| 00:09.55 | *** join/#htc-linux gw280 (i=authdeni@vm.gwright.org.uk) |
| 00:20.30 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-179-036.pools.arcor-ip.net) |
| 06:03.35 | *** join/#htc-linux m_tadeu (n=miguel@62.169.113.14.rev.optimus.pt) |
| 06:24.41 | *** join/#htc-linux rob_w (n=rob_w@p549B970D.dip0.t-ipconnect.de) |
| 08:00.14 | *** join/#htc-linux rob_w (n=rob_w@p549B970D.dip0.t-ipconnect.de) |
| 09:07.24 | *** join/#htc-linux florian_kc (n=fuchs@dns.kernelconcepts.de) |
| 09:11.34 | florian_kc | good morning |
| 11:19.35 | *** join/#htc-linux dullard (n=jim@adsl-static-1-30.uklinux.net) |
| 12:13.52 | rob_w | lamikr, are u alive ? |
| 13:47.38 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 16:22.33 | *** join/#htc-linux rob_w (n=bob@p213.54.218.193.tisdip.tiscali.de) |
| 16:36.45 | *** join/#htc-linux pH5 (n=ph5@e178229111.adsl.alicedsl.de) |
| 18:00.10 | *** join/#htc-linux jeanseb (n=jeanseb@88.164.32.155) |
| 18:02.21 | *** join/#htc-linux pleemans (n=peter@d51A5E421.access.telenet.be) |
| 18:11.28 | pH5 | cr2: it looks like I have a ds2762 in the battery after all. the serial communication seems to match the ds2726 data sheet. |
| 18:14.25 | cr2 | pH5: it was reasonable to expect such outcome. |
| 18:24.19 | *** join/#htc-linux goxboxlive (n=jrs@9.80-202-160.nextgentel.com) |
| 18:27.03 | rob_w | pH5, what does the "battery tell " you ? |
| 18:27.08 | rob_w | anything interesting ? |
| 18:27.13 | rob_w | capacity ? |
| 18:29.16 | pH5 | rob_w: I don't have working code yet. Are there cases where the ds2762 doesn't have useful information in all registers? |
| 18:32.40 | rob_w | what is your plan ? |
| 18:33.54 | rob_w | cool the specs really look as if it is all about capacity monitoring and battery treatment |
| 18:39.02 | rob_w | i added the pdf link to the hardwarexref wiki |
| 18:41.02 | pH5 | the wince code seems to use all the interesting bits |
| 18:41.17 | pH5 | voltage, current, ACR, temperature and both eeprom blocks |
| 18:43.30 | rob_w | hmm yet you still need to know the battery specs in order to decode the capacity for all those , right ? |
| 18:49.05 | pH5 | rob_w: it's in the datasheet. for temperature and voltage the units are given, for currents there are two possibilites depending on how it's wired. |
| 18:49.58 | pH5 | but as I understand there is a ds2762 driver already - looks like I just need to provide the w1_read/write interface? |
| 18:54.33 | cr2 | pH5: it seems yes. |
| 18:55.46 | cr2 | rob_w: i don't think blueangel uses ds276x, even if it is in the battery. because the asic3 w1 controller is disabled, and a lot of information is not shown in the powermon log. |
| 18:56.45 | pH5 | cr2: how does ba measure the voltage, then? |
| 18:58.00 | cr2 | ADC in the ts2200 |
| 18:58.19 | cr2 | i.e. directly. |
| 18:58.38 | cr2 | and it does not measure or show the current, btw. |
| 18:59.12 | pH5 | ah, ok. I think I have the backup battery voltage on ts2046 aux |
| 19:00.03 | cr2 | actually i don't understand why they did not use w1 on blueangel. maybe because it's an old model. |
| 19:00.14 | cr2 | pH5: like on hx4700 ? |
| 19:01.35 | cr2 | goxboxlive: are you listening ? |
| 19:01.54 | pH5 | I think so, yes. |
| 19:02.22 | cr2 | it seems oe moved to glibc-2.4 and i have rebuild the whole tree. i'm wondering how good is this rootfs for the universal. |
| 19:03.11 | pH5 | no idea. I use glibc2.4 for ages, same image for hx4700 and magician. |
| 19:04.34 | goxboxlive | no, not realy, should i? I am fixing my desktop pc |
| 19:07.25 | cr2 | goxboxlive: can i just dump this rootfs on the SD and expect it to work ? |
| 19:07.45 | goxboxlive | probably as far as i remember |
| 19:07.48 | goxboxlive | OPIE |
| 19:07.52 | goxboxlive | not GPE |
| 19:08.54 | cr2 | yes. |
| 19:15.29 | *** join/#htc-linux skodde (n=skodde@213-140-16-178.fastres.net) |
| 19:22.44 | cr2 | pH5: i have an interesting debugging question here. wince uses pxa gpio110 for some phone function. can i find out which program/dll/whatever controls it without disassembling everything ? |
| 19:24.22 | pH5 | cr2: you could look at the wi output, get the instruction address and look at all the headers to find the matching dll |
| 19:24.37 | cr2 | it is not an interrupt. |
| 19:24.46 | pH5 | is it an output? |
| 19:24.51 | cr2 | yes. |
| 19:25.08 | pH5 | so you can find the address of the instruction where it is set? |
| 19:25.20 | cr2 | how ? |
| 19:25.54 | pH5 | with a memory trace on the GPSR |
| 19:26.28 | cr2 | hmm. isn't it too much output ? |
| 19:27.32 | pH5 | hm... that could be. |
| 19:32.42 | pH5 | cr2: try set trace 0xaee00118 - I don't think there should be too much output on GPSR3 |
| 19:33.30 | cr2 | 115 is pwm1 output ;) |
| 19:33.40 | cr2 | ah. SR3. |
| 19:33.57 | cr2 | i'll try. |
| 19:41.01 | cr2 | btw, can i demux the pxa irq ? |
| 19:42.49 | cr2 | 02fc85ed: debug 0935f844: e5832118(str) 00004000 aee00000 |
| 19:47.26 | cr2 | it is always appearing in this combination: |
| 19:47.29 | cr2 | 046b0b73: irq 22(FFUART) |
| 19:47.29 | cr2 | 046b101a: debug 0935f844: e5832118(str) 00004000 aee00000 |
| 19:47.29 | cr2 | 046b5df9: irq 22(FFUART) |
| 19:49.56 | cr2 | heh. |
| 19:50.02 | cr2 | 01ae661a: debug 0935f770: e58e3124(str) 00004000 deadbeef |
| 19:50.32 | pH5 | I'd search for a dll that has 0135f844 inside its .data section. |
| 19:52.24 | cr2 | 09-01 ? |
| 19:52.26 | cr2 | ah. ok. |
| 19:52.54 | cr2 | but it's forth process slot. |
| 19:53.25 | cr2 | dump ps |
| 19:53.28 | cr2 | ps: pid=b3cb478a ppid=0 pmem=8000000 hand=b3cb478a procname=device.exe |
| 19:55.21 | pH5 | I don't understand. |
| 19:55.43 | *** join/#htc-linux BabelOued (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 19:55.51 | cr2 | dump modules |
| 19:56.10 | cr2 | does not tell very much. |
| 19:56.22 | cr2 | 0135 . it should be some XIP dll. |
| 19:57.17 | pH5 | you could try objdump -h |
| 19:57.40 | cr2 | afaik a dll is always loaded 64K aligned |
| 19:57.54 | cr2 | for objdump you need it dumped. |
| 19:58.06 | cr2 | which is very difficult for wince5. |
| 19:58.13 | pH5 | oh.. |
| 19:58.27 | pH5 | ~hail 2003se |
| 19:58.36 | apt | ACTION bows down to 2003se and chants, "I'M NOT WORTHY!!" |
| 19:58.54 | cr2 | but it's the same 0135 dll that uses SR and CR. |
| 19:59.17 | cr2 | now i will learn that it is rilgsm ;-) |
| 19:59.41 | pH5 | sounds likely, somehow. |
| 19:59.51 | cr2 | then the problem will be: who calls the exported RIL_DevSpecific function. |
| 20:02.23 | cr2 | hehe. rilgsm is 0140-1043 |
| 20:20.37 | cr2 | start address 0x01352644 |
| 20:22.31 | cr2 | ohci.dll |
| 20:22.31 | cr2 | start address 0x01352644 |
| 20:22.43 | cr2 | not bad. |
| 20:23.31 | cr2 | i thought it is a highlevel dll. |
| 20:35.48 | pH5 | nice |
| 20:48.36 | cr2 | hmm. did not help... |
| 20:52.54 | cr2 | what does it mean: |
| 20:52.59 | cr2 | debug 0935f780: e1dee1b8(ldrh) deadbeef deadbeef |
| 20:57.10 | pH5 | return nr < 14 ? regs->regs[nr] : 0xdeadbeef; |
| 20:57.16 | pH5 | in haret/src/irq.cpp |
| 21:00.59 | cr2 | hmm. nr=14 or 15. |
| 21:12.59 | goxboxlive | cr2: What did you do to build Opie-bluepin? Did you change anything in the OE? |
| 21:14.51 | cr2 | goxboxlive: yes. |
| 21:17.28 | cr2 | OE_QMAKE_LINK="${CXX}" |
| 21:17.43 | goxboxlive | where? |
| 21:17.46 | cr2 | in ./org.openembedded.dev/classes/qmake-base.bbclass |
| 21:17.50 | goxboxlive | thx |
| 21:20.14 | goxboxlive | cr2: did u replace this one? export OE_QMAKE_LINK="${CCLD}" |
| 21:22.55 | cr2 | yes. |
| 21:26.05 | goxboxlive | ok, my buil machine is up again. I am trying to buil nano now, just to check if everything works. |
| 21:31.08 | goxboxlive | cr2: In the hh.org cvs. Why is the MAIN realeas newer that *.17-hh1 ? |
| 21:31.27 | goxboxlive | isent the 17-hh1 the latest one? |
| 21:32.56 | cr2 | i have 17-hh1 |
| 21:46.54 | goxboxlive | me to, but the HEAD shows newer files : http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-pxa/htcuniversal/?only_with_tag=MAIN |
| 21:49.26 | cr2 | udevd is unhappy: invalid KERNEL operation |
| 21:50.27 | cr2 | don't use the tag. |
| 21:51.19 | cr2 | the acx debugging is crazy. |
| 21:54.11 | cr2 | it seems i need a reset. the lcd is flooded by the debug messages. |
| 21:59.57 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.179) |
| 22:02.11 | goxboxlive | i know. The acx debug should be shut off |
| 22:36.36 | *** join/#htc-linux WizMaui_ (n=WizMaui@62.112.90.179) |
| 22:52.31 | *** join/#htc-linux awelux (n=awelux__@dslb-084-058-188-237.pools.arcor-ip.net) |
| 22:56.52 | *** join/#htc-linux WizMaui (n=WizMaui@62.112.90.179) |
| 23:40.55 | *** join/#htc-linux booba (n=booba@81.56.252.135) |
| 23:59.08 | *** join/#htc-linux lamikr (n=chatzill@aragorn.kortex.jyu.fi) |