00:00.15 | dcordes | ginge: is "void vibrate" the thing that triggers the vibrate in vogue-ts.c? |
00:01.08 | ginge | yes. dont forget to include proc_comm.h too |
00:02.54 | ginge | Replace 0x4 with 0x17 and 0x5 with 0x18 in vibrate(). To use: vibrate(1) to turn vibrate on. vibrate(0) to turn off. |
00:10.23 | dcordes | ginge: So I'd just copy premy's updated kaiser-smd.c to enable pc on kaiser, copy the void vibrate and the following vibrate function thing to board-kaiser-keypad.c and then, where to put vibrate(1) and vibrate(0) |
00:10.34 | dcordes | in order to make it vibrate on each key press? |
00:10.48 | ginge | erm, lets see... |
00:11.23 | ginge | any particular key, or all/ |
00:11.37 | dcordes | nevermind, just for testing. let's say all |
00:11.53 | *** join/#htc-linux RockTheHouse (n=Rock_The@78.146.196.213) |
00:12.06 | RockTheHouse | Hi folks |
00:12.12 | marajin | 'lo |
00:13.15 | RockTheHouse | I've been reading a lot about putting Linux on the BA (something I've managed to do with Angstrom/GPE via an SD Card). But there's talk of a lot of other Linux distros for the BA. I wondered which was the best? |
00:13.51 | dcordes | RockTheHouse: qtopia I'd say |
00:13.53 | ginge | in kaiser_kbd_queuework line 383 somewhere in that ifblock |
00:14.07 | marajin | I tend to agree, I much prefer the qtopia interface |
00:14.20 | dcordes | http://linuxtogo.org/~htcpxa/htcblueangel/Qtopia/ |
00:14.26 | RockTheHouse | I heard that was getting quite old now |
00:14.40 | ginge | dcordes: or put it in kaiser_update_leds |
00:15.33 | marajin | I still like it better than GPE |
00:15.46 | RockTheHouse | Isn't GPE the only one that has the working phone though? |
00:15.59 | dcordes | RockTheHouse: the one in the link is quite stable |
00:16.13 | RockTheHouse | Yeah thanks I'll give it a try |
00:16.20 | dcordes | you should try it if you are aiming to use ba with linux as a real phone |
00:16.28 | marajin | On the BA, I have no idea |
00:16.46 | RockTheHouse | It would be nice. I have managed to make calls with the BA on GPE |
00:17.10 | dcordes | wie gommunicator eh? |
00:17.18 | RockTheHouse | Yes |
00:17.24 | dcordes | I could try and make openmoko work |
00:17.34 | dcordes | just need to put the setty.c |
00:17.43 | dcordes | htc-setty or what that is |
00:18.19 | RockTheHouse | So Angstrom is still the best actual distro to use for the BA to date? |
00:18.48 | marajin | bah, I hate it when these things don't light evenly |
00:19.20 | dcordes | use spit |
00:19.58 | dcordes | RockTheHouse: I like angstrom much more than qtopia but I find qtopia more usable to make calls on ba |
00:20.04 | marajin | eh? |
00:21.37 | marajin | dcordes: "spit" ? |
00:23.00 | RockTheHouse | I will definitely give it a go. But what of Xanadux? |
00:24.17 | dcordes | ginge: http://rafb.net/p/f6csWA15.html ? |
00:25.34 | ginge | thats gonna vibrate forever and never stop |
00:25.40 | ginge | I guess thats a good test tho |
00:26.04 | dcordes | what about unsigned k = 0xb22; doesn't that need changing? |
00:27.05 | ginge | lol |
00:27.05 | ginge | typedef cheque int |
00:27.05 | ginge | yeah, unsigned what? unsigned cheque j; |
00:27.09 | ginge | dont do that |
00:27.15 | ginge | unsigned int |
00:27.51 | dcordes | I have no clue what you are talking about |
00:28.07 | ginge | did you change thevibrate function? |
00:28.49 | dcordes | I just copied it from vogue-ts.c changing the proc_comm addresses to 17 and 18 |
00:29.07 | dcordes | see my paste l 226 following |
00:29.12 | ginge | should be fine then |
00:29.17 | dcordes | 266 |
00:29.18 | ginge | I only glanced over it |
00:29.38 | RockTheHouse | dcordes: Is Qtopia-Linux-htcblueangel.exe a HaRET-esque bootloader or something? |
00:35.33 | dcordes | RockTheHouse: iirc, put the exe and the tar.bz2 into the same folder somewhere on your b |
00:35.53 | dcordes | then run the exe and wait patiently until it creates some loopfile out of the rootfs |
00:37.05 | dcordes | alternatively make and SD card with two partitions, one fat where you put haret.exe, zImage and startup.txt from fodler sdinstall, one ext2 where you extract the .tar.bz2 |
00:38.52 | RockTheHouse | I already have a fat and reiserfs partition (ext3 lol) |
00:39.41 | dcordes | is journaling so good on sd? |
00:40.27 | RockTheHouse | The reason I went for reiserfs was because of dancing trees and the fact I can grow and shrink the fileing system while it's live |
00:40.38 | RockTheHouse | Which has been very useful for titting around with |
00:41.07 | dcordes | interesting |
00:41.23 | dcordes | I wasn't aware you could just resize kaiser while in use |
00:41.30 | RockTheHouse | Yes lol |
00:41.32 | dcordes | s/kaiser/reiserfs/ |
00:43.27 | RockTheHouse | Yes you can, it's quite an old feature now |
00:44.04 | RockTheHouse | I've had lots of success doing away with traditional partitions and simply having multiple reiserfs filing systems on disk which I can move and resize as I want |
00:44.08 | RockTheHouse | A poor man's (complex) LVM |
00:44.58 | RockTheHouse | Although of course if you combine it with proper LVM like I have with my media server, I can simply grow a single actual live filing system in real time as I increase storage capacity |
00:45.06 | dcordes | I use reiser on both my laptop and desktop computer |
00:45.16 | RockTheHouse | If I need to take physical disks out I can write the blocks to other disks, shrink the live partition and retire the disk |
00:46.14 | RockTheHouse | It's definitely my favourite fs and I'm sure you've never had any problems with it |
00:47.16 | dcordes | ginge: board-kaiser-keypad.c:271: error: implicit declaration of function âmsm_proc_commâ |
00:48.18 | ginge | did you include proc_comm.h? |
00:48.23 | dcordes | oh |
00:48.29 | marajin | I ned to sort out my server =/ but I don't have 1000 in cash to buy new drives |
00:48.32 | marajin | heh |
00:49.23 | ginge | I gave up with local storage. I pay amazon s3 $9 a month to host my 40 gigs |
00:49.40 | RockTheHouse | marajin: If you have older hardware around you can reuse it in an LVM approach to combine, virtualuse and utilise what would otherwise be unpracitcally small disks for a server |
00:49.43 | marajin | ginge: My fileserver has 1.2 GB of TB free |
00:49.51 | marajin | s/TB/2TB |
00:50.22 | ginge | right. i have 1TB in my server, but the "archive" stuff gets pushed to amazon. Backed up and available |
00:50.42 | marajin | I wanna do a 6TB LVM but alas I'dneed to win the lottery first |
00:50.52 | ginge | not that I am paranoid after all harddisks exploded in my server twice in a year ;) |
00:51.04 | RockTheHouse | marajin: What a shmuck :p You don't need the lottery, you need compound interest :D |
00:51.07 | marajin | only uses Seagate NS drives. |
00:51.08 | dcordes | ginge: ok the first error disappeared now only 298: error: expected expression before âelseâ and smae for line 304 left |
00:51.35 | marajin | RockTheHouse: Considering I don't play the lottery, I'm not likely to win it, but right now I'm paying interest on debt, not gaining it on savings so.. |
00:51.56 | RockTheHouse | marajin: Usury ftl |
00:52.09 | marajin | RockTheHouse: 'usury' ? |
00:52.13 | RockTheHouse | :-o |
00:52.30 | RockTheHouse | http://www.thefreedictionary.com/usury |
00:52.52 | marajin | he |
00:52.53 | marajin | h |
00:53.07 | marajin | THe interest rate isn't bad but I have a long term illness and ergo no proper job |
00:53.22 | marajin | so I don't really get enough cash to repay it sanely |
00:53.37 | marajin | the sad thing is, if I had a full time job I'd clear my debt in a month or two |
00:53.39 | ginge | dcordes: what is before the else? |
00:53.42 | RockTheHouse | Maybe it's the fact I'm very tired but I swear I just read "long term illness and negro problems" |
00:53.43 | RockTheHouse | :-| |
00:53.45 | dcordes | the vibrate(1) |
00:53.55 | marajin | Heh |
00:54.00 | marajin | No, no negro problems |
00:54.00 | dcordes | if (alternate_key) |
00:54.00 | dcordes | gpio_set_value(EGPIO_3_2_FN_LED,1); |
00:54.00 | dcordes | <PROTECTED> |
00:54.00 | dcordes | else |
00:54.12 | marajin | the Pakis next door are an irritation but nothing more |
00:54.14 | marajin | grins. |
00:54.22 | RockTheHouse | lol |
00:54.30 | ginge | youu need braces |
00:55.14 | marajin | but yeah, I only owe about 1400-1500 quid right now but my income is sweet FA really |
00:55.23 | marajin | ~200 a month I get realistically |
00:55.24 | dcordes | ginge: where? |
00:55.32 | ginge | if ... { ... } else |
00:55.38 | marajin | out of which I have to pay for regular bills, etc. |
00:56.14 | ginge | <PROTECTED> |
00:56.15 | ginge | { |
00:56.15 | ginge | <PROTECTED> |
00:56.15 | ginge | <PROTECTED> |
00:56.15 | ginge | } |
00:56.15 | ginge | else |
00:56.41 | marajin | In the end I can only put a maximum of er.. 100/month towards repaying debt so.. 1500 quid takes a while |
00:56.41 | RockTheHouse | marajin: Well if you're in teh UK and you're currently receiving disability benefit then you can write to the creditors and stop repayment or reduce it to a figure you can afford (allthe way to zero with benefits) since that it what the government deems is the minimum you need to live on so therefore no creditors should be able to take that off you, and they cannot take you to the county court to get a CCJ against you - all judges would stay any |
00:57.12 | dcordes | ginge: is that always required if there's more than one line in the if else sort of situation? |
00:57.24 | marajin | RockTheHouse: Meh I don't want to stiff the companies of what I owe, I am pissed at the government for paying me so little |
00:57.28 | ginge | if you dont see braces on an if statement, only the next line of code will run in that if |
00:57.47 | ginge | dcordes: yeah what you said :) |
00:58.00 | marajin | RockTheHouse: They used to pay me a bit more.. about 300/month but they mysteriously lost my records (they just claim that I never gave any medical evidence and 18 months of payments were just accidental.. yeah right) |
00:58.02 | RockTheHouse | marajin: Well I doubt your illness is your fault so why should you be penalised for it? The creditors won't go hungry if you don't pay them - trust me :-| |
00:58.11 | marajin | RockTheHouse: But since it's now a "new claim" I get less again |
00:58.25 | marajin | RockTheHouse: And they won't give me DLA so I just get IB/IS which is FA |
00:58.44 | RockTheHouse | marajin: I'm sorry to hear that but you should check out moneysavingexpert.com for help on dealing with this - there's a lot of clever people there who can get you out of all sorts of problems |
00:59.15 | marajin | RockTheHouse: Ah yes, I know they won't go hungry but I can live like this, just about. It wil just take a long while to repay my CC/OD balance |
00:59.15 | RockTheHouse | marajin: I used to volunteer at CAB for debt help advice you know, if you wanted to, there are a great many loopholes you can use to get out of paying it completely |
00:59.25 | ginge | RockTheHouse: +5 Insightful |
00:59.32 | dcordes | ginge: it builds |
00:59.51 | ginge | moneysavingexpert helped me level my credit card and get some cash back off the bank through the courts :) |
00:59.54 | marajin | RockTheHouse: yeah I'm just too honest to stiff them even though it's buggerall to them and it makes my li harder |
01:00.16 | RockTheHouse | marajin: And after the Statute of Limitations Act kicks in after 6 years and you don't acknowlege any consumer debt then they can no longer chase you for it. On top of that, if it is consumer debt then you have a great many powers under the Consumer Credit Act 1974 which you can use to stop them hassling you |
01:00.54 | RockTheHouse | marajin: Creditors biggest collection tools are ignorance and fear. Seriously check out your rights |
01:00.56 | ginge | dcordes: fingers crossed its the same address |
01:01.02 | dcordes | ginge: if we had usb host, you could make kaiser control your copter lol and you could control it through gprs |
01:01.26 | RockTheHouse | Well good luck to both of you on that :) |
01:01.34 | ginge | dcordes: yeah. I have gumstix on there at the moment. At least until my custom board arrives :) |
01:01.41 | marajin | RockTheHouse: I neither fear them nor am ignorant of how I could not pay, I simply just don't WANT to forfeit the debt. I spent the money so I'll repay it.. eventually |
01:01.51 | marajin | RockTheHouse: But your advice is appreciated nevertheless |
01:02.05 | RockTheHouse | marajin: You're welcome lol I've helped write off over £500k for other people :) |
01:02.39 | marajin | RockTheHouse: Yeah but it's only like... lemme check.. |
01:02.43 | *** join/#htc-linux Othello__ (i=Magorium@gateway/tor/x-9a8f03963ab52d75) |
01:02.44 | dcordes | ginge: still waiting for the pcb? |
01:03.06 | marajin | RockTheHouse: 1450 quid currently |
01:03.07 | ginge | The first design was fubar. I got some traces on the omap wrong |
01:03.41 | dcordes | what omapbrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr |
01:03.45 | RockTheHouse | marajin: If you're just being stiffed by interest payments you know you can repay your debt with an interest freeze. Also in your situation you may find the "cumulative accellerator margins" approach will speed your repaymente even with the same amounts. Definitely do check it out, no point paying more than you have to |
01:03.48 | dcordes | brrr |
01:03.56 | ginge | lol |
01:04.02 | dcordes | rrrrrrrrrrrr |
01:04.07 | ginge | awesome :) |
01:04.23 | dcordes | rrr<softreset> |
01:04.30 | marajin | RockTheHouse: Interest on 1450 quid isn't that high. well I guess 14% PA but by the end of the year most of it would be repaid anyway so.. |
01:04.31 | dcordes | on fn press |
01:04.43 | ginge | anything that triggers an led |
01:04.49 | dcordes | yep |
01:05.35 | marajin | RockTheHouse: Really my best alternative is to find work |
01:05.44 | RockTheHouse | marajin: Pfft |
01:05.59 | dcordes | ginge: how much weight will your copter be able to carry? |
01:06.25 | marajin | RockTheHouse: I have no objection to working but I get mentally/physically fatigued too easily so there's almost nobody who'd want to hire me |
01:06.37 | dcordes | ginge: wouldn't it be great to gprs control it? |
01:06.40 | RockTheHouse | marajin: Then work for yourself |
01:07.03 | RockTheHouse | marajin: If you have CFS then self employment is definitely the way to go. Do what you want when you want ftw |
01:07.18 | marajin | RockTheHouse: I've thought about it but I've never been able to think up anything inside my limits |
01:07.31 | RockTheHouse | marajin: Then change your limits? |
01:07.46 | marajin | RockTheHouse: That would require not being ill ? :P |
01:08.00 | RockTheHouse | marajin: You don't have to work harder or faster, just smarter |
01:08.20 | ginge | dcordes: omap... omap3530. its got a gprs radio module onboard. |
01:08.29 | marajin | RockTheHouse: With microscopic creatures eating my brains I'm not sure if I'm all that smart anymore ;) |
01:08.44 | dcordes | ginge: oh :) you're hardcore |
01:08.53 | RockTheHouse | marajin: :-| |
01:09.12 | RockTheHouse | marajin: Still using Linux. I'm sure there's hope |
01:09.15 | marajin | RockTheHouse: It sounds worse than it is really but it really does affect my mental status at times |
01:09.48 | ginge | marajin: are we talking about linux here ;) |
01:10.01 | marajin | ginge: Hah, maybe sometimes ;) |
01:10.06 | RockTheHouse | :-o |
01:10.15 | dcordes | marajin: what sort of creatures do you have? |
01:10.25 | RockTheHouse | marajin: Please don't say mudkips |
01:10.38 | marajin | dcordes: Microfilarial worms called Crypto Strongylus Pulmoni |
01:10.54 | marajin | Yes I know pulmoni -> Lungs but the move up the artery into the brain too |
01:11.31 | RockTheHouse | marajin: Interesting I've not heard of that before |
01:11.42 | marajin | http://marajin.w-ww.org/wurm.jpg <-- A very bad shot using my hermes down a microscope.. but that was a slide of my blood |
01:12.03 | marajin | RockTheHouse: It's the true creature behind some (maybe majority, they're still researching) CFS cases |
01:12.25 | dcordes | marajin: no way to kill em? |
01:12.33 | marajin | RockTheHouse: It smacks the lungs, digestive system and brain about, causing opportunistic additional infections |
01:12.40 | RockTheHouse | Have you heard of the miracle of HH2O? |
01:12.46 | RockTheHouse | Hydrogen peroxide |
01:12.57 | RockTheHouse | It's a wonder substance when ingested (!) |
01:13.16 | marajin | dcordes: Ivermectin will kill the adult form but the worm has a really complex lifespan that involves at least 4 stages, so anythig in the other 3 stages is immune to the treatment. |
01:13.31 | marajin | dcordes: So you can dose the adults and a month later, more've hatched |
01:13.46 | RockTheHouse | It was trialled by the Nazi scientists and found to cure a great many things including cancer, but has been mainly supressed since the war although there is a very large body of people now who have research showing it cures a great many things |
01:13.53 | RockTheHouse | Including viruses, bacerteria, parasites and even cancer |
01:14.09 | marajin | mate, if it kills cancer, it kills people too :P |
01:14.31 | RockTheHouse | No, it makes the blood too oxygen rich for cancer to survive so it dies. The body actually like oxygen ftw |
01:14.48 | RockTheHouse | It affects many things in the body this way, definitely check it out |
01:15.01 | marajin | er, isn't cancer usually caused by mutation in the body that causes rampant unchecked growth of cells? |
01:15.17 | dcordes | I'd take any hint. The slackware person also cured himself |
01:15.20 | RockTheHouse | Yes but cancer doesn't like oxygen rich environments |
01:15.28 | RockTheHouse | Many things don't |
01:15.38 | marajin | then why isn't hyperbaric chamber a treatment for cancer |
01:15.57 | RockTheHouse | HH2O is like water with an extra oxygen atom that is released in to the blood stream when ingested |
01:16.01 | dcordes | I need some sleep. night or whatever all |
01:16.04 | RockTheHouse | Stops wheezing and breathlessness instantly |
01:16.53 | marajin | My lungs are kinda ok nowadays actually. I took some azithromycin (for pneumonia) and the ivermectin I take periodically. So I generaly don't fight for oxygen and whatnot. |
01:17.04 | marajin | Problem is I do fight for blood sugar amongst other things |
01:17.34 | dcordes | ginge: tell premy about pc on polaris= pc on kaiser when you cross his way |
01:17.49 | ginge | will do |
01:17.54 | ginge | good night |
01:17.59 | RockTheHouse | Laters |
01:18.47 | marajin | Otherwise I consult with a great guy who's researching the stuff... so yeah we're working on it |
01:19.00 | marajin | but still, the symptoms are kinda too much to deal with at times |
01:19.59 | RockTheHouse | H2O2 btw not HH2O (am tired) :$ |
01:21.43 | marajin | yeah I already looked it up |
01:22.03 | RockTheHouse | I am a big believer in alternative medicines, I haven't take any drugs for several years, not evne pain killers, and use only natural remedies |
01:22.20 | marajin | I never take pain killers |
01:36.20 | *** join/#htc-linux surgex (i=surge@pool-71-186-163-107.bflony.fios.verizon.net) |
01:45.41 | *** join/#htc-linux Othello (i=Magorium@gateway/tor/x-10e6e7e9eed3c50d) |
01:59.23 | *** part/#htc-linux RockTheHouse (n=Rock_The@78.146.196.213) |
03:00.23 | *** join/#htc-linux xeroOTG (n=xero@75-173-64-182.albq.qwest.net) |
03:00.39 | xeroOTG | ok... i need some help here... when i run haret, i get an error message, Unknown Keyword: `' |
03:04.41 | *** join/#htc-linux ewon (n=ewan@skynet.skynet.ie) |
03:04.46 | ewon | \o/ |
03:04.53 | ewon | I just discovered this exists |
03:11.32 | xeroOTG | descovered what exists? |
03:13.50 | ewon | linux for the Kaiser |
03:14.16 | xeroOTG | oh... |
03:16.47 | ewon | is there currently any way to execute native code, or does one have to use android java vm? |
03:21.58 | xeroOTG | i think theres an sdk avalable at the google code page |
03:24.57 | ewon | but that's still using the android vm, right? |
03:25.19 | ewon | I'm more thinking of a console app written in C I can feed to a cross compiling gcc |
03:28.08 | xeroOTG | i have no clue mate... |
03:28.24 | xeroOTG | i just make stuff work, i dont bring it to life... |
03:50.48 | *** join/#htc-linux Thanatonaute (n=Than992@c-24-1-251-218.hsd1.il.comcast.net) |
03:57.39 | marajin | ewon: You can just run console linux if you like |
03:58.03 | marajin | ewon: Display works and there's a keyboard driver |
04:00.48 | marajin | ewon: Ping? |
04:20.25 | *** join/#htc-linux goxboxlive (n=goxboxli@195.159.97.196) |
04:47.23 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
05:24.42 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
05:33.07 | *** join/#htc-linux kiozen (n=oeichler@p54929FA2.dip0.t-ipconnect.de) |
05:42.52 | *** part/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
05:50.14 | *** join/#htc-linux kiozen (n=oeichler@p54929FA2.dip0.t-ipconnect.de) |
06:42.38 | *** join/#htc-linux rmoravcik (n=rmoravci@gtsgw.ttc.cz) |
06:48.40 | kiozen | cr2: please try http://www.qlandkarte.org/rootfs.tar.bz2 |
07:08.54 | *** join/#htc-linux gundam (n=gundam@slackware.it/staff/gundam) |
07:37.52 | *** join/#htc-linux hagisbasheruk (n=hagisbas@78.148.131.1) |
07:46.39 | marajin | alas poor hagises |
07:49.43 | hagisbasheruk | lol |
07:50.05 | hagisbasheruk | good morning marajin |
07:51.01 | hagisbasheruk | was reading #beagle logs @ http://www.beagleboard.org/irclogs/ ,forgive me for slow reply |
07:51.27 | marajin | No worries |
07:51.37 | marajin | and morning it is, dunno about good though ;) |
07:52.16 | hagisbasheruk | what you up to yourself ,yes it is quite wet here in Clydebank |
07:53.02 | *** join/#htc-linux ginge_ (n=baz@80.229.188.31) |
07:57.07 | *** join/#htc-linux rob_w (n=rob@p549B904F.dip0.t-ipconnect.de) |
08:32.58 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
08:44.21 | *** join/#htc-linux rob_w (n=rob@p549B904F.dip0.t-ipconnect.de) |
08:46.07 | marajin | ginge_: Morning :) |
08:46.17 | ginge_ | norning |
08:46.33 | ginge_ | as you can see I can't type yet |
08:52.56 | ginge_ | d'oh. I forgot to logout at home. I wonder is I can use DBUS to knock me offline. I suppose there is always... |
08:53.12 | ginge_ | heh |
08:56.36 | marajin | heh |
08:56.47 | marajin | that's why I have NX/RDP on all my boxes :P |
08:57.03 | ginge_ | meh. I just use ssh for everything :) |
08:57.07 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
09:00.40 | marajin | Oh don't get me wrong, Nearly everything is in a screen session for me, but I got annoyed with finch/centericq/et al so I basically gave up just so I could use a less shite IM client |
09:01.29 | ginge_ | kopete (once you kill the nasty defaults) is just about bearable |
09:04.48 | marajin | 's what I use now yep |
09:05.02 | marajin | People seem to have forgot the point of an IM client |
09:05.40 | ginge_ | in what way? |
09:05.55 | marajin | it's for sending realtime text communication peer to peer.. so wth do we need animated emoticons, display pictures, backgrounds, rich text/html rendering, VoIP, videostreaming, etc, etc, etc built into the sameclient/protocol in a very very hacky wa |
09:06.24 | ginge_ | gotcha. My Kopete is class white on black. Just like a proper terminal. |
09:06.34 | ginge_ | s/class/classic/ |
09:06.57 | ginge_ | I suppose you could count is as minimal. Everything funky is broken |
09:09.21 | marajin | aye :P |
09:52.25 | ginge_ | where would I find the universal repos? |
09:58.35 | *** join/#htc-linux zoolooc (n=lucian@nrbg-4dbfab9d.pool.einsundeins.de) |
10:10.31 | ginge_ | someone needs to beat me with a clue stick here. Anyone? |
10:12.11 | *** join/#htc-linux dcordes (n=dcordes_@unaffiliated/dcordes) |
10:13.10 | ginge_ | dcordes: morning. |
10:13.21 | zoolooc | ginge_: which reminds me of the joke they broadcasted yesterday all day long on the radio |
10:13.43 | zoolooc | morning dcordes, ginge_ ;) |
10:13.57 | ginge_ | zoolooc: oh? and morning |
10:14.02 | dcordes | morning/whatever all |
10:14.52 | zoolooc | ginge_: it went like: "do you remember what the number one sentence our teacher had to hear in school? - no idea! - yea right, that was it!" |
10:14.55 | dcordes | ginge_, are you looking for the htc universal kernel? |
10:15.09 | ginge_ | dcordes: yes |
10:15.16 | dcordes | handhelds.org cvs |
10:15.23 | ginge_ | zoolooc: heh. |
10:15.28 | ginge_ | dcordes: link? |
10:15.41 | dcordes | http://handhelds.org/sources.html |
10:15.58 | ginge_ | thanks. I have brain slugs today ;? |
10:16.11 | dcordes | http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-pxa/htcuniversal/ |
10:16.39 | dcordes | the futurama ones? |
10:16.47 | ginge_ | yeah |
10:17.50 | dcordes | no offence they came to my mind yesterday |
10:18.23 | dcordes | when marajin said something eats up his brains |
10:19.15 | ginge_ | ah, yeah. No offense marajin |
10:20.21 | dcordes | ginge_, http://handhelds.org/moin/moin.cgi/UniversalKernel |
10:21.32 | *** join/#htc-linux rmoravcik (n=rmoravci@gtsgw.ttc.cz) |
10:21.37 | ginge_ | I want to look at the touchscreen code. I have it almost calibrated perfectly, but I am starting to wonder if I am missing something obvious, like I am sampling noise. I was just checking the other implementations, and I believe the universal touchscreen has a similar mechanism |
10:22.11 | dcordes | you could also look at blueangel. that's qvga |
10:32.27 | *** join/#htc-linux Marex (n=vasum7am@195.113.26.203) |
11:16.21 | *** join/#htc-linux exception13 (n=exceptio@testdrive.kgts.ru) |
11:16.25 | exception13 | hi all |
11:21.30 | dcordes | ginge_, is the hh code of any help? |
11:22.14 | ginge_ | no, it doesnt compensate for the y changes based on the x value |
11:23.09 | ginge_ | the problem being that the touch screen reports xy as a distorted parallelogram |
11:27.48 | dcordes | ginge_, maybe looking at what the wince calibration program does would help? |
11:28.16 | ginge_ | I doubt we could trace it to that level. |
11:29.05 | ginge_ | If I can get it looking somethng like a rectangle again (I will check it out in a while when on lunch) we can then use a userspace application to calibrate |
11:30.22 | dcordes | ginge_, to pick that point up again: in the angstrom paint app thingy, it actually is rectangle like for me |
11:30.30 | dcordes | it's just that like 1cm at the bootom are cut off |
11:30.42 | dcordes | s/angstrom/android/ |
11:30.50 | dcordes | that calibration test app, you know what I mean |
11:30.56 | ginge_ | yeah |
11:31.07 | dcordes | I'll fire it up again to make sure |
11:31.19 | ginge_ | on mine it looked like a rhombus |
11:31.56 | dcordes | probably we really have different displays |
11:32.01 | dcordes | touchscreens |
11:32.04 | ginge_ | that would suck mightil |
11:32.18 | ginge_ | s/mightil/mightily/ |
11:32.36 | dcordes | because you have the shity one then? :) |
11:32.44 | ginge_ | lol. |
11:33.06 | dcordes | btw do you have any clue how premy found out the correct pc regs? |
11:33.18 | ginge_ | no, but I can guess |
11:33.35 | ginge_ | my guess == haret pdump and mmutrace |
11:34.03 | dcordes | ginge_, ok I'm in the android pointercal test app now |
11:34.30 | ginge_ | try iy with your finger too |
11:34.33 | dcordes | I just moved along the corners with my fingertip and it paints a nice square in the middle |
11:35.11 | ginge_ | maybe your fix is as simple as changing MAXY1 |
11:36.05 | ginge_ | if you feel brave... |
11:36.07 | ginge_ | int XMIN=575;int XMAX=3775;int YMIN=1600;int YMAX=2032;int YMIN1=2272;int YMAX1=3760;int PMIN=0;int PMAX=600; |
11:36.08 | dcordes | lines parallel to the y axis draw perfectly clean |
11:37.58 | dcordes | also x |
11:38.25 | dcordes | it's just an offset of a few mms on the y axis to the top |
11:38.46 | ginge_ | if is parallel along all the lines? |
11:38.55 | ginge_ | s/if is/is it/ |
11:38.57 | dcordes | that gets obvious when I move all the way to the top and Y values go negative |
11:38.59 | dcordes | up to -8 |
11:39.14 | ginge_ | ahh, yeah you have the same tuochscreen then |
11:39.31 | ginge_ | does y go go lower at x=0 |
11:39.47 | dcordes | no I can't reach x=0 |
11:39.52 | ginge_ | as close to |
11:40.01 | dcordes | it stops few mm in front of the left screen corner |
11:40.55 | ginge_ | if you move from x=0 to x=240 y looks like this:- -130 .............................. 0 |
11:41.37 | ginge_ | and y=320 x=0 to x=240 300 .............................. 320 |
11:41.43 | ginge_ | its not even uniform |
11:41.59 | *** join/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
11:42.04 | premy | Hi |
11:42.11 | ginge_ | when x=240 y tracks perfectly |
11:42.11 | ginge_ | when x=0 its fubar |
11:42.16 | ginge_ | premy: hi |
11:42.31 | dcordes | ginge_, now that you say it I notice a slight offset on the y axis when I move from x0 to x22? |
11:42.42 | dcordes | it's pushed up a bit to the right then at the end |
11:42.47 | dcordes | but that is in the corner only |
11:43.13 | dcordes | when I move inside the square that is drawn when I move along the corners everything is fit |
11:44.01 | dcordes | I'd shoot a vid but I have no good camera except of the one in kaiser |
11:44.15 | ginge_ | you think the one in kaiser is good? |
11:44.17 | ginge_ | lol |
11:44.22 | dcordes | well... |
11:44.24 | ginge_ | ;) |
11:44.32 | dcordes | it's better than the one in my d600 at least |
11:44.51 | dcordes | and vga webcam |
11:44.52 | ginge_ | not to worry. I know what you are talking about |
11:45.14 | premy | ginge_: that's right. mmutracing 0xb51FC100 0x40. That's probably what we're going to have to do to identify the other commands too. |
11:46.09 | dcordes | premy, did you try and feed polaris with power while tracing 0xb51FC100 0x40 ? |
11:46.31 | dcordes | or change the lcd backlight powerstate? |
11:47.37 | premy | dcordes: nope. I'm currently trying to backport dzo's new keyboard, to see if it's more usable. |
11:47.44 | cr2 | kiozen: this one boots. |
11:47.59 | kiozen | hehe :) root rules! |
11:48.04 | ginge_ | premy: should work fine if you have a pendown irq |
11:48.27 | kiozen | can you try to get M to startup and sen me the patches |
11:48.45 | cr2 | root@htcuniversal:~# QLandkarteM |
11:48.45 | cr2 | QLandkarteM: error while loading shared libraries: libts-1.0.so.0: cannot open shared object file: No such file or directory |
11:48.58 | cr2 | ok, i need to write the script |
11:49.09 | dcordes | hi cr2 kiozen |
11:49.17 | premy | ginge_: is there some work you've done I could reuse ? |
11:49.20 | cr2 | to export the qtopia variables, and link /dev/ttyS2 to /dev/gps |
11:49.26 | kiozen | hi dcordes |
11:49.36 | cr2 | kiozen: are they in /etc/profile ? |
11:49.39 | kiozen | cr2: yes and to start with inittab |
11:50.02 | cr2 | dcordes: if i'll make it work flawlessly today, you can try it on kaiser. |
11:50.07 | cr2 | kiozen: yes. |
11:50.09 | kiozen | cr2: nothing in profile yes |
11:50.20 | cr2 | kiozen: nothing ? |
11:50.34 | dcordes | cr2, qlandkarte? We need an itsy bitsy qtopia rootfs then |
11:50.37 | kiozen | cr2 no qtoipa stuff, just plain oe |
11:50.46 | cr2 | kiozen: you told me to copy them from /etc/profile before. |
11:51.00 | kiozen | cr2: yes, but they are not in there yet |
11:51.05 | dcordes | cr2, I tried shrinking BabelO's latest ba build, but didn't get it small enough to fit into kaiser ram at boottime |
11:51.39 | cr2 | kiozen: ok. what TS* and QT* variables should i export ? and the keyboard. |
11:51.53 | kiozen | moment.. |
11:51.54 | cr2 | dcordes: hm. yes, you don't have SD... |
11:52.15 | dcordes | cr2, but we have proc_comm working now. iirc a function in sd depended on it |
11:52.18 | dcordes | vreg something |
11:52.22 | ginge_ | premy: not really. The kaiser has no irq for pen down so its only half using dzo's code |
11:53.08 | kiozen | cr2: my old records tell me: http://pastebin.ca/1062182 |
11:53.16 | cr2 | dcordes: there was some trace file with more smd channels. are they DEX / m2a_6 related ? |
11:53.16 | kiozen | cr2: but handle with care ;) |
11:53.21 | cr2 | kiozen: ok. |
11:53.40 | cr2 | kiozen: i'll make it work, think about the battery now :) |
11:53.43 | cr2 | root@htcuniversal:/etc# cat /proc/apm |
11:53.43 | cr2 | 1.13 1.2 0x02 0xff 0xff 0xff -1% -1 ? |
11:54.28 | kiozen | cr2: no libts on the image, think I have to add it |
11:54.31 | dcordes | cr2, no idea. martin__ found them and added them up to the kaiser-smd.c. I could boot it and try and probe them for what they are |
11:55.06 | *** join/#htc-linux Othello (i=Magorium@gateway/tor/x-b700b9f5dec1c0d1) |
11:55.32 | ginge_ | premy: http://rafb.net/p/9sNdmv77.html it will make your eyes bleed |
11:56.49 | cr2 | dcordes: yes, it's something to think about. dex uses a2m_6 irq, so it should be related to channel6. |
11:57.01 | cr2 | kiozen: yes, i need the libts .ipk. |
11:58.09 | kiozen | cr2 does libts need xorg? |
11:59.05 | kiozen | cr2: and do we need tsclient? |
11:59.24 | cr2 | kiozen: they had weird gtk dependency before ;) don't know if it's resolved now. |
12:00.04 | cr2 | kiozen: may i take the libts ipk from angstrom feed ? |
12:00.19 | kiozen | oh my god! now they compile python and stuff !!! |
12:00.26 | cr2 | i.e. are you binary/version comaptible with the current angstrom ? |
12:00.48 | cr2 | kiozen: these are probably some calibration tools. |
12:01.13 | kiozen | cr2: don't know |
12:01.26 | cr2 | kiozen: the oe is pretty gtk-biased ($$$), so don't be surprised. bluez-utils depend on gtk too. |
12:02.53 | dcordes | I wonder what dzo's status with mmc is |
12:03.02 | dcordes | (on the vogue) |
12:03.33 | kiozen | cr2: ok now I got libts |
12:03.44 | kiozen | cr2: tsclient needs all that stuff |
12:04.23 | *** join/#htc-linux rob_w (n=rob@p549B904F.dip0.t-ipconnect.de) |
12:04.37 | kiozen | cr2: uploading a new tar ball |
12:10.42 | premy | cr2: DEX ? what does it mean ? |
12:11.01 | *** join/#htc-linux Othello__ (i=Magorium@gateway/tor/x-5fcf9185dc113a91) |
12:11.40 | kiozen | cr2: http://www.qlandkarte.org/rootfs.tar.bz2 with tslib |
12:15.47 | cr2 | kiozen: ok. |
12:16.06 | dcordes | msm_sdcc init on kaiser with proc_comm enabled in kaiser-smd.c: |
12:16.08 | dcordes | msmsdcc_init_dma: DM non-cached buffer at ffc01000, dma_addr 0x150e3000 |
12:16.08 | dcordes | msmsdcc_init_dma: DM cmd busaddr 353251328, cmdptr busaddr 353251352 |
12:16.08 | dcordes | msm_sdcc: probe of msm_sdcc.1 failed with error-1 |
12:16.48 | cr2 | premy: DataEXchanger, this is some shared memory IPC, similar to smem, but using the SD registers and BT/UART registers. |
12:17.11 | cr2 | premy: look at the kaisermemorymap wiki |
12:17.38 | premy | cr2: ok thanks. |
12:22.38 | dcordes | cr2, do you know how googles realize a sound device? |
12:22.59 | dcordes | or where to look for it |
12:30.19 | dcordes | boots with sdcc built in and sdcc dma debug enabled |
12:42.15 | dcordes | hm nothing extra with dma debug in dmesg. |
12:42.40 | dcordes | sdc used to print loads of info, including clock and stuff |
12:55.54 | cr2 | dcordes: i didn't look at the recent patches, but afair it was not implemented before. |
12:56.28 | dcordes | cr2, more detailed sdc debug? |
12:57.01 | dcordes | I'm sure I had it printing lots of dma messages and clock messages beginning with mmc: |
13:03.18 | dcordes | and I enabled all debug switches I found. the two in drivers/mmc/host/msm_sdcc.c and in kernel config (mmc trace) |
13:07.46 | premy | dcordes: my msm_proc_comm will not do anything more than vibrate on / off. If you want it to try other commands, you should remove lines 187 and 188 (if (cmd!=0x17)....) |
13:08.20 | premy | But my guess is that it will not work, because the command codes seem to be different than what's in the current code. |
13:13.56 | *** join/#htc-linux zwox (n=zwox@par69-7-82-243-234-94.fbx.proxad.net) |
13:16.40 | dcordes | premy, didn't you mention yesterday you already tried with extra commands and that caused oopses? |
13:17.32 | dcordes | will give that a go nevertheless |
13:21.33 | dcordes | premy, do those lines mean that proc comm does nothing but return -1 if other commands than vibrate 0/1 are triggered? |
13:21.48 | premy | dcordes: yes |
13:25.21 | dcordes | premy, looks like it doesn't boot with that function commented out |
13:25.30 | dcordes | did it also hang for you after fb takeover? |
13:28.02 | premy | dcordes_: I'm not sure, but it's possible, yes. Might be sdcc_init calling some clock related msm_proc command. |
13:29.04 | dcordes | I will try the same with sdcc off |
13:31.23 | dcordes | premy, same with mmc disabled in .config |
13:31.43 | dcordes | so it's not mmc specific proc comms |
13:33.30 | dcordes | g2g |
13:47.23 | *** part/#htc-linux exception13 (n=exceptio@testdrive.kgts.ru) |
14:02.55 | *** join/#htc-linux LunohoD_ (n=alex@e180071190.adsl.alicedsl.de) |
14:08.26 | *** join/#htc-linux AstainZZZZZZ (n=Administ@unaffiliated/astainhellbring) |
15:01.17 | *** join/#htc-linux Othello (i=Magorium@gateway/tor/x-bfd5bbbfed9d87ed) |
15:07.34 | *** part/#htc-linux zoolooc (n=lucian@nrbg-4dbfab9d.pool.einsundeins.de) |
15:13.08 | *** join/#htc-linux rob_w (n=rob@p549B904F.dip0.t-ipconnect.de) |
15:18.59 | *** join/#htc-linux goxboxlive (n=goxboxli@242.84-48-176.nextgentel.com) |
15:32.30 | Marex | seems like udc is screwed in mainline :E~ |
15:40.10 | Marex | looks like rndis issue |
15:51.21 | *** join/#htc-linux exception13 (n=exceptio@testdrive.kgts.ru) |
15:52.13 | *** join/#htc-linux kaboofa (n=tom@rain.angrycoder.org) |
15:53.26 | *** join/#htc-linux p3t3r__ (n=peter@wlan.net-you.de) |
16:05.53 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d86533b.pool.einsundeins.de) |
16:10.49 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbff42f.pool.einsundeins.de) |
16:14.02 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
16:24.46 | *** join/#htc-linux pikapika (n=pikapika@mar75-8-88-164-227-147.fbx.proxad.net) |
16:34.49 | *** join/#htc-linux xeroOTG (n=xero@75-173-64-182.albq.qwest.net) |
16:41.53 | *** join/#htc-linux ginge (n=baz@cpc6-ward3-0-0-cust579.manc.cable.ntl.com) |
16:53.20 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbf9138.pool.einsundeins.de) |
17:16.44 | *** join/#htc-linux Alex[sp3dev] (n=Alexande@86.110.163.25) |
17:36.10 | *** join/#htc-linux JohnnyK (n=johnnyk@nfx-nat-229.pilsfree.net) |
17:38.07 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
18:36.37 | *** join/#htc-linux marmotta (n=skodde@unaffiliated/skodde) |
18:49.02 | *** part/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
19:04.45 | *** join/#htc-linux bwr (n=william@67.60.252.132) |
19:07.20 | *** join/#htc-linux Marex (n=vasum7am@auryn.karlin.mff.cuni.cz) |
19:46.28 | cr2 | kiozen: still here ? |
19:46.44 | *** join/#htc-linux pH5 (n=ph5@p5485D014.dip.t-dialin.net) |
19:49.43 | pH5 | hi Marex, any news on the pxamci issue? I've just found a card that fails on my magician with that SCR error. |
19:51.46 | Marex | cr2: I just wrote my first asoc driver, hi ;) |
19:55.05 | Marex | pH5: no, no news :( |
19:55.33 | Marex | pH5: I was tinkering with the dma piece on pxamci, but still havent found where the error is |
19:55.43 | Marex | looks like the pointer gets crippled somewhere |
19:55.51 | Marex | pH5: dod the patch fixed it for you ? |
19:56.23 | Marex | s/dod/did/ |
19:56.44 | pH5 | Marex: didn't try yet, I just found out about the error. it's a kingston 1gb sd produced by toshiba. |
20:00.55 | Marex | pH5: I have the same one :) |
20:01.05 | Marex | also kingston-toshiba |
20:02.24 | cr2 | Marex: nice :) i have a lot of other troubles. |
20:02.37 | pH5 | hi cr2 |
20:02.43 | Marex | cr2: it wasnt for n560, dont worry |
20:07.53 | cr2 | hi pH5 |
20:08.19 | pH5 | Marex: are you sure? just moving the type field in struct mmc_card fixes the problem? |
20:08.32 | Marex | pH5: yea |
20:08.43 | Marex | pH5: the dma pointer is off by 4 bytes |
20:10.54 | pH5 | :/ |
20:10.57 | pH5 | <-- dma noob |
20:16.18 | pH5 | but looking at the diff between old hh.org pxamci and current mainline, there don't seem to be so big differences in the dma code |
20:18.44 | cr2 | kiozen: bt gps |
20:18.47 | cr2 | root@htcuniversal:~# cat /dev/rfcomm0 |
20:18.47 | cr2 | $PSRFTXTVersion 311.000.000-LD01*22 |
20:19.25 | cr2 | $PSRFTXTCLK: 96250*09 |
20:19.25 | cr2 | $PSRFTXTCHNL: 12*5F |
20:19.25 | cr2 | $PSRFTXTBaud rate: 57600 System clock: 0.000MHz*57 |
20:19.25 | cr2 | $PSRFTXTDatum: 214, WGS84*68 |
20:19.26 | cr2 | $PSRFTXTDF PRO: NMEA, 57600, Msg:1,0,1,1,1,1,0,0,0,0*47 |
20:20.43 | cr2 | kiozen: but the ts does not work (in M, od -x /dev/input/event2 works), not quite sure about the buttons. |
20:22.38 | cr2 | kiozen: hehe. get the GSOD on pressing the sat button again. |
20:26.24 | cr2 | kiozen: http://lists.openstreetmap.org/pipermail/dev/2008-July/010894.html |
20:26.52 | cr2 | Can display OSM data (XML) |
20:26.56 | cr2 | insane |
20:29.55 | Marex | pH5: indeed ... that's also why I wasnt able to find anything (errr ... rather because I dont like tinkering with DMA ;) ) |
20:30.22 | Marex | pH5: but it has to be the pxamci, since sdhci works |
20:31.36 | *** join/#htc-linux Guest71218 (i=annonymo@cpc5-stok8-0-0-cust898.bagu.cable.ntl.com) |
20:32.03 | Marex | pH5: btw what does Marc want with my aSoC patch ... jeez ... I dont understand that code myself and I understand his comments even less ;D |
20:32.25 | pH5 | let me have a look.. |
20:32.35 | cr2 | kiozen: lol. "was not accepted for google SoC". right, some java crap is also welcomed there ;) |
20:33.24 | Guest71218 | hi sorry to ask a really bone question ive copied all the files under \my device\android run haret and when i press run i get "failed to load file \android\android.bin |
20:33.48 | Marex | pH5: well they want me to use some new functions, but they arent in "devel" yet as I understood from it |
20:34.09 | pH5 | Marex: hmm? on alsa-devel? I didn't get that mail yet. |
20:34.39 | Marex | pH5: I mean this "Ideally this should use the new snd_soc_dapm_new_controls() function..." |
20:35.13 | Marex | btw Im not subscribed to alsa-devel ... time to do so |
20:35.28 | pH5 | ugh, the joy of concurrent development :) |
20:36.55 | pH5 | Marex: could it be something simple like a bug in 4bit mode? |
20:37.01 | pH5 | (the sd issue I mean) |
20:38.45 | Marex | pH5: bug in 4bit mode ? how ? |
20:39.03 | Marex | pH5: it's using 4 bit mode in .21-hh20 as well |
20:39.11 | Marex | and the pointer is off by 4 BYTES |
20:40.25 | pH5 | oh, I wasn't aware of that. |
20:41.26 | Marex | yea |
20:41.34 | Marex | u32 type; is 4 bytes ;) |
20:44.29 | pH5 | I wasn't aware that hh.o uses 4bit mode already. I thought maybe those cards trigger trigger a pxamci hardware bug in 4bit mode. |
20:44.40 | pH5 | s/trigger // |
20:46.16 | Marex | pH5: hh doesnt use 4 bit mode ? |
20:46.39 | Marex | pH5: well you can disable 4 bit and see what it does, it wont be very hard |
20:46.49 | pH5 | ok, you patch fixes my problem, too. |
20:46.58 | pH5 | I'll now recompile without that change and with 4bit disabled. |
20:47.03 | Marex | pH5: btw -hh doesnt have the "type" field in "card" structure |
20:47.18 | Marex | pH5: it was added sometime later |
20:47.35 | Marex | that's why I tried fixing it this way and it actually worked |
20:48.58 | ginge | Guest71218: rename the file initrd-android to android.bin |
20:50.01 | pH5 | Marex: can you explain to me where does it DMA into that mmc_card structure? |
20:50.27 | pH5 | hrrn, whytf does magician hang in i2c_pxa_reset (but hx4700 doesn't ...) |
20:50.44 | Guest71218 | thank you |
20:52.25 | annonymouse | also thank you very much for doing what your doing |
20:53.07 | Marex | pH5: it does dma into card->scr |
20:53.14 | Marex | pH5: that's what's even worse |
20:53.33 | Marex | see sd_ops and search for "scr(" |
20:53.39 | Marex | s/ops/ops.c/ |
20:57.21 | Marex | pH5: found that "sg_init_one(...." there ? |
20:57.27 | Marex | pH5: that's the place |
20:59.11 | *** join/#htc-linux TimRiker (n=timr@64-255-80-56.nrp7.dsl.d3c.infowest.net) |
21:00.16 | pH5 | Marex: ok. so sg->offset is wrong by 4 bytes in this call? |
21:01.22 | Marex | what sg->offset ? |
21:01.35 | marajin | 'lo |
21:01.42 | pH5 | pokes around cluelessly in sd.c to see how the parts fit together |
21:02.19 | Marex | sd_ops.c:285 |
21:02.26 | pH5 | Marex: you said that some pointer is off by 4 bytes. which one is it? |
21:02.29 | Marex | pH5: there it sets up the DMA |
21:02.51 | Marex | sg_init_one( ... the scr pointer is off by 4 there |
21:02.55 | Marex | sd_ops.c:285 |
21:03.10 | pH5 | scr itself? but it's called as mmc_app_send_scr(card, card->raw_scr) from sd.c |
21:03.32 | Marex | pH5: right ... card->raw_scr is off as well |
21:03.42 | Marex | the data is put 4 bytes before it |
21:04.41 | Marex | pH5: just put a printk at sd_ops:290 and dump the scr ... you'll see the error |
21:05.10 | Marex | btw what about the aSoC ? I dont have these functions they want me to use in my kernel tree :/ |
21:05.23 | pH5 | Marex: thanks, of course still no idea why, but I think I understand what the issue actually is now. |
21:06.40 | Marex | pH5: what it is ? |
21:06.53 | Marex | cr2: any idea about the asoc ? ;-) |
21:07.14 | cr2 | Marex: for n560 ? |
21:07.44 | Marex | cr2: snd_soc_dapm_new_controls() etc. |
21:12.06 | Marex | pH5: can you tell me where can I obtain kernel tree that supports this garbage ? :( |
21:21.33 | *** join/#htc-linux exco (n=exco@e181127048.adsl.alicedsl.de) |
21:23.41 | pH5 | Marex: I don't know what is the official alsa tree. did you have a look at tiwai/sound-2.6 on git.kernel.org? |
21:23.46 | pH5 | I think it's an alignment issue |
21:24.04 | pH5 | disabling 4bit didn't help, but the dma code changed completely between hh.o and mainline |
21:24.37 | pH5 | the pxamci only supports single byte and 32byte burst dma descriptors. what are the alignment constraints? |
21:24.50 | pH5 | I see the scr address as 0xc2eb964c, that doesn't look very aligned |
21:25.17 | cr2 | c is 4byte aligned |
21:25.40 | pH5 | good enough? |
21:25.53 | cr2 | Marex: i have no clue about alsa, i've just copied the magician driver from magician :) |
21:25.55 | pH5 | we're 4 bytes off... |
21:26.17 | cr2 | hmm. |
21:27.51 | Marex | cr2: in -hh it's at 0xXXXXX000 |
21:27.56 | Marex | errr pH5 ^ |
21:28.05 | Marex | that's kinda better alignment |
21:28.46 | Marex | but this is completely out of my understanding ;) |
21:29.38 | cr2 | Marex: 000 is page aligned, not 4byte |
21:30.14 | cr2 | Marex: kmalloc/dma will certainly return page aligned addresses. |
21:31.11 | cr2 | kiozen: bt gps got a fix, works as expected. fix your fscked qtopia stuff :) |
21:34.16 | pH5 | yup, so basically the pxa27x DMA controller is a 64bit device, 8byte alignment is needed unless you take special measures (involving the DALGN register and performance loss) |
21:34.43 | Marex | pH5: so we know where the problem is |
21:34.43 | pH5 | I wonder why some cards work at all, then. |
21:34.44 | Marex | I see |
21:35.20 | Marex | pH5: probably since it's an accident |
21:35.21 | Marex | ;) |
21:35.53 | cr2 | with 2.6.26 kernel ? |
21:36.34 | pH5 | yup, actually everything after the mmc revamp (was it .22-rc1/2 or .23-rc1/2?) |
21:37.58 | cr2 | .24+ was sdio ? |
21:39.28 | pH5 | cr2: I think so, yes. ancient history :) |
21:42.04 | cr2 | hmm. |
22:33.25 | pH5 | Marex: http://en.pastebin.ca/1062588 <-- this also did the trick for me |
22:33.46 | pH5 | Marex: can you take it up with Pierre to get us properly 64bit aligned dma targets? |
22:34.29 | Marex | pH5: join #libertas |
22:34.42 | Marex | pH5: ossman resides there all the time |
22:34.52 | pH5 | Marex: will do tomorrow, right now I need some sleep |
22:35.01 | Marex | pH5: btw that solution seems pretty ... ewwwww |
22:35.03 | Marex | pH5: :) |
22:35.08 | pH5 | yup |
22:35.17 | Marex | btw pxamci seems to be kinda unliked driver |
22:35.20 | pH5 | there has to be a proper way to do this |
22:35.22 | Marex | noone wants to touch it :D |
22:35.28 | Marex | pH5: right |
22:38.32 | pH5 | Marex: just allocating mmc_card 8byte aligned and moving type (or another word) out of the way like in your patch is probably the cleanest way. there just has to be a kernel api for that kind of thing. |
22:38.35 | pH5 | good night |
22:45.19 | cr2 | kiozen: FYI -> http://lists.openmoko.org/pipermail/community/2008-July/020552.html |
23:32.41 | *** part/#htc-linux p3t3r__ (n=peter@wlan.net-you.de) |
23:34.43 | *** join/#htc-linux lpotter (n=ljp@CPE-124-191-144-181.vic.bigpond.net.au) |