**** BEGIN LOGGING AT Fri Jun 15 02:59:59 2012
Jun 15 05:07:57 <PaulFertser> radekp_: lindi-: after upgrading Debian udev forced me to go away from my "trusty" 2.6.29 and now i'm on 2.6.34-qtmoko-v34.
Jun 15 05:08:45 <PaulFertser> radekp_: lindi- quite often it seems to misdetect usd size on resume forcing me to reboot :(
Jun 15 06:01:50 <prahal> jeepingben: yes ( I  could not tell what it means yet though)
Jun 15 06:07:23 <PaulFertser> radekp_: coupled with the fact it takes about 10 minutes to boot that made my beloved fr considerably more hard to use...
Jun 15 06:09:20 <PaulFertser> I know that's not a proper bugreport. But... Probably i'm just using a wrong kernel and sholud be on 3.something?
Jun 15 06:15:19 <prahal> PaulFertser: using kernel-image-2.6.39.4_2.6.39-r12_om_gta02.ipk . works well (at least w/o the issues you encountered)
Jun 15 06:16:03 <prahal> though using shr bitbake .
Jun 15 06:17:32 <PaulFertser> prahal: thanks, i'll consider it.
Jun 15 06:19:14 <prahal> with debian I was using 2.6.34 also worked fine
Jun 15 06:21:27 <prahal> my main issue was with libcap-ng applications crashing (bluez for one) ... I documented how I worked around this one as told in debian bug 631807
Jun 15 06:23:55 <prahal> this 2.6.34 might be more appropriate for debian freerunner ... uImage.bin-2.6.34-20101212.git049b71de it is and I believe it was grabbed by the debian installer
Jun 15 06:44:19 <lindi-> PaulFertser: 3.6.34 is the kernel to use still
Jun 15 06:44:55 <lindi-> PaulFertser: 2.6.39 has these suspend issues
Jun 15 06:47:10 <PaulFertser> lindi-: i've read some bugreport of yours related to uSD. But i think my issues looks a bit different. I sometimes get something like "attempt to read beyond the end of the device".
Jun 15 07:16:11 <jake42> jeepingben: prahal: the orange light should indicate if the battery is charged
Jun 15 07:16:48 <jake42> the blue one if the battery is fully charged
Jun 15 07:19:35 <jake42> though at some point there is a bug
Jun 15 07:19:54 <jake42> and it mixes up the states
Jun 15 07:47:46 <prahal> jake42: I found out that the red one is set bu u-boot when it start its sequence .
Jun 15 07:48:51 <prahal> I only get other than red before u-boot start
Jun 15 07:49:20 <prahal> though I am running 3.4.0 from shr bitbake ... 
Jun 15 07:50:27 <prahal> by the way is there a page with the scripts to rebuild bootargs.scr boot.scr . I did with the gumstix ones
Jun 15 07:50:33 <prahal> the build scripts I mean
Jun 15 07:50:41 <prahal> mkimage -A arm -O linux -T script -C none -a 0 -e 0 -n "myscript" -d myscript.cmd boot.scr
Jun 15 07:51:24 <prahal> though the result differs in the headers (other than the label) ...
Jun 15 07:57:05 <morphis> prahal: not really
Jun 15 08:09:12 <CIA-76> freesmartphone.org: 03morphis 07morphis/multi-device * raff5090c5f8d 10cornucopia/fsogsmd/src/ (3 files in 2 dirs): fsogsmd: lib: remove access of theModem variable from call class
Jun 15 08:27:32 <prahal> the usb plugged on the fso battery is opposite on gta02 and gta04 (also the led power is off when usb is plugged and red(gta04)/orange(gtaà2) when unplugged) . ie the usb icon on the battery shows up when the cable is unplugged and is discarded when the cable is plugged
Jun 15 08:28:59 <prahal> seems cosmetic though (the sys entries are ac offline usb online when usb cable is plugged ... might be that the fso battery and led is based on ac power
Jun 15 08:29:12 <prahal> but as there is no ac power ...
Jun 15 08:31:45 <morphis> prahal: for battery status we're reading energy_full, energy_now and capacity sysf entries in the fsodeviced kernel26_powersupply plugin
Jun 15 08:35:16 <prahal> thanks
Jun 15 08:36:25 <prahal> you mean there is a poll ? to not depend on hotplug events ?
Jun 15 08:37:29 <CIA-76> freesmartphone.org: 03morphis 07morphis/multi-device * r1f2bd6052cf2 10cornucopia/fsogsmd/src/lib/modem.vala: fsogsmd: lib: invert status update sending for NetworkTimeReport class
Jun 15 08:37:29 <CIA-76> freesmartphone.org: 03morphis 07morphis/multi-device * reb8ee12fe4ff 10cornucopia/fsogsmd/tests/ (testatcommand.vala testsmsstorage.vala): fsogsmd: test: adjust for recent changes to get tests working again
Jun 15 08:37:29 <CIA-76> freesmartphone.org: 03morphis 07morphis/multi-device * r82f52ccd2367 10cornucopia/fsogsmd/src/lib/modem.vala: fsogsmd: lib: completely remove theModem variable
Jun 15 08:39:06 <morphis> prahal: we're using uevents
Jun 15 08:39:16 <morphis> no polling
Jun 15 08:39:58 <prahal> ok
Jun 15 08:41:07 <prahal> still the led power and fso-battery ac indicator only trigger randomly (ie only after an unplug/plug sequence I did not managed to reproduce on purpose)
Jun 15 08:41:54 <prahal> or else it only works when charging (I am fully charged now)
Jun 15 08:42:53 <morphis> you mean the LED on the right side of the device?
Jun 15 08:43:02 <prahal> yes
Jun 15 08:43:07 <morphis> on GTA04?
Jun 15 08:43:50 <morphis> I am currently unsure who controls the LED in SHR
Jun 15 08:44:25 <morphis> when there is no updates for a change to energy_now there is maybe a bug in the kernel driver
Jun 15 08:46:26 <prahal> I do not know cozybit and derosier though I owe them a truck of beers for libertas_tf . Only missed a small line to fix an issue when fw failed to load and the module is unloaded otherwise is working great
Jun 15 08:46:48 <prahal> morphis: ok , yes GTA04
Jun 15 08:47:56 <prahal> seems reliable on gta02
Jun 15 08:49:06 <morphis> ok
Jun 15 09:38:17 <jake42> prahal: oh you've initally been talking about the AUX-Led (red)
Jun 15 09:38:34 <jake42> s/initally/initially/
Jun 15 09:38:34 <apt> jake42 meant: prahal: oh you've initially been talking about the AUX-Led (red)
Jun 15 09:39:34 <jake42> the one on the lower right side is called power button and the one on the upper left is called aux button
Jun 15 09:52:14 <CIA-76> freesmartphone.org: 03morphis 07morphis/multi-device * rfc5661d53a9f 10cornucopia/fsogsmd/src/ (3 files in 2 dirs): fsogsmd: dbus_service: rework to add modem management support
Jun 15 10:00:00 <CIA-76> SHR: 03Martin.Jansa 07shr-chroot * rf064768a4833 10/ (160 files in 24 dirs): system upgrade
Jun 15 10:00:19 <CIA-76> SHR: 03Martin.Jansa 07shr-chroot * r24ebf8453b18 10/ (4555 files in 165 dirs): system depclean
Jun 15 10:00:45 <CIA-76> SHR: 03Martin.Jansa 07shr-chroot * r476dce481b42 10/ (8271 files in 1111 dirs): system upgrade - perl-5.16.0
Jun 15 10:03:24 <jeepingben> it is the power led that is red when the phone is turned on.  It can be turned off by plugging and unplugging usb.  I can also echo 0>/sys/class/leds/gta04:red:power/brightness on startup
Jun 15 10:04:16 <jeepingben> I see no screen dimming except illume screensaver after ~10 minutes and I have not suspended successfully yet.
Jun 15 10:14:53 <prahal> jeepingben: running 3.4.0 kernel too ?
Jun 15 10:15:18 <prahal> ie I do not remind if suspend was working with the "stable" kernel
Jun 15 10:16:40 <prahal> also I have 3 state that can be reached on gta02 and gta04 ... "ac" (usb plugged) , charging (usb unplugged and the sys entry still at charging but change soon) and discharging (2 seconds after usb unplugged)
Jun 15 10:19:40 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * rd575e20ec43b 10cornucopia/fsogsmd/ (8 files in 6 dirs): fsogsmd: modem_hfp_hf: add initial support for a HFP HF modem type
Jun 15 10:19:40 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r851fe8f93315 10cornucopia/fsogsmd/src/lib/ (Makefile.am bluetooth.vala): 
Jun 15 10:19:40 <CIA-76> freesmartphone.org: fsogsmd: lib: add a bluetooth manager class for bluez
Jun 15 10:19:40 <CIA-76> freesmartphone.org: With this management class you can register different bluetooth profile handlers which will
Jun 15 10:19:40 <CIA-76> freesmartphone.org: be automatically associated with bluetooth devices handled by the bluez daemon if they
Jun 15 10:19:40 <CIA-76> freesmartphone.org: supports the registered profile.
Jun 15 10:19:41 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r22970aa33bb3 10cornucopia/fsogsmd/src/lib/bluetooth.vala: fsogsmd: lib: add interface definition for bluez handsfree related APIs
Jun 15 10:19:41 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r01d7548020a5 10cornucopia/fsogsmd/src/ (7 files in 2 dirs): fsogsmd: dbus_service: expose service registration method of our manager to the public
Jun 15 10:19:42 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r6b4ee29fa5cc 10cornucopia/fsogsmd/src/plugins/modem_hfp_hf/ (Makefile.am agent.vala plugin.vala profile.vala): fsogsmd: modem_hfp_hf: register bluez handsfree agent when device with HFP HF profile is found
Jun 15 10:19:42 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r6af70ea90509 10cornucopia/fsogsmd/src/plugins/modem_hfp_hf/agent.vala: fsogsmd: modem_hfp_hf: register handfree dbus agent with bluez
Jun 15 10:20:33 <prahal> I am a bit lost into the frameworkd ... odeviced call power_status (status) which thorugh dbus goes to frameworkd oeventsd triggers ... then  ?
Jun 15 10:20:34 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * rc7c72ff79ff9 10cornucopia/fsogsmd/ (4 files in 3 dirs): fsogsmd: modem_hfp_hf: uses dbus file descriptor passing feature for org.blue.HandsfreeAgent interface
Jun 15 10:20:34 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * ree000e994f8b 10cornucopia/fsogsmd/src/plugins/modem_hfp_hf/ (Makefile.am agent.vala modem.vala plugin.vala profile.vala): fsogsmd: modem_hfp_hf: rework to support new multi-device style core
Jun 15 10:20:34 <CIA-76> freesmartphone.org: 03morphis 07morphis/hfp * r5bdfacb0c2b3 10cornucopia/fsogsmd/src/plugins/dbus_service/device_servicemanager.vala: fsogsmd: dbus_service: accessing method via base identifier causes compile-time errors
Jun 15 10:22:11 <morphis> prahal: it's rather simple
Jun 15 10:22:18 <morphis> prahal: in SHR is fsodeviced
Jun 15 10:22:42 <morphis> which has a kernel26_powersupply plugin which implements the org.freesmartphone.Device.PowerSupply API
Jun 15 10:23:07 <morphis> it sents a dbus signal whenever something related to the power supply has changed
Jun 15 10:23:07 <prahal> this part works fine (ie it sends proper statuses )
Jun 15 10:23:10 <morphis> ok
Jun 15 10:23:15 <morphis> then there is oeventsd
Jun 15 10:23:26 <morphis> it's one of the last parts used from the old frameworkd in SHR
Jun 15 10:24:07 <prahal> calls self._trigger (status) in PowerStatusTrigger
Jun 15 10:24:13 <morphis> you have to look into the rules.yaml found in /etc/freesmartphone/conf/oevents
Jun 15 10:24:20 <prahal> ok
Jun 15 10:24:32 <morphis> prahal: yes thats the implementation of the trigger
Jun 15 10:24:40 <morphis> the connecting bits are in the rules.yaml
Jun 15 10:24:52 <morphis> there is a line with while: PowerStatus()
Jun 15 10:25:02 <morphis> filters: HasAttr(status, "charging")
Jun 15 10:25:15 <morphis> actions: SetLed("gta04_red_power", "light")
Jun 15 10:25:23 <morphis> thats where the magic happens
Jun 15 10:26:09 <prahal> hum it cannot work
Jun 15 10:26:16 <prahal> ie when charging the status is "ac"
Jun 15 10:27:13 <prahal> the status "charging only happens for a short while when unplugged ... ie ac is seen as offline but charging is still on because the kernel takes time to notice
Jun 15 10:27:59 <prahal> make me wonder why "ac" was included in the same entity than charging / full / ....
Jun 15 10:28:52 <prahal> there is a second bug ... when going from ac->charging the led turns red ... but when going from charging->discharging it does not change
Jun 15 10:29:35 <morphis> prahal: http://git.freesmartphone.org/?p=specs.git;a=blob_plain;f=html/org.freesmartphone.Device.PowerSupply.html;hb=HEAD#PowerStatus
Jun 15 10:31:08 <morphis> prahal: fell free to rework a bug with attached patch for this
Jun 15 10:31:45 <jeepingben> I am running shr-core 063 build (not sure which kernel that is).  I recognize the led behavior now as similar to gta02.  I guess the more important issue for me is that uboot leaves it on, so if I boot the phone outside away from power, the red led will stay lit
Jun 15 10:31:58 <prahal> ok then the issue is about the technology not behing filled by the kernel ... the log tells it takes the thing as "ac" as no technology was defined
Jun 15 10:32:21 <jeepingben> I put echo ->...brightness in one of my startup scripts (don't remember which one)
Jun 15 10:33:49 <morphis> prahal: yes
Jun 15 10:34:01 <morphis> prahal: that should be fixed somehow on the kernel side
Jun 15 10:34:12 <morphis> prahal: maybe you can to neil about this
Jun 15 10:36:44 <prahal> I  take a few minutes to review the logs ... the thing is the log does not tell which items has no type and no technology ... (could be twl4030_ac ...
Jun 15 10:37:35 <prahal> as twl4030_usb has type USB , twl4030_ac has type Mains , the bq27000-battery has type Battery and technology Li-ion
Jun 15 10:48:57 <prahal> POWER_SUPPLY_TYPE is not in any of the power_supply/*/uevents
Jun 15 10:49:38 <prahal> thus the issue . I am currently looking after if this is a kernel change on purpose or else
Jun 15 11:24:20 <prahal> hum it was not done on purpose but this uevent property is not available anymore and it looks like it dates back to 2010 commit 5f487cd34f4337f9bc27ca19da72a39d1b0a0ab4 (with an attempt to restore this prop in commit bbabb158f0e9d41174ae5c2183a8e4f981daf6ce
Jun 15 11:25:10 <prahal> still I miss it on every of my boxes , all arches (x86 , x86_64, armel
Jun 15 11:28:51 <prahal> or maybe the second commit was only there to restore the plain "type" sysfs attribute read ability
Jun 15 11:35:13 <prahal> the comment in the kernel headers tells to use the "type" attribute instead ... and we do but only half the time in fsodeviced src/plugins/kernel26_powersupply/plugin.vala
Jun 15 11:36:08 <prahal> so it might be good to fix this on both sides 
Jun 15 12:40:39 <prahal> well I fixed the above issue and stumble upon the next one in fsodeviced kernel26_powersupply
Jun 15 12:40:48 <prahal> " // if we have charger inserted and present, AC is our status" is first
Jun 15 12:42:40 <prahal> ie if we have a charge , the status is always "ac" (else it means the charger list is wrong
Jun 15 12:43:02 <prahal> thus never charging, full when charging
Jun 15 17:25:41 <GNUtoo-hplaptop> pespin, hi
Jun 15 19:14:55 <GNUtoo-desktop> Alex[sp3dev], hi
Jun 15 19:25:53 <nschle85> GNUtoo-desktop: hi, thank you for the mail regarding n900 forwarder.
Jun 15 19:26:25 <GNUtoo-desktop> nschle85, hi
Jun 15 19:27:02 <nschle85> GNUtoo-desktop: yesterday i tried to  fix system time setting on N900
Jun 15 19:27:09 <GNUtoo-desktop> ok
Jun 15 19:27:56 <nschle85> i have a "working solution" but it may break other phones so i need some help regarding bitbake files
Jun 15 19:28:04 <GNUtoo-desktop> ok
Jun 15 19:28:10 <GNUtoo-desktop> what's the solution?
Jun 15 19:29:00 <nschle85> i created a busybox-hwclock-systemd package  which must be installed manually after installing an image
Jun 15 19:29:44 <GNUtoo-desktop> ok
Jun 15 19:29:47 <nschle85> it simply contains a busybox-hwclock-systemd.service file
Jun 15 19:29:50 <GNUtoo-desktop> do you have a patch to look at?
Jun 15 19:29:55 <GNUtoo-desktop> ah ok
Jun 15 19:30:16 <nschle85> i extended busuybox???.bbappend
Jun 15 19:30:28 <GNUtoo-desktop> no
Jun 15 19:30:37 <GNUtoo-desktop> I think there is a machine unit for systemd
Jun 15 19:30:39 <nschle85> i have 2  "problems"
Jun 15 19:32:22 <nschle85> 1. i have hardcoded a hwclock call in the  service file. a better solution would be to call old hwclock.sh its some wore intelligent
Jun 15 19:32:31 <Alex[sp3dev]> GNUtoo-desktop: hi here
Jun 15 19:32:56 <GNUtoo-desktop> Alex[sp3dev], hi
Jun 15 19:33:13 <GNUtoo-desktop> Alex[sp3dev], I've some questions about galaxy nexus
Jun 15 19:33:25 <GNUtoo-desktop> at #replicant we're trying to make the framebuffer work
Jun 15 19:33:30 <GNUtoo-desktop> and we have a hard time
Jun 15 19:33:46 <GNUtoo-desktop> specially because neither paulk nor me have a galaxy nexus
Jun 15 19:33:48 <nschle85> 2. i would like to have automatically installed the packe on systemd enabled systems when busybox is installed
Jun 15 19:33:57 <Alex[sp3dev]> GNUtoo-desktop: yes, what's the problem
Jun 15 19:34:23 <Alex[sp3dev]> GNUtoo-desktop: btw, I have a question about omap4. do you have open composer, gralloc and gps replacements
Jun 15 19:34:34 <GNUtoo-desktop> Alex[sp3dev], gps probably, it seem SIRF
Jun 15 19:34:43 <GNUtoo-desktop> gralloc, we're working on it
Jun 15 19:34:59 <GNUtoo-desktop> basically we have that:
Jun 15 19:35:11 <GNUtoo-desktop> http://dl.dropbox.com/u/2757918/temp/27914997-54bc-4232-a77d-14f4b3e9c916/file.avi
Jun 15 19:35:43 <Alex[sp3dev]> GNUtoo-desktop: wrong resolution?
Jun 15 19:36:03 <GNUtoo-desktop> resolution seem correct
Jun 15 19:36:07 <Alex[sp3dev]> should be 720@1280 and 3bpp stride
Jun 15 19:36:13 <Alex[sp3dev]> yes, not 4
Jun 15 19:36:23 <GNUtoo-desktop> what's stride already?
Jun 15 19:36:43 <Alex[sp3dev]> so, 3bpp and 720 * 3 is the stride. stride is one line
Jun 15 19:36:56 <GNUtoo-desktop> ok
Jun 15 19:37:17 <GNUtoo-desktop> let me pastebin
Jun 15 19:37:23 <GNUtoo-desktop> we're in 16bit 565
Jun 15 19:37:40 <Alex[sp3dev]> GNUtoo-desktop: do you have the code to run?
Jun 15 19:37:57 <GNUtoo-desktop> yes one second
Jun 15 19:38:44 <GNUtoo-desktop> http://pastie.org/private/lqzch5kxbbbtfnuaboypa
Jun 15 19:38:48 <GNUtoo-desktop> and the code is here:
Jun 15 19:39:00 <Alex[sp3dev]> GNUtoo-desktop: it's rgb888. you can use rgb565, but you need to align each color to its own byte
Jun 15 19:39:09 <GNUtoo-desktop> ok
Jun 15 19:39:15 <GNUtoo-desktop> what version of 888
Jun 15 19:39:19 <GNUtoo-desktop> what's the #define ?
Jun 15 19:39:53 <Alex[sp3dev]> https://gitorious.org/i9100-proper-linux/i9100-uboot/blobs/pandacrap/board/ti/panda/panda.c#line320 that's weird. 4bpp, but 24bit rgb, otherwise fonts are corrupted..
Jun 15 19:40:00 <Alex[sp3dev]> GNUtoo-desktop: what do you need? a logcat?
Jun 15 19:40:10 <GNUtoo-desktop> wait a second for the code
Jun 15 19:40:57 <GNUtoo-desktop> https://gitorious.org/replicant/hardware_ti_omap3_libgralloc/commits/replicant-4.0-maguro
Jun 15 19:41:41 <GNUtoo-desktop> ok I'll try
Jun 15 19:41:55 <GNUtoo-desktop> I just hope that it works fast
Jun 15 19:42:01 <GNUtoo-desktop> fast in time
Jun 15 19:42:10 <Alex[sp3dev]> GNUtoo-desktop: basically what we're trying to do is to get rid of as much of closed code as possible. still, we'll probably use gps/powervr but we'll try to protect physical memory and prohibit blobs from controlling radio
Jun 15 19:42:11 <GNUtoo-desktop> because the user is kind of impatient
Jun 15 19:42:24 <GNUtoo-desktop> ok
Jun 15 19:43:13 <Alex[sp3dev]> GNUtoo-desktop: what should I do with that? if i patch the code, could you recompile? I've just cloned cm9 and it will take ages to build (well, probably just 2-3 hours on my laptop, but nevertheless)
Jun 15 19:43:35 <GNUtoo-desktop> if you send me a patch I'll give it to the user to test
Jun 15 19:43:42 <GNUtoo-desktop> he's able to recompile stuff
Jun 15 19:43:47 <GNUtoo-desktop> I can recompile too
Jun 15 19:43:56 <nschle85> Alex[sp3dev]: GNUtoo-desktop: which phone is it you are working on or is it for panda board ?
Jun 15 19:44:15 <GNUtoo-desktop> nschle85, I don't have the phone, it's for replicant and it's galaxy nexus
Jun 15 19:44:36 <Alex[sp3dev]> nschle85: it is tuna (galaxy nexus), i've just not started cleaning up uboot code. i was hacking inside panda code because i used panda to debug when tuna hanged
Jun 15 19:44:36 <GNUtoo-desktop> nschle85, apart that I'm also working on n900
Jun 15 19:45:30 <GNUtoo-desktop> basically there is that in framebuffer.cpp:
Jun 15 19:45:44 <Alex[sp3dev]> I think i'll need to work on galaxy s2 ril some week or two later. both because i want it for my personal fun and because nexus has the same kind of ril
Jun 15 19:45:53 <GNUtoo-desktop>      * Explicitly request 5/6/5
Jun 15 19:45:57 <nschle85> GNUtoo-desktop: this i recognized and this makes me happy :-)
Jun 15 19:46:25 <GNUtoo-desktop> nschle85, are you able to write stuff in vala?
Jun 15 19:46:38 <GNUtoo-desktop> since you come from the java world it shound't be too much problematic
Jun 15 19:47:11 <nschle85> GNUtoo-desktop: http://pastebin.com/qpnpqfYu this is the patch for the recipe itself the   service file is not included here but its no magic
Jun 15 19:47:45 <nschle85> GNUtoo-desktop: not yet but you are right  i have to learn (google) the different API
Jun 15 19:48:07 <GNUtoo-desktop> Alex[sp3dev], is that ok: http://pastie.org/private/q7zg9thxzv0aieyi1p8i9w
Jun 15 19:48:16 <Alex[sp3dev]> GNUtoo-desktop: this is the most obvious fix http://pastebin.com/s86pkCZv but I think we can use rgb888
Jun 15 19:48:44 <Alex[sp3dev]> GNUtoo-desktop: nope. you should change offset as well
Jun 15 19:48:50 <GNUtoo-desktop> I did
Jun 15 19:48:57 <Alex[sp3dev]>     info.red.offset     = 11; 
Jun 15 19:49:01 <Alex[sp3dev]> wrong
Jun 15 19:49:04 <GNUtoo-desktop> ah right
Jun 15 19:49:19 <Alex[sp3dev]> and i think green offset is 8 and blue is 0. why not?
Jun 15 19:49:31 <GNUtoo-desktop> http://pastie.org/private/namle42j4j5zb3xqnf6xpw ?
Jun 15 19:50:10 <Alex[sp3dev]> http://pastie.org/4094207
Jun 15 19:50:13 <Alex[sp3dev]> try this
Jun 15 19:50:22 <Alex[sp3dev]> i can try the binaries on the nexus
Jun 15 19:50:43 <GNUtoo-desktop> ok
Jun 15 19:51:05 <GNUtoo-desktop> I've pasted in #replicant for the user
Jun 15 19:51:19 <Alex[sp3dev]> GNUtoo-desktop: ping me when there are binaries
Jun 15 19:51:29 <GNUtoo-desktop> Alex[sp3dev], ok
Jun 15 19:51:50 <Alex[sp3dev]> what's the gps on omap? sirf star iv? is it supported by gpsd?
Jun 15 19:53:02 <GNUtoo-desktop> seem so
Jun 15 19:53:11 <GNUtoo-desktop> the proprietary libs have sirf in their name
Jun 15 19:56:44 <Alex[sp3dev]> uboot actually is such a pile of poo. it is unreadable (ifdefs everywhere), it has neither fastboot nor dfu, patch acceptance criteria are very strict. and hush scripting is a weird idea, I'd be better off with plain C. but since all the support for tegra/omap/samsung is there, I'm not complaining
Jun 15 19:58:03 <GNUtoo-desktop> else there is barebox...
Jun 15 19:58:35 <GNUtoo-desktop> <mirage335> Now I get greyscale wavylines on the top two-thirds of the screen.
Jun 15 19:59:22 <Alex[sp3dev]> GNUtoo-desktop: ok, how does android build system work? is there a way to compile a single module (like mmm module-path) but with dependencies?
Jun 15 20:00:59 <GNUtoo-desktop> you need to compile everything and then use mm
Jun 15 20:02:17 <GNUtoo-desktop> altough on a core i7 it's 20min
Jun 15 20:02:35 <Alex[sp3dev]> well, core i7 is at work.. i only have a dual-core i5 at home
Jun 15 20:03:14 <GNUtoo-desktop> ok
Jun 15 20:03:52 <GNUtoo-desktop> I'll build on dual core AMD XII
Jun 15 20:05:00 <nschle85> GNUtoo-desktop: model name      : AMD Athlon(tm) II X4 640 Processor 
Jun 15 20:05:18 <GNUtoo-desktop> ok
Jun 15 20:05:44 <nschle85> GNUtoo-desktop: have you read the patch ?
Jun 15 20:05:45 <GNUtoo-desktop> the problem is the RAM here....2G only
Jun 15 20:05:52 <GNUtoo-desktop> not yet
Jun 15 20:06:17 <nschle85> GNUtoo-desktop: should i sent it to shr ML and  tell about my problems ?
Jun 15 20:07:19 <GNUtoo-desktop> nschle85, yes
Jun 15 20:07:31 <GNUtoo-desktop> I think JaMa will take care of it
Jun 15 20:07:59 <nschle85> GNUtoo-desktop: is he available next time ?
Jun 15 20:08:58 <GNUtoo-desktop> nschle85, what do you mean by "next time"
Jun 15 20:09:26 <nschle85> GNUtoo-desktop: during next days
Jun 15 20:09:35 <GNUtoo-desktop> write a mail then
Jun 15 20:09:43 <GNUtoo-desktop> because I've no idea
Jun 15 20:09:45 <Alex[sp3dev]> GNUtoo-desktop: any idea how to set up ramdisk aufs/unionfs so that android compiles in ram and does not write to the ssd but it is synced to ssd when ram is getting low? well, given that i rarely compile anything, I can just compile as is, won't probably kill the disk within two-three years at least
Jun 15 20:09:49 <nschle85> ah ok
Jun 15 20:10:31 <GNUtoo-desktop> Alex[sp3dev], the compilation is setup here
Jun 15 20:11:15 <Alex[sp3dev]> i think i'll compile cm9 on my laptop just in case. at least i'll have to have it compiled for building libs and ril
Jun 15 20:14:08 <GNUtoo-desktop> ok I've the binary for you:
Jun 15 20:14:41 <GNUtoo-desktop> http://gnutoo.homelinux.org/downloads/people/Alex%5bsp3dev%5d/gralloc.omap4.so
Jun 15 20:15:16 <nschle85> does anybody know :  is the RTC (clock) running in UTC or local time on android phones ?
Jun 15 20:15:48 <GNUtoo-desktop> nschle85, I don't know sorry, how to check?
Jun 15 20:19:08 <Alex[sp3dev]> nschle85: likely in UTC. only insane software like windows uses localtime
Jun 15 20:19:23 <Alex[sp3dev]> although android is insane. need to figure out
Jun 15 20:20:26 <Alex[sp3dev]> yeah, likely utc, because java API is said to return UTC via the RTC method
Jun 15 20:22:05 <nschle85> ok so RTC is running  in UTC on N900 Maemo, and Android,
Jun 15 20:23:21 <JaMa|Off> nschle85: pong
Jun 15 20:23:45 <nschle85> JaMa|Off: got notified :-)
Jun 15 20:23:57 <GNUtoo-desktop> Alex[sp3dev], do you have another change for me to compile?
Jun 15 20:26:18 <nschle85> JaMa|Off: i tried to enable  time sync between  rtc and system time on boot so i created a service file for systemd and extended busybox??.bbappend but the result is not what i want.
Jun 15 20:26:58 <Alex[sp3dev]> GNUtoo-desktop: it just segfaults
Jun 15 20:27:02 <GNUtoo-desktop> ah ok
Jun 15 20:27:05 <GNUtoo-desktop> hmmm
Jun 15 20:27:14 <nschle85> JaMa|Off: the sync itself works but handling bitbake package creation and integration is not what i expect
Jun 15 20:27:16 <GNUtoo-desktop> what does it prints in logs?
Jun 15 20:27:33 <Alex[sp3dev]> GNUtoo-desktop: let's get back to it later, it's 0.30 AM, so I need to finish some school stuff and go to bed
Jun 15 20:27:44 <GNUtoo-desktop> ok ouch
Jun 15 20:27:46 <Alex[sp3dev]> GNUtoo-desktop: it prints a typical stack trace of a native crash
Jun 15 20:27:51 <GNUtoo-desktop> ok
Jun 15 20:27:59 <GNUtoo-desktop> so it doesn't setup the fb first....
Jun 15 20:28:00 <GNUtoo-desktop> ok
Jun 15 20:29:05 <Alex[sp3dev]> GNUtoo-desktop: http://pastebin.com/Gu69LwH3
Jun 15 20:29:13 <GNUtoo-desktop> thanks
Jun 15 20:29:42 <GNUtoo-desktop> ok it's logic why
Jun 15 20:29:59 <GNUtoo-desktop> it's because you didn't use a full cyanogenmod build without proprietary stuff
Jun 15 20:30:04 <GNUtoo-desktop> at least it'
Jun 15 20:30:04 <Alex[sp3dev]> yes
Jun 15 20:30:06 <GNUtoo-desktop> oops
Jun 15 20:30:15 <Alex[sp3dev]> powervr doesn't find its compatible surface
Jun 15 20:30:21 <GNUtoo-desktop> indeed
Jun 15 20:30:24 <GNUtoo-desktop> see you later then
Jun 15 20:30:31 <GNUtoo-desktop> maybe launch a build
Jun 15 20:30:44 <GNUtoo-desktop> and finish the school work when it's building
Jun 15 20:30:52 <Alex[sp3dev]> yes
Jun 15 20:31:12 <Alex[sp3dev]> it's still downloading some tuna-specific device folders
Jun 15 20:31:37 <GNUtoo-desktop> ok
Jun 15 20:40:04 <JaMa|Off> nschle85: what you mean by "handling bitbake package creation and integration is not what i expect" ?
Jun 15 20:42:27 <nschle85> JaMa|Off: ok what i have  done results in: creating a n empty package "busybox-hwclock" and a package "busybox-hwclock-systemd" which contains the service file for busybox-hwclock.service
Jun 15 20:43:43 <nschle85> JaMa|Off: if you install shr you need manually install busybox-hwclock-systemd package to get system time setting to work (thats not user friendly)
Jun 15 20:45:13 <nschle85> JaMa|Off: what i wanted to do is to create a busybox-systemd package but  as i know its not possible to have more than one systemd file per packe so this way i wont go
Jun 15 20:46:13 <Alex[sp3dev]> GNUtoo-desktop: is cm9 brunch parallel by default or do i need to specify the number of -j threads for make somehow?
Jun 15 20:46:16 <nschle85> JaMa|Off: did you understand my last sentece ? (kvirk in ubuntu does not work as i expect )
Jun 15 20:47:36 <nschle85> JaMa|Off: did you read my patch ?  http://pastebin.com/qpnpqfYu
Jun 15 20:49:34 <GNUtoo-desktop> Alex[sp3dev], no idea, look here:
Jun 15 20:49:58 <JaMa|Off> nschle85: multiple -systemd packages should work and feel free to add it to task-shr-systemd recipe to rdepend it in image
Jun 15 20:49:59 <GNUtoo-desktop> http://wiki.cyanogenmod.com/wiki/Galaxy_Nexus_%28GSM%29:_Compile_CyanogenMod_%28Linux%29
Jun 15 20:50:24 <GNUtoo-desktop> ah there is no such thing here
Jun 15 20:51:09 <GNUtoo-desktop> Alex[sp3dev], basically follow the procedure without the extract files
Jun 15 20:51:15 <nschle85> JaMa|Off: ok ill do
Jun 15 20:51:36 <GNUtoo-desktop> Alex[sp3dev], but then it would fails
Jun 15 20:51:39 <JaMa|Off> nschle85: ah right, better to keep busybox-hwclock-systemd in separate package, do not add that .service file to ${PN}-syslog-systemd
Jun 15 20:51:43 <GNUtoo-desktop> Alex[sp3dev], maybe checkout replicant....
Jun 15 20:52:49 <nschle85> JaMa|Off: i am not doing it so i hope or is there an error in  the patch ?
Jun 15 20:53:07 <GNUtoo-desktop> Alex[sp3dev], the branch we use for ICS is ics-preparation
Jun 15 20:54:28 <GNUtoo-desktop> git://gitorious.org/replicant/android_vendor_cm.git
Jun 15 20:59:51 <nschle85> JaMa|Off: is there a way to avoiud creating the empty  busybox-hwclock package ?
Jun 15 21:03:46 <JaMa|Off> nschle85: why did you add "PACKAGES =+ "${PN}-hwclock"" in first place?
Jun 15 21:06:32 <nschle85> JaMa|Off: without it and without  ALLOW_EMPTY_${PN}-hwclock = "1"  the busybox-hwclock-systemd cannot be installed because there is some dependency generated i cannot influence
Jun 15 21:07:12 <JaMa|Off> it should add only dependency on ${PN} not ${PN}-hwclock
Jun 15 21:07:59 <JaMa|Off> ah true.. pkg_systemd_base = pkg_systemd.replace('-systemd', '') is added to RDEPENDS_pkg_systemd
Jun 15 21:08:02 <nschle85> JaMa|Off: my answer was somehow incorrect: i added  PACKAGES =+ "${PN}-hwclock"  because bitbake was missing that package during build
Jun 15 21:09:28 <JaMa|Off> you can add RDEPENDS_${PN}-hwclock-systemd = "${PN}" and drop ${PN}-hwclock
Jun 15 21:09:38 <nschle85> JaMa|Off: and then i added ALLOW_EMPTY_${PN}-hwclock = "1" for installation
Jun 15 21:28:46 <nschle85> JaMa|Off: how can i update busybox without breaking the installation ?  so opkg install --force-reinstall busybox does not work because important commands are removed  during installation
Jun 15 21:42:13 <JaMa|Off> nschle85: install coreutils and wget before doing that
Jun 15 21:44:39 <nschle85> JaMa|Off: ok i think wget was the command which failed..   is there any other implementation of hwclock apart of busybox i have to care about ?
Jun 15 21:45:44 <JaMa|Off> util-linux also provides hwclock
Jun 15 21:46:51 <nschle85> JaMa|Off: does it also provide hwclock.sh  script ? 
Jun 15 21:47:05 <JaMa|Off> so it's not bad idea to add hwclock-systemd recipe shared by both, using just ${base_sbindir}/hwclock without care if it's provided by busybox or util-linux
Jun 15 21:47:32 <JaMa|Off> no, just hwclock binary afaik
Jun 15 22:01:04 <nschle85> JaMa|Off: hmm i recived the message but i have to think about  how to implement .. so i should create a new bb file for that ? 
Jun 15 22:11:19 <nschle85> JaMa|Off: ill change my code as you told me and send it to shr-ml  so we can talk about later 
Jun 15 22:14:51 <nschle85> JaMa|Off: hmm bitbake -c cleansstate  busybox does now bring an error , do you have time to support me ?
Jun 15 22:21:55 <nschle85> JaMa|Off: http://pastebin.com/bZyYqvbt this is the new patch, the error follows
Jun 15 22:25:52 <nschle85> JaMa|Off: http://pastebin.com/xSz2qVCx this is the current error
**** ENDING LOGGING AT Sat Jun 16 02:59:58 2012