| 00:00.07 | dcordes_ | if you want to look into how that works, might wanna look at that source |
| 00:00.07 | tmzt_ | yeah, that was a while ago |
| 00:00.20 | marajin | quack |
| 00:00.31 | dcordes_ | oink |
| 00:00.42 | tmzt_ | found, accidently, the remote control commands for most cdma phones (not pdas) |
| 00:00.43 | dcordes_ | tmzt_: I can email it? |
| 00:00.51 | tmzt_ | yeah. |
| 00:00.58 | dcordes_ | <PROTECTED> |
| 00:00.59 | tmzt_ | do you need an email? |
| 00:01.37 | ginge | marajin: wb |
| 00:01.48 | ginge | marajin: do you fancy trying sd card support :) |
| 00:01.56 | ginge | marajin: sdhc that is |
| 00:05.29 | tmzt | used bitpim to try to backup a phone, and it started pressing buttons, so I looked in the protocol log, and it decoded the messages |
| 00:05.57 | tmzt | might use it in telepathy or something, with bluetooth and audio gateway |
| 00:06.29 | dcordes_ | what phone? |
| 00:06.40 | tmzt | simple LGs |
| 00:06.58 | tmzt | VX8300 |
| 00:07.35 | dcordes_ | in wince you can log the AT communication with a registry switch ( at least on my kaiseR ) |
| 00:07.38 | marajin | ginge: Sure :) |
| 00:07.53 | ginge | marajin: http://headfuzz.co.uk/files/android/zImage-sdhctest mounting and ls would be a bonus |
| 00:08.20 | tmzt | yeah, I thought I could access it through a tcp port (rndis), or with the modem mode, but that is only an AT emulator/translator |
| 00:08.21 | marajin | ginge: mounting is what we're aiming for dude :P it ddi everything else anyway no? |
| 00:08.37 | tmzt | so will have to write a shunt for ce |
| 00:09.15 | tmzt | then there is usbipc and whatever connection is between the pxa and the msm (I think its the full usb host) |
| 00:09.44 | tmzt | Not that phone, on a wince phone |
| 00:11.14 | marajin | note to self: NEVER put a microSD card down, you may never find it again. |
| 00:11.16 | dcordes_ | is confused |
| 00:11.27 | tmzt | was going to get htc dash (excalibur) but didn''t want to leave Verizon |
| 00:11.38 | marajin | dcordes_: More beer! You won't be less confused but you will care less |
| 00:11.40 | dcordes_ | marajin: I thought mine was lost for many weeks |
| 00:12.00 | dcordes_ | I'm drunk enough |
| 00:12.09 | tmzt | I think I have one in a MinSD adapter somewhere, otherwise it is gone |
| 00:12.10 | ginge | I lost a 4GB one up a vacuum cleaner |
| 00:12.24 | tmzt | my sansa corrupted the fat anyway |
| 00:12.30 | marajin | ginge: Oh that's why you have no SDHC huh? |
| 00:12.30 | tmzt | 1G |
| 00:12.45 | ginge | luckily they are cheaper than alcohol at the moment |
| 00:12.55 | ginge | marajin: yeah |
| 00:12.57 | marajin | < 15 quid for a 4GB yeah |
| 00:13.38 | marajin | wow.. tehere's 6 zImages on my kaiser now |
| 00:13.55 | ginge | zImage(5) |
| 00:14.35 | marajin | yes ginge, zImage(5) |
| 00:15.49 | marajin | um what ws the mknod again? mknod /dev/mmctest0 b 179 1? |
| 00:16.55 | marajin | ginge: You may have a cookie |
| 00:17.07 | marajin | ginge: mounts and lses |
| 00:17.17 | dcordes_ | wooha |
| 00:18.09 | ginge | :) |
| 00:18.16 | marajin | No error at all that I saw |
| 00:18.19 | *** join/#htc-linux deet (n=detran11@adsl-69-228-94-184.dsl.pltn13.pacbell.net) |
| 00:18.21 | marajin | not even the codepage nag |
| 00:18.25 | dcordes_ | ginge: is that with the patch you handed? |
| 00:18.39 | ginge | dcordes_: no, you want another? |
| 00:18.56 | marajin | May as well have working SHDC in the tree |
| 00:18.56 | dcordes_ | yes please. |
| 00:19.05 | ginge | marajin: i added the right codepage in the kernel. |
| 00:19.13 | ginge | you probably see some cmd timeout, but you can ignore them |
| 00:19.18 | marajin | that'd make the kaiser the best supported MSM device at the moment |
| 00:19.34 | dcordes_ | how'd we top vogue? |
| 00:19.41 | dcordes_ | we being ginge |
| 00:19.42 | marajin | SDHC works? :P |
| 00:19.52 | dcordes_ | dzo also has it :(( |
| 00:19.57 | marajin | oh =/ |
| 00:20.02 | marajin | nevermind, we're equal hen I guess |
| 00:20.20 | marajin | Or does dzo hav.. oh he has screen dimming and vibrate |
| 00:20.20 | dcordes_ | also he has more proc comm regs |
| 00:20.22 | dcordes_ | we need to find out the remaining functions |
| 00:20.28 | marajin | we can do it! |
| 00:20.29 | tmzt | not equal, more devices is good, everybody wins |
| 00:20.35 | ginge | yeah, proc comm tracing is easy |
| 00:20.44 | marajin | all that takes is time |
| 00:21.00 | ginge | I found one by accident the other day. What was it now... power change signal maybe |
| 00:21.09 | marajin | dcordes_: How many beers d'ya think we owe Ginge now? |
| 00:21.31 | dcordes_ | depending on how many he can drink |
| 00:21.39 | marajin | haha |
| 00:21.47 | ginge | can sink 5 pints before death |
| 00:21.49 | dcordes_ | too many anyway |
| 00:21.54 | marajin | 5 pints? |
| 00:21.56 | marajin | /5/ ? |
| 00:22.04 | marajin | You disappoint me greatly |
| 00:22.05 | ginge | is a lightweight |
| 00:22.21 | ginge | I had to give up for a number of years because I drank myself to ulcers in my youth |
| 00:22.40 | marajin | Heh, fair enough reason really mate |
| 00:23.49 | marajin | We'll just ... space the pints out |
| 00:24.13 | dcordes_ | ginge: does the patch also have proc_comm? |
| 00:24.16 | ginge | yes |
| 00:24.49 | dcordes_ | could you also put the .config you use as defconfig? |
| 00:25.31 | ginge | thats quite subjective. What should we include a bare minimum in the defconfig? |
| 00:25.39 | ginge | should we just distribute my .config instead? |
| 00:26.30 | ginge | dcordes_: according to this you have the latest gitdiff. I must have diffed after the changes |
| 00:26.48 | dcordes_ | I think it should have a minimum filesystems because of the sd |
| 00:26.53 | dcordes_ | mmc w/o fs makes no sense |
| 00:27.12 | marajin | well yeah we need at least vfat an ext2/3 |
| 00:27.26 | dcordes_ | but I think that's not so critical |
| 00:27.31 | ginge | so msdos vfat ext2 ext3 and block devices? |
| 00:27.45 | dcordes_ | yea |
| 00:27.53 | dcordes_ | jffs2 would be useful also |
| 00:27.57 | dcordes_ | often used in embedded |
| 00:28.00 | dcordes_ | jfs |
| 00:28.17 | marajin | do we care about thing like cramfs |
| 00:28.39 | dcordes_ | I think people can decide really |
| 00:28.48 | dcordes_ | who can fetch the git, can configure the kernel |
| 00:29.05 | dcordes_ | I think that's even in the small hwoto in xdadev |
| 00:29.07 | ginge | yeah, this is about bare minimum for defconfig |
| 00:29.16 | ginge | which I think should be msdos and vfat only |
| 00:29.24 | ginge | why force ext2/3 on people |
| 00:29.32 | marajin | heh |
| 00:29.43 | marajin | because on linux vfat/msdos don't much make sense |
| 00:30.05 | ginge | well get rid of those too then |
| 00:30.08 | tmzt | how exactly do you put your root on vfat (and even if you use loop you still need the fs) |
| 00:30.21 | ginge | the defconfig should only really contain options that are platform specific |
| 00:30.38 | marajin | tmzt: exactly... we pretty much need ext too |
| 00:30.42 | tmzt | if they were inherited, yeah |
| 00:30.46 | dcordes_ | ginge: that would only make msm_sdcc out of it and the dependency tree |
| 00:31.15 | ginge | dcordes_: right |
| 00:31.41 | ginge | so we distribute a kernel with all reasonable filesystems compiled in, provide a .config and fill defconfig with msmsdcc and deps |
| 00:32.08 | tmzt | doesn't defconfig replace .config? |
| 00:33.04 | tmzt | now that you have sd on kaiser/vogue/titan, you could use a small initrd for the device specific stuff |
| 00:33.19 | ginge | tmzt: yes. we have menuconfig and oldconfig for adding optionals like additional filesystems. thats my pov on how it should be used anyway |
| 00:33.20 | tmzt | and put other modules in there |
| 00:33.35 | dcordes_ | we have a monolithic kernel atm |
| 00:33.52 | dcordes_ | tmzt: it's only ~ 1M with all drivers |
| 00:33.56 | dcordes_ | with the filesystems like 1,3 |
| 00:33.57 | ginge | dcordes_: most of the google code works really badly as a module |
| 00:34.10 | tmzt | of course if you have a ce rom with a large enough heap, and all the other problems fixed, you can boot as big of an intrd as ram will hold |
| 00:34.13 | tmzt | and ce is not using |
| 00:34.31 | dcordes_ | initd maximum is 20M |
| 00:34.34 | tmzt | but you really want ram for programs at runtime, not bins |
| 00:34.54 | tmzt | is that with Dutty's and similar big heap roms? |
| 00:35.20 | dcordes_ | I use dutty's may 22nd rom :-D |
| 00:35.21 | ginge | tmzt: I can run an android 20MB initrd in my bloated stock orange |
| 00:35.26 | dcordes_ | I like that description |
| 00:35.42 | tmzt | right, but what was the upper limit in ce5.5 stock |
| 00:36.03 | dcordes_ | I can boot nothing in vodafone.de stockl |
| 00:36.26 | ginge | dcordes_: you probably would be able to with remote haret. Nothing boots with the run button |
| 00:36.28 | tmzt | initrd has to be loaded before the mmu pagetables and other changes can be made, right? |
| 00:36.44 | dcordes_ | ginge: I could bet I tried that |
| 00:36.44 | ginge | tmzt: yes, as far as I understand it |
| 00:36.57 | tmzt | does it use the heap for malloc? |
| 00:37.13 | ginge | tmzt: I reallt don't know too much about the haret magic |
| 00:37.33 | tmzt | keven and cr2 then? |
| 00:37.49 | dcordes_ | Kevin2 is the haret author/maintainer |
| 00:37.59 | tmzt | yeah, Kevin2 |
| 00:38.09 | dcordes_ | both will be able to tell you |
| 00:38.30 | tmzt | i've used his hermes initrd for something, it was the other thing that worked |
| 00:38.41 | tmzt | s/other/only/ |
| 00:39.50 | tmzt | I thought my kernel was dead because pxafb / lcd didn't come up, but it worked. |
| 00:40.09 | dcordes_ | tmzt: I'm still in the tslib/raw ts image. any clue what I can try additionally to get response from ts in the calibration app/X? |
| 00:40.25 | tmzt | what have you tried? |
| 00:40.42 | tmzt | maybe evtest /dev/input/event0 |
| 00:40.47 | tmzt | look for absolute |
| 00:40.50 | tmzt | Abs |
| 00:41.43 | tmzt | did you get the strings on Xfbdev? |
| 00:41.54 | tmzt | is strings in busybox? |
| 00:42.20 | ginge | yes built in |
| 00:42.26 | tmzt | sometimes the symlinks are'nt there, but you can call busybox <applet> |
| 00:42.30 | tmzt | ok |
| 00:42.37 | tmzt | you do not have pipe |
| 00:43.00 | tmzt | strings /usr/bin/Xfbdev > tmp |
| 00:43.04 | tmzt | grep ts_ tmp |
| 00:43.09 | ginge | tmzt: I have a file on my server with all of the missing keys in. I just wget it :) |
| 00:43.33 | dcordes_ | ok |
| 00:43.58 | *** join/#htc-linux eBug (n=eric@100.Red-83-50-226.dynamicIP.rima-tde.net) |
| 00:44.08 | dcordes_ | tmzt: what to look for? |
| 00:44.16 | tmzt | ts_open ts_close |
| 00:44.16 | eBug | dcordes_ i finally was able to make it to the android main screen |
| 00:44.27 | tmzt | and then grep libts tmp |
| 00:44.44 | tmzt | look for libts0 |
| 00:44.53 | ginge | eBug: off sd? |
| 00:45.03 | dcordes_ | libts-1.0.so.0 |
| 00:45.09 | tmzt | okay. |
| 00:45.23 | tmzt | evtest /dev/input/event0 |
| 00:45.33 | tmzt | look for Abs when you touch the screen |
| 00:45.40 | dcordes_ | ginge: eBug is on sedna. we tried the kaiser zImage with sd and it doesn't boot. normal kaiser zImage boots- so sedna has some sd differences |
| 00:45.44 | dcordes_ | eBug: how did you reach it? |
| 00:45.49 | dcordes_ | tmzt: ok |
| 00:46.30 | dcordes_ | tmzt: that looks neat. lots of Abs |
| 00:46.49 | tmzt | does it make any sense (are the coords right?) |
| 00:46.58 | tmzt | at least, upper left, lower right |
| 00:47.02 | tmzt | pressure? |
| 00:47.40 | dcordes_ | it is always 3 lines followed by -----------Report Sync--------------------- |
| 00:48.02 | tmzt | I'll see if in can compile it native on my hpc |
| 00:48.11 | dcordes_ | the 3 lines are on X one Y and one pressure |
| 00:48.12 | ginge | x y p is the layout in the touchscreen reporting order |
| 00:48.29 | tmzt | hostap is staying up again with wpa |
| 00:48.40 | tmzt | it's been dropping all morning |
| 00:49.15 | tmzt | maybe irssi will work now |
| 00:49.23 | deet | dcordes: sorry to interrupt you but do you have an ad for sunnyvale and can I speak to you about that? |
| 00:50.02 | dcordes_ | deet: you don't interrupt. is that a city? |
| 00:50.15 | dcordes_ | tmzt: the values change, but I am not sure on the correctness |
| 00:50.24 | deet | dcordes: yes sunnyvale calif |
| 00:50.34 | tmzt | that would be raw from the kernel |
| 00:50.43 | ginge | dcordes_: the touchscreen driver reports odd values. y and p are not lineat |
| 00:50.53 | tmzt | try TSLIB_TSDEVICE=/dev/input/event0 ts_print_raw |
| 00:51.01 | dcordes_ | I think there is some rough connection between position and valu size but I'm not too sure |
| 00:51.21 | dcordes_ | deet: so what about that place? |
| 00:52.05 | deet | dcordes: i saw your name and channel in an ad , I thought it was your ad |
| 00:53.21 | deet | the ad was for a linux developer |
| 00:53.46 | dcordes_ | deet: ha that's funny man :D I doubt that's me |
| 00:54.16 | tmzt | same type of values? |
| 00:54.19 | eBug | dcordes_ as we said yesterday, the input buffer is somehow filled with e's when android launches. Pressing repeatedly the 'hang up' button while android is booting flushes the buffer |
| 00:54.33 | dcordes_ | deet: never made such an ad, and I'm not much of a developer. can you point me to that ad? |
| 00:54.46 | deet | dcordes i guess they linked your page on linuxtogo? |
| 00:55.08 | dcordes_ | tmzt: the binary is broken somehow. I copied them from OE workdir but they say odd things |
| 00:55.26 | tmzt | somebody's in sunnyval, but I can't think who (palm? before) |
| 00:55.44 | deet | they linked this page http://linuxtogo.org/~lgorris/kaiser-bootkit/README.txt |
| 00:56.01 | dcordes_ | deet: yes I wrote this readme |
| 00:56.26 | dcordes_ | tmzt: shall I get that binary right or can we try something else? |
| 00:56.44 | tmzt | what does it say |
| 00:56.57 | tmzt | (the command) |
| 00:57.32 | deet | dcordes: ok sorry to bother you then. I am trying to contact "those people" in sunnyvale |
| 00:58.38 | dcordes_ | deet: maybe we can also help if you have a problem with the linux on htc stuff :) you're welcome |
| 00:59.22 | eBug | dcordes_ ts also somehow works |
| 00:59.22 | dcordes_ | tmzt: first line in ts_print_raw is |
| 00:59.40 | eBug | but it's not calibrated at all |
| 00:59.43 | dcordes_ | #ts_print_raw - temprary wrapper script for .libts/ts_printraw |
| 00:59.59 | dcordes_ | eBug: how did you manage to get to the main screen? |
| 01:00.07 | eBug | dcordes_ as we said yesterday, the input buffer is somehow filled with e's when android launches. Pressing repeatedly the 'hang up' button while android is booting flushes the buffer |
| 01:00.39 | eBug | so i reach the main screen with an empty buffer |
| 01:00.41 | dcordes_ | great. did you get to the dialer? |
| 01:00.46 | eBug | no |
| 01:00.54 | eBug | i can't press that button |
| 01:01.03 | eBug | ts isn't calibrated at all |
| 01:01.09 | dcordes_ | ginge: has the kaiser touchscreen working now so you'd try the on screen keyboard in angstrom |
| 01:01.20 | tmzt | there is a buffer in hw, or in linux driver? |
| 01:01.58 | eBug | pressing center of the ts triggers the upper left corner button |
| 01:02.08 | tmzt | 900c seems to be linear (its in a pic on the btuart) |
| 01:02.30 | dcordes_ | tmzt: some script ts_print_raw is, but it has paths from my buildhost. |
| 01:02.52 | tmzt | the ts.conf says linear modules, so maybe you need something else |
| 01:03.06 | tmzt | there is collie, corgie, a couple of tsc's and some other chips |
| 01:03.16 | tmzt | maybe wm something |
| 01:03.24 | ginge | tsc2003 and 2007 are compatible if you see those |
| 01:04.09 | dcordes_ | ok let's see |
| 01:04.34 | tmzt | /usr/lib/ts/ and no |
| 01:04.51 | tmzt | but h3600 h2200 ucb1x00 |
| 01:05.07 | dcordes_ | do they use tsc2003? |
| 01:05.13 | tmzt | I don't know |
| 01:06.02 | tmzt | I don't work on libts, I have just used it |
| 01:06.17 | tmzt | the list has a few questions answered |
| 01:06.22 | tmzt | it is written my rmk |
| 01:06.33 | tmzt | maybe arm list, not arm-kernel? |
| 01:07.54 | tmzt | but, even still, X cursor should be jumping around |
| 01:08.21 | tmzt | not standing still |
| 01:08.51 | tmzt | what do you have for modules (the bottom) of /etc/ts.conf |
| 01:09.13 | dcordes_ | second |
| 01:09.17 | tmzt | module_raw input should be chosen at the top |
| 01:09.45 | tmzt | that gets the input and passes it to the other modules to process |
| 01:10.02 | dcordes_ | tmzt: http://rafb.net/p/chSwVU70.html |
| 01:10.11 | tmzt | it looks like those device modules are pre input subsystem |
| 01:10.18 | tmzt | not used any more |
| 01:11.12 | dcordes_ | it's just the moko default stuff which mostly is fic-gta0? (moko phone) specific |
| 01:11.14 | tmzt | get rid of grab_events= (comment that line and copy or retype) |
| 01:11.20 | ginge | I think we might need to kernel level linearisation to get tslib working |
| 01:11.30 | tmzt | you are using a moko image? |
| 01:11.44 | dcordes_ | tmzt: yes. the one from the link I posted |
| 01:12.02 | tmzt | mine doesn't have grab_events |
| 01:12.21 | tmzt | right, I know it has moko bins, I didn't know it was for the gta |
| 01:12.44 | dcordes_ | I just assume that because we have not much defined for the kaiser in OE |
| 01:12.50 | dcordes_ | maybe stupid |
| 01:13.10 | tmzt | my file is the same except grab_events |
| 01:13.17 | tmzt | I don't know what it does |
| 01:13.26 | dcordes_ | so just comment the whole line? |
| 01:13.31 | tmzt | even linear should give you something, a little jerky |
| 01:13.41 | tmzt | no, just grab_events=0 |
| 01:13.52 | dcordes_ | ok |
| 01:13.52 | tmzt | module_raw input |
| 01:13.57 | tmzt | should be the line |
| 01:14.00 | dcordes_ | is |
| 01:14.13 | tmzt | then try TSLIB_TSDEVICE=/dev/input/event0 ts_calibrate |
| 01:14.43 | tmzt | I see, you are getting multiple Touch events, which is now a key instead of a pressure reading |
| 01:15.04 | tmzt | that would quickly confuse the calibrate |
| 01:15.05 | dcordes_ | I still get the pendown irqs |
| 01:15.11 | dcordes_ | s/irqs/printks/ |
| 01:15.20 | dcordes_ | I hope they do not confuse the calibration program |
| 01:15.29 | tmzt | event layer should be giving you Touch key events |
| 01:15.33 | tmzt | look at evtest |
| 01:15.50 | dcordes_ | evtest is also a bit hard to read because of the printks |
| 01:16.32 | dcordes_ | I can only read it when I do no touch |
| 01:16.52 | dcordes_ | because while holding down, the screen is flooded by "pen down 1" lines |
| 01:16.57 | dcordes_ | blames ginge |
| 01:17.16 | tmzt | dmesg -n 1 |
| 01:17.23 | ginge | sorry, you only had to say :) suppress all messages |
| 01:17.28 | exco | ;-) |
| 01:18.05 | dcordes_ | oh useful that dmesg -n 1 |
| 01:18.38 | tmzt | moko has patched tslib, added grab_events |
| 01:18.44 | tmzt | not sure what for yet |
| 01:19.18 | tmzt | look in oe/packages/tslib/tslib-1.0/tslib-input_raw-grab_events.patch |
| 01:19.22 | tmzt | in oe src |
| 01:19.29 | dcordes_ | tmzt: yep also get the touch key (when I tap shortly w/o holding down moving I see the 4th line with key something) |
| 01:19.34 | tmzt | if you are using moko oe |
| 01:19.48 | dcordes_ | pendown event |
| 01:19.56 | tmzt | you turned grab_events off? |
| 01:20.05 | dcordes_ | yes in the .conf as you said |
| 01:20.23 | dcordes_ | is that config passed when you use? |
| 01:20.23 | ginge | http://www.headfuzz.co.uk/files/android/zImage-tsraw |
| 01:20.36 | ginge | should be nice and quiet :) |
| 01:20.49 | dcordes_ | ginge: thanks. I think we shoulÃd definetly put those printks in the git |
| 01:21.00 | dcordes_ | not |
| 01:21.11 | ginge | this is why we test :) |
| 01:21.25 | ginge | and wy you don't have this code :P |
| 01:21.52 | tmzt | can you build libts from a non moko repo? |
| 01:22.01 | dcordes_ | I can try |
| 01:22.42 | tmzt | well I'm confused, there's no metadata on openmoko svn that I can find |
| 01:22.51 | tmzt | so it must be upstream oe? |
| 01:23.33 | dcordes_ | they have a web thing but I'm not sure if that's up to date |
| 01:23.41 | dcordes_ | but I have the whole stuff locally |
| 01:24.29 | dcordes_ | oh I did not notice they have a git now! |
| 01:24.46 | dcordes_ | http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=tree;f=packages;h=9610ccacc04717497ee8bc5ecf6570bce9fa6f9e;hb=refs/heads/org.openembedded.dev |
| 01:25.08 | dcordes_ | folder tslib has all the metadata |
| 01:26.53 | tmzt | huh, the grab_events just prevents them from being treated as mouse events |
| 01:27.58 | tmzt | comment out everything (at the bottom) except module linear |
| 01:28.14 | dcordes_ | ginge: I just hit the camera button and I'm able to tap the on screen keyboad keys |
| 01:28.19 | dcordes_ | ok |
| 01:28.38 | ginge | dcordes_: yeah, its still a bit off tho |
| 01:28.51 | ginge | I hope you guys figure this out :) |
| 01:28.58 | tmzt | you have touchscreen keyboard enabled? |
| 01:29.12 | dcordes_ | ginge: it does not output anything, does it? |
| 01:29.17 | tmzt | or is that a different device |
| 01:29.39 | ginge | maybe :) |
| 01:29.40 | tmzt | what kernel src are you working with |
| 01:29.49 | ginge | tmzt: I ported the fb keyboard to kaiser |
| 01:29.54 | dcordes_ | tmzt: so I only got input and linear? |
| 01:29.54 | tmzt | drivers/input/* |
| 01:30.08 | pigeon | ah, glad to see haret works on the wallaby. |
| 01:30.16 | tmzt | yeah, see if that works, it might jump around but you'll no what the problem is (hopefully) |
| 01:30.26 | dcordes_ | tmzt: http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/htc-msm |
| 01:30.44 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 01:31.42 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-742150c5e92c3c0a) |
| 01:31.43 | tmzt | what driver *.c |
| 01:32.07 | dcordes_ | tmzt: the touchscreen keyboard is not in yet |
| 01:32.20 | tmzt | not keyboard, the touchscreen itself |
| 01:32.26 | dcordes_ | ginge has the relevant patches |
| 01:32.38 | dcordes_ | the tsc2003.c in arch/arm/mach-msm/ |
| 01:32.53 | tmzt | although it would be interesting to know how the keyboard is hooked up, since it has to be doing the translation |
| 01:32.58 | tmzt | for itself |
| 01:34.00 | dcordes_ | when I evtest /dev/input1 (keyboard) I get a Report Sync line verytime I tap the ts. but only on tap |
| 01:34.18 | dcordes_ | tmzt: I think ts_calibrate just doesn |
| 01:34.28 | dcordes_ | tmzt: I think ts_calibrate just doesn't move the crosshair, still it calibrates |
| 01:34.41 | ginge | dcordes_: the touchscreen penirq is on the keyboard gpios |
| 01:34.42 | dcordes_ | because when I do a certain amount of taps, it exits |
| 01:35.28 | dcordes_ | tmzt: 5 times. I suspect the ts_calibrate to simply be broken. |
| 01:36.11 | tmzt | you do 5 taps for it to quit |
| 01:36.16 | dcordes_ | yep |
| 01:36.24 | tmzt | if you click on the first one does it go away |
| 01:36.33 | dcordes_ | just like in the gpe screen calibration program |
| 01:36.42 | dcordes_ | no it doesn't want to move. |
| 01:37.04 | tmzt | it doesn't move, it just goes away and another one comes |
| 01:37.12 | dcordes_ | not happening |
| 01:37.26 | tmzt | even if you tap 5 times it should go away each time |
| 01:37.39 | dcordes_ | so it's obviously broken |
| 01:37.46 | tmzt | yeah |
| 01:38.10 | tmzt | what names do you get from /sys/class/input/*/name |
| 01:38.24 | tmzt | and what is the .c file for the driver? |
| 01:39.02 | dcordes_ | ginge: could you pastebin your latest tsc2003.c? |
| 01:40.07 | dcordes_ | tmzt: the output is kaiser-ts |
| 01:40.22 | dcordes_ | and kaiser kbd |
| 01:40.24 | tmzt | that is the only input device |
| 01:40.31 | tmzt | ok. so event0 is ts |
| 01:40.35 | tmzt | event1 is kbd |
| 01:40.45 | tmzt | do you have a /dev/input/touchscreen0 |
| 01:41.06 | dcordes_ | yep |
| 01:41.23 | tmzt | mickeyl has an event debug program for moko, I just saw it in the svn (not git) |
| 01:42.33 | tmzt | what is the file under devices/input in that kernel git |
| 01:42.51 | dcordes_ | which? |
| 01:43.00 | tmzt | the input driver for ts |
| 01:43.20 | dcordes_ | arch/arm/mach-msm/tsc2003.c |
| 01:43.37 | dcordes_ | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=arch/arm/mach-msm/tsc2003.c;h=e9a3efec561b8b1a27fe2e4d1c70fb5c9464c607;hb=refs/heads/htc-msm |
| 01:43.37 | ginge | http://pastebin.com/m9321df |
| 01:44.15 | dcordes_ | tmzt: use the one in the pastebin it's more up to date |
| 01:44.32 | tmzt | its not in git anyway |
| 01:44.34 | ginge | that wont actually compile on its own |
| 01:44.40 | tmzt | unless the head is wrong |
| 01:44.50 | ginge | goodnight all |
| 01:45.13 | tmzt | vogue or kaiser, this is for vogue? |
| 01:45.17 | tmzt | both? |
| 01:45.19 | dcordes_ | ginge: one sec. can you give me the final patch? |
| 01:45.30 | ginge | I did |
| 01:46.05 | ginge | tmzt: kaiser. Vogue touchscreen is different and linear |
| 01:46.21 | tmzt | comment at top was wrong then. |
| 01:47.03 | ginge | its a wip, but thanks. thats where the port is based |
| 01:47.20 | ginge | dcordes_: that link to the gitdiff is the latest with sdhc etc |
| 01:47.55 | ginge | http://pastebin.com/m7f6cd418 ;) |
| 01:48.06 | ginge | night |
| 01:48.30 | tmzt | input_report_abs() is the output from ts? |
| 01:48.37 | dcordes_ | tmzt: that's getting to abstract for me. fear I can't help |
| 01:48.43 | ginge | tmzt: yes |
| 01:48.51 | tmzt | osk and ts come from the same event device? |
| 01:49.08 | dcordes_ | refetches current head, patches |
| 01:49.28 | tmzt | and then input_sync() |
| 01:49.37 | tmzt | that would be all those events, I guess |
| 01:50.11 | ginge | nothing happens while the touchscreen is not displayed |
| 01:50.17 | tmzt | XMIN, XMAX, YMIN, YMAX are right for kaiser? |
| 01:50.18 | ginge | nothing special anyway |
| 01:50.31 | ginge | tmzt: for this particular driver implementation, yes |
| 01:50.38 | tmzt | yeah, I see that now |
| 01:50.38 | ginge | they are also not used when reporting raw |
| 01:51.09 | ginge | i really must go. night all |
| 01:51.17 | dcordes_ | night |
| 01:52.46 | tmzt | it figures out the key on pendown, and reports on penup, but only if enabled |
| 01:53.30 | dcordes_ | tmzt: if it actually works in the testev program, I don't get the point of the problem |
| 01:53.57 | tmzt | yeah, it is reported raw events from adc (tsc) |
| 01:54.30 | dcordes_ | and that's what tslib needs, no? |
| 01:54.41 | tmzt | do you have showkey? |
| 01:54.48 | tmzt | yes, depending on the modules |
| 01:54.57 | dcordes_ | yes I have it |
| 01:55.38 | dcordes_ | one screen tap yields the following sequence |
| 01:55.45 | dcordes_ | 0 press |
| 01:55.50 | dcordes_ | 2 release |
| 01:55.53 | dcordes_ | 74 release |
| 01:55.56 | dcordes_ | 0 release |
| 01:55.58 | dcordes_ | 2 release |
| 01:56.02 | dcordes_ | 74 release |
| 01:56.42 | tmzt | gpm is running, blocking those for me, hold on |
| 01:57.02 | tmzt | not getting those at all |
| 01:57.28 | dcordes_ | when I oly hold down it oly shows the first 3 lines from above |
| 01:57.36 | dcordes_ | and then on release the other 3 |
| 01:58.52 | tmzt | i think Touch and maybe stylus are reported by showkey, but it is busybox, so know keysyms |
| 01:59.36 | tmzt | http://svn.openmoko.org/trunk/src/target/debug/input/input.c |
| 01:59.46 | tmzt | but it needs glib |
| 02:02.03 | *** join/#htc-linux tmzt_hpc (n=root@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) |
| 02:02.46 | dcordes_ | if I had a way into the device it would be useful to keep a native build enviornment |
| 02:14.19 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-67-49-214-162.bak.res.rr.com) |
| 02:29.06 | *** join/#htc-linux tmzt_hpc (n=root@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) |
| 02:31.22 | dcordes_ | tmzt_hpc: on the go now? |
| 02:31.39 | tmzt | back, in the kitchen, and testing hostap_cs |
| 02:31.42 | tmzt | only wifi |
| 02:32.29 | tmzt | I'm really not sure what else to do, mayble look at some source for libts and libts-utils |
| 02:32.51 | dcordes_ | I think I will dig through arm mailing lists and stuff tomorrow |
| 02:33.05 | dcordes_ | and see if I can get info on tsc2003 vs tslib |
| 02:33.33 | tmzt | any other htc with working ts in Xfbdev? |
| 02:34.06 | dcordes_ | there is exco and pH5 on this channel who use openmoko on magician |
| 02:34.13 | dcordes_ | no clue if they use fbdev |
| 02:34.18 | tmzt | eabi |
| 02:34.32 | tmzt | opemoko would be Xfbdev, or Xomap, or similar |
| 02:34.43 | tmzt | well, it's s3c |
| 02:35.04 | tmzt | any of the gpe's should be Xfbdev |
| 02:36.56 | dcordes_ | here's a moko magician rootfs http://www.linuxtogo.org/~htcpxa/htcmagician/images/OpenMoko/openmoko.linux.rootfs.tar.bz2 I will look into this |
| 02:37.11 | tmzt | I got to find some space first |
| 02:37.37 | tmzt | Also why no oe or bitbake |
| 02:38.09 | tmzt | have 160gb for my laptop, but has bad blocks so I have been restricted to a tiny partition, and now nothing |
| 02:38.18 | tmzt | this computer has 10gb |
| 02:43.21 | tmzt | is there ts driver for kaiser with no osk |
| 02:44.06 | tmzt | should there be penup pendown messages in userspace, not trapped by osk |
| 03:07.55 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 03:13.38 | dcordes | any POLARIS user present by chance? |
| 03:48.36 | CPress | Hey guys |
| 03:54.15 | *** join/#htc-linux AstainZZZZZZ (n=Administ@unaffiliated/astainhellbring) |
| 04:25.43 | tmzt | instructions to build kdrive: http://rene.rebe.name/2007-04-29/faq-where-has-k-drive-aka-tinyx-gone |
| 04:26.13 | tmzt | ./configure --enable-kdrive --disable-dri --diasable-xorg --disable-xorgcfg |
| 04:26.19 | tmzt | probably more for tslib |
| 04:37.56 | tmzt | looks like oe has all that though |
| 04:38.22 | tmzt | will look into building an Abs only evtest |
| 04:53.16 | tmzt | multievent patch: http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob;f=packages/tslib/tslib/multievent.patch;hb=HEAD |
| 06:38.47 | *** join/#htc-linux goxboxlive (n=goxboxli@31.84-48-212.nextgentel.com) |
| 06:49.17 | *** join/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
| 06:53.28 | premy | dcordes: hi |
| 06:54.31 | premy | dcordes: just tested the zImage you posted on xda/polaris. Works for me. The best would be to replace the one in the zip file, to make sure everyone uses and reports problems against the same image. |
| 06:55.12 | premy | dcordes: and thanks for the uploads. |
| 07:18.13 | pigeon | has anyone experienced an issue where no matter how you format an mmc/sd card the damn pocketpc (on the wallaby in this case) just won't recognize it and also fails to format it? |
| 07:19.51 | tmzt | what's your partition table look like? |
| 07:19.54 | pigeon | the card works under linux and winxp, and i've also tried formatting it under linux and winxp. same thing. |
| 07:20.11 | pigeon | however, i have another card which works on the device, but it's not mine, so i can't use it. |
| 07:20.15 | tmzt | fdisk -l /dev/sdb (or /dev/mmcblock0) |
| 07:20.23 | pigeon | hmm, what in particular? |
| 07:20.51 | tmzt | this is 2G or under card? |
| 07:21.11 | pigeon | yes |
| 07:21.25 | pigeon | a 32mb mmc, and a 256mb sd |
| 07:21.26 | tmzt | fat32 should be type 'b' I think |
| 07:21.57 | tmzt | that might be fat16 |
| 07:22.00 | pigeon | the card that works, has this: |
| 07:22.12 | pigeon | /dev/mmcblk0p1 3 10112 495369+ 6 FAT16 |
| 07:22.35 | pigeon | i tried exactly that on my mmc and sd, doesn't work. |
| 07:23.10 | pigeon | well, being FAT16, and mkfs.vfat -F 16 |
| 07:23.19 | tmzt | 6 would be fat16, right |
| 07:23.24 | tmzt | had to look it up |
| 07:23.28 | pigeon | unless there's something else i need |
| 07:23.41 | tmzt | have you tried backing it up and formating on the device? |
| 07:24.06 | pigeon | i have, i plug the card it, it asks if i want to format it, then it fails to format it. |
| 07:24.29 | tmzt | yeah, I saw that up above |
| 07:24.52 | tmzt | maybe clear the partition dd if=/dev/zero of=/dev/mmcblk0p1 |
| 07:25.11 | pigeon | i tried that too :\ |
| 07:25.33 | tmzt | try fdisk and option o to create a new table |
| 07:26.26 | tmzt | there are formating tools for ce, but then it would be the same as formating on Windows or Linux |
| 07:27.17 | pigeon | yeah, i think i saw some people having the same issue, and they had to use a special formating tool on windows. |
| 07:27.32 | pigeon | i'm hoping someone here can tell me the linux way, that's all :) |
| 07:27.33 | tmzt | what Win CE/Win Mobile is on the device |
| 07:27.43 | pigeon | pocket pc 2002 |
| 07:28.06 | tmzt | is that ram based (does a reboot lose data) |
| 07:28.47 | tmzt | might be an update for ppc2002, or update to a newer ppc, but you'll have to ask someone here who knows the device |
| 07:29.22 | pigeon | it has 32mb flash, in ppc there's a "permanent save" settings, whatever that's for. |
| 07:29.25 | pigeon | ok. |
| 07:33.08 | pigeon | ahhhhh |
| 07:33.37 | pigeon | zero-ing it and then let the device formating does the trick. |
| 07:33.52 | pigeon | i did zero-ing and fdisk + mkfs.vfat on linux last time, and didn't work. |
| 07:33.58 | pigeon | tmzt: thanks! |
| 07:34.19 | tmzt | sure |
| 07:34.29 | tmzt | you zeroed the whole device? |
| 07:34.35 | pigeon | yeah |
| 07:35.05 | tmzt | I didn't suggest that because I have had issues woth sd cards before, the never seem to recover. |
| 07:35.30 | tmzt | Usually they are preformatted and have a particular partition table |
| 07:35.48 | tmzt | It doesn't matter much to linux, but Windows seems to care. |
| 07:35.59 | pigeon | yeah, i suspect my friend's card which worked is a preformatted one too. |
| 07:36.36 | tmzt | Windows concerns itself with the partition types (maybe not NT), linux doesn't |
| 07:36.45 | pigeon | ahhh |
| 07:37.10 | pigeon | after formating the card on this ppc, i realized something. |
| 07:37.23 | pigeon | it's creating the fs on the whole card. |
| 07:37.31 | pigeon | it didn't bother creating a partition. |
| 07:37.38 | tmzt | that was why I asked about fdisk |
| 07:37.52 | tmzt | some cards with fat16 are treated like floppies |
| 07:38.28 | tmzt | it might be a ce3/4 thing, with its simple devices |
| 07:39.00 | pigeon | my friend's 512mb has a partition, while this one doesn't. fun. |
| 07:39.33 | tmzt | yeah, at that size I'm not sure fat16 will work, and I think fat32 requires the partition |
| 07:40.36 | *** join/#htc-linux marajin (n=marajin@87-194-102-189.bethere.co.uk) [NETSPLIT VICTIM] |
| 07:40.36 | *** join/#htc-linux pronto (n=pronto@pool-71-179-232-167.bltmmd.fios.verizon.net) [NETSPLIT VICTIM] |
| 07:40.36 | *** join/#htc-linux NiZoX (n=none@2a01:e35:8a13:a2b0:21c:c0ff:fe25:ff68) |
| 07:40.36 | *** join/#htc-linux thinko (n=jdoe6alp@smaug.rackdragon.com) [NETSPLIT VICTIM] |
| 07:40.36 | *** join/#htc-linux paulproteus (n=paulprot@2002:cbb2:8293:0:0:0:0:1) |
| 07:40.36 | *** join/#htc-linux hlbot (n=adm@iclem.net) [NETSPLIT VICTIM] |
| 07:50.03 | *** join/#htc-linux Othello_ (i=Magorium@gateway/tor/x-bac919b330109d62) |
| 07:51.59 | premy | Rroet: are you around ? |
| 08:01.08 | *** part/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
| 08:06.41 | Rroet | premy ... |
| 08:06.48 | Rroet | Premy!!!!.. yes I'm just in. |
| 08:23.27 | Rroet | YES!.. grps worsks, phone works.. sms works |
| 08:32.26 | *** join/#htc-linux Othello_ (i=Magorium@gateway/tor/x-1387a9a9b66eda2e) |
| 08:35.45 | Rroet | Guys, incase I'm afk lateron. Tell Premy it all worked :) |
| 09:09.39 | *** join/#htc-linux goxboxlive (n=goxboxli@31.84-48-212.nextgentel.com) |
| 09:30.05 | *** join/#htc-linux amiral_ (n=root@gob75-2-81-56-65-10.fbx.proxad.net) |
| 09:35.12 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
| 09:39.33 | *** join/#htc-linux exco (n=exco@e181095143.adsl.alicedsl.de) |
| 10:30.16 | *** join/#htc-linux LunohoD (n=alex@e180077138.adsl.alicedsl.de) |
| 10:52.17 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 11:07.06 | *** join/#htc-linux pH5 (n=ph5@p5485BECC.dip.t-dialin.net) |
| 11:08.48 | ginge | morning all |
| 11:09.22 | pH5 | morning ginge |
| 11:10.01 | exco | good morning |
| 11:12.08 | pH5 | hej exco |
| 11:12.42 | exco | hej pH5 |
| 11:12.50 | exco | :-) |
| 11:13.45 | pH5 | exco: I'll build another fso image now, will try and see if I can get at least the frameworkd running. |
| 11:14.20 | exco | any reason some htc devices are called htcxxxxx and some just xxxxx on oe |
| 11:14.38 | exco | I saw you were active building |
| 11:16.30 | pH5 | exco: different people with different naming preferences. I don't like seeing/typing/thinking/saying htcmagician at all. |
| 11:17.26 | exco | well if there comes an e.g. HP Magician ... :-) |
| 11:17.37 | pH5 | exco: too late, we're first :P |
| 11:17.44 | exco | yes |
| 11:22.07 | exco | I like the fso ui |
| 11:22.51 | pH5 | I don't care about the ui so much, I like the dbus API. |
| 11:23.12 | exco | well the looks count |
| 11:23.47 | exco | is so shallow |
| 11:24.00 | pH5 | :D |
| 11:25.03 | pH5 | looks do count though, especially if you want to attract new people |
| 11:25.07 | *** join/#htc-linux kiozen (n=oeichler@Mbdc6.m.pppool.de) |
| 11:25.13 | exco | exactly |
| 11:25.15 | pH5 | hi kiozen |
| 11:25.27 | kiozen | hi pH5 et al |
| 11:27.03 | exco | if there's an image with working gsm (not just from cmd), pm and sound -> revival for Magician? |
| 11:28.01 | pH5 | exco: you decide. as far as I'm concerned, magician is not dead (well, except the bricked one, until I can make use of your jtag discovery) |
| 11:28.51 | pH5 | gosh, I need to port suspend/resume to 2.6.26 and clean up the TS driver :/ |
| 11:28.54 | exco | I still use mine as daily phone, alarmclock and mp3player |
| 11:29.11 | exco | so to me it's very much alive |
| 11:29.57 | exco | doesn't sound like you're enjoying it too much |
| 11:30.19 | *** join/#htc-linux rob_w (n=bob@M9c58.m.pppool.de) |
| 11:32.22 | pH5 | nah, I just seem to like whining about it a lot. of course things always could advance faster.. |
| 11:34.18 | exco | I'm still trying to find my place - where I can be of most help |
| 11:42.06 | pH5 | exco: probably where you have the most fun doing it. |
| 11:42.51 | exco | the way is the goal? |
| 11:43.33 | pH5 | just write a kernel driver, fix the gsm midlayer in userspace, write an application, integrate something into the distribution, create a bit of artwork and then prepare ready-to-install images from current sources. then decide which was the most interesting part :P |
| 11:43.34 | exco | so management then :-P |
| 11:43.36 | pH5 | it sure is |
| 11:43.56 | pH5 | the way the goal I mean |
| 11:44.00 | exco | *rofl* - literally |
| 11:44.15 | pH5 | I'm not so sure about management, right now there's nobody to manage here except me :P |
| 11:44.27 | exco | "just" was good |
| 11:45.47 | exco | what kernel drivers are missing? camera and irda? |
| 11:46.13 | pH5 | camera |
| 11:46.31 | pH5 | irda should be working with the pxafixp_ir driver, but that needs to be tested |
| 11:47.18 | exco | you're making those of thin air? |
| 11:47.23 | pH5 | oh and recording support for the audio driver. right now only playback works, really |
| 11:48.22 | exco | If I had any idea ... I certainly would write you some drivers |
| 11:48.28 | pH5 | nope. I'm mostly waiting until other people do the work. the pxa camera interface is in mainline now for example, we need to write an i2c driver for the camera module and do a bit of register tracing in wince |
| 11:49.13 | pH5 | very few drivers are completely original, magician reuses a lot of known components. |
| 11:49.24 | exco | well that I could do... but I'd have to read up on how to do that first |
| 11:50.01 | exco | that's good for us |
| 11:51.03 | pH5 | it sure is. and except for the cpld, pasic3, the ov camera module and the calypso gsm chip, most of the stuff had nice documentation. |
| 11:52.19 | exco | what I find really cool is that there's gonna be a HTC Android phone before it's officially released... |
| 11:53.17 | pH5 | android won't start do interest me until the sources are in the open |
| 11:54.02 | exco | I guess that's not going to happen, is it? |
| 11:54.55 | pH5 | I wouldn't know. But I'm planning to be pleasantly surprised if it should happen. |
| 11:55.06 | exco | :-) |
| 11:56.58 | *** join/#htc-linux miknix (n=miknix@bl4-86-166.dsl.telepac.pt) |
| 11:57.47 | ginge | does anyone here know about the h2200? |
| 12:05.15 | *** join/#htc-linux dzo (n=dzo@121.98.128.127) |
| 12:06.24 | kiozen | BabelO: how is M working for you? :) |
| 12:07.21 | BabelO | not tried at the moment |
| 12:08.16 | ginge | ho dzo |
| 12:08.28 | dzo | ho ginge |
| 12:08.46 | ginge | dzo: was a typo, but ho anyway :) |
| 12:09.50 | dzo | how's it going with kaiser? |
| 12:10.15 | ginge | dzo: good. sd+sdhc working well now. |
| 12:10.38 | dzo | yes seems 4bit mode makes sdhc work. |
| 12:10.39 | Rroet | o/ |
| 12:10.46 | kiozen | BabelO: roooohhhh :( |
| 12:11.03 | ginge | just investigating wifi on the other sdio port now |
| 12:11.14 | Rroet | You guys have done a nice job in the last few days :) |
| 12:11.16 | BabelO | kiozen: booting auto install image :) |
| 12:11.18 | Rroet | Everything seems to be working. |
| 12:11.24 | ginge | dzo: how is vogue coming along? |
| 12:11.42 | Rroet | ginge: how does the sd media scanning work btw? Premy asked me to test it on my polaris.. I see a /sdcard ... now what? |
| 12:11.44 | kiozen | BabelO: your *.cab thingy? |
| 12:11.58 | BabelO | kiozen: no .exe + bz2 |
| 12:12.01 | ginge | Rroet: what initrd image are you using? |
| 12:12.08 | Rroet | Premy's |
| 12:12.25 | Rroet | http://forum.xda-developers.com/showpost.php?p=2422876&postcount=252 |
| 12:12.27 | ginge | Rroet: is that angstrom? |
| 12:12.43 | dzo | same as ever, stableish, mostly android bugs are the problem, need a new SDK, just looking at audio. |
| 12:12.43 | kiozen | ah, me playing with a Sony walkman |
| 12:12.44 | ginge | ahh android |
| 12:12.46 | Rroet | Premy put up a full packaage on the polaris thread yesterday |
| 12:13.33 | ginge | dzo: Android is a showstopper for me. Crashes all the time |
| 12:13.41 | ginge | dzo: a new SDK would be very welcome |
| 12:13.55 | ginge | Rroet: so what are you trying to do? |
| 12:14.36 | Rroet | Premy asked me if sd was working in his image I'm testing.. Yet He didn't supply any clues to how to test it |
| 12:14.49 | dzo | doesn't crash for me but silly things are annoying like if caller hangs up before you answer, same happens on emu though. |
| 12:15.01 | Rroet | I opened up the dev tools app and went for the mediascanner.. it shows something like /sdcard .. but I don't know if that' enough for him to know. |
| 12:15.16 | ginge | Rroet: can you copy and paste a file into there |
| 12:15.23 | ginge | Rroet: not used android much yet, too unstable |
| 12:15.44 | ginge | Rroet: and our touchscreen is still a bit wonky : |
| 12:15.45 | Rroet | ok, how do I get into the sdcard from the android ui? |
| 12:15.53 | dzo | are you using moko then? |
| 12:16.05 | ginge | dzo: angstrom and moko, yes |
| 12:16.39 | dzo | does that work OK? it won't do cdma though will it. |
| 12:17.06 | ginge | dzo: yes, seems ok(ish) here. Still early days for getting features to work. Not too sure about CDMA |
| 12:17.22 | Rroet | ginge: everything else seems to be working in his image for me. Dialing, sms and grps (no freezups there) |
| 12:17.36 | Rroet | gprs even works nicely with a bad reception. |
| 12:17.59 | ginge | Rroet: give me a moment, just booting into android to see about this sd |
| 12:18.59 | dzo | ginge: i'll stick to android for now, it does all i really use my phone for. |
| 12:19.16 | ginge | dzo: if it ain't broke... ;) |
| 12:19.19 | Rroet | sorry for my nubquestion, but when I start android... all I get is a spinning globe.. that's it? |
| 12:20.01 | ginge | I think so |
| 12:20.08 | Rroet | mkay |
| 12:20.21 | dzo | Rroet: thats the global time app. |
| 12:20.41 | dzo | press space to find time in different cities. |
| 12:20.49 | Rroet | ok. All I get is the spinning globe.. after a few spings it doesnt stop spinning anymore :) |
| 12:21.13 | ginge | wow. Windows Mobile reboot number 1272. This thing will wear out in no time. I already had to resolder my reset button. |
| 12:22.29 | Rroet | bad reception here again.. browser doesn't get any data. I get a action suspended notification. |
| 12:23.34 | ginge | okay, so in android. Where is this media thingy? |
| 12:23.51 | Rroet | dev tools |
| 12:23.58 | Rroet | media scanner is on the bottom |
| 12:24.13 | ginge | scanning /sdcard |
| 12:24.13 | ginge | finished |
| 12:24.41 | Rroet | All I have as well. So I guess I'll have to wait for premy then to come up with more specific guidelines. |
| 12:25.45 | ginge | Rroet: all I can tell you is this... the code is kaiser and polaris compatible (1:1) and I know it works on kaiser as I just booted android from the sd card. |
| 12:25.54 | ginge | doesn't really help you though :( |
| 12:26.30 | Rroet | how can I escape the lollypop icky sweet graphical environment back to the shell? |
| 12:26.42 | ginge | unknown at this time ;) |
| 12:26.59 | ginge | if your device give you a real ip address you could telnet in |
| 12:27.29 | ginge | or you could load an angstrom image from the kaiserbootkit |
| 12:27.32 | Rroet | tmobile.nl ... neh they don't do that :) |
| 12:27.40 | ginge | Rroet: same here |
| 12:27.45 | Rroet | dutch? |
| 12:27.52 | ginge | Orange Uk |
| 12:28.11 | Rroet | ahh. ok.. so the angstrom image.. is that on the vogue site? |
| 12:28.20 | ginge | no, its on the kaiser site... |
| 12:28.25 | Rroet | ok. |
| 12:28.33 | ginge | http://linuxtogo.org/~lgorris/kaiser-bootkit/ |
| 12:28.42 | Rroet | Is that an image with a seperate system file located on the sdcard? |
| 12:28.47 | ginge | if you grab angstrom.bin |
| 12:29.03 | ginge | its still an initrd at this time, so it doesnload load off the sd card directly |
| 12:29.17 | ginge | but it will allow you to manually mount the sd card and poke around it in |
| 12:29.43 | Rroet | angstrom.bin.. ok I'll grab that one icm with the zImage of premy.. lemme find a usb cable here to hook polly up to the laptop |
| 12:30.07 | ginge | damned android crashed :/ |
| 12:32.03 | Rroet | mine too.. I reopened the browser after a minute or 5 since last use. |
| 12:36.47 | Rroet | btw, on loading the angstrom.bin what will I get to see? gpe ? |
| 12:37.10 | ginge | Rroet: nothing but console :) |
| 12:37.27 | Rroet | ok, how'll I connect to it? hyperterminal? |
| 12:37.50 | ginge | well it will boot into a shell, so you won't need to connect to it |
| 12:38.08 | Rroet | here goes nothing. |
| 12:38.53 | ginge | so you should see the sd card detected on boot |
| 12:39.14 | ginge | if you do, you need to do a bit of manual typing to get it initted |
| 12:39.31 | ginge | mknod /dev/mmc0 b 179 1 |
| 12:39.41 | ginge | mount /dev/mmc0 /mnt/card |
| 12:41.44 | Rroet | stuck in haret I think... lemme reboot again. |
| 12:42.15 | Rroet | Does it require the zImage from your site as well, or should I work with the zImage from premy? |
| 12:43.08 | ginge | you should use premys zImage. It has the onscreen keyboard that you will nee |
| 12:43.13 | ginge | otherwise very similar |
| 12:49.12 | Rroet | tres cool, premy already posted a new package with a console initrd. |
| 12:50.31 | ginge | link? |
| 12:50.50 | *** join/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
| 12:50.55 | ginge | speak of the devil |
| 12:51.16 | premy | Hi guys. |
| 12:51.19 | ginge | hi premy |
| 12:51.59 | Rroet | Premmeh! |
| 12:52.08 | premy | Rroet: I posted on xda a link to an angstrom initrd that you can use to further test SD. |
| 12:52.28 | ginge | premy does it just have the device nodes in it, or did you go further |
| 12:52.28 | Rroet | on it |
| 12:52.48 | premy | ginge: no absolutely nothing new in there. Sorry |
| 12:53.12 | ginge | oh so you pointed to the kaiser bootkit angstrom? |
| 12:54.11 | Rroet | he did by the looks of it. |
| 12:54.16 | ginge | ok. I will ask dcordes to add the device nodes into the image to make it easier. |
| 12:54.32 | Rroet | premy.. grps should work for me too shouldn't it? I've already confirmed browsing works. |
| 12:54.54 | premy | ginge: just added some scripts to save people from struggling with the keyboard.... |
| 12:57.27 | Rroet | This initrd needs some tough love. the android one booted immediately 3 times in a row.. with the angstrom initrd attached it doesn't boot. |
| 12:58.01 | premy | Rroet: well you will need to update /sbin/gprs, /etc/ppp/chap-secret and /etc/ppp/pap-secret to reflect your operator's GPRS settings. |
| 12:58.28 | Rroet | mm.. Your angstrom initrd is different from the kaiser angstrom file.. your's only 7mb instead of 15... I'll check your angstrom image |
| 12:58.40 | ginge | premy: oh so you didn't use the ppp.sh? |
| 12:59.42 | premy | ginge: the angstrom initrd I posted is only intended to help debug a problem imfloflo from the xda forum is having. |
| 13:00.24 | premy | ginge: it's the initrd I was using when I was debugging the smd code. |
| 13:00.24 | ginge | premy: ok. btw did dcordes send you that patches for the polaris sd updates I did? |
| 13:01.38 | premy | ginge: no, but I checked out a whole new ltg repository. I had missed a udelay in dma.c. Was anything else missing ? |
| 13:02.11 | ginge | only the clock init code. Because I put the clocks in a 2d array, the loop needed he upper bound changing. Thats pretty much it |
| 13:02.57 | Rroet | ginge: I changed to premy's initrd-android (the differend angstrom image) and now it boots. |
| 13:03.16 | ginge | Rroet: our angstrom is quite large, so ymmv |
| 13:03.27 | ginge | premy: clock.c for(n=0;n |
| 13:03.32 | Rroet | time to log in |
| 13:03.52 | ginge | premy: damn. you get the idea |
| 13:04.25 | premy | Rroet: my angstrom initrd is based on a rather old one. In between, dcordes might have started trying to add some X stuff. |
| 13:04.41 | Rroet | premy: FAT: Codepage cp437 not found :) |
| 13:04.48 | Rroet | So I think it sees it. |
| 13:04.56 | ginge | premy: no X as yet in the image. |
| 13:05.41 | premy | Rroet: I'm having the same problem. Do you have a plain SD card or SDHC ? |
| 13:05.47 | diogene31 | pH5: Hi there. Just saw your conversation about the camera. It just happens I wrote the last days the camera module for the Mio, based on I2C commands. I have a lot of feedback to give to save a bit of time ... |
| 13:05.55 | ginge | premy: you need to add support in your kernel for it |
| 13:06.27 | Rroet | premy: 2GB microSD from Sandisk. I suspect regular SD |
| 13:06.32 | ginge | premy: SDHC should work now anyways |
| 13:06.50 | Rroet | I have device nodes mmcb1k0 mmcb1kop1 |
| 13:06.59 | Rroet | in /dev that is. |
| 13:07.04 | ginge | premy: CONFIG_FAT_DEFAULT_CODEPAGE=437CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1" |
| 13:07.08 | pH5 | diogene31: perfect, that's what I mean with "waiting for other people to do the work" :) |
| 13:07.19 | ginge | premy: and CONFIG_NLS_DEFAULT="iso8859-1"CONFIG_NLS_CODEPAGE_437=y |
| 13:07.24 | diogene31 | pH5: Do you know which model is in the magician ? |
| 13:07.33 | diogene31 | pH5: I mean, the camera chip model. |
| 13:07.42 | premy | ginge: oops, shame on me. Will recompile right away. |
| 13:07.45 | ginge | :) |
| 13:08.13 | Rroet | Shame on premy.... ? |
| 13:08.22 | Rroet | anyway, the device nodes I have are the correct ones ? |
| 13:08.32 | ginge | Rroet: yes, they are correct |
| 13:08.53 | Rroet | cool, it w0rsks then :P time to mod the ppp scripts |
| 13:08.53 | ginge | you can try doing fdisk /dev/mmcblk0 |
| 13:09.00 | pH5 | diogene31: it's omnivision OV9650 |
| 13:09.36 | diogene31 | pH5: Ah, too bad. Mio's is Micron MT9M111 ... you won't use mine :) |
| 13:10.25 | Rroet | mm odd, I can't do an fdisk on the mmc device. |
| 13:11.14 | ginge | error? |
| 13:12.02 | Rroet | unable to open ... but when I follow premy's instrctions from previously in the irc chat (about 4 pages up) |
| 13:12.07 | Rroet | I |
| 13:12.17 | Rroet | I "mknod /dev/mmc b 179 1" |
| 13:12.25 | Rroet | mmc0 that is |
| 13:12.37 | Rroet | I can fdisk /dev/mmc0 and it shows me a 2gb drive |
| 13:12.42 | Rroet | yes no partitions on it. |
| 13:12.46 | Rroet | should be ok |
| 13:12.53 | Rroet | s/yes/yet/ |
| 13:13.01 | diogene31 | pH5: If you have the technical specification, it takes about 5 hours to write a simple driver to get images. It integrates naturally with QTopia, so you don't have to bother with userland :) But it may take days to test through ... |
| 13:13.11 | ginge | Rroet: can you check the mmcblkxxx file with an ls -l |
| 13:13.21 | ginge | Rroet: they should be very similar to the one you just created |
| 13:14.18 | Rroet | how do I get the - sign jotted in? alt and sym don't seem to work.. |
| 13:14.37 | ginge | Rroet: not sure on polaris |
| 13:14.55 | premy | Rroet: it's on the TODO list :-( |
| 13:15.08 | Rroet | hehehe |
| 13:15.21 | ginge | premy: if its anything like my todo list, you need to write it on a toilet roll :) |
| 13:15.47 | Rroet | that smells |
| 13:16.10 | ginge | long and smelly |
| 13:17.05 | pH5 | diogene31: that's the ugly part, I think we have a 2.4 driver for ov9640 and specs for even older chips, but nothing for ov9650 itself. |
| 13:17.50 | pH5 | also ov has the habit to use a lot of magic value initialisation that isn't even documented in the spec sheets. |
| 13:18.37 | diogene31 | pH5: Ah, I see. Waiting for a shinny knight to make the driver :) Perhaps in this forum someone will raise up to the task ... |
| 13:19.33 | pH5 | might be, I'm just not interested enough in the camera to spend a lot of time there. |
| 13:20.44 | diogene31 | pH5: That's because your wife/girlfriend doesn't insist upon having her face appearing when she calls you :) |
| 13:21.08 | ginge | diogene31: mine insists on the opposite ;) |
| 13:21.16 | exco | if I find you a datasheet, diogene31 - will you be the knight in shining armor? |
| 13:22.14 | *** join/#htc-linux NetRipper (n=netrippe@netripper.nl) |
| 13:22.38 | diogene31 | exco: Nope. I have enough work with my hardware. And the person writing the driver must have a magician at hand to test. |
| 13:24.46 | diogene31 | ginge: Tell her you can take several pictures, and she can choose the one. After a dozen of snapshots, she'll be happy ... probably. |
| 13:25.32 | ginge | diogene31: I am not sure I have enough storage to cover the amount of snapshots it would need to convince her ;) |
| 13:26.13 | pH5 | :D |
| 13:26.24 | pH5 | I fear nobody will be happy with the images produced by magician's cam |
| 13:26.34 | exco | so true |
| 13:26.37 | ginge | the kaiser cam is not much better |
| 13:26.50 | ginge | if it is 2 megapixel I will eat my hat |
| 13:26.55 | Rroet | Neither is the polaris.. it's crappy slow |
| 13:27.01 | premy | ginge, Rroet: ok, configuring properly the kernel does help solving the cp437 and iso8859-1 error messages. I'll send a patch to dcordes. |
| 13:27.26 | premy | premy: I'm still getting "bogus number of reserved sectors" messages though. |
| 13:27.43 | Rroet | note to self? |
| 13:27.44 | ginge | not seen that one |
| 13:27.58 | ginge | Rroet: dual personalities |
| 13:28.33 | exco | can't we salvage the driver from the m5602 that also uses the ov9650? |
| 13:28.48 | ginge | premy: did you do the clock.c fix too? if not you will be stuck at 100khz read and write speeds |
| 13:29.45 | ginge | premy: also I don't know if you patch enabled 4 bit mode needed for sdhc |
| 13:29.58 | premy | Rroet: ;-) need some sleep, rather. |
| 13:31.28 | ginge | i'm off out for a bit. I need to scavenge some boxes for my impending house move. |
| 13:32.30 | premy | I'm off too. |
| 13:32.34 | *** part/#htc-linux premy (n=pr@73.205.76-86.rev.gaoland.net) |
| 13:33.31 | pH5 | exco: you could start with hunting for an ov9650 spec sheet (or ov9640 or ov9630, but those are known to have a slightly different register layout and featureset) |
| 13:33.46 | pH5 | exco: and then trace the i2c commands sent to the ov9650 with haret |
| 13:33.55 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 13:34.19 | exco | fine |
| 13:34.33 | diogene31 | pH5: Yep, especially register IDR (0x40300008) (set trace p2v(0x40300008) |
| 13:37.14 | pH5 | diogene31: is that for pxa25x? on pxa27x, we should look at IDBR, 0x40301688 |
| 13:37.27 | diogene31 | pH5: exco: Note that WinCE wrongly assumes that I2C base is at 0x40300000 while is really is 0x40301680. I didn't found out how it works, but have not investigated much anyway, since after traces and disassembly I got the chip model => specification => easy coding. |
| 13:38.04 | *** join/#htc-linux Marex (n=vasum7am@auryn.karlin.mff.cuni.cz) |
| 13:38.41 | pH5 | diogene31: sounds like you have some strange mapping there. I used "set trace 0xae301688; wi 10" to trace i2c traffic in haret |
| 13:38.54 | diogene31 | pH5: That was for pxa27x. And yes, I mispelled IDBR (and not IDR of course). But you really should consider 0x40300008 instead of 0x40301688 for tracing, as this is the address my WinCE used. |
| 13:39.18 | cr2 | hi |
| 13:39.21 | pH5 | hi cr2 |
| 13:39.34 | Marex | pH5, cr2: hi ;) |
| 13:39.51 | diogene31 | pH5: Yes, sure. Did you see that thread in haret ML about MMUMerge ? It applies to the Mio A701. So I can't rely on mmutrace in haret :( |
| 13:39.53 | cr2 | diogene31: htc hardcodes 0x40301680 |
| 13:40.01 | diogene31 | pH5: cr2: Hi |
| 13:40.24 | diogene31 | cr2: Then HTC developpers were brighter than mio's ones. |
| 13:40.31 | pH5 | hi Marex |
| 13:40.32 | cr2 | lol |
| 13:40.49 | pH5 | :) cr2 has seen too much |
| 13:42.40 | cr2 | yeah. |
| 13:43.01 | cr2 | hmm. even garmin made pxa_camera work... |
| 13:44.30 | kiozen | lol |
| 13:44.36 | kiozen | even garmin .... :P |
| 13:46.56 | exco | pH5: http://www2.hawaii.edu/~toyl/CubeSAT/docs/OV9650_DS%20(1.91).pdf page 12 what we need? |
| 13:48.24 | pH5 | exco: heh, perfect |
| 13:48.49 | miknix | BabelO, hey |
| 13:49.00 | pH5 | "NOTE: All other registers are factory-reserved. Please contact OmniVision Technologies for reference register settings." <-- that's what we'll need the traces for |
| 13:49.16 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 13:52.56 | cr2 | diogene31: a780 kernel had the 9650 driver (2.4, afair by intel), but it was a complete mess |
| 13:53.10 | cr2 | diogene31: btw, the 9650 is used on many other devices. |
| 13:54.03 | diogene31 | cr2: The new soc_camera framework is quite clean now. Integration should be pretty straightforward. |
| 13:54.12 | cr2 | kiozen: the SSCR data you've sent me many years ago are very strange. can you pick haret from http://jornada820.sf.net/files/haret-w.exe and do 'dump pxa27x' ? |
| 13:54.42 | cr2 | diogene31: hehe. that will make universal support all the hardware :) |
| 13:55.23 | cr2 | diogene31: no, there is a 6650 there too. with some funny omap driver. |
| 13:55.27 | diogene31 | cr2: Hah, as the MIO :) The only leftover is the Flash DoC G3 chip, but I didn't give up yet. |
| 13:55.42 | kiozen | lol, many years ago :) hm, no empty SD available ... how much space will it need? |
| 13:55.44 | cr2 | diogene31: yes. mdoc |
| 13:55.57 | cr2 | kiozen: 1K ? |
| 13:56.16 | kiozen | ok, I can do that with my golden master :) |
| 13:56.38 | cr2 | diogene31: does the docg3.c work for reading from mdoc ? |
| 13:57.22 | diogene31 | cr2: I don't know that docg3.c. I use a test module of mine to read mdoc sectors, and partition info. |
| 13:58.03 | kiozen | cr2: got a 404 |
| 13:58.09 | cr2 | diogene31: ok. |
| 13:58.21 | cr2 | http://jornada820.sf.net/files/haret/haret-w.exe |
| 14:00.30 | miknix | cr2, do you know where the TI wifi devices are connected on the omap850? |
| 14:01.09 | miknix | cr2, there is a driver for wizard's wifi device but is using pci |
| 14:04.47 | cr2 | miknix: i don't know where is it mapped. |
| 14:05.24 | diogene31 | cr2: BTW, if you have that docg3.c at hand, would you send it to me please for crosscheking ? |
| 14:05.42 | cr2 | miknix: you can't really use this pci driver. you should start with the acx100 driver in hh.org CVS. it's known to work more or less stable at 1Mb |
| 14:06.08 | cr2 | diogene31: it's in the hh.org CVS -> htcuniversal/docg3.c |
| 14:09.24 | miknix | cr2, ok thanks.. I was expecting to find some useful information on the pci driver |
| 14:10.52 | diogene31 | cr2: With correct CVSROOT, "cvs co htcuniversal" fails. Is htcuniversal a valid alias ? |
| 14:11.53 | kiozen | cr2: http://pastebin.no/11847 |
| 14:13.39 | cr2 | diogene31: http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-pxa/htcuniversal/docg3.c |
| 14:13.54 | cr2 | kiozen: thanks. |
| 14:14.24 | cr2 | SSCR0_P1 : 0x0000818e = 00000000 00000000 10000001 10001110 |
| 14:14.27 | cr2 | hmm. |
| 14:14.42 | kiozen | cr2: welcome, if that battery stuff is working, I will do a rootfs release of M |
| 14:14.49 | miknix | the acx100 looks like a TNETW1450.. wizard has an TNETW1350. I'll fork the driver |
| 14:16.00 | cr2 | miknix: 1250 |
| 14:16.48 | cr2 | miknix: you should find the location of the registers, and perform the right sdio init sequence. |
| 14:17.11 | *** join/#htc-linux mads- (n=madsravn@0x5733e5ca.aaanxx4.dynamic.dsl.tele.dk) |
| 14:17.13 | cr2 | miknix: there are 10 devices that use 1250 or 1251. |
| 14:17.25 | cr2 | miknix: the msm* included. |
| 14:17.50 | mads- | Hi. I'm thinking about buying a new phone, and I thought it could be fun to get a phone with linux on it. Is there a table over which HTC phones that supports linux...? |
| 14:17.59 | diogene31 | cr2: Thanks. |
| 14:18.27 | miknix | cr2, you are right. the wizard wiki is wrong. I just looked to the TRM, its an TNETW1230 |
| 14:18.49 | cr2 | mads-: it's all work in progress. |
| 14:19.20 | miknix | http://acx100.sourceforge.net/matrix.html |
| 14:19.24 | cr2 | miknix: 1250. trm is not an authoritative source on external devices. |
| 14:19.28 | mads- | cr2 : damn :( I thought there was something I could use :) |
| 14:19.59 | miknix | hum.. ok. fixing the wiki |
| 14:21.30 | mads- | You know of any phone where linux runs good? |
| 14:21.51 | *** join/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
| 14:31.04 | diogene31 | is away: Away |
| 14:43.01 | *** part/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
| 15:03.17 | *** join/#htc-linux JEEB (n=kanakana@a88-114-192-27.elisa-laajakaista.fi) |
| 15:06.14 | *** join/#htc-linux dwaradzyn (n=dwaradzy@damian8155.net.autocom.pl) |
| 15:09.38 | *** join/#htc-linux pH5_ (n=ph5@p5485BECC.dip.t-dialin.net) |
| 15:09.53 | *** join/#htc-linux AstainZZZZZZ (n=Administ@unaffiliated/astainhellbring) |
| 15:10.52 | pH5 | ecxo_grabs_a_bit: raw haret I2C log of turning on the camera http://en.pastebin.ca/1076691 |
| 15:11.44 | exco | just fetched my Magician and a cable |
| 15:26.42 | exco_ | bbl |
| 15:56.02 | *** join/#htc-linux Othello_ (i=Magorium@gateway/tor/x-08e58dfd67571688) |
| 16:03.14 | *** join/#htc-linux resuba (i=relisys@ool-18bed275.dyn.optonline.net) |
| 16:04.04 | resuba | hi |
| 16:04.06 | resuba | <PROTECTED> |
| 16:09.32 | marajin | resuba: http://www.htcclassaction.org |
| 16:10.02 | marajin | resuba: In particular, http://www.htcclassaction.org/driverprogress.php |
| 16:11.03 | marajin | resuba: Note: The Tilt is the AT&T rebranding of the HTC TyTn II, so it is talking about the video drivers you want |
| 16:11.10 | kaboofa | resuba: in short, d3d is kinda working, video/windows stuff is still kinda D: |
| 16:12.03 | kaboofa | resuba: you might want to try flashing hyperdragoniii rom, i don't know if it's in my head or not, but it seems way faster than the others i have tried |
| 16:13.29 | ginge | marajin: what network you on? |
| 16:13.41 | marajin | ginge: O2? as well you know |
| 16:13.47 | ginge | marajin: I forget |
| 16:14.04 | marajin | opens Ginge's head and installs some extra RAM. |
| 16:14.26 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbfc87b.pool.einsundeins.de) |
| 16:14.27 | ginge | lol |
| 16:14.38 | diogene31 | is back |
| 16:14.38 | kaboofa | might want to slap in another hard disk too -- as soon as he falls asleep ram is purged! |
| 16:14.43 | marajin | hah |
| 16:14.46 | marajin | no |
| 16:14.51 | marajin | we just keep him up 24/7 |
| 16:14.51 | kaboofa | ;) |
| 16:14.55 | marajin | installs the caffeine IV drip. |
| 16:15.08 | kaboofa | then cpu performance will suffer after about 2 days D: |
| 16:15.13 | ginge | marajin: just building a new android image with hopefully working sms and data without a bodge job of manual ppp |
| 16:15.24 | pH5 | exco_: the same haret log piped through this (http://en.pastebin.ca/1076745) script: http://en.pastebin.ca/1076748 |
| 16:15.35 | ginge | marajin: could do with the drip. Sleeping is such a waste |
| 16:15.49 | marajin | ginge: Why oh why do you keep going back to android? It's clearly borkened badly |
| 16:16.09 | ginge | marajin: is there a viable om or gpe yet? |
| 16:16.21 | marajin | ginge: Nah, X is half working still |
| 16:16.29 | ginge | marajin: thats why :) |
| 16:16.40 | marajin | ginge: Yeah but android crashes every 5 minutes :P |
| 16:16.44 | ginge | marajin: I will drop it as soon as we have a good os |
| 16:17.04 | marajin | ginge: And without PM it's not a viable phone under any other OS anyway yet :P |
| 16:17.32 | ginge | marajin: I got a whole day out of it on friday :) if you dont use data, android for the phone works ok |
| 16:18.08 | marajin | ginge: You're mad, do you understand this? MAD! :P |
| 16:18.27 | ginge | so i'm told.... often. |
| 16:18.35 | marajin | Its the hair, we can't help it |
| 16:18.45 | ginge | lol |
| 16:20.17 | *** join/#htc-linux kiozen (n=oeichler@Mbdc6.m.pppool.de) |
| 16:22.08 | marajin | ginge: Still, why's my network relevant? you have an app to change the settings after boot or are we just going to pass them the kernel still? |
| 16:22.25 | ginge | I am writing a switcher app |
| 16:22.35 | marajin | fairy nuff |
| 16:22.50 | marajin | mobile.o2.co.uk mobileweb/password as far as I remember |
| 16:23.08 | ginge | just want the testers to share the punishment |
| 16:23.10 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 16:23.21 | ginge | good day to you mr dcordes |
| 16:23.24 | marajin | ginge: speak of the victim and he shall surely appear ;) |
| 16:23.31 | ginge | as if by magic |
| 16:23.38 | marajin | yep |
| 16:23.43 | Marex | pH5: hi, how's the serial_power patch going ? |
| 16:23.46 | marajin | hides the cattleprod behind his back. |
| 16:23.46 | dcordes | good day |
| 16:23.54 | Marex | pH5: any idea when can we expect it in mainline ? |
| 16:24.00 | ginge | ahh, the shopkeeper appeared. |
| 16:24.06 | Rroet | Do I have to blow up my polly already? |
| 16:24.17 | ginge | dcordes: did you see the spreadsheet is not public? |
| 16:24.53 | dcordes | yes I updated the link |
| 16:25.01 | ginge | ok |
| 16:25.15 | dcordes | I have seen I messed something up with clock.c while commiting? |
| 16:25.39 | ginge | you did? |
| 16:25.53 | dcordes | or something was missing or so |
| 16:26.36 | ginge | premy missed them from his patch, but my review caught them before you got them |
| 16:26.38 | ginge | so its ok |
| 16:26.44 | ginge | yep, gitweb confirms its ok |
| 16:27.51 | pH5 | Marex: no idea. not until dmitry (or somebody else) tries to push it again. I guess that's post-merge stuff |
| 16:28.07 | dcordes | ginge: don't you think it would really be better if you would commit the stuff directly from your working tree? |
| 16:28.17 | dcordes | also premy |
| 16:28.26 | ginge | I would break more than I fix because I know nothing about git |
| 16:29.01 | dcordes | I mean I have no problem at all and I'm sure OK with keeping it that way in the future |
| 16:29.20 | cr2 | kiozen: http://tile.openstreetmap.org/tile-downloads-week-of-2008-07-01.png |
| 16:29.36 | dcordes | but it would probably be better if the actual authors would check in the changes |
| 16:29.49 | ginge | true. I guess I should learn |
| 16:29.58 | kiozen | cr2: cool |
| 16:30.34 | Marex | pH5: damned ... even simple gsm module for hw6515d depends on it ... |
| 16:30.53 | Marex | looks like we are pretty stuffed without it |
| 16:32.15 | cr2 | kiozen: i don't understand why the SPI is set for 15bit transfers in wince. |
| 16:32.25 | dcordes | ginge: you just check for git-status what needs to be git-added, then with git-commit file1 file2 file3 put everything that goes into the first commit, $EDITOR will open and ask for a description. git-commit file3 file4 etc.. then git push to push through your account |
| 16:32.41 | ginge | hmm ok. |
| 16:32.47 | dcordes | and that's about it |
| 16:33.16 | dcordes | just need to register for an account at linuxtogo or you can use mine |
| 16:33.31 | kiozen | cr2: because it can be used like this? |
| 16:33.47 | dcordes | or I can of course keep commiting everything. no problem |
| 16:33.48 | kiozen | cr2: you are sure it is? |
| 16:34.15 | cr2 | look at the SSCR0_P1 bits |
| 16:35.30 | cr2 | page 481 |
| 16:35.51 | ginge | dcordes: what network you on? |
| 16:36.08 | dcordes | ginge: vodafone |
| 16:36.28 | cr2 | kiozen: sorry, it's 485 for the speed table |
| 16:39.43 | ginge | marajin: dcordes: catch you laters |
| 16:39.47 | dcordes | ginge: (germany) |
| 16:40.01 | dcordes | cya ginge |
| 16:40.05 | marajin | catcha man |
| 16:40.54 | kiozen | cr2: ok, I would say the same |
| 16:41.52 | cr2 | kiozen: on universal et al. it was 24bit. strange... |
| 16:42.38 | cr2 | kiozen: bizarre machine :) |
| 16:42.43 | cr2 | bbl |
| 16:43.42 | kiozen | ok |
| 17:01.47 | exco_ | pH5: did you use telnet or are you using haretconsole? |
| 17:08.01 | exco | so haretconsole it is |
| 17:11.57 | dwaradzyn | dcordes: i've tested the kernel you have built from merged sources (zImage-polaris_test.bin) and it works just fine (call, sms, data connection) |
| 17:18.44 | *** join/#htc-linux surgex (n=surge@70.57.72.6) |
| 17:22.52 | *** join/#htc-linux SmallR2002 (n=SmallR20@c-67-162-60-33.hsd1.il.comcast.net) |
| 17:32.29 | *** join/#htc-linux gudz666 (n=gudz666@damian8155.net.autocom.pl) |
| 17:34.00 | Marex | cr2: Im going to prepare mainline patch for hw6515d |
| 17:34.18 | Marex | but Im missing that damned pxa-serial-power patch :/ |
| 17:34.36 | *** join/#htc-linux miknix (n=miknix@bl4-86-166.dsl.telepac.pt) |
| 17:34.47 | miknix | nickserv identify 808189alabama! |
| 17:34.53 | miknix | shit |
| 17:34.55 | miknix | : D |
| 17:35.07 | exco | :-) |
| 17:36.01 | exco | I really hope for you that's not a standardpw with logging ... |
| 17:36.27 | miknix | hehe |
| 17:36.32 | miknix | I dont trust IRC sniffers |
| 17:39.12 | miknix | this '/' needs to be hammered |
| 17:39.15 | Marex | su -c "rm -rf" ;) |
| 17:39.26 | miknix | the key is too hard |
| 17:39.28 | exco | yeeehaaa |
| 17:42.55 | Marex | pH5: is there already working pxa-qc interface in mainline ? |
| 17:43.45 | Marex | oh :) pxa_camera looks nice |
| 17:53.51 | *** join/#htc-linux OpnSrc1 (i=harisha1@gateway/tor/x-ffe244f68f5a53a1) |
| 18:40.31 | *** join/#htc-linux LunohoD_ (n=alex@e180069186.adsl.alicedsl.de) |
| 18:49.53 | cr2 | Marex: asic3-mmc ? |
| 19:01.18 | dwaradzyn | how can i make "-" with android green kernel-level keyboard? |
| 19:04.03 | ginge | dwaradzyn: polaris? |
| 19:05.20 | dwaradzyn | yes |
| 19:05.37 | ginge | I believe it is on premy's todo list |
| 19:05.47 | dwaradzyn | ok :) |
| 19:06.19 | dwaradzyn | i'm able to mount sd under polaris, i just saw my files in angstrom |
| 19:12.09 | Rroet | o/ |
| 19:12.27 | tmzt | dcordes: ping? |
| 19:12.31 | Rroet | pong |
| 19:12.40 | Rroet | right... |
| 19:21.28 | marajin | ginge: wb there |
| 19:21.41 | ginge | marajin: lo |
| 19:27.42 | Marex | cr2: what about it |
| 19:28.10 | Marex | (asic3-mmc) |
| 19:33.00 | cr2 | Marex: is it working in mainline ? |
| 19:38.57 | *** join/#htc-linux tsdogs (n=tsdogs@62.13.169.86) |
| 19:41.29 | marajin | ginge: Did we get android booting from SD ? |
| 19:44.30 | Marex | cr2: looks like we are missing tmio mmc there |
| 19:53.09 | *** join/#htc-linux BabelO_ (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 20:11.21 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 20:14.35 | cr2 | Marex: tmio mmc is only for some strange palms :) |
| 20:15.58 | dcordes | ~ping tmzt |
| 20:15.59 | apt | pong tmzt |
| 20:16.08 | tmzt | hi |
| 20:16.15 | dcordes | hola ginge |
| 20:16.17 | tmzt | do you have the bb for tslib? |
| 20:16.24 | dcordes | sure |
| 20:16.53 | tmzt | can you paste it? (rafb) |
| 20:17.30 | dcordes | tmzt: http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=tree;f=packages;h=9610ccacc04717497ee8bc5ecf6570bce9fa6f9e;hb=refs/heads/org.openembedded.dev tslib/tslib_1.0.bb |
| 20:17.49 | dcordes | http://rafb.net/p/n4Ge9735.html |
| 20:18.36 | tmzt | I think we are going to have to find out from moko people what the multievent patch does. |
| 20:19.14 | pH5 | cr2: AIUI the mmc core in asic3 is basically the same as in tmio, just clock handling and register spacing is different. |
| 20:19.29 | pH5 | there's a note about this in asic3_mmc.c |
| 20:20.01 | cr2 | pH5: tmio looks as a primitive driver compared to asic3-mcc |
| 20:20.42 | cr2 | pH5: and the atiw228x is also very similar, but still there any many annoying differences. |
| 20:21.40 | cr2 | pH5: and then there is the sdio problem. |
| 20:22.01 | pH5 | right, this has yet to be solved |
| 20:22.25 | cr2 | pH5: atiw228x uses only one register block for sd/sdio. asic3-mmc has 2 separate (+the control block) |
| 20:23.25 | cr2 | so it's a bit of wishful thinking to write 1 driver for all those. |
| 20:23.44 | tmzt | http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob;f=packages/tslib/tslib/multievent.patch;hb=HEAD |
| 20:25.08 | cr2 | pH5: do you think we should use pxa_serial_power or standardize the external /sys interface for power ? |
| 20:27.43 | dcordes | ginge: which android image shall we pick for the sd loop image? |
| 20:28.16 | Marex | well it looks like Im still missing a few things while registering asic3 in mainline |
| 20:28.22 | Marex | it dies with crashdump |
| 20:41.52 | *** join/#htc-linux TimRiker (n=timr@rikers.org) |
| 21:02.08 | Marex | pH5: ping |
| 21:11.08 | *** join/#htc-linux AstainHellbring (n=Administ@unaffiliated/astainhellbring) |
| 21:17.58 | pH5 | Marex: pong |
| 21:18.48 | pH5 | cr2: my current problem with serial pxa power is that it doesn't seem to make a difference between power down due to closing of the tty and due to suspend. |
| 21:19.59 | Marex | pH5: looks like asic3 has some issues with irqs |
| 21:23.53 | pH5 | Marex: this is a bit vague... there might well be a few issues. I think the mainline code only booted on universal, hx4700 and blueangel so far. |
| 21:24.19 | Marex | asic3.c:316 ... it dies there |
| 21:24.34 | Marex | at least for me |
| 21:24.44 | Marex | do you have those patches for mainline universal somewhere ? |
| 21:28.39 | ginge | hola |
| 21:32.23 | pH5 | Marex: I haven't seen any universal/mainline code yet, ask samuel about that. |
| 21:32.43 | pH5 | the blueangel code is in my tree dump on git.linuxtogo.org |
| 21:42.25 | Marex | pH5: magician uses asic3 as well ? |
| 21:42.37 | Marex | doesnt it have some weird pasic3 ? |
| 21:44.27 | pH5 | magician doesn't use asic3, it's a very simple platform |
| 21:45.38 | pH5 | pasic3 seems to be the successor of pasic2 in blueangel, it is much smaller than asic3 and only contains led and ds1wm parts |
| 21:45.59 | pH5 | no mmc, no configurable gpios, no spi |
| 21:48.55 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) |
| 21:53.47 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 21:54.01 | pH5 | good night |
| 21:54.51 | tmzt | did you get that patch? |
| 21:57.09 | dcordes | tmzt: yes already seen it yesterday but I have no idea what it does |
| 21:57.16 | dcordes | and what the problem about multievent thing is |
| 21:58.08 | ginge | dcordes: yo |
| 21:58.14 | tmzt | it seems that openmoko uses it, at least in one bb |
| 21:58.19 | ginge | dcordes: I was reading up about tslib |
| 21:59.01 | ginge | apparently the hp h2200 has a non-linear touchscreen and tslib requires a kernel level correction algorithm |
| 21:59.14 | ginge | I couldn't find this supposed source, but could be a lead |
| 21:59.31 | tmzt | I thnk the modules are supposed to correct that |
| 22:01.09 | tmzt | the ts_cal crashing has to be two many events, too quickly |
| 22:01.17 | tmzt | s/two/too |
| 22:03.32 | ginge | those wifi chipsets are power hungry little bitches. That and your battery is clearly fubar |
| 22:03.45 | ginge | sorry wrong window |
| 22:04.02 | tmzt | relevant enough here |
| 22:04.32 | ginge | someone with a wizard that shuts off when wifi is on. |
| 22:05.07 | marajin | heh |
| 22:05.36 | dcordes | ginge: http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-pxa/h2200/ |
| 22:05.46 | dcordes | but no clue what ts it uses |
| 22:06.20 | ginge | all the code is in drivers/input/touchscreen in that tree. I read all the source and nothing fancy going on there. |
| 22:06.27 | ginge | Must be in tslib |
| 22:07.48 | tmzt | there's tsc2200, maybe similar to 2003/2007? |
| 22:08.35 | ginge | its not the chipset thats the problem, its what the touschreen itself outputs is wrong/different |
| 22:09.08 | tmzt | do we have a kaiser ts with out onscreen keyboard? |
| 22:09.20 | tmzt | it looks like osk outputs all events on penup |
| 22:09.21 | ginge | its in the tree |
| 22:09.47 | ginge | tmzt: it reports all events as they happen |
| 22:09.59 | ginge | but there could be a delay sending penup |
| 22:11.24 | ginge | why not compile up a stock rom? |
| 22:11.25 | tmzt | we need a simple evtest with only Abs and Touch (key event) |
| 22:11.45 | tmzt | and to touch the stylus to the ts only once |
| 22:12.00 | tmzt | hopefully getting one X, one Y, and a Touch |
| 22:12.33 | ginge | but touching generates a new position all the time the pen is down |
| 22:12.36 | goxboxlive | BabelO_: awake? |
| 22:12.47 | BabelO_ | goxboxlive: yes |
| 22:12.48 | tmzt | I think dejitter should handle that |
| 22:13.02 | goxboxlive | BabelO_: i got my iPhone this morning. |
| 22:13.05 | tmzt | But it might need the multievents patch |
| 22:13.22 | ginge | tmzt: sure you can filter for it if you are not moving the pen easily enough, but is that right? |
| 22:13.29 | goxboxlive | BabelO_: The menus is great, though it misses a lot of apps, like we have in wince or linux |
| 22:13.42 | ginge | tmzt: doesnt tslib do that |
| 22:13.47 | goxboxlive | gps and some wifi apps |
| 22:13.54 | tmzt | how do you get to htc-msm head on gitweb? |
| 22:14.02 | BabelO_ | goxboxlive: like ? |
| 22:14.26 | BabelO_ | goxboxlive: ok, for gps i ve read that already |
| 22:14.28 | goxboxlive | BabelO_: TomTom, Wififofum |
| 22:14.30 | goxboxlive | and so |
| 22:14.43 | BabelO_ | goxboxlive: so what you can do with gps ? |
| 22:15.06 | goxboxlive | BabelO_: well you can use google maps. |
| 22:15.27 | tmzt | dejitter is in tslib as a module, multievents changes it to read I think 64 events and average |
| 22:15.31 | goxboxlive | It works great, but it's expensive. |
| 22:15.34 | ginge | tmzt: http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=tree;h=refs/heads/htc-msm;hb=refs/heads/htc-msm |
| 22:15.36 | tmzt | up to 64 |
| 22:15.40 | BabelO_ | goxboxlive: ok |
| 22:15.57 | goxboxlive | BabelO_: and i cant use it with linux yet |
| 22:16.05 | goxboxlive | sync with amarok |
| 22:16.29 | goxboxlive | ANd neither use it with my icelink in my car. Thats a shame, because iphone 1th generation did work |
| 22:17.12 | goxboxlive | But websurfing, byuing music and such is outstanding |
| 22:17.20 | BabelO_ | goxboxlive: :' too much problem |
| 22:17.41 | goxboxlive | :-) |
| 22:18.44 | BabelO_ | goxboxlive: so it run fast ? |
| 22:18.55 | BabelO_ | goxboxlive: video as well? |
| 22:19.31 | goxboxlive | yes, it's fast, Yes, i havent tried uploading vidoe to it, just running from youtube via wifi and it works great. |
| 22:20.04 | BabelO_ | good :) |
| 22:20.18 | BabelO_ | so waiting for tomtom then :( |
| 22:21.24 | tmzt | looks like that has calibration in the kernel |
| 22:23.23 | tmzt | maybe not |
| 22:23.34 | tmzt | can you build zImage with htc-msm tsc2003? |
| 22:23.49 | cr2 | goxboxlive: lol. [00:15] <goxboxlive> It works great, but it's expensive. |
| 22:24.05 | cr2 | goxboxlive: have you tried roadmap on ephone ? |
| 22:25.10 | goxboxlive | cr2: You are not able to install apps that are not certified by apple. A hack is onit's way, but it'snot released yet :-( |
| 22:26.05 | tmzt | isn't it just baseband unlock that's not released? |
| 22:26.25 | ginge | tmzt: yes, I have one pre-built here:http://www.headfuzz.co.uk/files/android/zImage |
| 22:26.48 | tmzt | dcordes: can you try that zImage from ginge? |
| 22:26.57 | goxboxlive | tmzt: no, not last time i checked |
| 22:27.00 | ginge | that does the android conversion, not raw |
| 22:27.04 | tmzt | same userland |
| 22:27.09 | dcordes | k |
| 22:27.12 | tmzt | Pwned? or something |
| 22:27.33 | cr2 | goxboxlive: ok. |
| 22:31.38 | dcordes | oh that hang in haret. it didn't freeze in quite a while now |
| 22:33.27 | ginge | dcordes: how do I update my local git to googles .25 source? |
| 22:33.40 | dcordes | good question |
| 22:34.03 | dcordes | best would be to git-diff the two |
| 22:34.19 | ginge | shifts uncomfortably |
| 22:35.43 | tmzt | is android evdev? |
| 22:37.30 | ginge | I guess it will be, probably via usb? |
| 22:39.00 | dcordes | ginge: I can't boot with the posted zImage |
| 22:39.14 | dcordes | it freezes in haret while fb takes over |
| 22:39.20 | ginge | hmm. you booting off the sd? |
| 22:39.25 | dcordes | yep |
| 22:39.29 | ginge | one sec |
| 22:39.33 | dcordes | I just e2fscked |
| 22:40.25 | ginge | try now |
| 22:41.02 | dcordes | ok |
| 22:43.07 | dcordes | ginge: booting. |
| 22:44.38 | dcordes | tmzt: ok logged in |
| 22:44.42 | tmzt | cool |
| 22:44.57 | tmzt | can you try ts_calibrate |
| 22:46.37 | dcordes | TSLIB_TSDEVICE=/dev/input/event0 ts_calibrate |
| 22:46.46 | tmzt | yes, sry |
| 22:46.52 | dcordes | like that? |
| 22:46.52 | dcordes | ok |
| 22:47.18 | dcordes | I can tap as much as I want and it won't quit now. looks like it doesn't recognize the pendown |
| 22:47.32 | dcordes | I should have probed for any data before |
| 22:47.55 | dcordes | is there any way to quit? |
| 22:48.03 | tmzt | probably not |
| 22:48.10 | tmzt | I never found one |
| 22:48.15 | dcordes | ok booting again |
| 22:48.57 | tmzt | I guess I should have had you do cat /sys/class/input/*/name |
| 22:48.57 | tmzt | just to make sure |
| 22:48.57 | ginge | would you like to prepare an image based on git that reports raw events? |
| 22:48.57 | tmzt | and evtest |
| 22:50.47 | dcordes | portmap is taking ages to start |
| 22:51.08 | tmzt | any network devices up? |
| 22:51.28 | dcordes | yes loopback is there when I logged in |
| 22:51.33 | dcordes | it's just taking long |
| 22:52.01 | dcordes | ok I'm in evtest /dev/input/event0 now |
| 22:52.24 | dcordes | the coords look way more sane now |
| 22:52.46 | tmzt | do you get touch events as key events? |
| 22:52.51 | dcordes | I also get the pendown key and wonder why ts_calibrate didn't react now |
| 22:53.09 | dcordes | type 1 |
| 22:53.14 | tmzt | i think it wants Touch |
| 22:53.22 | dcordes | type 1 (Key), code 330 (touch), value 0 |
| 22:53.27 | dcordes | is penup |
| 22:53.44 | dcordes | type 1 (Key), code 330 (touch), value 1 down |
| 22:53.59 | tmzt | that makes sense |
| 22:54.06 | dcordes | the coordes are type 3 Absolute values. they are between 0 and about 250 |
| 22:54.33 | tmzt | yeah, I saw it did some of that in the driver |
| 22:54.46 | tmzt | not sure if it has to be calibrated in the kernel though |
| 22:54.52 | dcordes | ok in the corners they get negative. the values look very similar in an android touchscreen test application |
| 22:55.04 | tmzt | do you have /etc/pointercal |
| 22:55.12 | ginge | http://www.headfuzz.co.uk/files/android/zImage-tsc2003raw with the 0-320/240 correction |
| 22:55.12 | ginge | http://www.headfuzz.co.uk/files/android/zImage-tsc2003-corr-raw with just the y correction |
| 22:55.43 | tmzt | kaiser is qvga? |
| 22:55.46 | ginge | the first will report all raw events |
| 22:55.46 | ginge | the second one will report raw events with a sane y |
| 22:56.01 | dcordes | tmzt: yes |
| 22:56.12 | tmzt | yes on pointercal |
| 22:56.14 | dcordes | kaiser is qvga and I don't have pointercal |
| 22:56.44 | dcordes | I wonder what changed that makes ts_calibrate not recognize the 'pendowns' or whatever is the right term. |
| 22:56.52 | tmzt | ts_calibrate did not exit, it did not move the targets when you touch the screen? |
| 22:56.58 | dcordes | right |
| 22:57.09 | dcordes | yesterday it didn't move the targets but exited after 5 taps |
| 22:57.29 | dcordes | <PROTECTED> |
| 22:57.44 | ginge | I think its confused by our screen |
| 22:57.54 | ginge | can you try with the first and secod links I pasted? |
| 22:57.56 | tmzt | can you try ts_print_raw, ts_print |
| 22:58.04 | tmzt | they should exit on ctrl-c |
| 22:58.22 | dcordes | wait, does fbcon rotation 3 constitute a problem? |
| 22:58.25 | tmzt | yeah, the raw zImage might work better |
| 22:58.46 | dcordes | tmzt: ok need to install that into the image. 4 minutes |
| 22:58.57 | tmzt | probably, but the coords should match the phys orientation |
| 22:59.02 | ginge | rotate sounds like a bad idea for now |
| 22:59.14 | tmzt | type sideways -) |
| 22:59.17 | dcordes | removing it |
| 22:59.22 | ginge | I do |
| 22:59.38 | dcordes | hm? |
| 22:59.40 | ginge | its not like I do much typing locally |
| 23:00.00 | tmzt | I had a 240x320 on one until I found fbcon rotate |
| 23:00.04 | tmzt | that was fun |
| 23:04.01 | dcordes | booting |
| 23:05.22 | ginge | I think what we are looing for are values int he range of 0-4095 but realistically more like 1800-3500 or so |
| 23:05.33 | ginge | I think thats what tslib wants |
| 23:05.40 | dcordes | I had that yesterday in the rawts image |
| 23:05.47 | dcordes | such sort of values |
| 23:05.49 | ginge | looking at how everyone else does it anyway |
| 23:05.55 | tmzt | the calibrate is supposed to figure out the range |
| 23:06.09 | tmzt | it just is not working |
| 23:06.15 | ginge | I also think it expects y to be sane, which is a problem |
| 23:07.02 | tmzt | the options are ts_calibrate or xtscal, which uses an X extension in kdrive, I think |
| 23:07.28 | tmzt | at least poky uses it, but it never worked in qemu so I disabled it |
| 23:07.31 | dcordes | ok got tslib-test now |
| 23:07.49 | tmzt | TSLIB_TSDEVICE=/dev/input/event0 ts_print_raw |
| 23:09.51 | dcordes | Selected device is not a touchscreen I understand |
| 23:10.18 | tmzt | I think it probes something, not sure what it means |
| 23:10.38 | tmzt | there are some IOCTLs to find out stuff about the driver |
| 23:10.51 | dcordes | ts_print understands it the same way |
| 23:12.11 | dcordes | tmzt: why does evtest print coordinates and stuff while ts_print doesn't recognize it as a ts? |
| 23:12.31 | tmzt | I don't have the patched source, trying to recreate it |
| 23:13.04 | dcordes | tmzt: which? |
| 23:13.18 | tmzt | the bb you gave me, and the urls it downloads |
| 23:14.08 | dcordes | I can email you a tar of the patched tslib OE source |
| 23:14.19 | tmzt | allright, you have my email? |
| 23:14.25 | dcordes | yes |
| 23:14.29 | tmzt | ok |
| 23:19.56 | tmzt | it's not here |
| 23:22.13 | dcordes | tmzt: I only just sent it. had to clean and refetch/patch the source |
| 23:22.17 | dcordes | should be there |
| 23:22.29 | tmzt | yeah, just using mutt and wasn't sure |
| 23:28.06 | tmzt | does evtest give you the event flags at the top (don't press anything)? |
| 23:28.17 | tmzt | yeah, got it |
| 23:31.01 | dcordes | tmzt: Input driver, device etc? |
| 23:31.07 | dcordes | brb |
| 23:31.18 | tmzt | evtest /dev/input/event0 |
| 23:34.59 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
| 23:36.05 | tmzt | if (i->sane_fd == -1) |
| 23:36.47 | tmzt | it doesn't return -1 because that confuses kdrive, but it still fails |
| 23:37.11 | tmzt | plugins/input-raw.c |
| 23:38.12 | tmzt | ABS_X, ABS_Y, ABS_PRESSURE have to be set |
| 23:45.29 | dcordes | tmzt: don't know if that's relevant but in evdev pressure is not shown anymore |
| 23:45.41 | tmzt | in evtest? |
| 23:46.09 | dcordes | evtest, sorry |
| 23:46.56 | ginge | dcordes: which image? |
| 23:47.19 | dcordes | zImage |
| 23:47.56 | tmzt | does evtest report the input subsystem version? |
| 23:47.59 | dcordes | 00:26 ginge tmzt: yes, I have one pre-built here:http://www.headfuzz.co.uk/files/android/zImage |
| 23:48.03 | dcordes | that one |
| 23:48.24 | ginge | ok, so the one that reports android corrected values |
| 23:48.25 | dcordes | Input driver version is 1.0.0 |
| 23:48.51 | dcordes | ginge: so that's why the values look like in android |
| 23:49.05 | ginge | yeah |
| 23:49.50 | ginge | sync |
| 23:51.01 | tmzt | the question is why is check_fd failing in input-raw.c? |
| 23:51.20 | tmzt | it seems to be getting values from the driver, and not liking them |
| 23:52.27 | tmzt | that prevents any events from getting through because sane_fd is -1 |
| 23:52.54 | tmzt | it looks like it might be in a moko patch though |
| 23:56.15 | dcordes | tmzt: all patches OE applies should be in the tar I sent |
| 23:56.25 | tmzt | yeah, they are |
| 23:56.27 | dcordes | or do you mean moko phone specific patching? |
| 23:56.33 | tmzt | right |
| 23:56.58 | tmzt | it might be in the grab-events patch, check the metadata |
| 23:57.33 | tmzt | I can't find where EV_VERSION is defined |
| 23:57.45 | tmzt | did you get warnings about that? |
| 23:59.08 | dcordes | not sure I can build and check the do_compile log |
| 23:59.18 | dcordes | if useful |
| 23:59.28 | tmzt | it might be in linux/input.h, but that does not make sense |
