IRC log for #htc-linux on 20070511

17:23.50*** join/#htc-linux apt (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
17:23.50*** topic/#htc-linux is HTC Linux Channel: Find logs at http://apt.rikers.org/%23htc-linux/ | please check http://handhelds.org/moin/moin.cgi/HTC_2dPhones | http://wiki.xda-developers.com/index.php?pagename=Xanadux | <cr2> let's define a common setup.
17:47.12BabelOgood evening ;)
17:48.26BabelOpH5: can i do some test for you with BA ?
18:12.36*** join/#htc-linux goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
18:20.00*** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be)
18:22.36*** join/#htc-linux scorpio16v (n=scorpio1@AC9EF97F.ipt.aol.com)
18:34.27*** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
18:43.17*** join/#htc-linux awelux (n=awelux@lvps87-230-8-217.dedicated.hosteurope.de)
19:17.14*** part/#htc-linux scorpio16v (n=scorpio1@AC9EF97F.ipt.aol.com)
19:17.32*** join/#htc-linux gundam (n=gundam@slackware.it/staff/gundam)
19:31.41*** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be)
19:42.06*** part/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt)
19:46.51*** join/#htc-linux gundam (n=gundam@slackware.it/staff/gundam)
19:51.40*** join/#htc-linux JuNuKN (n=chatzill@p549A8E38.dip0.t-ipconnect.de)
20:03.42*** join/#htc-linux pancake (n=xxx@55.pool85-52-227.static.orange.es)
20:03.57pancakehi*
20:44.25pH5BabelO: could you turn on the phone in wince and check haret's wi output?
20:44.41pH5(set ii(x) 1 the annoying irqs that always fire)
20:45.08pH5I'd like to know if there is a second irq gpio that fires during gsm powerup besides gpio6
20:45.12BabelOpH5: sure, i ve a special exe to power on / off phone ;)
20:45.35pH5:) cool
20:46.11BabelOneed to do some set trace 0xXXXXXX ?
20:47.02cr2BabelO: SetSystemPowerState(NULL,POWER_STATE_SUSPEND, POWER_FORCE);
20:47.08pH5BabelO: i think haret can trace the gpio irqs just like that.
20:47.56BabelOcr2: what this function doing ?
20:48.00cr2pH5: where are you going to store the output for 'suspend' ?
20:48.25cr2BabelO: suspends the device. == apm --suspend
20:48.44BabelOok with your latest are i suppose
20:48.58cr2BabelO: not yet.
20:49.12cr2i don't know if it makes sense to implement it.
20:49.14pH5cr2: I'm sorry, I didn't follow. What output and why suspend?
20:49.43cr2pH5: this wince command can be called for force suspend in software.
20:49.56cr2but i don't understand how it can be traced in haret.
20:50.10cr2because the wifi/usb/LCD et al. will be suspended too.
20:50.41pH5ah, ok. I don't know. does wince freeze haret before doing the actual suspend?
20:50.59cr2maybe it is possible to allocate a big buffer in RAM, and store the output there.
20:51.03cr2i guess it should.
20:58.22BabelOpH5, cr2 : got two irq only this one irq 25(DMA) and irq 22(FFUART)
20:58.39BabelOi don't know what is 25 irq ?
21:00.42pH5hm.. those aren't gpio irqs. what uses dma - did it play a sound?
21:01.21cr2probably. FFUART does not use DMA (although it can).
21:01.32BabelOand also irq 10(GPIOx)
21:01.45BabelOfor sound i try to shutdown sound, maybe it is a click
21:01.57cr2GPIOx must be demux'd
21:02.05pH5doesn't haret do that automatically?
21:02.11cr2into pxa or asic3.
21:02.20cr2pxa - yes. asic3 - no.
21:02.21pH5I always get a second irq after GPIOx on pxa27x
21:02.30pH5ah, asic goes3 into gpio 10, too?
21:02.35cr2no.
21:02.40pH5s/asic goes3/asic3 goes/
21:02.55cr2there should be a pxa asic3 mux gpio.
21:02.55pH5but then there should be the asic3 gpio irq after irq 10, right?
21:03.01cr2yes.
21:03.07pH5ok
21:03.39cr2i forgot the number. need to look into wiki.
21:03.57*** join/#htc-linux florian_ (n=fuchs@84.245.185.57)
21:04.02pH5should be gpio 3
21:04.10cr2yes.
21:04.11BabelOso i can mask irq 10 too
21:04.11pH5(from htcblueangel-gpio.h)
21:04.37pH5BabelO: yes, but look out for the irq of gpio 3
21:04.50cr2the only unknown irq is gpio38.
21:05.07cr2i've observed it, since it's called 'NO INFO' in the wiki.
21:05.35cr2BabelO: which gpio is the DCD ?
21:05.50cr2it must be in the logs, but i've forgot.
21:06.00BabelO36
21:06.15BabelOfor FFUART right ?
21:06.17cr2either 60 or 63.
21:06.18cr2yes.
21:06.43cr2hm.
21:06.58BabelOok i try again , restarting haret
21:07.05cr2it's an ALT gpio on ba.
21:09.13BabelOyes got some gpio 3 witj gpio 10
21:09.31BabelOsorry irq 10
21:10.39pH5oh great, where are the asic3 irq registers on ba?
21:11.24BabelOyes after two power on / off got irq22(ffuart) and irq 37(gpio3)
21:12.23BabelOdon't know, where can i check ?
21:13.43pH5htcblueangel-asic.h says it is at CS3+0x01800000 (CS3 is 0x0C000000)
21:14.02cr2BabelO: they are decoded by gnu-haret with 'wasic3gpio'
21:14.18cr2together with the pxa gpios.
21:14.19BabelOok check with this command
21:14.20pH5woo, very good :)
21:14.34cr2you only need to mask the junk bits.
21:14.40cr2like asic3 pwm0 and 1.
21:14.46cr2CONTROL_CX
21:14.54BabelOyes but never ermember how ? ;)
21:15.04cr2set iasic3gpioa(11) 1
21:15.24cr2set igpio(x) 1
21:16.35pH5the great thing with haret is somebody always already implemented whatever I happen to need.
21:18.59BabelOeh eh ;)
21:18.59cr2:)
21:19.00BabelOso for power on
21:20.31BabelOgot gpio59 gpio102 gpio6 gpio36 gpio63 gpio35 gpio60 gpio 34 gpio 39
21:20.41BabelOmaybe there is some parasit, need to clean
21:22.09BabelOgot a lot of gpio 85 too, but not related to phone power on or off
21:22.24BabelO102 is not on pxa263
21:24.36cr2yes, these are stray bits from pxa27x code.
21:25.06BabelOgpio59= gsm reset,  gpio6 = gsm alert, gpio 36 is DCD, 35 is CTS, 60 and 63 unknown, 34 =RXD GSM and 39 txd GSM
21:25.17cr236 and 63 are together ? that's good.
21:25.49BabelOGPIO6 S 1 1631109
21:25.53BabelOGPIO36 S 0 1631632
21:25.53BabelOGPIO63 S 0 1631633
21:25.53BabelOGPIO35 S 0 1631692
21:25.54BabelOGPIO36 S 1 1631863
21:25.54BabelOGPIO60 S 0 1631864
21:25.54BabelOGPIO63 S 1 1631865
21:26.38BabelOand after toggle 34 and and 39
21:27.41pH5so there is no asic3 gpio now at all?
21:28.55BabelOseems no
21:30.32BabelOdon't know what it is this gpio 85 :(
21:31.27BabelOso pH5 your question is why asic3 related irq are toggle during power on / off and there is no asic3 related here
21:31.38pH5BabelO: I guess it's noise from the pxa27x implementation. does pxa263 have gpio 85 at all?
21:32.04BabelOyes have 89 gpio
21:32.06pH5BabelO: exactly. are 60,63 inputs or outputs?
21:32.17BabelOinput
21:34.41BabelOpH5: is it possible GPIO85 is ds1wm related ? because it toggle periodically 10 times
21:37.05pH5BabelO: I don't think so. The OWM irq is known, and ds1wm only has one interrupt line.
21:37.19BabelOok
21:38.34pH5so now we know that the gsm ready irq could be one of 60,63 or some asic3 irq?
21:38.58BabelOi think asic3 irq are due to sound
21:38.59cr2BabelO: gpio85 has nothing to do with ds1wm.
21:39.22cr2pH5: i think it's 63.
21:39.28BabelOcr2: ok, it is only because i show it periodically like a call to something
21:39.49cr2BabelO: i've looked at the battery.dll code.
21:40.09cr2if you'll tell me which dll/exe uses gpio85, i'll tell you what it can be.
21:40.44BabelOok , with your friend IDA ;)
21:41.45cr2does haret show the address of an instruction for wi ?
21:42.09cr2then we can guess which dll it is.
21:42.26cr2by the dll base load addreess.
21:42.41BabelOno str or ldr during wi
21:43.32cr2ok.
21:43.49pH5cr2: does the wasic3 code work for blueangel? maybe the memory address is wrong?
21:44.20pH5BabelO: for that you'd have to trace the GPLR2 register and check the ldr's.
21:44.56BabelOok
21:47.12pH5hm.. where is the ASIC3GPIOBASE haret variable set?
21:47.13BabelOpH5: i set trace 0x40E00008 0x8; wi 30
21:47.21BabelObut nothing appear
21:49.48cr2pH5: they are hardcoded. depending on the machine.
21:50.27cr2pH5: wasic3 works on ba/hima/universal. i've tested all of them
21:50.49pH5ok, then I'll trust you that they are correct :)
21:50.56cr2the only device where there were some problems recently is ipaq.
21:51.05cr2it's all in the wiki.
22:39.23*** join/#htc-linux cr2 (n=konversa@crpl22.physik.uni-wuppertal.de)
22:56.43pofanyone knows how can i redirect 'cu' output to a file ?
22:56.56pancakehi pof
22:57.03pofhi pancake :)
22:57.11pancakemaybe with libfdsniff
22:57.33cr2pof: run 'script'
22:58.08pofthanks!
22:58.13*** part/#htc-linux rmoravcik (n=rmoravci@adsl-d181.84-47-56.t-com.sk)
22:59.21pofnah... not working :(
22:59.24pofhangs my terminal
22:59.28pofit's binary data
22:59.34pofi'm trying to capture a 'rbmc' dump
22:59.49pancakepof: let me try ;)
23:04.20pancakehey pof
23:04.23pancakeuse this: # ((sleep 1 && echo rbmc file 50000000 80000 ) | cu -l /dev/ttyUSB0 ) > pop
23:06.00pofthanks pancake, it works :)
23:06.49cr2you can pipe to hexdump, and then recover.
23:07.35BabelOcr2: gpio 4 is toggle web camera image is displayed, maybe ati capture irq ?
23:09.34pancakepof. you can embed this system() call in your flasher
23:10.14pancakebu would be good not to depend on cu
23:11.00cr2BabelO: wiki ?
23:11.28BabelOcr2: yes wiki say this ATI irq?  ...
23:11.49cr2yes. it's how wince kernel names it.
23:12.18BabelOi see it only when i activate camera
23:13.13cr2so it's capture-related.
23:13.27BabelOyes ;)
23:13.46BabelOi m looking for recording sound
23:15.08*** join/#htc-linux cyrill62 (n=user@ble59-5-82-233-205-36.fbx.proxad.net)
23:38.35*** join/#htc-linux bd2 (n=cbou@89.113.85.104)
23:43.29*** join/#htc-linux psokolovsky (n=psokolov@empty-crossing.volia.net)
23:48.27pancakecr2: what's the next step on hermes-linux?
23:57.13cr2pancake: to fix the 2.6.21 kernel
23:59.23pancakewhat's the problem?

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.