| 02:58.39 | *** join/#htc-blueangel hlbot (n=adm@iclem.net) |
| 07:19.09 | *** join/#htc-blueangel awelux_ (n=awelux__@dslb-084-058-174-184.pools.arcor-ip.net) |
| 09:15.42 | *** join/#htc-blueangel jeanseb (i=jeanseb@ipv4.sabine-et-benoit.com) |
| 11:52.28 | *** join/#htc-blueangel zamboni (n=zamboni@206-248-129-72.dsl.teksavvy.com) |
| 11:55.29 | zamboni | coffee time.... |
| 11:55.53 | zamboni | so anyone here mind answering a potentailly dumb question? |
| 11:56.26 | zamboni | Can I make a phone call with any of the HTC handsets flashed with linux? |
| 12:03.46 | *** part/#htc-blueangel zamboni (n=zamboni@206-248-129-72.dsl.teksavvy.com) |
| 15:06.46 | *** join/#htc-blueangel jeanseb (n=jeanseb@lns-bzn-47f-62-147-133-94.adsl.proxad.net) |
| 15:30.59 | *** join/#htc-blueangel rob_w (n=bob@p213.54.223.223.tisdip.tiscali.de) |
| 15:30.59 | *** mode/#htc-blueangel [+o rob_w] by ChanServ |
| 15:48.10 | *** join/#htc-blueangel jeanseb (n=jeanseb@lns-bzn-47f-62-147-133-94.adsl.proxad.net) |
| 15:54.28 | *** join/#htc-blueangel lamikr (n=chatzill@aragorn.kortex.jyu.fi) |
| 17:15.38 | *** join/#htc-blueangel BabelOued (n=BabelOue@lun34-2-82-238-28-28.fbx.proxad.net) |
| 17:25.16 | lamikr | rob_w: Any changes to get gomunicator patch in to cvs? |
| 17:29.27 | rob_w | no time yet... hang on pls |
| 17:29.40 | rob_w | probably today evening |
| 17:41.48 | lamikr | ok. Have you asked btw. from the gomunicator repository? |
| 17:43.22 | rob_w | asked what ? |
| 17:44.38 | lamikr | To putting the gomunicator to own repo in handheld or some other location? |
| 17:48.06 | rob_w | yeah , that i did , i am talking to some people ,,, but we need to wait maybe 2 weeks .. thats fine as then i am done with my house |
| 17:49.08 | rob_w | by then i will note the other cvs admin that we move to a differen primary |
| 17:49.43 | lamikr | rob_w: ok, To which repository you are planning to move the gomunicator? |
| 17:50.07 | rob_w | to hh.org or similar ;-) |
| 17:50.24 | lamikr | angstrom? |
| 17:50.41 | rob_w | also i would love to make images with using the hh.org kernels .. patched ones though |
| 17:51.38 | rob_w | it should and has tobe possible to use one head and branch for all htc devices |
| 17:52.39 | lamikr | Are you currently sharing a lot of code with other hh devices? |
| 17:54.14 | rob_w | well the code should be fairly comparable and already divided but the developments happen on different BRANCH `s in the xanadux cvs |
| 17:54.14 | lamikr | As h6300 is omap, we desired to use omap kernel as a base. |
| 17:54.14 | rob_w | i find this not very convienient .. |
| 17:54.53 | rob_w | but it was basicly necessary as the devices come up ,, then you each of them is over the basic problems and boots we should think to move togethere as hard as possible |
| 17:55.07 | lamikr | I have actually liked to git with kernel. It is much easier to keep in sync with mainline in that way as almost all are nowadays using it. |
| 17:56.20 | lamikr | Currently I prepare and test our upstream updates by using git... And after everything is fine, I make diff against our svn repo and then patch and update the repo. |
| 17:58.16 | lamikr | rob_w: The svn is beats than cvs but git is most handiest at least with kernel. |
| 17:58.46 | rob_w | i am not used to git yet .. first contacts where down knocking |
| 18:00.02 | awelux_ | an uninitialized variable should always be 0 in C right? |
| 18:00.34 | lamikr | awelux: I think that's true in nowadays compilers. |
| 18:00.49 | awelux_ | Any suggestions why this could be not the case for a variable here? |
| 18:01.10 | awelux_ | becasue the console_drivers pointer contains garbadge. |
| 18:01.37 | rob_w | awelux_, no if it is unititialised then you never know whats in there ! |
| 18:01.55 | rob_w | if you test for it before that , you will see "garbage" |
| 18:02.17 | awelux_ | ok. Now I explicitely stated struct console *console_drivers = 0; |
| 18:02.23 | rob_w | specially pointers i think |
| 18:02.43 | awelux_ | but it's still garbadge. |
| 18:02.46 | awelux_ | strange |
| 18:03.00 | awelux_ | I will try to export it and check from a different code point. |
| 18:03.40 | lamikr | awelux: I was googling and it seems that they are NOT set to zero by default. |
| 18:04.33 | awelux_ | lamikr, oh. thanks. |
| 18:05.29 | lamikr | Yes, I was also remembering otherwise. I am used to init them always in my c-code as java requires that also. In java that is consirered to be an error. |
| 18:06.11 | awelux_ | yep. I thought I read in some c documentation that tit'S always initialized. But I usually don'T trust this either. |
| 18:06.39 | awelux_ | But I've no idea why the struct console *console_drivers = 0; doesn'T work |
| 18:06.53 | awelux_ | maybe I'm corrupting my memory somehow. |
| 18:07.48 | rob_w | awelux_, -Wall or -pedantic as compiler switch should show you warnings about that |
| 18:08.05 | awelux_ | k. good idea. |
| 18:08.06 | rob_w | maybe it is accessed even before your init |
| 18:11.17 | awelux_ | I'm currently at the first printk. |
| 18:11.42 | awelux_ | That's very early and the idea is that unless a console driver is loaded the messages will be stored |
| 18:11.53 | awelux_ | and then printed once the console driver is available. |
| 18:12.34 | awelux_ | Right now the check that the console driver is uninitialized fails because of the garbadge |
| 18:13.46 | rob_w | so the check for console comes back wrong ß |
| 18:14.08 | awelux_ | yep. |
| 18:14.41 | awelux_ | if (msg_log_level < console_loglevel && |
| 18:14.41 | awelux_ | console_drivers && start != end) { |
| 18:14.59 | awelux_ | should be false untill we really have a console driver. |
| 18:15.24 | rob_w | is there a else part on that if ? |
| 18:15.45 | awelux_ | no. |
| 18:16.10 | awelux_ | and, I've checked the value of console_drivers. |
| 18:16.18 | awelux_ | was 0x22 something |
| 18:16.30 | awelux_ | not a valid memory location. |
| 18:17.36 | lamikr | rob_w: So it would take 2 weeks for the gomunicator repo move. Do you mind if I ask write access for that from the xanadux admin? |
| 18:17.44 | lamikr | for that time. |
| 18:18.18 | rob_w | is it that urgent ,, but sure i dont mind that .. free world , right .. |
| 18:18.49 | rob_w | i would like to ask you to wait , as then we can toghether give it a concentrated push |
| 18:19.56 | lamikr | rob_w: Well, ok. I am just quite ancient to start hacking that now when I have phone working. |
| 18:19.58 | rob_w | man the xanadux cvs kernel is also broken .. tsts |
| 18:20.39 | rob_w | well go hack it , but just wait with the cvs commits .. |
| 18:20.47 | *** join/#htc-blueangel jeanseb (n=jeanseb@lns-bzn-47f-62-147-133-94.adsl.proxad.net) |
| 18:20.52 | lamikr | rob_w: Ok. |
| 18:20.58 | rob_w | cool |
| 18:21.54 | lamikr | rob_w: Has you or others thinked btw. the separation of the the command layer to own lib from the gui? |
| 18:23.02 | lamikr | It seems that some people in gpe are trying to move non-ui libs to lgpl while keeping apps itself in gpl. |
| 18:38.01 | awelux_ | ok. I checked the variable right after entering start_kernel and it's already messed up. |
| 18:38.32 | awelux_ | Maybe a mistake while copying the kernel. |
| 18:39.49 | awelux_ | c01fb3e8 <console_drivers>: |
| 18:39.49 | awelux_ | c01fb3e8: 00000000 andeq r0, r0, r0 |
| 18:40.12 | awelux_ | so it should be 0 when untouched. |
| 19:05.13 | rob_w|bath | awelux_, it could be "messed up " as it is not initialized .. |
| 19:05.30 | awelux_ | rob_w|bath, no |
| 19:05.47 | awelux_ | when the kernel is copied the 0 should be copied as well |
| 19:06.01 | awelux_ | and when this memory location is accessed the 0 should be there |
| 19:06.04 | awelux_ | but it is not |
| 19:06.14 | awelux_ | I've to check the kernel copy. |
| 19:06.35 | rob_w | go make it 0 on your own .. |
| 19:06.57 | awelux_ | yes. That's a plan. |
| 19:07.10 | awelux_ | But I doubt that this is the only messed up variable |
| 19:12.47 | rob_w | i dont get why you need to depend on a preset variable even before start |
| 19:13.23 | rob_w | lamikr, did you mean a command layer for at and such ? |
| 19:14.17 | awelux_ | rob_w, the kernel seams to depend on the variable bein zero when untouched. |
| 19:15.18 | rob_w | lamikr, maybe we need to define action profiles .. so if a gui wants to turn off gsm or such calls a function and all is handled behind that |
| 19:15.44 | lamikr | rob_w: Yes, somekind of abstraction of commands like init, make call, send sms, receive sms, etc... |
| 19:15.50 | rob_w | awelux_, i wouldnt like this .. as i dont get it that this is good ! |
| 19:16.15 | awelux_ | rob_w, hehe. But that's how the linux kernel works |
| 19:16.38 | lamikr | rob_w: Yes, and how about sim-number input.. |
| 19:16.43 | awelux_ | there is no real data segment |
| 19:17.04 | awelux_ | the kernel is copied together with space for variables to memory |
| 19:17.18 | lamikr | rob_w: Then the interface between alsa and gomunicator should be specsed. |
| 19:17.30 | awelux_ | and then the kernel can use the known address of this memory to store something. |
| 19:17.36 | rob_w | lamkir , those profiles only be a help if we need to do many different commands on such situatiuons.. but that is only the case at real init or so |
| 19:17.47 | awelux_ | that's fine since we don't have malloc at the beginning. |
| 19:19.08 | awelux_ | and zeros compress well with gzip |
| 19:19.15 | rob_w | but then the if loop checking console_drivers existing m doesnt make sense to me .. as they "exist" prior "the big bang" |
| 19:19.28 | lamikr | I tracked the sms message send in last year. Even the h6300 and ba used basically same at commands, they required different sequence for sending the message. |
| 19:19.45 | awelux_ | rob_w, just the pointer exists prior to big bang |
| 19:19.49 | rob_w | lamikr, aha |
| 19:19.59 | awelux_ | the structure is filled much later |
| 19:20.03 | rob_w | then this is a nice need example |
| 19:20.27 | rob_w | awelux_, interesting and new |
| 19:21.02 | awelux_ | I'm probably the first in a few month who had to look at this part ;) |
| 19:21.17 | rob_w | but what if that if loop will be forced to break until those structures are filled acuratly ? |
| 19:21.40 | rob_w | say you zero it asap .. or so |
| 19:21.58 | awelux_ | then the buffer will not be written to console. |
| 19:22.30 | awelux_ | but it will be written to console once the console driver is up. |
| 19:22.57 | rob_w | until the console structure is generated , right ? |
| 19:22.58 | awelux_ | the very first kernel messages are not written immediately but delayed. |
| 19:23.03 | awelux_ | right |
| 19:23.20 | rob_w | sounds logical |
| 19:23.22 | awelux_ | The console structure is generated early nbut not that early |
| 19:23.42 | rob_w | so and your kernel is stopping there as it already adresses some dummy space .. |
| 19:24.20 | rob_w | ~seen cr2 |
| 19:24.32 | apt | cr2 <n=konversa@crpl22.physik.uni-wuppertal.de> was last seen on IRC in channel #htc-blueangel, 1d 19h 19m 49s ago, saying: 'awelux, i have identified the IR transceiver on universal.'. |
| 19:24.50 | awelux_ | rob_w, yep. I guess that's what is happening |
| 19:31.25 | awelux_ | ah. might have found the problem. there is a seperate section for the variables after all. |
| 19:31.34 | awelux_ | so I told some bogus ;) |
| 19:32.02 | awelux_ | I had a few instructions cleaning that area. |
| 19:32.21 | awelux_ | And I deleted them because they failed and I didn't understand them |
| 19:32.56 | lamikr | awelux_: Let's hope that will work! |
| 19:32.58 | awelux_ | hmm. soun d a bit strange if you have to explain the reasoning. |
| 19:33.25 | awelux_ | That's the funny thing when you drink wine and program at the same time. |
| 19:33.34 | awelux_ | You never know what youw ill find the next day. |
| 19:36.18 | awelux_ | much better. |
| 19:57.25 | rob_w | cool |
| 19:58.49 | rob_w | awelux_, sofar , how would you estimate is your branch away from the blueangel one ? would you be able to code it so it goes in parallel with the other htc devices .. board id , CPU type like ? |
| 19:59.04 | awelux_ | so far it's almost identical |
| 19:59.22 | awelux_ | But I changed a lot of code to debug the boot process |
| 19:59.34 | awelux_ | if I remove that it's almost the same |
| 19:59.52 | rob_w | would you be able to make it either identical or be able to decided upon the device ? |
| 20:00.05 | awelux_ | identical should work |
| 20:00.22 | awelux_ | So far I first had the problem with USB messing up memory |
| 20:00.43 | awelux_ | And while I was trying to find that I messed up other parts |
| 20:01.34 | awelux_ | So if the USB is switched of properly by haret the kernel is basically the same |
| 20:02.03 | rob_w | haret still only works right after a soft reset .. does anyone know details for that ? |
| 20:02.27 | awelux_ | so one soft reset is needed for haret ? |
| 20:02.48 | awelux_ | I always soft reset to circumvent the vendor customizations |
| 20:02.56 | awelux_ | So I didn't notice |
| 20:03.19 | awelux_ | But we noticed that having flight mode on/off makes a difference |
| 20:03.35 | awelux_ | ift doesn't boot with flight mode _on_ |
| 20:04.13 | awelux_ | which is strange because you would thing in flight mode the phone would do less. |
| 20:04.38 | rob_w | if i soft reset and dont enter the pin , it mostly works .. if i soft reset and enter the pin on ce it mostly doesnt work |
| 20:05.16 | awelux_ | We din't find a difference depending on pin. |
| 20:05.16 | rob_w | crap .. my gpe-dm is not starting .. |
| 20:05.36 | awelux_ | if you boot without sim card it shouldn't work. |
| 20:06.22 | rob_w | never did that |
| 21:32.12 | *** join/#htc-blueangel AlphaPi[jr] (i=aaa@host226-180.pool870.interbusiness.it) |
| 21:48.51 | rob_w | something still wakes my blueangel up from suspend .. |
| 22:32.18 | *** join/#htc-blueangel lamikr (n=chatzill@aragorn.kortex.jyu.fi) |