00:00.01 | BabelO | Kevin2: ok, can you advice me about an omap i2c thing ? |
00:00.10 | BabelO | it is short |
00:00.20 | Kevin2 | Sure - what's happening with it? |
00:01.09 | BabelO | today i disassemble htc_lib.dll from wm6 and compare i2c_init function to i2c-omap.c in current kernel |
00:03.00 | BabelO | i m looking at my office mail ;) |
00:03.15 | BabelO | where an i send you my debug file ? |
00:04.50 | Kevin2 | Can you pastebin it? |
00:05.50 | BabelO | Kevin2 sure |
00:12.23 | *** join/#htc-linux GvzEvxre (n=timr@pdpc/supporter/bronze/TimRiker) |
00:16.45 | BabelO | Kevin2: sorry i m looking for my email i sent |
00:19.52 | BabelO | Kevin2: found http://pastebin.ca/701350 |
00:25.04 | BabelO | and register for i2c are at virtual 0xB80B38xx |
00:29.26 | *** join/#htc-linux jott_ (n=j@unaffiliated/jott) |
00:32.21 | Kevin2 | Whoa.. Bitbake took over my system for the last five minutes.. |
00:32.46 | BabelO | Kevin2: often for me too on my poor athlon xp 1800 |
00:34.03 | Kevin2 | BabelO: Okay, I see your pastebin - what am I looking for? |
00:34.46 | BabelO | i don't understand the bus shift, pH5 say it is 2 |
00:35.07 | BabelO | at start i was agree with him, but not sure now |
00:36.55 | Kevin2 | Bus shift for i2c? |
00:37.21 | BabelO | Kevin2: yes, if you look at i2c-omap.c it is +4 |
00:37.44 | BabelO | why it is strange for you ? |
00:39.50 | Kevin2 | I'm not sure what a bus shift for i2c is. Where in i2c-omap is it set/used? |
00:41.21 | *** join/#htc-linux Hawk|- (n=Hawk@p5B17CC29.dip0.t-ipconnect.de) |
00:41.37 | BabelO | Kevin2: http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/drivers/i2c/busses/i2c-omap.c |
00:42.17 | BabelO | register are here and all are +4, but according to asm file i ve +2 with a strange register in middle |
00:46.46 | Kevin2 | Hrmm. How did you disassemble this? Some of the stuff looks odd - LDR R5, =0xB80B3802 - looks like it would throw an alignment fault. |
00:48.19 | BabelO | Ah Ah you point me on a bug ? i disassemble with IDA and just take interesting part of code |
00:48.29 | BabelO | why an alignement fault ? |
00:49.19 | Kevin2 | You're only supposed to do 4byte accesses on a 4-byte aligned address. |
00:51.01 | BabelO | yes, agree with you i just thinking that it was a compiler optim, using LDR one time instead of two successive retrieve |
00:57.29 | *** join/#htc-linux Babel (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
01:00.19 | Kevin2 | BabelO: I don't know, but it seems odd to me. You may want to confirm the asm output by just running cegcc's objdump on the dll. |
01:00.57 | BabelO | Kevin2: yes maybe they use LDR R2, =0xB80B3812 or LDRH R2,[3] with R3 correctly filled |
01:01.21 | BabelO | ok i ll try that tomorrow |
01:01.32 | BabelO | ah , maybe it is just a fault from testwm5.exe |
01:02.52 | Kevin2 | BabelO: Could be. But I doubt testwm5 would show something mostly correct.. |
01:03.23 | Kevin2 | There is no asm statement "LDR R2, =0xB80B3812" -- you can only load a value from a memory location - so IDL must be trying to make it look pretty for you.. |
01:04.16 | BabelO | ah ok, i ask cr2 about LDRH this afternoon because i just see LDR and STR in ARM asm code documentation |
01:06.50 | BabelO | Kevin2: good night and thanks for your help, i will try to continu compare kernel code to this dll and adjust |
01:07.04 | Kevin2 | Ok. Good night. |
01:08.07 | BabelO | just one more thing, what is cegcc command line to disassemble ? |
01:09.35 | Kevin2 | I use -- /opt/mingw32ce/bin/arm-wince-mingw32ce-objdump -d foo.dll |
01:11.52 | BabelO | ok, no more argument, thanks again ang good night again ;) |
02:06.58 | *** join/#htc-linux Hawk||- (n=Hawk@p5B17CC29.dip0.t-ipconnect.de) |
02:53.52 | ImCoKeMaN | anything new and exciting? |
02:56.56 | ImCoKeMaN | ok guess not |
05:27.45 | *** join/#htc-linux sadeness_ (n=LamersIn@nat2.rlan.ru) |
05:37.25 | *** join/#htc-linux pleemans (n=peter@leemanpe.ugent.be) |
06:14.40 | *** join/#htc-linux rob_w|laptop (n=rob_w@p549B9651.dip0.t-ipconnect.de) |
06:26.17 | *** join/#htc-linux TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
06:34.01 | *** join/#htc-linux youshe (n=youshe@jaalon.org) |
06:52.29 | *** join/#htc-linux oeichler (n=oeichler@p54922BFD.dip0.t-ipconnect.de) |
07:14.44 | *** join/#htc-linux kiozen (n=oeichler@p54922BFD.dip0.t-ipconnect.de) |
07:39.43 | *** join/#htc-linux snua12 (n=admin@113-70-207-85.strcechy.adsl-llu.static.bluetone.cz) |
08:33.29 | *** join/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt) |
09:17.59 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
10:10.25 | *** join/#htc-linux BabelO (n=Miranda@81.80.236.75) |
10:13.49 | BabelO | hi |
10:14.13 | *** join/#htc-linux _law_ (n=law@mail.stiftadmont.at) |
10:30.05 | BabelO | cr2 : for Qt / CE you have to wait a little ;) http://labs.trolltech.com/blogs/2007/06/ |
10:30.51 | BabelO | rohhh http://wiki.openmoko.org/wiki/Qtopia_on_Neo_1973 |
10:33.04 | BabelO | and this one, not same look but for gnome lover http://pokylinux.org/#about |
11:05.52 | *** join/#htc-linux rob_w (n=bob@X07b9.x.pppool.de) |
11:08.24 | cr2 | BabelO: where is htc_lib.dll ? |
11:10.12 | BabelO | i think you already downloaded it from my ftp server |
11:13.53 | BabelO | cr2: : it is in Modules directory, so loaded in memory, i m looking at objdump from cegcc now |
11:32.53 | cr2 | BabelO: i'm downloading the qtopia4.3.0 preview now |
11:33.11 | cr2 | hmm. can you post the link to htc_dll again? |
11:34.11 | sadeness | guys, any progress with htc prophet? |
11:35.57 | BabelO | cr2 : i check preview yesturday night, everything is here :) very happy now, we just need a multiplexer plug for tty GPRS and tty modem and a phonevendors plugin for TI based modem, few hour of work i think |
11:36.15 | cr2 | BabelO: that's great. |
11:36.22 | cr2 | sadeness: it's omap850 ? |
11:36.39 | sadeness | cr2, i guess yes. |
11:36.54 | cr2 | i think we'll have some progress soon. |
11:37.08 | cr2 | it seems BabelO found the core problem with i2c. |
11:37.36 | BabelO | cr2: look here http://linuxbox2.free.fr/Htc_Lib.dll |
11:41.50 | *** join/#htc-linux rob__w (n=bob@X0fbb.x.pppool.de) |
11:42.21 | BabelO | cr2 : i think i go back to home very soon today, to test preview release ;) |
11:45.56 | BabelO | cr2 : do you understand what i mean about i2c register last night ? i don't know how to explain it |
11:47.35 | *** join/#htc-linux AC_Mast88 (i=s@gateway/tor/x-4ead732bf812d7a5) |
11:52.25 | *** join/#htc-linux rob_w (n=bob@X13e6.x.pppool.de) |
12:18.34 | dcordes | somebody want prophet image booting kernel? |
12:19.18 | sadeness | me me! |
12:19.21 | sadeness | pick me! |
12:19.22 | sadeness | :D |
12:19.42 | *** join/#htc-linux rob__w (n=bob@X1cf7.x.pppool.de) |
12:24.01 | *** join/#htc-linux youshe (n=youshe@jaalon.org) |
12:35.41 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
12:46.52 | *** join/#htc-linux kiozen (n=oeichler@p54922BFD.dip0.t-ipconnect.de) |
12:51.27 | *** join/#htc-linux _law_ (n=law@mail.stiftadmont.at) |
14:10.16 | *** join/#htc-linux rob_w (n=bob@X1cf7.x.pppool.de) |
14:44.21 | *** join/#htc-linux goxboxlive (n=goxboxli@139.80-202-160.nextgentel.com) |
15:15.40 | *** join/#htc-linux Othello (i=nobody@gateway/tor/x-3899b901e59f76f0) |
15:29.20 | *** join/#htc-linux dcordes_ (n=tsingtao@f049006007.adsl.alicedsl.de) |
15:42.14 | *** join/#htc-linux TimRiker (n=timr@216.49.181.128) |
16:07.36 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
16:18.00 | *** join/#htc-linux the_sys0p (i=nroberts@123.57.211.201) |
16:27.01 | *** join/#htc-linux CrAsH_x64 (n=crash@84.78.99.40) |
16:44.31 | goxboxlive | well, it's configured without bluetooth and dbus. At least i will have most of it compiled, and then just reconfigure half of it |
16:50.24 | *** part/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt) |
16:57.43 | BabelO | goxboxlive : ok :( how many time it take to compile ??? 3 hours ? |
16:58.14 | goxboxlive | yes as far as i remember |
16:58.33 | BabelO | good, so i go home and try myself ;) |
16:58.52 | goxboxlive | good |
16:58.59 | BabelO | bbl |
17:25.51 | *** join/#htc-linux Babel (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
17:37.23 | *** join/#htc-linux CrAsH_x64 (n=crash@84.78.99.40) [NETSPLIT VICTIM] |
17:37.23 | *** join/#htc-linux Kevin2 (n=Kevin@207-237-65-88.c3-0.avec-ubr12.nyr-avec.ny.cable.rcn.com) [NETSPLIT VICTIM] |
17:41.20 | goxboxlive | and it failes compiling too |
17:59.57 | Babel | goxboxlive: i never rememeber for depot path ? |
18:04.10 | goxboxlive | export QTOPIA_DEPTH_PATH=/PATH_TO_QTOPIA |
18:04.25 | goxboxlive | s/DEPTH/DEPOTH |
18:04.41 | Babel | goxboxlive: i m on it ;) |
18:05.21 | goxboxlive | export QTOPIA_DEPOT_PATH=~/projects/opie |
18:08.08 | *** join/#htc-linux snua12 (n=admin@113-70-207-85.strcechy.adsl-llu.static.bluetone.cz) |
18:18.10 | *** join/#htc-linux kiozen (n=oeichler@rgnb-4db08057.pool.einsundeins.de) |
18:21.09 | goxboxlive | Babel: how far are you? |
18:21.22 | goxboxlive | Have you manage to configure with dbus? |
18:31.23 | goxboxlive | custom.cpp fails , some power and brightness control stuff |
18:38.06 | *** join/#htc-linux dcordes (n=tsingtao@f049006007.adsl.alicedsl.de) |
18:38.36 | *** join/#htc-linux SmallR2004 (n=SmallR20@79-74-87-33.dynamic.dsl.as9105.com) |
18:39.02 | Babel | goxboxlive: yes ;) |
18:39.32 | Babel | i think i found why it fail, i m trying again now |
18:40.04 | Babel | goxboxlive: in custom.cpp you can use same as fic device, it is very simple :) |
18:56.18 | Babel | goxboxlive: dbus : OK :) now it is bluetooth :) |
19:11.01 | goxboxlive | hmm, i am trying to use the fic device and changing everything to htcphone :-) |
19:12.17 | Babel | goxboxlive: now it look for dbus.pc in lib/pkgconfig |
19:13.03 | goxboxlive | Babel: and not in lib/dbus ? |
19:13.44 | Babel | goxboxlive: yes too, but they use another file to know dbus version create in your dbuspath |
19:14.10 | Babel | <dbuspath>/lib/pkgconfig/dbus-1.pc |
19:14.20 | goxboxlive | k |
19:15.00 | Babel | with something like that http://pastebin.ca/702261 |
19:15.30 | Babel | then just try if it detect ok with<qtopia src>config.tests/dbus_versions.test script ;) |
19:17.36 | Babel | goxboxlive: and for bluetooth it look for bluetooth directory with include in mkspecs/qws |
19:17.45 | goxboxlive | ok |
19:24.59 | goxboxlive | Babel: i dont have dbus.pc |
19:25.04 | goxboxlive | I have never used it |
19:25.30 | Babel | goxboxlive: i know, it is new detection mechanism |
19:25.50 | goxboxlive | Where did you get it? OE? |
19:26.18 | Babel | goxboxlive: now i look at the detection script and create one from an other ;) |
19:26.29 | Babel | i pastebin it few lines ago |
19:28.05 | goxboxlive | ahh i get it |
19:28.18 | goxboxlive | thats why i have had problmes with tslib. |
19:28.28 | goxboxlive | it's hardcoded to /usr/local/arm :-) |
20:21.48 | goxboxlive | Babel: at last dbus is ok. Have you sorted out ssl? |
20:21.55 | goxboxlive | I have the same problme with thattoo |
20:22.14 | Babel | no slls for the moment ;) i want to make first test before bed ;) |
20:22.51 | goxboxlive | ok, seems like the openssl folder is not commited. It where commited in the 4.2.x . I aks ljp |
20:24.45 | goxboxlive | Babel: do we use multplexers? |
20:24.55 | Babel | goxboxlive: on ba yes because gprs data are on tts8 and modem is on ttyS0 |
20:25.52 | goxboxlive | SO actually i dont need muliplexers for universal and magician? But does it screw up the system if i have it? |
20:39.20 | Babel | goxboxlive: i don't think, it is just variable settings i f i remember well |
20:40.20 | goxboxlive | ok, i started to build it anyway. Its all an copy of GTA01, just changed the name (gta01->htcphone) and the tty ports |
21:03.44 | *** join/#htc-linux cr2 (n=konversa@crpl6.physik.uni-wuppertal.de) |
21:18.16 | *** join/#htc-linux bd2_ (n=cbou@89.113.85.104) |
21:25.18 | *** join/#htc-linux DTMonterrey (n=evandro@bl6-221-16.dsl.telepac.pt) |
21:37.40 | *** join/#htc-linux dcordes (n=tsingtao@f049172007.adsl.alicedsl.de) |
21:42.08 | cr2 | Babel: still here ? |
21:42.58 | Babel | Cryes |
21:43.10 | Babel | oups, no battery one keyboard |
21:43.14 | Babel | Cr2 yes |
21:43.57 | cr2 | what was the problem with the omap850 i2c ? |
21:44.35 | cr2 | there is no unaligned access there, it's just a style ida uses for picking the addresses |
21:44.36 | Babel | cr2: if i start counting i2c bus for omap-i2c.c with register each +2 |
21:45.12 | Babel | then disassembled code get a value between two register ... |
21:45.31 | Babel | maybe i miss something else |
21:47.34 | Babel | cr2: are you agree that 0xfffe1084 access setup i2c pins ?, and if i boot over haret , this setup is lost ? |
21:50.07 | cr2 | what is 1084 ? |
21:51.00 | cr2 | 0xfffe1084OMAP730_IO_CONF5 |
21:52.00 | cr2 | 3800-3840, shifted 3800-3820 |
21:52.39 | cr2 | do you see something outside of this range ? |
21:53.35 | Babel | cr2: i m looking again, because i see a register 3811 if i shift all bus and disassmbled code access to 3812 |
21:53.40 | Babel | maybe a bas calcul ;) |
21:54.07 | cr2 | 1084 is provided as a parameter to some subroutines. |
21:54.22 | cr2 | i don't see 3811 in your subroutines. |
21:54.52 | cr2 | looks like a very clean and simple asm anyway. |
21:55.52 | Babel | ok, i ve to look better when i count register shift... maybe i forgot to switch between 0x3808 to 0x380a ... |
21:56.17 | cr2 | ? |
21:56.32 | cr2 | these addresses are all hardcoded |
21:56.38 | cr2 | i don't see any offsets. |
21:56.58 | cr2 | OV R0, #0xA |
21:56.58 | cr2 | BL Sleep |
21:56.58 | cr2 | LDR R5, =0xB80B3802 |
21:56.59 | Babel | cr2: yes 0x3808 + 0x02 = 0x380a and not 0x3811 like i do i think |
21:57.07 | cr2 | LDRH R2, [R5] |
21:57.21 | cr2 | i wish all the wince code looked the same. |
21:57.57 | Babel | no, i just want to mean i have made a bad calcul whan adding haxadecimal values |
21:58.30 | cr2 | i'm only worried by such code |
21:58.32 | cr2 | if (cpu_is_omap16xx() || cpu_is_omap24xx()) { |
21:58.51 | cr2 | but i guess you can just extend it to your cpu. |
21:59.00 | cr2 | to include yours. |
21:59.27 | Babel | cr2: yes i was thinking to do like psokolovsky with platform parameters |
21:59.48 | Babel | a . shift and .cpu_type |
22:00.30 | cr2 | hm. try the same as omap MMC does |
22:00.36 | cr2 | then you'll know if it works. |
22:01.54 | *** join/#htc-linux huhlig (n=UhligH@68-189-77-13.dhcp.wtvl.ca.charter.com) |
22:02.00 | Babel | yes i can look |
22:02.12 | huhlig | /join #eclipse |
22:02.24 | huhlig | whoops |
22:02.40 | *** join/#htc-linux the_sys0p (i=nroberts@123.57.209.221) |
22:03.23 | cr2 | this is ugly |
22:03.26 | cr2 | if (cpu_is_omap15xx()) |
22:03.26 | cr2 | dev->rev1 = omap_i2c_read_reg(dev, OMAP_I2C_REV_REG) < 0x20; |
22:03.49 | Babel | cr2: yes and it seem omap850 do the same |
22:07.02 | Babel | cr2: i have xpanel to a comport under wince :( is that normal ? |
22:08.41 | cr2 | don't know. |
22:17.49 | ImCoKeMaN | did that second memory dump help out much? |
22:31.10 | ImCoKeMaN | guess cr2 left |
23:04.04 | *** part/#htc-linux DTMonterrey (n=evandro@bl6-221-16.dsl.telepac.pt) |
23:43.12 | Babel | good night |