| 00:00.05 | tmzt | rpc user driver |
| 00:00.24 | cr2 | ok |
| 00:00.27 | tmzt | must be linked to fd |
| 00:00.39 | tmzt | so that userspace can select bthe events |
| 00:01.00 | phh | http://husson.hd.free.fr/dm.6 |
| 00:01.00 | cr2 | ok |
| 00:01.11 | cr2 | canyou dump the packet in E/RPC ? |
| 00:01.26 | phh | not as easily but yes |
| 00:02.00 | cr2 | end sigkill to 1472 (ndroid.calendar), adj 14, size 4341 |
| 00:02.02 | cr2 | lol |
| 00:02.46 | cr2 | 3100005b 00000000 0000000f |
| 00:02.48 | cr2 | aha |
| 00:03.25 | cr2 | a970bdc1 003c18c8 00000004 00000009 |
| 00:03.27 | phh | cr2: should I expect 3000005b and 3100005b have the same version ? |
| 00:03.45 | cr2 | don't know |
| 00:03.56 | cr2 | but it's your server, you register it |
| 00:04.08 | cr2 | and rpcrouter tells arm9 about it |
| 00:04.20 | phh | the problem is that the messages are supposed to go to libgps.so. |
| 00:04.28 | cr2 | arm9 must get and honor this data |
| 00:04.55 | cr2 | without g1 rpc fifo dump you will never know |
| 00:05.13 | cr2 | i think GNUtoo made the gps smem dump for me. |
| 00:05.21 | cr2 | hm.. where did it disappear... |
| 00:05.31 | phh | I might ask ida. |
| 00:05.46 | cr2 | ok |
| 00:06.36 | tmzt | userspace has to be registering the callback to get called |
| 00:07.18 | cr2 | i should grep my RPC fifo logs for 3100005b messages |
| 00:07.28 | phh | cr2: how can I see the calls to svc_register function in IDA ? |
| 00:09.31 | cr2 | alt-t ? |
| 00:09.33 | herpez|away | well, i see that today is a day full of work for you guys!!! are you worki at something specific? (yes, i can read the logs, but i cant understand them ;) ) |
| 00:09.53 | cr2 | herpez|away: started with cam, and moved to gps ;) |
| 00:09.58 | phh | seems to work thanks |
| 00:10.07 | phh | herpez|away: and I tried 3D earlier |
| 00:10.10 | phh | so nothing specific at all :p |
| 00:10.27 | cr2 | svc with plugger, and svc with proxy. wince is strange |
| 00:10.44 | phh | expected version = 0x728B7ED4 for 0x31000001d |
| 00:10.45 | herpez|away | when you talked about NAND, the goal is to to run android from internal memory? |
| 00:11.10 | phh | herpez|away: yes |
| 00:11.21 | herpez|away | ok, no more questions :) |
| 00:11.26 | cr2 | 2 is plugger,, 13 is proxy |
| 00:12.00 | phh | cr2: how vbus detection doesn't seem stupid btw. |
| 00:12.06 | phh | well it detects vbus when it's here |
| 00:12.12 | phh | (I don't know for the opposite yet :D) |
| 00:12.23 | cr2 | phh: ? |
| 00:12.40 | cr2 | 1d is plugger |
| 00:12.54 | phh | cr2: I putted your vbus detection with readl() |
| 00:12.56 | phh | and it seems to work. |
| 00:13.01 | cr2 | lol |
| 00:13.11 | phh | you already knew that for a fact. ok. |
| 00:13.37 | cr2 | it is in the wiki for a year :D |
| 00:13.40 | phh | arf |
| 00:14.17 | cr2 | +0xfc00c Vbus, usb/ac charging? |
| 00:14.41 | cr2 | usb/ac charging? should go away |
| 00:15.17 | cr2 | however it's certainly linked to vbus :) |
| 00:15.40 | tmzt | vbus without handshake is ac |
| 00:15.43 | cr2 | phh: it shouldbe checked for rhod/top |
| 00:15.46 | tmzt | or at least 1A |
| 00:15.49 | phh | cr2: and cdma ? |
| 00:16.15 | cr2 | tmzt: i think it's reported by pmic, and arm9 puts it into smem |
| 00:16.28 | cr2 | phh: need to ask MrPippy |
| 00:16.35 | cr2 | phh: but i think it is. |
| 00:16.36 | tmzt | rhod400/500? |
| 00:16.42 | phh | ok |
| 00:16.59 | cr2 | tmzt: start the MemorMap wiki page for it. |
| 00:17.03 | MrPippy | hmm? |
| 00:17.18 | tmzt | can't do wiki on phone |
| 00:17.23 | cr2 | MrPippy: can you check the smem location for vbus ? |
| 00:17.33 | cr2 | tmzt: lol |
| 00:17.40 | tmzt | actually don't understand your wiki too well |
| 00:17.46 | tmzt | what should be different |
| 00:17.49 | cr2 | tmzt: nc10 is 1024x600 |
| 00:18.02 | MrPippy | is haret or linux easier for that? |
| 00:18.10 | tmzt | if I can forward the port back through putty I think I can check it |
| 00:18.14 | phh | MrPippy: linux :p |
| 00:18.16 | tmzt | cr2: is it via? |
| 00:18.35 | tmzt | nc20 can boot an external bios chip from lpc bus |
| 00:18.48 | cr2 | MrPippy: phh can commit the pqatch, and you can test it in linux |
| 00:18.55 | cr2 | tmzt: nc10 ? |
| 00:19.10 | tmzt | not sure, you could ask in #coreboot |
| 00:19.22 | tmzt | (linuxbios) |
| 00:19.23 | cr2 | 00:00.0 Host bridge: Intel Corporation Mobile 945GME Express Memory Controller Hub (rev 03) |
| 00:19.25 | cr2 | 00:02.0 VGA compatible controller: Intel Corporation Mobile 945GME Express Integrated Graphics Controller (rev 03) |
| 00:19.26 | cr2 | 00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03) |
| 00:19.28 | cr2 | 00:1b.0 Audio device: Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller (rev 02) |
| 00:19.35 | phh | cr2: not now. |
| 00:19.35 | tmzt | oh |
| 00:19.40 | phh | my tree is broken because of gps |
| 00:19.41 | tmzt | guess not |
| 00:20.03 | cr2 | tmzt: and |
| 00:20.07 | cr2 | 00:1f.3 SMBus: Intel Corporation 82801G (ICH7 Family) SMBus Controller (rev 02) |
| 00:20.08 | cr2 | 02:00.0 Ethernet controller: Atheros Communications Inc. AR5001 Wireless Network Adapter (rev 01) |
| 00:20.10 | cr2 | 03:00.0 Ethernet controller: Marvell Technology Group Ltd. 88E8040 PCI-E Fast Ethernet Controller (rev 13) |
| 00:20.16 | cr2 | tmzt: everything works |
| 00:20.22 | tmzt | but ich7 is becoming reality for them |
| 00:20.27 | tmzt | in linux? |
| 00:20.30 | tmzt | you mean |
| 00:20.35 | phh | MrPippy: http://husson.hd.free.fr/0001-Better-vbus-detection.patch |
| 00:20.41 | cr2 | tmzt: yes. and usb |
| 00:20.45 | cr2 | Bus 001 Device 004: ID 0bb4:0303 High Tech Computer Corp. |
| 00:20.46 | cr2 | Bus 001 Device 003: ID 0ac8:c326 Z-Star Microelectronics Corp. |
| 00:20.48 | cr2 | Bus 001 Device 002: ID 04e8:6773 Samsung Electronics Co., Ltd |
| 00:20.56 | phh | MrPippy: push it if it works |
| 00:21.03 | cr2 | raph100, cam and umts modem |
| 00:21.08 | tmzt | yep |
| 00:21.12 | cr2 | phh: it's 1 line patch ? |
| 00:21.16 | phh | yes |
| 00:21.18 | tmzt | ar5k? |
| 00:21.19 | cr2 | ok |
| 00:21.21 | phh | cr2: I know it can be done better. |
| 00:21.27 | cr2 | tmzt: wifi |
| 00:21.47 | tmzt | yeah I know |
| 00:21.52 | tmzt | ar5k driver works? |
| 00:22.01 | cr2 | phh: there is something in the usb code too |
| 00:22.09 | cr2 | tmzt: wifi ? yes |
| 00:22.11 | phh | cr2: uh ? |
| 00:22.20 | cr2 | phh: that uses vbus |
| 00:22.35 | phh | cr2: in htc_battery_smem there is a msm_hsusb_set_vbus |
| 00:22.39 | phh | is that what you're talking about ? |
| 00:22.49 | tmzt | is that real now or still fake vbus? |
| 00:22.58 | phh | real |
| 00:23.06 | tmzt | good |
| 00:23.08 | cr2 | tmzt: vbus reported by pmic |
| 00:23.16 | cr2 | i guess it's "real" |
| 00:23.17 | tmzt | so if it changes hsusb gets updated? |
| 00:23.27 | tmzt | yeah, there's a comment to fix that function |
| 00:23.39 | tmzt | from Google I think |
| 00:23.47 | cr2 | yes, somewhere in the code |
| 00:24.09 | phh | ok. last try before bed now. |
| 00:24.13 | cr2 | maybe it's the same even on leo |
| 00:24.45 | cr2 | phh: let's finish the cam tomorrow morning |
| 00:24.59 | phh | ok :D |
| 00:26.27 | phh | btw, is it me or the set_time_gps thing sounds stupid ? |
| 00:26.36 | phh | I mean, we get the date from the modem ... |
| 00:26.47 | phh | most of the time. |
| 00:26.54 | cr2 | phh: it helps gps |
| 00:27.06 | phh | I know but modem already got date with GSM or UMTS |
| 00:27.12 | cr2 | time from modem is from network. |
| 00:27.23 | cr2 | nd it's not always available |
| 00:27.28 | cr2 | as a gsm service |
| 00:28.51 | cr2 | 5b is plugger |
| 00:29.09 | phh | grr I can't get divert working |
| 00:29.19 | phh | read divert |
| 00:29.31 | phh | anyway. time to sleep. |
| 00:32.03 | cr2 | good night |
| 00:47.15 | *** part/#htc-linux DarkMasterHalo (n=DarkMast@modemcable187.118-80-70.mc.videotron.ca) |
| 00:53.49 | *** part/#htc-linux we350z (n=bkusnir@adsl-99-8-225-30.dsl.pltn13.sbcglobal.net) |
| 00:55.04 | *** join/#htc-linux Traveler0 (n=traveler@cpe-066-057-232-249.nc.res.rr.com) |
| 00:58.28 | *** join/#htc-linux TheMuffinMan (n=themuffi@ool-4575b0a0.dyn.optonline.net) |
| 01:01.42 | *** join/#htc-linux Kevin2 (n=Kevin2@207-172-165-101.s101.tnt1.nywnj.ny.dialup.rcn.com) |
| 01:55.16 | *** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821) |
| 01:59.28 | *** join/#htc-linux ALoGeNo (n=alogeno@unaffiliated/alogeno) |
| 02:08.58 | *** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a) |
| 02:10.44 | *** join/#htc-linux sse2 (n=sse2@41.104.111.48) |
| 02:14.45 | *** join/#htc-linux droid001 (n=g1@p4FDCA526.dip.t-dialin.net) |
| 02:14.55 | *** join/#htc-linux ali1234 (n=al@s15238489.onlinehome-server.info) |
| 02:23.40 | *** join/#htc-linux zolikaHUN (n=zolikaHU@dsl51B6CEEF.pool.t-online.hu) |
| 02:23.58 | zolikaHUN | Enter text here...ha |
| 02:24.02 | zolikaHUN | hi |
| 02:24.05 | *** join/#htc-linux kam187 (n=kam187@81-179-8-102.dsl.pipex.com) |
| 02:24.16 | zolikaHUN | magyar vagy? |
| 02:24.33 | zolikaHUN | are you here? |
| 02:25.30 | tmzt | looking for somebody? |
| 02:25.56 | zolikaHUN | hey |
| 02:26.03 | zolikaHUN | i'm here |
| 02:27.16 | zolikaHUN | have you hot htc pda? |
| 02:27.44 | zolikaHUN | g |
| 02:28.55 | zolikaHUN | i have a htc touch diamond with win os, but i would like to change it for linux os |
| 02:48.50 | *** join/#htc-linux bartman` (n=bart@tau.jukie.net) |
| 02:50.20 | *** join/#htc-linux bartman` (n=bart@tau.jukie.net) |
| 04:13.26 | *** join/#htc-linux ali12341 (n=al@s15238489.onlinehome-server.info) |
| 04:26.24 | *** join/#htc-linux zycho_ (n=zycho@hnvr-4d078086.pool.mediaWays.net) |
| 04:26.36 | kupo | damn i wish i would have talked to we350z earlier |
| 04:43.55 | AstainHellbring | NetRipper you around?> |
| 05:18.33 | *** join/#htc-linux droid0011 (n=g1@p4FDCB496.dip.t-dialin.net) |
| 06:05.43 | *** join/#htc-linux jumoit (n=samuel@121.77.87.145) |
| 06:36.31 | tmzt | AstainHellbring: he's on nexus tree I think |
| 06:36.58 | tmzt | not pushed anywhere yet I think |
| 06:39.25 | *** part/#htc-linux jumoit (n=samuel@121.77.87.145) |
| 06:46.00 | *** join/#htc-linux randomblame (n=kevin@c-71-59-184-96.hsd1.wa.comcast.net) |
| 06:46.23 | randomblame | anyone on? |
| 06:47.12 | tmzt | yeah |
| 06:47.24 | randomblame | how's it going? |
| 06:47.50 | tmzt | ok |
| 06:48.00 | randomblame | I just got a waterdamaged incite in the mail free |
| 06:48.11 | randomblame | I'm playing frankenstein with the parts |
| 06:48.21 | randomblame | can't identify the lcd though bah |
| 06:50.30 | randomblame | I'm thinking of soldering it up to a serial cable and writing a driver for it on the pc |
| 06:51.30 | tmzt | not that kind of lcd |
| 06:52.08 | randomblame | why not, if you supply the right voltage and the right signal to the proper pins it will display output |
| 06:52.51 | tmzt | how many pins? |
| 06:52.59 | tmzt | your serial can drive on tx |
| 06:53.13 | tmzt | even parallel is too slow |
| 06:53.18 | randomblame | 24, only 8 are for signal the rest are power and ground |
| 06:53.26 | tmzt | hmm, mddi? |
| 06:53.29 | randomblame | yes |
| 06:53.49 | tmzt | you need a high speed serial link like sata |
| 06:54.13 | randomblame | bah that would be way too much work |
| 06:54.40 | tmzt | essentially impossible |
| 06:55.11 | randomblame | yeah, I want to get the fb working |
| 06:55.31 | randomblame | but I'm tired of working the problem |
| 06:55.45 | tmzt | just get jtag working |
| 06:55.49 | tmzt | much more useful |
| 06:55.57 | randomblame | yeah we have a jtag port |
| 06:56.16 | tmzt | cool |
| 06:56.26 | randomblame | I don't really want to solder on my phone though |
| 06:56.37 | randomblame | those are some tiny tiny points |
| 06:56.47 | tmzt | I thought the phone was broken |
| 06:57.09 | randomblame | that phone is broken, what good is jtagging to a dead phone though |
| 06:58.35 | randomblame | I'm sure there is a connector somewhere out there that will click right onto this jtag port |
| 06:58.45 | randomblame | I just don't know where to start looking |
| 07:00.59 | randomblame | I think the panel driver is the only thing keeping us from a real working frame buffer |
| 07:01.27 | randomblame | and I think I know exactly what lcd this is |
| 07:01.36 | randomblame | but I can't get the whitepaper for it |
| 07:02.29 | randomblame | samsung: S6D04 |
| 07:03.27 | tmzt | oh |
| 07:03.47 | tmzt | you could maybe let a device drive it mddi and talk to it for lcm init |
| 07:03.53 | tmzt | but I don't think so |
| 07:04.02 | tmzt | because mddi chip should be in lcm |
| 07:07.33 | randomblame | is mddi just a direct memory link |
| 07:09.10 | *** join/#htc-linux makkonen (n=makkonen@cpe-66-69-229-9.austin.res.rr.com) |
| 07:09.43 | tmzt | no, it's a serial link driven by dma |
| 07:10.12 | randomblame | thats what I meant |
| 07:10.43 | tmzt | high speed |
| 07:14.51 | randomblame | samsung has me irritated |
| 07:15.17 | randomblame | I requested the white paper from them and they refused to give it to me |
| 07:15.24 | tmzt | yeah |
| 07:15.27 | tmzt | they can't |
| 07:15.38 | tmzt | but qualcomm.com has some information |
| 07:15.54 | tmzt | just don't init it, use the nop version |
| 07:16.04 | tmzt | assuming you boot from ce |
| 07:16.15 | randomblame | no I meant the lcd whitepaper |
| 07:17.07 | randomblame | I'd rather have straight ce than winmo |
| 07:17.37 | randomblame | mayhaps I'll find a way to cook a rom like that one of these days |
| 07:17.52 | *** join/#htc-linux rashire (n=ed1112wa@pool-98-114-205-192.phlapa.fios.verizon.net) |
| 07:18.11 | *** join/#htc-linux marcin_ (n=marcin@chello089077185059.chello.pl) |
| 07:35.40 | *** join/#htc-linux Bry8Star_ (n=Bry8Star@adsl-99-32-1-4.dsl.lsan03.sbcglobal.net) |
| 07:39.42 | *** join/#htc-linux Bry8Star__ (n=Bry8Star@adsl-99-48-209-136.dsl.lsan03.sbcglobal.net) |
| 07:56.22 | randomblame | is there any way to query the lcd id |
| 07:57.00 | randomblame | so I could write a quick patch to the panel driver to not kill itself when it detects it |
| 07:59.57 | *** part/#htc-linux Psychonerd (n=Casey@67.143.1.240) |
| 08:27.24 | *** join/#htc-linux ALoGeNo (n=alogeno@unaffiliated/alogeno) |
| 08:28.15 | *** join/#htc-linux MSM (n=michael@222-155-71-177.jetstream.xtra.co.nz) |
| 08:40.17 | randomblame | how was this information found for diamond |
| 08:40.19 | randomblame | .product_id = (0xd263 << 16 | 0), |
| 08:40.20 | randomblame | .name = "TC358720XBG", |
| 08:40.20 | randomblame | .id = 0, |
| 08:40.20 | randomblame | .client_data = &toshiba_client_data, |
| 08:40.22 | randomblame | .clk_rate = 0, |
| 08:40.42 | randomblame | product_id especially |
| 08:55.03 | *** part/#htc-linux MSM (n=michael@222-155-71-177.jetstream.xtra.co.nz) |
| 09:00.48 | *** join/#htc-linux leobaillard (n=leobaill@leobaillard.org) |
| 09:11.00 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d879ec5.pool.mediaWays.net) |
| 09:18.12 | *** join/#htc-linux cr2 (n=cr2@ip-109-85-35-207.web.vodafone.de) |
| 09:19.53 | phh | hi cr2 |
| 09:29.25 | randomblame | is there somewhere in the code to manually enter your frame buffer memory address? |
| 09:29.55 | phh | randomblame: htc_fb_console ? |
| 09:30.05 | randomblame | thats the console fb |
| 09:30.27 | randomblame | I want to make the fully featured frame buffer work |
| 09:30.28 | phh | which framebuffer then ? |
| 09:30.44 | phh | hum it depends |
| 09:30.50 | phh | on htc-msm-2.6.27 it's done in pmem.c |
| 09:31.02 | randomblame | thats the kernel I'm working off |
| 09:31.55 | *** join/#htc-linux toi (n=toi@d54C2A96D.access.telenet.be) |
| 09:32.17 | randomblame | but I don't see that file |
| 09:32.29 | phh | arch/arm/mach-msm/pmem.c |
| 09:33.06 | randomblame | nope |
| 09:33.15 | randomblame | not there |
| 09:34.52 | randomblame | pm.c? |
| 09:35.58 | phh | no pmem.c |
| 09:36.02 | phh | randomblame: you're using linuxtogo ? |
| 09:36.34 | phh | if you want to use linuxtogo (but I think you don't.), it's in the board-blabla.h |
| 09:36.37 | randomblame | I found pmem.c under drivers/misc/pmem.c |
| 09:36.44 | phh | no it's not that. |
| 09:36.58 | phh | it's arch/arm/mach-msm/pmem.c |
| 09:37.05 | phh | but it's only in the gitorious htc-msm-2.6.27 tree |
| 09:37.27 | randomblame | well then I don't have it haha |
| 09:37.56 | randomblame | should I be working off a different git |
| 09:38.22 | randomblame | yes I think it's a linuxtogo git |
| 09:38.36 | phh | the one that is the most commonly used now is http://gitorious.org/linux-on-qualcomm-s-msm/linux-msm |
| 09:39.15 | randomblame | I'll try that one out |
| 09:39.36 | randomblame | if it depends less on these damned panel drivers |
| 09:39.43 | randomblame | then it's worth every second of the download |
| 09:39.52 | phh | framebufffer address is in pmem.c for this one |
| 09:39.57 | randomblame | awesome |
| 09:40.15 | randomblame | I'll bet that will fix it |
| 09:40.33 | phh | I can't see how |
| 09:40.41 | phh | framebuffer address can be whereever you want |
| 09:41.09 | randomblame | the physical address? |
| 09:41.14 | phh | yes |
| 09:41.37 | randomblame | then why doesn't it work when I change it in htc_fb_console |
| 09:41.43 | randomblame | it must be a specific value |
| 09:41.59 | phh | htc_fb_console uses wince's framebuffer settings |
| 09:42.03 | phh | it does no init at all |
| 09:42.20 | phh | msm_fb does the init and tell the mdp where is the fb |
| 09:42.58 | randomblame | well setting the framebuffer size would make a difference right |
| 09:43.18 | phh | yes |
| 09:43.23 | randomblame | good |
| 09:43.56 | randomblame | is that also in pmem.c |
| 09:44.02 | phh | yes |
| 09:45.01 | randomblame | with my current kernel build htc_fb console works up until enabling smd |
| 09:45.17 | phh | which board ? |
| 09:45.26 | randomblame | lg incite |
| 09:45.33 | phh | ah. |
| 09:45.36 | phh | you know which amss is that ? |
| 09:45.47 | randomblame | amss? |
| 09:45.54 | phh | radio protocol |
| 09:46.08 | randomblame | gsm |
| 09:46.11 | phh | ....;; |
| 09:46.37 | phh | at winmo's startup you have in bottom right three numbers |
| 09:46.39 | phh | it's the 2° one |
| 09:47.45 | randomblame | at winmos startup I have no numbers in bottom right :D |
| 09:48.01 | phh | bah ... |
| 09:48.10 | phh | randomblame: it's a copy of which device ? |
| 09:48.59 | randomblame | I'm not understanding |
| 09:49.07 | phh | it's not an htc-made device ? |
| 09:49.17 | randomblame | no |
| 09:49.22 | phh | ah. |
| 09:49.27 | randomblame | thats why I said lg incite |
| 09:49.28 | randomblame | lg |
| 09:49.34 | phh | yers and ? |
| 09:49.38 | phh | that doesn't mean lg made it /o\ |
| 09:49.41 | randomblame | lol |
| 09:49.44 | randomblame | true enough |
| 09:50.05 | randomblame | I have one apart in front of me, maybe I should compare it to some naked htc pictures |
| 09:50.15 | phh | AFAIK radio are different per constructor (even for the same constructor...), so you'll have to RE it by yourself ... |
| 09:50.33 | randomblame | I'm not too worried about the radio |
| 09:50.42 | phh | lol |
| 09:50.44 | phh | radio does everything |
| 09:51.03 | randomblame | you mean the arm9 that's built into the msm7201a |
| 09:51.07 | phh | yes |
| 09:51.26 | randomblame | that should be set up just like every other arm9 in every other msm7200a and 7201a |
| 09:51.38 | phh | lol |
| 09:51.41 | phh | they are all different |
| 09:51.47 | phh | and I know only HTC devices |
| 09:52.58 | randomblame | it detects that it has crashed as soon as smd is enabled which panics the kernel |
| 09:53.06 | randomblame | pain in the ass |
| 09:53.13 | phh | randomblame: smd is radio-dependant. |
| 09:53.20 | randomblame | I know |
| 09:53.27 | randomblame | smd is communication to the arm9 |
| 09:53.33 | phh | if it's not an HTC AMSS, you have to RE smd by yourself |
| 09:53.38 | phh | is *one* way to communicate with arm9 |
| 09:53.42 | phh | there are several others. |
| 09:53.47 | randomblame | reallllly |
| 09:53.54 | phh | cr2: so for camera, how are muxed the gpio ? |
| 09:54.02 | randomblame | could I use one of the others to find the reason it crashes? |
| 09:54.15 | phh | no, ARM9 never says why it crashes. |
| 09:54.31 | randomblame | arm9 is a whore |
| 09:54.38 | phh | randomblame: when has this device been released ? |
| 09:54.51 | phh | cr2: meaning what do I have to do ? :D |
| 09:54.53 | randomblame | a year or two ago |
| 09:55.02 | phh | randomblame: it's a 7201a ? |
| 09:55.08 | randomblame | yes sir |
| 09:55.18 | phh | try 5200 AMSS then |
| 09:56.14 | randomblame | where do I set that |
| 09:56.21 | randomblame | defconfig |
| 09:57.09 | randomblame | yes |
| 09:57.18 | *** join/#htc-linux onen|openBmap (n=quassel@vbo91-1-89-87-201-85.dsl.club-internet.fr) |
| 09:58.01 | phh | randomblame: which mtype are you using ? |
| 09:58.09 | phh | what memory layout do you have ? |
| 09:58.43 | randomblame | amss is set to 5200 already in my defconfig |
| 09:58.50 | randomblame | mtype is diamond lol |
| 09:58.51 | phh | it's not in defconfig |
| 09:59.05 | phh | randomblame: you have 64MB SMI ?!? |
| 09:59.48 | randomblame | smi? |
| 10:00.01 | randomblame | captain hooks assistant? |
| 10:00.01 | phh | pff you seriously know nothing about your board ... |
| 10:00.06 | randomblame | lol |
| 10:00.09 | randomblame | 76mb |
| 10:00.16 | phh | what 76mb ? |
| 10:00.27 | randomblame | usable memory |
| 10:00.28 | phh | randomblame: how much memory LG assert your device has ? |
| 10:00.55 | randomblame | they assert it is 128mb |
| 10:00.57 | randomblame | they lie |
| 10:01.23 | phh | why do you say that ? |
| 10:01.45 | phh | can you dump 0x0 physical address with haret ? |
| 10:02.19 | randomblame | sure |
| 10:02.29 | phh | I mean, does it work |
| 10:02.31 | phh | I don't care about the result |
| 10:02.41 | randomblame | let me check |
| 10:04.01 | *** join/#htc-linux mrmoku|away (n=mrmoku@ppp-188-174-7-126.dynamic.mnet-online.de) |
| 10:05.09 | randomblame | yes |
| 10:05.15 | phh | ok so it's not 128MB |
| 10:05.35 | phh | can you dump 0x02000000 ? |
| 10:05.46 | phh | and 0x20000000 ? |
| 10:08.39 | phh | if you can for both, then read the dump files and see wether it's random data or 0x00/0xff |
| 10:10.58 | phh | randomblame: ? |
| 10:11.13 | randomblame | sorry my son was yelling at me |
| 10:11.20 | phh | yell looder |
| 10:11.35 | randomblame | lol he's 5months old he wins the loudness contest |
| 10:11.39 | phh | ah. |
| 10:11.42 | *** join/#htc-linux Bry8Star_ (n=Bry8Star@adsl-99-182-76-24.dsl.lsan03.sbcglobal.net) |
| 10:13.29 | randomblame | omg ok to dump a memory location with haret console you type? |
| 10:13.39 | randomblame | because I'm not happy with these odd results |
| 10:14.19 | phh | what is the result ? |
| 10:14.21 | phh | short write ? |
| 10:14.58 | phh | the command is something like pd address size iirc |
| 10:15.22 | phh | well I use pwf |
| 10:16.33 | randomblame | short write for 0x02000000 |
| 10:17.13 | phh | and what about the second one ? |
| 10:17.49 | randomblame | the second does nothing |
| 10:17.58 | phh | ? |
| 10:18.04 | phh | what does that mean ? |
| 10:18.04 | randomblame | seems to have crashed haret console |
| 10:18.11 | phh | uh ? |
| 10:18.22 | randomblame | or it's still writing |
| 10:18.42 | phh | which size did you use ? |
| 10:18.47 | randomblame | 10000 |
| 10:19.18 | phh | it should be instant // |
| 10:20.07 | randomblame | I see the phone crashed |
| 10:20.08 | phh | maybe try with 0x20010000 |
| 10:20.28 | phh | anyway it sounds like you have 128MB EBI+32MB SMI |
| 10:21.53 | phh | so diamond's pmem won't work |
| 10:22.56 | phh | randomblame: I don't remember how linuxtogo's tree was, but if you use gitorious, edit pmem.c and force the tuse of 128EBI+32SMI layout |
| 10:23.07 | randomblame | ok |
| 10:23.13 | randomblame | I'm looking through it now |
| 10:24.30 | *** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl) |
| 10:29.16 | phh | cr2: wake up ! :p |
| 10:31.06 | chamonix | hey guys |
| 10:31.07 | phh | randomblame: don't forget to edit the board file to the corresponding available ram |
| 10:31.09 | phh | hey chamonix |
| 10:31.18 | chamonix | hey phh |
| 10:31.32 | phh | chamonix: if you want camera, help me waking up cr2 :p |
| 10:31.38 | chamonix | lol |
| 10:31.46 | chamonix | shakes cr2 |
| 10:32.39 | randomblame | wow htc fb console is much different on this branch |
| 10:32.47 | chamonix | phh: if you look into locale you will love ASE, it's a scripting platform |
| 10:32.52 | phh | randomblame: way cleaner you mean. |
| 10:32.59 | phh | chamonix: uh ? |
| 10:33.03 | phh | another day maybe :p |
| 10:33.29 | chamonix | locale is the app that will allow you to control e.g. your profiles depending on loc/time/name-it |
| 10:33.34 | phh | I know |
| 10:33.39 | phh | is it yours ? |
| 10:34.06 | chamonix | nah, I'm writing a tool like that but it's not that advanced yet |
| 10:34.09 | phh | ok |
| 10:34.47 | chamonix | but now that loc works I've a motivation to keep working ;) |
| 10:34.55 | phh | :) |
| 10:35.03 | phh | I'm still playing with GPS. |
| 10:35.08 | chamonix | diving into pm this week tho |
| 10:35.18 | *** join/#htc-linux ICEMANno1 (n=ICEMANno@ip-109-90-220-39.unitymediagroup.de) |
| 10:35.55 | chamonix | I remember you saying something about WLAN being always powered at a minimum... is that correct and how can I avoid that? |
| 10:36.14 | phh | no longer true |
| 10:36.24 | chamonix | ok |
| 10:36.24 | phh | well unless you start it |
| 10:36.32 | chamonix | sure |
| 10:36.35 | phh | (meaning once it's activated you can't disactivate it) |
| 10:37.21 | chamonix | ok, so I don't have to bother about that for now.... I'm trying to get hands on whatever part I'll be looking at |
| 10:37.36 | phh | I have no clue what eats battery now |
| 10:37.54 | chamonix | lol, me neither, that's y I want to scratch the surface |
| 10:37.57 | *** join/#htc-linux kvaster (n=kvaster@93.84.112.82) |
| 10:38.26 | phh | I'd bet there is just one little thing eating the greatest part |
| 10:38.28 | phh | but I can't see what |
| 10:38.36 | phh | maybe microp ? |
| 10:38.43 | phh | or navi pad |
| 10:39.14 | phh | hum it won't consumes 75mA |
| 10:39.35 | phh | (75mA is a hand-made approximation of the consumption considering battery life.) |
| 10:41.02 | randomblame | this is all much cleaner |
| 10:41.04 | randomblame | I like it |
| 10:41.04 | chamonix | well I'll try approaching that from logs... looking at how long phone sleeps, what waked it up etc. |
| 10:41.18 | phh | randomblame: that's the idea. |
| 10:41.28 | phh | chamonix: it's not arm11 the problem now |
| 10:41.41 | phh | arm11 is awake <2minutes/10hours |
| 10:42.59 | chamonix | well I want to be sure about that.... it may be my settings/software but I've seen many "sleeping for" < 200s in my last log and I feel that's not optimal |
| 10:43.15 | phh | even if it's not optimal it's not the greatest drain |
| 10:43.28 | phh | again we have > 50mA of drain |
| 10:43.52 | *** join/#htc-linux randomblame (n=kevin@c-71-59-184-96.hsd1.wa.comcast.net) |
| 10:43.59 | randomblame | virtual box is evil |
| 10:47.59 | *** join/#htc-linux emwe (n=michael@86.56.50.180) |
| 10:48.39 | phh | chamonix: any clue why multi-bytes characters doesn't work in SMSes ? |
| 10:52.58 | chamonix | ummmm, no except it's maybe restricted to utf-8 |
| 10:53.22 | *** join/#htc-linux c2d (n=Mike@port-92-200-10-104.dynamic.qsc.de) |
| 10:53.28 | *** part/#htc-linux c2d (n=Mike@port-92-200-10-104.dynamic.qsc.de) |
| 10:54.27 | chamonix | Short messages can be encoded using a variety of alphabets: the default GSM 7-bit alphabet (see GSM 03.38 for details), the 8-bit data alphabet, and the 16-bit UTF-16 alphabet.[29] Depending on which alphabet the subscriber has configured in the handset, this leads to the maximum individual Short Message sizes of 160 7-bit characters, 140 8-bit characters, or 70 16-bit characters (including spaces) |
| 10:55.11 | chamonix | the question is what alphabet do we have |
| 10:57.34 | chamonix | I guess the default would be 7-bit: http://en.wikipedia.org/wiki/GSM_03.38 |
| 10:57.44 | phh | é works but not à for example. |
| 10:57.48 | phh | so I don't think it's 7-bit |
| 10:58.02 | phh | oh well it's maybe not ascii |
| 10:58.27 | phh | ok fine it's 7bit. |
| 10:59.13 | chamonix | so we'll have to find out if it can be changed or if the java app sending SMS isn't just not taking care of proper encoding |
| 10:59.23 | phh | gps is going to work. gps is going to work. gps is going to work. gps is going to work. |
| 10:59.28 | phh | chamonix: I don't think android devices has such a big problem :p |
| 11:00.20 | chamonix | well, it may just be a matter of settings, e.g. if java uses utf-16 and whatever-below doesn't expect that |
| 11:00.44 | phh | I told leobaillard to try with htc's ril |
| 11:00.57 | chamonix | RIL logs whetever goes to the wire right? |
| 11:01.09 | phh | uh ? |
| 11:01.22 | phh | oh |
| 11:01.28 | phh | wire = /dev/smd 0? |
| 11:01.34 | chamonix | yes |
| 11:01.37 | chamonix | :) |
| 11:01.43 | phh | our ril does that yes |
| 11:01.46 | phh | not htc's one |
| 11:01.55 | phh | you have to strace it if you want his commands |
| 11:02.24 | chamonix | well looking at ours and the ref there may be some clues... I mean I'm sure android guys don't think they only have us keyboards |
| 11:02.32 | phh | I have a dream, that one day ARM9 stops crashing. |
| 11:02.38 | chamonix | that reminds me of the times of novell 3.1 |
| 11:03.54 | phh | chamonix: maybe a huge problem is SD card :p |
| 11:04.09 | phh | (aka, what about trying to write android on nand ?) |
| 11:04.40 | phh | \o/ |
| 11:04.48 | phh | ARM11 sent a pdsm_get_position \o/ |
| 11:04.55 | phh | and arm9 crashed, obviously |
| 11:05.25 | chamonix | wow |
| 11:06.08 | chamonix | bbl, I have to get rid of the snow outside |
| 11:06.34 | phh | chamonix: haha, no snow at all here :p |
| 11:06.40 | phh | cr2: http://husson.hd.free.fr/dm.7 |
| 11:08.40 | phh | ARM9 crashes as soon as ARM11 ack's ARM9's 0xb procedure callback |
| 11:08.54 | phh | (I wonder if my position is in my dump. I hope not.) |
| 11:09.43 | phh | oh I know why it crashes. still this stupid offset |
| 11:09.56 | phh | it acks 0xa, while ARM9 sent 0xb :D |
| 11:14.57 | phh | hum no that's not the problem |
| 11:16.20 | *** join/#htc-linux bakkie (n=bakkie@84-106-91-26.cable.quicknet.nl) |
| 11:23.30 | cr2 | phh: i've added the CB functions from my raph100 to wiki |
| 11:23.39 | phh | ok |
| 11:24.03 | phh | uh, where ? |
| 11:24.15 | cr2 | phh: i don't understand 1 thing. why do we never get any details about arm9 crash ? |
| 11:24.33 | phh | cr2: maybe there are two ARM9 modes |
| 11:24.39 | cr2 | is the crash parser correct ? |
| 11:25.00 | cr2 | the gpsdriver has the "modem" reboot handler |
| 11:25.25 | phh | ah. |
| 11:25.33 | cr2 | but our arm11 kernel makes an oops when arm9 is unavailable |
| 11:25.42 | cr2 | even if it's temporary |
| 11:26.35 | cr2 | maybe we should disassemble amss after all |
| 11:26.57 | cr2 | and to check the crash log parser too |
| 11:27.12 | cr2 | i'll be back in 20min |
| 11:28.16 | phh | hum when ARM9 restarts we send a RESET_ARM9 command ? |
| 11:28.57 | phh | that's not what trout do |
| 11:29.17 | phh | trout only takes gpio25 and set it to 0 |
| 11:29.22 | cr2 | yes, need to fix the halt/reboot/softreboot handler |
| 11:30.05 | cr2 | this one maybe the most urgent thing to do. |
| 11:30.27 | phh | I try to see what's the last reply ARM11 sends |
| 11:30.38 | cr2 | ok. bbl |
| 11:33.39 | *** join/#htc-linux GlemSom (n=glemsom@0x5da34bca.cpe.ge-1-1-0-1105.sdnqu1.customer.tele.dk) |
| 11:34.06 | phh | hi GlemSom |
| 11:34.18 | GlemSom | hi phh :) |
| 11:38.42 | GlemSom | Seems the commitlog isn't quite uptodate at http://gitorious.org/linux-on-qualcomm-s-msm/linux-msm ? (Or am I missing something?) |
| 11:38.48 | phh | yes |
| 11:38.50 | phh | I don't know why |
| 11:42.01 | GlemSom | Other projects at gitorious.org have commits today... So, guess it's isolated to our "linux-on-qualcomm-s-msm" project? |
| 11:43.33 | phh | GlemSom: argh |
| 11:43.47 | leobaillard | don't tell me the git is broken again :'( |
| 11:43.50 | leobaillard | taht would be sad |
| 11:44.03 | phh | it's only the GUI, I don't really care about it :p |
| 11:46.49 | emwe | GlemSom: at least linux-msm-topaz suffers the same ;) |
| 11:47.02 | *** join/#htc-linux bakkie (n=bakkie@84-106-91-26.cable.quicknet.nl) |
| 11:56.05 | *** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl) |
| 11:56.47 | *** join/#htc-linux m3dlg (n=m3dlg@bb-87-81-252-83.ukonline.co.uk) |
| 11:59.18 | NetRipper | AstainHellbring, am now |
| 12:01.02 | *** join/#htc-linux GNUtoo (n=GNUtoo@host82-133-dynamic.54-79-r.retail.telecomitalia.it) |
| 12:07.01 | cr2 | NetRipper: haret patches ? |
| 12:07.22 | cr2 | phh: the reset modes |
| 12:07.44 | phh | oh the restart_reason stuff ? |
| 12:08.10 | *** join/#htc-linux kvaster_ (n=kvaster@live.bn.by) |
| 12:09.02 | phh | hum maybe not |
| 12:10.06 | cr2 | phh: i think we have 2 no |
| 12:10.14 | cr2 | s/no/now/ |
| 12:10.25 | phh | 2 what ? |
| 12:10.38 | cr2 | 2 reset modes |
| 12:11.00 | phh | I can't see anything about that :/ |
| 12:11.05 | cr2 | hm. my big lcd eats 240mA |
| 12:11.20 | cr2 | phh: in the board file and in pm.c |
| 12:11.46 | phh | msm_pm_restart isn't linked with that, is it ? |
| 12:11.52 | MrPippy | phh: i tried the vbus patch, but i'm not sure that it made any difference |
| 12:12.14 | MrPippy | plugging and unplugging usb (for adb) did work, but the charger status in android didn't change when the cable was out |
| 12:12.27 | phh | MrPippy: boot with usb connected, dmesg -c, unplug usb, wait >10seconds, replug, see what dmesg says |
| 12:12.31 | cr2 | MrPippy: charger is not vbus |
| 12:12.48 | MrPippy | vbus is just cable connected to usb? |
| 12:12.54 | phh | cr2: afaik there is no way to tell android there is vbus but no charger :/ |
| 12:12.58 | cr2 | MrPippy: you can disable/enable it independently from vbus |
| 12:13.25 | cr2 | vbus is external power availability on the usb +5V pin |
| 12:13.45 | cr2 | phh: why should android care about vbus ? |
| 12:14.03 | phh | cr2: show when adb is activated |
| 12:14.12 | phh | and mass storage |
| 12:14.54 | cr2 | it's driver business, not androids |
| 12:16.15 | *** join/#htc-linux pH5 (n=ph5@e178235013.adsl.alicedsl.de) |
| 12:17.32 | phh | makkonen: have you tried calibration ? |
| 12:18.09 | phh | cr2: I can't see what is supposed to happen when ARM9 reboots if not crashing :/ |
| 12:18.55 | cr2 | phh: 1 reset: put 0x55aa00ff to +fc08c and gpio25=0 |
| 12:19.08 | cr2 | this is full shutdown |
| 12:19.41 | cr2 | gpio25 needs to be assigned to arm11 first, and then configured as output |
| 12:20.05 | phh | ok |
| 12:21.15 | phh | but all restart/reset commands ends with a for(;;); so the kernel can't get out.. I'm wrong ? |
| 12:21.52 | cr2 | then there is soft reset |
| 12:22.40 | *** join/#htc-linux rzk_ (n=rzk@daemonet.ru) |
| 12:23.48 | phh | it is possible to kill this forever loop without killing the kernel ? |
| 12:23.56 | cr2 | check if dex is ready (fc13c=1), send dex_0x8e |
| 12:24.03 | cr2 | aka notify_a9_reboot |
| 12:24.32 | randomblame | cool beans |
| 12:25.05 | randomblame | I applied the changes to that new branch and it is working just the same, right up to smd init |
| 12:25.11 | cr2 | phh: hmm. |
| 12:25.13 | randomblame | but this should be easier to work with |
| 12:26.29 | phh | cr2: so I notify arm9 reset but I don't pull down the 25 gpio ? |
| 12:27.25 | cr2 | in the soft case yes |
| 12:27.27 | phh | cr2: dex ready is fc13c ?!? according to source code I'd say fc104 |
| 12:28.17 | cr2 | +0xfc104 PC_STATUS |
| 12:28.22 | phh | ok |
| 12:28.33 | cr2 | 0xfc13c DEX ready=1 |
| 12:28.50 | cr2 | wiki is your friend :) |
| 12:28.57 | phh | I prefer source code :p |
| 12:29.07 | cr2 | i'ts more buggy |
| 12:29.21 | phh | ok |
| 12:29.35 | phh | so in htcraphael_reset I do: |
| 12:29.40 | phh | <PROTECTED> |
| 12:29.40 | phh | <PROTECTED> |
| 12:29.40 | phh | <PROTECTED> |
| 12:29.40 | phh | <PROTECTED> |
| 12:29.41 | phh | and that's it ? |
| 12:29.46 | cr2 | PC_STATUS is command sent ready |
| 12:29.53 | phh | ok |
| 12:29.58 | cr2 | fc13c is dex ready in general. |
| 12:30.20 | cr2 | if dex is not ready, it does not make sense to send anything with dex. |
| 12:30.44 | cr2 | yes, it looks like softreset |
| 12:31.12 | phh | but this function is only used in: |
| 12:31.14 | phh | <PROTECTED> |
| 12:31.14 | phh | <PROTECTED> |
| 12:31.14 | phh | <PROTECTED> |
| 12:31.14 | phh | <PROTECTED> |
| 12:31.14 | phh | <PROTECTED> |
| 12:31.17 | phh | <PROTECTED> |
| 12:31.21 | phh | so it won't ever exit ?!? |
| 12:31.27 | cr2 | wince also does storing "CQRG" and "REST" in smem |
| 12:31.46 | cr2 | phh: it's not hard reboot here. |
| 12:31.52 | cr2 | only softreset |
| 12:32.17 | phh | so I should bypass the for(;;); ? |
| 12:32.18 | cr2 | the hard reset is [13:18] <cr2> phh: 1 reset: put 0x55aa00ff to +fc08c and gpio25=0 |
| 12:32.33 | cr2 | wrong function i guess |
| 12:32.40 | cr2 | in the current source |
| 12:32.52 | phh | hum it's check_for_modem_crash |
| 12:32.54 | cr2 | phh: what does pm.c have ? |
| 12:33.21 | cr2 | so we have hard reset, and soft reset functions. |
| 12:33.33 | cr2 | afair there was one more. maybe done in pm.c |
| 12:33.45 | phh | there is msm_pm_restart, which also does for(;;); |
| 12:33.48 | cr2 | some pseudo-dex call |
| 12:35.08 | *** join/#htc-linux luminoso (n=lumos@av-217-129-128-10.netvisao.pt) |
| 12:35.55 | phh | [13:24:35] <cr2> phh: 1 reset: put 0x55aa00ff to +fc08c and gpio25=0 <----- I have to wiritel(cpu_to_be32(0x55aa00ff)) or writel(0x55aa00ff); ? |
| 12:36.37 | cr2 | phh: be32 only for (onc)rpc |
| 12:36.41 | phh | ok |
| 12:37.32 | cr2 | phh: the nk.exe waits for dex_ready in a loop, and does softreset |
| 12:37.49 | cr2 | funny |
| 12:37.59 | phh | nk.exe ? |
| 12:38.05 | cr2 | because softreset needs dex ready itself. |
| 12:38.08 | cr2 | wince kernel |
| 12:38.10 | NetRipper | cr2, which patches? |
| 12:38.29 | phh | cr2: so when it detects ARM9 resets, I just add a loop to wait for dex_ready, and then just return ? |
| 12:38.31 | cr2 | NetRipper: bank5 tracing is missing on 7201A |
| 12:38.49 | cr2 | NetRipper: i'd like to implement more functions though |
| 12:38.55 | NetRipper | aha |
| 12:38.57 | NetRipper | like what? |
| 12:39.08 | cr2 | NetRipper: like the rpc get_khz call |
| 12:39.15 | NetRipper | make a patch for bank5 ;) |
| 12:39.18 | cr2 | did we have a source for that ? |
| 12:39.44 | cr2 | NetRipper: i don't have git, only cvs tree. but i think it'll be ok |
| 12:40.01 | NetRipper | checkout the git tree then ;) |
| 12:40.17 | cr2 | 0x64 =100, 0xc8=200 ? |
| 12:40.25 | cr2 | NetRipper: link ? |
| 12:40.41 | NetRipper | http://git.linuxtogo.org/?p=groups/haret/haret.git;a=summary |
| 12:41.03 | NetRipper | git clone git://git.linuxtogo.org/home/groups/haret/haret.git |
| 12:41.27 | cr2 | yeah. for a 1 dev project cvs is more than enough ;) |
| 12:41.52 | NetRipper | the svn repository is going to be obsolete |
| 12:41.53 | cr2 | - " Plays chord.wav") |
| 12:41.54 | cr2 | + " Plays chord.wav and Default.wav") |
| 12:41.58 | cr2 | err |
| 12:42.21 | cr2 | NetRipper: playsound does not have a filename as its argument ? |
| 12:42.40 | NetRipper | eh? |
| 12:42.42 | NetRipper | what file? |
| 12:42.54 | phh | cr2: it means no I guess :p |
| 12:43.17 | cr2 | diff -u -r1.16 wincmds.cpp |
| 12:43.18 | NetRipper | "Plays" is not part of filename anyway :p |
| 12:43.33 | cr2 | the names are hardcoded |
| 12:43.47 | cr2 | <PROTECTED> |
| 12:43.49 | cr2 | <PROTECTED> |
| 12:43.50 | cr2 | - " Plays chord.wav") |
| 12:43.51 | NetRipper | ah like that |
| 12:44.06 | *** join/#htc-linux MethoS- (n=clemens@134.102.106.250) |
| 12:44.11 | NetRipper | then you answered your own question :P |
| 12:44.18 | cr2 | obviously there was some device without chord.wav ;) |
| 12:44.20 | NetRipper | but i suppose you want to add support for a filename |
| 12:44.29 | cr2 | yeah |
| 12:46.32 | cr2 | i'll write VIDEOLFB support after your patch |
| 12:46.59 | NetRipper | what do you mean with my patch? :) |
| 12:47.49 | cr2 | NetRipper: https://privatepaste.com/8b0cb760d4 |
| 12:47.52 | phh | cr2: am I supposed to do something to get DEX up again ? |
| 12:48.17 | cr2 | i'll checkout the git after this one |
| 12:48.30 | cr2 | phh: wince loops 200 times for 100ms |
| 12:48.40 | cr2 | phh: loking what it does after that |
| 12:49.30 | phh | I haven't had a working DEX command for more than 40seconds |
| 12:50.01 | cr2 | phh: with DEX ready=0 ? |
| 12:50.47 | phh | theorically it should be 1 |
| 12:51.01 | cr2 | haha. phh, it printk's DEX not ready, and goes on :) |
| 12:51.19 | phh | [ 412.178894] msm_proc_comm_wince: DEX cmd timed out. status=0x0, A2Mcntr=29, M2Acntr=21 |
| 12:51.21 | cr2 | if it's =1, send the dex command |
| 12:51.23 | phh | that's what it printks |
| 12:51.29 | phh | <PROTECTED> |
| 12:51.29 | phh | <PROTECTED> |
| 12:51.29 | phh | <PROTECTED> |
| 12:51.40 | phh | that's what I do when I get the ARM9 reset |
| 12:51.49 | cr2 | is it for hard rebooot ? |
| 12:52.22 | phh | bah I put it whenever I get ARM9 crashed |
| 12:52.28 | phh | I don't check if it's hard or soft reboot yet. |
| 12:52.31 | cr2 | the PCOM_ prefix is misleading btw. need to fix that too. |
| 12:52.32 | phh | so it acts like soft |
| 12:52.37 | *** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net) |
| 12:52.53 | phh | hum |
| 12:52.56 | phh | it sent a pdsm_end_position |
| 12:53.23 | phh | when it set screen off |
| 12:55.07 | phh | cr2: am I supposed to call msm_proc_comm_wince_init after arm9 reboot ?(writing 0 to DEX registers) |
| 12:55.09 | cr2 | phh: wince dex driver supports wat queue, btw |
| 12:55.33 | cr2 | phh: let me find the 3rd reset mode |
| 12:55.49 | phh | maybe it's the good reset mode, but we are missing something ? |
| 12:56.00 | phh | arm9 is alive, RPC works |
| 12:56.55 | phh | it's just that DEX isn't correctly reinited |
| 13:00.15 | cr2 | NetRipper: do we include gpio owner regs in haret ? |
| 13:00.57 | cr2 | phh: can't find yet |
| 13:01.22 | cr2 | x8e notify arm9 reboot |
| 13:01.28 | cr2 | it's softreset |
| 13:01.50 | phh | ok but when do we send that ? |
| 13:01.55 | cr2 | but there is also this one |
| 13:01.58 | cr2 | 0x14 shutdown by arm9, power suspend |
| 13:02.12 | cr2 | need to find this one now |
| 13:02.46 | cr2 | if dex is not working for too long ? |
| 13:03.52 | phh | http://husson.hd.free.fr/dm.8 but I don't think it's of any help. |
| 13:04.16 | cr2 | M2A_NOTIFY_ARM9_REQ_RESTART is a DEX irq which we don't do afaik |
| 13:04.29 | phh | and we are supposed to do it ? |
| 13:04.59 | randomblame | phh where in pmem.c are the fb base address, offset, size, etc |
| 13:05.06 | randomblame | I don't see |
| 13:05.06 | phh | oh maybe we are supposed to wait for ARM9 alive again |
| 13:05.19 | cr2 | wait. i think i've told something wrong |
| 13:05.19 | phh | randomblame: they are in CALC_PMEM |
| 13:06.14 | randomblame | thanks, seems it doesn't directly translate |
| 13:06.40 | cr2 | phh: softreset sends dex 8e if dex is ready, and does gpio25=0 if not. |
| 13:07.05 | phh | ok |
| 13:07.13 | cr2 | and loops on gpio25=0 |
| 13:07.16 | *** join/#htc-linux dcordes (n=dcordes_@unaffiliated/dcordes) |
| 13:07.24 | dcordes | hi |
| 13:07.33 | phh | cr2: loops on it and checks dex ready ? |
| 13:07.35 | phh | or just loops ? |
| 13:07.51 | cr2 | looks like endless loop |
| 13:07.54 | phh | ok |
| 13:08.13 | phh | cr2: ok so dex not ready after softreset means softreset failed then. |
| 13:08.22 | cr2 | ah |
| 13:08.56 | cr2 | you can't get out of this loop ? |
| 13:09.07 | cr2 | only on some irq |
| 13:09.10 | cr2 | ? |
| 13:09.14 | phh | which loop ? |
| 13:09.42 | phh | dex timeout ? |
| 13:09.43 | IceBone | What's a good program to give android 2 more homescreens? |
| 13:09.50 | IceBone | I don't need any extra features. |
| 13:10.41 | IceBone | Just tried HomeExtended and it just keeps crashing. |
| 13:12.10 | cr2 | looks like buggy code to me |
| 13:12.12 | phh | cr2: I tried setting dex registers to 0, it's not better |
| 13:13.23 | cr2 | you set gpio25=0 |
| 13:13.32 | cr2 | if dex is not ready |
| 13:13.40 | phh | dex is ready ... |
| 13:14.01 | cr2 | if |
| 13:14.16 | phh | well, I have to check for dex ready before or after the NOTIFY_ARM9_REBOOT ? |
| 13:14.16 | cr2 | how will you get out of this softreset ? |
| 13:14.20 | phh | I'd assume before but well ... |
| 13:14.26 | cr2 | before |
| 13:14.54 | cr2 | i think we need to track the dex irqs |
| 13:15.08 | phh | when it sees the modem reset, it check dex ready , send the notify_arm9_reboot, sleeps 300ms, and just continue |
| 13:15.42 | cr2 | +0xfc138 irq |
| 13:15.53 | cr2 | print this register too, on all debug printks |
| 13:16.50 | cr2 | 300ms ? |
| 13:17.10 | phh | that's what is in htcraphael_reset |
| 13:21.17 | cr2 | static void msm_pm_power_off(void) |
| 13:21.20 | cr2 | in pm.c |
| 13:22.36 | phh | what about it? |
| 13:22.49 | cr2 | should be hard reset imho |
| 13:24.51 | cr2 | static void htcraphael_reset(void) |
| 13:25.07 | cr2 | is softreset |
| 13:25.52 | phh | hu ? |
| 13:26.04 | phh | so I have to do pcom_reset_arm9 instead of pcom_notify_reset_arm9 ? |
| 13:27.52 | phh | cr2: dex irq is always 0 |
| 13:28.29 | cr2 | ok, so you don't have dex irq pending |
| 13:28.41 | cr2 | what is gpio 0x24 ? |
| 13:29.07 | cr2 | 6 0x24 3,0 20 I reset button, irq. |
| 13:29.17 | cr2 | which button is that ? |
| 13:29.27 | *** join/#htc-linux rzk__ (n=rzk@188.162.48.126) |
| 13:29.31 | phh | red little thing under stylus ? |
| 13:30.01 | cr2 | 0x21 33 INT_GPIO_GROUP2 |
| 13:30.04 | cr2 | red ? |
| 13:30.19 | phh | bah the soft reboot button |
| 13:31.10 | cr2 | do we have a handler for it ? |
| 13:31.43 | phh | we can't handle it |
| 13:32.03 | cr2 | wince has, and sends 'notify a9 to reboot' + dex 8e |
| 13:32.12 | cr2 | why ? |
| 13:32.27 | phh | ok so it's not this button I guess. |
| 13:32.49 | phh | this button reboots the device even if the OS is crashed ... |
| 13:32.53 | phh | especially if it's crashed. |
| 13:33.04 | cr2 | strange |
| 13:33.13 | phh | but it's maybe not this button then |
| 13:33.34 | phh | but that's the only "reset button" I can think of |
| 13:34.15 | *** join/#htc-linux Markinus (n=Miranda@gtng-4db0412a.pool.mediaWays.net) |
| 13:36.25 | cr2 | MBAT_IN irq sets 2c=1 |
| 13:36.48 | cr2 | 4 0x2c 0 1 O battery charging (0=Enable, 1=Disable) |
| 13:37.58 | cr2 | so if you pull out the battery, charging is switched off :) |
| 13:38.08 | phh | rho why ? :p |
| 13:38.14 | phh | that'd be fun. |
| 13:38.59 | Markinus | cr2: Hi! I tested the mmtrace from ADSP5 register location when i playing sound but I can see nothing .. The only thing what is comming is IRQ 23 |
| 13:39.36 | tmzt | when starting playback no writes? |
| 13:39.42 | cr2 | Markinus: yes, i have the same on raph100 |
| 13:40.00 | cr2 | Markinus: so it's difficult to check ;) |
| 13:41.31 | Markinus | yes :-( . did you have a idea more what I could check why this doesn't work? |
| 13:42.15 | cr2 | Markinus: no |
| 13:42.30 | phh | Markinus: your changes in gps-rfkill.c with amss_get_num_value are wrong btw |
| 13:42.41 | cr2 | phh: let's check the crash handler again |
| 13:43.02 | phh | you do pdapi_rpc_server.prog = amss_get_num_value(RPC_PDAPI_CB_PROG); while RPC_PDAPI_CB_PROG has been defined in the same file as #define RPC_PDSM_ATL_PROG 0x3000001d |
| 13:43.08 | Markinus | cr2: ohh, I will look on it .. |
| 13:43.56 | cr2 | phh: i thing PROG values do not depend on amss |
| 13:44.05 | Markinus | hmm, I forgot to remove it. . . |
| 13:44.09 | phh | cr2: I think too actually |
| 13:44.13 | phh | Markinus: and to add it to amss thing :p |
| 13:44.26 | Markinus | ;-) okay |
| 13:44.47 | phh | Markinus: you'd better revert this change, as cr2 say, it doesn't depend on amss at all |
| 13:45.27 | cr2 | Markinus: you may strsep(,".") the amss version from smem g23 sting in the board file at boot |
| 13:45.53 | phh | cr2: see proc_comm_wince.c to know how to get amss string id |
| 13:46.19 | cr2 | Markinus: then you will get rid on 6120 in favor of 6125 everywhere |
| 13:46.28 | cr2 | phh: in the current code ? |
| 13:46.35 | phh | yes |
| 13:47.04 | phh | err it was meant for Markinus not cr2 |
| 13:47.14 | phh | cr2: you already know how to get amss id :p |
| 13:47.20 | Markinus | ok, I will look on it .. |
| 13:47.36 | cr2 | printk(KERN_INFO "AMSS version: %s\n", amss_ver); |
| 13:48.02 | cr2 | the first and third field together are amss version |
| 13:48.10 | tmzt | 6120 to 6125 is a small layout change |
| 13:48.22 | cr2 | but it will break our cdma friends though. with their 6150 ;) |
| 13:48.34 | cr2 | tmzt: 6120 does not exist |
| 13:48.44 | cr2 | tmzt: at least i've not seen it. |
| 13:48.52 | tmzt | it did |
| 13:49.03 | tmzt | there's a not in android's git about it |
| 13:49.06 | cr2 | tmzt: where ? |
| 13:49.17 | Markinus | dou you mean to replace this machine specific code with AMSS specific? |
| 13:49.24 | tmzt | hmm, have to look |
| 13:50.06 | tmzt | it might have never shipped |
| 13:50.07 | cr2 | Markinus: you can detect the amss version lready in the board file, and put it into some global varialble i think. |
| 13:50.15 | tmzt | so for our purposes it doesn't exist |
| 13:51.00 | cr2 | tmzt: yes. once it will pop out in the real world (tm) we may add it. |
| 13:51.26 | tmzt | it's been replaced, but support did exist in git for testing |
| 13:51.47 | cr2 | Markinus: btw, you have some strange gpios in *topaz.h |
| 13:51.47 | tmzt | anyway |
| 13:52.02 | cr2 | tmzt: 6220 ? |
| 13:52.08 | tmzt | has fixing haret helped with gpios? |
| 13:52.23 | tmzt | what's 6220? |
| 13:52.33 | cr2 | tmzt: i was spammed by ulpi data immediately ;) |
| 13:52.45 | phh | :) |
| 13:53.36 | cr2 | [ 245.141754] ARM9 has CRASHED |
| 13:53.38 | cr2 | [ 245.141876] smem: DIAG '' |
| 13:53.55 | cr2 | phh: i'm confused that the diag is always empty. |
| 13:54.15 | cr2 | something is wrong here. |
| 13:54.46 | phh | maybe it's some security "featuer" ? |
| 13:55.47 | tmzt | is the qcdiag? |
| 13:55.49 | *** join/#htc-linux Markinus (n=markus@gtng-4db0412a.pool.mediaWays.net) |
| 13:56.08 | *** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821) |
| 13:56.52 | tmzt | diag smd_tty channel? |
| 13:57.16 | cr2 | tmzt: already forgot it |
| 13:57.27 | phh | smd1 |
| 13:57.30 | cr2 | phh: there are 2 dex irq regs: 128 and 138 |
| 13:57.48 | cr2 | phh: do we have smd1 ? |
| 13:58.54 | tmzt | what chip? |
| 14:00.45 | cr2 | tmzt: not in wiki |
| 14:00.50 | cr2 | so it does not exist |
| 14:01.16 | tmzt | huh? |
| 14:03.23 | *** join/#htc-linux sse2 (n=sse2@41.104.85.18) |
| 14:03.48 | phh | cr2: you want me to dump 128 too ? |
| 14:03.56 | cr2 | phh: irq is 128, 138 is some lock |
| 14:04.51 | phh | ok |
| 14:05.29 | cr2 | the 6 irq is acked |
| 14:05.52 | cr2 | 0x06 6 INT_A9_M2A_6 + DEX |
| 14:05.53 | sse2 | Hi there. Anyone here familiar with qxdm ? |
| 14:06.06 | cr2 | i think g1 has a bit different implementation |
| 14:06.16 | tmzt | somewhat, the program or the protocol? |
| 14:06.27 | sse2 | both |
| 14:06.28 | tmzt | android does less with dex |
| 14:06.55 | tmzt | sse2: I'm looking to add support to ofonod/gatcmd |
| 14:08.09 | sse2 | tmzt: ofonod ? |
| 14:08.11 | cr2 | Markinus: google amss uses INT_ADSP_A9_A11 for adsp5, we use INT_ADSP_A11. you may add it too |
| 14:09.08 | sse2 | I'm getting a OEM_HOOK_RAW error for no particular reason on the g1 |
| 14:09.22 | phh | cr2: 1004 |
| 14:09.26 | phh | 0x1004 |
| 14:09.47 | phh | and it was 4 before trying to send the notify_arm9_reboot |
| 14:10.01 | cr2 | phh: i had a list somewhere |
| 14:10.13 | tmzt | see: ofono.org basically I'm looking to have a daemon on linux talk qxdm based on dbus calls |
| 14:10.13 | cr2 | in the logs ;) |
| 14:10.20 | phh | cr2: irq.c ? |
| 14:10.37 | phh | oh *dex* irq |
| 14:10.47 | tmzt | cr2: nice, one of thhose useful changes in amss |
| 14:10.48 | cr2 | 4 is ignored |
| 14:11.10 | cr2 | tmzt: yeah. thank you googel |
| 14:11.22 | tmzt | sse2: that see: was for you |
| 14:11.41 | sse2 | yep i'm loking into it |
| 14:11.44 | tmzt | it reall6y depends if we need it for GPS |
| 14:11.56 | tmzt | we is merphone in this case |
| 14:13.27 | cr2 | phh: yes, 0x1000 is [14:04] <cr2> M2A_NOTIFY_ARM9_REQ_RESTART is a DEX irq which we don't do afaik |
| 14:13.34 | phh | ok |
| 14:14.34 | cr2 | phh: so marm9 acks the 8e cmd ? |
| 14:14.56 | phh | I don't think so |
| 14:14.56 | sse2 | tmzt: well, there's already an interface to qxdm on android, but the rild is returning errors when sending the hook command |
| 14:14.58 | Markinus | cr2: I tested this already because I saw this in my logs, but this doesn't change somethings on my sound problems. Is this only for Topaz or all devices? |
| 14:15.11 | phh | cr2: the 0x8e command times out |
| 14:15.39 | cr2 | Markinus: on all our devices |
| 14:15.53 | cr2 | phh: times out means ? |
| 14:16.10 | cr2 | phh: we have requested arm9 reboot here ? |
| 14:16.14 | tmzt | sse2: what's the hook command? |
| 14:16.16 | phh | cr2: pc_serial_check is never incremented |
| 14:16.25 | phh | cr2: it's "notify arm9 reboot", I can't see what this actuall means |
| 14:16.33 | cr2 | phh: i guess arm9 needs some time |
| 14:16.55 | cr2 | phh: who has written the timeout code ? |
| 14:16.59 | tmzt | that REQ RESTART could be wakeup? |
| 14:17.04 | tmzt | meaning of arm11 |
| 14:17.08 | phh | cr2: maejrep I think |
| 14:17.13 | cr2 | tmzt: no idea |
| 14:17.31 | cr2 | tmzt: what do you think it means ? |
| 14:17.48 | cr2 | phh: it's in dex driver ? aka proc_comm_wince ? |
| 14:17.52 | phh | yes |
| 14:17.56 | cr2 | hm |
| 14:18.05 | phh | <PROTECTED> |
| 14:18.08 | phh | I try adding that ? |
| 14:18.10 | tmzt | I think it means arm9 wants arm11 to start running |
| 14:18.16 | tmzt | but I don't know |
| 14:18.22 | sse2 | tmzt: there are basically 2 parameters to send to the interface: size of the log file and num of rotating files |
| 14:18.25 | *** join/#htc-linux x29a (n=x29a@unaffiliated/x29a) |
| 14:18.40 | tmzt | this is wird QCT terminology |
| 14:18.46 | tmzt | sse2: what log? |
| 14:18.52 | sse2 | tmzt: qxdmlog |
| 14:18.54 | tmzt | you can get us amss's log? |
| 14:19.21 | *** join/#htc-linux leaigor (n=laigor@188.134.16.241) |
| 14:19.30 | cr2 | phh: add a comment there |
| 14:19.46 | cr2 | phh: otherwise nobody will rememeber what's that in a week |
| 14:19.53 | phh | TRUE. |
| 14:19.56 | phh | oops |
| 14:20.09 | cr2 | tmzt: how do i disassemble amss ? |
| 14:20.19 | cr2 | phh: we need a new dex driver anyway... |
| 14:20.32 | tmzt | talk to viper |
| 14:20.36 | phh | cr2: we need everything brand new. |
| 14:20.37 | sse2 | cr2: look at the partition table at the beginning of the file |
| 14:20.38 | cr2 | phh: i need to strip some oob data |
| 14:20.48 | tmzt | or sse2 :) |
| 14:21.21 | cr2 | sse2: i don't get all relocations right. most annoying are the text string relocations. |
| 14:22.38 | cr2 | tmzt: NOTIFY_ARM9_AT_CMD_READY |
| 14:22.56 | tmzt | what? |
| 14:23.01 | tmzt | interpreter? |
| 14:23.06 | cr2 | tmzt: it's also a dex irq. arm9 tells us that AT interpreter is ready ? |
| 14:23.16 | tmzt | yeah that's my guess |
| 14:23.19 | sse2 | cr2:just strip down the amss image and only load the elf in ida |
| 14:23.24 | tmzt | what about rpc |
| 14:24.00 | cr2 | sse2: strip down ? what are the block+oob sizes ? |
| 14:24.30 | cr2 | sse2: i've done that with oemsbl once, but forgot the details |
| 14:24.35 | tmzt | in radio.nb? |
| 14:24.42 | sse2 | cr2:dump the sram to have the coorect offsets/sizes |
| 14:24.51 | sse2 | oops |
| 14:24.53 | sse2 | dump |
| 14:24.59 | cr2 | sse2: need to disable mmu first |
| 14:25.10 | cr2 | can i do it with haret ? |
| 14:25.16 | sse2 | tmzt: not the radio nb, just the amss, use qmat |
| 14:25.20 | cr2 | s/mmu/mpu/ |
| 14:26.19 | cr2 | i think it's in the suspend code |
| 14:26.33 | phh | cr2: waiting for M2A_NOTIFY_ARM9_REQ_RESTART before sendint notify_arm9_reboot doesn't help |
| 14:27.46 | cr2 | phh: i think it's the answer to dex 8e |
| 14:27.48 | sse2 | cr2: I don't think you need that for oemsbl |
| 14:28.06 | phh | cr2: btw, arm9 sent a 0x3100005b, proc=0xa, do you know what is it ? |
| 14:28.12 | phh | just before crashing |
| 14:28.23 | cr2 | ah. tvcable inout is a dex irq too |
| 14:28.39 | phh | cr2: I didn't call dex 0x8e .... |
| 14:28.50 | cr2 | phh: i've seen only 1,4 and b in the "normal" wince operation |
| 14:29.17 | tmzt | inout? |
| 14:29.30 | tmzt | is this stuff in the uboot patch? |
| 14:29.32 | phh | cr2: I get this irq *before* sending dex 0x8e |
| 14:29.43 | cr2 | phh: but you did an pdsm_init(1) instead of pdsm_init(2) in wince |
| 14:29.47 | cr2 | phh: ok |
| 14:30.18 | tmzt | still on gps? |
| 14:30.20 | cr2 | phh: msm_check_for_modem_crash |
| 14:30.26 | cr2 | phh: where is this defined ? |
| 14:30.29 | phh | tmzt: yes |
| 14:30.42 | cr2 | 30 |
| 14:30.44 | cr2 | <PROTECTED> |
| 14:30.44 | phh | cr2: msm_smd_probe |
| 14:30.45 | cr2 | <PROTECTED> |
| 14:30.47 | cr2 | <PROTECTED> |
| 14:30.48 | cr2 | <PROTECTED> |
| 14:30.50 | cr2 | <PROTECTED> |
| 14:30.51 | cr2 | <PROTECTED> |
| 14:30.52 | sse2 | btw, anyone here succeded in patching the amss and rebuilding th rsa keys ? |
| 14:30.53 | cr2 | <PROTECTED> |
| 14:31.05 | cr2 | phh: we need to demux this irq |
| 14:31.37 | cr2 | sse2: we need to make linux run first |
| 14:31.47 | cr2 | phh: in smd.c ? |
| 14:31.55 | phh | cr2: yes |
| 14:32.19 | sse2 | cr2: what device are you working on ? |
| 14:32.38 | cr2 | all amss5225 |
| 14:32.50 | cr2 | and 6125 too |
| 14:33.05 | cr2 | phh: this is different on leo : unsigned base = (unsigned)(MSM_SHARED_RAM_BASE + 0xfc100); |
| 14:33.07 | sse2 | ah |
| 14:33.15 | sse2 | i'm with 7200 |
| 14:33.21 | phh | cr2: may I say that leo can wait a bit ? :D |
| 14:33.25 | tmzt | same |
| 14:33.30 | tmzt | 7200 non-A? |
| 14:33.38 | sse2 | nope, a |
| 14:33.43 | tmzt | sorry, I mean all of this is 7k |
| 14:33.57 | sse2 | anyway yes |
| 14:33.58 | tmzt | 7201a 7200a but the differences are in amss |
| 14:34.06 | tmzt | between android and wince |
| 14:34.30 | phh | (and some devices) |
| 14:34.52 | cr2 | phh: we may add a note, so NetRipper will have it easier :) |
| 14:34.59 | phh | cr2: sure |
| 14:35.01 | phh | but still :p |
| 14:35.14 | sse2 | been trying to patch g1 amss some time ago |
| 14:35.57 | sse2 | can't seem to locate the private rsa key |
| 14:37.00 | cr2 | sse2: can you patch it in ram ? |
| 14:37.52 | cr2 | tmzt: AUD_REC_GET_BUF has a dex irq too |
| 14:38.22 | sse2 | cr2: you can, but useless in my case |
| 14:39.05 | tmzt | GNUtoo: ping |
| 14:39.06 | sse2 | cr2: basically for a persistant patch, you can flash it the loader will accept it, but it won't load it to sram |
| 14:39.12 | GNUtoo | tmzt, hi |
| 14:39.20 | cr2 | sse2: ok |
| 14:39.22 | tmzt | sse2: tried mpu? |
| 14:39.34 | sse2 | tmzt: no |
| 14:39.59 | tmzt | talk to cmonex, she can do it for gsm unlock |
| 14:40.13 | tmzt | from sspl |
| 14:41.24 | GNUtoo | tmzt, what's going on? |
| 14:41.30 | GNUtoo | why patching amss is needed? |
| 14:41.55 | tmzt | GNUtoo: the dex irq cr2 found |
| 14:42.11 | GNUtoo | what's that? |
| 14:42.24 | GNUtoo | the GPS thing? |
| 14:42.30 | cr2 | tmzt: actually i'd like to see the amss dex and rpc handler code |
| 14:42.30 | tmzt | AUD_REC_GET_BUF |
| 14:43.03 | tmzt | is there a playback equivalent? |
| 14:43.14 | GNUtoo | I'll look |
| 14:43.18 | GNUtoo | in alsa or qdsp? |
| 14:43.20 | sse2 | cr2: disass the elf |
| 14:43.21 | cr2 | dex seems to be a very stable api. unlike rpc ;) |
| 14:43.50 | cr2 | sse2: yes. need to disable 0xa protection first |
| 14:44.26 | GNUtoo | I don't even know what is dex |
| 14:44.48 | phh | GNUtoo: proc_comm |
| 14:45.24 | sse2 | cr2: right |
| 14:45.31 | cr2 | 2 |
| 14:45.33 | cr2 | <PROTECTED> |
| 14:45.34 | cr2 | <PROTECTED> |
| 14:45.36 | cr2 | <PROTECTED> |
| 14:45.37 | cr2 | <PROTECTED> |
| 14:45.39 | cr2 | <PROTECTED> |
| 14:45.40 | cr2 | <PROTECTED> |
| 14:45.42 | cr2 | <PROTECTED> |
| 14:45.55 | cr2 | phh: hmm. how long is that ? |
| 14:46.00 | phh | cr2: 10s |
| 14:47.33 | GNUtoo | ok |
| 14:48.23 | cr2 | phh: can you add the state dump here -> |
| 14:48.27 | cr2 | 114 |
| 14:48.28 | cr2 | <PROTECTED> |
| 14:48.30 | cr2 | <PROTECTED> |
| 14:48.31 | cr2 | <PROTECTED> |
| 14:49.17 | phh | which state ? |
| 14:49.18 | phh | irq ? |
| 14:49.34 | cr2 | arm9 and arm11 |
| 14:49.55 | cr2 | ID_SHARED_STATE |
| 14:50.25 | cr2 | diag is ID_DIAG_ERR_MSG |
| 14:51.07 | tmzt | smsm is astruct matching part of smem right? |
| 14:51.15 | GNUtoo | so where should I help |
| 14:51.15 | GNUtoo | ? |
| 14:52.13 | tmzt | GNUtoo: does this help you with alsa_qdsp? |
| 14:52.17 | cr2 | struct smsm_shared |
| 14:52.17 | tmzt | this dex irq |
| 14:52.30 | tmzt | or these dex irqs |
| 14:52.41 | tmzt | could it explain what's happening with buffers? |
| 14:53.16 | GNUtoo | tmzt, basically for alsa_qdsp the problem I was dealing is that one: http://mailman.alsa-project.org/pipermail/alsa-devel/2010-January/024445.html |
| 14:53.24 | GNUtoo | so... |
| 14:53.39 | GNUtoo | what is unknow is why the first buffer is INVALID_LELN |
| 14:53.44 | GNUtoo | *INVALID_LEN |
| 14:53.52 | tmzt | right, but there might be more or different notification proceedures on wince amss |
| 14:54.05 | tmzt | done through dex/proc_comm |
| 14:55.11 | GNUtoo | so: I don't know a thing about the proc_comm or dex...I only know that it's "The lowest layer IPC" and that it "is used for clock/power control" |
| 14:55.25 | GNUtoo | where can I learn about it? |
| 14:55.28 | tmzt | yeah |
| 14:55.36 | GNUtoo | and will it affect the alsa driver? |
| 14:55.40 | tmzt | but it might be involved in qdsp |
| 14:55.46 | GNUtoo | ok |
| 14:55.54 | cr2 | phh: smd handles M2A_5 only |
| 14:55.55 | tmzt | cr2 found a dex irq called AUD_REC_REQ_BUF |
| 14:55.58 | cr2 | r = request_irq(INT_A9_M2A_5, smsm_irq_handler |
| 14:56.07 | cr2 | IRQF_TRIGGER_RISING, "smsm_dev", 0); |
| 14:56.14 | GNUtoo | ok |
| 14:56.56 | cr2 | phh: int smd_core_init(void) |
| 14:57.13 | cr2 | phh: probably need a dex_core_init() too |
| 14:57.28 | phh | cr2: I init dex when receiving this interrupt ? |
| 14:59.58 | cr2 | phh: no, a more general dex driver |
| 15:00.02 | cr2 | static int debug_read_stat(char *buf, int max) |
| 15:00.22 | cr2 | phh: check this one for crash logging |
| 15:01.26 | cr2 | phh: the main question for me is: do we need to queue dex calls ? |
| 15:01.53 | tmzt | sse2: is debug_read_stat related to what you're talking about? |
| 15:02.12 | tmzt | see2: we need this qxdm command |
| 15:02.17 | tmzt | to enable logging |
| 15:02.21 | tmzt | and how to send it |
| 15:02.30 | tmzt | can you strace rild? |
| 15:03.00 | tmzt | oh |
| 15:03.04 | phh | tmzt: how useful tracing rild would be ? |
| 15:03.08 | tmzt | OEM_HOOK_RAW is something sent to theclosed ril? |
| 15:03.11 | GNUtoo | I'll try something |
| 15:03.18 | tmzt | phh: you've seen this? |
| 15:03.22 | phh | no |
| 15:03.29 | phh | what is it ? |
| 15:03.30 | phh | a rpc ? |
| 15:03.39 | sse2 | tmzt: the hook command is sent by thr rild yeah |
| 15:03.39 | tmzt | ril wire command I think |
| 15:04.05 | tmzt | who was working on ficing libreference_ril? |
| 15:04.11 | sse2 | nope, not talking about debug_read_stat |
| 15:04.50 | phh | smsm: ARM9 HAS CRASHED |
| 15:04.50 | phh | smsm: a9: 00000829 a11: 00010c41 |
| 15:04.50 | phh | diag: '' |
| 15:05.22 | phh | hum, arm9 has crashed stays forever |
| 15:05.28 | phh | maybe we have to unset it ? |
| 15:05.30 | tmzt | sse2: let's find this and add to kernel |
| 15:05.40 | cr2 | phh: lol. decode the bits |
| 15:05.50 | tmzt | also, can we add raw smd dump to dbgfs? |
| 15:06.23 | sse2 | smd dump would be useful |
| 15:06.27 | cr2 | tmzt: into ramconsole buf |
| 15:06.39 | tmzt | why? |
| 15:06.52 | tmzt | just a dbgfs node to cat |
| 15:07.17 | tmzt | if qxdm is enabled we can add that to logcat radio or ramconsole |
| 15:07.31 | phh | cr2: init & smdinit & reset & timeinit ? |
| 15:07.40 | sse2 | if it's enabled |
| 15:07.40 | tmzt | so we really need this binary command |
| 15:08.27 | cr2 | 800, 20, 8 , 1 and 10000, 800, 400, 40, 1 |
| 15:09.54 | phh | timeinit & rpcinit & smdinit & init for a9 and unknown&timewait & reset & init for A11 ? |
| 15:09.55 | phh | hum |
| 15:10.35 | cr2 | yeah. looks quite dead ;) |
| 15:10.52 | phh | kind of yes. |
| 15:10.58 | cr2 | where did i see the dex 0x14 call ?? |
| 15:11.54 | phh | cr2: hum, PCOM_PMIC_CAM_ON is the same as the three vreg enable I have ? |
| 15:13.37 | cr2 | phh: don't mix PCOM_ and DEX_ |
| 15:13.54 | phh | that's what in proc_comm_wince.h, not my fault :p |
| 15:13.54 | cr2 | i think we need to separate them properly again. |
| 15:14.12 | cr2 | yeah. comes from old times ;) |
| 15:14.52 | cr2 | http://htc-linux.org/wiki/index.php?title=RaphaelDEX |
| 15:14.59 | cr2 | the dex cmd list |
| 15:15.19 | *** join/#htc-linux BabelO (n=fcr@unaffiliated/babelo) |
| 15:18.07 | phh | cr2: btw I can see an arm9 low speed, is that a power saving feature ? |
| 15:18.24 | cr2 | maybe. |
| 15:18.42 | cr2 | i'm looking at it too. the syntax is not documented in wiki |
| 15:18.43 | phh | :) |
| 15:18.54 | phh | wiki says there is no arg |
| 15:18.56 | cr2 | there are 2 calls |
| 15:19.05 | phh | (which seems strange) |
| 15:19.10 | cr2 | lock speed, and low speed. |
| 15:19.21 | cr2 | both 1d, but params are different. |
| 15:19.27 | cr2 | wiki has a bug probably. |
| 15:19.34 | phh | chamonix: that's something for you |
| 15:20.00 | cr2 | update audio |
| 15:20.10 | cr2 | is also not well documented. |
| 15:20.41 | tmzt | cr2: could this be something we neexd for qdsp to work that g1 doesn't? |
| 15:27.14 | chamonix | oO |
| 15:27.38 | chamonix | low speed is good |
| 15:27.42 | chamonix | :) |
| 15:27.49 | phh | maybe |
| 15:28.14 | chamonix | did I dream or was there a powertop port in some other version of eclair/donut? |
| 15:28.30 | phh | chamonix: donut had powertop |
| 15:29.17 | chamonix | ty, I'm not nuts :) would it run on Eclair as is or do I need to recompile the kernel with some specific settings? |
| 15:29.34 | phh | run as is iirc |
| 15:30.03 | chamonix | cool, I'll transfer |
| 15:30.25 | phh | cr2: is there some "cam sel" gpio to set for camera ? |
| 15:30.30 | chamonix | I just have to find a machine with old squahfs |
| 15:30.33 | cr2 | hm. cpu frequency table |
| 15:30.44 | chamonix | s/ah/ash |
| 15:30.46 | cr2 | phh: cam sel ? |
| 15:31.06 | phh | cr2: to select if the gpios goes to MI380 or MT9T013 |
| 15:31.28 | dcordes | is the MI one aptina? |
| 15:31.32 | *** join/#htc-linux zycho (n=zycho@f054160186.adsl.alicedsl.de) |
| 15:34.35 | cr2 | phh: 1e=1 for mi380 or idle. 1e=0 for mt9p013 |
| 15:34.41 | chamonix | cool, it works |
| 15:35.00 | phh | chamonix: means it doesn't crash ? :p |
| 15:35.08 | chamonix | Suggestion: Enable the CONFIG_NO_HZ kernel configuration option <--- lmfao |
| 15:35.10 | cr2 | ok, need to chec the freq table now. |
| 15:35.37 | chamonix | wow, 1 Watt is at stake |
| 15:35.44 | tmzt | that should work, if not we need to fix timer |
| 15:35.55 | phh | chamonix: lol. |
| 15:36.04 | chamonix | well, let's go back to lagging time |
| 15:36.18 | phh | chamonix: I thaught you were talking about arm9 slow clock :( |
| 15:36.23 | chamonix | HeHeHe |
| 15:36.29 | chamonix | it wants ACPI too |
| 15:36.55 | chamonix | sry phh :-/ |
| 15:36.56 | phh | :) |
| 15:37.05 | phh | it's an intel tool you know. |
| 15:37.37 | chamonix | indeed |
| 15:42.08 | chamonix | well I'll try to follow some suggestions..... see if I get something that compiles and runs |
| 15:42.58 | phh | chamonix: first check the highest interrupt maker |
| 15:43.10 | chamonix | gp_timer |
| 15:43.18 | phh | screen on ? |
| 15:43.22 | chamonix | yes |
| 15:43.27 | phh | 10/s ? |
| 15:43.44 | phh | iirc it's kxsd |
| 15:43.51 | chamonix | 100 per 15s |
| 15:44.08 | chamonix | and msm_i2c 75 / 15s |
| 15:44.26 | chamonix | what is kxsd? |
| 15:44.30 | phh | accelerometer |
| 15:44.34 | phh | which uses i2c bus :p |
| 15:45.40 | phh | oooooohhhhhhhhhh |
| 15:45.43 | phh | it seems camera works |
| 15:45.47 | phh | now how do i test it :p |
| 15:45.48 | chamonix | OO |
| 15:46.43 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 15:46.45 | phh | works means i2c commands work. |
| 15:47.18 | phh | hum it still created no msm_camera devices |
| 15:49.31 | cr2 | phh: with 1e=0 you can talk to mt9p013 ? |
| 15:49.37 | phh | yup |
| 15:49.47 | cr2 | phh: that means you can talk to mi380 too :) |
| 15:49.50 | phh | well I'm not sure, i have to double check |
| 15:49.57 | cr2 | if you properly set the gpio alt |
| 15:50.25 | phh | i have no mention to mt9t013 at all in the dmesg log |
| 15:50.29 | phh | which seems strange. |
| 15:50.34 | phh | but the clocks and gpio_set_function are called |
| 15:50.34 | cr2 | phh: the lowd clock id dex_1d(2); |
| 15:50.49 | cr2 | 2 seems to be the 0-7 voltage index ?? |
| 15:50.50 | *** join/#htc-linux DarkMasterHalo (n=DarkMast@modemcable187.118-80-70.mc.videotron.ca) |
| 15:51.22 | phh | haha. I booted to eclair while the only distrib I have with correct android camera driver is donut. and I did that three times. |
| 15:51.23 | cr2 | ok, looks like khz block |
| 15:51.33 | cr2 | lol |
| 15:51.48 | phh | wimo is so slow to boot that I forget what I want to do ... |
| 15:53.04 | cr2 | 61440000 speed is low |
| 15:53.12 | cr2 | lpp0/4 ? |
| 15:53.18 | cr2 | pll0/4 ? |
| 15:53.21 | phh | not for sleep |
| 15:53.39 | phh | not for wakeup either I think actually for a modem |
| 15:53.40 | cr2 | dex_1d(2) |
| 15:53.52 | cr2 | ARM9 low speed |
| 15:54.48 | chamonix | ok, back to GENERIC_TIME |
| 15:57.41 | maejrep | that's annoying that I can't change my nick when I'm in #android-root |
| 15:57.42 | phh | time to cross finger really hard |
| 15:58.15 | phh | black screen. |
| 15:58.55 | maejrep | cr2: long time no see :) |
| 15:59.22 | cr2 | hi maejrep |
| 15:59.33 | cr2 | maejrep: need to rewite the dex driver :) |
| 15:59.42 | maejrep | :X |
| 15:59.49 | maejrep | i have a cdma hero now :p |
| 16:00.05 | tmzt | irqchip driver? |
| 16:00.10 | tmzt | mfd? |
| 16:00.10 | cr2 | still cdma ? |
| 16:00.12 | maejrep | working on reverse engineering the linux kernel :( |
| 16:00.24 | cr2 | hehe |
| 16:00.38 | tmzt | hmm, android devices don't need all this |
| 16:00.42 | cr2 | maejrep: haret for linux ? |
| 16:00.42 | maejrep | once again, they have a very weird smem layout for msm75xx |
| 16:00.48 | maejrep | hah i wish :p |
| 16:00.51 | maejrep | i don't even get a console |
| 16:01.09 | phh | cr2: without /dev/mem that's hard |
| 16:01.09 | cr2 | maejrep: port uboot |
| 16:01.15 | maejrep | I have to trigger a kernel panic, force restart_reason = "recovery", just to see the ram console |
| 16:01.34 | cr2 | maejrep: htc posted all msm_clk layout now. |
| 16:01.48 | maejrep | for raph? |
| 16:01.55 | cr2 | for 7201A |
| 16:01.58 | phh | cr2: uh ? |
| 16:02.00 | maejrep | ah |
| 16:02.21 | cr2 | phh: 7500*.h |
| 16:02.28 | phh | cr2: that's for 7500 then :p |
| 16:02.33 | cr2 | lol |
| 16:02.47 | cr2 | they differ only in the dsp |
| 16:03.19 | tmzt | cr2: where? |
| 16:03.21 | phh | no the problem is with VFE i think. |
| 16:03.39 | cr2 | maejrep: fixed /dev/mem ? |
| 16:03.52 | cr2 | phh: post the code, since you don't use rpc for it. |
| 16:03.54 | maejrep | i don't know :o? |
| 16:04.04 | cr2 | maejrep: on hero |
| 16:04.11 | phh | cr2: uh ? |
| 16:04.18 | maejrep | there is no /dev/mem |
| 16:04.53 | phh | cr2: you have clock dumps for camera ? the VFE NS bit 14 is set to 0 or 1 ? |
| 16:05.06 | cr2 | phh: wiki. |
| 16:05.19 | cr2 | phh: but i may connect and recheck it |
| 16:05.26 | cr2 | maejrep: why ? |
| 16:05.36 | maejrep | i don't know |
| 16:05.42 | maejrep | ls /dev/mem == not there :p |
| 16:05.45 | cr2 | maejrep: i also need to fix your buggy ksc leds on raph100 :) |
| 16:05.56 | maejrep | pfft |
| 16:06.00 | maejrep | they worked for me :p |
| 16:06.16 | tmzt | switching to android microp driver? |
| 16:06.22 | cr2 | maejrep: the same problem as everywhere. no /dev/mem on android |
| 16:06.30 | cr2 | maejrep: on raph800 ;) |
| 16:07.07 | cr2 | maejrep: on raph100 they oops, and Fn led is Caps |
| 16:07.12 | phh | tmzt: their driver isn't usable as is for us |
| 16:07.18 | maejrep | weird |
| 16:07.22 | phh | and it's a really big driver |
| 16:07.27 | phh | I'd prefer cutting it in several files |
| 16:07.47 | tmzt | it should be mfd |
| 16:07.49 | cr2 | maejrep: they updated the firmware now, so a lot more commands are needed. |
| 16:07.56 | maejrep | doh |
| 16:08.02 | maejrep | only if you applied the update? :p |
| 16:08.06 | phh | cr2: uh ? |
| 16:08.09 | cr2 | maejrep: but it is used everywhere, and nexus has a new microp driver |
| 16:08.24 | maejrep | so more uniform? |
| 16:08.31 | cr2 | maejrep: for the newer phones |
| 16:08.43 | phh | maejrep: hero hasn't the same microp "api" though |
| 16:08.46 | cr2 | i still edit the raph page for all commands |
| 16:09.26 | phh | cr2: you've seen how they handle irqs btw ? |
| 16:09.33 | cr2 | maejrep: but we got most things working now. (on gsm). struggling with the last bits of cam and gps |
| 16:09.40 | cr2 | phh: where ? |
| 16:09.42 | phh | and fm. |
| 16:09.46 | phh | cr2: mahimahi's microp |
| 16:09.50 | cr2 | fm is userspace |
| 16:09.56 | chamonix | phh: have you found anything about SMS encoding or should I have a look? |
| 16:09.57 | phh | right |
| 16:10.05 | phh | chamonix: you should have a look |
| 16:10.07 | cr2 | phh: not yet. need a nexus first. |
| 16:10.08 | chamonix | k |
| 16:10.20 | phh | cr2: you need a nexus to read its code ? :p |
| 16:10.24 | cr2 | btw. what is the arm9 clock speed ? |
| 16:10.37 | chamonix | I need a nexus too, preferably the one with keyboard |
| 16:10.38 | maejrep | cr2: http://xda.privatepaste.com/bc505aeb6b -- that's the best we've been able to get so far with compiling the gsm hero source for cdma heroc |
| 16:10.48 | phh | chamonix: :p |
| 16:10.49 | cr2 | phh: need a nexus anyway :) |
| 16:10.57 | tmzt | microp irq? |
| 16:11.14 | phh | cr2: and we still need hardware video decoding :D |
| 16:11.41 | cr2 | maejrep: new smem layout ? |
| 16:11.50 | maejrep | yes (again) |
| 16:12.10 | maejrep | only for heroc though.. gsm hero doesn't use the weird large fifo buffers |
| 16:12.19 | tmzt | did start of ram move? |
| 16:12.21 | phh | maejrep: delete diag usb function driver. |
| 16:12.22 | cr2 | maejrep: 0x4000 is 0x2000*2 it's the new SMD TX/RX pair |
| 16:12.23 | phh | you don't want it anyway. |
| 16:12.28 | maejrep | yes |
| 16:12.33 | maejrep | same as raph800, essentially |
| 16:12.38 | cr2 | maejrep: check out 6125 patch |
| 16:12.43 | maejrep | except it's also contained in smem toc |
| 16:12.43 | tmzt | phh: it's pointless without qxdm start |
| 16:12.52 | phh | tmzt: yup |
| 16:13.06 | tmzt | if we haqd that it could be quite useful |
| 16:13.17 | phh | greeee |
| 16:13.23 | tmzt | heroc uses qmi/rmnet |
| 16:13.26 | phh | I booted eclair instead of donut AGAIN |
| 16:13.34 | cr2 | maejrep: can you dump smem ? |
| 16:13.54 | maejrep | not easily.. i guess i could write a kernel module to do it |
| 16:13.54 | phh | it's going to time forever if i keep doing such stupid mistakes .. |
| 16:13.59 | cr2 | maejrep: maybe in the bootloader |
| 16:14.12 | cr2 | yes. kernel module may be the easiest. |
| 16:14.13 | maejrep | i don't know how |
| 16:14.19 | cr2 | ok |
| 16:14.28 | maejrep | and heroc bootloader (fastboot) is locked down |
| 16:14.44 | maejrep | phh: buggy kernel.. doesn't honor CONFIG_USB_FUNCTION_DIAG ;x |
| 16:14.51 | phh | ah. |
| 16:14.53 | phh | remove it by hand :p |
| 16:14.55 | maejrep | drivers/usb/function/msm_hsusb.c:1041: undefined reference to `g_bUsbDiagMode' |
| 16:15.03 | tmzt | but you can flash it right? |
| 16:15.11 | maejrep | only from recovery |
| 16:15.21 | maejrep | cat /dev/zero >> /dev/mtd/mtd2 && flash_image boot /sdcard/boot.img |
| 16:15.22 | maejrep | :p |
| 16:15.30 | tmzt | that sounds like a gadget header |
| 16:15.31 | phh | zero > mtddevice ?!? |
| 16:15.38 | phh | that sounds crazy |
| 16:15.46 | phh | a mtddevice should be set to 0xff, not 0x00 |
| 16:15.57 | tmzt | it will failback to recovery |
| 16:16.20 | maejrep | http://android-dls.com/wiki/index.php?title=HOWTO:_Unpack%2C_Edit%2C_and_Re-Pack_Boot_Images#Flashing_your_new_image_back_onto_the_phone |
| 16:16.28 | maejrep | that's what i was going by |
| 16:16.41 | sse2 | nand default byte state is 0xff not 0 |
| 16:17.26 | cr2 | phh: video decoder is tough |
| 16:17.35 | phh | cr2: I know |
| 16:17.54 | cr2 | phh: fmradio is doable |
| 16:18.07 | cr2 | phh: gps annoys me ;) |
| 16:18.09 | phh | everything is :p |
| 16:18.25 | cr2 | yeah. but gps |
| 16:18.39 | phh | bah it will be done I'm sure. |
| 16:18.49 | maejrep | well there isn't a /dev/one |
| 16:20.00 | phh | now we have to play with adsp. cool. |
| 16:21.04 | phh | cr2: but it won't be done today but next week then :p |
| 16:21.55 | cr2 | phh: your gpio setup ? |
| 16:22.12 | cr2 | for cam |
| 16:22.19 | phh | <PROTECTED> |
| 16:22.19 | phh | <PROTECTED> |
| 16:22.19 | phh | <PROTECTED> |
| 16:22.19 | phh | <PROTECTED> |
| 16:22.57 | phh | <PROTECTED> |
| 16:22.57 | phh | <PROTECTED> |
| 16:23.02 | phh | <PROTECTED> |
| 16:23.07 | phh | <PROTECTED> |
| 16:23.29 | phh | meaning everybody is set to 5 but: |
| 16:23.35 | phh | [ 15.784667] msm_gpio_set_function(c, 1c5) |
| 16:23.35 | phh | [ 15.823760] msm_gpio_set_function(f, 1c5) |
| 16:24.08 | *** join/#htc-linux slight (n=23hdci27@154.112.221.87.dynamic.jazztel.es) |
| 16:24.32 | cr2 | 1c5 ? |
| 16:24.49 | phh | 1c5. |
| 16:24.52 | cr2 | should be 5 for c |
| 16:24.56 | cr2 | and 1c4 for f |
| 16:25.01 | GNUtoo | mmm |
| 16:25.55 | cr2 | phh: http://htc-linux.org/wiki/index.php?title=Raphael_GPIO |
| 16:25.55 | GNUtoo | if (frame->used == BUF_INVALID_LEN){break;} now works but I have horrible sound : [ 341.368591] audpp: DMA missed obj=20 even with aplay |
| 16:26.46 | phh | cr2: http://husson.hd.free.fr/dm-cam.1 I think there is nothing relevant but the gpio funcs in it |
| 16:26.55 | tmzt | GNUtoo: with dex? |
| 16:27.03 | GNUtoo | no without |
| 16:27.13 | GNUtoo | what should I do with dex? |
| 16:27.17 | GNUtoo | rewrite all the drivers? |
| 16:27.33 | cr2 | Unbalanced IRQ 158 wake disable ? |
| 16:27.52 | phh | cr2: bah that's just a slowpath |
| 16:28.00 | tmzt | GNUtoo: I think amss tells you when it wants the next buffer |
| 16:28.07 | tmzt | or you tell it when your ready |
| 16:28.29 | cr2 | [ 1.724182] BT RFK register |
| 16:28.30 | cr2 | [ 1.724243] vreg_get: rftx -> 11 |
| 16:28.32 | cr2 | [ 1.724395] BT RFK probe |
| 16:28.38 | cr2 | phh: on diam ? |
| 16:28.41 | phh | yes |
| 16:28.51 | GNUtoo | ok but I don't understand why there is this frame->used of invalid len |
| 16:28.54 | phh | cr2: it's just a get |
| 16:29.00 | phh | on diam it does nothing with it |
| 16:29.07 | cr2 | http://htc-linux.org/wiki/index.php?title=MSM_VREG |
| 16:29.20 | phh | and I was talking about camera -_-' |
| 16:30.25 | tmzt | GNUtoo: only parts of smsm can trigger irqs on arm9 or qdsp cores |
| 16:30.43 | cr2 | phh: change alt |
| 16:30.53 | cr2 | phh: maybe ifdef |
| 16:30.57 | GNUtoo | you mean smem? |
| 16:31.07 | cr2 | phm. just change then in board file |
| 16:31.08 | tmzt | yeah |
| 16:31.21 | GNUtoo | ok |
| 16:31.21 | tmzt | sort of, I'm not sure what the difference is |
| 16:32.33 | phh | cr2: f to 1c4 ?so GPIO_NO_PULL ? |
| 16:32.52 | cr2 | phh: bte, do you check for vbus presence and battery presence in the charger enable ? |
| 16:33.05 | phh | no |
| 16:33.24 | cr2 | 1c4=0001.1100.0100 |
| 16:33.35 | *** join/#htc-linux cr2 (n=cr2@ip-109-85-35-207.web.vodafone.de) |
| 16:33.43 | phh | cr2: yes and the first two bits are for gpio pull |
| 16:33.57 | cr2 | what is the byte divis |
| 16:33.59 | cr2 | ion ? |
| 16:34.09 | cr2 | 000111000100 |
| 16:34.18 | phh | config = (cfg.drvstr<<6) | (cfg.func<<2) | (cfg.pull); |
| 16:34.35 | cr2 | 000111.0001.00 ? |
| 16:34.38 | *** join/#htc-linux c2d (n=Mike@port-92-200-10-104.dynamic.qsc.de) |
| 16:34.39 | phh | yup |
| 16:34.46 | cr2 | ok. |
| 16:34.50 | cr2 | 7,1,0 |
| 16:35.11 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 16:35.39 | phh | so no pull on it ? |
| 16:36.00 | cr2 | yes |
| 16:36.07 | cr2 | do what wince does |
| 16:36.39 | cr2 | 5 is 0101 |
| 16:36.51 | cr2 | alt=1, pull=1 |
| 16:39.26 | *** join/#htc-linux Billyonthemounta (n=Miranda@stiitcsc409.epfl.ch) |
| 16:40.30 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 16:40.40 | maejrep | cr2: i have an enum bitfield setup in ida that decodes the PCOM_GPIO_CFG values ;o |
| 16:41.56 | phh | ~ramconsole-diam |
| 16:41.57 | apt | well, ramconsole-diam is pwf dm 0x00800000 0x00100000 |
| 16:42.10 | cr2 | maejrep: too complex for me :) |
| 16:43.05 | *** join/#htc-linux herpez|away (n=herpez@87.196.51.213) |
| 16:43.18 | cr2 | maejrep: i don't even know how to replace =#0xFOOOBA00 with MY_REG_VALUE |
| 16:43.33 | cr2 | from a list |
| 16:43.36 | phh | you just know reading asm and text ? :p |
| 16:43.52 | *** join/#htc-linux rashire (n=ed1112wa@pool-98-114-205-192.phlapa.fios.verizon.net) |
| 16:43.57 | cr2 | phh: i remember register locations. |
| 16:44.05 | cr2 | of same the .asm and use 'sed' |
| 16:44.18 | cr2 | s/same/save/ |
| 16:44.19 | maejrep | i don't either ;x (replacing =#* with a symbol) |
| 16:44.24 | maejrep | cr2: http://xda.privatepaste.com/00b309bc25 |
| 16:45.07 | maejrep | it saves to a .idc file you can import, if you want |
| 16:45.27 | cr2 | looks good |
| 16:46.24 | maejrep | http://xda.privatepaste.com/95d9556f0f -- idc script |
| 16:46.33 | phh | cr2: exactly the same result |
| 16:46.45 | maejrep | (it has all my structs, but only really need gpio_table_record, and the gpio_* enums |
| 16:46.57 | phh | cr2: http://husson.hd.free.fr/dm-cam.2 |
| 16:47.18 | phh | V/videocamera( 1502): startPreview |
| 16:47.18 | phh | I/ActivityManager( 1180): Process com.android.mms (pid 1290) has died. |
| 16:47.18 | phh | D/CameraService( 1149): CameraService::connect E (pid 1502, client 0x1c400) |
| 16:47.18 | phh | D/CameraService( 1149): Client::Client E (pid 1502) |
| 16:47.19 | phh | D/QualcommCameraHardware( 1149): createInstance: E |
| 16:47.21 | phh | D/BatteryWidget( 1461): Updating Views |
| 16:47.23 | phh | D/BatteryWidget( 1461): Updating Views |
| 16:47.25 | phh | W/WindowManager( 1180): App freeze timeout expired. |
| 16:47.32 | phh | that's what android says |
| 16:47.36 | phh | pretty much nothing. |
| 16:47.44 | cr2 | phh: dmesg ? |
| 16:47.57 | phh | cr2: http://husson.hd.free.fr/dm-cam.2 |
| 16:48.37 | cr2 | phh: maybe call rpc for clocks ? |
| 16:50.13 | phh | ok |
| 16:50.41 | cr2 | there are 3 calls |
| 16:50.52 | cr2 | i'll connect raph100 and check now. |
| 16:50.56 | phh | 28/29/33 ? |
| 16:50.58 | phh | on 0x7 |
| 16:51.06 | cr2 | wiki sequence |
| 16:51.07 | phh | and 0 or 1 on 0x26 |
| 16:51.23 | cr2 | pmem is a different issue. |
| 16:51.30 | phh | 0x7 is pmem ? |
| 16:52.17 | cr2 | x28,x33,x29 for cam |
| 16:52.29 | cr2 | i'll write a haret function for it now. |
| 16:52.35 | phh | cr2: with command 0x7 ? |
| 16:52.41 | phh | but it needs also command 0x26 |
| 16:52.54 | cr2 | PMEM_REMOTE 0x30000055 pmem_* |
| 16:52.55 | cr2 | 1,1 pmem_remote_imem_enable (1) |
| 16:52.57 | cr2 | 2,1 pmem_remote_imem_disable (1) |
| 16:53.14 | cr2 | ah. ok |
| 16:53.18 | cr2 | 0x26,1 clk_regime_sec_sel_vfe_src (0/1) |
| 16:54.12 | phh | oh we might need pmem_remote for 3D init |
| 16:54.13 | cr2 | i think in CIF.dll it's mclk_ something |
| 16:54.29 | phh | cr2: mclk_internal ? |
| 16:56.30 | phh | cr2: same thing with rpc |
| 16:56.33 | *** join/#htc-linux TrXuk (n=TrXuk@host81-151-234-104.range81-151.btcentralplus.com) |
| 16:58.16 | phh | oh not exactly |
| 16:58.21 | cr2 | lol.wakeup_by_removing_battery |
| 16:58.36 | cr2 | phh: it's translated into something else |
| 16:58.42 | cr2 | by msm_clk probably |
| 16:59.08 | tmzt | cr2: those are dex? |
| 17:01.55 | cr2 | no. rpc |
| 17:02.29 | cr2 | startig haret |
| 17:08.14 | cr2 | Markinus: still here ? |
| 17:08.32 | phh | cr2: btw, some amss has a special clock for adsp, doesn't that apply for us as well ? |
| 17:09.14 | phh | see adsp_6225.c for example |
| 17:09.19 | tmzt | sse2: when we send amss something it doesn't like it can crash |
| 17:09.36 | phh | s/can/will/ |
| 17:09.38 | sse2 | like what ? |
| 17:09.42 | cr2 | phhYES. BUT IT'S TRICKY |
| 17:09.46 | cr2 | lol |
| 17:10.00 | phh | I'll let you do your stuff :p |
| 17:10.21 | sse2 | tmzt: sending over the oem hook comands ? |
| 17:10.33 | tmzt | we haven't tried yet |
| 17:10.41 | tmzt | what device does rild send those on? |
| 17:10.55 | tmzt | if it's encoded AT I can tell you how to try that |
| 17:12.59 | sse2 | android ones and it's not encoded at commands |
| 17:14.04 | tmzt | well you can still try it |
| 17:14.16 | *** join/#htc-linux makkonen (n=makkonen@cpe-66-69-229-9.austin.res.rr.com) |
| 17:14.25 | tmzt | but it seems to be limited to efs and cfg access |
| 17:14.37 | tmzt | okay, what device then? |
| 17:14.42 | tmzt | can you lsof it |
| 17:15.53 | phh | makkonen: have you tried anything for calibration ? |
| 17:16.04 | makkonen | phh: not yet. |
| 17:16.08 | phh | ok |
| 17:16.14 | phh | I'll then |
| 17:18.44 | makkonen | I'm getting a kernel panic coming from: [ 355.494964] kernel BUG at drivers/char/tty_ldisc.c:199! |
| 17:19.04 | makkonen | I think it happens when data comes up. anyone else seen it at any point? |
| 17:20.00 | *** join/#htc-linux rayman18 (n=opera@residL-102-227.uc3m.es) |
| 17:20.01 | tmzt | makkonen: device? |
| 17:20.05 | makkonen | raph800 |
| 17:20.14 | phh | cr2: is spl on nand protected by default ? |
| 17:20.37 | cr2 | no |
| 17:20.38 | sse2 | phh: no |
| 17:20.55 | phh | ouch. |
| 17:20.58 | cr2 | it's below 36MB |
| 17:21.05 | phh | uh ? then it's protected |
| 17:21.24 | cr2 | http://htc-linux.org/wiki/index.php?title=RaphaelNAND |
| 17:21.24 | sse2 | it's at 2400000 |
| 17:21.39 | cr2 | below is wrong word :) |
| 17:21.47 | cr2 | below in the wiki table |
| 17:21.59 | phh | and the lowest address I can read is 220000000. so I have to take good care when writing to nand. |
| 17:22.22 | phh | hum too many 0 |
| 17:22.28 | phh | oh it's in octal |
| 17:22.39 | cr2 | i need to entrer the calibration values there, and the crc algo. didn't find which one though, |
| 17:23.07 | cr2 | disable nand mpu, and you can do everything |
| 17:23.27 | phh | I just want not brick my phone :p |
| 17:23.47 | cr2 | spl is reflashable |
| 17:23.49 | chamonix | [11:59:28] PUBMSG #htc-linux: <= phh> chamonix: I don't think android devices has such a big problem :p <--- you have no idea.... there was a donut bug open from may till end of last year about multipart messages with different encoding (forbidded by the spec) and chars like c-cedille sucking 100 of SMS length |
| 17:24.17 | phh | cr2: only if I don't reboot meanwhile :p |
| 17:24.18 | cr2 | haha |
| 17:24.24 | phh | chamonix: ah. |
| 17:24.40 | phh | cr2: or if I have jtag |
| 17:24.45 | chamonix | didn't find a fix yet but some proposed patch.... have to see if they were meerged |
| 17:25.29 | cr2 | chamonix: you switch to utf16 if weird chars are entered ? |
| 17:25.46 | cr2 | chamonix: wince does the same btw. even for ' |
| 17:26.23 | phh | ouch |
| 17:26.24 | chamonix | we are back at slackware 0.2 where u did have to recompile the kernel to get anything else than a-z0-9 keyboard :-D |
| 17:26.40 | chamonix | :-X now u know how old I am ;) |
| 17:26.58 | phh | chamonix: have you flashed the light wince rom ? |
| 17:27.05 | chamonix | nope |
| 17:27.14 | phh | groumpf. |
| 17:27.22 | chamonix | no need, I never reboot |
| 17:27.24 | cr2 | chamonix: lol. i don't rememebr the slackware version, bt the kernel was 1.2.8 ? |
| 17:27.34 | cr2 | err |
| 17:27.40 | cr2 | it was 1.0.somethng |
| 17:27.49 | cr2 | because 1.1.xx was dev :) |
| 17:27.56 | chamonix | wow, u r that old too? |
| 17:28.13 | cr2 | heh. using linux snce 1994. |
| 17:28.26 | phh | I was 5 in 1994. |
| 17:28.34 | cr2 | lol |
| 17:28.44 | cr2 | and ued linux ? |
| 17:28.47 | chamonix | graduated in 93 |
| 17:28.49 | chamonix | HeHeHe |
| 17:28.51 | cr2 | s/ues/used/ |
| 17:29.01 | phh | cr2: I was so young I don't really remember :/ |
| 17:29.01 | cr2 | chamonix: '92 |
| 17:29.15 | chamonix | ... for the seond time :) |
| 17:29.16 | phh | you're damn old. |
| 17:29.19 | chamonix | second |
| 17:29.24 | cr2 | lol |
| 17:29.32 | phh | makkonen: I'll do it. |
| 17:31.11 | makkonen | phh: ok. sorry I haven't gotten to it. I'm probably going to be in and out today, but if I can help with anything, just shout. |
| 17:31.33 | cr2 | phh: certified windows hater since 3.1 |
| 17:31.44 | phh | cr2: I haven't even used windows 3.1 .... |
| 17:31.51 | phh | I've just seen 3.11 for less than 10minutes |
| 17:32.06 | cr2 | phh: i used x11 before 3.1 == 3.1 hater |
| 17:32.13 | makkonen | heh. back then I was rocking the Mac -- System 7. |
| 17:32.21 | phh | makkonen: oh right same |
| 17:32.28 | phh | I even updated it to Mac OS 8 |
| 17:32.31 | phh | 8.1 iirc |
| 17:32.32 | cr2 | but x11 was on vax |
| 17:35.02 | cr2 | phh: btw, the xserver was 286 with 3mb ram |
| 17:35.09 | phh | lol |
| 17:35.28 | cr2 | or 640+2mb isa ext card. |
| 17:35.38 | cr2 | don't remember nymore |
| 17:36.30 | cr2 | phh: and you tell me that raph with 256mb+32ms is not a server ;) |
| 17:36.47 | phh | it has no serial port. |
| 17:36.58 | cr2 | it has |
| 17:37.00 | sse2 | cr2: about the basband crash, try a sram dump (arm9 uses 32mb), or just place some arm interrupts there |
| 17:37.09 | cr2 | you just need a cable |
| 17:37.20 | phh | cr2: baahhh |
| 17:37.23 | phh | it has no ethernet :p |
| 17:37.46 | cr2 | phh: 10mbit aui ? |
| 17:37.51 | phh | aui ? |
| 17:37.56 | cr2 | lol |
| 17:38.04 | tmzt | I know :) |
| 17:38.14 | cr2 | aui is older than coax |
| 17:38.18 | tmzt | forget a though |
| 17:38.37 | phh | cr2: waw. |
| 17:38.43 | phh | kind of old then. |
| 17:38.50 | *** join/#htc-linux zycho (n=zycho@f054160186.adsl.alicedsl.de) |
| 17:39.30 | tmzt | older than coax ethernet |
| 17:39.40 | cr2 | http://en.wikipedia.org/wiki/Attachment_Unit_Interface |
| 17:39.49 | tmzt | and that's intergrated phy, transceiver/taps must have already been in use |
| 17:40.45 | cr2 | "subminiature" |
| 17:42.57 | phh | if you want to see my ugly rpc command patches btw, I pushed it in http://gitorious.org/linux-on-qualcomm-s-msm/linux-msm/commit/69cc611a9bec67b47abdf92eee2548a071976546 |
| 17:45.55 | cr2 | smd_rpcrouter_device.c are ifdef'd ? |
| 17:46.11 | phh | no. I want something working before doing other devices |
| 17:46.21 | phh | that's why I put it in another branch |
| 17:46.22 | cr2 | ok |
| 17:47.57 | cr2 | phh: can you try the "normal" board-htcraphael-gps ? |
| 17:48.16 | phh | cr2: we will have the same problem no ? |
| 17:48.39 | cr2 | maybe |
| 17:48.56 | phh | bah we always had ARM9 reboot with it |
| 17:49.00 | phh | and we still can't handle it properly ... |
| 17:49.05 | cr2 | btw, running linux with dead arm9 can be considered a security feature ? |
| 17:49.14 | cr2 | shuld apply for B2 cert :) |
| 17:49.20 | phh | uh ? why ? |
| 17:49.28 | phh | I can't see how "secure" this is |
| 17:49.42 | cr2 | you can't attack it over RF |
| 17:49.49 | phh | lol |
| 17:49.55 | phh | arm9 isn't done btw. |
| 17:49.57 | phh | it's only dex |
| 17:50.04 | cr2 | ok |
| 17:50.35 | phh | I'm not sure of what arm9 does though, but rpc still works |
| 17:50.39 | cr2 | dead dex means you can't use pmic |
| 17:51.02 | phh | it means the cpu is blocked for 10s at every dex try :p |
| 17:51.59 | cr2 | phh: i'll add your get_position call to wiki |
| 17:53.28 | *** part/#htc-linux rayman18 (n=opera@residL-102-227.uc3m.es) |
| 17:54.27 | *** join/#htc-linux writemefree (n=writemef@207-47-237-190.regn.hsdb.sasknet.sk.ca) |
| 17:55.36 | cr2 | phh: has 28 args too |
| 17:57.56 | cr2 | compared to kaiser, it looks like they reshuffle the params a lot |
| 17:58.12 | phh | and compared to your raph dumps ? |
| 17:59.06 | cr2 | too |
| 17:59.18 | phh | arf |
| 18:00.54 | cr2 | but your get position was a failure |
| 18:01.08 | cr2 | [ 230.034149] oncrpc read |
| 18:01.09 | cr2 | [ 230.034179] 0000000c 00000001 00000000 00000000 00000000 00000000 00000001 |
| 18:01.50 | cr2 | but crashed onl after a cb |
| 18:03.48 | cr2 | the first 2params are like in the get_pos call |
| 18:03.51 | *** join/#htc-linux malgon (n=simon@ABordeaux-257-1-105-90.w90-50.abo.wanadoo.fr) |
| 18:04.17 | phh | I forgot that a pixel on this screen is really tiny. |
| 18:05.00 | tmzt | where are you drawing pixels? |
| 18:05.07 | phh | msm_fb |
| 18:05.11 | phh | for calibration software |
| 18:05.14 | tmzt | at what point |
| 18:05.15 | tmzt | oh |
| 18:05.30 | phh | with direct mmap on fb device |
| 18:05.51 | tmzt | yeah |
| 18:07.00 | cr2 | phh: you may divert the init to x2 instead of 1 now |
| 18:08.43 | phh | drawing while android is drawing is everything but a good idea. |
| 18:08.54 | *** join/#htc-linux MrPippy (n=pip@adsl-75-37-167-3.dsl.sndg02.sbcglobal.net) |
| 18:09.08 | chamonix | this is crazy, I can't find any commit fÃÅr the SMS encoding thing |
| 18:09.19 | chamonix | s/fÃÅr/for/ |
| 18:12.52 | cr2 | phh: you have managed several get_position calls before arm9 was unhappy |
| 18:13.02 | phh | really ?!? |
| 18:13.45 | phh | which dmesg are you reading ? i can see only one |
| 18:14.43 | *** join/#htc-linux babijoee (n=joee@110.33.178.90) |
| 18:14.55 | phh | hi BabelO |
| 18:14.57 | phh | babijoee: * |
| 18:15.00 | babijoee | hi guys |
| 18:15.11 | phh | babijoee: you're running the light rom ? |
| 18:15.20 | babijoee | nope |
| 18:15.22 | babijoee | i'll flash it now |
| 18:15.35 | phh | arf |
| 18:15.40 | babijoee | :) |
| 18:15.42 | phh | I need a tester for the backlight thing |
| 18:16.07 | babijoee | gimme 2 mins |
| 18:17.38 | babijoee | backlight for? |
| 18:18.51 | babijoee | ok done |
| 18:19.40 | phh | bah you have no backlight diming with this ROM normally |
| 18:20.06 | babijoee | true |
| 18:20.16 | phh | try doing echo 3 > /sys/class/htc_hw/test |
| 18:20.51 | babijoee | -.-" |
| 18:20.58 | babijoee | i'll need to flash a full rom |
| 18:21.22 | phh | uh ? |
| 18:21.23 | phh | why ? |
| 18:21.35 | babijoee | backlight is on automatic |
| 18:22.01 | babijoee | or it wont matter |
| 18:22.02 | babijoee | ? |
| 18:22.04 | phh | it won't matter |
| 18:22.07 | babijoee | okay |
| 18:22.10 | phh | it should stop auto. |
| 18:22.16 | babijoee | ooo |
| 18:22.29 | babijoee | put that line in rootfs/init |
| 18:22.31 | babijoee | right |
| 18:22.38 | phh | bah run it whenever for test |
| 18:24.35 | phh | babijoee: we had some progress on gps and camera today, but it's still not working. |
| 18:24.52 | babijoee | ah ok |
| 18:25.03 | babijoee | still progress :) |
| 18:26.36 | phh | ToAsTcfh: did you get 3D working btw ? |
| 18:29.01 | babijoee | phh: when you plan on buying nexus one? |
| 18:29.21 | phh | when I'll have money for it. in years I guess. |
| 18:29.53 | babijoee | :( |
| 18:30.05 | babijoee | everyone i show is so envious |
| 18:30.56 | cr2 | lol |
| 18:32.48 | babijoee | phh: it works |
| 18:32.57 | phh | cool |
| 18:35.22 | babijoee | is ToAsTcfh backporting the donut 3d? |
| 18:35.32 | phh | babijoee: yes |
| 18:35.40 | phh | to hero cdma, but it would be the same result for us :D |
| 18:35.46 | babijoee | ooo |
| 18:35.48 | babijoee | cool |
| 18:36.18 | babijoee | doesnt hero cdma have working 3d? |
| 18:36.24 | phh | not on eclair |
| 18:36.29 | babijoee | ooo |
| 18:36.36 | phh | no camera either |
| 18:36.43 | babijoee | waaa HTC fail |
| 18:36.55 | phh | lol |
| 18:38.29 | phh | don't worry you'll have the same problem with 2.5 |
| 18:38.35 | babijoee | at the moment all we need is big developer for games on Android |
| 18:39.14 | babijoee | 2.5? |
| 18:39.33 | babijoee | is that already announced |
| 18:39.40 | phh | not yet. |
| 18:39.55 | babijoee | whats its called? |
| 18:40.19 | phh | not even called yet I guess :p |
| 18:40.31 | babijoee | ah k |
| 18:40.43 | babijoee | do u feel there is any need to update to 2.1 Flan? |
| 18:40.51 | babijoee | i honestly dont think there is |
| 18:41.00 | phh | I don't know |
| 18:41.10 | babijoee | i cant see any features that are better |
| 18:44.54 | sse2 | called froyo - 2.1 is still eclair |
| 18:51.28 | babijoee | sse2: 2.1 = flan 2.0.1 is eclair |
| 18:51.37 | GNUtoo | hi leviathan |
| 18:52.08 | GNUtoo | leviathan, someone named Thingol in #alsa told me how to fix it....and it seem to work flawlessly!!!!!!!!!!!!!!!! |
| 18:52.11 | GNUtoo | -ao alsa |
| 18:52.14 | GNUtoo | let's try intone |
| 18:56.50 | *** join/#htc-linux zycho_ (n=zycho@f054160186.adsl.alicedsl.de) |
| 18:57.32 | *** join/#htc-linux x29a (n=x29a@unaffiliated/x29a) |
| 18:57.34 | *** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a) |
| 18:58.31 | sse2 | babijoee: nope - actually a -b eclair sync gives 2.1 features (3d gallery...) |
| 18:59.58 | Markinus | phh: I corrected the patch and switched to AMSS detection, all is checked in! |
| 19:03.22 | cr2 | Markinus: have you tried to enable gps ? |
| 19:04.36 | Markinus | cr2: today? No, this isn't work, or? I can do it . . |
| 19:05.07 | *** join/#htc-linux luminoso_ (n=lumos@av-217-129-128-10.netvisao.pt) |
| 19:05.45 | cr2 | Markinus: it will probably crash, but you will get an interesting log |
| 19:07.15 | Markinus | ahh, yes, to much to do in the moment, I forgot it ... I do it now! |
| 19:08.13 | cr2 | phh: btw, the PDAPI_CB responses are fragmented on raph (len=111). there was some rpcrouter bug fixed in the codeaurora tree |
| 19:09.13 | phh | cr2: oh |
| 19:09.36 | cr2 | phh: or in the 2.6.29 tree, don't remember |
| 19:10.03 | cr2 | afair th rpc packet must be <512 bytes |
| 19:12.42 | phh | I can't see anything like that :/ |
| 19:13.30 | Markinus | cr2: need a kernel start with angstrom or a activate with a android GPS application too? |
| 19:13.31 | phh | cr2: any info about battery ? |
| 19:13.41 | phh | Markinus: angstrom + cat /dev/vogue_gps |
| 19:13.51 | Markinus | o.k. |
| 19:14.07 | leviathan | GNUtoo: really? |
| 19:14.09 | ToAsTcfh | phh: any luck on the camera? |
| 19:14.22 | leviathan | how can we fix this issue? |
| 19:14.29 | phh | ToAsTcfh: it won't be of any help for you, but it's progressing |
| 19:14.33 | leviathan | GNUtoo: please send me the patch! |
| 19:14.51 | phh | ToAsTcfh: now I do have /dev/msm_camera/{control0,config0,frame0} :p |
| 19:15.00 | ToAsTcfh | yeah i know. i just seen the logcats u posted a bit ago |
| 19:15.19 | ToAsTcfh | thats sweet though |
| 19:15.47 | ToAsTcfh | are u using the dream hacked libcamera |
| 19:15.57 | phh | no idea :D |
| 19:16.09 | phh | ToAsTcfh: oh but I do my tests on donut |
| 19:16.52 | ToAsTcfh | ok yeah the dream guys just forward ported the 1.6 libcamera to 2.0 |
| 19:16.59 | Markinus | cr2: I activate this driver in the topaz config: |
| 19:17.01 | Markinus | static struct platform_device topaz_gps = { |
| 19:17.02 | Markinus | <PROTECTED> |
| 19:17.04 | Markinus | }; |
| 19:17.11 | phh | Markinus: not gps_rfkill but htcraphael_gps |
| 19:17.18 | Markinus | ahh, o.k. |
| 19:17.25 | leviathan | GNUtoo: online? |
| 19:17.28 | leviathan | GNUtoo: if yes |
| 19:17.29 | ToAsTcfh | phh: thats good though. good luck |
| 19:17.49 | phh | Markinus: it's raphael_gps |
| 19:17.50 | leviathan | GNUtoo: just send me the commit as mbox, i'll import it and push it into the repo |
| 19:18.02 | phh | ToAsTcfh: so no result for you on 3D ? |
| 19:18.09 | ToAsTcfh | no |
| 19:18.25 | phh | maejrep: you tried ? |
| 19:18.57 | *** join/#htc-linux zycho (n=zycho@f054160186.adsl.alicedsl.de) |
| 19:19.11 | *** join/#htc-linux zhbvgt (n=zhbvgt@84-75-143-9.dclient.hispeed.ch) |
| 19:19.15 | ToAsTcfh | i guess the dream guys fixed up their kernel to work with the new opengl stuff |
| 19:19.15 | zhbvgt | http://www.pandamailer.de/?bettel=pimbolli |
| 19:19.15 | *** part/#htc-linux zhbvgt (n=zhbvgt@84-75-143-9.dclient.hispeed.ch) |
| 19:19.59 | GNUtoo | leviathan, I was eating |
| 19:20.04 | GNUtoo | ok |
| 19:20.05 | leviathan | oh |
| 19:20.06 | leviathan | ok |
| 19:20.07 | leviathan | sry |
| 19:20.16 | GNUtoo | 5s I've something to do but basically I'll explain |
| 19:20.21 | leviathan | ok |
| 19:20.26 | ToAsTcfh | the gsm hero guys are gonna steal apage or two from what cyanogen did to the dream kernel.and put it into their .27 kernel |
| 19:20.35 | GNUtoo | size_t alsa_send_buffer(struct msm_audio *prtd, const char __user *buf, size_t count, loff_t *pos,int copy_count) |
| 19:20.39 | GNUtoo | note the copy_count |
| 19:20.41 | phh | ToAsTcfh: no need to stole it from cyanogen. |
| 19:20.41 | GNUtoo | then do: |
| 19:21.02 | ToAsTcfh | exspression |
| 19:21.03 | GNUtoo | alsa_audio_configure(prtd); |
| 19:21.06 | GNUtoo | at the end |
| 19:21.16 | GNUtoo | (move it from where it was) |
| 19:21.19 | phh | ToAsTcfh: I mean that you can just use codeaurora's code |
| 19:21.20 | GNUtoo | and it works |
| 19:21.51 | ToAsTcfh | yeah i guess there not into it that deep |
| 19:22.06 | ToAsTcfh | but thats an idea |
| 19:22.26 | leviathan | GNUtoo: ok, thx! |
| 19:23.43 | leviathan | GNUtoo: uhhm |
| 19:23.51 | leviathan | which line do you call "end" |
| 19:23.52 | leviathan | ? |
| 19:24.06 | ToAsTcfh | phh: i guess joe is getting much closer on our kernel. so thats a positive |
| 19:24.15 | leviathan | GNUtoo: just before return |
| 19:24.19 | GNUtoo | I 'll send a commit |
| 19:24.24 | leviathan | ok |
| 19:24.24 | GNUtoo | no,at the end of while |
| 19:24.26 | leviathan | thx |
| 19:24.27 | leviathan | :) |
| 19:24.27 | phh | ToAsTcfh: wasn't your kernel supposed to be released last week end ? :p |
| 19:24.33 | ToAsTcfh | lol |
| 19:24.36 | ToAsTcfh | yeah |
| 19:24.42 | GNUtoo | also I've tested only pure alsa |
| 19:24.47 | ToAsTcfh | it was fraud |
| 19:24.53 | ToAsTcfh | a hoax |
| 19:24.53 | leviathan | GNUtoo: ok |
| 19:25.40 | ToAsTcfh | i guess if he can get smd right we will beat htc to the punch |
| 19:29.05 | Markinus | cr2 phh: Here is dmesg, but I see nothing.http://pastebin.com/md45f2e |
| 19:29.20 | Markinus | the cat to device is hanging, there is nothing comming |
| 19:29.37 | phh | something's wrong. |
| 19:29.58 | phh | Markinus: did you added the board-htcraphael-gps in makefile ? |
| 19:30.11 | Markinus | yes, I have the vague_gps under /dev |
| 19:30.28 | Markinus | and activated in the topaz board file |
| 19:30.53 | phh | ah. |
| 19:31.27 | phh | and the dmesg is complete ? |
| 19:32.05 | cr2 | Markinus: something is missing |
| 19:32.14 | cr2 | Markinus: btw, mddi: registering panel S1D13774 |
| 19:32.28 | cr2 | you need to create a new panel file |
| 19:33.00 | cr2 | for your match mddi.mfr_name= 0xb9f6 mddi.prod_code= 0x5582 |
| 19:33.11 | cr2 | it's not epson |
| 19:33.11 | Markinus | yes, dmesg -s 1000000 |
| 19:33.25 | phh | hum, codeaurora has some changes to mmc driver |
| 19:33.27 | Markinus | no? |
| 19:33.29 | phh | might have some things for us |
| 19:33.41 | cr2 | maybe novatek, but not the one used by google |
| 19:34.29 | cr2 | epson is used on blackstone and raph800 |
| 19:34.44 | cr2 | Markinus: and you need to query the lcd panel id |
| 19:35.03 | cr2 | set gpio0x55=1 |
| 19:35.15 | cr2 | 0x39 and 0x3a to input |
| 19:35.31 | cr2 | and then dump the 0x39 and 0x3a values |
| 19:35.31 | Markinus | in the panel file was: |
| 19:35.34 | Markinus | { |
| 19:35.36 | Markinus | .product_id = (0xb9f6 << 16 | 0x5582), |
| 19:35.37 | Markinus | .name = "S1D13774", /* TODO */ |
| 19:35.39 | Markinus | .id = 0, |
| 19:35.40 | cr2 | id0 and id1 |
| 19:35.40 | Markinus | .client_data = &epson_client_data, |
| 19:35.41 | chamonix | NO_HZ kernel and potential fix for navipad ready... testing |
| 19:35.41 | Markinus | .clk_rate = 0, |
| 19:35.43 | Markinus | }, |
| 19:35.44 | Markinus | so I thought this is a epson . . |
| 19:36.05 | cr2 | <PROTECTED> |
| 19:36.17 | cr2 | just to make it work |
| 19:36.27 | chamonix | babijoee: around? |
| 19:36.28 | cr2 | epson panel code is a dummy too. |
| 19:36.34 | phh | chamonix: fix how ? |
| 19:36.41 | cr2 | so you can pocypaste it |
| 19:36.42 | phh | fix what I mean |
| 19:37.01 | Markinus | ok |
| 19:37.03 | phh | the fail at start, or the random behaviour ? |
| 19:37.10 | cr2 | for epson we have a datasheet, but nobody found time to write a proper panel driver |
| 19:37.39 | chamonix | phh: I've reduced some enable/disable stuff of irq to my sceanrio with wake on hard only to see if the phenomenon that I have (lights on sleep when mobile is in my pocket) can be fixed that way |
| 19:38.36 | chamonix | what I get is navi lights op during sleep, reacts on touch and breathe does not work, back arrow blinks instead |
| 19:39.20 | cr2 | phh: the crc32 table is the posix/ieee one. but they are doing something dirty with crc32 |
| 19:39.58 | chamonix | phh: did someone else confirm / deny the clock lagging with NO_HZ? |
| 19:40.00 | *** join/#htc-linux toi (n=toi@d54C2A96D.access.telenet.be) |
| 19:40.02 | Markinus | I'm offline now . .. bye |
| 19:40.23 | phh | chamonix: you ? :p |
| 19:40.40 | chamonix | nah, I confirmed it lagging with GENERIC_TIME :P |
| 19:40.48 | phh | NO_HZ needs GENERIC_TIME :p |
| 19:40.53 | chamonix | I know |
| 19:41.29 | chamonix | that's y I wanted to know the result from other testers.... it may work as we don't stay in idle for long anymore |
| 19:42.14 | chamonix | anyway, I'll test that because I want to see what powertop says once I start following its recommendations :) |
| 19:42.28 | chamonix | ACPI may be trickier tho |
| 19:42.30 | phh | :p |
| 19:42.42 | phh | why ? :D |
| 19:44.02 | chamonix | well I don't even know if it's supposed to work, I can't find much specific in msm tho |
| 19:44.22 | phh | ACPI is x86 specific /o\ |
| 19:44.23 | chamonix | s/much/anything/ |
| 19:45.35 | chamonix | that would explain an intel tool recommending it |
| 19:45.37 | chamonix | :) |
| 19:46.13 | chamonix | it's rather a tip for getting more info than a recommentation to save anything so I'll jump that one |
| 19:47.03 | phh | iirc every other tips are wrong for us |
| 19:47.21 | chamonix | next would be CONFIG_USB_SUSPEND |
| 19:48.21 | phh | LOL |
| 19:48.23 | phh | sorry. |
| 19:48.35 | phh | oh well. we might try to kill vbus. |
| 19:48.39 | phh | err not vbus |
| 19:48.44 | phh | but usb phy when no vbus |
| 19:50.34 | cr2 | phh: they crc32 the whole buffer. |
| 19:50.41 | cr2 | for wifi cal |
| 19:55.01 | cr2 | eor is xor ? |
| 19:56.37 | cr2 | crc32-n, n is "layer" |
| 20:02.38 | phh | chamonix: would you be able to do some nand tests ? |
| 20:03.01 | chamonix | what r the requirements? |
| 20:03.16 | phh | flashing michyprima's light rom V3 |
| 20:04.04 | chamonix | well, I never flashed my mobile but there is a start for everything.... do I need a windows box to do that or can I flash from SD? |
| 20:04.21 | phh | you can flash from sd |
| 20:04.28 | phh | and there is some linux tools |
| 20:04.33 | phh | chamonix: but you don't even have hspl then ? |
| 20:04.45 | chamonix | nope |
| 20:04.54 | chamonix | but I've read about it |
| 20:05.20 | GNUtoo | leviathan, email sent...so based against your kernel...+ bad mail address inside |
| 20:05.26 | GNUtoo | I was in a hurry |
| 20:05.37 | leviathan | ok |
| 20:05.54 | phh | chamonix: then do :p but for HSPL I don't think it's doable without windows |
| 20:06.28 | GNUtoo | leviathan, btw do you have more details about your fixes....your commit messages were not great |
| 20:07.55 | GNUtoo | leviathan, because with this fix I wonder which ones are necessary and which ones are not |
| 20:07.56 | leviathan | uhhm |
| 20:08.03 | leviathan | hmm |
| 20:08.04 | leviathan | yes |
| 20:08.19 | cr2 | phh: the gsensor cal data is some custom ioctl. like wifi, but with another number |
| 20:08.27 | leviathan | but my silent fix has nothing to do with yours |
| 20:08.33 | leviathan | its a totaly different problem |
| 20:08.39 | GNUtoo | leviathan, you've made many fixes |
| 20:08.43 | leviathan | hmm |
| 20:08.45 | leviathan | right |
| 20:08.59 | leviathan | but only to get cleaner code and better debug info |
| 20:09.01 | GNUtoo | good commit messages are important |
| 20:09.04 | GNUtoo | ok |
| 20:09.14 | leviathan | and to get rid of android specific stuff |
| 20:09.19 | GNUtoo | ah ok |
| 20:09.26 | leviathan | like this interface for the binary driver |
| 20:09.41 | GNUtoo | leviathan, mine is not binary |
| 20:09.42 | leviathan | /dev/smd |
| 20:09.46 | leviathan | yes |
| 20:09.51 | leviathan | sure |
| 20:10.00 | GNUtoo | leviathan, maybe I should try to rebase the changes against the old patch |
| 20:10.09 | leviathan | uhh? |
| 20:10.13 | GNUtoo | to see what works and what doesn't |
| 20:10.24 | leviathan | ahh, ok |
| 20:10.27 | GNUtoo | because it works fine but I've nearly imperceptibles crackling sometimes |
| 20:10.28 | leviathan | btw: I've pushed |
| 20:10.35 | GNUtoo | I wonder if it's not the silent thing |
| 20:10.40 | *** join/#htc-linux Dinde (i=kayser@sur-internet.net) |
| 20:10.45 | leviathan | IT IS the silent thing |
| 20:10.49 | GNUtoo | ok |
| 20:10.54 | leviathan | It should make it silenct |
| 20:10.58 | GNUtoo | so if I remove the silent thing I would get rid of it? |
| 20:11.05 | leviathan | that you do not hear the cracking on the end |
| 20:11.24 | leviathan | you would get rid of the working /dev/dsp too |
| 20:11.26 | *** part/#htc-linux Syrefaen (n=fistings@213.184.217.156) |
| 20:11.31 | GNUtoo | no I hear during songs but it's nearly imperceptible...I'm not even shure they are real |
| 20:11.34 | leviathan | as long as you do not implement silent format |
| 20:11.47 | GNUtoo | we would have to do that at the end |
| 20:11.52 | leviathan | removing the silent patch will kill /dev/dsp |
| 20:12.00 | leviathan | but ok |
| 20:12.01 | GNUtoo | I know |
| 20:12.23 | leviathan | ok |
| 20:12.27 | GNUtoo | it's just to try... |
| 20:12.54 | GNUtoo | basically you tried dsp with huge buffers right? |
| 20:15.13 | leviathan | yes |
| 20:16.00 | leviathan | buffer was not recognized correct |
| 20:19.04 | leviathan | GNUtoo: wonderfull |
| 20:19.09 | *** join/#htc-linux googleman (n=googlema@41.105.30.170) |
| 20:19.11 | leviathan | managed to change routing too! :D |
| 20:19.14 | googleman | hi all :) |
| 20:19.20 | leviathan | googleman: buuh |
| 20:19.37 | leviathan | I am GNU Man |
| 20:19.48 | googleman | i got an xperia X1 from where i begin with this phone to run linux on it ? |
| 20:20.19 | GNUtoo | leviathan, and so you fixed? |
| 20:20.27 | GNUtoo | with unecessary fixes |
| 20:20.32 | leviathan | googleman: http://wiki.openmoko.org/wiki/OpenMoko_on_HTC-Dream |
| 20:20.35 | GNUtoo | (that are unecessary now) |
| 20:20.55 | leviathan | googleman: no, it just worked with your fixes somehow |
| 20:21.07 | leviathan | but I would prefer to have something like a selection |
| 20:21.12 | leviathan | a listing in alsamixer |
| 20:21.15 | *** part/#htc-linux DarkMasterHalo (n=DarkMast@modemcable187.118-80-70.mc.videotron.ca) |
| 20:21.19 | leviathan | for every output |
| 20:21.26 | googleman | is dream and X1 some ? |
| 20:21.47 | leviathan | should work too, I guess ^^ |
| 20:21.55 | leviathan | is xperia a never one? |
| 20:21.56 | leviathan | mom |
| 20:23.16 | googleman | i check pdadb |
| 20:23.40 | googleman | any one tried to play with ril in linux phone ? |
| 20:23.52 | googleman | ril = radio interface |
| 20:23.58 | GNUtoo | googleman, I did but only basic things |
| 20:24.05 | GNUtoo | for htcdream |
| 20:24.16 | googleman | is there way to emulate simcard ? |
| 20:24.25 | GNUtoo | cals,pin,ussd,sms worked but not much more |
| 20:24.59 | googleman | emulating simcard is nice :D |
| 20:27.16 | *** join/#htc-linux ToAsTcfh (n=toast@110-43.200-68.tampabay.res.rr.com) |
| 20:27.23 | googleman | leviathan, xperia is wm based os ? |
| 20:27.37 | *** part/#htc-linux c2d (n=Mike@port-92-200-10-104.dynamic.qsc.de) |
| 20:28.00 | leviathan | I thought it would also be either msm or omap based |
| 20:28.11 | leviathan | so have a look on the page of openembedded |
| 20:28.14 | leviathan | uhhm |
| 20:28.18 | leviathan | wait a moment |
| 20:28.39 | googleman | it have some cpu |
| 20:28.54 | googleman | only x1 have 512/256 memory |
| 20:31.24 | leviathan | uhhm |
| 20:31.27 | leviathan | btw |
| 20:31.32 | leviathan | its a sony device |
| 20:32.14 | MrPippy | its made by htc http://wiki.xda-developers.com/index.php?pagename=HTC_Kovsky |
| 20:33.20 | chamonix | phh: I will, I just got to get hands on a windows machine with local admin rights at work as I assume installing that won't work from VMWare will it? |
| 20:33.29 | phh | chamonix: it will |
| 20:33.58 | chamonix | cool, let's search for a winxd iso then :) |
| 20:34.06 | chamonix | s/xd/xp |
| 20:35.55 | leviathan | googleman: try haret |
| 20:36.57 | MrPippy | and support for kovsky is in our tree, don't know how well it works but its there |
| 20:37.07 | cr2 | phh: how many values do you need for calibration ? |
| 20:37.14 | phh | MrPippy: there is a fork tree |
| 20:37.17 | phh | cr2: screen calib ? |
| 20:37.22 | googleman | with rootfs of openmoko ? |
| 20:37.26 | cr2 | phh: yes. 10 values in nand |
| 20:37.40 | phh | cr2: 10 ? 5 values *{x,y} ? |
| 20:37.51 | cr2 | phh: maybe |
| 20:38.09 | phh | if we don't have the (x,y) screen points it's not useful |
| 20:39.01 | cr2 | phh: the gsensor cal is stored at the end of conf area. but there are 3 options (12 values, with crc32) |
| 20:39.15 | cr2 | phh: corners and the center ? |
| 20:39.25 | phh | cr2: not real corners |
| 20:39.30 | phh | it's something like 40pixel from the border |
| 20:39.53 | cr2 | phh: screen cal has some weird parity algorithm |
| 20:40.21 | phh | cr2: maybe we have them. |
| 20:40.33 | cr2 | phh: it think it's 0,0 ; 640,0 ; 480,0 ; 640,480 |
| 20:40.48 | phh | with 5 points ? :p |
| 20:40.53 | phh | I'd bet it's not. |
| 20:40.57 | cr2 | phh: middle too |
| 20:41.14 | cr2 | phh: dump the conf data (0x400) |
| 20:41.45 | cr2 | and look at +0x3a4,3a8,4ac |
| 20:41.58 | *** join/#htc-linux Hellie (i=soulreap@wield.the.scythe.of.the.soulreaper.net) |
| 20:42.36 | cr2 | touch cal may be 0xf4-0x118, ir i got the offsets right. |
| 20:43.58 | cr2 | phh: anyway, the data is there. weird+crc32(gsensor)+crc32(wifi) |
| 20:44.12 | cr2 | we may even check the checksums |
| 20:44.48 | cr2 | i had a full nand dump somewhere |
| 20:45.01 | phh | cr2: I have one if you need |
| 20:45.12 | phh | http://htcandroid.xland.cz/dump-nand.7z |
| 20:45.21 | phh | (yes I'm crazy) |
| 20:45.25 | cr2 | how big ? |
| 20:45.42 | phh | 110MB |
| 20:46.08 | cr2 | i need the data behind spl http://htc-linux.org/wiki/index.php?title=RaphaelNAND |
| 20:46.12 | chamonix | phh: this can take a while, I must get hands on winxp sp2 to install .net fw |
| 20:46.16 | cr2 | and before wince et al. |
| 20:46.54 | cr2 | 4MB+128K |
| 20:47.26 | cr2 | it contains a lot of tech data, including ids, and debug |
| 20:47.43 | *** join/#htc-linux Sti_0239 (n=Where_is@155.214-244-81.adsl-dyn.isp.belgacom.be) |
| 20:48.52 | GNUtoo | leviathan, ok I think that most of your patches are necessary |
| 20:49.16 | GNUtoo | it doesn't work with the old linuxtogo + the rebased qualcomm patch |
| 20:49.54 | cr2 | phh: cmn r0,1 ? |
| 20:50.08 | cr2 | ~r0 what ? |
| 20:50.11 | GNUtoo | hi mickey|office |
| 20:50.17 | GNUtoo | we're close for the sound |
| 20:51.02 | cr2 | GNUtoo: not so fast :) are you familiar with audiopara3.csv? what is CE there ? |
| 20:51.19 | GNUtoo | I didn't say we were there...I said we were close |
| 20:51.41 | GNUtoo | I've no windows mobile/ce devices |
| 20:51.57 | GNUtoo | but maybe there is something similar under android |
| 20:52.33 | cr2 | GNUtoo: look for audiopara.cvs |
| 20:52.38 | GNUtoo | ok |
| 20:54.21 | tmzt | GNUtoo: so what's new? |
| 20:55.07 | GNUtoo | tmzt, alsa doesn't block anymore thanks to someone named Thingol in #alsa |
| 20:55.16 | googleman | is there chan for WM section ? |
| 20:55.31 | GNUtoo | on the sound quality issue we can still hear crackeling if you concentrate to ear them |
| 20:55.42 | GNUtoo | barbely noticiable but they are there |
| 20:58.42 | *** join/#htc-linux zycho_ (n=zycho@f054218011.adsl.alicedsl.de) |
| 20:58.49 | cr2 | GNUtoo: /system/etc/AudioPara4.csv ? |
| 20:58.57 | GNUtoo | I'll look |
| 20:59.02 | GNUtoo | I was booting android |
| 20:59.40 | GNUtoo | cr2, maybe that was a proprietary file that was removed... |
| 20:59.51 | GNUtoo | I don't know where I put it |
| 20:59.57 | *** join/#htc-linux zycho_ (n=zycho@f054218011.adsl.alicedsl.de) |
| 21:00.07 | cr2 | GNUtoo: these are audio parameters |
| 21:00.15 | GNUtoo | ok |
| 21:01.12 | cr2 | CE_PLAYBACK_HEADSET, CE_PLAYBACK_HANDSFREE, CE_SYS |
| 21:01.23 | cr2 | i'm interested what is "CE" here |
| 21:02.43 | tmzt | chamonix: you might be able to hspl with htcflasher |
| 21:02.50 | tmzt | after running sspl |
| 21:03.01 | tmzt | flash the full nbh |
| 21:03.17 | tmzt | in the ruu package |
| 21:04.34 | GNUtoo | cr2, I've only them in my backups....that is to say system.img |
| 21:04.49 | GNUtoo | should I mount it on ram? |
| 21:05.03 | tmzt | cr2: can we add a dbgsfs way to soft reset mddi pnale from linux userland? |
| 21:05.20 | chamonix | well, I'm only 104 KBs from SP2 on my WinXP image so I'll stick to go thru windows tmzt :) |
| 21:05.30 | cr2 | tmzt: /sys for fb ? |
| 21:05.31 | tmzt | maybe push a panel-init struct to be used? |
| 21:05.35 | GNUtoo | cr2, someone pasted me his file |
| 21:05.40 | cr2 | GNUtoo: on ram ? |
| 21:05.46 | cr2 | GNUtoo: ok |
| 21:05.49 | GNUtoo | yes there is mtd emulation on ram |
| 21:05.56 | GNUtoo | can I paste the link here? |
| 21:06.22 | cr2 | yes, why not |
| 21:06.32 | GNUtoo | http://pastebin.com/d768649ca |
| 21:06.52 | *** join/#htc-linux m3dlg (n=m3dlg@bb-87-81-252-83.ukonline.co.uk) |
| 21:07.09 | cr2 | F27,CE_SYS |
| 21:07.21 | cr2 | F26,CE_PLAYBACK_HANDSFREE |
| 21:07.31 | cr2 | what's "CE" ? |
| 21:09.05 | cr2 | phh: the dump was with mpu on, it seems |
| 21:09.14 | phh | yes |
| 21:09.20 | phh | you want with it off ? |
| 21:09.20 | cr2 | but i'm not interessted in this part anyway |
| 21:09.24 | phh | ok |
| 21:09.29 | phh | who cares about amss ? :p |
| 21:09.53 | cr2 | we could have discussed how to split the blocks and oob |
| 21:10.09 | cr2 | without lame shareware |
| 21:13.31 | sse2 | oob should be 16 bytes |
| 21:13.42 | sse2 | size* |
| 21:14.15 | cr2 | hm. 28, 3d, 6d and then crc32 |
| 21:15.18 | cr2 | sse2: looking for other things now. |
| 21:15.47 | cr2 | wlan eeprom |
| 21:16.08 | tmzt | what shareware? |
| 21:16.40 | tmzt | cr2: I think CE is pcm/wave |
| 21:17.26 | tmzt | cr2: to test init sequences without rebuilding/booting kernel |
| 21:22.15 | tmzt | GNUtoo: so what's different for alsa to work? |
| 21:22.26 | tmzt | what devices do you think will have it working |
| 21:22.41 | GNUtoo | I don't know yet I've only dream |
| 21:23.13 | GNUtoo | and you can still ear some cracks that are not very noticiable so not ready yet but we are close |
| 21:23.18 | GNUtoo | and no silence implemetnation |
| 21:23.28 | GNUtoo | I bet it would work on qdsp5 devices |
| 21:23.36 | GNUtoo | but I've not tested on other devices than dream |
| 21:28.34 | leviathan | GNUtoo: and this way of changing routing by changing a bar ist not very cool |
| 21:28.42 | GNUtoo | ok |
| 21:28.45 | leviathan | I move it once routing works |
| 21:28.51 | leviathan | second time |
| 21:29.01 | leviathan | every thing hangs with NULL pointer |
| 21:29.05 | leviathan | inacceptable |
| 21:29.09 | leviathan | needs some change |
| 21:29.10 | GNUtoo | it works for me |
| 21:29.13 | GNUtoo | strange... |
| 21:29.18 | leviathan | yes but |
| 21:29.31 | leviathan | I'd like to have something like a direct mute control |
| 21:29.33 | leviathan | you know |
| 21:29.39 | GNUtoo | ok |
| 21:29.41 | leviathan | boxes for every channel |
| 21:29.50 | leviathan | which you can mute/unmute |
| 21:29.58 | leviathan | with builtin logic |
| 21:30.06 | GNUtoo | good idea but we still need something like a bar which is not a bar |
| 21:30.07 | leviathan | mutes if some other chanel is unmutes |
| 21:30.17 | leviathan | uhhm |
| 21:30.25 | leviathan | no, ohne volume control over pcm |
| 21:30.28 | GNUtoo | in replicant I had this bug |
| 21:30.33 | leviathan | yes |
| 21:30.37 | leviathan | I have it atm |
| 21:30.46 | GNUtoo | it crashed becuase it was putting the audio in 2 output at the same time |
| 21:30.57 | GNUtoo | how do you fix that then? |
| 21:31.27 | leviathan | I'll create separate audionodes with only one volume control |
| 21:31.30 | leviathan | that would be PCM |
| 21:31.49 | cr2 | there is no crc32sum tool like md5sum and sha* ? |
| 21:32.08 | leviathan | means [PCM] bar [MIC] Nobar [HANDSET] nobar [..] nobar |
| 21:32.31 | leviathan | then you can go over MIC/HANDSET and mute/unmute |
| 21:32.37 | *** join/#htc-linux Boydell (n=Boydell@69.49.58.28) |
| 21:32.49 | leviathan | cr2: dunno |
| 21:34.26 | cr2 | libarchive-zip-perl: /usr/bin/crc32 |
| 21:35.44 | cr2 | omg, c bytes. i m dumb |
| 21:36.27 | leviathan | GNUtoo: what do you think of my idea? |
| 21:37.14 | GNUtoo | leviathan, what would prevent someone from unmuting mic and handset? |
| 21:37.27 | cr2 | leviathan: do you have the google calibration ATAGS dumped somewhere ? |
| 21:37.42 | leviathan | ATAGS? |
| 21:37.45 | leviathan | no |
| 21:37.47 | leviathan | why? |
| 21:38.28 | cr2 | we hve them in nand, i'd like to compare |
| 21:38.49 | cr2 | touchscreen calibration, acceleroemter calibration, and wifi nvram |
| 21:40.15 | *** join/#htc-linux yoyey (n=yoann@78.242.110.43) |
| 21:41.48 | cr2 | lol |
| 21:41.50 | cr2 | dd if=dump-nand of=gsensor.cal bs=1 skip=$((0x2400000+0xc1000+0x3c8)) count=$((0xc)) |
| 21:43.22 | leviathan | okee |
| 21:43.25 | leviathan | :D |
| 21:43.46 | leviathan | anyway |
| 21:43.54 | leviathan | I'll sleep now |
| 21:44.06 | leviathan | I need to get up early in the morning tomorrow |
| 21:44.19 | leviathan | I've got a meeting |
| 21:44.24 | leviathan | need to travel 1hour |
| 21:44.28 | leviathan | so |
| 21:44.29 | leviathan | gn8 |
| 21:44.33 | leviathan | :) |
| 21:44.41 | leviathan | oh |
| 21:44.45 | leviathan | and I need to reboot |
| 21:44.53 | leviathan | so if theres something new |
| 21:45.07 | leviathan | just contact me over email while I'm offline |
| 21:45.07 | leviathan | cu |
| 21:46.18 | cr2 | sse2: 16 after 512 block ? i have a samsung file here |
| 21:47.08 | cr2 | or was it 0x2000 block... |
| 21:48.25 | cr2 | 22:13:36 [D:GSNR] Pass CRC32-3.M |
| 21:48.26 | cr2 | 22:13:36 [D:GSNR] Assign g-sensor offset successM |
| 21:48.28 | cr2 | 22:13:36 [D:GSNR] Calibrate Offset: x = -60,y = 74,z = 107M |
| 21:48.29 | cr2 | 22:13:36 [D:GSNR] Read G-sensor calibration data from flash.M |
| 21:49.27 | cr2 | 28 3d 6d |
| 21:49.48 | *** join/#htc-linux Markinus (n=markus@gtng-4db0412a.pool.mediaWays.net) |
| 21:50.40 | cr2 | phh: 40 61 109 in your rom |
| 21:57.06 | *** join/#htc-linux c2d (n=Mike@port-92-200-10-104.dynamic.qsc.de) |
| 22:19.09 | Markinus | cr2: I looked a second time to the GPS on my topaz |
| 22:19.16 | Markinus | I fond somethink like sound |
| 22:19.22 | Markinus | found |
| 22:19.42 | Markinus | I have this mesage here |
| 22:19.43 | Markinus | vogue_gps_read waiting |
| 22:19.49 | tmzt | 7/wi 16 |
| 22:20.05 | Markinus | it is like sound, he is maiting for a answer or event but there is nothing cimming |
| 22:20.20 | Markinus | waiting |
| 22:20.23 | phh | huhu |
| 22:20.58 | tmzt | msm should be 4k blocks |
| 22:21.39 | tmzt | cr2: cat /proc/atag maybe but it might be disabled |
| 22:22.15 | cr2 | tmzt: the gsensor cal looks ok but i need to compare it with raw nand |
| 22:22.15 | tmzt | cr2: openssl might have one |
| 22:22.55 | cr2 | 00000028 0000003d 0000006d c9b9c474 |
| 22:23.04 | cr2 | the last one should be crc32 |
| 22:23.30 | tmzt | gnu: meaning all qdsp5 devices? this is alsa_qdsp? |
| 22:23.36 | tmzt | have you pushed it? |
| 22:23.51 | cr2 | Markinus: hmm. strange |
| 22:24.04 | tmzt | we need a cia bot |
| 22:24.12 | tmzt | for gitorious |
| 22:24.21 | phh | tmzt: i haven't found how to get it |
| 22:24.37 | tmzt | rss? |
| 22:24.46 | tmzt | cia.vc |
| 22:26.00 | phh | tmzt: http://gitorious.org/linux-on-qualcomm-s-msm/linux-msm.atom |
| 22:28.42 | *** join/#htc-linux rayman18 (n=opera@163.117.102.227) |
| 22:29.33 | makkonen | that commit from 2019 bugs me. |
| 22:29.52 | makkonen | ...people from the future should be doing more exciting things. |
| 22:30.08 | phh | :) |
| 22:30.10 | *** join/#htc-linux MSM (n=michael@222-155-71-177.jetstream.xtra.co.nz) |
| 22:32.26 | Markinus | I've a question, could someone explain me what the difference is between the normal RPC servers 30xxxxxx and the CB servers 31xxxxxx ? I don't see the CB servers in the debug rpc server output, there are anly 30xxxxxx. It's normal? |
| 22:33.19 | phh | CB are A9 called to A11 |
| 22:34.00 | cr2 | Markinus: the most 30* servers are remote (amss), with only 1 exception: 300000b is local (arm11) CB server registered dynamically by adsp code |
| 22:34.29 | phh | cr2: I love exceptions. |
| 22:34.44 | cr2 | the 31* CB servers are local (arm11), and you need to have the code which registers them, and processes the data sent my arm9/amss clients |
| 22:34.58 | cr2 | phh: blame qualcomm |
| 22:35.09 | phh | as usual |
| 22:35.16 | cr2 | if they will publish amss source, we may fix it :) |
| 22:35.26 | phh | that's a big if :p |
| 22:35.39 | cr2 | lol |
| 22:35.44 | sse2 | cr2: not gonna happen even in your wildest dreams |
| 22:36.01 | emwe | gnight. |
| 22:36.45 | Markinus | so I don't see the registraton of the CB servers in the logfile . . . |
| 22:38.09 | Markinus | what is with the AUDMGR_CB_VERS and AUDMGR_CB_PROG, the defines are there but there are never used, it's okay? |
| 22:38.40 | Markinus | or can we remove them? |
| 22:38.46 | cr2 | Markinus: braindead google code ;) |
| 22:38.58 | Markinus | ahh, :-=) |
| 22:39.16 | cr2 | Markinus: the registered local servers just don't spam the log |
| 22:39.43 | cr2 | but if amss will send something to them, you will see it in rpc logs |
| 22:39.56 | Markinus | ok |
| 22:41.51 | cr2 | Markinus: you may also see the registration message sent by rpcrouter to arm9 |
| 22:42.37 | cr2 | sse2: hehe. then we have removed a lot of junk from there too |
| 22:45.16 | Markinus | cr2: is the infrastucture completly closed or are there some homepages or documents, books, where I could lern more about this? |
| 22:45.38 | phh | there is only source code I think |
| 22:52.11 | *** part/#htc-linux rayman18 (n=opera@163.117.102.227) |
| 22:53.51 | *** join/#htc-linux mc (n=nathan@66.189.214.135) |
| 23:03.46 | *** part/#htc-linux yoyey (n=yoann@78.242.110.43) |
| 23:06.28 | *** join/#htc-linux Tinyboom_ (n=nahh@ti0121a340-dhcp0200.bb.online.no) |
| 23:11.33 | *** join/#htc-linux josch (n=josch@vr142.rayconnect.de) |
| 23:11.35 | *** join/#htc-linux tech9iner (n=hacim@unaffiliated/tech9iner) |
| 23:12.50 | josch | hi there :) |
| 23:13.52 | tmzt | hey |
| 23:14.27 | makkonen | phh: neat. someone on ppcgeeks found out that the calibration data stored in the winmo registry (HKLM\HARDWARE\DEVICEMAP\TOUCH] is of basically the same format as the calibration for the tssc_manager. The tuples are just reordered. |
| 23:15.00 | phh | makkonen: we "just" need to know the points used by wimo. |
| 23:15.11 | phh | anyway I don't want to rely on wimo |
| 23:15.11 | makkonen | heh |
| 23:15.15 | phh | I progressed on calibration |
| 23:15.33 | Boydell | see, this type of thing i was wondering, why cant we see how WinMo handles things, and try and use them(kinda).. |
| 23:15.33 | phh | it's almost working |
| 23:15.39 | phh | I just have to add a FB blit |
| 23:15.49 | phh | Boydell: we try. |
| 23:15.54 | phh | it's just not easy at all |
| 23:16.08 | *** join/#htc-linux theknigh (n=cfnzin@89.180.2.154) |
| 23:16.12 | Boydell | kernal-hardstuff isnt easy... |
| 23:16.14 | theknigh | hey everyone |
| 23:16.17 | Boydell | *hardware |
| 23:16.26 | theknigh | I'm sorry to bother, but is anyone here working on mustymod? |
| 23:16.41 | makkonen | theknigh: I've never heard that word here. |
| 23:17.25 | theknigh | its Eclair 2.0 for Samsung Galaxy... a person told me there wassomeone here working on it |
| 23:17.33 | *** part/#htc-linux tech9iner (n=hacim@unaffiliated/tech9iner) |
| 23:17.37 | makkonen | which means... i don't know. but not many, if so. |
| 23:17.53 | theknigh | kk... sorry to bother then :/ |
| 23:17.59 | tmzt | searc the logs |
| 23:18.11 | tmzt | site:irclog.netripper.com galaxy |
| 23:18.21 | tmzt | we haven't seen any patches |
| 23:18.30 | phh | makkonen: http://husson.hd.free.fr/calibrate.c |
| 23:18.32 | phh | that's my current code |
| 23:18.37 | tmzt | most of the 'eclair' stuff is similar to what ppp is working on |
| 23:18.39 | phh | I think all it lacks is a BLIT |
| 23:18.40 | tmzt | phh :) |
| 23:18.57 | tmzt | that would be awesome if you got blit working |
| 23:18.58 | toer | 22:41:49 < cr2> dd if=dump-nand of=gsensor.cal bs=1 skip=$((0x2400000+0xc1000+0x3c8)) count=$((0xc)) |
| 23:19.01 | toer | 22:43:21 < leviathan> okee |
| 23:19.02 | phh | tmzt: uh ? |
| 23:19.08 | toer | 22:43:24 < leviathan> :D |
| 23:19.08 | toer | 22:43:45 < leviathan> anyway |
| 23:19.09 | toer | 22:43:53 < leviathan> I'll sleep now |
| 23:19.09 | phh | tmzt: blit is damn easy |
| 23:19.11 | toer | 22:44:05 < leviathan> I need to get up early in the morning tomorrow |
| 23:19.12 | tmzt | phh: eclair |
| 23:19.13 | toer | 22:44:18 < leviathan> I've got a meeting |
| 23:19.16 | toer | 22:44:23 < leviathan> need to travel 1hour |
| 23:19.18 | phh | tmzt: blit in eclair ? |
| 23:19.19 | toer | 22:44:28 < leviathan> so |
| 23:19.21 | toer | 22:44:29 < leviathan> gn8 |
| 23:19.22 | phh | ok that's not damn easy :p |
| 23:19.23 | phh | toer: ? |
| 23:19.24 | toer | 22:44:32 < leviathan> :) |
| 23:19.26 | toer | 22:44:40 < leviathan> oh |
| 23:19.28 | toer | 22:44:44 < leviathan> and I need to reboot |
| 23:19.31 | toer | 22:44:52 < leviathan> so if theres something new |
| 23:19.34 | toer | 22:45:05 < leviathan> just contact me over email while I'm offline |
| 23:19.36 | toer | 22:45:06 < leviathan> cu |
| 23:19.38 | toer | 22:45:15 -!- leviathan [n=quassel@c-82-192-226-27.customer.ggaweb.ch] has quit [Client Quit] |
| 23:19.40 | phh | hum. |
| 23:19.42 | toer | 22:46:17 < cr2> sse2: 16 after 512 block ? i have a samsung file here |
| 23:19.44 | toer | 22:47:07 < cr2> or was it 0x2000 block... |
| 23:19.46 | toer | 22:48:24 < cr2> 22:13:36 [D:GSNR] Pass CRC32-3.M |
| 23:19.49 | toer | 22:48:26 < cr2> 22:13:36 [D:GSNR] Assign g-sensor offset successM |
| 23:19.50 | phh | toer: kill your stupid IRC client. |
| 23:19.51 | toer | 22:48:27 < cr2> 22:13:36 [D:GSNR] Calibrate Offset: x = -60,y = 74,z = 107M |
| 23:19.54 | toer | 22:48:29 < cr2> 22:13:36 [D:GSNR] Read G-sensor calibration data from flash.M |
| 23:19.56 | toer | ops |
| 23:20.15 | tmzt | phh: by fbflip29 doesn't work |
| 23:20.16 | toer | it was more than 5 lines so it should have ask me |
| 23:20.18 | toer | :\ |
| 23:20.23 | phh | tmzt: fbflip29 ? |
| 23:20.35 | tmzt | people.openezx.org/tmzt |
| 23:20.35 | phh | tmzt: it's MSMFB_BLIT |
| 23:20.42 | tmzt | oh |
| 23:21.02 | phh | read copybit.c for an implementation sample |
| 23:21.06 | phh | or my calibrate.c in 1m :p |
| 23:21.15 | tmzt | cool |
| 23:21.25 | tmzt | maybe getting X to work is doable |
| 23:21.49 | phh | hum, english homework to finish first |
| 23:21.56 | tmzt | gnutoo come back |
| 23:22.09 | tmzt | ? |
| 23:22.12 | tmzt | ? :) |
| 23:22.23 | chamonix | phh: async connection drop when installing hspl from vmware so I have to get hands on a box with windows host |
| 23:22.35 | phh | chamonix: arf |
| 23:22.54 | tmzt | chamonix: try htcflasher? |
| 23:23.05 | phh | tmzt: without even HSPL ? |
| 23:23.19 | chamonix | it seems to be due to the phone returning another usb id/vendor once the updater starts |
| 23:23.22 | tmzt | sspl |
| 23:23.41 | phh | chamonix: anyway to flash hspl, you do want a system as stable as possible |
| 23:23.51 | tmzt | true |
| 23:23.53 | chamonix | I guess so :) |
| 23:23.57 | phh | (meaning not windows, ok.) |
| 23:24.12 | chamonix | now where am I gonna find an xp sp2? |
| 23:28.12 | theknigh | anyone here worked on andromnia? or with sonique? |
| 23:29.20 | makkonen | ah. the touchscreen calibration settings from winmo give worse results for me anyway. oh well. I think winmo's calibration screen points are slightly closer to center. |
| 23:30.57 | makkonen | chamonix: I assume it's not too hard, if you're not looking for something legal. |
| 23:31.41 | chamonix | I meant a box with one running :) I guess I have to get an old disk and reinstall+update |
| 23:33.06 | Markinus | phh: I looked a bit more why the gps isn'T starting. The driver is registering successfully but the vogue_gps_ioctl is never called |
| 23:33.19 | Markinus | so the gps in't enabled and the rpc servers not started |
| 23:37.42 | *** join/#htc-linux leaigor (n=laigor@188.134.16.241) |
| 23:46.56 | IceBone | http://farm5.static.flickr.com/4011/4256841390_c765c2d597.jpg |
| 23:47.24 | makkonen | awww |
| 23:47.42 | IceBone | I want that so much, I'm afraid I'll start growing ovaries. |
| 23:54.50 | theknigh | lool |
| 23:55.03 | *** part/#htc-linux theknigh (n=cfnzin@89.180.2.154) |
