| 00:02.20 | *** join/#htc-linux Othello_ (i=Magorium@gateway/tor/x-5f7faecc34dcbc09) |
| 00:15.52 | ginge | anyone titan wanna run a zImage? |
| 00:16.11 | *** join/#htc-linux dcordes_ (n=dcordes@unaffiliated/dcordes) |
| 00:16.42 | ginge | http://www.headfuzz.co.uk/files/android/zImage-titan |
| 00:17.44 | marajin | aaand I'm back |
| 00:17.49 | ginge | mmm nicotine |
| 00:17.54 | ginge | brb |
| 00:19.55 | marajin | heh :) |
| 00:21.43 | wasup | im thinking about getting some nicotine right now too |
| 00:23.31 | marajin | you're only 19, you don't need to be slowly killing yourself with tobacco yet. |
| 00:23.39 | marajin | Alcohol, now that's a different thing ;) |
| 00:23.55 | ginge | okay, do we have a titan victim yet? |
| 00:24.06 | wasup | im all yours |
| 00:24.16 | wasup | i hate alcohol |
| 00:24.28 | wasup | now menthol cigarettes....those hit the spot |
| 00:24.30 | ginge | grab that file above and boot angstrom |
| 00:24.38 | ginge | menthol?? maybe herbal ;) |
| 00:25.31 | ginge | although I can't really say anything about tase. I just put vanilla iceream in my tea because I ran out of milk. clearly have none |
| 00:25.34 | marajin | methols are pointless unless you're hopped up :P |
| 00:25.38 | marajin | *menthols |
| 00:26.45 | marajin | I kinda wish I had some djarums at the moment |
| 00:27.32 | wasup | maybe i am |
| 00:27.35 | wasup | hopped up |
| 00:28.05 | marajin | Well each to his own, I stick to the legal (caffeine, nicotine, alcohol) |
| 00:28.34 | wasup | Im just kidding |
| 00:28.36 | wasup | im at work |
| 00:28.42 | marajin | Hmm my friend promised me some ziganovs, I oughta chase him up on that too |
| 00:30.04 | wasup | k ginge...im booting it |
| 00:30.06 | marajin | wasup: Where'd you say you were from? |
| 00:30.42 | wasup | Yuma, AZ... a little pathetic crap town in the us |
| 00:30.51 | wasup | well i live here |
| 00:30.54 | wasup | im from atlanta |
| 00:30.59 | marajin | Ah so you can't get proper cigars or stuff like ziganovs |
| 00:31.32 | wasup | nope |
| 00:31.51 | wasup | but the shift key makes my camera flash two times |
| 00:31.54 | wasup | :D |
| 00:32.22 | marajin | I feel for ya man ;) You're missing out on a lot |
| 00:32.31 | wasup | as for...q i get >mmki (with a camera flash) |
| 00:33.02 | wasup | w = bhju (camera flash) |
| 00:33.25 | wasup | e = vgytr(camera flash) |
| 00:33.41 | wasup | r= cfde(camera flash) |
| 00:34.16 | *** join/#htc-linux surgex (i=surge@pool-71-186-163-107.bflony.fios.verizon.net) |
| 00:34.23 | *** join/#htc-linux TeringTu1y (n=maarten@195-241-125-243.ip.telfort.nl) |
| 00:34.26 | ginge | so one of those gpios is for camera :) |
| 00:35.00 | ginge | damn where are they |
| 00:35.12 | wasup | enter= xzsw(camera flash) |
| 00:35.37 | wasup | hmm... cant you just look at what is the same in all of those and take it out? |
| 00:35.40 | marajin | how'd you get so many keys to flash the camera? |
| 00:35.58 | wasup | cause all the keys (working) do the camera flash |
| 00:36.15 | marajin | I meant the question to ginge :P |
| 00:36.34 | wasup | no i know... i meant that to ginge too |
| 00:36.36 | wasup | hahaha |
| 00:36.38 | ginge | doesnt work like that |
| 00:37.00 | marajin | I imagine most of the keys may be doing something you just dunno _what_ :p |
| 00:37.07 | ginge | marajin: looks like one of te gpios I am pulling high on the scan is for the camera |
| 00:37.16 | ginge | some are right though |
| 00:37.30 | wasup | but actually if i press the keys alot (any of them) even the ones that dont work... give me a flash |
| 00:37.33 | wasup | on the camera |
| 00:37.44 | marajin | O.o |
| 00:37.51 | marajin | that sounds like it's going to be interesting |
| 00:38.17 | marajin | Giiiinge, are you trying to use the titan as a secret morse encoder? :p |
| 00:38.33 | wasup | hahaha |
| 00:38.59 | wasup | the keys that do work though give me 2 fast flashed |
| 00:39.04 | wasup | fhlased = flashes |
| 00:39.20 | wasup | fhlased = flashed = flashes |
| 00:39.22 | wasup | dang |
| 00:39.42 | wasup | the other ones only give me a single flash |
| 00:39.53 | marajin | ..wow, kreteks are banned in parts of the US. you live in a very silly country you know that? |
| 00:40.11 | wasup | what are kreteks? |
| 00:40.19 | marajin | clove cigarettes |
| 00:40.25 | wasup | hey im half brazilian... so i have dual citizenship |
| 00:40.41 | wasup | i just have to find out what i want to be by the time i am 21 |
| 00:40.50 | wasup | brazilian or american |
| 00:41.00 | marajin | heh |
| 00:41.10 | marajin | neither? :P |
| 00:41.24 | wasup | that wouldn work |
| 00:41.30 | wasup | hmmm... maybe though. |
| 00:41.39 | marajin | In theory it could but you'd have to get citizenship in a 3rd country |
| 00:41.48 | wasup | europe? |
| 00:41.51 | wasup | ;) |
| 00:42.01 | marajin | Europe is a /continent/ man |
| 00:42.07 | wasup | i know |
| 00:42.23 | wasup | but still you guys cant travel from country to country? |
| 00:42.29 | marajin | though you can technically be a european citizen you have to be a citizen of a country inside europe first |
| 00:42.31 | wasup | i thought they were gonna pass a law like that |
| 00:42.37 | marajin | yeah |
| 00:42.49 | wasup | european union or something |
| 00:42.50 | marajin | I can go most anywhere in europe without passport I think |
| 00:42.58 | wasup | well thats what i meant |
| 00:43.04 | wasup | saved himself |
| 00:43.13 | wasup | ;) |
| 00:43.26 | marajin | Not sure if I can LIVE in another european country without applying for residency/citizenship there tho' |
| 00:43.37 | wasup | how do you direct a question to someone without private message? |
| 00:43.43 | wasup | oh |
| 00:43.49 | wasup | i should look into that |
| 00:43.52 | marajin | er, their name followed by a colon |
| 00:43.55 | marajin | like this.. |
| 00:43.58 | marajin | wasup: Boo? |
| 00:44.05 | wasup | oh... |
| 00:44.21 | wasup | ginge: so any ideas about the keyboard? |
| 00:44.36 | wasup | did it work? |
| 00:44.41 | wasup | jk |
| 00:45.10 | marajin | Should have done but ginge'll be buried in what he's doing |
| 00:45.20 | wasup | oh |
| 00:45.23 | marajin | but it'll highlight the line so he sees it when he gets back |
| 00:45.28 | wasup | oh cool |
| 00:45.36 | wasup | wasup: this is cool |
| 00:45.39 | marajin | nods. |
| 00:45.43 | marajin | hah :P |
| 00:45.47 | marajin | talking to yourself eh? |
| 00:45.51 | wasup | it didnt highlight on mine |
| 00:45.52 | marajin | That's fairly novel even for IRC |
| 00:46.08 | marajin | Yeah but you're using some dodgy java client aren't you? |
| 00:46.22 | wasup | i was before.... now i use colloguy on my mac |
| 00:46.35 | marajin | ah well, mac application for you |
| 00:46.41 | marajin | Mac applications are _special_ |
| 00:46.42 | wasup | not a mac fan? |
| 00:47.25 | marajin | the iPod classic is a decent deal and the mac mini was a cute idea, otherwise I have very little interest in Apple products (overpriced) and I find Mac OSX very taxing to use |
| 00:47.45 | wasup | oh... yeah i have two ipod classics |
| 00:47.52 | marajin | ..why? |
| 00:48.30 | wasup | i used to work for best buy and customer abandoned them... so i took them home and fixed them |
| 00:48.34 | *** join/#htc-linux Kevin2 (n=Kevin2@207-237-52-122.c3-0.avec-ubr12.nyr-avec.ny.cable.rcn.com) |
| 00:48.42 | marajin | anyway, mostly I just don't like OS X, I find it hard to use because it won't behave the way I expect |
| 00:48.47 | wasup | i can fix things... i just cant program very well :) |
| 00:49.20 | wasup | oh... really... It's simple...plus i use the terminal for alot of my wanna be programming |
| 00:49.30 | marajin | Yeah that's the _problem_ |
| 00:49.35 | marajin | it's simple |
| 00:49.40 | marajin | it makes too many assumptions |
| 00:49.45 | wasup | windows gives me issues with ruby and python |
| 00:49.55 | ginge | wasup: no, it failed. |
| 00:49.57 | wasup | true... |
| 00:50.05 | wasup | aw.. what do you need? |
| 00:50.07 | marajin | ah well console work then it's basically just BSD/Unix stuff |
| 00:50.18 | marajin | from the console OS X is fine :P |
| 00:50.25 | ginge | wasup: you realise this will take a while to get right? |
| 00:50.39 | ginge | btw can you download the image again and see what happens this time :) |
| 00:50.42 | wasup | yeah |
| 00:51.03 | marajin | wasup: A few years ago some friends and I went to PC world and locked them out of all their demonstration model macs :P |
| 00:51.51 | ginge | marajin: esc dg on the pc ones :) |
| 00:52.10 | wasup | oh... |
| 00:52.19 | wasup | u have a deep passion for windows? |
| 00:52.26 | marajin | (Changed the user account names and then turned it to a login prompt rather than a click the picture) |
| 00:52.27 | wasup | or do u use linux and stuff? |
| 00:52.33 | ginge | distaste in my case |
| 00:52.35 | marajin | No we boke the windows computers too |
| 00:52.41 | marajin | I use windows and linux both |
| 00:53.00 | wasup | well i can screw up a windows computer with some virii |
| 00:53.13 | wasup | got to love dos |
| 00:53.18 | wasup | and its permissions |
| 00:53.20 | marajin | you can screw a mac up with virii but the market is too small for anyone to /bother/ |
| 00:53.31 | ginge | I can screw up a mac by just using it |
| 00:53.35 | marajin | ginge: That too |
| 00:53.48 | wasup | true... but still atleast it normally asks if its ok to "destroy this mac" |
| 00:53.50 | wasup | hahaha |
| 00:54.09 | wasup | or type in your password to kill your mac |
| 00:54.26 | ginge | can you drag the whole macos into the trash can? |
| 00:54.36 | wasup | i lol |
| 00:54.36 | marajin | ginge: Anyway what we did to the PCs was make an infinite loop script that launched iexplore and set it to run every 15 minutes... then we stood around and watched |
| 00:55.14 | ginge | marajin: a long time ago i wrote a program that moves the windows away fom the mouse and put them on some machines. funy to watch |
| 00:55.34 | marajin | ginge: So suddenly a half dozen PCs opened a bunch of IE windows till they hardlocked and the PC world cretins went "huuuh? **prods keyboard a bit** **hits reset**...." 15 minutes later... "huuuh?" |
| 00:55.57 | marajin | ginge: Heh =) I had a program in the days of the Sinclair QL that turned the screen upside down. |
| 00:55.57 | ginge | I bet they set it up to reboot every 14 :) |
| 00:56.06 | marajin | ginge: Nah they just did it manually. |
| 00:56.14 | ginge | marajin: I had some games that did that. dizzy? |
| 00:56.24 | marajin | ginge: ooh dizzy I remember dizzy.. |
| 00:56.38 | marajin | ginge: But nah I jut had it on key press the screen inverted. |
| 00:57.07 | ginge | i'm sure there was a level you had to do standing on your head in that game... (goes to google) |
| 00:57.10 | marajin | ginge: One guy actually put his screen upside down so he could see it again, pressed his next key which promptly flipped it back again.. |
| 00:57.20 | ginge | haha |
| 00:57.30 | wasup | ginge: whats the link to the new zImage? |
| 00:57.35 | wasup | i hate tech support |
| 00:57.45 | ginge | same as before |
| 00:58.05 | ginge | http://www.headfuzz.co.uk/files/android/zImage-titan |
| 00:58.24 | wasup | k |
| 00:58.33 | marajin | My dad used to do goofy stuff with the QLs.. they used them where he worked... so he wrote preprogrammed scripts that made them appear to have conversations |
| 00:58.47 | marajin | cause there was another guy that believed the computer was really talking to him |
| 00:59.25 | wasup | lol |
| 00:59.31 | marajin | He spent an hour trying to comfort the computer because it'd been left on all night and it was lovesick about another computer across the room and alsorts |
| 00:59.36 | ginge | heh |
| 00:59.51 | ginge | thats pretty funny |
| 01:00.15 | marajin | well it was the 80s |
| 01:00.16 | marajin | heh |
| 01:00.45 | wasup | ginge: k booting it again |
| 01:00.47 | ginge | uh oh tv is starting to loop arouund |
| 01:01.28 | wasup | ginge: no keys work now |
| 01:01.38 | ginge | wasup: thought that might happen |
| 01:01.55 | wasup | oh..well im not blinded from the reflection of my flash anymore |
| 01:01.58 | wasup | :D |
| 01:02.14 | *** join/#htc-linux matsuu_ (n=matsuu@eb46.opt2.point.ne.jp) |
| 01:06.13 | *** join/#htc-linux Othello_ (i=Magorium@gateway/tor/x-892b0727afdd78fd) |
| 01:12.09 | pigeon | cr2: looking at haret's code, arch/mach/arch-sa.cpp is pretty much just a stub, which might explain why lots of commands/features are not present on the wallaby. |
| 01:12.49 | ginge | wasup: you fancy another download? |
| 01:14.05 | pigeon | cr2: and i probably want to add wallaby to machlist.txt too |
| 01:14.24 | wasup | yep.. im down. |
| 01:14.57 | wasup | hahaha... dude everyone i work with is alcoholics... my first beer in a few months |
| 01:15.34 | wasup | ginge: is the download up? |
| 01:16.00 | ginge | yeah |
| 01:16.03 | wasup | k |
| 01:17.50 | wasup | unable to register kbd irq!!! |
| 01:17.52 | wasup | ? |
| 01:18.00 | wasup | and none of the keys work |
| 01:19.15 | ginge | okay, try download again |
| 01:19.34 | wasup | k |
| 01:22.54 | wasup | DUDE! |
| 01:22.59 | wasup | it works but backwards |
| 01:23.07 | wasup | q = r |
| 01:23.08 | ginge | define backwards? |
| 01:23.12 | wasup | w =e |
| 01:23.19 | wasup | e= w |
| 01:23.23 | wasup | r=1 |
| 01:23.30 | ginge | 1? |
| 01:23.32 | wasup | r=q |
| 01:23.33 | wasup | lol |
| 01:23.44 | ginge | well, I kinda expected that |
| 01:23.49 | ginge | do they all work now? |
| 01:24.08 | ginge | i'm sure its getting warmer. marajin? |
| 01:24.16 | wasup | no just those and shift = U? |
| 01:24.19 | marajin | yeah? |
| 01:24.20 | wasup | or v? |
| 01:24.27 | ginge | how hot man? |
| 01:24.33 | wasup | i cant tell the screen is small? |
| 01:24.48 | ginge | so some keys dont work |
| 01:24.53 | marajin | how hot what? |
| 01:24.57 | ginge | is it |
| 01:24.57 | wasup | just the first 4 |
| 01:24.59 | wasup | qwer |
| 01:25.09 | wasup | and shift |
| 01:25.41 | marajin | 25 C apparently? |
| 01:28.16 | ginge | wasup: I have a minor update if you want to test |
| 01:28.21 | wasup | k |
| 01:28.47 | ginge | the keys will still report wrong, just want to know if any more or less work |
| 01:28.54 | wasup | k |
| 01:32.20 | wasup | no the same...but shift now moves the cursor back |
| 01:33.08 | wasup | do u think you have the gpios data? |
| 01:33.13 | wasup | i mean irq |
| 01:33.40 | wasup | i can re do it in haretconsole if you want |
| 01:34.56 | ginge | I need the gpios for the colums but I think wm does it by masking the gpio irq for the scan point... basically not sure it will help |
| 01:35.32 | wasup | hm. |
| 01:35.44 | wasup | well just let me know |
| 01:40.21 | ginge | last one before I head off to sleep |
| 01:40.26 | wasup | ]k |
| 01:40.29 | wasup | k |
| 01:41.37 | ginge | i think the camera may flash :) |
| 01:42.13 | wasup | ill put my sunglasses on |
| 01:43.24 | wasup | r no longer works |
| 01:43.30 | wasup | the rest is still backwards |
| 01:43.34 | wasup | with caps? |
| 01:43.41 | wasup | i dont know if it was caps before |
| 01:43.49 | wasup | shift still moves cursor back |
| 01:44.12 | ginge | no flash? |
| 01:44.23 | wasup | nope? |
| 01:44.27 | ginge | hmm |
| 01:44.33 | wasup | nope |
| 01:45.41 | marajin | Hunger strikes.. |
| 01:46.13 | wasup | hahaha |
| 01:46.26 | marajin | I could kill for a curry atm =/ |
| 01:46.53 | wasup | whats that? |
| 01:47.07 | marajin | what's a curry? |
| 01:47.20 | wasup | atm? |
| 01:47.30 | marajin | At The Moment |
| 01:47.39 | wasup | oh..so whats a curry? |
| 01:48.06 | marajin | Indian food |
| 01:48.37 | wasup | ok thats what i thought |
| 01:48.47 | wasup | spicey....that does sound good |
| 01:49.10 | marajin | I do love my chilli |
| 01:49.21 | wasup | lol |
| 01:49.33 | marajin | There's a curry they sell in the UK where you have to sign a waiver before they'll serve it to you, in case it kills you |
| 01:49.41 | wasup | what? |
| 01:49.45 | wasup | r u serious/ |
| 01:49.49 | marajin | yeah |
| 01:49.54 | wasup | wtf... hahaha |
| 01:49.59 | wasup | i need to try that |
| 01:50.22 | marajin | If you have a weak heart then it could trigger a heart attack/failure |
| 01:50.32 | wasup | i got a good heart |
| 01:51.16 | marajin | well if you eat the curry completely then they don't charge you for it |
| 01:52.31 | wasup | dang |
| 01:52.55 | marajin | The manager comes out to pose with you and they put the photo on the wall too |
| 01:52.55 | marajin | h |
| 01:53.10 | wasup | that must be hot... there is a place here that does the same but for their buffalo wings... |
| 01:53.18 | wasup | but no waiver i dont think |
| 01:53.50 | marajin | Well they didn't used to have one but I think someone got hospitalized once |
| 01:53.57 | marajin | so they thought theyd play it safe ya know/ |
| 01:54.59 | wasup | oh |
| 01:55.06 | wasup | that sucks |
| 01:57.23 | marajin | yah :P |
| 01:59.45 | wasup | yep. |
| 02:00.35 | marajin | semi afk, gonna get to cooking some fuds |
| 02:01.25 | wasup | afk? |
| 02:01.39 | wasup | oh |
| 02:01.43 | wasup | away from keyboard |
| 02:02.52 | ginge | wasup: one last one? |
| 02:03.04 | wasup | ok |
| 02:03.30 | wasup | when i close my keyboard on the other one... i get a crap load of VVVVVVVVVVV |
| 02:03.33 | wasup | hahaha |
| 02:05.42 | ginge | okay, at least we know which that one is. |
| 02:05.52 | wasup | we do? |
| 02:06.27 | wasup | nothing works :( |
| 02:06.59 | wasup | request_irq failed for input 41 |
| 02:07.08 | wasup | ,irq 105 |
| 02:07.30 | ginge | okay |
| 02:07.58 | wasup | whats that mean? |
| 02:08.45 | ginge | I wrote some debug code to dump a whole load of stuff, one of the things I want to dump failed |
| 02:08.57 | wasup | oh |
| 02:10.18 | ginge | oh well, I will get back to this at some point. good night |
| 02:10.58 | wasup | ginge: good night |
| 02:12.18 | marajin | pah |
| 02:19.18 | AstainHellbring | bah |
| 02:22.31 | marajin | I agree! |
| 02:29.18 | marajin | So what's up Astain? |
| 02:34.59 | AstainHellbring | not much |
| 02:35.03 | AstainHellbring | just playing at home |
| 02:35.16 | marajin | Sounds.. ominous |
| 02:35.18 | Kevin2 | cr2: wirq (and hence mmutrace) requires AllocPhysMem. That's not typically on a strongarm cpu. It should be possible to emulate AllocPhysMem, but that hasn't been done. |
| 02:36.00 | AstainHellbring | eh playing with my son |
| 02:36.14 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-76-88-212-37.bak.res.rr.com) |
| 02:40.36 | marajin | How old is he? |
| 02:42.47 | AstainHellbring | 13 months |
| 02:42.54 | *** join/#htc-linux tmzt_mp900c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 02:43.07 | marajin | Cute, but ok then, switch ominous for dangerous ;) |
| 02:43.13 | marajin | kids that age always poke me inna eye |
| 02:43.50 | *** join/#htc-linux the_sys0p` (n=the_sys0@cpe-76-88-212-37.bak.res.rr.com) |
| 02:50.53 | AstainHellbring | mines good with me |
| 02:51.29 | marajin | wears his glasses just in case ;) |
| 02:51.41 | AstainHellbring | lol |
| 02:51.51 | AstainHellbring | he loves to steal those |
| 02:52.18 | marajin | superglues 'em to his ears... |
| 03:03.35 | *** join/#htc-linux LunohoD_ (n=alex@e180065083.adsl.alicedsl.de) |
| 03:17.02 | *** join/#htc-linux tmzt_mp900c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 03:30.06 | *** join/#htc-linux tmzt_mp910c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 04:20.34 | *** join/#htc-linux tmzt_mp900c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 04:22.28 | tmzt_mp900c | irssi is really smart, 910 does come after 900 ! |
| 04:23.32 | tmzt_mp900c | dcordes: hi |
| 04:23.39 | tmzt_mp900c | dcordes: or, ping? |
| 04:33.12 | marajin | tmzt_mp900c: 910 DOES come after 900 |
| 04:33.24 | marajin | tmzt_mp900c: It uses old QBASIC line numbering ;) |
| 04:49.19 | *** join/#htc-linux zwox (i=zwox@par69-7-82-243-234-94.fbx.proxad.net) |
| 05:17.33 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 05:17.51 | dcordes | morning |
| 05:18.05 | dcordes | tmzt_mp900c: pong |
| 05:18.50 | tmzt_mp900c | hey, |
| 05:18.58 | dcordes | hi there |
| 05:19.17 | tmzt_mp900c | how are things going with X? |
| 05:19.36 | dcordes | to be honest, I didn't do anything since yesterday |
| 05:19.37 | tmzt_mp900c | I was reading logs, then found they were old, then started reading again, |
| 05:20.01 | tmzt_mp900c | alright, I can't really remember what we were working on, though |
| 05:20.33 | dcordes | don't have the data flatrate yet |
| 05:20.34 | tmzt_mp900c | if you start without -dumb, do you get update messages that match where on the screen you tap? |
| 05:20.38 | tmzt_mp900c | right |
| 05:21.08 | dcordes | that's a strange thing with the Invalid update messages. sometimes they appear, sometimes not |
| 05:21.09 | tmzt_mp900c | do you have to build everything to get a zImage? |
| 05:21.26 | dcordes | noth happens when just starting with the TS_* variables. |
| 05:21.27 | tmzt_mp900c | ok, when they do have the matched where you tapped on the screen? |
| 05:21.52 | tmzt_mp900c | it was before we added all those command line options though, right? |
| 05:22.21 | tmzt_mp900c | did you get the straces, do you see event anywhere in an open? |
| 05:22.45 | tmzt_mp900c | and do you have mdev, hal, and dbus installed in the image? |
| 05:23.01 | tmzt_mp900c | and lshal/lshw? |
| 05:23.03 | dcordes | yes, they appeared before we added those command line options. maybe they also appear with them, I think I didn't look after them. |
| 05:23.14 | tmzt_mp900c | even with -dumb? |
| 05:23.46 | tmzt_mp900c | do you have xev? |
| 05:24.07 | dcordes | I don't know if they appear with the options. I think I did not see them with them yet |
| 05:24.22 | tmzt_mp900c | the events might not be core, so they would not affect the cursor, but should be in xev |
| 05:24.46 | tmzt_mp900c | have you been watching #xorg, (sorry for all these questions) |
| 05:25.12 | dcordes | ./usr/lib/libhal.so.1 |
| 05:25.44 | dcordes | dbus is present |
| 05:25.47 | dcordes | no mdev |
| 05:25.52 | tmzt_mp900c | or udev? |
| 05:26.03 | dcordes | udev is present |
| 05:26.07 | tmzt_mp900c | ok |
| 05:26.15 | tmzt_mp900c | (mdev is small udev) |
| 05:26.26 | tmzt_mp900c | (used in some initrds on pc, at least) |
| 05:26.35 | dcordes | and no xev |
| 05:26.40 | dcordes | ok |
| 05:26.56 | tmzt_mp900c | we need xev, should be in xorg-apps or xorg-utils, its an old x program |
| 05:27.37 | dcordes | now I will check for the error messages, first w/o, then with the options. redirecting them is Xfbdev 2>fbdev.txt? |
| 05:27.48 | tmzt_mp900c | for stderr, yes |
| 05:28.05 | tmzt_mp900c | can we get those in oe, maybe for tommorow? |
| 05:28.58 | dcordes | xev? |
| 05:29.47 | tmzt_mp900c | x event dumper program, shows what events the x server is processing |
| 05:29.59 | dcordes | no I mean which ones do we need? |
| 05:30.10 | tmzt_mp900c | yes xev |
| 05:30.15 | tmzt_mp900c | lshal or lshw |
| 05:31.27 | dcordes | oh I just noticed I copied the movie before I turned off yesterday |
| 05:31.41 | dcordes | the pc |
| 05:33.29 | dcordes | mmc errors trying to play it |
| 05:35.37 | tmzt_mp900c | sync? |
| 05:35.43 | tmzt_mp900c | did you use sync? |
| 05:35.53 | dcordes | after I wrote it on the host? |
| 05:36.02 | tmzt_mp900c | yes, or umount properly |
| 05:36.09 | dcordes | I always umount the sds |
| 05:36.26 | dcordes | just did an fsck.ext2 and it fixed some errors. |
| 05:36.30 | dcordes | I will put a different video |
| 05:37.13 | tmzt_mp900c | ok |
| 05:37.45 | tmzt_mp900c | might be dma or something though, so proably as ginge about it |
| 05:37.51 | tmzt_mp900c | s/as/ask |
| 05:38.08 | dcordes | yes I should report it to him. it said something about dma |
| 05:38.16 | dcordes | and FLUSH failed |
| 05:38.18 | tmzt_mp900c | get those messages if you can |
| 05:38.34 | tmzt_mp900c | sounds like a mmc command, probably printks in mmc driver |
| 05:39.02 | tmzt_mp900c | (not a standard linux thing, that I know of) |
| 05:39.33 | tmzt_mp900c | playing movies is a good test of mmc as well, but we are checking the fbdev |
| 05:40.08 | tmzt_mp900c | can you add CONFIG_OABI_COMPAT to zImage if you are going to rebuild |
| 05:40.17 | tmzt_mp900c | to rebuild it |
| 05:40.43 | dcordes | ok |
| 05:40.56 | tmzt_mp900c | did you find xev and lshal? |
| 05:41.26 | dcordes | let me check |
| 05:41.30 | ptl[lamisil] | is there a way for me to get one of these ROMs with videocall and install only the dialer program on my HTC Athena? |
| 05:41.48 | tmzt_mp900c | that would be an xda question wouldn't it? |
| 05:42.51 | ptl[lamisil] | is there a channel for that? |
| 05:42.54 | ptl[lamisil] | it that xda? |
| 05:43.30 | ptl[lamisil] | no #xda channel |
| 05:44.15 | dcordes | ptl[lamisil]: check xda-developers.com |
| 05:44.35 | tmzt_mp900c | forums |
| 05:44.49 | dcordes | it would be great if OE would be able to parse the metadata faster |
| 05:44.55 | tmzt_mp900c | there are threads for all the different "cooked" roms |
| 05:45.01 | dcordes | bitbaking xev |
| 05:45.35 | tmzt_mp900c | is athena with the x86 and arm? |
| 05:45.47 | tmzt_mp900c | or is that the phone/tablet |
| 05:45.52 | ptl[lamisil] | I know that, tmzt_mp900c, thanks. I have also downloaded the ROM of interest. But I couldn't find any way to extract just the dialler to install on my device |
| 05:45.53 | dcordes | I think that is shift what you mean |
| 05:46.02 | ptl[lamisil] | tmzt_mp900c: athena X7501 |
| 05:46.04 | ptl[lamisil] | not shift |
| 05:46.12 | ptl[lamisil] | Intel XScale - ARM processor |
| 05:46.14 | tmzt_mp900c | ok, was confused by that one |
| 05:46.31 | tmzt_mp900c | the people who make the roms know how to extract dialers or whatever |
| 05:46.48 | tmzt_mp900c | does it use sip or is it a special thing on your network? |
| 05:46.51 | ptl[lamisil] | I'll try posting there then. thanks. |
| 05:47.00 | ptl[lamisil] | no, I have a good SIP phone already |
| 05:47.11 | ptl[lamisil] | I want a regular dialer that can have videocall using my primary camera |
| 05:47.12 | ptl[lamisil] | that's all |
| 05:47.13 | ptl[lamisil] | :) |
| 05:47.20 | tmzt_mp900c | the back camera? |
| 05:47.33 | ptl[lamisil] | yes |
| 05:47.40 | tmzt_mp900c | what I meant was is there a special way that you network handles video calls? |
| 05:47.47 | tmzt_mp900c | or is it sip based |
| 05:47.48 | ptl[lamisil] | because I dont' have a frontone |
| 05:47.56 | ptl[lamisil] | hm |
| 05:48.17 | tmzt_mp900c | might be a registry setting which one, or the dialer might be specific to one device |
| 05:48.30 | ptl[lamisil] | I don't know enough of the 3G protocol to say that. I just know my 3G provider supports it. |
| 05:48.43 | tmzt_mp900c | is say that because you asked before whether the network had to support it (reading logs) |
| 05:48.52 | tmzt_mp900c | I say that |
| 05:49.06 | ptl[lamisil] | I've googled for it and couldn't find any way to change it on the registry... I found a "videocall_out.cab" which I installed and it almost worked, but it's buggy as hell yet. |
| 05:49.28 | tmzt_mp900c | but it works? |
| 05:49.54 | ptl[lamisil] | no, it connects on the videocall but no image appears and a few seconds later it bails out. |
| 05:50.22 | tmzt_mp900c | you might have to have dedicated ip (static ip) or sip proxy to make it work |
| 05:50.45 | tmzt_mp900c | same as your sip phone, I would imagine, you could even try that account |
| 05:50.53 | tmzt_mp900c | but not all proxies support video |
| 05:50.54 | ptl[lamisil] | ? |
| 05:50.57 | ptl[lamisil] | 3G uses SIP? |
| 05:51.06 | ptl[lamisil] | hm |
| 05:51.21 | tmzt_mp900c | no, video might use sip or it might be native over 3g somehow, but I would assume sip |
| 05:51.29 | ptl[lamisil] | I am sorta confused here, maybe I should know more about this subject... |
| 05:51.39 | tmzt_mp900c | do you have to be connected to internet (does IE work), to try to make that call? |
| 05:51.51 | ptl[lamisil] | oh, it seems that it is native under 3G. My operator does not require any specific configuration for it (if you have a front camera, that's it). |
| 05:51.55 | ptl[lamisil] | no |
| 05:52.01 | tmzt_mp900c | SIP is a protocol for setting up sessions, usually voice but also video |
| 05:52.02 | ptl[lamisil] | I don't have to |
| 05:52.30 | ptl[lamisil] | I know. Microsoft Portrait for PPC which I have installed on my athena supports SIP calls and even SIP video. |
| 05:52.58 | tmzt_mp900c | is there a way to change the device name through a website or something, can you say I have whatever device not athena? |
| 05:54.04 | ptl[lamisil] | i don't think I understand your question, do you want me to change some device identifier? So that the dialer tries to make a videocall misidentifying it as, say, a X7500? |
| 05:54.22 | tmzt_mp900c | what is X7500? |
| 05:55.44 | ptl[lamisil] | there are different models of HTC Athena/Advantage: X7500 has a front camera besides the back camera. X7500, sold in the US, doesn't. X7510 is a later model with 16 GB hard drive and also back and front cameras. |
| 05:55.57 | ptl[lamisil] | sorry, let me rephrase that |
| 05:56.04 | ptl[lamisil] | there are different models of HTC Athena/Advantage: X7500 has a front camera besides the back camera. X7501, sold in the US, doesn't. X7510 is a later model with 16 GB hard drive and also back and front cameras. |
| 05:56.09 | ptl[lamisil] | now it's right :) |
| 05:56.59 | tmzt_mp900c | the dlls or usually specific to the camera, and often the camera.exe, which xda forums often call a camera, is also specific |
| 05:57.19 | tmzt_mp900c | there doesn't seem to be one standard on win CE for cameras |
| 05:58.19 | ptl[lamisil] | hmm |
| 05:58.30 | ptl[lamisil] | well.. maybe it's impossible then |
| 05:58.34 | ptl[lamisil] | :( |
| 06:01.01 | tmzt_mp900c | dcordes: are you building those images (OE and zImage)? |
| 06:01.41 | ptl[lamisil] | I hope then that when HTC Linux is ready, we'll have an universal driver for the cameras so that in it I could make videocalls using the back or the front camera. |
| 06:01.49 | tmzt_mp900c | I will have to see you in a few hours, |
| 06:02.49 | tmzt_mp900c | ptl: I would suggest tracking down that cab and asking who ever made it what to do, or asking in the thread for your device (athena/advantage) on xda, if you have network connections you might try howardforums.com |
| 06:04.30 | ptl[lamisil] | tmzt_mp900c: that's the problem, I could never find anything about this cab. I just found it in a file listing on pocketpt.net |
| 06:06.48 | dcordes | tmzt_mp900c: yea OE is still working, needed to upgrade cross toolchain |
| 06:07.57 | dcordes | and the old eabi kernel is done |
| 06:19.54 | dcordes | tmzt_mp900c: xev still takinga while, bbiab. let's see if I can get me a cheap data plan |
| 06:21.48 | tmzt_mp900c | ok, talk to you later then |
| 06:30.17 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbf9f09.pool.einsundeins.de) |
| 06:35.15 | *** join/#htc-linux no2chem2 (n=user@h-72-245-84-225.lsanca54.dynamic.covad.net) |
| 06:49.29 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 07:04.14 | *** join/#htc-linux CVirus (n=Satan@82.201.178.25) |
| 07:04.25 | *** part/#htc-linux CVirus (n=Satan@82.201.178.25) |
| 07:07.04 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 07:09.56 | *** join/#htc-linux Bendy (n=sniveri@bb-81-175-215-228.dsl.phnet.fi) |
| 07:15.37 | *** join/#htc-linux Zoolooc_ (n=fredsiba@nrbg-4dbf9f09.pool.einsundeins.de) |
| 07:34.33 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 07:41.49 | *** join/#htc-linux tmzt_mp900c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 08:00.39 | *** join/#htc-linux TeringTuby (n=maarten@195-241-125-243.ip.telfort.nl) |
| 08:29.32 | *** join/#htc-linux goxboxlive (n=goxboxli@31.84-48-212.nextgentel.com) |
| 08:59.34 | *** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo) |
| 09:59.11 | *** join/#htc-linux smev (n=smev@S010600215514f029.cg.shawcable.net) |
| 10:20.43 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
| 11:38.00 | *** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo) |
| 11:43.14 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 11:56.43 | *** join/#htc-linux pH5 (n=ph5@p5485D253.dip.t-dialin.net) |
| 11:59.59 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 12:05.05 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d866565.pool.einsundeins.de) |
| 12:10.36 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
| 12:22.47 | *** join/#htc-linux no2chem2 (n=user@h-68-165-18-84.lsanca54.dynamic.covad.net) |
| 12:39.27 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 12:44.10 | *** join/#htc-linux BabelO (n=Fabrice@2a01:e35:2ee1:c1c0:21e:8cff:fe2d:7d3f) |
| 12:57.25 | BabelO | hi kiozen |
| 13:07.52 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
| 13:20.09 | *** join/#htc-linux Marex (n=vasum7am@auryn.karlin.mff.cuni.cz) |
| 13:24.13 | cr2 | Kevin2: how much RAM do you need for mmutrace with AllocPhysMem ? |
| 13:25.29 | Kevin2 | Hi cr2. Not much - it is the sizeof(struct irqData). |
| 13:27.55 | Kevin2 | I don't know the exact size, but it's probably around 300K. |
| 13:30.48 | cr2 | Kevin2: hi. ok, so it will probably need some recursive attempts. |
| 13:31.59 | BabelO | hi |
| 13:32.15 | BabelO | cr2: any idea how can i control suspend devices order ? |
| 13:32.57 | Kevin2 | cr2: Yeah - you'd want to call allocPages with the "memory_required * 2" or something, sort by physical address and see how big a consecutive address you can find. If it's big enough - great. If not, retry with "memory_required *4" or something... |
| 13:42.42 | pigeon | cr2: btw, there's no AllocPhysMem and FreePhysMem in the coredll on my wallaby, and hence mmutrace and some other commands are not available in haret. |
| 13:50.20 | cr2 | pigeon: ok. |
| 13:50.58 | cr2 | BabelO: you can't, and you will try to add some locks, it may end in a deadlock. |
| 13:52.04 | cr2 | Kevin2: i don't think an 1MB allocation will really fail. |
| 13:52.17 | pigeon | i've added wallaby to machlist.txt, filled in some irqs and gpios in arch-sa.cpp, but that's pretty much it for now. |
| 13:52.50 | cr2 | pigeon: have you modified the python gpio/irq decoder ? |
| 13:52.53 | Kevin2 | cr2: It has to be physically continuous. But yes - I think it should be possible to emulate AllocPhysMem for that size. |
| 13:53.39 | pigeon | cr2: ah, nope not yet |
| 13:53.41 | cr2 | Kevin2: yes, i remember watching the big allocations, they are very continuous |
| 13:54.34 | cr2 | pigeon: i see from the wiki that the buttons and touchscreen are bitbang spi-based ? |
| 13:55.17 | cr2 | pigeon: i think we should document more asic registers in the memory map |
| 13:56.49 | pigeon | cr2: if i know how, i would :) |
| 13:57.17 | cr2 | pigeon: do you have a bootloader somewhere around ? |
| 13:59.16 | cr2 | pigeon: i have put the cpu docs here http://jornada820.sourceforge.net/docs/arm/ |
| 14:01.06 | Kevin2 | pigeon / cr2: If you're willing to test - I'll try coding up an emulated AllocPhysMem. |
| 14:01.43 | cr2 | pigeon: so if you will dump the registers, you can add IRQ here http://wiki.xda-developers.com/index.php?pagename=WallabyGPIO |
| 14:02.02 | cr2 | Kevin2: thanks, that will be great. i can test it on jornada820 |
| 14:04.53 | pigeon | Kevin2: sure. |
| 14:05.02 | pigeon | cr2: wait, i'll upload it. |
| 14:10.30 | pigeon | cr2: http://pigeond.net/~pigeon/wallaby/wallaby.nb0 |
| 14:13.24 | cr2 | pigeon: got it. |
| 14:16.00 | cr2 | pigeon: send your haret path to Kevin2 / haret ML |
| 14:16.28 | cr2 | teqp r1, #d3 |
| 14:16.34 | cr2 | hehe. old good times :) |
| 14:17.18 | pigeon | cr2: i also have somewhat annotated asm of the bootloader from someone, it might not match the one i've got though. |
| 14:17.35 | pigeon | and also an ida idb file, but i cannot open it anymore. |
| 14:18.13 | cr2 | wow. i wish the resume was so simple on n560 ;) |
| 14:18.22 | cr2 | jump to 0x41000 |
| 14:19.14 | cr2 | pigeon: if you can't open it, it may be from some very old version. |
| 14:19.50 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
| 14:19.52 | pigeon | yeah, i tried in a recent version of ida, it says it wants to convert it, then it reckons it's created by a pirate version of ida :) |
| 14:21.43 | cr2 | mine is legitimate. but old. |
| 14:22.10 | cr2 | should not be a problem for sa1100 |
| 14:22.26 | cr2 | works very well even for the diamond. |
| 14:23.07 | pigeon | cr2: http://pigeond.net/~pigeon/wallaby/bl517.asm.gz is a generated asm |
| 14:23.56 | pigeon | cr2: http://pigeond.net/~pigeon/wallaby/idb517.zip is the idb i got |
| 14:24.30 | cr2 | have it. |
| 14:24.46 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d866565.pool.einsundeins.de) |
| 14:24.57 | pigeon | bed time for me here... nite all. |
| 14:25.10 | pigeon | cr2: thanks for you help, i'll catch you again soon. |
| 14:25.35 | pigeon | i'll send the haret patch later. |
| 14:25.40 | cr2 | ok |
| 14:26.18 | cr2 | hi kiozen |
| 14:26.26 | kiozen | hi |
| 14:28.44 | cr2 | kiozen: i wish the resume sequence on n560 was like on wallaby. 10 machine commands. |
| 14:29.00 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 14:29.14 | kiozen | cr2: me too, especially as booting takes > 1min |
| 14:29.24 | cr2 | kiozen: ok, the battery monitor. |
| 14:29.32 | kiozen | but I am happy that I can boot without failure |
| 14:30.20 | kiozen | yes let's focus on the monitor, one step at a time |
| 14:30.47 | *** join/#htc-linux jeanseb (n=jeanseb@88.164.32.155) |
| 14:34.18 | cr2 | kiozen: i have 3-4 gpio combinations followed by 2 spi functions |
| 14:34.38 | kiozen | ok |
| 14:34.40 | cr2 | 2 spi functions should send some command, and get the response |
| 14:35.03 | cr2 | the commands should refer to 2 channels i guess. |
| 14:35.11 | cr2 | batt/temp ? |
| 14:35.20 | cr2 | ADC channels. |
| 14:35.34 | cr2 | because the TS ADC functions are separate. |
| 14:36.06 | kiozen | ok |
| 14:36.26 | cr2 | the gpio combinations are controlled by CPLD GPIOG0-2 |
| 14:37.22 | cr2 | if "both" ADC channels are queried, then |
| 14:37.38 | cr2 | you do |
| 14:37.46 | cr2 | cpld6(0,0,0) |
| 14:37.49 | cr2 | msleep 2 |
| 14:38.13 | cr2 | write ssp cmd, get ssp answer1 |
| 14:38.22 | cr2 | msleep 100 |
| 14:38.31 | cr2 | cpld6(1,0,0) |
| 14:38.42 | cr2 | <PROTECTED> |
| 14:38.46 | cr2 | msleep 100 |
| 14:39.10 | cr2 | this is probably for the battery voltage and temperature ? |
| 14:40.16 | cr2 | then there is some initial sequence. |
| 14:40.26 | cr2 | cpld6(0,0,0) |
| 14:40.28 | cr2 | msleep 2 |
| 14:40.40 | cr2 | write ssp cmd, get ssp answer |
| 14:40.51 | cr2 | msleep 100 |
| 14:41.04 | cr2 | cpld6(0,0,1) |
| 14:41.10 | cr2 | msleep 2 |
| 14:41.16 | cr2 | write ssp cmd, get ssp answer |
| 14:41.22 | cr2 | msleep 100 |
| 14:41.49 | cr2 | sometimes, on init the cpld6(1,1,0) is set |
| 14:42.21 | cr2 | and finally there is some gpio83-state dependent sequence. |
| 14:42.34 | cr2 | cpld6(1,1,0) |
| 14:42.39 | cr2 | msleep 100 |
| 14:42.51 | cr2 | write ssp cmd1, get ssp answer1 |
| 14:42.57 | cr2 | msleep 100 |
| 14:43.24 | cr2 | some kioctl (probably the cpu frequency/voltage related) |
| 14:43.35 | cr2 | cpld6(0,1,0) |
| 14:43.38 | cr2 | msleep 2 |
| 14:43.52 | cr2 | write ssp cmd1, get ssp answer1 |
| 14:44.41 | cr2 | get_cpld(3,2) <- this one may be set and unset inside the battery monitor module. stop charging ? |
| 14:45.04 | cr2 | ok, that's all what i've gathered. |
| 14:45.34 | kiozen | sounds like black magic to me |
| 14:45.42 | cr2 | imho it should be possible to decode the write ssp cmd1, get ssp answer1 |
| 14:45.49 | cr2 | and write ssp cmd2, get ssp answer2 |
| 14:46.12 | cr2 | to know which ADC is addressed and read here. |
| 14:46.34 | cr2 | but this 15bit story makes it a real PITA. |
| 14:46.52 | kiozen | you have an idea why it is 15bit? |
| 14:46.57 | cr2 | no. |
| 14:47.01 | kiozen | surely not because of the sport |
| 14:47.25 | cr2 | on other devices the tsc2046 is addressed in 24bit mode. |
| 14:47.41 | cr2 | such mode is documented in the tsc2046 datasheet. |
| 14:47.56 | cr2 | but we are able to use it in the 8bit mode. |
| 14:48.11 | cr2 | the ADC can be set to 8bit or 12bit mode. |
| 14:48.37 | kiozen | ok, let's try to record the values from the ssp access, maybe their meaning is obvious |
| 14:48.39 | cr2 | to read a 12bit value in one bus cycle, you need a >12bit mode |
| 14:49.00 | cr2 | you've already tried it, but the result was not obvious for me. |
| 14:49.23 | cr2 | ok, do you have the tsc2046 datasheet ? you can pick it from TI site. |
| 14:50.01 | kiozen | did I? so far I just remember doing powermon |
| 14:50.32 | cr2 | you've traced the ssp1 together with cpld6 |
| 14:51.05 | cr2 | i'm not sure about the purpose of cpld(3,2) and gpio83 and gpio11. |
| 14:51.17 | cr2 | but the monitoring itself can be decoded RSN. |
| 14:51.54 | cr2 | because we can find out which channels are read, and which cpld gpios are enabled to read this data. |
| 14:52.42 | *** join/#htc-linux EA2 (n=chipper@cpe-024-074-138-191.carolina.res.rr.com) |
| 14:52.49 | kiozen | ah, ok |
| 15:08.02 | *** join/#htc-linux kiozen_ (n=oeichler@rgnb-5d866565.pool.einsundeins.de) |
| 15:26.50 | Kevin2 | cr2 / pigeon : http://www.handhelds.org/~koconnor/haret/haret-20080726.exe |
| 15:27.44 | cr2 | Kevin2: downloading |
| 15:27.50 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
| 15:34.42 | cr2 | Kevin2: mmutrace is not in the 'help' commandlist ? |
| 15:34.57 | Kevin2 | cr2: help vars? |
| 15:35.32 | cr2 | yes, it's there :) |
| 15:35.56 | *** join/#htc-linux no2chem2 (n=user@h-67-101-50-22.lsanca54.dynamic.covad.net) |
| 15:37.25 | cr2 | Kevin2: got an error |
| 15:37.29 | cr2 | Found 62 continuous pages by allocating 124 virtual pages |
| 15:37.29 | cr2 | Can't find vm addr of alloc'd physical ram 00000000 |
| 15:37.29 | cr2 | Can't allocate memory for irq code |
| 15:37.29 | cr2 | UnmapViewOfFile failed 43200000 (code 487) |
| 15:39.32 | Kevin2 | cr2: Hrmm - it couldn't find a handy wince map of all memory.. Can you post full haretlog.txt. |
| 15:40.06 | cr2 | yes, i'll rerun it now. |
| 15:42.23 | Kevin2 | http://www.handhelds.org/~koconnor/haret/haret-20080726b.exe -- should fix confusing debug messages. |
| 15:42.51 | cr2 | Welcome, this is HaRET pre-0.5.2-20080726_112625 running on WindowsCE v2.11 |
| 15:42.58 | cr2 | ok, i'll pastebin the output |
| 15:44.37 | Kevin2 | cr2: "dump mmu 1" would help too. |
| 15:45.57 | cr2 | Kevin2: i have the full list |
| 15:46.49 | cr2 | http://pastebin.no/13478 |
| 15:47.22 | cr2 | 16mb. hmm |
| 15:48.40 | cr2 | Kevin2: afair the LockPages was the only reliable way to pinpoint the address. |
| 15:51.39 | Kevin2 | cr2: allocPages does use LockPages() |
| 15:51.59 | Kevin2 | cr2: Hrmm. Can you rerun "wirq" with the "b" version above? |
| 15:54.30 | *** join/#htc-linux dcordes_ (n=dcordes@unaffiliated/dcordes) |
| 15:55.06 | cr2 | Kevin2: ok |
| 15:56.32 | cr2 | HaRET(1)# wirq 5 |
| 15:56.32 | cr2 | Found 62 continuous pages by allocating 124 virtual pages |
| 15:56.32 | cr2 | Can't find vm addr of alloc'd physical ram 4327B000 |
| 15:56.32 | cr2 | Can't allocate memory for irq code |
| 15:57.20 | cr2 | Kevin2: 4327B000 looks like a virtual address |
| 15:58.05 | Kevin2 | Yeah. We successfully got a range of memory. Now we need to find a virtual address that is outside the haret process. It's strange that it is failing there. |
| 15:58.28 | Kevin2 | (We have a virtual address inside haret, we just need to translate that to a virtual address outside of haret.) |
| 15:59.58 | cr2 | ok |
| 16:00.27 | Kevin2 | cr2: Can you grab "c" version? This just adds a little more debugging. |
| 16:00.36 | cr2 | Kevin2: user mode ? |
| 16:00.37 | cr2 | ok |
| 16:02.18 | cr2 | Found 62 continuous pages by allocating 124 virtual pages |
| 16:02.18 | cr2 | cachedMVA: addr=4327B000 paddr=ffffffff newaddr=00000000 |
| 16:02.18 | cr2 | Can't find vm addr of alloc'd physical ram 4327B000 |
| 16:02.18 | cr2 | Can't allocate memory for irq code |
| 16:14.33 | Kevin2 | cr2: Odd. That's reporting that virtToPhys() is not working. Can you try running "show v2p(0x80000000)" and "show v2p(0x09201000)"? |
| 16:15.10 | cr2 | HaRET(2)# show v2p(0x80000000) |
| 16:15.10 | cr2 | 0x00000000 |
| 16:15.26 | cr2 | HaRET(3)# show v2p(0x09201000) |
| 16:15.26 | cr2 | 0xffffffff |
| 16:15.55 | Kevin2 | What about "show v2p(0x80100000)" ? |
| 16:16.09 | cr2 | HaRET(4)# show v2p(0x80100000) |
| 16:16.10 | cr2 | 0x00100000 |
| 16:16.41 | Kevin2 | show v2p(0x02032000) ? |
| 16:16.55 | cr2 | HaRET(5)# show v2p(0x02032000) |
| 16:16.55 | cr2 | 0xffffffff |
| 16:17.32 | Kevin2 | It appears we can v2p 1meg maps, but nothing else. If you go through the output of "dump mmu" can you see if v2p resolves any 4K pages? |
| 16:19.31 | *** join/#htc-linux tmzt_mp900c (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 16:20.31 | cr2 | yes |
| 16:21.03 | cr2 | 091a0000 | c0dea000 | Small (4K) | CB AP=3333 |
| 16:21.17 | cr2 | and many others. |
| 16:21.32 | Kevin2 | And it returns the correct addr (eg, 0xc0dea000)? |
| 16:25.43 | cr2 | no, because it's the zero bank |
| 16:26.32 | cr2 | HaRET(7)# show v2p(0xffffc000) |
| 16:26.33 | cr2 | 0xc0079000 |
| 16:26.36 | cr2 | this one is ok. |
| 16:27.00 | cr2 | ffffc000 | c0079000 | Small (4K) | CB AP=1120 |
| 16:27.35 | cr2 | 43180000 | c0deb000 | Small (4K) | CB AP=3333 |
| 16:27.49 | cr2 | HaRET(8)# show v2p(0x43180000) |
| 16:27.50 | cr2 | 0xffffffff |
| 16:27.52 | Kevin2 | cr2: Can you grab "d" version of haret. This one wont free after an error. So we can dump the mmu table after the error and see why it wont resolve. |
| 16:27.57 | cr2 | ok, this one does not work. |
| 16:28.08 | cr2 | ok. |
| 16:29.56 | cr2 | HaRET(2)# wirq 5 |
| 16:29.56 | cr2 | Found 62 continuous pages by allocating 124 virtual pages |
| 16:29.56 | cr2 | cachedMVA: addr=4327B000 paddr=ffffffff newaddr=00000000 |
| 16:29.56 | cr2 | Can't find vm addr of alloc'd physical ram 4327B000 |
| 16:29.57 | cr2 | Can't allocate memory for irq code |
| 16:30.24 | Kevin2 | Okay - you should be able to run "dump mmu" and see what was at 0x4327B000 |
| 16:30.53 | tmzt_mp900c | pxa25x is supported without this, right? |
| 16:31.01 | cr2 | <PROTECTED> |
| 16:31.02 | cr2 | <PROTECTED> |
| 16:31.02 | cr2 | <PROTECTED> |
| 16:31.02 | cr2 | <PROTECTED> |
| 16:31.03 | cr2 | <PROTECTED> |
| 16:31.03 | cr2 | <PROTECTED> |
| 16:31.17 | cr2 | 43200000 | | UNMAPPED | |
| 16:31.20 | cr2 | that's all |
| 16:31.48 | cr2 | tmzt_mp900c: it's sa1100 |
| 16:32.12 | tmzt_mp900c | is there some reason that lsmod does not work on ce.net then? |
| 16:32.19 | tmzt_mp900c | 4.2 |
| 16:33.04 | cr2 | tmzt_mp900c: is it registered ? |
| 16:33.07 | tmzt_mp900c | if I remember right, I could only trace about 64 bytes of i/o with mmutrace |
| 16:33.26 | Kevin2 | cr2: Can you restart haret and run "alloctest 124"? |
| 16:33.26 | tmzt_mp900c | no, I don't know if it recognizes Plat |
| 16:33.39 | cr2 | Kevin2: afair LockPages should be run only on the memory allocated with VirtualAlloc |
| 16:33.50 | cr2 | Kevin2: yes |
| 16:34.46 | cr2 | HaRET(1)# alloctest 124 |
| 16:34.50 | cr2 | pg#: <virt> <phys> |
| 16:34.54 | cr2 | 000: 43200000 c0dea000 |
| 16:34.58 | cr2 | 001: 43201000 c0de7000 |
| 16:35.02 | cr2 | 002: 43202000 c0de6000 |
| 16:35.06 | cr2 | 003: 43203000 c0de5000 |
| 16:35.06 | cr2 | 004: 43204000 c0de4000 |
| 16:35.07 | cr2 | 005: 43205000 c0de3000 |
| 16:35.07 | cr2 | 006: 43206000 c0de0000 |
| 16:35.07 | cr2 | ... |
| 16:35.37 | cr2 | tmzt_mp900c: what machine do you have ? |
| 16:35.46 | cr2 | ah, motoQ |
| 16:35.56 | cr2 | have you added it in machlist ? |
| 16:37.23 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 16:39.04 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-87-174.dsl.akrnoh.ameritech.net) |
| 16:43.49 | tmzt | cr2: this is the NEC mp900c not motoq, the device I was logged into irc with linux, rebooting it to ce.net |
| 16:44.15 | tmzt | motoq is pxa27x and wm5, and it is detected, but I have had to set mtype to boot |
| 16:45.40 | cr2 | ok |
| 16:45.59 | tmzt | what is latest haret for pxa? |
| 16:46.46 | cr2 | compiled by Kevin2 |
| 16:47.00 | cr2 | it#s for all (arm) devices |
| 16:47.06 | tmzt | hh/~koconnor ? |
| 16:47.09 | cr2 | yes |
| 16:47.13 | tmzt | ok |
| 16:47.24 | cr2 | tmzt: no video accel on mp900c ? |
| 16:47.36 | tmzt | no, it is sis something |
| 16:47.48 | tmzt | that's not right |
| 16:48.01 | cr2 | 1 x VGA-out |
| 16:48.03 | cr2 | ok |
| 16:48.16 | cr2 | you can see the fbaddr ? |
| 16:48.32 | tmzt | I have to get a console first |
| 16:48.40 | tmzt | (haret) |
| 16:48.58 | tmzt | vga is special connector so I don't know if it works |
| 16:49.08 | cr2 | i think mp900c is on oe. |
| 16:49.29 | tmzt | when haret boots linux, the boot messages appear on console bot not the haret boot messages |
| 16:49.52 | tmzt | I have to hide cursor or it leaves a blank spot on the fb in linux |
| 16:50.19 | tmzt | but I am looking for the usb chip which is four byte registers in i/o somewhere |
| 16:50.33 | cr2 | "so far keyboard, touchscreen, lcd/framebuffer, pcmcia/cf (the basics) are working" |
| 16:50.35 | tmzt | I have datasheet for it |
| 16:51.10 | cr2 | tmzt: run 'dump mmu' first |
| 16:51.15 | tmzt | yes, I have been using jlime and debian with jlime Xfbdev, and hostap with wpa quite successfully until recently |
| 16:51.38 | tmzt | I am going to try my pppd line, since synce-hal doesn't seem to want to work |
| 16:51.44 | tmzt | hold on |
| 16:57.13 | Kevin2 | cr2: Can you pull "e" version? It has more debugging info. |
| 17:00.24 | cr2 | Kevin2: hh.org does not respond now. |
| 17:00.31 | cr2 | ok, got it |
| 17:01.34 | cr2 | HaRET(1)# wirq 5 |
| 17:01.38 | cr2 | Found 62 continuous pages by allocating 124 virtual pages |
| 17:01.39 | cr2 | from phys c0d69000=88D69000 |
| 17:01.39 | cr2 | p2v(4327B000)=ffffffff |
| 17:01.39 | cr2 | with retry=c0d69000 |
| 17:01.40 | cr2 | cachedMVA: addr=4327B000 paddr=c0d69000 newaddr=88D69000 |
| 17:01.40 | cr2 | irq:8003c7d4@A80783F8=88da50ec abort:8003c778@A80783F0=88da5110 prefetch:8003c5b8@A80783EC=88da513c data=88d6c000 sizes=c:000013e4,t:0003d3e4 |
| 17:02.02 | cr2 | and then |
| 17:02.06 | cr2 | Beginning memory tracing. |
| 17:02.07 | cr2 | Replacing windows exception handlers... |
| 17:02.07 | cr2 | Finished installing exception handlers. |
| 17:02.07 | cr2 | Restoring windows exception handlers... |
| 17:02.07 | cr2 | Finished restoring windows exception handlers. |
| 17:02.08 | cr2 | Handled 9295 irq, 5163 abort, 48605 prefetch, 0 lost, 0 errors |
| 17:03.19 | cr2 | Kevin2: now it locked the machine. |
| 17:05.20 | Kevin2 | Did you trace something that locked the machine, or it just locked up on second run? |
| 17:07.14 | tmzt | 20080726? |
| 17:07.36 | cr2 | Kevin2: i've added mmutrace 1MB for the gpio area |
| 17:07.43 | cr2 | Kevin2: then it locked. |
| 17:07.56 | cr2 | p2v(0x90000000) |
| 17:08.04 | cr2 | ah. |
| 17:09.13 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-76-88-212-37.bak.res.rr.com) |
| 17:09.22 | tmzt | Detected machine Generic Intel PXA/PXA (Plat='Jupiter' OEM='NEC') |
| 17:09.43 | Kevin2 | Well, it's weird. LockPages returned successfully and gave a whole bunch of physical addresses. But, the pages weren't actually in the mmu table. It's no big deal to touch the page before resolving, but it is odd. |
| 17:10.36 | cr2 | Kevin2: old wince was delaying mapping the pages as long as possible |
| 17:10.45 | tmzt | is it same git with seabios? |
| 17:11.15 | cr2 | seabios ? |
| 17:11.17 | Kevin2 | cr2: Sure - but after calling LockPages! That's odd. |
| 17:11.37 | tmzt | it says last change 4 weeks ago |
| 17:12.03 | Kevin2 | tmzt: What are you looking for? Seabios is an unrelated project I have at linuxtogo.org. |
| 17:12.17 | tmzt | haret source so I can add Plat |
| 17:12.25 | tmzt | Detected machine Generic Intel PXA/PXA (Plat='Jupiter' OEM='NEC') |
| 17:13.04 | Kevin2 | tmzt: It's on cvs server at handhelds.org. See: http://www.handhelds.org/moin/moin.cgi/HaRET |
| 17:13.15 | tmzt | ok |
| 17:18.28 | Kevin2 | cr2: Does wirq work on your device, or it still just hangs? |
| 17:19.42 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
| 17:20.22 | tmzt | should I use cegcc 4.3.0 or an earlier version? |
| 17:20.51 | tmzt | (it's been a while since I've build haret) |
| 17:21.08 | *** join/#htc-linux CPress (n=cpress@CPE-72-129-218-217.new.res.rr.com) |
| 17:21.20 | Kevin2 | tmzt: I'm using the v4.1.0 package. I haven't tried the v4.3 one yet. As far as I know it should work though. |
| 17:21.54 | tmzt | the sf downloads makes no sense, I don't even see 4.1.0 |
| 17:22.04 | tmzt | and it is mingw32ce? |
| 17:22.35 | Kevin2 | tmzt: Yeah, you need mingw32ce. |
| 17:23.09 | *** join/#htc-linux CPress (n=cpress@CPE-72-129-218-217.new.res.rr.com) |
| 17:23.22 | Kevin2 | I use: mandriva-cegcc-mingw32ce-0.51.0-1.i586.rpm |
| 17:24.15 | cr2 | Kevin2: plain wirq works. |
| 17:25.47 | cr2 | Kevin2: does not work after softreset |
| 17:26.04 | cr2 | HaRET(1)# wirq 60 |
| 17:26.08 | cr2 | Found 62 continuous pages by allocating 62 virtual pages |
| 17:26.12 | cr2 | from phys c0dc3000=88DC3000 |
| 17:26.16 | cr2 | p2v(4323D000)=ffffffff |
| 17:26.17 | cr2 | with retry=c0dc3000 |
| 17:26.17 | cr2 | cachedMVA: addr=4323D000 paddr=ffffffff newaddr=00000000 |
| 17:26.17 | cr2 | Can't find vm addr of alloc'd physical ram 4323D000 |
| 17:26.17 | cr2 | Can't allocate memory for irq code |
| 17:27.09 | Kevin2 | cr2: grab "f" version. This has the debugging cleaned up and uses "retryVirtToPhys" for real. It should prevent the above. |
| 17:27.59 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
| 17:32.05 | dcordes | tmzt: hi I just ordered 300mb rate |
| 17:32.35 | dcordes | tmzt: got the X program thing and oabi kernel compiled. |
| 17:37.15 | cr2 | Kevin2: yes, mmutrace works now :) |
| 17:37.30 | cr2 | addlist mmutrace p2v(0x18000000) 0x1000 |
| 17:37.37 | cr2 | it's the sa1101 |
| 17:37.52 | cr2 | 000.000 80059ab8: str r3, [r2] # a4000400 =00000062 |
| 17:37.52 | cr2 | 000.000 80059ac4: streq r7, [r6] # a4000000 =00000019 |
| 17:38.02 | Kevin2 | cr2: Wow, that's odd. I didn't really change anything with 'f' version. |
| 17:38.34 | cr2 | bbl, need to buy some food for the weekend |
| 17:39.57 | dcordes | ginge: playing _any_ sort of videof rom mmc crashes |
| 17:40.25 | dcordes | ginge: it gives DMA and flush errors. (I'm in the git kernel) |
| 17:43.24 | dcordes | cr2: just get pizza tonno and nizza salad |
| 17:45.59 | *** join/#htc-linux smev (n=smev@S010600215514f029.cg.shawcable.net) |
| 17:49.39 | dcordes | bbl |
| 17:54.17 | tmzt | what's password for hh.org anon cvs? |
| 17:54.36 | tmzt | it was right there in the wiki, sorry |
| 17:56.29 | tmzt | <PROTECTED> |
| 18:00.31 | tmzt | do I add to machlist.txt and generate somehow? |
| 18:05.06 | ginge | dcordes: any large sd operation gives dma errors. flush is the cleanup after the kernel warn |
| 18:05.23 | tmzt | ginge: have you tried mplayer also? |
| 18:06.05 | ginge | tmzt: no, I can cause the crash backtrace 100% of the time |
| 18:06.15 | tmzt | on .25? |
| 18:06.22 | tmzt | I think that was on .24 |
| 18:06.25 | ginge | tmzt: .24 and .25 |
| 18:06.51 | tmzt | oh, we triggered it with mplayer trying to test fbdev |
| 18:06.57 | tmzt | large as in? |
| 18:07.10 | tmzt | is FLUSH your debugging info |
| 18:07.14 | ginge | tmzt: it varies. It depends how how heavily you load irqs |
| 18:07.22 | ginge | tmzt: flush is the buffer purge |
| 18:07.36 | ginge | tmzt: you can crash sooner if you tap the screen and hammer the keys |
| 18:07.37 | tmzt | your printk? |
| 18:07.49 | ginge | tmzt: nope. Part of the sd driver |
| 18:08.01 | tmzt | is it an mmc command then, flush? |
| 18:08.07 | tmzt | or is that the controller? |
| 18:09.21 | ginge | its part of the host controller msm_sdcc.c its not am mmc command. the flush is triggered after the mmc returns error code 0x8000002. Flush is literally what it says. It is purging the buffers of the data and freeing the handles |
| 18:09.36 | tmzt | ok |
| 18:09.44 | tmzt | we just were not sure before |
| 18:10.04 | ginge | it is caused by handling an invalid irq somehwhere |
| 18:10.44 | ginge | ... I think |
| 18:10.50 | tmzt | the dma controller and sd are on arm11 side? |
| 18:11.03 | ginge | yeah |
| 18:11.49 | tmzt | how close are we on bluetooth or wifi on kaiser? |
| 18:12.12 | tmzt | we are going to try a gprs ssh -R trick, but dcordes has to get data working first |
| 18:12.20 | ginge | bluetooth should be easy enough, just not looked into it. wifi is on sdio, which has issues |
| 18:12.36 | ginge | tmzt: yeah, I use that to get into my phone |
| 18:12.39 | tmzt | ssh -R 2222:localhost:22 remotemachine |
| 18:13.06 | tmzt | can we get bt, at least serial |
| 18:13.18 | tmzt | so scp will work |
| 18:13.29 | ginge | tmzt: probably. whats the problem with scp? |
| 18:13.30 | tmzt | or nc, or cat /dev/rfcomm0, or whatever |
| 18:13.33 | Kevin2 | tmzt: Just add to machlist.txt -- the makefile will automatically build the appropriate files. |
| 18:13.44 | tmzt | ok, thanks, Kevin2 |
| 18:14.02 | tmzt | the addlist gpios are optional? |
| 18:14.19 | tmzt | keyboard and touchscreen are on BTUART |
| 18:14.29 | tmzt | so that would be defined by pxa then? |
| 18:14.48 | Kevin2 | tmzt: addlist gpios is optional. |
| 18:14.51 | tmzt | ok |
| 18:15.07 | tmzt | keyboard and touchscreen work anyway |
| 18:15.34 | tmzt | Jornada is also Jupiter, that is ce.net 4.2 right? |
| 18:15.37 | ginge | gtg. cya later. |
| 18:15.43 | tmzt | ok |
| 18:16.02 | Kevin2 | The "addlist gpios" allows one to add gpios that aren't part of the standard pxa list. (For example, a cpld gpio expander present on many htc models.) |
| 18:16.44 | Kevin2 | The jornada is a ce2.11 version. cr2 posted the init log at: http://pastebin.no/13478 |
| 18:17.16 | tmzt | PLATFORM=Jupiter |
| 18:17.17 | tmzt | Jornada820, SA ,"HP, Jornada 820", JORNADA820 |
| 18:17.59 | Kevin2 | You don't need to specify the wince version in machlist.txt (haret doesn't care). You just need to have the cpu model type (eg, "SA" for jornada). |
| 18:18.27 | tmzt | right, how does it detect then, since platform and oem are the same? |
| 18:18.34 | tmzt | it's pxa25x |
| 18:18.54 | tmzt | Detected machine Generic Intel PXA/PXA (Plat='Jupiter' OEM='NEC') |
| 18:18.55 | tmzt | CPU is Intel XScale ARM arch 5TE revision 3 product 16 stepping 6 running in system mode |
| 18:19.11 | tmzt | OEM is NEC! |
| 18:19.31 | tmzt | this should be the only pxa-based oem NEC |
| 18:19.42 | Kevin2 | Use something like: MyMachine, PXA,"NEC", NEC_MP900 |
| 18:19.48 | tmzt | the second field |
| 18:20.32 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d866565.pool.einsundeins.de) |
| 18:20.45 | tmzt | ok |
| 18:21.12 | tmzt | I just copy cegcc to /opt/ ? |
| 18:21.26 | tmzt | minw32ce |
| 18:21.31 | tmzt | mingw32ce |
| 18:22.25 | Kevin2 | Make expects to find: /opt/mingw32ce/bin/arm-wince-mingw32ce-g++ --- You can modify BASE in Makefile if you want to put it somewhere else. |
| 18:23.16 | tmzt | there's no g++ only gcc |
| 18:23.27 | tmzt | and c++filt |
| 18:24.04 | tmzt | change CXX? |
| 18:24.17 | Kevin2 | $ rpm -qf /opt/mingw32ce/bin/arm-wince-mingw32ce-g++ |
| 18:24.17 | Kevin2 | cegcc-mingw32ce-0.51.0-1.i586 |
| 18:24.38 | Kevin2 | You should have gotten a g++ compiler. |
| 18:24.40 | tmzt | I am on Ubuntu, and I don't see debs |
| 18:24.46 | tmzt | I can alien though |
| 18:27.01 | tmzt | there's only a mandriva rpm of mingw32ce 0.51 |
| 18:27.13 | Kevin2 | I don't know. Does the tar.gz from sf.net have the g++ compiler in it? |
| 18:28.46 | tmzt | also mandriva, but I am downloading, should be lsb anyway |
| 18:29.48 | tmzt | if I am connected through the usb and I want to trace access to usb, is that good or bad? |
| 18:31.23 | Kevin2 | tmzt: I've traced wifi over wifi. Just be aware that you can end up getting tons of traces (because a trace report can generate more trace reports). |
| 18:31.44 | tmzt | I am logging with haretconsole |
| 18:32.48 | tmzt | I can try to get one of my 16-bit pcmcia wired nic working if I need to, but the device hardresets everytime I boot linux so I have not done it |
| 18:35.31 | tmzt | I would then need a usb keyboard or something to generate access? |
| 18:35.57 | tmzt | (about the only other class that works is storage, not even serial) |
| 18:38.08 | Kevin2 | You're trying to trace usb host? If your device has that, then I guess you'd need to plug something into the usb host. |
| 18:38.42 | tmzt | usb host and usb device are on same chip, four sequential byte registers |
| 18:39.17 | tmzt | driver for linux exists, but we do not have the addresses |
| 18:39.18 | cr2 | Kevin2: have you added the XDA aka HTC wallaby to the machlist ? |
| 18:39.59 | cr2 | tmzt: 'dump mmu' first, and check the RAM. if you can dump the nk.exe from RAM it will help a lot. |
| 18:40.07 | cr2 | Kevin2: here it is -> |
| 18:40.10 | cr2 | Trying to detect machine (Plat='PocketPC' OEM='PW10A1') |
| 18:40.10 | cr2 | Wince reports processor: core=ARM name=SA1110 cat= vend=INTEL |
| 18:41.03 | cr2 | xda SA1100_XDA XDA 357 |
| 18:41.04 | tmzt | still building new haret |
| 18:41.36 | tmzt | the tar has g++ |
| 18:42.30 | tmzt | /opt/mingw32ce/libexec/gcc/arm-wince-mingw32ce/4.1.0/cc1plus': execv: Text file busy |
| 18:43.14 | tmzt | sorry |
| 18:46.37 | Kevin2 | cr2: Okay - I checked it in. |
| 18:46.40 | tmzt | yes, I am looking for usb host, which is about the only thing not working in linux, and about useless in ce (no drivers) |
| 18:46.40 | Kevin2 | bbl |
| 18:47.20 | cr2 | Kevin2: thanks. |
| 18:48.04 | cr2 | tmzt: you should dump the bootloader/nk.exe |
| 18:48.18 | cr2 | pwf nec900 0x0 0x40000 |
| 18:48.50 | cr2 | and let the trained monkey (me :) look at it. |
| 18:50.01 | tmzt | I have haret! |
| 18:52.00 | *** join/#htc-linux LunohoD_ (n=alex@e180068176.adsl.alicedsl.de) |
| 18:54.27 | tmzt | paste mmu or whole log (and is .no longer term?) |
| 18:56.20 | tmzt | should nk.exe be 256k? |
| 18:56.23 | cr2 | paste as much as you can :) |
| 18:56.31 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 18:56.38 | cr2 | no, it's (may be) the bootloader |
| 18:56.51 | cr2 | i'll let you know when i'll see the 256K |
| 18:57.14 | cr2 | it depends on whether your wince kernel is xip of not. |
| 18:57.26 | tmzt | I wasn't pasting the 256k, just the mmu |
| 18:57.44 | tmzt | it's in nor as far as I know |
| 18:58.16 | cr2 | yes. i'd like to see both. |
| 18:58.50 | cr2 | xip means in nor, not xip - sdram at 0xa0000000 |
| 19:01.51 | tmzt | cmonex is in #hpcdev and she would know a lot more about then I do |
| 19:02.03 | tmzt | I have asked her about the xip question |
| 19:03.00 | cr2 | tmzt: let me see the 256K |
| 19:03.06 | tmzt | is bootloader often a seperate ce on these devices |
| 19:03.11 | tmzt | ok, where can I paste it? |
| 19:03.16 | tmzt | as bin |
| 19:03.24 | cr2 | you can't paste, it's binary |
| 19:03.35 | *** join/#htc-linux LunohoD__ (n=alex@e180068059.adsl.alicedsl.de) |
| 19:03.36 | cr2 | can i download it ? |
| 19:04.18 | tmzt | I can message you an ip |
| 19:05.34 | cr2 | ok |
| 19:11.01 | *** join/#htc-linux pikapika (n=pikapika@mar75-8-88-164-227-147.fbx.proxad.net) |
| 19:11.40 | pikapika | yo |
| 19:24.27 | tmzt | now if I can only find a port that isn't blocked |
| 19:33.47 | tmzt | "Filter NAT internet redirection" disables everything, not just bad source routes, how nice |
| 19:37.57 | cr2 | tmzt: ok. looks like a bootloader :) |
| 19:38.15 | tmzt | you use ida? |
| 19:38.23 | cr2 | MP900 BSQUARE boot loader |
| 19:38.27 | cr2 | i use 'strings' |
| 19:38.28 | tmzt | strings |
| 19:38.31 | tmzt | right |
| 19:38.48 | tmzt | would there be usb stuff in the bootloader? |
| 19:39.08 | cr2 | strings -el looks fun. |
| 19:39.40 | cr2 | it decompresses zlib-compressed code into ram, it seems. |
| 19:40.03 | cr2 | you should dump the first 2 megabytes of RAM, and run 'strings -el' on it. |
| 19:40.20 | *** join/#htc-linux Tanyafortinburg (i=d169c34b@gateway/web/ajax/mibbit.com/x-d022387eea8374f5) |
| 19:42.03 | cr2 | logo.bmp :) |
| 19:42.15 | cr2 | it seems 256K is a bit small- |
| 19:47.21 | cr2 | tmzt: i think i have the virtual mapping |
| 19:47.32 | tmzt | for phys io? |
| 19:48.00 | Tanyafortinburg | anyone knwo in what channel i could find heavensrevenge? |
| 19:49.04 | cr2 | tmzt: full table |
| 20:16.11 | tmzt | do you still need dump mmu? |
| 20:18.22 | cr2 | tmzt: no. |
| 20:18.33 | cr2 | tmzt: do you have some wiki ? |
| 20:19.01 | cr2 | tmzt: i can post the list, you may tell me what is at these address spaces. |
| 20:19.26 | cr2 | need to format it first. |
| 20:24.18 | tmzt | wiki? |
| 20:24.29 | tmzt | I have tomboy |
| 20:24.51 | tmzt | that's where all my q stuff is anyway |
| 20:25.16 | tmzt | csv would be ok though |
| 20:25.26 | tmzt | or html tables |
| 20:25.43 | cr2 | tmzt: http://pastebin.no/13494 |
| 20:26.47 | cr2 | pcmcia and pxa internals are good visible |
| 20:26.54 | cr2 | the rest - don't know. |
| 20:46.33 | tmzt | so i/o isn't just under 0x40000000? |
| 20:46.49 | *** join/#htc-linux LunohoD_ (n=alex@e180073149.adsl.alicedsl.de) |
| 20:47.17 | cr2 | where is your framebuffer located ? |
| 20:47.38 | tmzt | what is the command if wince stuff doesn't work? |
| 20:49.01 | cr2 | print "%x" VRAM |
| 20:49.36 | tmzt | Unable to detect frame buffer address |
| 20:49.46 | tmzt | I can check the jlime source though |
| 20:50.26 | cr2 | hm. interesting |
| 20:50.52 | tmzt | and lsmod doesn't work |
| 20:51.03 | tmzt | I was going to try to find the usb driver and dump it |
| 20:53.32 | tmzt | Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF |
| 20:53.32 | tmzt | Detected machine Generic Intel PXA/PXA (Plat='Jupiter' OEM='NEC') |
| 20:53.32 | tmzt | CPU is Intel XScale ARM arch 5TE revision 3 product 16 stepping 6 running in system mode |
| 20:53.49 | tmzt | this is my haret with OEM NEC in the machlist.txt |
| 20:54.19 | tmzt | what is the replacement for lsmod? |
| 20:54.31 | tmzt | and ps |
| 20:55.12 | cr2 | if you don't have the toolhelp.dll there is no portable way to do it. |
| 20:55.37 | cr2 | does testwm5.exe work for you ? |
| 20:55.39 | tmzt | HaRET(4)# print "%x" mod2addr("nk.exe") |
| 20:55.39 | tmzt | line 4: Unknown variable 'mod2addr' in expression |
| 20:55.39 | tmzt | 3ce59bc |
| 20:55.39 | tmzt | HaRET(5)# print %s" addr2mod(0x40000000) |
| 20:55.39 | tmzt | line 5: Unknown variable 'addr2mod' in expression |
| 20:55.56 | tmzt | where is testwm5.exe? |
| 20:56.41 | cr2 | hands.com/~lkcl |
| 20:56.41 | tmzt | toolhelp is in the image? |
| 20:56.50 | cr2 | http://hands.com/~lkcl |
| 20:57.12 | cr2 | toolhelp should be in your image. |
| 20:57.49 | tmzt | it's not in \windows or \windows\system, so it must be in the rom image |
| 20:58.20 | tmzt | is there a way to find a dmesg type buffer in memory? |
| 20:58.50 | cr2 | only on certain htc devices. |
| 20:59.09 | cr2 | some images don't have toolhelp.dll |
| 21:00.09 | tmzt | windows icon in explorer, cannot find testwm5.exe or one of its components |
| 21:00.44 | tmzt | would the xip metadata help? |
| 21:03.26 | cr2 | ok |
| 21:03.50 | cr2 | no, the problem is more difficult. |
| 21:03.57 | tmzt | it didn't detect the machine either then, did it? |
| 21:04.11 | tmzt | Generic PXA |
| 21:04.41 | tmzt | mtype is 0 |
| 21:05.39 | tmzt | NEC is not in machines.cpp |
| 21:06.10 | tmzt | MP900, PXA, NEC MP900, NEC_MP900 |
| 21:06.15 | tmzt | is missed the quotes |
| 21:06.18 | tmzt | I |
| 21:07.16 | tmzt | no, thats only if there are commas in the name |
| 21:12.53 | tmzt | would it be enough to get a toolhelp.dll for ce.net 4.2 and put in cwd or \windows? |
| 21:14.32 | cr2 | it's device-specific |
| 21:14.38 | tmzt | ok |
| 21:14.51 | cr2 | if the producer left it out, you need to do some hardcore stuff instead. |
| 21:15.08 | cr2 | it's doable, just much more difficult |
| 21:15.39 | tmzt | I had a log of what addresses I was able to dump/trace I will see if I can find it if we need it |
| 21:16.24 | tmzt | I tried dumping most of 0x4.. before reading the datasheet and finding I was looking for 4 bytes, I saw repeating patterns in most of it |
| 21:17.00 | tmzt | what can we do with those maps? |
| 21:18.14 | cr2 | where can i find the nec900c kernel code ? |
| 21:21.53 | tmzt | +#define S1D13MP900_FB_PHYS 0x0C200000 |
| 21:22.02 | tmzt | http://www.jlime.org/gitweb?p=jlime.git;a=blob;f=kernel/patches-vanilla/v2.6.24-stable.patch;h=6fb0602c4922b49afcff552e7331d34063e66165;hb=HEAD |
| 21:22.10 | tmzt | second half is mostly mp900c |
| 21:22.33 | tmzt | the git has the patches applied but they are hard to sort out except by commit |
| 21:29.05 | cr2 | btw, how does one boot on sh3 wince devices ? |
| 21:29.31 | cr2 | Kevin2 said haret is too arm-specific to port it there. |
| 21:30.12 | tmzt | like jornadas? |
| 21:30.18 | cr2 | yes |
| 21:31.43 | cr2 | s1d3* is epson |
| 21:32.32 | tmzt | shlo.exe |
| 21:32.54 | cr2 | does it work on sh4 ? |
| 21:32.56 | tmzt | http://jlime.com/downloads/releases/donkey/bootloaders/shlo.exe for Jornada 680/690 |
| 21:33.02 | Kevin2 | cr2: Some parts of haret could be ported to other architectures - but someone would need to port cegcc first. |
| 21:33.03 | tmzt | I don't know |
| 21:33.21 | tmzt | do the dreamcast people have gcc? |
| 21:33.50 | tmzt | you asked about the fb memory |
| 21:34.05 | cr2 | Kevin2: i'd like to backport regs_pxa.py to regs_sa.py is the ARCH:SA autodetected ? |
| 21:34.23 | tmzt | fb, touchscreen, keyboard, cf work, probably serial port on side as well |
| 21:34.36 | cr2 | Kevin2: i was reading the sa1100 manual, it seems the 'set trace' and breakpoints should work there too. |
| 21:35.01 | cr2 | tmzt: i think there was sh-wince-gcc some time ago. |
| 21:35.12 | cr2 | even before arm-wince-gcc |
| 21:35.24 | cr2 | but i don't know if it's still working. |
| 21:35.40 | tmzt | mingw32ce is native pe? |
| 21:38.24 | cr2 | yes |
| 21:38.51 | cr2 | NMC1110 PRS/PRC register is at physical address 0x0900_0000 |
| 21:39.05 | tmzt | add these to the tab delim? |
| 21:39.07 | cr2 | do you have this datasheet ? |
| 21:39.21 | Kevin2 | cr2: Yeah - StrongArm should be autodetected. (It was on the Wallaby - your jornada isn't, but that's okay because we forced it to SA.) |
| 21:39.46 | cr2 | ok |
| 21:39.59 | cr2 | Kevin2: do i need to add GPCR and GPSR ? |
| 21:40.24 | cr2 | tmzt: here -> |
| 21:40.25 | Kevin2 | cr2: Yeah, I vaguely recall seeing that the StrongArm supported the debug unit. Someone would need to verify that the registers were the same - then it is just a matter of enabling it. |
| 21:40.27 | cr2 | 7575 +#define CF_CARD_REG 0x09000000 |
| 21:40.27 | cr2 | 7576 +#define PC_CARD_REG 0x0a000000 |
| 21:40.37 | Kevin2 | cr2: GPCR and GPSR ? |
| 21:40.59 | cr2 | Kevin2: yes, i don't see them being decoded by regs_pxa.py |
| 21:41.11 | tmzt | I've seen it, it must have been on devfiles.jlime.com before they reorg'd |
| 21:41.29 | cr2 | DBAR, DBVR, DBMR, DBCR are there. and p14 |
| 21:41.49 | tmzt | that is a multiplexer for the pxa cf host, if I understand it |
| 21:41.57 | tmzt | has cs and the bus |
| 21:42.23 | tmzt | there might be two of them, there are two cf/pcmcia slots |
| 21:42.44 | cr2 | tmzt: ok, so you can document these in the table |
| 21:43.02 | tmzt | ok |
| 21:43.17 | Kevin2 | cr2: Add archname = "SA"; |
| 21:43.17 | Kevin2 | to src/mach/arch-sa.cpp. |
| 21:43.30 | cr2 | Kevin2: ok. |
| 21:43.45 | Kevin2 | And then add memalias.RegsList['ARCH:SA'] = Regs_pxa |
| 21:43.45 | Kevin2 | to haretconsole/regs_pxa.py |
| 21:44.45 | cr2 | Kevin2: sa does not have so many gpios and CKEN as pxa. |
| 21:45.05 | tmzt | the git is broken out now and there is some usb related stuff I haven't seen before, but still only guesses I think |
| 21:45.49 | cr2 | Kevin2: sa1100 is missing |
| 21:45.53 | cr2 | <PROTECTED> |
| 21:45.53 | cr2 | <PROTECTED> |
| 21:46.06 | *** join/#htc-linux ChanServ (ChanServ@services.) |
| 21:46.06 | *** mode/#htc-linux [+o ChanServ] by irc.freenode.net |
| 21:46.34 | tmzt | + .start = 0x4c000000, |
| 21:46.34 | tmzt | + .end = 0x4c100000, |
| 21:46.44 | tmzt | http://www.jlime.org/gitweb?p=jlime-stable.git;a=blobdiff;f=arch/arm/mach-pxa/mp900.c;h=6cb60d9d25ceb020cdf2e847fececeb7b355dbe4;hp=ea47feecc55482e647812f8511bf3818e5ea3597;hb=d2774366574a9688d67814d492e9e8fd3cec8324;hpb=b8c5bbf19301abc3436f884d2a134c24e3cf1a31 |
| 21:47.11 | cr2 | Kevin2: hmm. my mingwce32 does not like sprintf() |
| 21:47.14 | cr2 | src/memory.cpp:538: error: 'sprintf' was not declared in this scope |
| 21:47.26 | tmzt | should I check dump mmu for that? |
| 21:47.28 | Kevin2 | cr2: SA110 will mach sa1100 -- it need only match the beginning of the string. |
| 21:47.34 | cr2 | ok |
| 21:50.32 | tmzt | 84c00000 | 04c00000 | 1MB section | CB AP=1 |
| 21:50.54 | Kevin2 | cr2: Are you having problems compiling haret? |
| 21:51.36 | cr2 | Kevin2: yes. |
| 21:51.47 | cr2 | tmzt: the framebuffer -> |
| 21:51.50 | cr2 | 8620 +#define S1D13MP900_FB_PHYS 0x0C200000 |
| 21:51.50 | cr2 | 8621 +#define S1D13MP900_REG_PHYS 0x0C000000 |
| 21:51.50 | cr2 | 8622 +#define S1D13MP900_FB_SIZE 0x00140000 |
| 21:53.34 | cr2 | Kevin2: it seems i need to update to the latest cegcc svn |
| 21:53.43 | Kevin2 | cr2: Hrmm. Can you grab the cegcc-mingw32ce-0.51.0 package? |
| 21:53.55 | Kevin2 | I'm not using svn - I just use the last published binaries. |
| 21:58.28 | *** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo) |
| 21:59.36 | cr2 | Kevin2: ok, i'll try |
| 21:59.49 | cr2 | tmzt: what about 'dump gpio' ? |
| 22:01.01 | cr2 | tmzt: you may try 'dump pxa27x' with this haret http://jornada820.sourceforge.net/files/haret/haret-w.exe |
| 22:01.22 | cr2 | tmzt: but some output will be incorrect, because it's for pxa27x |
| 22:04.19 | tmzt | mostly worked but got and exception and bl pwm is now blinking |
| 22:04.23 | tmzt | shutting it down |
| 22:04.55 | tmzt | the ts and probably bl or on a pic on btuart |
| 22:05.35 | tmzt | power led flashed too |
| 22:05.38 | *** join/#htc-linux LunohoD_ (n=alex@e180070035.adsl.alicedsl.de) |
| 22:07.21 | dcordes | aye |
| 22:07.28 | dcordes | jornada hacking? |
| 22:07.33 | tmzt | back in haret |
| 22:07.36 | tmzt | haret-w |
| 22:08.17 | cr2 | tmzt: dump gpio and dump pxa27x work ? |
| 22:08.31 | tmzt | that was dump pxa27x, haven't done gpio |
| 22:08.44 | cr2 | ok. |
| 22:08.48 | tmzt | it got an exception at the end and device went crazy |
| 22:09.07 | cr2 | for pxa255 it needs several adjustments. |
| 22:09.24 | tmzt | might just be this device, has to talk to pic all the time |
| 22:10.06 | cr2 | tmzt: you can trace the BTUART data register with haret mmutrace |
| 22:10.07 | tmzt | I have the dump |
| 22:10.13 | cr2 | pastebin ? |
| 22:12.56 | tmzt | http://pastebin.no/13510 |
| 22:14.25 | cr2 | very strange output. |
| 22:15.27 | tmzt | USBH and USBC are builtin to pxa? (or only pxa27x) |
| 22:16.09 | tmzt | STUART is ir? |
| 22:16.17 | tmzt | FFUART is the serial port? |
| 22:17.41 | tmzt | there is no mmc/sd smartmedia or anything else |
| 22:17.53 | tmzt | (except cf) |
| 22:18.29 | tmzt | dcordes: did you get those images built? |
| 22:18.35 | cr2 | pxa2[56]x does not have USBH |
| 22:19.52 | tmzt | are we looking for irq/gpio for usb to find the driver? |
| 22:20.02 | tmzt | let me get the dump gpio |
| 22:22.19 | tmzt | http://pastebin.no/13511 |
| 22:22.59 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 22:23.18 | tmzt | can we filter builtin pxa gpios? |
| 22:25.04 | tmzt | usb datasheet: http://www.jlime.org/gitweb?p=jlime.git;a=blob;f=datasheets/MobilePro900c/usb;h=96990c253ec1584cef8b76099795d78388d65dcb;hb=HEAD |
| 22:26.21 | dcordes | tmzt: yea all done |
| 22:26.40 | dcordes | tmzt: kernel image with oabi and the x application I just need to opkg install. |
| 22:26.57 | tmzt | alright |
| 22:27.05 | tmzt | hold on a second |
| 22:27.16 | tmzt | can you mount the sd |
| 22:27.43 | tmzt | you have the hal programs then |
| 22:27.47 | tmzt | ? |
| 22:28.23 | tmzt | how much space do you have on sd left? |
| 22:28.49 | dcordes | yea libhal and so |
| 22:28.54 | dcordes | I think it's normal in angstrom |
| 22:28.59 | dcordes | like 500 megabyte |
| 22:29.02 | tmzt | lshal or lshw |
| 22:29.09 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 22:29.09 | dcordes | both not present I think |
| 22:29.12 | tmzt | can you extract http://www.jlime.com/downloads/releases/henchman/icewm-image-Henchman-04.14.08.tar.bz2 to /jlime/ |
| 22:29.32 | dcordes | sure |
| 22:29.34 | tmzt | ok |
| 22:29.36 | tmzt | then boot |
| 22:30.06 | tmzt | careful though, its a full root so it will overwrite your image if you put it under root |
| 22:31.13 | dcordes | oh, ok |
| 22:31.17 | dcordes | putting it in jlime then |
| 22:31.38 | dcordes | but.. it's not so dramatic when it fscks up my image because I have backed up all the tslib settings |
| 22:31.45 | tmzt | mkdir /sd/jlime; cd /sd/jlime/; tar xjf /path/to/icewm*tar.bz2 |
| 22:31.46 | dcordes | and also your kernel patches |
| 22:31.52 | dcordes | done |
| 22:31.57 | tmzt | ok |
| 22:32.29 | tmzt | cr2: I will need to boot linux so if there is anything else I should do? |
| 22:34.48 | tmzt | but, thanks for your help |
| 22:35.15 | dcordes | tmzt: will I chroot into jlime? |
| 22:35.39 | tmzt | yeah, chroot /jlime |
| 22:35.48 | tmzt | wait, first |
| 22:36.10 | tmzt | mount -o bind /sys /jlime/sys |
| 22:36.21 | tmzt | mount -o bind /dev/ /jlime/dev |
| 22:36.27 | tmzt | mount -o bind /proc /jlime/proc |
| 22:36.31 | tmzt | mkdir if neccessary |
| 22:36.40 | tmzt | mount -o bind /dev /jlime/dev |
| 22:36.59 | tmzt | exit or ctrl-d will get you out |
| 22:39.30 | dcordes | I will put a script |
| 22:39.51 | tmzt | cr2: interesting while booting the output from haret (Detected ...) jumps to the left |
| 22:40.46 | tmzt | dcordes: I am in jlime, I think the same root as you |
| 22:41.50 | dcordes | ok cool |
| 22:41.54 | dcordes | will boot in a few |
| 22:42.10 | tmzt | ok |
| 22:42.43 | tmzt | we really are only using Xfbdev and /usr/lib/libts0*,/usr/lib/ts, which is what I use under debian |
| 22:42.49 | tmzt | but chroot will work fine |
| 22:45.26 | ginge | hi |
| 22:45.45 | tmzt | any luck on dma? |
| 22:45.53 | ginge | I only just got back in :) |
| 22:46.01 | tmzt | yeah, obviously |
| 22:46.45 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 22:47.23 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 22:48.27 | tmzt | actually, debian has libts now and that works so I use it, so only /usr/bin/Xfbdev from jlime |
| 22:48.39 | tmzt | for me |
| 22:51.58 | dcordes | ginge: hi there |
| 22:55.00 | dcordes | tmzt: ok oabi kernel seems to boot fine |
| 22:55.44 | marajin | huh... what actually counts as a sane torrent client for linux? |
| 22:55.56 | dcordes | I use transmission when it comes to gui |
| 22:56.04 | dcordes | otherwise rtorrent |
| 22:56.21 | ginge | oabi? |
| 22:56.28 | ginge | marajin: ktorrent works ok for me |
| 22:56.34 | dcordes | old abi |
| 22:56.37 | dcordes | support |
| 22:56.38 | tmzt | oabi compat |
| 22:56.40 | ginge | yeah seriously |
| 22:56.43 | marajin | ginge: It won't emerge the 2.x branch and 3.x needs KDE4 |
| 22:57.01 | dcordes | marajin: really gotta checkout transmission then |
| 22:57.02 | ginge | sucks to be on gentoo |
| 22:57.08 | dcordes | it's minimalistic, stable nice |
| 22:57.30 | ginge | until you want to install something really quick, which needs a full rebuild of x and kde ;) |
| 22:57.40 | dcordes | tmzt: ok chroot'D |
| 22:57.48 | tmzt | either the chroot, or (TSLIB_TSDEVICE=/dev/input/event0 /jlime/usr/bin/Xfbdev :0) & (sleep 20; chvt 1) |
| 22:57.51 | tmzt | ok |
| 22:58.03 | tmzt | same thing without the /jlime/ part |
| 22:58.08 | marajin | ginge: Quad xeon, emerge -ue world takes <= 6 hours :P |
| 22:58.23 | ginge | :O |
| 22:58.34 | dcordes | O |
| 22:58.40 | dcordes | I like ubuntu/debian |
| 22:59.10 | marajin | I use kubuntu on laptops and stuff but gentoo has been on my server for like 6 years |
| 22:59.38 | tmzt | until you start wondering about where the bins for i686 or armv6 or whatever are |
| 23:00.01 | marajin | I don't reinstall, I just cp -rP * |
| 23:01.28 | tmzt | marajin: I mv /* /old/ and debootstrap |
| 23:02.05 | marajin | notes his way is one less step? :p |
| 23:03.01 | tmzt | but renaming a few dirs on one fs is not that much, whats the cp -rP for? |
| 23:03.18 | dcordes | mmc error when I try and run Xfbdev |
| 23:03.27 | dcordes | is the Xfbdev binary in the jlime image like 10x biggerP? |
| 23:03.42 | marajin | tmzt: copy recursive preserve permissions. |
| 23:03.45 | tmzt | not sure, can you put it in ramdisk? |
| 23:03.56 | marajin | tmzt: Basically what I'm saying is, I move the rootfs to a new HDD when necessary |
| 23:04.04 | tmzt | oh |
| 23:04.33 | tmzt | thought you meant reinstall on one disk |
| 23:04.44 | marajin | nope, I'm saying I don't reinstall |
| 23:04.55 | marajin | I just move the rootfs where needed |
| 23:06.14 | marajin | that rootfs has seen everything from a P-166 to 1.2Ghz Athlon t-bird, 1.8Ghz Athlon-XP, Athlon64 3400, Athlon64 3200 (newer socket type/SSE3 instructions) and then to the quad xeon... |
| 23:06.18 | marajin | heh |
| 23:06.50 | dcordes | ginge: do you get the same dma mmc problems in .25? |
| 23:07.04 | ginge | yeah, pretty much |
| 23:07.06 | tmzt | there are obviously tradeoffs, |
| 23:07.08 | ginge | just looking at it now |
| 23:07.57 | dcordes | ok |
| 23:08.04 | tmzt | ls -lh /jusr/bin/Xfbdev ls -lh /jlime/usr/bn/Xfbdev (or du -h if you prefer) |
| 23:08.25 | tmzt | ls -lh /jusr/bin/Xfbdev /jlime/usr/bn/Xfbdev |
| 23:08.28 | ginge | I suggest using rd for now |
| 23:08.50 | tmzt | maybe tmpfs? |
| 23:08.59 | tmzt | dd bs=512 |
| 23:09.06 | tmzt | and not mmaping the image |
| 23:09.10 | marajin | ginge: ...well just to make no sense, ktorrent emerged |
| 23:09.16 | dcordes | tmzt: the jlime one is like 20K bigger |
| 23:09.28 | tmzt | ok |
| 23:09.37 | tmzt | are you in the chroot? |
| 23:09.44 | ginge | marajin: heh, gentoo for you |
| 23:09.45 | dcordes | no |
| 23:09.54 | ginge | marajin: not that any of the other are perfect |
| 23:10.01 | marajin | ginge: I checked 3 times with 3 different ebuilds.... and now just to try and find out what the error was, I emerged and it just did... |
| 23:10.04 | marajin | :p |
| 23:10.06 | tmzt | can you do just a mount? |
| 23:10.08 | ginge | marajin: I apt updated my laptop to a non boot last week |
| 23:10.11 | marajin | I've had a lot of issues with apt |
| 23:10.24 | marajin | yeah, I've apted a laptop into death before |
| 23:10.29 | ginge | marajin: yeah, like every damn update kills soemthing |
| 23:10.33 | dcordes | tmzt: you mean like mount /jlime to /mnt/jlime or so? |
| 23:10.46 | dcordes | :) |
| 23:10.56 | tmzt | I very intellegently dpkg -i /var/cache/apt/archives/* a machine with old and new packages, but still recovered it |
| 23:11.19 | tmzt | with a lot of dpkg --force-depends --purge |
| 23:11.38 | tmzt | no, just run mount |
| 23:12.00 | dcordes | ok what's the info required? |
| 23:12.08 | tmzt | is /tmp on tmpfs? |
| 23:12.46 | dcordes | there are 3 tmpfs |
| 23:13.08 | dcordes | on /var/volatile, /dev/shm, /media/ram |
| 23:13.24 | tmzt | dd if=/jlime/usr/bin/Xfbdev of=/media/ram/Xfbdev bs=512 |
| 23:13.29 | tmzt | use a smaller bs= if needed |
| 23:14.22 | dcordes | what's bs/ibs/obs? |
| 23:14.23 | *** join/#htc-linux nighthawk08 (n=a@212.143.154.159) |
| 23:14.36 | tmzt | block size |
| 23:14.39 | tmzt | in bytes |
| 23:15.36 | tmzt | basically, how much to copy at once, but it can make a difference on some media (such as tape) |
| 23:16.01 | dcordes | ok I see. why do you use dd for such an operation, and not e.g. cp? |
| 23:16.04 | tmzt | input, output, both |
| 23:16.05 | nighthawk08 | I am trying to run the Angstrom -- using the haret,. how do I know what is the name of the partition on the sd card? currently I've put mmcblk0p2 -- is that ok? because when I boot it gets stuck before loading the linux |
| 23:16.45 | tmzt | fdisk -l /dev/sd? in linux, disk management in windows |
| 23:17.23 | tmzt | do you have builtin or usb sd reader on you computer? |
| 23:17.24 | nighthawk08 | what should it specify? |
| 23:17.30 | dcordes | tmzt: done. |
| 23:17.33 | nighthawk08 | I got an sd reader... |
| 23:17.47 | cr2 | pigeon: are you sleeping ? |
| 23:18.08 | tmzt | the partition type should be 83 for ext2, b for fat32 |
| 23:18.26 | cr2 | pigeon: i've done a lot of dirty work here http://wiki.xda-developers.com/index.php?pagename=WallabyMemoryMap |
| 23:18.27 | tmzt | windows will recognize fat32 and say Unknown partition for ext2 |
| 23:18.58 | pigeon | cr2: morning! |
| 23:19.02 | nighthawk08 | I got two partitions on that SD -- one FAT second ext2 |
| 23:19.03 | cr2 | pigeon: now haret need some sa1110 changes (regs_sa.py) and mmutrace tracing |
| 23:19.07 | marajin | ginge: Ktorrent does seem ok, but can it be stopped from autostarting torrents? |
| 23:19.08 | cr2 | pigeon: lol |
| 23:19.26 | ginge | marajin: not sure |
| 23:19.45 | marajin | ginge: Nevermind, It seems to be a per-torrent option when you open the file |
| 23:19.53 | tmzt | somebody, is mmcblkp0 or mmcblkp1 the first partition? |
| 23:20.01 | dcordes | second |
| 23:20.10 | dcordes | mmcblk0p1 |
| 23:20.20 | tmzt | do you get a panic line? |
| 23:20.22 | ginge | marajin: check the plugins for the blocking of nasty plugin. |
| 23:20.24 | dcordes | in angstrom on zaurus that is |
| 23:20.49 | nighthawk08 | tmzt: when it sais jumping to Kernel.... -- then it is stuck |
| 23:21.08 | tmzt | that's not a root partition thing |
| 23:21.13 | cr2 | pigeon: the asic1 egpio will need a .py decoder too. like it's done for other htc egpio cplds. |
| 23:21.28 | nighthawk08 | tmzt: then what is it? |
| 23:21.33 | tmzt | what device, htc name? |
| 23:21.49 | nighthawk08 | htc himalaya - h3900 |
| 23:21.57 | cr2 | pigeon: the asic1/2 code is bitrotten, but it's there, and rather complete. |
| 23:21.57 | tmzt | stuck before loading the linux, I missed that |
| 23:22.37 | tmzt | what zImage are you using? |
| 23:22.48 | nighthawk08 | zImage-2.6.21-hh20-r6-h3900 |
| 23:23.03 | cr2 | pigeon: i'd say that with some hardcore patient tracing it'd be possible to support wallaby. |
| 23:23.04 | tmzt | is there newer for hima? |
| 23:23.06 | nighthawk08 | in the distro3of Angstrom |
| 23:23.18 | nighthawk08 | where can I check? |
| 23:23.39 | tmzt | dcordes: do you know if there is a bootkit for hima? |
| 23:24.29 | tmzt | where are the bootkits anyway? |
| 23:25.31 | tmzt | htcpxa? |
| 23:25.43 | nighthawk08 | imate ppc |
| 23:26.00 | nighthawk08 | I do not know whether it is the htcpxa.... |
| 23:28.13 | tmzt | I am looking for the url for a newer zImage for HIMA if anyone knows it |
| 23:30.37 | tmzt | you might try the logs at http://irclog.iclem.net a few days back, there was talk about HIMA |
| 23:31.46 | tmzt | nighthawk08: you might need a newer zImage or to set MTYPE in default.txt |
| 23:32.05 | tmzt | any current zImage should work with angstrom |
| 23:32.22 | nighthawk08 | ok,. I'll check that, and get back later -- if I won't succeed in any way.... thanks tmzt |
| 23:33.04 | pigeon | Kevin2: anything particular in the new haret i should test? |
| 23:33.27 | tmzt | I found ~htcpxa but htchimalaya is empty |
| 23:34.45 | tmzt | dcordes: did the dd work? |
| 23:35.14 | tmzt | I don't know how to limit the size of transfers with cp thats why I suggested dd |
| 23:35.27 | dcordes | 01:17 < dcordes> tmzt: done. |
| 23:35.32 | dcordes | yea |
| 23:35.42 | dcordes | no clue where himalaya images can be found |
| 23:35.47 | tmzt | try running it, chmod +x first and do the chvt trick |
| 23:36.00 | tmzt | the rest are in ~htcpxa now though? |
| 23:36.12 | tmzt | except, whatever it was |
| 23:36.20 | tmzt | had a bootkit |
| 23:36.20 | dcordes | I only know about ~htcpxa and ~lgorris |
| 23:36.23 | dcordes | and headfuzz |
| 23:36.37 | dcordes | maybe angstrom builds hold up |
| 23:37.14 | dcordes | http://www.angstrom-distribution.org/unstable/autobuild/ no joy |
| 23:37.55 | tmzt | only uni, ba, and magician appear to be there |
| 23:38.32 | tmzt | he was using -hh20 |
| 23:38.39 | tmzt | might have been mtype though |
| 23:39.12 | dcordes | uhm, I could build one? |
| 23:39.27 | dcordes | I'd bet the hh kernel on OE is broken though |
| 23:39.40 | tmzt | chmod +x /media/ram/Xfbdev; (TSLIB_TSDEVICE=/dev/input/event0 /media/ram/Xfbdev) & (sleep 30; chvt 1) |
| 23:39.45 | tmzt | outside of chroot |
| 23:39.45 | marajin | WHOA.. damn.... Disposable lighters are getting a bit cheery nowadays |
| 23:40.03 | ginge | lol |
| 23:40.05 | marajin | ginge: Since when did a 50p lighter from a newsagents have a 4 inch flame? |
| 23:40.23 | ginge | marajin: mine went up in a ball of flames today |
| 23:40.51 | marajin | ginge: Heh :) You don't know the fun I had with one of them butane torches. |
| 23:40.56 | tmzt | we don't need that zImage, I was referring to nighthawk08's question |
| 23:41.01 | marajin | ginge: Dual 2 inch torch flame :p |
| 23:41.04 | marajin | ginge: Yes /dual/ |
| 23:41.15 | marajin | ginge: Frightened the fuck outta anybody dumb enough to cadge a light off me |
| 23:41.19 | ginge | lol |
| 23:41.46 | marajin | "Hey mate, got a light?" "Sure... **chink** **click** **BCKKKKKKKKKKKKKKKKKKKKKKKKKKKKT**" |
| 23:42.06 | ginge | ... no hair |
| 23:42.30 | marajin | it will melt plastic at 6+ inches from the flame yeah :P |
| 23:43.19 | marajin | I made the mistake of trying to use it to strip a waxed wire of its coating once.. |
| 23:43.27 | marajin | POOF, no coating... no wire either though |
| 23:43.58 | ginge | heh. been there with a windproof and solder |
| 23:44.30 | marajin | Most lighters'll mess up solder man |
| 23:44.59 | ginge | it only needed to hold until I got home... it was to the computer core of my car |
| 23:45.09 | marajin | heh |
| 23:45.11 | marajin | fair enough |
| 23:45.25 | marajin | but yeah, those dual flame butane torches are fnu |
| 23:45.51 | marajin | turn an aluminium can to dust in seconds |
| 23:45.55 | ginge | http://www.headfuzz.co.uk/?q=vw_ccm checkt he pics for the hack |
| 23:46.55 | ginge | damnned vw crap |
| 23:49.35 | marajin | heh |
| 23:49.39 | marajin | it a diesel or petrol? |
| 23:50.33 | ginge | petrol |
| 23:50.48 | tmzt | dcordes: did X start? |
| 23:51.30 | ginge | marajin: drinks fuel like a beast |
| 23:51.45 | marajin | 100k on the clock IS a little tired then |
| 23:51.45 | ginge | marajin: I wonder if I could port Linux to it... |
| 23:52.10 | marajin | I know a guy witha diesel one though, similar mileage, you'd barely know it wasn't new from the showroom |
| 23:52.13 | ginge | marajin: got it for a bargain :) |
| 23:52.14 | marajin | 'cept the turbo died |
| 23:52.29 | ginge | marajin: passat 100k is ok, just worn in |
| 23:52.39 | ginge | marajin: yeah my tubo is onthe way out |
| 23:52.40 | cr2 | pigeon: There is a labor assistant of the writer, to also the device of the type turn which is similar âiPaq 3870â (HTC 30H800125) had entered |
| 23:52.43 | marajin | and maaaan you get some trundleage on a diesel passage with no turbo |
| 23:52.54 | ginge | marajin: randyly comes on and tries to propell you over the car infront |
| 23:53.08 | ginge | marajin: randomly |
| 23:53.09 | marajin | ginge: Hey just claim you're working on a BTTF style hover car ;) |
| 23:53.10 | dcordes | tmzt: -sh: /media/ram/Xfbdev, not found |
| 23:53.16 | cr2 | pigeon: http://pc.watch.impress.co.jp/docs/2005/1213/pda45.htm |
| 23:53.25 | ginge | marajin: or recreating an A-Team episode |
| 23:53.36 | marajin | ginge: Heh, true, /any/ A-team episode |
| 23:53.52 | pigeon | cr2: i thought you're sleeping! :) |
| 23:53.59 | tmzt | did you do the dd? |
| 23:54.30 | dcordes | tmzt: yes file is present, also chmod +x'D it |
| 23:54.38 | tmzt | dd if=/jlime/usr/bin/Xfbdev of=/media/ram/Xfbdev bs=512; chmod +x /media/ram/Xfbdev |
| 23:54.43 | tmzt | oh, ldd it |
| 23:54.58 | tmzt | you will need the chroot, can't mix oabi and eabi libs I think |
| 23:55.13 | tmzt | but if it works, we can build it from source |
| 23:55.20 | tmzt | (as eabi) |
| 23:55.26 | dcordes | can I just mv the klime Xfbdev and link my ram Xfbdev in place? |
| 23:55.27 | pigeon | cr2: i do have a 3870 |
| 23:55.50 | cr2 | pigeon: it has the same asic as wallaby ? |
| 23:55.55 | pigeon | cr2: and what about the rx3115? |
| 23:55.57 | tmzt | no, you would have to set LD_LIBRARY_PATH to the path of the oabi libs in /jlime/lib /jlime/usr/lib |
| 23:56.35 | tmzt | ldd /jlime/usr/bin/Xfbdev ?? |
| 23:56.40 | dcordes | no ldd. |
| 23:56.43 | pigeon | cr2: i don't know, the thing with the 3870 is linux is well supported in very early times, i didn't have to know any of those about it :) |
| 23:56.49 | cr2 | pigeon: no idea. |
| 23:57.27 | tmzt | it needs a lot of libs (I ldd'd it) |
| 23:57.29 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 23:58.08 | tmzt | so we have to get mmc working first, or build eabi Xfbdev from jlime source |
| 23:58.45 | cr2 | pigeon: hmm. i really think that asic1 is there, so we can reuse the existing codebase. you only need to trace the gpios (egpios) |
| 23:58.46 | smev | on wm6 is there anyway to disable that damn low battery vibrate ? |
| 23:58.51 | tmzt | mount -o bind /media/ram /jlime/tmp |
| 23:59.14 | tmzt | then do the chroot and run it from /tmp/Xfbdev |
| 23:59.32 | tmzt | what do yo get when you run free |
| 23:59.41 | cr2 | pigeon: can you dump the gpio area ? |
| 23:59.44 | pigeon | hmm trying the latest haret, wirq seems to do things, but it doesn't stop, and then it seems to hang it after a while. |