| 01:25.31 | *** join/#openezx [g2] (n=g2@nslu2-linux/g2) |
| 03:17.36 | *** join/#openezx leoncamel (n=leoncame@219.238.107.107) |
| 04:03.11 | *** join/#openezx proberts (n=proberts@67-20-86-248.vnnyca.adelphia.net) |
| 05:16.26 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 06:54.18 | *** join/#openezx LoneStar (n=xinzhen@211.144.200.92) |
| 07:58.36 | Hopscotch | good morning |
| 08:42.08 | robtaylor | 'morn |
| 10:11.17 | *** join/#openezx alphaone (n=alphaone@w0722.wlan.rz.tu-bs.de) |
| 10:11.50 | stefan_schmidt | hi |
| 10:12.01 | alphaone | good morning |
| 10:14.25 | *** join/#openezx florian_kc (n=fuchs@dns.kernelconcepts.de) |
| 10:14.39 | stefan_schmidt | hi florian_kc |
| 10:15.26 | florian_kc | good morning |
| 10:37.53 | *** join/#openezx ade|desk (n=ADAVEY@194.200.143.249) |
| 10:45.52 | stefan_schmidt | hi ade|desk |
| 10:46.19 | ade|desk | hi stefan_schmidt |
| 10:46.40 | ade|desk | good news about the weekend hacking ... great stuff! |
| 10:47.26 | stefan_schmidt | yes, but it could be better. :) |
| 10:47.36 | Hopscotch | hihi |
| 10:47.56 | Hopscotch | seems like we're getting a channel for 23c3 - i had several phonecalls with the "landesmedienanstalt" now... |
| 10:48.50 | stefan_schmidt | yeah |
| 10:48.51 | alphaone | cool :-) |
| 10:49.10 | Hopscotch | hehe, yeah - i guess so |
| 10:50.26 | Hopscotch | last year day and night were just inverted |
| 10:52.39 | stefan_schmidt | Hopscotch: heh |
| 10:52.55 | stefan_schmidt | hi mickeyl |
| 10:52.58 | mickeyl | morning guyds |
| 10:53.00 | mickeyl | guys, even |
| 10:53.36 | mickeyl | my wife won't let me attend anything that's later than 19:00 or else she won't come with me |
| 10:53.56 | alphaone | lol |
| 10:53.58 | Hopscotch | yeah, but nobody will be awake before 2pm |
| 10:54.04 | stefan_schmidt | mickeyl: That's the reason why my GF is at home. ;) |
| 10:54.16 | Hopscotch | mine as well |
| 10:54.42 | Hopscotch | but last year she inverted as well and was online in irc all the time... |
| 10:55.03 | Hopscotch | when i'm at home she wants to go to bed at 11pm - understand the girls... |
| 10:58.17 | stefan_schmidt | heh |
| 10:58.37 | stefan_schmidt | my GF don't even know what IRC is. :) |
| 10:58.57 | florian_kc | :-) |
| 10:59.04 | Hopscotch | hehe, i found mine on irc... |
| 10:59.29 | stefan_schmidt | wow, find GF at irc...interesting. :) |
| 10:59.34 | Hopscotch | well, i knew her before - but she used to be on irc all the time |
| 10:59.35 | florian_kc | Hopscotch: but you know her irl? ;) |
| 10:59.48 | Hopscotch | :P |
| 11:00.33 | Hopscotch | and she knows my real name and not only my nick |
| 11:00.49 | florian_kc | heh |
| 11:00.56 | ade|desk | it can't be done i tell you ! |
| 11:00.58 | stefan_schmidt | heh |
| 11:01.14 | Hopscotch | ade|desk: what can't be done? |
| 11:01.30 | ade|desk | my wife can't tell its me without a desk infront of me ;) |
| 11:01.38 | Hopscotch | rofl |
| 11:01.46 | ade|desk | real life ... it can't be done |
| 11:02.10 | Hopscotch | anything you can't ping isn't real |
| 11:02.43 | ade|desk | bridging not working ? |
| 11:02.59 | stefan_schmidt | it's getting geeky here. :) |
| 11:03.47 | Hopscotch | ade|desk: you can't ping your bridge? |
| 11:04.37 | ade|desk | if you can't ping real life ojects its cos your bridging isn't setup correctly |
| 11:04.58 | ade|desk | </geekness> |
| 12:20.25 | *** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1) |
| 12:31.20 | *** join/#openezx [g2] (n=g2@nslu2-linux/g2) |
| 13:10.54 | stefan_schmidt | re |
| 13:11.33 | florian_kc | stefan_schmidt: wb |
| 13:11.57 | stefan_schmidt | Some http://wiki.openezx.org/E680_Board updates |
| 13:31.46 | *** join/#openezx jonwil (n=jonwil@60-240-169-200.tpgi.com.au) |
| 13:31.47 | jonwil | hi |
| 13:32.00 | florian_kc | hi jonwil |
| 13:32.27 | jonwil | its good news that we found out the chips inside the EZX E378i phone :) |
| 13:32.31 | jonwil | all except one that is |
| 13:57.16 | stefan_schmidt | hi jonwil |
| 13:57.35 | stefan_schmidt | jonwil: We found out the one for _E680_ :) |
| 13:57.49 | jonwil | yeah, E680 |
| 13:57.52 | jonwil | my mistake |
| 13:57.55 | stefan_schmidt | I guess the last one could be flash for BP. |
| 13:58.00 | jonwil | could be |
| 13:58.14 | jonwil | although BP may have flash on board |
| 13:58.17 | stefan_schmidt | I added it with a questionmark |
| 14:00.06 | jonwil | if you look at the second board picture, next to IC 10 is a connector on the edge of the PCB, anyone know what that is? |
| 14:00.08 | jonwil | camera? |
| 14:00.37 | jonwil | and whats the other thing above it |
| 14:00.49 | jonwil | the silver one at the top looks like the USB port to me |
| 14:01.57 | jonwil | also, between IC 8 and the USB connector I see 3 smaller ICs, I dont know what they are but if we can, we should document them too |
| 14:03.03 | stefan_schmidt | jonwil: headphone jacket |
| 14:03.15 | stefan_schmidt | The other one is the mini-usb |
| 14:03.48 | jonwil | <looks at high res photo> |
| 14:04.07 | stefan_schmidt | The thing above headphone is a screen lock switch |
| 14:05.11 | jonwil | I can see 3 things on the high res photo that look like they have enough pins to be significant |
| 14:05.22 | stefan_schmidt | No idea about the small ICs. |
| 14:05.26 | jonwil | ok |
| 14:05.29 | jonwil | but there are 3 of them |
| 14:05.33 | jonwil | you see the 3 I mean? |
| 14:05.54 | jonwil | do you happen to know where the speaker and microphone connect to the PCB? |
| 14:06.16 | stefan_schmidt | It's my phone and my pictures. I should know. :) |
| 14:06.23 | jonwil | ok, where do they connect? |
| 14:07.20 | stefan_schmidt | Socket next to the screen lock switch is speacker connection. |
| 14:07.39 | jonwil | so thats the thing with 2 pins? |
| 14:07.49 | stefan_schmidt | microphone is bundled with the keypad. Black connector on the lower right edge. |
| 14:08.03 | stefan_schmidt | yes, the one with the 2 pins |
| 14:08.06 | jonwil | ok |
| 14:08.18 | jonwil | What graphics hardware does the E680 have? |
| 14:08.33 | stefan_schmidt | On the other side there are additional sockets for the "3D" speaker on the back side. |
| 14:08.59 | jonwil | btw, one of the 3 small ICs is most probobly an audio amplifier for the speaker |
| 14:09.00 | stefan_schmidt | Just an LCD interface from bulverde (PXA). |
| 14:09.01 | jonwil | or so |
| 14:09.13 | stefan_schmidt | It the white connector next to AP |
| 14:09.49 | stefan_schmidt | I can see if I can read the chip numbers once a have my phone back. |
| 14:10.01 | stefan_schmidt | (Friend of mine is working with it atm) |
| 14:10.12 | jonwil | ok |
| 14:11.12 | jonwil | actually, I can only see 2 ICs there, the first one is directly below the thing with a big 8 moulded into it |
| 14:11.23 | jonwil | between it and the speaker connector |
| 14:11.33 | jonwil | the second one is further in and slightly up from it |
| 14:11.47 | jonwil | below that is a black square |
| 14:11.51 | jonwil | I dont know if its an IC or not |
| 14:12.10 | jonwil | in any case we seem to have located most of the important chips |
| 14:13.25 | jonwil | if I had the right tools (and could be sure of not damaging the phone), I would take the cover off my L6 and write down chip numbers from that to compare with what an E680i has |
| 14:14.55 | jonwil | but I dont want to risk damaging the phone |
| 14:15.29 | jonwil | plus I dont have anything that even looks like it would fit those screws :) |
| 14:17.19 | stefan_schmidt | jonwil: I guees the srews are T6 (Torx) |
| 14:20.35 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 14:29.16 | jonwil | is that yamaha chip a MIDI FM synthisis chip? |
| 14:31.23 | stefan_schmidt | hmm, I don't know. |
| 14:31.39 | jonwil | Well its obviously MIDI since there is a hwmidi.o file that presumably drives it |
| 14:31.52 | jonwil | I would guess given that its only a mobile phone, its probobly FM synthisis |
| 14:32.01 | stefan_schmidt | It's a good guess, but I'm not really frim with the audio system on this phones. Looks really strange sometimes. |
| 14:32.31 | stefan_schmidt | Anyway, I have to do some work now. cu later. |
| 14:33.25 | jonwil | c uhttp://gsm-magic.com/ArchitRaj/Schematics/Motorola/A780_schematics.rar |
| 14:33.28 | jonwil | opps |
| 14:33.29 | jonwil | ignore that |
| 14:33.33 | jonwil | my mistake |
| 14:45.39 | jonwil | I was googling for schematics incase they could tell us more about the EZX phone hardware but the ones I found have "confidential" on them so I decide it best not to refer to them |
| 14:45.57 | jonwil | for obvious reasons |
| 15:01.46 | stefan_schmidt | Indeed. Pollution with illegal documents or code is bad. |
| 15:02.03 | stefan_schmidt | We will get our goals, but it will take longer the clean way. |
| 15:16.14 | jonwil | I just did more searches for 1020W0Y on various places and found this possibly relavent result |
| 15:16.16 | jonwil | http://www.datasheets.org.uk/pdf/datasheets/2670916.shtml |
| 15:16.24 | jonwil | could that be what chip 11 is? |
| 15:17.14 | jonwil | also, I think at least one of our smallish chips may well be some kind of audio amplifier or something |
| 15:17.47 | jonwil | I am not an electrical engineer but I know enough about audio to know that some kind of amplification is needed for a speaker that small |
| 15:18.41 | jonwil | once we see any numbers written on the small chips maybe we will know for sure what they are |
| 15:20.01 | jonwil | in any case knowing what the chips are doesnt help us all that much |
| 15:21.07 | jonwil | the only thing I can think of that is undocumented and not motorola custom is the Yamaha MIDI chip |
| 15:21.15 | jonwil | and knowing how that works doesnt help us all that much |
| 15:21.36 | jonwil | everything else is either moto specific, connected to the BP or both |
| 15:21.39 | *** join/#openezx lkcl_work (i=cgiircus@gateway/web/cgi-irc/chriscole.info/x-b4b0ff11b764e135) |
| 15:22.52 | lkcl_work | http://www.handhelds.org/hypermail/kernel-discuss/current/2218.html - phone audio routing infrastructure plan |
| 15:25.24 | robtaylor | lkcl_work: hmm, should all teh routing logic really be in-kernel? |
| 15:26.24 | vmaster | stefan_schmidt: the JTAGkey is on its way |
| 15:27.28 | *** join/#openezx letoto (n=paul@tla.xelerance.com) |
| 15:27.36 | stefan_schmidt | vmaster: Thanks a lot. Once I have it, I let you know. |
| 15:27.48 | robtaylor | lkcl_work: i guess the hardware usually has some restrictions on what outputs can be used simultaneously? |
| 15:27.55 | jonwil | cya |
| 15:27.55 | robtaylor | laters stefan_schmidt |
| 15:28.19 | letoto | any news on the 30s hacking weekend? |
| 15:38.20 | letoto | cool. I just got offered access to a linux phone for $200k plus an NDA |
| 15:38.39 | letoto | you friend, you special deal! |
| 15:46.29 | jonwil | I really think there is a market out there for a phone with an open architecture. |
| 15:46.59 | jonwil | Either a fully open architecture or (more likely given FCC regs and stuff) a closed part that handles radio communications and a totally open part for the phone itself |
| 15:47.01 | jonwil | like EZX |
| 15:47.07 | alphaone | jonwil: Very likely that that is the chip. It's definitely an intel |
| 15:47.08 | jonwil | only with the AP side documented and 100% open |
| 15:47.56 | jonwil | the size listed in the data sheet matched up with the size you get if you dump the flash from various P2K phones |
| 15:48.22 | jonwil | which makes sense since the EZX BP is a modified P2K phone really |
| 15:48.29 | alphaone | jonwil: Yeah that are going to be great times :-) |
| 15:49.07 | jonwil | We dont need (or want) to touch BP side other than by talking to it via PCAP, mux_cli etc |
| 15:49.38 | alphaone | jonwil: When I get home I'll see if I can find any numbers at all on that chip (I'm the one who has stefan_schmidt 's phone ATM) |
| 15:49.50 | jonwil | ok |
| 15:51.26 | jonwil | see if you can read the numbers on some of the smaller chips next to chips labeled 7 and 8 on board diagram |
| 15:51.39 | jonwil | I can see what looks like outline of RF shield I think |
| 15:52.03 | jonwil | with chips 7 and 8 (MIDI chip and FM radio) and 2 other small ones |
| 15:52.14 | jonwil | I think those small ones may be audio amplifiers |
| 15:52.16 | jonwil | or one of them may be |
| 15:52.25 | alphaone | Okay, I'll try |
| 15:52.27 | jonwil | if either of them have numbers on them, that would be usefull to have :) |
| 15:52.29 | jonwil | thanks :) |
| 15:52.53 | alphaone | np :-) |
| 15:53.15 | jonwil | probobly the biggest problem with making an "open" mobile phone, even one with a "blackbox air interface" (i.e. like BP on openezx) is the phone carriers |
| 15:53.15 | alphaone | Yeah, we soldered off the RF shielding to get to the chips |
| 15:54.24 | jonwil | not only would carriers never take a phone which is open enough for people to run their own stuff on (why do you think MS added locks to windows for smartphones to prevent running apps not approved by the carrier...) |
| 15:54.56 | alphaone | Well but they seem to be doing it with openmoko |
| 15:58.40 | robtaylor | jonwil: depends if they're american or not ;) |
| 15:58.57 | robtaylor | jonwil: the main concern tehy have is not allowing people to poison their network |
| 15:59.13 | jonwil | well thats why you build an open phone with a dual processor |
| 15:59.16 | jonwil | like with EZX |
| 15:59.32 | jonwil | so that you have a closed source system that you pass commands to |
| 15:59.36 | jonwil | such as "dial this number" |
| 15:59.48 | jonwil | or "answer the phone" |
| 16:00.02 | jonwil | and receieve envents like "incoming call" or "other end hang up" |
| 16:00.18 | jonwil | have as little logic as possible on this "bp" side |
| 16:00.36 | jonwil | but have enough so that you cant do anything that would cause network issues |
| 16:03.01 | alphaone | :-) |
| 16:03.42 | robtaylor | jonwil: indeed |
| 16:03.54 | jonwil | Based on what I have seen, what carriers want is the abillity to lock handsets to their network |
| 16:04.14 | jonwil | otherwise subsidised phones like prepaid (like my L6) wouldnt be available |
| 16:04.36 | robtaylor | jonwil: hmm, thats something thats changing afict |
| 16:05.02 | robtaylor | some uk providers actually use that the phones are unlocked as a sales point |
| 16:05.14 | jonwil | basicly, there are 3 kinds of phones |
| 16:05.22 | jonwil | phones sold outright (generally at "full price") |
| 16:05.33 | jonwil | these dont need to be locked as the phone co makes proffit on them |
| 16:05.39 | jonwil | phones sold with a contract |
| 16:06.09 | robtaylor | yeh, i'm talking about contract phones |
| 16:06.48 | robtaylor | does anyone care about locking on pay-as-you go? |
| 16:07.19 | jonwil | phone carriers that lock their phones sold with a contract are stupid |
| 16:07.34 | jonwil | if you break the contract, you have to pay an exit fee anyway |
| 16:07.44 | jonwil | my L6 is network locked |
| 16:07.47 | alphaone | robtaylor: Would be so you are bound to that company. |
| 16:08.06 | alphaone | Like a contract without calling it one |
| 16:08.11 | jonwil | but considering its subisised and on prepaid, I can understand |
| 16:09.09 | jonwil | the other thing some providers (especially in US) care about is stopping people from being able to access content (ringtones, wallpapers, games etc) not authorized by the carrier |
| 16:09.34 | jonwil | the worst carriers even make you pay to take pictures with the camera phone and move them to your PC |
| 16:10.11 | robtaylor | jonwil: yeah, i'd hope market forces stop that in the future ;) |
| 16:10.37 | robtaylor | i dont think we're gonna be looking at linux phones at the pay-as-you go level for a good while yet |
| 16:11.13 | robtaylor | if/when they do get there, i expect that providers will be using trusted boot to make sure noone modifies their system |
| 16:11.22 | robtaylor | and some kinda GPL fight will ensue ;) |
| 16:11.36 | jonwil | no, the answer is to put any network locking on the closed part |
| 16:11.45 | jonwil | where the GSM and GPRS and bluetooth code lives |
| 16:11.53 | jonwil | and where the SIM read code lives |
| 16:12.19 | robtaylor | yeah, i guess so |
| 16:12.28 | alphaone | Which is also the case in the Thinkpad models with UMTS |
| 16:14.04 | jonwil | if I was buying a laptop that was locked to a carrier, I would want that carrier to give me something resembling a discount on it... |
| 16:14.19 | alphaone | A friend of mine bought one and the UMTS card had the carrier locked to vodafone - fortunately he got the code to unlock it after making a few phone calls. |
| 16:14.54 | alphaone | jonwil: Well I think IBM and vodafone had some kind of contract |
| 16:15.05 | jonwil | so what, is this just a laptop with a UMTS PCMCIA card in it? |
| 16:15.18 | jonwil | or is this actual UMTS hardware built inside the laptop somewhere |
| 16:15.22 | alphaone | It's connected via USB |
| 16:15.33 | jonwil | ok |
| 16:16.03 | alphaone | But internal |
| 16:16.58 | jonwil | does anyone here have an EZX phone with software that works such that they can send AT commands to the BP side of the phone? |
| 16:17.04 | jonwil | Or do we not know how to do that yet? |
| 16:18.46 | jonwil | I think it would be usefull to do an AT+CLAC on the BP side of the phone then use google to find out what all the commands do |
| 16:18.47 | alphaone | Should be possible with the original kernel and usbnet |
| 16:19.05 | jonwil | so we can figure out what commands we can send to BP |
| 16:24.32 | alphaone | But there is to be more communication between AP and BP than AT commands |
| 16:24.57 | jonwil | as far as I know, communication happens through AT commands via mux_cli |
| 16:24.59 | stefan_schmidt | re |
| 16:25.07 | jonwil | and also through direct connections such as power |
| 16:25.09 | jonwil | and audio |
| 16:25.11 | jonwil | and other lines |
| 16:25.18 | jonwil | but AT commands is most usefull right now |
| 16:25.18 | stefan_schmidt | signalling is done via GPIO |
| 16:25.46 | alphaone | hi stefan_schmidt |
| 16:26.22 | letoto | stefan_schmidt: no good news yet? |
| 16:26.30 | letoto | stefan_schmidt: my phone will arrive today or tomorrow :) |
| 16:26.37 | jonwil | what phone you get? |
| 16:26.42 | letoto | a780 |
| 16:26.53 | jonwil | cool |
| 16:27.11 | stefan_schmidt | letoto: Power off problem is still there. |
| 16:27.36 | jonwil | I think that there is probobly either some GPIO line or some AT command that the BP expects |
| 16:27.44 | jonwil | something that says "AP is up and in working order" |
| 16:27.45 | stefan_schmidt | letoto: http://lists.openezx.org/pipermail/openezx-devel/2006-December/000871.html |
| 16:28.26 | jonwil | whats most wierd is that rebuilding the motorola provided 2.4.x kernel sources and flashing them onto the phone seems not to work |
| 16:28.40 | alphaone | The problem is that even the 2.4 kernel from Motorola has the power off problem |
| 16:28.54 | alphaone | exactly |
| 16:29.51 | alphaone | I meant a self compiled 2.4 kernel from the Motorola sources, not the binary kernel from Motorola of course |
| 16:30.37 | letoto | stefan_schmidt: so still unknown? |
| 16:31.02 | jonwil | Does that mean that mototrola didnt provide everything required? |
| 16:31.14 | jonwil | or that we are doing something wrong in building the 2.4 kernel |
| 16:31.15 | alphaone | jonwil: Thats the question |
| 16:31.15 | jonwil | or what? |
| 16:31.20 | letoto | depends |
| 16:31.34 | letoto | if it is the kernel doing this yes. if it is a non-gpl application on the OS, no |
| 16:31.38 | stefan_schmidt | We can not be sure yet. :( |
| 16:31.51 | letoto | I've been talking to RMS to explain to him this situation. and hopefully he can fix it in the gplv3 |
| 16:32.02 | letoto | but it took like 4 emails to him to understand the issue |
| 16:32.10 | alphaone | heh |
| 16:32.21 | letoto | stefan_schmidt: the gpio signal might be a checksum |
| 16:32.29 | letoto | we need to see what it sends |
| 16:32.37 | alphaone | There will always be loopholes. We have to get the companies to want to make the devices open |
| 16:32.46 | robtaylor | alphaone: *nod* |
| 16:33.01 | letoto | alphaone: want == money == not gettin gsued |
| 16:33.03 | letoto | unfortunately |
| 16:33.10 | letoto | let's hope a few good companies make it and cause the rest to open up |
| 16:33.22 | robtaylor | letoto: no, better that they appreciate the gains |
| 16:33.33 | robtaylor | but tbh, they're starting to get there |
| 16:33.39 | stefan_schmidt | letoto: They have to learn that they can make money with oss, too. |
| 16:33.54 | robtaylor | the guys at the top are starting to get it a bit.. not a whole lot of them, but some |
| 16:34.13 | alphaone | Which is a very good sign |
| 16:34.15 | stefan_schmidt | letoto: I don't think it is a checksum just a alive signal. |
| 16:35.11 | jonwil | if I was going to replace my L6, I would probobly get a L7, a KRZR K1 or an E2 ROKR |
| 16:35.11 | letoto | stefan_schmidt: think about it. if the GPL kernel sends a life signal that includes a checksum, the non-GPL kernel can cut power. and no GPL violation occurs |
| 16:35.23 | jonwil | wait, no, not KRZR K1, RIZR Z3 |
| 16:35.28 | letoto | but if the GPL kernel sends a secret signal, and the sources dont include that, then it is a gpl violation |
| 16:35.50 | jonwil | I dont want a flip phone |
| 16:36.13 | jonwil | or anything that doesnt have a normal dialpad and buttons |
| 16:36.19 | alphaone | The GPIO communication is done in kernel and so they would have to open that part up. |
| 16:36.26 | jonwil | not that I want to replace my L6 anytime soon :) |
| 16:37.16 | stefan_schmidt | letoto: The key is to get the 2.4 fully running, or proof that they haven't offer all code. It's difficult. |
| 16:37.55 | jonwil | maybe we could try compiling with the GCC published by motorola |
| 16:38.08 | letoto | stefan_schmidt: i know. and I want it working dec 28 :) |
| 16:38.22 | letoto | guess I can get away with running the motorola kernel |
| 16:38.23 | stefan_schmidt | letoto: heh |
| 16:38.39 | alphaone | We could try to get the original kernel to panic somewhere before the userland starts and if the phone stays alive we know it's something in the kernel |
| 16:39.05 | stefan_schmidt | alphaone: Go ahead, use my phone. ;) |
| 16:39.06 | jonwil | another idea is to back up the userland and then remove it |
| 16:39.16 | jonwil | i.e. we backup all libraries |
| 16:39.23 | jonwil | and then remove anything that isnt open source from the phone |
| 16:39.26 | jonwil | and somehow let it boot |
| 16:39.31 | jonwil | and see if it stays alive |
| 16:39.34 | alphaone | stefan_schmidt: I need to flash the orginal 2.4 kernel on it for that |
| 16:39.42 | letoto | init=/no/such/file ? |
| 16:39.49 | letoto | can we specify stuff to the bootloader? |
| 16:40.08 | stefan_schmidt | alphaone: right, I should have explained this to you before leaving. |
| 16:40.12 | alphaone | letoto: It's not that easy |
| 16:40.36 | stefan_schmidt | letoto: We can't give options to kernel in bootloader |
| 16:40.52 | alphaone | letoto: We could gunzip the kernel and change the commandline in the binary - then gzip it |
| 16:41.13 | jonwil | hmmm, that would also tell us if the phone does a checksum on the kernel |
| 16:41.32 | stefan_schmidt | alphaone: I hope to get the JTAG from vmaster at least on wednesday. Perhaps we can hvae some test with it this day. |
| 16:41.34 | alphaone | stefan_schmidt: Can you tarball the whole app so I can use it? |
| 16:41.50 | letoto | stefan_schmidt: do you have the kernel image of motorola? |
| 16:42.00 | letoto | and can you run their kernel? |
| 16:42.07 | stefan_schmidt | alphaone: Will do. Expect an upload in the evening. |
| 16:42.08 | letoto | as in one you had write access to? |
| 16:42.29 | stefan_schmidt | letoto: I have a image of /dev/mtd4, the part of the kernel. So yes. |
| 16:42.36 | letoto | you can flip the bit of the booting device with rdev or something |
| 16:42.46 | letoto | so boot from a bogus device |
| 16:42.47 | vmaster | stefan_schmidt: i didn't make it to the post office before 16:00, so it might be wednesday or thursday |
| 16:43.24 | jonwil | does anyone know what hatcreek.o is for? |
| 16:43.30 | stefan_schmidt | vmaster: OK, I rely on the Post. :) |
| 16:43.42 | stefan_schmidt | jonwil: Another flash file system. |
| 16:43.47 | jonwil | ok |
| 16:43.57 | stefan_schmidt | jonwil: It seems we don't have enough these days. ;) |
| 16:43.58 | jonwil | so hatcreek and TFFS are just filesystem modules |
| 16:43.59 | letoto | stefan_schmidt: if you mail me the kernel, I'll see about changing the root device on it |
| 16:44.47 | alphaone | stefan_schmidt: Okay, thanks |
| 16:45.25 | stefan_schmidt | letoto: Thanks for the offer. We should be able to do this on our own. Just strip the offset from zImage, gunzip, change, zip and add offset. |
| 16:46.07 | stefan_schmidt | or using hexedit. We just not tried it yet. |
| 16:46.07 | letoto | stefan_schmidt: yeah. then go for it. then you know if it is kernel or userland for sure |
| 16:46.12 | letoto | is the gpio a kernel module? |
| 16:46.15 | letoto | or inline? |
| 16:46.19 | stefan_schmidt | inline |
| 16:47.21 | jonwil | if you do whatever it is you asre going to do with this kernel, will mux_cli still be loaded? |
| 16:47.34 | jonwil | if not, wont that mean that things will fail because the AP doesnt talk to the BP? |
| 16:48.04 | stefan_schmidt | jonwil: That is another possibility. We will also check this. |
| 16:48.17 | letoto | well |
| 16:48.29 | jonwil | one idea that would be usefull would be to try to rebuild mux_cli but with some kind of log to log all communications |
| 16:48.31 | letoto | you can always change the rootfs and just replace /sbin/init with a custom script |
| 16:48.31 | stefan_schmidt | letoto: alphaone has my devel phone atm. We will check on wednesday. |
| 16:48.39 | letoto | that loads the mux_cli and just sleeps |
| 16:50.00 | stefan_schmidt | letoto: Changing the 2.4 userland is not easy. And the 2.6 mux_cli is the code we like to test in long-term |
| 16:50.19 | stefan_schmidt | letoto: But I get your point. |
| 16:52.08 | letoto | stefan_schmidt: changing the 2.4 userland isnt easy? |
| 16:52.14 | robtaylor | stefan_schmidt: whats the status on mux_cli? |
| 16:52.16 | letoto | hmm, i was counting on adding some stuff to it :P |
| 16:55.56 | alphaone | Okay, I'm off. cu |
| 16:58.35 | stefan_schmidt | robtaylor: It _should_ work if the phone would not power off. We are working on testing this. |
| 16:58.35 | stefan_schmidt | letoto: You can add some custom programs on it. But most areas are looked read only. |
| 16:59.18 | stefan_schmidt | cu |
| 17:18.41 | letoto | stefan_schmidt: locked readonly? |
| 17:18.42 | *** join/#openezx kenws (n=ken@dslb-088-064-150-118.pools.arcor-ip.net) |
| 17:29.57 | *** join/#openezx alphaone (n=alphaone@c120.apm.etc.tu-bs.de) |
| 17:30.12 | alphaone | re |
| 18:17.30 | *** join/#openezx vmaster_ (i=vmaster@p549B68C5.dip.t-dialin.net) |
| 18:23.40 | *** join/#openezx LoneStar (n=xinzhen@218.81.34.26) |
| 18:40.59 | *** join/#openezx TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 18:59.18 | *** join/#openezx blkhawk (i=blkhawk@pD9E04C17.dip.t-dialin.net) |
| 19:27.45 | *** join/#openezx jim_rayner (n=jim_rayn@beyondvoice.plus.com) |
| 21:02.27 | *** join/#openezx |dion| (n=dionua@xen.inhex.net) |
| 21:45.38 | *** join/#openezx florian__ (n=fuchs@87.193.63.169) |
| 22:28.22 | florian | re |
| 22:29.00 | stefan_schmidt | wb florian |
| 23:15.10 | letoto | sniff. |
| 23:15.16 | letoto | my motorola did not arrive today |
| 23:27.40 | mickeyl | suckers! |
| 23:27.43 | mickeyl | who's holding it back? |
| 23:28.09 | stefan_schmidt | heh |
| 23:28.18 | stefan_schmidt | anyone here know arm assembler? |
| 23:29.30 | stefan_schmidt | This way we could load it via boot_usb and just restart into passthrough. No need for booting kernel, log in via ssh and execute devmem2 |
| 23:37.09 | mickeyl | that'd rock |
| 23:37.22 | mickeyl | might ask RP on #oe if it's a small question |
| 23:37.26 | mickeyl | he's _very_ busy :) |
| 23:38.40 | stefan_schmidt | mickeyl: I think I will read some more about it first. I have never done assembler before. |
| 23:38.53 | mickeyl | wow. then it'll be fun :)) |
| 23:39.00 | mickeyl | imagine C being a high level language ;) |
| 23:39.11 | stefan_schmidt | But I only want: write 0x12345678 to address 0xA0000000 |
| 23:39.18 | stefan_schmidt | heh |
| 23:39.20 | mickeyl | well, that should be fair |
| 23:39.40 | stefan_schmidt | yes, should be only some lines of code. |
| 23:40.01 | stefan_schmidt | But I have to learn the basics first... |
| 23:40.01 | letoto | stefan_schmidt: I could get access to one |
| 23:40.09 | letoto | an arm-assembly writer that is |
| 23:40.23 | letoto | mickeyl: tigerdirect :) |
| 23:40.40 | stefan_schmidt | letoto: Is he online right now? |
| 23:41.03 | letoto | stefan_schmidt: on and off. he is "mcr", on #openswan. collegue of mine |
| 23:41.18 | letoto | but do you know yet what the magic is to write yet? |
| 23:41.44 | stefan_schmidt | yes, it _is_ 0x12345678 |
| 23:42.13 | stefan_schmidt | I already can enter the passthrough mode with this written at 0xA0000000 |
| 23:44.08 | stefan_schmidt | ok, no mcr in #openswan |
| 23:44.54 | stefan_schmidt | mickeyl: I try RP now. |
| 23:46.40 | letoto | *** mcr is n=mcr@desk.marajade.sandelman.ca (Michael Richardson) |
| 23:46.41 | letoto | *** on channels: @#openswan @#openswan-dev |
| 23:48.32 | stefan_schmidt | letoto: Sorry, he is a moderator. I just not looked at the top of the member list. |
| 23:50.22 | stefan_schmidt | letoto: I wait if RP can give me an answer before asking mcr |
| 23:50.39 | letoto | stefan: good idea |
| 23:50.50 | *** join/#openezx nevarrie_ (n=jgrant@CPE-72-128-35-75.kc.res.rr.com) |
| 23:51.03 | letoto | so if we know the magic, why can't we write it? |
| 23:51.43 | letoto | <PROTECTED> |
| 23:51.47 | letoto | that makes te phone stay on? |
| 23:52.05 | stefan_schmidt | I can write it. But need to boot kernel, open ssh connection to phone and use devmem2 to write it. |
| 23:52.28 | stefan_schmidt | letoto: Ah, sorry, my mistkae. No it don't make the phone stay on. |
| 23:52.36 | letoto | so what does it do? |
| 23:53.00 | letoto | do you need to write it once? or every minute? |
| 23:53.09 | stefan_schmidt | letoto: It is a new bootloader mode we found. In this mode you can communicate directly to the BP. |
| 23:53.17 | stefan_schmidt | letoto: once |
| 23:54.00 | letoto | so why assembly? |
| 23:54.06 | letoto | why not a tiny c program? |
| 23:55.20 | stefan_schmidt | I like to load this binary directly from the bootloader. No kernel is running, nothing. |
| 23:55.41 | letoto | and we dont have source of the bootloader? |
| 23:55.47 | stefan_schmidt | I'm not aware that I can write with c directly to the mem address 0xA0000000 |
| 23:55.48 | letoto | where are you hooking in your assembly code ? |
| 23:56.31 | letoto | i am confused though. you can you can do it using devmem2 from a booted linux? |
| 23:56.37 | stefan_schmidt | we have code of the bootloader, but nobody flashed a new one yet. |
| 23:57.04 | stefan_schmidt | OK, let me explain it. |
| 23:57.08 | letoto | please:) |
| 23:59.19 | stefan_schmidt | You need to write this magic in the first word of the AP mem. Starts at 0xA0000000 |
| 23:59.44 | stefan_schmidt | Reboot the phone. Halt down the power button while power one for some seconds. |