| 00:00.41 | dcordes | vilord, don't know where my AT stuff is. but it won't be of much help I'm sure. do you have your findings documented somewhere? |
| 00:01.37 | vilord | for now it is just documented in the code... github.com/vilord/android-htc-ril/tree |
| 00:02.30 | vilord | my primary findings were that although vogue doesn't seem to care what is issued when, gsm is pickier |
| 00:02.34 | dzo | vilord: there's no hardware long (or deep) press like on some phones, you would have to implement it in android or the driver. using it as a shift key is probably better so we get lots of combinations. |
| 00:03.33 | vilord | i was thinking of implementing the long press in the driver as a shift... short press would act as the "back" button, and long press would act as shift/meta |
| 00:05.50 | vilord | but first i have to figure out how to watch for a wait in your code, and only send back the short press upon release |
| 00:05.55 | dzo | that would work, i was thinking that the green key is not really used so that could be menu. |
| 00:06.22 | vilord | thats something where a keymap would be good IMO, as I use the green key all the time |
| 00:07.12 | dzo | you would have to change it to report back being pressed and then released on release and use a timer to check how long it had been pressed for. |
| 00:07.14 | vilord | how is wakeup handled? is that in the kernel or in android side? |
| 00:07.31 | vilord | thats what i was thinking too |
| 00:07.50 | dzo | wkeup from power collapse? |
| 00:07.53 | vilord | right |
| 00:08.27 | dzo | all in the kernel, you have to say which gpios can wake up the device but not all are allowed. |
| 00:09.27 | vilord | oh good. so changing the top button from home to menu (or making it remappable) won't affect wakeup |
| 00:10.43 | dzo | no, use irq_set_wakeup on the gpio you need to use for wakeup. |
| 00:10.55 | dzo | (or something similar) |
| 00:11.44 | dzo | it would probably get a menu keypress on wakeup though so not show the lock screen. |
| 00:12.13 | dzo | ok, got to go, see you. |
| 00:12.15 | vilord | right now you have debouncing in the code, is there a separate thread that triggers the button code for each time a button is pressed? If so, why does the loop need to hit all of the keys? |
| 00:12.26 | vilord | okay, I'll see what I can find, talk later |
| 00:12.48 | dzo | the debouncing is crap, it doesn't really work |
| 00:13.35 | dzo | there is one gpio for all the cpld keys and you have to poll to find which one is pressed. |
| 00:13.35 | vilord | i was thinking of adding a debounce attribute to each of the keys in the keys[] |
| 00:14.07 | dcordes | ~seen raster |
| 00:14.08 | apt | raster <n=raster@enlightenment/developer/raster> was last seen on IRC in channel #openmoko, 2d 17h 20m 24s ago, saying: 'just back that up - put it back on after reflahs and it will be as it was before'. |
| 00:14.32 | vilord | is there a separate call to the function for each key press? If so we could use a while loop and stop at the first key instead of a for/continue to check all keys |
| 00:14.40 | dzo | in fact it's more complicated thanthat because there is a gpio that says the cpld caused and interrupt then a set of cpld ints that need acknowledging then you poll the keys. |
| 00:15.27 | dzo | ok, really got to go, bye |
| 00:15.34 | vilord | thanks for the help cya |
| 00:16.19 | dcordes | bye dzo |
| 00:18.38 | dcordes | vilord, which driver do you use for kaiser wifi ? |
| 00:19.10 | vilord | they're using a slightly modified version of the wlan.ko in the open source distro... the chip is identical to that of the G1 |
| 00:19.34 | vilord | the only modifications are to turn off dma i/o, because it doesn't work right now |
| 00:20.14 | dcordes | ti1251 |
| 00:20.21 | vilord | right |
| 00:20.22 | dcordes | can you point me to the slight modifications ? |
| 00:20.35 | vilord | i'll see if he posted his source... |
| 00:22.18 | dcordes | thanks |
| 00:23.00 | dcordes | did you try the modified wlan.ko on polaris? |
| 00:23.27 | vilord | here's his post: http://forum.xda-developers.com/showpost.php?p=4260681&postcount=33 |
| 00:23.57 | *** join/#htc-linux dan-kaiser-TPro (n=none@d75-159-40-217.abhsia.telus.net) |
| 00:24.18 | *** join/#htc-linux Kez (n=chatzill@pool-72-88-86-158.bflony.fios.verizon.net) |
| 00:24.22 | vilord | looks like the source of the ti tree is here: http://drop.io/wifikaiser |
| 00:24.45 | vilord | ti-kaiserwifi.zip |
| 00:25.11 | vilord | i tried it and it wouldn't load at all, I don't think we have the memory addresses set right yet |
| 00:25.13 | Kez | Hello |
| 00:26.23 | Kez | Does anyone know if there is a way to root the new mytouch 3g? And if the hero rom runs smoothly on it? |
| 00:26.49 | dcordes | Kez, what's mytouch? |
| 00:27.00 | vilord | i think it is the same as the magic |
| 00:27.22 | vilord | (thats the one without the keyboard, right?) |
| 00:27.28 | Kez | Yeah |
| 00:27.34 | Kez | It's the htc magic |
| 00:27.40 | Kez | but for the usa |
| 00:27.49 | Kez | It is 32B I believe |
| 00:28.04 | vilord | i think the rogers magic has been rooted, so i imagine if it isn't already done, the t-mobile one shouldn't be long |
| 00:28.18 | Kez | Cool |
| 00:28.25 | Kez | Do you know if the hero rom is any good? |
| 00:28.26 | tmzt | dcordes: somewhat, we need to get dream working and merge our code for ce into it with proper abstractions |
| 00:29.07 | tmzt | the tmo has slightly different hardware and needs a kernel patch |
| 00:29.17 | Kez | what do you mean |
| 00:29.35 | vilord | kez: if i were u i'd check in the magic forum on xda-developers and do some searching |
| 00:29.42 | Kez | I've been looking around for a bit |
| 00:30.19 | Kez | and I've seen that there is a rom for the 32B (I'm guesing thats the same as the mytouch) I have read that it can be sluggish but I'm not sure if this is for the 32A or for teh 32B |
| 00:31.09 | vilord | tmzt's comment is that the hardware in the t-mobile 32B is a little bit different, so the rom won't work without a patch... right? |
| 00:31.59 | Kez | Yeah I see what your saying I'll take a look at that then. |
| 00:32.45 | dcordes | tmzt, I keep missing Kevin2. maybe you can ask him about the branches if you see him |
| 00:34.06 | tmzt | oh, that's what it is |
| 00:35.14 | dcordes | vilord, polaris and kaiser memory mapping are pretty much the same, no ? |
| 00:36.34 | vilord | they should be almost identical... but we don't have any really good kernel guys looking at it... from what i can tell the polaris and nike code is a mess right now, kaiser has had many more cleanups |
| 00:37.29 | vilord | (and as yet i'm not extremely confident in my kernel coding abilities... direct hardware i/o stuff isn't my strong suit yet) |
| 00:40.31 | dcordes | http://wiki.xda-developers.com/index.php?pagename=KaiserMemoryMap3 |
| 00:43.06 | *** join/#htc-linux Moku (n=John@f048140074.adsl.alicedsl.de) |
| 00:45.12 | dcordes | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=arch/arm/mach-msm/Makefile |
| 00:45.18 | dcordes | vilord, here's the problem |
| 00:45.56 | dcordes | s/board-htcpolaris-wifi.o/board-htckaiser-wifi.o/ |
| 00:46.22 | dcordes | if you take a quick look at http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=arch/arm/mach-msm/board-htcpolaris-wifi.c |
| 00:46.31 | dcordes | you see trout naming is replaced by polaris |
| 00:46.36 | dcordes | ti1251 might not like that |
| 00:47.23 | vilord | hmm, could be |
| 00:47.26 | vilord | what is trout? |
| 00:47.30 | vilord | i see it all over the place |
| 00:47.51 | tmzt | why are you using that driver? |
| 00:48.06 | tmzt | trout is dream/g1 |
| 00:48.19 | vilord | ah okay |
| 00:48.39 | vilord | is there a better ti1251 driver readily available? |
| 00:48.42 | dcordes | codename overdose |
| 00:48.50 | tmzt | what kernel? |
| 00:48.58 | vilord | 2.6.25 |
| 00:49.12 | tmzt | probably not |
| 00:49.21 | tmzt | what's the conflict? |
| 00:50.10 | vilord | trying to figure that out eventually... it works on kaiser but not polaris, the kernel code is very different tho, and so far I don't have access to the code for the kernel for kaiser that actually functions |
| 00:50.29 | tmzt | ? |
| 00:50.38 | tmzt | htc-vogue should |
| 00:51.34 | tmzt | what high-rez happens with kaiser mtype? |
| 00:51.39 | dcordes | vilord, wrt ti1251 wlan.ko from kaiser. you said it wouldn't load at all. modprobe fails ? |
| 00:51.40 | tmzt | happens |
| 00:52.06 | tmzt | what does dmesg say? |
| 00:52.07 | vilord | i agree it should... there's a kernel that was posted by user addisu that works with wlan.ko (minus the dma i/o), but if we bring in a regular release of the kernel it fails, so addisu changed something |
| 00:52.36 | vilord | i'll have to check... i've been working on RIL, not wifi... |
| 00:52.50 | tmzt | cool |
| 00:52.58 | tmzt | you have a gsm ril? |
| 00:53.01 | vilord | yeah |
| 00:53.06 | vilord | it works on both |
| 00:53.19 | vilord | s/both/gsm and cdma |
| 00:54.21 | vilord | lots of code rips from the ti idcc driver, lots of martin's code, lots of new code to more cleanly support connect/disconnect... |
| 00:54.37 | tmzt | idcc? |
| 00:54.54 | vilord | i'm not sure what it is, its in the main kernel tree |
| 00:54.55 | tmzt | for gprs? |
| 00:55.04 | vilord | or rather in the omap tree from ti |
| 00:55.17 | tmzt | oh |
| 00:55.18 | vilord | not sure what you mean |
| 00:55.29 | tmzt | <PROTECTED> |
| 00:55.29 | tmzt | <PROTECTED> |
| 00:55.45 | vilord | oh, yes for gprs |
| 00:55.54 | tmzt | cool |
| 00:56.09 | tmzt | can you make device configurable? |
| 00:56.16 | tmzt | ali1234: ping |
| 00:56.19 | vilord | i don't see why not |
| 00:56.26 | ali1234 | yes? |
| 00:56.32 | tmzt | what about audio |
| 00:56.52 | tmzt | vilord is working on a general android ril |
| 00:57.23 | ali1234 | right? |
| 00:57.23 | vilord | audio on our devices is controlled by the arm9 i think, so right now it just sends commands to the kernel via sysfs |
| 00:57.44 | vilord | (dzo's stuff there) |
| 00:57.53 | tmzt | would that work better than referwnce ril? |
| 00:57.59 | tmzt | on omap850 |
| 00:58.09 | vilord | you are running on omap850? |
| 00:58.27 | ali1234 | tmzt: i dont know anything about RIL, you need to talk to darkstar62 about this, he ported RIL |
| 00:58.43 | tmzt | ok |
| 00:59.30 | ali1234 | vilord: we have android running very well on omap850. if sound worked it would be very usable |
| 00:59.43 | ali1234 | on omap850 sound is split between mpu and gsm cores |
| 00:59.59 | ali1234 | basicallu the mpu has a soundcard and the gsm core is on one of it's inputs |
| 01:00.13 | ali1234 | so both ends have to be set up right for audio in calls |
| 01:00.35 | ali1234 | at least i think thats how it works |
| 01:00.37 | vilord | vogue/kaiser/polaris do it the same way... lots of gpio and cpld calls |
| 01:03.19 | darkstar62 | someone need ril support? |
| 01:03.41 | vilord | right now kaiser and polaris have audio out from the userspace as well as audio in/out from the gsm radio working. vogue also has audio in to userspace working |
| 01:04.52 | vilord | darkstar62: maybe we can work together to get a common RIL working for all these various htc devices... |
| 01:05.27 | darkstar62 | I'm up for that, I've got a library in Android that has most/all of the AT commands in place -- just a matter of making that more generic |
| 01:05.52 | vilord | the RIL I've been working on is finally very stable and supports lots of stuff, but there are still special htc-centric modem commands that I don't know what they are for |
| 01:06.19 | *** join/#htc-linux AstainZZZZZZ (n=AstainHe@unaffiliated/astainhellbring) |
| 01:06.19 | ali1234 | i hope whatever you guys comes up with makes gprs work for me :) |
| 01:06.33 | vilord | right now kaiser, polaris, and vogue are all working well with voice and data calls |
| 01:06.39 | darkstar62 | ali1234: have you had any luck setting up ppp from linux userspace? |
| 01:07.52 | ali1234 | darkstar62: you mean like running pppd myself? i have shown that can work on my own rootfs (not android) |
| 01:08.15 | darkstar62 | ok, so it's just an android issue then |
| 01:08.31 | ali1234 | darkstar62: yeah. android "knows" my correct apn and provider name, just wont connect to it |
| 01:09.26 | vilord | that was something specific that my RIL now handles much cleaner... using apn's from android |
| 01:11.00 | dcordes | bye |
| 01:11.16 | vilord | but it also works with a userspace tool that runs the ppp daemon... darkstar: is your RIL source in the wild somewhere? |
| 01:11.26 | vilord | bye |
| 01:11.54 | ali1234 | is RIL in kernel? |
| 01:12.17 | vilord | no its a userspace library run by rild... within the android tree |
| 01:13.07 | darkstar62 | yeah, it's in my repo |
| 01:13.22 | darkstar62 | it's based on libreference_ril, with some small changes |
| 01:15.31 | vilord | sorry... where's your repo? |
| 01:22.02 | darkstar62 | uh |
| 01:22.17 | darkstar62 | git://git.wing-linux.sourceforge.net/gitroot/wing-linux |
| 01:23.08 | darkstar62 | master branch is probably fine -- look in buildsys/oe/android-1.5 -- there's patches for the reference-ril, which is included in the Android source code |
| 01:27.31 | vilord | if its useful to you, take a look at github.com/vilord... there is userspace stuff in the initrd and ril stuff in the htc-ril |
| 01:30.12 | darkstar62 | cool, I'll take a look |
| 01:52.19 | *** join/#htc-linux AstainZZZZZZ (n=AstainHe@unaffiliated/astainhellbring) |
| 01:59.01 | darkstar62 | vilord: very cool, looks very similar to the reference ril code |
| 01:59.07 | darkstar62 | but very complete |
| 02:04.21 | vilord | it started with the reference ril code too |
| 02:05.53 | darkstar62 | figued |
| 02:05.57 | darkstar62 | *figured |
| 02:10.53 | vilord | okay the version I have on my phone is finding the device just fine, and it says that the device was "already mounted at startup" |
| 02:10.59 | *** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a) |
| 02:12.50 | Kez | Sorry, I'm sort of newb. I've decided to get the mytouch 3g (htc magic) however I don't really understand this whole rooting process. What exactly do i get out of it besides access to other roms such as hero and the like? |
| 02:13.03 | vilord | whoops, wrong forum sorry |
| 02:13.29 | tmzt | Kez: I think you enable adb mode and then follow some instructions, but I don't have a link for that |
| 02:13.48 | tmzt | if I understand it, for now, you have to use zImage and wlan.ko from the shipped image |
| 02:13.57 | Kez | Mk |
| 02:14.18 | Kez | What do I get out of rooting the phone? |
| 02:14.41 | Kez | thus far all I really understand is that i get access to some special roms and the like |
| 02:16.19 | tmzt | #android can probably tell you better |
| 02:18.37 | Kez | ALright thanks |
| 02:18.40 | Kez | Talking with them now |
| 02:23.41 | *** join/#htc-linux root (n=root@rgnb-5d874707.pool.einsundeins.de) |
| 02:32.27 | tmzt | vilord: you have that ril in git then? can we get that somewhere more stable, I'm trying to understand how you build it with oe though |
| 02:32.58 | vilord | is github not stable enough? |
| 02:33.25 | tmzt | well, that's not it. it's that we have to keep track of all these trees |
| 02:33.34 | tmzt | I guess it could just go in wiki somewhere |
| 02:33.45 | vilord | if you have an alternative, I'm open to it |
| 02:33.53 | vilord | linux-to-go doesn't make sense |
| 02:33.57 | tmzt | right |
| 02:34.10 | tmzt | we really need to get a clean android somewhere |
| 02:34.36 | tmzt | with patches for working on different kernels than than the g1/magic/hero |
| 02:34.44 | tmzt | well, ok |
| 02:34.51 | tmzt | how does oe work as build system? |
| 02:35.00 | vilord | what i do to build it now, is I just put a folder into hardware/ril called htcgeneric-ril and I just make that my git source |
| 02:35.01 | tmzt | can you build the ril standalone? |
| 02:35.21 | vilord | then i make the whole tree |
| 02:35.23 | vilord | very annoying |
| 02:35.28 | tmzt | right |
| 02:35.44 | tmzt | but it shouldn't really have any dependencies |
| 02:35.44 | vilord | and if i update any libraries on my desktop, it takes an hour |
| 03:01.22 | *** join/#htc-linux mrb__ (n=mrb@160.89-10-30.nextgentel.com) |
| 03:33.31 | *** join/#htc-linux verizonguy (n=verizong@c-24-63-231-35.hsd1.ct.comcast.net) |
| 04:40.28 | *** join/#htc-linux droid0011 (n=mc@p4FDCC1C7.dip.t-dialin.net) |
| 04:44.26 | *** join/#htc-linux dilinger (n=dilinger@204.132.39.17) |
| 04:48.30 | tmzt | dilinger: oh, hi. when you are awake again we would be happy to help with native debian (not android root) or whatever |
| 04:54.33 | dilinger | tmzt: i kinda want to play around more w/ the bootloader once i'm done w/ this ofono project (which will probably take a few months). openfirmware on the g1 would be pretty awesome :) |
| 04:54.41 | dilinger | really goes to sleep now |
| 06:09.26 | *** part/#htc-linux ragtimer (n=chatzill@p57AE2411.dip0.t-ipconnect.de) |
| 06:21.05 | *** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net) |
| 07:40.28 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d874230.pool.einsundeins.de) |
| 07:41.24 | *** join/#htc-linux ArteK (n=Artur@81.15.241.96) |
| 07:46.55 | *** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821) |
| 08:10.13 | *** part/#htc-linux ArteK (n=Artur@81.15.241.96) |
| 08:17.57 | *** join/#htc-linux leaigor (n=laigor@188.134.36.14) |
| 08:41.38 | *** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl) |
| 09:11.39 | *** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net) |
| 09:12.55 | *** join/#htc-linux marex (n=marex@thor.hackndev.com) |
| 09:41.53 | *** join/#htc-linux xperia (n=chatzill@77-56-91-2.dclient.hispeed.ch) |
| 09:59.43 | tmzt | cool, xrandr --output LVDS1 --panning 1280x800+0+400 |
| 10:02.39 | Squarc | ?" |
| 10:04.15 | tmzt | if you have dri enabled it shouldn't hurt anything |
| 10:04.25 | tmzt | I'm trying to see how to make msm graphics faster in X |
| 10:31.34 | Squarc | ok |
| 10:31.39 | Squarc | I get it :) |
| 10:32.18 | Squarc | tmzt, what was that Open Mobile project you where also working on... that htc-linux is kinda part of |
| 10:37.17 | tmzt | which? |
| 10:38.44 | *** join/#htc-linux Aiv- (n=aiv@office.mdsoftware.pl) |
| 10:38.46 | Aiv- | hello |
| 10:39.00 | Aiv- | hello xperia |
| 10:39.07 | Aiv- | i have good news ;] http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=commit;h=b2604231f692ae8c4ed96bfe505857511ac967c6 |
| 10:40.01 | tmzt | cool, so it's going to be intergrated? |
| 10:40.41 | Aiv- | yap, i use preprocessor in microksc |
| 10:41.10 | Aiv- | it's not looks gr8 but it's working fine and didn't change enything in other devices |
| 10:41.13 | tmzt | preprocessor? |
| 10:41.26 | tmzt | is it not possible to read version from microp? |
| 10:41.30 | tmzt | or put that in pdata? |
| 10:42.23 | Aiv- | in my opinion - no, i was asking you about preprocessor yesterday |
| 10:42.44 | Aiv- | i will write to madCoder- and maybe in future we will remote this preprocessor code |
| 10:43.07 | Aiv- | or maybe someone with raphael, blackstone can check this on his device |
| 10:43.48 | xperia | hey aiv great work man. |
| 10:43.54 | xperia | :-) |
| 10:44.16 | *** join/#htc-linux Captnoord (n=Captnoor@dc5147a47b.adsl.wanadoo.nl) |
| 10:44.19 | tmzt | well, if nothing else, machine_is_htckovsky() would be better than preprocessor |
| 10:44.24 | tmzt | since that's at runtime |
| 10:44.38 | tmzt | then we can build zImage's with kovs support by default |
| 10:45.33 | Aiv- | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blobdiff;f=drivers/i2c/chips/microp-ksc.c;h=693de2982efe2bea9fc01f2664d356453dd4b7da;hp=a2702676d99bc27be46945d8a10dc9c00fc28827;hb=b2604231f692ae8c4ed96bfe505857511ac967c6;hpb=31483ef2576d34bfb54e9fd8658d8d51cfd26bbb |
| 10:45.55 | Aiv- | here is that code |
| 10:46.18 | tmzt | right, you can change those to if (machine_is_htckovsky()) { ... } |
| 10:46.21 | Aiv- | IMO we need to add one argument to micropksc_read_scancode function |
| 10:46.32 | tmzt | instead of #if defined(MACH_HTCKOVSYKY) |
| 10:46.38 | tmzt | what is that? |
| 10:48.22 | Aiv- | where u find KOVSYKY? |
| 10:48.30 | xperia | i itself i am not a big friend of runtime checks. for me preprocessors settings are better. |
| 10:48.41 | tmzt | that's the codename |
| 10:48.52 | tmzt | well, arm list expects that these things are runtime |
| 10:49.13 | tmzt | but in reality, if there is only one machine defined, it changes to a compile time check automatically |
| 10:49.32 | Aiv- | changing function definition in runtime? |
| 10:50.06 | tmzt | gcc optimizes it out because it's changed to a macro that returns a constant 0 or 1 depending on if that's the machine being compiled |
| 10:50.35 | xperia | i have thinked the prepocessor just include or exclude the code that is inside the definition. checks are not maked at runtime. |
| 10:50.41 | tmzt | so you lose nothing in the base case and don't break the build in the other |
| 10:50.46 | tmzt | you are correct |
| 10:50.51 | tmzt | but if gcc sees: |
| 10:50.54 | tmzt | if (0) { } |
| 10:50.59 | tmzt | it will remove that whole block |
| 10:51.06 | tmzt | if it sees if (1) { } |
| 10:51.13 | tmzt | it won't check the value at all |
| 10:51.45 | tmzt | if only KOVS is defined in .config, for example, machine_is_htckovsky() gets #defined to be (1) |
| 10:52.02 | tmzt | if it's not defined, it's #defined to be (0) |
| 10:52.07 | Squarc | isnt it better to use precompile directions then ? |
| 10:52.24 | tmzt | it is a precompile directive, or the equivalent |
| 10:52.46 | tmzt | unless multiple macines are defined (which is allowed), then it becomes a runtime check |
| 10:52.53 | Squarc | oke |
| 10:54.27 | tmzt | anyway, xperia, what did you need an new parameter for? |
| 10:54.41 | tmzt | in microp_read_scancode() |
| 10:55.05 | xperia | tmzt: does this also count for other build software than gcc like arm-none |
| 10:55.51 | tmzt | arm-none is gcc |
| 10:56.44 | xperia | ahh okay well in this case it makes sense. i didnt know that gcc optimaze the source code such good before it compile |
| 10:56.46 | tmzt | if you look in the patch it's installed in, you will see arm-none-eabi-gcc or similar |
| 10:56.49 | Aiv- | tmzt i need new parameter for clamshell in kovsky. i told u many times that madCoder- didn't know what he is reading from irq "//TODO: Find out what channel 0x11 is for" |
| 10:56.50 | tmzt | that's the compile |
| 10:56.57 | tmzt | I know that |
| 10:57.01 | tmzt | what is the parameter |
| 10:57.06 | tmzt | sorry, asked the wrong dev |
| 10:57.15 | tmzt | can you export these: |
| 10:57.19 | parmaster | headed to japan, catch y'all in a couple of weeks |
| 10:57.23 | tmzt | microp_get_switch() |
| 10:57.34 | tmzt | microp_get_backlight() |
| 10:57.39 | tmzt | microp_set_backlight() |
| 10:57.43 | tmzt | etc. |
| 10:57.54 | tmzt | and leave scancode for actual keys? |
| 10:58.16 | tmzt | or led, maybe, not backlight |
| 10:59.06 | tmzt | pwm, etc. |
| 10:59.38 | tmzt | well, that won't really work either |
| 10:59.57 | tmzt | since you shouldn't be polling the clamshell status, you must have an interrupt handler already |
| 11:00.16 | Aiv- | you talking about solution: create other function (and export it) for clamshell, and leave micropksc_read_scancode ony for keyboces? i'm understand correctly? |
| 11:00.22 | tmzt | yes |
| 11:00.32 | Aiv- | ok but second question |
| 11:00.34 | tmzt | we need backlight also |
| 11:00.39 | Aiv- | madCoder- read data from irq two times |
| 11:00.49 | tmzt | I havea driver for it but it needs an export |
| 11:01.07 | Aiv- | he allways read data after keypress |
| 11:01.15 | tmzt | yeah |
| 11:01.23 | Aiv- | so i need to change this function to save clamshell status localy in this module |
| 11:01.36 | Aiv- | but function definiction will be clear |
| 11:01.45 | xperia | tmzt sorry to unterup you with the discussion. i have spend yesterday more than 12 Hours to get the f***ing Keyboard backlight to work. Can you say me what this here mean for you |
| 11:02.07 | xperia | KOVS110 |
| 11:02.09 | xperia | SPL-0.72.OliNex |
| 11:02.10 | xperia | MicroP1-787 |
| 11:02.12 | xperia | MicroP2-707 |
| 11:02.19 | tmzt | does microp_read_scancode already read from the i2c bus directly? |
| 11:02.25 | tmzt | I forget how he implemented it |
| 11:02.29 | tmzt | xperia: huh? |
| 11:02.34 | tmzt | you mean keyled? |
| 11:02.38 | tmzt | I mean screen backlight |
| 11:03.05 | Aiv- | it's use micropksc_read() to read from i2c directly |
| 11:03.10 | tmzt | ok |
| 11:03.16 | xperia | normaly we should have MicroP -Led and MicroP-Key printed for Microp-KSC and MicroP-KLt |
| 11:03.18 | tmzt | I think we should do something different anyway |
| 11:03.21 | tmzt | just thought about it |
| 11:03.29 | Aiv- | i'm think that madCoder- rebuild driver using deassambled windows driver |
| 11:03.40 | tmzt | we should have a msm_get_microp_klt/ksc() function that returns a struct |
| 11:03.48 | tmzt | this should be in devices.c |
| 11:03.55 | xperia | yeah i think we need to dissamble the drivers too |
| 11:03.58 | Aiv- | and he put this second read (first it's for keycode), but hw don't know why .. |
| 11:04.01 | tmzt | or maybe htc_get_microp |
| 11:04.11 | tmzt | but anyway, just do the exports for now |
| 11:04.31 | tmzt | you will have to ask him then, he might have a few minutes some time |
| 11:04.45 | tmzt | where is this second read being done? |
| 11:04.59 | Aiv- | pleas open this url: http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=drivers/i2c/chips/microp-ksc.c;h=693de2982efe2bea9fc01f2664d356453dd4b7da;hb=b2604231f692ae8c4ed96bfe505857511ac967c6 |
| 11:05.04 | Aiv- | i will operate with line numbers |
| 11:05.21 | Aiv- | #77 he read data but never use it |
| 11:05.21 | tmzt | ok |
| 11:05.40 | Aiv- | so i add code in #if .. #endif |
| 11:06.35 | tmzt | this? |
| 11:06.36 | tmzt | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=drivers/i2c/chips/microp-ksc.c;h=693de2982efe2bea9fc01f2664d356453dd4b7da;hb=b2604231f692ae8c4ed96bfe505857511ac967c6 |
| 11:06.40 | tmzt | sorry |
| 11:06.48 | tmzt | <PROTECTED> |
| 11:06.59 | tmzt | so that's for clamshell on kovs? |
| 11:07.04 | Aiv- | so i will replace preprocesor code with save read data in line 77 to local var |
| 11:07.22 | Aiv- | and this local var will be accessible for read via exportted function |
| 11:07.26 | Aiv- | it will be ok? |
| 11:07.27 | tmzt | it has to be in data |
| 11:07.34 | tmzt | it can't be local, or things will break |
| 11:07.47 | tmzt | and you have to lock/unlock around it |
| 11:08.28 | tmzt | wait, maybe not |
| 11:08.29 | Aiv- | but AFAIR data struct contains only information about keycode |
| 11:08.38 | tmzt | how long do you expect it to survive? |
| 11:08.48 | tmzt | if you intend to read it from somewhere else |
| 11:09.00 | tmzt | I think it would be eaiser to duplicate this function and export it |
| 11:09.07 | tmzt | then call both of them on irq |
| 11:09.17 | tmzt | check keydown and report the key if it's down |
| 11:09.25 | tmzt | check clamshell and report that |
| 11:09.54 | tmzt | best would be to ask maejrep though, it's his code |
| 11:09.58 | tmzt | sorry I can't help more |
| 11:10.14 | Aiv- | u can read one irq signal many times? |
| 11:10.43 | tmzt | no, I mean read from i2c multiple times, in different functions |
| 11:10.46 | tmzt | same irq |
| 11:12.53 | Aiv- | hmm |
| 11:13.52 | Aiv- | i will do some test and try to rewrite this little part od code |
| 11:15.50 | xperia | is captnord here or others who have experince with reverse enginering some stuff |
| 11:18.28 | Aiv- | last question: if u duplicate this function only for kovsky, we can use runtine check and use "special" function for kovsky and original function for other devices |
| 11:18.33 | Aiv- | i can do it fast |
| 11:18.34 | *** join/#htc-linux UnknownArtists (n=unknown@84.19.198.113) |
| 11:18.38 | Aiv- | this will be ok too? |
| 11:18.44 | tmzt | yes |
| 11:18.49 | Aiv- | gr8 |
| 11:18.57 | Aiv- | i will commit this today |
| 11:20.12 | xperia | tmzt: what does MicroP1 at 787 and MicroP2 at 707 mean. some of this 7x7 number must be the xex adress 0x67 for the microp-ksc module be |
| 11:22.55 | tmzt | those are version |
| 11:22.57 | tmzt | versions |
| 11:23.01 | tmzt | huh? |
| 11:23.18 | tmzt | that's how you should be detecting the version, but checking board works as well |
| 11:23.24 | tmzt | firmware versions |
| 11:24.56 | xperia | hmm there is no chance getting the second i2c address then from this output. as i know it print normally also i2c addresses |
| 11:25.13 | tmzt | it should be in i2c_board_info |
| 11:25.21 | tmzt | why do you need it? |
| 11:25.38 | xperia | To get the Leds and Keybaordbacklight working |
| 11:26.18 | tmzt | those are diferent drivers I think |
| 11:26.23 | tmzt | klt and ksc |
| 11:26.40 | tmzt | hold on, why is this a seperate i2c address? |
| 11:27.02 | xperia | ksc is working allready on 0x67 in raph for klt we have 0x66 |
| 11:27.34 | xperia | i have inside the klt driver a bruteforce check done to find the other i2c adress |
| 11:27.45 | tmzt | the i2c addresses should not be in the driver |
| 11:27.49 | tmzt | only in i2c_board_info |
| 11:27.59 | tmzt | maejrep already did that part |
| 11:28.11 | tmzt | if you have the addresses, add them to i2c_board_info in kovs board file |
| 11:28.36 | xperia | yeah i know but if you dont know the right i2c address you need in the driver to chnage the address to probe if it works. |
| 11:29.06 | tmzt | sorry? |
| 11:29.38 | tmzt | oh, to override it |
| 11:29.39 | tmzt | ok |
| 11:29.44 | tmzt | (for debuging) |
| 11:30.10 | xperia | right: look in the board file of htc-raphael you have two i2c addresses. one is for micropksc and the other is for microp-klt |
| 11:30.42 | xperia | the i2c adress for microp-ksc from raphael works also in htc0kovsky |
| 11:30.56 | tmzt | and klt? |
| 11:31.11 | xperia | the i2c address for microp-klt from raphael dont work in htckovsky |
| 11:31.50 | tmzt | so you need to trace with haret |
| 11:31.54 | tmzt | the i2c write register |
| 11:32.05 | *** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net) |
| 11:32.29 | xperia | yeah that would be the best but nobody could give me the haret console command how to trace it |
| 11:33.15 | tmzt | you need to look ath linux headers to get the address |
| 11:33.20 | tmzt | then: |
| 11:33.31 | tmzt | addlist mmutrace <address> |
| 11:33.53 | tmzt | <address> being all of the interesting ones in dump mmu 2 <phsyical) |
| 11:36.32 | xperia | hey aiv do you understand what tmzt is talking about. do i need first to run dump mmu 2 in haret console to get the range of the interessting addresses and then addlist mmutrace the specific address |
| 11:37.29 | xperia | linux headers is that the header files in include/linux or the haret include header files |
| 11:37.36 | tmzt | yes |
| 11:37.51 | tmzt | in arch/arm/mach-msm/include |
| 11:38.01 | tmzt | you need to git grep to find it |
| 11:38.13 | tmzt | otherwise you can look through the logs here from when we worked on this |
| 11:39.44 | xperia | okay what i dont understand at the moment is the part if i have the adress from linux headers where can i use the and how. is this address needed for this dump mmu 2 < linuxheaderadress |
| 11:41.28 | xperia | ahh okay now i think i understand. first find the linuxheader adress. then addlist mmutrace linuxheader address. |
| 11:41.59 | xperia | but for what is dump mmu 2 good to watch only this address. is this right |
| 11:42.06 | tmzt | yes |
| 11:42.24 | tmzt | mmutrace watches virtual addresses |
| 11:43.22 | xperia | goood thanks a lot tmzt fot the helpfull info. i have now to research some stuff |
| 11:43.24 | tmzt | since you need to know what virtual addresses wm uses for the physical addresses |
| 11:44.29 | tmzt | arch/arm/mach-msm/include/mach/msm_iomap.h:#define MSM_I2C_PHYS 0xA9900000 |
| 11:44.32 | tmzt | arch/arm/mach-msm/include/mach/msm_iomap.h:#define MSM_I2C_PHYS 0xA9900000 |
| 11:44.35 | tmzt | there is is |
| 11:46.49 | xperia | thanks for finding it. i have checkd it also and i can confirm that is the same |
| 11:48.40 | tmzt | so |
| 11:48.54 | tmzt | dump mmu 2 0xa9900000 |
| 11:49.01 | xperia | now i will start haretconsole and execute the command "dump mmu 2 < 0xA9900000" for finding the interessting virtual adresses. and then addlist mmutrace <specificvirtaddress> |
| 11:49.03 | tmzt | just use the 1M mappings by default |
| 11:49.09 | tmzt | no < |
| 11:49.18 | tmzt | right |
| 11:49.19 | xperia | ah okay |
| 11:49.26 | tmzt | then wirq 2 or 20 |
| 11:49.31 | tmzt | and flip the clamshell |
| 11:49.41 | tmzt | or whatever turns on the leds |
| 11:52.38 | xperia | <PROTECTED> |
| 11:52.40 | xperia | <PROTECTED> |
| 11:52.42 | xperia | ----------+----------+---------------+---------------------- |
| 11:52.43 | xperia | 92300000 | a9900000 | 1MB section | AP=1 T=2 |
| 11:52.45 | xperia | b2300000 | a9900000 | 1MB section | AP=1 T=2 |
| 11:53.13 | xperia | this is what dump mmu 2 0xa9900000 gives |
| 11:53.44 | tmzt | addlist mmutrace 0x92300000 0x10 |
| 11:53.47 | tmzt | addlist mmutrace 0xb2300000 0x10 |
| 11:53.52 | tmzt | that should work |
| 11:53.54 | tmzt | wirq 20 |
| 11:54.10 | tmzt | the address should be the first thing written to the address |
| 11:54.18 | tmzt | if you want the exact address, hold on |
| 11:54.54 | xperia | okay one second please |
| 11:57.05 | xperia | i am getting spaming the whole screen with 000.000 IRQS IRQ1: GPIO2(33)=1 |
| 11:58.53 | xperia | i think i need ibit irq 1 right to remove this |
| 11:59.20 | tmzt | yes |
| 11:59.25 | tmzt | not 1 though, ibit 33 |
| 11:59.40 | tmzt | your machine is not being detected correctly |
| 12:02.02 | xperia | it gives |
| 12:02.04 | xperia | line 3: Expected <watch list var |
| 12:02.24 | tmzt | ibit irqs 33 |
| 12:03.29 | xperia | tmzt it works great now. the IRQS are not mor showing |
| 12:04.56 | xperia | now i see only adresses. the problem is that it output heavy a lot of addresses even if nothing is doing on the phone |
| 12:04.58 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00005b01 |
| 12:05.00 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00005b01 |
| 12:05.01 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00007b01 |
| 12:05.02 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00005b01 |
| 12:05.04 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00009b01 |
| 12:05.06 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00007b01 |
| 12:05.07 | xperia | 004.378 7806000c: e5923000(ldr) # b2300008==00005b01 |
| 12:05.16 | tmzt | you need to use haretconsole for this |
| 12:05.32 | tmzt | well, we never added it |
| 12:05.37 | xperia | this adresses are spaming the whole time. i am using haretconsole for this |
| 12:05.45 | tmzt | this is where I need madCoder- |
| 12:05.50 | tmzt | ok |
| 12:06.06 | tmzt | I think that's it though |
| 12:06.21 | tmzt | madCoder-: have a minute? |
| 12:06.30 | xperia | i will now look if i can separate the adresses when i open slider. it output some other adresses as well |
| 12:07.01 | *** join/#htc-linux imfloflo (n=moi@153.25.71-86.rev.gaoland.net) |
| 12:09.34 | imfloflo | hi dzo just test your lastest zimage-polaris and have some issue with it |
| 12:09.56 | xperia | could it be that the number at the end 67 is the ksc |
| 12:09.57 | xperia | 002.092 7805f9f4: e5975000(ldr) # b2300008==00004700 |
| 12:09.59 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.01 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.04 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.06 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.07 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.08 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.10 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.12 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.13 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.15 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.16 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00008700 |
| 12:10.18 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00004700 |
| 12:10.20 | xperia | 002.092 780602fc: e5963000(ldr) # b2300008==00006700 |
| 12:10.41 | xperia | i have to prove it with a key stroke. one second please |
| 12:10.57 | *** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net) |
| 12:22.10 | Aiv- | tmzt http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=commitdiff;h=abfb31f3e1d5e8096eb6668401963d1ec5212889 |
| 12:22.26 | Aiv- | tmzt now u can include kovsky support with other devices without problems |
| 12:24.23 | tmzt | cool |
| 12:25.15 | tmzt | I guess that works |
| 12:30.02 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 12:31.11 | *** join/#htc-linux BUGabundo (n=bugabund@ubuntu/member/bugabundo) |
| 12:31.34 | BUGabundo | waves |
| 12:31.40 | tmzt | hello |
| 12:31.55 | BUGabundo | just checking out the place |
| 12:32.04 | BUGabundo | likes the word GNU/Linux |
| 12:32.31 | tmzt | logs are at irclog.netripper.com |
| 12:33.10 | BUGabundo | I see it |
| 12:33.46 | BUGabundo | not sure its appropriated here, but how come I had to use Windows to update my fw ?? |
| 12:34.02 | BUGabundo | why doesn't the HTC site have a linux package and instructions? |
| 12:34.50 | tmzt | on what device? |
| 12:35.01 | tmzt | there's fastboot for linux, #android supports that stuff |
| 12:35.26 | tmzt | I mean google/android develops it |
| 12:35.33 | xperia | tmzt: i have now uploaded the haret log to pastebin |
| 12:35.35 | xperia | http://pastebin.com/f1e8d6158 |
| 12:35.48 | tmzt | good, maybe madCoder- can look at it later |
| 12:35.59 | BUGabundo | tmzt: magic. I saw on htc site the upgrade, but no Linux info. just a wind exe |
| 12:36.05 | tmzt | cna you annotate were the fllip changes or led or whatever |
| 12:36.26 | BUGabundo | it would be great for all of my linux friends to be able to upgrade too |
| 12:36.34 | tmzt | they must have packaged it |
| 12:36.36 | tmzt | they can :) |
| 12:36.43 | BUGabundo | wiki link please??? |
| 12:36.49 | BUGabundo | pretty please? |
| 12:36.50 | tmzt | I don't have it |
| 12:36.56 | BUGabundo | with cherry on top? |
| 12:36.58 | BUGabundo | ohh .( |
| 12:37.02 | tmzt | you can always try 7za x the .exe |
| 12:37.06 | tmzt | that usually works |
| 12:37.12 | tmzt | then you get update.zip/nbh or whatever |
| 12:37.16 | BUGabundo | ok... then what's need to do? |
| 12:37.24 | BUGabundo | right |
| 12:37.36 | BUGabundo | what's this fastboot you say? |
| 12:37.52 | tmzt | a program to talk to the bootloader |
| 12:38.01 | BUGabundo | ok |
| 12:38.14 | BUGabundo | so I get the exe, extract, get the zip, and then? |
| 12:39.15 | BUGabundo | tmzt: ? |
| 12:39.39 | tmzt | I don't even know if that will work |
| 12:39.44 | tmzt | but that's what I would try |
| 12:41.33 | BUGabundo | ahahah |
| 12:41.38 | BUGabundo | then I better not :) |
| 12:41.58 | tmzt | it won't hurt anything, it just unzips the file |
| 12:42.11 | tmzt | I mean I don't know the format, it will even unzip |
| 12:42.38 | BUGabundo | lunch. bbl |
| 12:46.31 | *** join/#htc-linux _ibz (n=ibz@91.104.180.180) |
| 12:46.38 | *** join/#htc-linux nebi (n=nebi@c83-255-51-86.bredband.comhem.se) |
| 12:50.50 | *** join/#htc-linux Gnutoo (n=gnutoo@ABordeaux-152-1-26-178.w83-193.abo.wanadoo.fr) |
| 12:52.10 | Gnutoo | hi |
| 12:52.20 | Gnutoo | for the touchscreen: |
| 12:52.36 | Gnutoo | calibrate with unmodified openmoko-msm kenrel |
| 12:52.43 | Gnutoo | ->2* click |
| 12:52.56 | Gnutoo | then: modify the kenrel |
| 12:52.58 | Gnutoo | ->better |
| 12:53.02 | Gnutoo | then recalibtate |
| 12:53.13 | Gnutoo | ->terrible inaccuracy |
| 12:53.21 | Gnutoo | so I suspect the calibration tool |
| 12:53.36 | tmzt | hello again |
| 12:53.49 | Gnutoo | could leviathan_ publish his /etc/pointercal ? |
| 12:54.02 | tmzt | ask him |
| 12:54.06 | Gnutoo | via mail? |
| 12:54.13 | tmzt | leviathan_: ping |
| 12:54.16 | Gnutoo | or does he come on irc sometime? |
| 12:55.34 | *** part/#htc-linux _ibz (n=ibz@91.104.180.180) |
| 12:55.34 | Gnutoo | btw was someone able to receive calls? |
| 12:55.41 | *** join/#htc-linux _ibz (n=ibz@91.104.180.180) |
| 12:55.43 | *** join/#htc-linux dcordes (n=luke@unaffiliated/dcordes) |
| 12:55.54 | dcordes | hi |
| 12:56.18 | xperia | hello dcordes |
| 12:56.32 | Gnutoo | hi |
| 12:58.19 | dcordes | hey how's everybody? |
| 12:59.38 | xperia | good thanks for the question. how are you. |
| 13:00.30 | Gnutoo | good too |
| 13:01.27 | *** join/#htc-linux StarLite (n=nnscript@84-245-35-223.dsl.cambrium.nl) |
| 13:01.27 | dcordes | I'm fine. except for my fridge being empty :) |
| 13:04.12 | dcordes | xperia, did you read the log from yesterday? I think we should also try wifi on msm7xxxA |
| 13:04.13 | xperia | ohh man thats no good. time for shoping drinks and food in this case or doing the old way going to jungle in tarzan costume and getting the food :-) |
| 13:04.40 | dcordes | where I live grocery stores are closed sundays |
| 13:05.05 | dcordes | Aiv-, hi, good to see you pushed everything |
| 13:05.47 | tmzt | with wl12xx? |
| 13:06.48 | xperia | thats bad really in this case only some pizza service :-) for me i have it more easy. the mainstaition train just halt at my house and go directly to the mainstaition where the stroes are nearly 24 hours open. about wifi i will for sure try wifi but at the moment i am stuck on microp-klt |
| 13:07.29 | xperia | i am having big problems getting this microp-klt working |
| 13:07.29 | dcordes | tmzt, I'm curious if we can get it work easily with ti1251. Just wondering becuase it seems so trivial for kaiser with the trout module. wl12xx of course is preferred. |
| 13:07.57 | tmzt | it works on kaiser? |
| 13:08.10 | *** join/#htc-linux miknix (n=miknix@gentoo/developer/miknix) |
| 13:08.13 | dcordes | yea. see log from yesterday. vilord explained everything |
| 13:08.25 | tmzt | so mmc works |
| 13:08.32 | dcordes | and put the releveant links |
| 13:08.44 | xperia | some clarification if x1 has microp-klt and having the right i2c adress would help a lot |
| 13:08.45 | tmzt | what device does he have? I thought it was omap |
| 13:09.05 | dcordes | tmzt, the kaiser is msm7200 |
| 13:09.17 | dcordes | and lives in htc-vogue |
| 13:10.28 | dcordes | xperia, I think I will go for kebab |
| 13:10.56 | xperia | well kebab is for sure also very good :-) |
| 13:11.44 | dcordes | tmzt, vilord has a vogue and a polaris and he works on a unified android ril for both, gsm and cdma |
| 13:12.48 | dcordes | tmzt, the author of the kaiser wifi patches anounces it here http://forum.xda-developers.com/showpost.php?p=4260681&postcount=33 |
| 13:13.06 | dcordes | tmzt, I inqiured about needed patches in htc-vogue |
| 13:16.35 | Gnutoo | I started a page at: http://www.htc-linux.org/wiki/index.php?title=Dream |
| 13:18.58 | Aiv- | dcordes hello |
| 13:19.40 | Aiv- | dcordes as u can see, we solved problem with microp clamshell support without breaking other devices |
| 13:20.09 | dcordes | yea I took a quick glance at everything seems nice |
| 13:21.37 | dcordes | Gnutoo, ok I added the link on Main_Page |
| 13:22.24 | Gnutoo | ok |
| 13:22.53 | Aiv- | dcordes did u try to run wifi on kovsky? |
| 13:23.48 | Gnutoo | btw how do I accelerate the trackball? |
| 13:23.50 | dcordes | Aiv-, not but I want to try with both, ti1251 (google git) and wl12xx drivers |
| 13:24.13 | dcordes | Gnutoo, is it a mouse input device? |
| 13:24.15 | Gnutoo | btw still no chances with wifi? |
| 13:24.17 | Gnutoo | yes it is |
| 13:24.22 | dcordes | nice |
| 13:24.30 | dcordes | does it work with Xfbdev? |
| 13:24.35 | Gnutoo | yes it does |
| 13:24.39 | Gnutoo | but it's slow |
| 13:24.40 | dcordes | that's even nicer |
| 13:24.53 | Gnutoo | I have to wait a lot of time to go from one point of the screen to another |
| 13:25.07 | Gnutoo | it works fine but...you have to click a special way |
| 13:25.14 | Gnutoo | you need to go up while clicking |
| 13:26.46 | dcordes | Aiv-, not yet I mean |
| 13:27.23 | dcordes | Gnutoo, I suspect we must change the click thing in the driver, also for mainline |
| 13:27.31 | Aiv- | dcordes k, did u try to run gps from board-htcraphael-gps.c? |
| 13:27.32 | dcordes | so it's comptabile with normal userspace |
| 13:27.34 | Gnutoo | ok |
| 13:27.37 | dcordes | Aiv-, no |
| 13:27.57 | Gnutoo | Aiv-, I also want gps...where can i find thins file? |
| 13:28.03 | Gnutoo | can I try it on G1? |
| 13:29.04 | Aiv- | Gnutoo ltg git? |
| 13:29.12 | Gnutoo | ok thanks |
| 13:29.34 | Aiv- | Gnutoo tell me, is g1 isn't delivered with android ? |
| 13:29.50 | Gnutoo | g1 comes with android |
| 13:29.52 | Aiv- | Gnutoo so why u tru to run this kernel ? ;] |
| 13:29.53 | Gnutoo | but I prefer SHR |
| 13:30.00 | Gnutoo | so I've put SHR on it |
| 13:30.13 | Gnutoo | I used android's kenrel |
| 13:30.14 | Gnutoo | but... |
| 13:30.29 | Gnutoo | I didn't find power_on sysfs node |
| 13:30.33 | Gnutoo | for gps |
| 13:30.42 | Gnutoo | and I also grepped for 0 in /sys |
| 13:30.54 | *** join/#htc-linux MethoS- (n=clemens@dyndsl-085-016-163-201.ewe-ip-backbone.de) |
| 13:31.00 | Gnutoo | ah board-htcraphael-gps.c is in the kenrel |
| 13:31.05 | dcordes | Aiv-, shr is GPL distro |
| 13:31.06 | Gnutoo | I thought it was an userland program |
| 13:31.15 | dcordes | Aiv-, originated from openmoko project |
| 13:31.18 | Aiv- | ok, now i understand |
| 13:31.21 | Gnutoo | SHR is the best distro for the openmoko |
| 13:31.29 | Gnutoo | but it evolves every days |
| 13:31.34 | Gnutoo | no stable release |
| 13:31.49 | dcordes | Aiv-, android kernel drivers for g1 use non standard userspace interfaces. so we must change the kernel. |
| 13:32.02 | Aiv- | ok |
| 13:32.06 | Gnutoo | is there an usespace program that could power on the GPS? |
| 13:32.27 | tmzt | echo? |
| 13:32.34 | Aiv- | ;] |
| 13:32.37 | Gnutoo | yes but on what /sys node? |
| 13:32.56 | *** join/#htc-linux StarLite (n=nnscript@84-245-35-223.dsl.cambrium.nl) |
| 13:33.46 | dcordes | Gnutoo, have a look in the htc-vogue git. somebody made gps work for non-A kaiser device there |
| 13:33.52 | dcordes | grep commit log |
| 13:33.54 | Gnutoo | ok thanks |
| 13:34.08 | dcordes | I think it's for the android use, but it might give us some hints |
| 13:34.37 | *** join/#htc-linux junalmeida (n=junalmei@201.19.121.81) |
| 13:35.10 | Gnutoo | ok thanks a lot |
| 13:35.29 | dcordes | Gnutoo, http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=commit;h=1b9d1eb2e688574604f29014a79546a2a24cc54b this is the first commit with gps |
| 13:36.14 | dcordes | #include <asm/arch/msm_rpcrouter.h> |
| 13:36.26 | dcordes | it's all rpc |
| 13:36.43 | tmzt | but that needs to be in kernel right? |
| 13:36.52 | dcordes | Aiv-, might be easy for us to use it for kovsky |
| 13:37.30 | Aiv- | we dont need any gpios/irq if it's only rpc calls |
| 13:37.32 | Aiv- | i will try to run it |
| 13:37.35 | dcordes | the commit shortlog explains it |
| 13:37.47 | dcordes | Aiv-, I'm sure you can handle this |
| 13:37.50 | Aiv- | i have startgps.c from our last discussion about gps |
| 13:39.26 | dcordes | where's startgps.c ? |
| 13:39.53 | Gnutoo | that would be great!!! |
| 13:40.01 | Aiv- | w8 i will put it on my www |
| 13:40.17 | Gnutoo | because we have free kernel implementation of gps but lack the userspace thing |
| 13:40.49 | Aiv- | http://aiv-dev.info/opensource/xperiax1/ |
| 13:40.57 | Aiv- | source and compiled for arm |
| 13:41.08 | dcordes | what will be the best way to send the commands ? add it in gpsd ? |
| 13:41.43 | Aiv- | there isn only remote call to vouge_gps device |
| 13:41.54 | marajin | yo |
| 13:41.57 | Gnutoo | btw for the trackball I have: -mouse driver [,n,,options] Specify the pointer driver and its options (n is the number of buttons) |
| 13:42.09 | Gnutoo | what could be the good option for accelerating it? |
| 13:42.14 | Gnutoo | ok thanks for sources |
| 13:42.58 | dcordes | marajin, arisen from the non-geek underworld eh |
| 13:43.09 | marajin | dcordes: yeah been busy |
| 13:43.18 | marajin | dcordes: the new business is eating my time |
| 13:43.29 | dcordes | what are you doing now? |
| 13:43.41 | marajin | own/run a private/independent school |
| 13:44.08 | dcordes | nice. for what? |
| 13:44.16 | marajin | whaddya mean 'for what?' |
| 13:44.28 | dcordes | marajin's private business college? |
| 13:44.39 | marajin | nah it's 11-16 year olds |
| 13:45.06 | dcordes | cool |
| 13:45.24 | marajin | my mum worked there for like 20 years but it went up for sale so we bought it |
| 13:50.31 | dcordes | Gnutoo, is the trackball driver a single module or does it come with the touchscreen driver? |
| 13:50.39 | Gnutoo | no idea |
| 13:50.44 | Gnutoo | it just woked |
| 13:50.52 | Gnutoo | s/woked/worked |
| 13:50.52 | dcordes | where is -mouse driver [,n,,options] from ? |
| 13:51.00 | Gnutoo | from Xfbdev --help |
| 13:51.11 | dcordes | ok thought that was a paramter for loading the module |
| 13:51.24 | dcordes | so we should find out where the driver is first |
| 13:51.55 | Gnutoo | ok I'll look |
| 13:53.25 | Gnutoo | how could I find that |
| 13:55.31 | tmzt | which driver? |
| 13:55.49 | tmzt | trackball is another evdev device |
| 13:55.51 | tmzt | input device |
| 13:56.07 | tmzt | I believe it has KEY_UP, etc. but I don't know |
| 13:56.25 | tmzt | cat /sys/class/input/*/name should tell you what it supports |
| 13:56.29 | tmzt | what the names are |
| 13:57.27 | dcordes | kernel-google/msm/arch/arm/mach-msm$ grep -ri mouse . | grep trout |
| 13:57.27 | dcordes | ./board-trout.c:{ TROUT_GPIO_NAVI_ACT_N, BTN_MOUSE } |
| 13:57.27 | dcordes | ./board-trout.c:BTN_MOUSE, |
| 13:58.41 | tmzt | not exactly |
| 13:58.53 | tmzt | actually, BTN_MOUSE was always the wrong event |
| 13:59.42 | dcordes | the movement is a combination of direction key events? |
| 13:59.54 | tmzt | I think so |
| 14:00.03 | tmzt | it's not analog, it's just a set of switches |
| 14:00.09 | tmzt | according to swetland I think |
| 14:00.30 | tmzt | you should have a driver in drives/input/mouse or somewhere |
| 14:02.02 | dcordes | drives/input/mouse doesn't have any google, dream or trout |
| 14:02.44 | xperia | have to go. till next time bye |
| 14:03.36 | *** join/#htc-linux gnutoo_ (n=gnutoo@ABordeaux-152-1-64-172.w82-125.abo.wanadoo.fr) |
| 14:05.35 | dcordes | tmzt, do you have any idea why GNUtoo has to perform a special trick in order to do a click with trout_axis ? |
| 14:05.41 | dcordes | tmzt, he needs move up and then click |
| 14:06.06 | dcordes | http://www.htc-linux.org/wiki/index.php?title=Dream#Status |
| 14:06.17 | Kevin2 | dcordes: Hi |
| 14:06.38 | dcordes | hi Kevin2 |
| 14:06.46 | tmzt | probably because of BTN_MOUSE |
| 14:06.51 | tmzt | it's just the wrong code |
| 14:07.11 | dcordes | Kevin2, I wanted to ask you two things, provide Aiv- with mobile-linux commit access which is done already, and about new branches in kernel.git |
| 14:07.28 | tmzt | 16:42 < tmzt> trout axis |
| 14:07.28 | tmzt | 16:42 < tmzt> look at the board |
| 14:07.29 | tmzt | 16:42 < tmzt> you had it |
| 14:07.29 | tmzt | 16:43 < dcordes> tmzt, thanks. yea I've spotted the MOUSE lines but assumed the |
| 14:07.32 | tmzt | <PROTECTED> |
| 14:07.34 | tmzt | 16:44 < tmzt> .info.func = gpio_event_axis_func, |
| 14:07.37 | tmzt | 16:44 -!- iGama [n=igama@ubuntu/member/igama] has joined #android |
| 14:07.39 | tmzt | 16:44 < tmzt> what google did is a little unusual, it's odd to see Pavel's |
| 14:07.42 | tmzt | <PROTECTED> |
| 14:07.45 | tmzt | 16:44 < tmzt> they tried to be so generic with the gpio_ drivers |
| 14:07.47 | tmzt | 16:44 < tmzt> ah, it's EV_REL |
| 14:07.50 | tmzt | 16:45 < tmzt> like a mouse |
| 14:08.18 | Kevin2 | dcordes: By agreement, pH5 approves the mobile-linux access. |
| 14:08.54 | dcordes | Kevin2, didn't dzo approve it for Aiv- ? |
| 14:09.13 | dcordes | Kevin2, as I've seen on the project site, he's listed as maintainer now |
| 14:09.41 | Aiv- | dcordes i have access, and dzo told me that i can create new branch |
| 14:10.10 | Aiv- | but tmzt don't want to create new branch specially for kovsky |
| 14:10.21 | dcordes | yea because there's no reason |
| 14:10.27 | Aiv- | that's right |
| 14:10.36 | Aiv- | so now we have all to work and develop |
| 14:10.41 | dcordes | yep |
| 14:11.09 | Aiv- | dcordes btw i run startgps, and after few debug lines it's crashed |
| 14:12.26 | dcordes | Aiv-, our tree doesn't know the vogue rpc calls |
| 14:12.27 | Kevin2 | dcordes: Yes. dzo is also a maintainer. I'm okay if he wishes to create accounts. Since I'm not involved that much, I just feel I shouldn't be creating accounts. |
| 14:13.32 | Aiv- | dcordes ahh |
| 14:14.22 | dcordes | Kevin2, ok then. can we add a new branch that tracks android-msm-2.6.29 branch at git://android.git.kernel.org/kernel/msm.git ? |
| 14:14.44 | Kevin2 | It sounds okay to me. |
| 14:16.28 | GNUtoo | so no one knows kdrive/Xfbdev? |
| 14:16.42 | tmzt | what's the questions? |
| 14:17.47 | dcordes | GNUtoo, you have two problems, cursor moves to slow, click is only possible while doing the up movement, right |
| 14:18.02 | GNUtoo | yes |
| 14:18.13 | GNUtoo | the first I'd like to solve is the cursor moving too slow |
| 14:18.50 | dcordes | I think the click problem must be solved in arch/arm/mach-msm/board-trout.c and the slow cursor in userspace/X |
| 14:19.00 | GNUtoo | yes indeed |
| 14:19.14 | GNUtoo | but I don't know the good option for Xfbdev to solve the speed problem |
| 14:19.19 | GNUtoo | the manual isn't great |
| 14:21.58 | tmzt | speed problem? |
| 14:22.41 | tmzt | evtest should help |
| 14:22.45 | tmzt | what driver are you using? |
| 14:22.48 | GNUtoo | mouse |
| 14:22.54 | tmzt | oh |
| 14:22.55 | tmzt | try evdev |
| 14:23.00 | GNUtoo | ok thakns a lot |
| 14:24.48 | GNUtoo | -mouse evdev,/dev/input/mouse1 didn't work |
| 14:25.03 | GNUtoo | the trackball doesn't work anymore with that |
| 14:25.03 | tmzt | -mouse evdev,,device=/dev/input/mouse1 |
| 14:25.07 | tmzt | and it shouldn't be mouse |
| 14:25.12 | GNUtoo | ok |
| 14:25.12 | tmzt | it should /dev/input/eventX |
| 14:25.50 | dcordes | I use hexdump /dev/input/eventX to probe for the right one |
| 14:26.35 | GNUtoo | I use cat |
| 14:32.10 | GNUtoo | is there a manual? |
| 14:32.13 | GNUtoo | a complete one? |
| 14:32.44 | tmzt | manual for what? |
| 14:32.47 | *** join/#htc-linux leviathan (n=quassel@c-82-192-226-27.customer.ggaweb.ch) |
| 14:32.55 | GNUtoo | Xfbdev |
| 14:32.55 | tmzt | kdrive isn't really even supported |
| 14:33.09 | GNUtoo | ouch...should I change for xorg? |
| 14:33.36 | GNUtoo | ARGS="$ARGS -mouse evdev,1,,device=/dev/input/event2" ;; works |
| 14:33.36 | tmzt | no |
| 14:33.39 | GNUtoo | but still slow |
| 14:33.39 | tmzt | this is faster |
| 14:33.42 | GNUtoo | ok |
| 14:34.00 | GNUtoo | so now that I've switched to evdev |
| 14:34.03 | tmzt | I hope we can make it even faster, but still researching that |
| 14:34.06 | GNUtoo | how do I accelerate the mouse |
| 14:34.30 | GNUtoo | accelerate what? |
| 14:34.32 | tmzt | you will have to look at X source |
| 14:34.37 | GNUtoo | :( |
| 14:34.59 | GNUtoo | there is some options |
| 14:35.02 | GNUtoo | I can set optionsx |
| 14:35.14 | GNUtoo | so I look at the source for finding the options? |
| 14:35.30 | tmzt | yes |
| 14:35.40 | dcordes | Xfbdev configuration and adjustment make me wanna puke a bit |
| 14:35.55 | *** join/#htc-linux thebishop (n=thebisho@c-68-80-81-99.hsd1.pa.comcast.net) |
| 14:37.46 | dcordes | brb |
| 14:39.59 | tmzt | GNUtoo: http://cgit.freedesktop.org/xorg/xserver/tree/hw/kdrive/src/kinput.c?h=dri2-swapbuffers&id=c69e1eed41b64ffe96e765f70002e70232bcff39 |
| 14:40.29 | GNUtoo | I was in the source but I didn't found a main |
| 14:40.31 | GNUtoo | thanks a lot |
| 14:41.48 | tmzt | hw/kdrive is the main |
| 14:41.56 | tmzt | but it uses parts from the rest of tree |
| 14:42.12 | GNUtoo | ok thanks |
| 14:43.59 | tmzt | I don't velive kdrive uses module tree though |
| 14:44.05 | tmzt | but I'm not sure |
| 14:47.01 | tmzt | oh, it must be related to dpi and mouse accel |
| 14:47.05 | tmzt | it thinks it's a mouse |
| 14:47.35 | *** join/#htc-linux dcordes-kais (n=dcordes-@ip-77-24-34-64.web.vodafone.de) |
| 14:49.03 | tmzt | <PROTECTED> |
| 14:49.09 | tmzt | I ownder if flasgs comes from kernel |
| 14:49.56 | *** part/#htc-linux BUGabundo (n=bugabund@ubuntu/member/bugabundo) |
| 14:56.06 | GNUtoo | I'll try rawcoord for the touchscreen |
| 14:57.00 | GNUtoo | but I didn't find something for acceletating the pointer |
| 14:57.19 | tmzt | well, how many buttons are you using? |
| 14:57.23 | GNUtoo | 1 |
| 14:57.28 | tmzt | this code is so hard to read |
| 14:57.55 | GNUtoo | mmm there should be some X program to accelerate the pointer no? |
| 14:57.59 | GNUtoo | like kde or gnome do |
| 14:57.59 | GNUtoo | ? |
| 14:58.04 | GNUtoo | here I've illume |
| 14:58.10 | GNUtoo | so enlightment |
| 14:59.10 | tmzt | <PROTECTED> |
| 14:59.15 | tmzt | try xinput |
| 14:59.17 | GNUtoo | thanks |
| 14:59.23 | GNUtoo | as driver? |
| 15:00.06 | tmzt | no, that's a user utility |
| 15:00.11 | tmzt | but I don't know if kdrive suports it |
| 15:00.32 | GNUtoo | ok thanks a lot |
| 15:02.50 | GNUtoo | I'll bitbake it |
| 15:05.50 | tmzt | <PROTECTED> |
| 15:05.54 | tmzt | hah, ok |
| 15:06.03 | tmzt | so they do come from kernel |
| 15:06.07 | tmzt | (the flags) |
| 15:06.35 | tmzt | no |
| 15:06.37 | tmzt | wrongs ones |
| 15:06.56 | GNUtoo | ok |
| 15:12.41 | tmzt | #if 0 pScreen->backingStoreSupport = Always; miInitializeBackingStore (pScreen); |
| 15:12.44 | tmzt | #endif |
| 15:12.49 | tmzt | interesting, something else to look into |
| 15:24.54 | GNUtoo | is installing xinput |
| 15:27.43 | tmzt | actually, you probably want acceleration |
| 15:28.26 | GNUtoo | done |
| 15:28.30 | *** join/#htc-linux MethoS- (n=clemens@host-091-097-242-041.ewe-ip-backbone.de) |
| 15:28.32 | GNUtoo | but what should I do now? |
| 15:28.33 | tmzt | anthing? |
| 15:28.38 | tmzt | run xinput |
| 15:28.40 | tmzt | xinput list |
| 15:28.44 | GNUtoo | yes but with what args |
| 15:28.45 | GNUtoo | ok |
| 15:28.51 | GNUtoo | I've mouse touchscreen etc... |
| 15:28.56 | GNUtoo | I'll pastebin it |
| 15:29.26 | tmzt | I think we should add some new drivers, including a dpad->mouse mode |
| 15:29.27 | GNUtoo | http://pastebin.com/m4ec01115 |
| 15:29.29 | tmzt | and this trackpad |
| 15:29.56 | tmzt | xipnut could be used to enable/disable it |
| 15:30.00 | *** join/#htc-linux dilledalle (n=quassel@N246-145.surffi.net) |
| 15:33.01 | GNUtoo | ah ok ). |
| 15:33.02 | GNUtoo | xinput set-ptr-feedback |
| 15:33.12 | GNUtoo | s/).// |
| 15:34.11 | *** join/#htc-linux juke71 (n=quassel@N246-145.surffi.net) |
| 15:34.31 | tmzt | what is this? |
| 15:36.09 | tmzt | xinput list-prop 3 |
| 15:36.35 | GNUtoo | https://forum.fukt.bsnet.se/cgi-bin/man/man2html?xinput+1 |
| 15:36.42 | tmzt | accleration is not exposed |
| 15:36.49 | tmzt | interesting |
| 15:36.53 | *** join/#htc-linux juke71 (n=quassel@N246-145.surffi.net) |
| 15:36.56 | GNUtoo | ouch!!! |
| 15:37.03 | GNUtoo | the pointer is slow as hell |
| 15:37.09 | tmzt | does it work? |
| 15:37.15 | GNUtoo | no |
| 15:37.18 | *** join/#htc-linux juke71 (n=quassel@N246-145.surffi.net) |
| 15:37.23 | tmzt | I think your ranges in kernel are wrong |
| 15:37.29 | tmzt | is it -1..0? |
| 15:37.31 | GNUtoo | no |
| 15:37.36 | GNUtoo | I've restored to default |
| 15:37.38 | GNUtoo | z |
| 15:37.40 | tmzt | wait, you could just report any size jump you want |
| 15:37.47 | tmzt | well |
| 15:37.47 | GNUtoo | ??? |
| 15:37.51 | GNUtoo | as Z |
| 15:37.54 | GNUtoo | yes |
| 15:37.57 | tmzt | Z? |
| 15:38.19 | tmzt | I don't know how g1 trackball works |
| 15:38.19 | GNUtoo | 11 sec to get from bottom to top of the screen |
| 15:38.19 | tmzt | hold on |
| 15:38.26 | GNUtoo | ok z was for the touchscreen |
| 15:38.29 | GNUtoo | it was the pressure |
| 15:38.37 | tmzt | well, we should look at how android does it |
| 15:38.40 | GNUtoo | ok |
| 15:38.45 | tmzt | in theird window manager |
| 15:38.50 | GNUtoo | btw we should get a roadmap |
| 15:38.52 | GNUtoo | ok |
| 15:38.53 | tmzt | but isn't touchscreen ok? |
| 15:38.56 | GNUtoo | no |
| 15:39.05 | tmzt | why is that? |
| 15:39.05 | GNUtoo | touchscreen clickes twice |
| 15:39.09 | GNUtoo | don't know |
| 15:39.13 | tmzt | what about with evdev? |
| 15:39.18 | GNUtoo | and leviathan is not there |
| 15:39.21 | tmzt | how new is your X server? |
| 15:39.22 | GNUtoo | I'll try |
| 15:39.27 | GNUtoo | I'll check |
| 15:40.10 | GNUtoo | lol |
| 15:40.11 | GNUtoo | -a # mouse acceleration (pixels) |
| 15:40.32 | tmzt | hwere? |
| 15:40.42 | tmzt | that's not for kdrive I think |
| 15:40.48 | GNUtoo | it's for Xfbdev |
| 15:40.54 | tmzt | oh, ok |
| 15:40.56 | tmzt | try that |
| 15:40.59 | GNUtoo | ok |
| 15:42.45 | *** join/#htc-linux dcordes (n=luke-g@unaffiliated/dcordes) |
| 15:43.26 | GNUtoo | no change |
| 15:44.10 | tmzt | with evdev or tslib? |
| 15:45.52 | tmzt | trying to see how kdrive builds again |
| 15:46.25 | tmzt | Kevin2: we should consider a userspace git tree as well, it looks like it could be useful for some of this stuff |
| 15:46.54 | tmzt | or is that just not part of the mobile-linux project |
| 15:48.06 | GNUtoo | evdev |
| 15:48.22 | GNUtoo | tmzt, no we should push to oe |
| 15:48.29 | GNUtoo | tmzt, for the patches |
| 15:48.33 | GNUtoo | and make a repo for the rest |
| 15:48.59 | GNUtoo | I should do it but I have issued with oe's git-native |
| 15:49.02 | tmzt | well, that's a possibility |
| 15:49.05 | GNUtoo | so I can't push the kernel |
| 15:49.07 | Kevin2 | tmzt: That's not my call. It's pretty easy to setup a new tree, but it's also very easy to create a new project. I'd say GNUtoo has a good point also. |
| 15:49.07 | *** join/#htc-linux |juke71| (n=Somebody@N246-145.surffi.net) |
| 15:49.12 | GNUtoo | ok |
| 15:49.19 | GNUtoo | I've oe commit access |
| 15:49.27 | tmzt | well, ok |
| 15:49.37 | GNUtoo | so if that's usefull I can push things |
| 15:49.39 | tmzt | I would hope are stuff will build on debian/gentoo/etc. though |
| 15:49.49 | GNUtoo | but the problem is that I'm trying to do too much things at once today |
| 15:49.49 | tmzt | but if we could get some patches against oe's kdrive |
| 15:49.56 | GNUtoo | ok |
| 15:50.04 | GNUtoo | 1)put in oe |
| 15:50.08 | tmzt | I know dcordes had started on this, building machine defs for oe |
| 15:50.08 | GNUtoo | 2)push upstream |
| 15:50.17 | GNUtoo | ok |
| 15:50.24 | GNUtoo | I've machine def and msm tune |
| 15:50.30 | GNUtoo | I've done it |
| 15:50.37 | tmzt | ah, ok |
| 15:50.40 | dcordes | correct |
| 15:50.43 | GNUtoo | but I had problems with kernel fetching |
| 15:50.56 | GNUtoo | dcordes, does your kenrel recipe work? |
| 15:50.57 | tmzt | we can always build kernel seperate |
| 15:51.03 | GNUtoo | I fetched openmoko-msm-2.6.29 |
| 15:51.16 | dcordes | GNUtoo, yes |
| 15:51.27 | GNUtoo | wow |
| 15:51.27 | dcordes | GNUtoo, but it uses linuxtogo git, no github or so |
| 15:51.31 | GNUtoo | ah ok |
| 15:51.33 | dcordes | they |
| 15:51.50 | dcordes | but dream should go linuxtogo kernel.git anyway |
| 15:51.54 | GNUtoo | I've issues with github's openmoko-2.6.29 + git-native of oe |
| 15:51.54 | tmzt | well, let's define some things then |
| 15:51.57 | GNUtoo | yes |
| 15:52.01 | GNUtoo | when will it be imported? |
| 15:52.10 | GNUtoo | so I could make recipe and push |
| 15:52.13 | tmzt | dcordes: can we get a wiki page for Roadmap or Plan or soemthing? |
| 15:52.21 | GNUtoo | tmzt, we've just got one |
| 15:52.29 | GNUtoo | http://www.htc-linux.org/wiki/index.php?title=Dream |
| 15:52.52 | tmzt | good double meaning :) |
| 15:53.28 | tmzt | well, not quite the structure I can work with |
| 15:53.42 | GNUtoo | ok what do you need then? |
| 15:53.43 | tmzt | the top stuff is known, I hope we can get beyoind it |
| 15:53.44 | GNUtoo | trac? |
| 15:53.49 | GNUtoo | ok |
| 15:53.50 | tmzt | this is fine |
| 15:54.13 | tmzt | but I think using g1 as a baseline will help a lot for the project |
| 15:54.31 | GNUtoo | ok |
| 15:54.36 | tmzt | we can "commit" to using/testing the same userspace on both |
| 15:54.43 | tmzt | and hopefully kernel at some point |
| 15:55.06 | tmzt | this is the approach taken by wine/reactos |
| 15:55.16 | tmzt | they test everything on windows, more than one version |
| 15:55.31 | tmzt | build up tests |
| 15:56.10 | tmzt | okay, if oe is are build system, we should start with a minimal task |
| 15:56.37 | tmzt | dcordes: what did you have working for angstrom? |
| 15:57.33 | tmzt | we should assume that when we boot android, it will be exec'd from oe init |
| 15:57.53 | GNUtoo | ok you want free android too? |
| 15:58.00 | GNUtoo | me too |
| 15:58.05 | tmzt | if you can look at wing-linux git on wing-linux on sf |
| 15:58.15 | GNUtoo | what's that? |
| 15:58.17 | tmzt | you'll see what darkstar did |
| 15:58.24 | tmzt | sf.net/projects/wing-linux |
| 15:58.27 | tmzt | click on a commit |
| 15:58.28 | tmzt | then tree |
| 15:59.04 | tmzt | that's being developed for omap, but we can share most of the things with it |
| 15:59.29 | GNUtoo | ok |
| 15:59.29 | tmzt | the nice part is that it allows patching android git repos using bitbake |
| 16:00.45 | tmzt | well, we will need: |
| 16:00.58 | tmzt | a libhardware that works with 2.6 standard interfaces |
| 16:01.17 | *** join/#htc-linux gouki (n=gouki@fsf/member/ubuntu.member.gouki) |
| 16:01.19 | GNUtoo | look at opnemoko then |
| 16:01.21 | tmzt | input event support with scaled coordinates, but the in kenrel approach using pointercal should work |
| 16:01.34 | tmzt | wich, the andorid project for fr? |
| 16:01.41 | GNUtoo | yes |
| 16:01.44 | tmzt | ok |
| 16:01.47 | GNUtoo | for instance they use alsa |
| 16:01.53 | GNUtoo | or more standard things |
| 16:01.59 | tmzt | we need to be able to transition to that |
| 16:02.04 | GNUtoo | but they implement some android driver in the kenrel too |
| 16:02.08 | tmzt | probably using #defines, such as |
| 16:02.12 | GNUtoo | ok |
| 16:02.17 | tmzt | LIBHARDWARE_AUDIO_ALSA |
| 16:02.24 | tmzt | LIBHARDWARE_AUDIO_MSM_PCM_CONFIG |
| 16:02.31 | GNUtoo | ok |
| 16:02.39 | GNUtoo | it's defined in the build system |
| 16:02.49 | tmzt | great |
| 16:04.02 | GNUtoo | ~/koolu/vendor/neo/freerunner $ cat BoardConfig.mk | grep ALSA => BOARD_USES_ALSA_AUDIO:= true |
| 16:04.14 | tmzt | hah, greate |
| 16:08.22 | tmzt | gouki: I would suggest not upgrading to htc firmware |
| 16:08.45 | tmzt | gouki: to preserve root access |
| 16:09.32 | GNUtoo | what should I do now for the input,touchscreen is unusable,kernel patch did nothing,calibration tool is suspect and acceleration of the trackball doesn't work |
| 16:10.09 | tmzt | there might be bad kernel flags |
| 16:10.13 | GNUtoo | ah ok |
| 16:10.16 | tmzt | but you really don't need tslib |
| 16:10.23 | GNUtoo | ok I'll try evdev |
| 16:10.30 | GNUtoo | unless I want gimp |
| 16:10.43 | GNUtoo | but who needs gimp on a 320x480 screen |
| 16:10.43 | tmzt | pressure should work without it |
| 16:10.46 | tmzt | without tslib |
| 16:10.49 | GNUtoo | ok |
| 16:10.53 | GNUtoo | I'll try evdev |
| 16:13.35 | GNUtoo | I get no event with evdev |
| 16:13.40 | GNUtoo | I'll try mouse |
| 16:13.44 | GNUtoo | s/try/retry |
| 16:13.48 | tmzt | no events? |
| 16:13.55 | GNUtoo | that is to say: |
| 16:13.58 | tmzt | can you paste Xfbdev output? |
| 16:14.01 | GNUtoo | it doesn't move on the screen |
| 16:14.04 | GNUtoo | log? |
| 16:14.08 | tmzt | yes |
| 16:15.35 | GNUtoo | Pointer option key () of value ((null)) not assigned! |
| 16:15.46 | tmzt | hold on |
| 16:16.46 | tmzt | paste your cmdline |
| 16:16.49 | GNUtoo | ok |
| 16:16.50 | tmzt | command line |
| 16:17.28 | GNUtoo | ARGS="$ARGS -mouse evdev,1,,device=/dev/input/event2 -mouse evdev,1,,device=/dev/input/event1";; |
| 16:17.34 | GNUtoo | or |
| 16:17.36 | tmzt | emulatemiddle, noemulatemiddle, transformcoord, rawcoord, device, protocol |
| 16:17.39 | tmzt | that's all it supports |
| 16:17.42 | GNUtoo | yes |
| 16:17.43 | tmzt | too many , |
| 16:18.11 | tmzt | add the noemulatemiddle |
| 16:18.51 | GNUtoo | I'll add -mouse evdev,1,,device=/dev/input/mouse0 |
| 16:18.52 | tmzt | it's driver,buttons(,option)+ |
| 16:18.56 | tmzt | no |
| 16:19.01 | GNUtoo | ok |
| 16:19.11 | tmzt | -mouse evdev,1,device=/dev/input/eventX |
| 16:19.17 | GNUtoo | s/evdev/mouse |
| 16:19.38 | tmzt | whicever, it sees an empty option |
| 16:20.34 | GNUtoo | so what should I try |
| 16:20.41 | GNUtoo | mouse /dev/mouse0 works too badly |
| 16:21.10 | GNUtoo | when I'm on the midle the mouse pointer is on the bottom |
| 16:23.00 | tmzt | we need to patch X |
| 16:23.05 | GNUtoo | ok |
| 16:23.08 | tmzt | do you know how oe builds kdrive? |
| 16:23.13 | GNUtoo | more or less |
| 16:23.18 | GNUtoo | I'll look |
| 16:23.32 | tmzt | evdev shoudl work though |
| 16:23.43 | tmzt | I think I need to see how it determines if it's abs or rel, etc. |
| 16:23.59 | tmzt | how small can oe be now? |
| 16:24.09 | tmzt | what does it have to pull down? |
| 16:24.24 | GNUtoo | there are several patches |
| 16:24.32 | GNUtoo | ok |
| 16:24.45 | GNUtoo | small in what sense? |
| 16:24.49 | GNUtoo | size on hdd |
| 16:24.50 | GNUtoo | ? |
| 16:24.56 | GNUtoo | speed of parsing? |
| 16:25.12 | *** join/#htc-linux madCoder` (n=madcoder@c-71-225-60-178.hsd1.pa.comcast.net) |
| 16:25.14 | GNUtoo | because you can rm_work to gain a lot of size |
| 16:25.33 | tmzt | sometihng like the MokoMakefile |
| 16:25.42 | GNUtoo | ah ok |
| 16:25.43 | tmzt | I mean, we might need something like that |
| 16:25.46 | GNUtoo | so try shr's makefile |
| 16:25.52 | tmzt | but I think it hides a little too much |
| 16:26.06 | tmzt | yeah, size of downloading and on hdd |
| 16:26.10 | GNUtoo | yes |
| 16:26.45 | GNUtoo | ah you need the whole git of oe |
| 16:26.45 | GNUtoo | I'll look for the size |
| 16:27.10 | tmzt | it the git stuff smart enough to only pull one rev? not clone |
| 16:27.14 | GNUtoo | df -h gives 419M |
| 16:27.19 | tmzt | rev and dependencies |
| 16:27.20 | gouki | tmzt, thanks |
| 16:27.29 | GNUtoo | mmm I don't know |
| 16:27.32 | GNUtoo | maybe ask #git |
| 16:27.49 | GNUtoo | you're on 3g or something with GB limits? |
| 16:29.28 | tmzt | for now |
| 16:29.31 | GNUtoo | ok |
| 16:31.38 | GNUtoo | mmm...maybe with a pre-built toolchain? |
| 16:31.44 | GNUtoo | + the sources |
| 16:31.57 | GNUtoo | or maybe you should find a wifi caffe |
| 16:32.32 | GNUtoo | mmm bad idea |
| 16:32.37 | tmzt | I have a toolchain |
| 16:32.37 | GNUtoo | there seem to be a lot of deps |
| 16:32.39 | GNUtoo | ok |
| 16:32.54 | tmzt | there always are |
| 16:33.08 | GNUtoo | you could use your toolchain in oe but you would still need to download oe+ deps and kdrive sources |
| 16:40.03 | GNUtoo | the last column in ts_print_raw is the pression? |
| 16:40.14 | GNUtoo | I bet there is double click because: |
| 16:40.25 | tmzt | I think so |
| 16:40.26 | GNUtoo | 1)lot of incomming values |
| 16:40.33 | GNUtoo | 2)pression = 0 |
| 16:40.39 | GNUtoo | 3)pression !=0 |
| 16:40.45 | GNUtoo | the 3 does a second click no? |
| 16:40.54 | GNUtoo | that's for one touch and removing the finger |
| 16:41.18 | GNUtoo | with z?0:1 it does the same |
| 16:41.25 | GNUtoo | oops |
| 16:41.56 | GNUtoo | z?1:0 |
| 16:42.05 | GNUtoo | anyway it does the same: |
| 16:42.09 | GNUtoo | lot of 1 |
| 16:42.11 | GNUtoo | 0 |
| 16:42.12 | GNUtoo | 1 |
| 16:42.14 | GNUtoo | and nothing |
| 16:42.23 | GNUtoo | so the second 1 counts for a click no? |
| 16:42.32 | tmzt | second 1? |
| 16:42.34 | tmzt | I would think so |
| 16:42.41 | GNUtoo | ok |
| 16:42.45 | GNUtoo | how can we remove that? |
| 16:43.00 | tmzt | I don't know what's causing it |
| 16:43.05 | GNUtoo | hardware? |
| 16:43.38 | tmzt | I don't think so |
| 16:43.46 | GNUtoo | strange then |
| 16:43.46 | tmzt | but your hardware doesn't support pressure |
| 16:44.25 | tmzt | disable reporting pressure and try mouse/evdev again |
| 16:44.31 | GNUtoo | I've a theory |
| 16:44.55 | GNUtoo | when you're close enough to the screen but don't touch it it reports a pressure of ~10/15 |
| 16:44.57 | GNUtoo | maybe it's that |
| 16:45.04 | GNUtoo | ok thanks a lot |
| 16:45.14 | tmzt | well, ok |
| 16:45.25 | GNUtoo | it supports pressure |
| 16:45.46 | GNUtoo | the more I press,bigger is the pressure value |
| 16:46.05 | GNUtoo | or is it unrelated |
| 16:46.19 | GNUtoo | ok I'll try mouse + removing the pressure |
| 16:46.23 | GNUtoo | I didn't try it yet |
| 16:46.35 | GNUtoo | I tried removing the pressure+ tslib |
| 16:46.55 | GNUtoo | all I have to do is to replace z by (z?1:0) |
| 16:47.05 | GNUtoo | or should I change some others things? |
| 16:50.18 | tmzt | it's capc. that is correct |
| 16:50.19 | tmzt | tslib won't work at all though |
| 16:51.48 | GNUtoo | ok |
| 16:52.05 | tmzt | I mean don't report pressure from kernel |
| 16:52.07 | tmzt | at all |
| 16:52.22 | tmzt | just ABS_X, ABS_Y, EV_SYNC |
| 16:52.55 | *** join/#htc-linux Zy (n=zy@84.238.80.215) |
| 16:55.20 | GNUtoo | ok |
| 16:55.44 | tmzt | also remove the setbit or set_param call for ABS_PRESSURE |
| 16:56.26 | GNUtoo | ok |
| 16:58.18 | GNUtoo | I also comment ABS_TOOL_WIDTH, and finger2 |
| 16:59.51 | GNUtoo | thanks a lot |
| 17:00.11 | GNUtoo | re-swtiching to 2.6.30,cleaning,and building |
| 17:00.21 | tmzt | mrproper? |
| 17:00.25 | tmzt | not needed |
| 17:00.32 | GNUtoo | no make clean |
| 17:00.39 | GNUtoo | because I switch branch |
| 17:00.43 | GNUtoo | ma |
| 17:00.43 | tmzt | oh |
| 17:00.44 | tmzt | ok |
| 17:00.44 | GNUtoo | oops |
| 17:01.03 | GNUtoo | and I'm not shure the usual time based makefile work if I switch branch |
| 17:01.04 | tmzt | which are you working on? |
| 17:01.14 | GNUtoo | openmoko-msm-2.6.29 |
| 17:01.17 | GNUtoo | branch 2.6.30 |
| 17:01.27 | tmzt | if it's mtime it should be fine |
| 17:01.29 | GNUtoo | I was on master |
| 17:01.31 | GNUtoo | ok |
| 17:01.35 | GNUtoo | it is |
| 17:01.35 | tmzt | why were you on master? |
| 17:01.40 | GNUtoo | to try |
| 17:01.42 | GNUtoo | to get wifi |
| 17:01.59 | GNUtoo | but it didn't pass the beginning of the compilation cause a header was missing |
| 17:02.06 | GNUtoo | and I was on ARCH=arm |
| 17:02.08 | GNUtoo | and on msm |
| 17:02.25 | tmzt | did you get OE to build kdrive? |
| 17:03.36 | GNUtoo | not yet |
| 17:03.45 | tmzt | I can't figure out how it calculates flags or the b parameter |
| 17:03.47 | GNUtoo | my laptop is slow |
| 17:03.53 | tmzt | if I knew that |
| 17:04.05 | GNUtoo | and the internet connection is at the wifi caffe |
| 17:04.09 | GNUtoo | so not 24h/24 |
| 17:04.16 | GNUtoo | and now I must go...sorry |
| 17:04.45 | GNUtoo | bye |
| 17:12.07 | *** join/#htc-linux surgex0 (i=surge@72.75.233.50) |
| 17:14.06 | *** join/#htc-linux darkstar62_away (n=darkstar@97-126-106-35.tukw.qwest.net) |
| 17:48.09 | *** join/#htc-linux stickboy (n=anonymou@cpe-74-68-145-150.nyc.res.rr.com) |
| 18:07.34 | *** join/#htc-linux stickboy (n=anonymou@cpe-74-68-145-150.nyc.res.rr.com) |
| 18:10.34 | *** join/#htc-linux MethoS (n=clemens@host-091-097-240-045.ewe-ip-backbone.de) |
| 18:27.03 | *** join/#htc-linux googleman (n=sdfddd@41.105.96.15) |
| 18:31.51 | *** join/#htc-linux nebi (n=nebi@c83-255-51-86.bredband.comhem.se) |
| 18:49.13 | *** join/#htc-linux vilord (n=jesse@ip68-227-61-169.lv.lv.cox.net) |
| 19:24.13 | *** join/#htc-linux Zy (n=zy@84.238.80.215) |
| 19:29.41 | *** join/#htc-linux MethoS-- (n=clemens@host-091-096-209-253.ewe-ip-backbone.de) |
| 20:03.44 | *** join/#htc-linux verizonguy (n=verizong@c-24-63-231-35.hsd1.ct.comcast.net) |
| 20:43.54 | *** join/#htc-linux MethoS- (n=clemens@host-091-096-209-253.ewe-ip-backbone.de) |
| 20:44.13 | *** join/#htc-linux stefan_schmidt (n=stefan@p5B036A0D.dip.t-dialin.net) |
| 21:42.43 | *** join/#htc-linux Captnoord (n=Captnoor@dc5147a47b.adsl.wanadoo.nl) |
| 22:08.52 | *** join/#htc-linux MethoS- (n=clemens@dyndsl-085-016-165-101.ewe-ip-backbone.de) |
| 22:18.23 | *** join/#htc-linux alphaone (n=alphaone@2001:638:602:af01:0:0:0:1) [NETSPLIT VICTIM] |
| 22:18.23 | *** join/#htc-linux jos (i=Jos@Lithium.Jossie90.info) [NETSPLIT VICTIM] |
| 22:19.50 | *** join/#htc-linux Venny (n=Travis@h216-45-127-2.dynamic.platinum.ca) |
| 22:20.21 | Venny | Can someone in here help me with a sync problem im having with my phone? or direct me to the right IRC room lool |
| 22:25.54 | *** join/#htc-linux MethoS- (n=clemens@dyndsl-085-016-165-101.ewe-ip-backbone.de) |
| 23:43.30 | *** join/#htc-linux BabelO (n=fcr@unaffiliated/babelo) |