55 | BabelO | something change ? |
| 18:02.57 | Kevin2 | BabelO: Huh? Running haret causes it to try and boot linux? |
| 18:03.34 | BabelO | yes don't understand why :( i ve no default.txt file |
| 18:04.18 | BabelO | ok it is good now, it try to load zImage if it is here ;) |
| 18:06.50 | BabelO | hum interesting new stuff |
| 18:06.59 | BabelO | and i see atidbg too |
| 18:07.23 | Kevin2 | BabelO - why do you think acx is at 0x2c000000 physical? |
| 18:07.59 | BabelO | Kevin2: because it pcmcia port 0 and it is detected by acx module when i set port here |
| 18:10.41 | BabelO | Kevin2: i do that |
| 18:10.44 | BabelO | HaRET(4)# ibit irqs 17 ; addlist mmutrace 0x8c200000 1024*1024 w ; wi 10 |
| 18:10.44 | BabelO | Bit 0 is past max found of 0 |
| 18:11.27 | Kevin2 | BabelO - don't run the "ibit" command - that is specific to hermes. |
| 18:11.35 | BabelO | ok |
| 18:12.02 | BabelO | and it is wrong, it is gpio 17 an irq 51 :) |
| 18:13.25 | Kevin2 | Try running "set trace 0xac000000 ; set tracemask 0xffffff" |
| 18:15.01 | BabelO | nothing, i see only irq 51 |
| 18:15.17 | BabelO | i ll try on ati for test |
| 18:17.51 | Kevin2 | What about: set trace 0x8c000000 ; set tracemask 0xffffff |
| 18:18.12 | BabelO | already tryed, but same only irq 51 |
| 18:19.38 | Kevin2 | BabelO - was wifi running when you ran "dump mmu"? |
| 18:19.58 | BabelO | yes, i m running haret over wifi |
| 18:20.15 | BabelO | maybe it is masked because i use it like this ? |
| 18:21.16 | Kevin2 | Should work - I do the same. |
| 18:22.02 | Kevin2 | Maybe it uses dma to write to acx? |
| 18:22.44 | BabelO | maybe it is a potential solution, i see some pcmcia driver which have DMA range memory parameter |
| 18:22.52 | BabelO | how can i trace DMA access ? |
| 18:23.00 | BabelO | directly to DMA port ? |
| 18:23.06 | BabelO | register i meam |
| 18:24.09 | Kevin2 | BabelO - lookup DMA registers and then trace them. Something like "set trace p2v(0x40000000) ; set tracemask 0xffff" |
| 18:24.36 | BabelO | ok |
| 18:25.00 | Kevin2 | Try running with the original tracing settings (the 0xac... one) and then suspend/resume the device during wi. |
| 18:25.44 | BabelO | hum, but if i do this, wifi doesn't come up, i have to wait for automatic shutdoiwn, if i want wifi resume |
| 18:26.08 | Kevin2 | Huh, why doesn't wifi come back up? |
| 18:26.42 | BabelO | if i use power button to shut off, it didn't come back, it is a ROM cooked from xanadux.... |
| 18:27.28 | Kevin2 | You have this problem when running haret, or always? |
| 18:28.17 | Kevin2 | In either case then, can you turn wifi off/on instead of powering down/up? (If you're afraid it wont come back up - add a "log foo.txt" before wi and then pull up the file afterwards.) |
| 18:28.52 | BabelO | ok, but i need some unlog to flush buffer ? right ? |
| 18:29.16 | BabelO | whitout haret it do the same, i ll try your latest solution |
| 18:29.21 | Kevin2 | Buffer flush should be automatic when wi completes. |
| 18:29.56 | BabelO | ok |
| 18:30.08 | BabelO | trying |
| 18:32.34 | BabelO | ;) if i shutdown wifi during tracing it resume after |
| 18:34.18 | BabelO | no more trace like this, ok i ve to trace with p2v now |
| 18:35.47 | BabelO | no more on DMA register |
| 18:38.46 | Kevin2 | So, you still don't see any traffic? |
| 18:38.46 | Kevin2 | BTW, can you trace other things? |
| 18:38.47 | BabelO | yes i can trace my ds1wm |
| 18:38.47 | BabelO | it is on aic2 |
| 18:38.47 | BabelO | and i see irq 51 |
| 18:38.59 | Kevin2 | You're using my haret, right? |
| 18:39.37 | BabelO | right, i m sure, because i didn't see wasic3gpio command |
| 18:40.27 | *** part/#htc-linux scorpio16v (n=scorpio1@ACB36DAE.ipt.aol.com) |
| 18:40.38 | Kevin2 | I guess you could try "addlist mmutrace 0xac000000 1024*1024 rw" |
| 18:41.08 | Kevin2 | You should cover every range from 0xac000000 to 0xacf00000. |
| 18:43.35 | Kevin2 | It is always possible that the acx driver is mapping in the registers and then unmapping when done - but that seems unlikely. |
| 18:50.15 | BabelO | i see nothing... i try to trace with your haret something like ds1wm and it is working ... so there is something wrong with acx |
| 19:51.15 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250) |
| 20:17.22 | cr2 | hi |
| 20:18.05 | cr2 | BabelO: it seems that we need to find the code location for writing to the acx100 registers. |
| 20:23.02 | cr2 | Kevin2: it seems i've found an i2c bitbang driver on 2 GPC gpio pins. |
| 20:23.31 | BabelO | hi cr2 |
| 20:24.24 | BabelO | my atempt running acx under linux show me that rx buffer is full after a moment |
| 20:25.29 | cr2 | ok, but what should the driver do at such event ? |
| 20:25.52 | BabelO | that s why i m thinking about an acx register mapping to pcmcia that are not tx control |
| 20:26.22 | BabelO | it is like it did not receive correct irq |
| 20:28.26 | BabelO | maybe we need to notiify something else about irq |
| 20:29.18 | BabelO | kevin2 was thinking about dma, but if i look at dma register nothing appear |
| 20:29.55 | BabelO | maybe it is a dma memory copy |
| 20:31.03 | cr2 | there are no DMA registers for pcmcia. at least i don't see them in the manual. |
| 20:32.46 | BabelO | ok, i say that because some wireless pcmcia driver have dma settings |
| 20:48.28 | cr2 | do you see any DREQ changes ? |
| 20:48.50 | cr2 | i doubt that acx uses DMA. |
| 20:49.38 | BabelO | so it has a strange memory mapping |
| 20:53.45 | cr2 | there are 2 files. |
| 20:53.55 | cr2 | tiacxwln.dll and pcmcia.dll |
| 20:54.24 | cr2 | the last one includes power/reset gpios. |
| 20:54.41 | cr2 | we were able to locate them. |
| 20:55.14 | cr2 | tiacxwln is much more obfuscated. |
| 20:56.03 | Kevin2 | Hi cr2 |
| 20:56.27 | cr2 | the only direct references are AEA and AE9 |
| 20:56.34 | cr2 | hi Kevin2 |
| 20:57.03 | cr2 | pancake made some changes to HermesGPIO wiki page, but i think some of them are wrong. |
| 20:59.31 | cr2 | AEA and AE9 are RTC and OStimer |
| 20:59.39 | Kevin2 | GPC12/GPC13 are bit-banging? Any idea what type of info is sent through it? |
| 21:00.29 | cr2 | at least LCD setup. |
| 21:00.42 | cr2 | but imho he has shifted them. |
| 21:01.06 | Kevin2 | cr2: Hrmm. LCD power one? That sounds like Apache. |
| 21:01.09 | cr2 | it should be 11 and 12. |
| 21:01.11 | Kevin2 | s/one/on/ |
| 21:01.44 | cr2 | maybe not the power, but some internal LCD config. |
| 21:02.02 | cr2 | BA boardid6 has something similar. |
| 21:02.52 | cr2 | on hermes it is some sync serial 2line protocol. |
| 21:03.02 | cr2 | with addr/data |
| 21:03.11 | cr2 | looks like i2c to me. |
| 21:03.57 | Kevin2 | Apache controls lcd power, front led power, and can read front keypad using an i2c like protocol that it bitbangs on two gpio ports. |
| 21:04.29 | cr2 | ok. |
| 21:04.43 | cr2 | maybe it's the same. |
| 21:05.25 | Kevin2 | The apache code is in htcapache-mc.c |
| 21:05.29 | cr2 | the acx100 on hermes is using SDIO port on s3c24xx. |
| 21:05.32 | cr2 | ok. |
| 21:06.05 | cr2 | the microSD must use ATI port. that will be tough. |
| 21:06.26 | cr2 | maybe we'll debug NAND instead. |
| 21:07.27 | cr2 | Kevin2: is it possible to set a haret breakpoint in tiacxwln.dll ? |
| 21:12.19 | Kevin2 | On PXA or Hermes? It should be straight forward on PXA, but there is no breakpoint support on s3c processors. |
| 21:13.39 | cr2 | on PXA |
| 21:14.25 | cr2 | BabelO: use "dump modules" to see where tiacxwln.dll is mapped. |
| 21:14.40 | Kevin2 | set insn 0xaddr ; wi 10 |
| 21:15.51 | BabelO | HaRET(1)# dump modules |
| 21:15.52 | BabelO | line 1: No dumper modules available, see HELP DUMP for a list |
| 21:16.53 | cr2 | lsmod ? |
| 21:16.59 | Kevin2 | BabelO: it is "lsmod" on mainline haret. |
| 21:17.03 | cr2 | i forgot the name |
| 21:17.04 | cr2 | ok. |
| 21:17.11 | BabelO | better thx |
| 21:17.17 | Kevin2 | I'll be back later. |
| 21:18.44 | BabelO | 1076 fl=00000000 mid=955c9a94 pid=00000000 gusg= 1 pusg=00000004 base=00DD0000 size=00056000 hmod=955C9A94 mod=tiacxwln.dll exe= |
| 21:20.38 | BabelO | so it is "set insn 0x955C9A94 ; wi 10" ? |
| 21:20.53 | cr2 | ok, we know hmod, but it will be painful to decode the struct manually. |
| 21:21.02 | cr2 | no, there is an offset. |
| 21:21.15 | cr2 | this is the address of the "MODULE" structure. |
| 21:21.22 | BabelO | ok |
| 21:21.29 | cr2 | hm. looking for a link. |
| 21:21.47 | cr2 | so you can decode some core members of the struct. |
| 21:22.52 | BabelO | how ? using pv .... |
| 21:22.52 | cr2 | http://mamaich.uni.cc/wm_re |
| 21:22.52 | cr2 | get TestWM5-src.rar |
| 21:23.07 | cr2 | there are some bugs in the headers, but the MODULE structure is ok afaik. |
| 21:24.48 | cr2 | here is the mdoc g3 related file http://nah6.com/~itsme/cvs-xdadevtools/itsutils/dll/CEIOCTL.H |
| 21:26.17 | cr2 | found, |
| 21:26.19 | cr2 | http://nah6.com/~itsme/cvs-xdadevtools/itsutils/old/cenk.h |
| 21:26.56 | cr2 | typedef struct Module { |
| 21:26.56 | cr2 | LPVOID lpSelf;/* Self pointer for validation */ |
| 21:27.19 | cr2 | BasePtr |
| 21:27.26 | cr2 | look for it. |
| 21:28.23 | cr2 | and 'startip' |
| 21:29.35 | BabelO | ok, if i undrstand i have to download dll to computer look into it for this ? |
| 21:30.17 | cr2 | no. |
| 21:30.49 | cr2 | pastebin the 'pd 0xb55c9a94 0x100' |
| 21:31.22 | cr2 | this is the struct Module * |
| 21:32.01 | cr2 | or (better) look where 0x955c9a94 is mapped in RAM. |
| 21:32.07 | BabelO | it shutdown haret :( |
| 21:32.11 | cr2 | it's the virtual address. |
| 21:32.16 | cr2 | yes, soory :) |
| 21:32.49 | cr2 | dump mmu. |
| 21:33.11 | BabelO | http://linuxbox2.free.fr/blueangel/mmu-blueangel.txt |
| 21:33.56 | cr2 | 95500000 | a5500000 | CB | 1MB section |
| 21:34.07 | BabelO | yes i found it too |
| 21:34.32 | cr2 | pd 0xa55c9a94 0x200 |
| 21:34.49 | BabelO | better ;) |
| 21:34.59 | cr2 | it should have beed 'vd 0x955*' ;) |
| 21:35.43 | BabelO | :( shutdown haret vd |
| 21:35.56 | cr2 | vd 0xb99* |
| 21:36.14 | cr2 | ok. but 'pd 0xa55c9a94' should work. |
| 21:36.50 | BabelO | http://pastebin.ca/485123 |
| 21:37.18 | cr2 | 955c9a94 this is the lpSelf. good. |
| 21:38.16 | cr2 | 951da074 |
| 21:38.35 | cr2 | this should be the pointer to the name == tiacxwln.dll |
| 21:38.53 | cr2 | inuse=4 |
| 21:39.53 | cr2 | then the array of refcounts. |
| 21:40.33 | BabelO | ;) good it is my first lesson |
| 21:41.35 | cr2 | MAX_PROCESSES is 32. |
| 21:41.45 | cr2 | 16*4 values. |
| 21:42.56 | cr2 | it should be built in haret ;) |
| 21:44.28 | cr2 | 955c97d6 |
| 21:44.37 | cr2 | this is 'startip' |
| 21:46.01 | cr2 | well, it's too painful to do manually ;) |
| 21:46.21 | cr2 | when the code is actually available. |
| 21:47.32 | BabelO | what can i do then ? |
| 21:48.04 | cr2 | you will know where is the dll code in RAM. |
| 21:48.10 | cr2 | for 'insn' |
| 21:48.16 | cr2 | http://www.osronline.com/ddkx/network/103ndisx_4m42.htm |
| 21:48.26 | cr2 | virtual... |
| 21:49.08 | cr2 | nice m$ docs: |
| 21:49.13 | cr2 | NdisAllocateMemoryWithTag allows the caller to supply a tag that appears in any crash dump of the system that occurs subsequently. |
| 21:49.25 | cr2 | lol |
| 21:49.52 | BabelO | yes, usefull for mail to :) |
| 21:50.21 | cr2 | subsequently=inevitably |
| 21:50.51 | cr2 | the vaddr is taken from stack... |
| 21:51.34 | cr2 | length is 0xd48 |
| 21:51.47 | cr2 | 0xd40 | 0x8 |
| 21:52.21 | cr2 | tag is 0x5449574c |
| 21:52.28 | cr2 | looks like ascii to me. |
| 21:54.02 | BabelO | set insn 0x955c97d6 ; wi 10 (no value) |
| 21:55.14 | cr2 | it's only the start PC. you need to trace the called function. |
| 21:55.42 | cr2 | can yout dump 0x1345 ? |
| 21:56.00 | cr2 | RAM. |
| 21:57.53 | cr2 | 1350 is also acx100 related. |
| 21:59.21 | BabelO | it is that right ? pd 0x1345 0x200 |
| 21:59.32 | cr2 | no. |
| 21:59.49 | cr2 | pd 0xa0001344 4 |
| 22:00.03 | BabelO | it is to low level to uinderstand... |
| 22:00.20 | BabelO | a0001344 | 0000010c |
| 22:01.30 | cr2 | ok, so it's zero. cleared on init. good. |
| 22:01.51 | cr2 | the 0xd48 buffer is allocated in RAM. so it's not pcmcia. |
| 22:02.02 | BabelO | what is this address ? |
| 22:02.23 | *** join/#htc-linux WizMaui_ (n=WizMaui@62.112.90.250) |
| 22:02.52 | cr2 | it is some acx100 control structure. 0xd48 is its size. |
| 22:03.24 | BabelO | ok, so you are close to the right adress ;) |
| 22:04.28 | cr2 | then mmapiospace is called. |
| 22:04.46 | *** part/#htc-linux rmoravcik (n=rmoravci@adsl-d181.84-47-56.t-com.sk) |
| 22:05.12 | cr2 | http://msdn2.microsoft.com/en-us/library/ms958815.aspx |
| 22:05.48 | cr2 | R2 is physical. |
| 22:06.23 | cr2 | hm. R2=R4=0 |
| 22:06.58 | cr2 | R3=R4=0 |
| 22:07.02 | cr2 | length. |
| 22:08.05 | cr2 | MiniportInitialize gets the PhysicalAddress value either from the driver's Parameters registry key |
| 22:08.14 | cr2 | ok, it does not help very much. |
| 22:08.35 | cr2 | it will be probably the pcmcia memory address. |
| 22:10.30 | BabelO | ok, so memory split is not good for us |
| 22:10.52 | cr2 | NdisMSetAttributes |
| 22:11.14 | cr2 | it probably just points to the pcmcia. |
| 22:11.53 | cr2 | but we know that already. |
| 22:12.19 | BabelO | yes, we are pointing into pcmcia mem or io ? |
| 22:14.21 | cr2 | need to look for other ndis* functions. |
| 22:22.10 | cr2 | nothing useful. |
| 22:24.11 | BabelO | can we try to keep acxconfigured when booting linux, to see if acx module can drive it ? |
| 22:25.35 | BabelO | only to know if it is miss configured or really not possible with current driver and need something else |
| 22:27.38 | cr2 | i can only think about some missing asic3 power gpio, but it's unlikely. |
| 22:28.29 | BabelO | i can try to power on all output gpio i found which are not used |
| 22:29.55 | cr2 | we only need to check those referenced by pcmcia.dll |
| 22:30.39 | cr2 | gpioa 0x10 wifi reset. ok. |
| 22:30.54 | BabelO | i see psokolovsky doing some adjust to h4000 driver in the sam way |
| 22:32.20 | cr2 | 0x20 power2 |
| 22:32.33 | BabelO | ok |
| 22:32.41 | BabelO | 0x8 power1 |
| 22:33.35 | cr2 | gpiob 0x2000 power3 ok. |
| 22:33.50 | psokolovsky | there's nothing missing in initial gpio setup, period |
| 22:35.25 | cr2 | sleep(0x64) for reset. |
| 22:36.11 | cr2 | then sleep(0x32) and check wifi pxa irq status. |
| 22:38.07 | BabelO | i can try this sequance |
| 22:40.25 | cr2 | the rest is some LCD power stuff. |
| 22:41.24 | BabelO | actually we do power on then mdelay(100), reset then mdelay(100) then remove reset |
| 23:03.40 | *** join/#htc-linux leOn (n=xxx@85.139.79.205) |
| 23:03.43 | leOn | hello |
| 23:04.23 | leOn | can anyone tell me what's the preferred X windowing system for the htc devices ? |
| 23:04.42 | leOn | plain X, nano-x, tinyx, other ? |
| 23:49.06 | *** join/#htc-linux g3gg0__ (n=g3gg0@ppp-62-245-162-211.dynamic.mnet-online.de) |