IRC log of #maemo-ssu for Thursday, 2013-10-17

*** NIN101 has quit IRC00:00
*** Pali has quit IRC00:04
*** _d0t has joined #maemo-ssu00:20
freemangordonyop00:20
freemangordonyup even :)00:21
_d0ti also compiled the watchdog timer into the kernel00:21
freemangordonok, you should use Pali's uboot00:21
_d0tlast time I tried it had problems with booting a kernel from ubi00:21
freemangordonI guess it was a while ago :)00:22
freemangordonnow there are no problems00:22
_d0tcould you give me the uboot boot script then?00:22
freemangordonnothing in there, everything is in the cmdline, check rx51_defconfg00:23
freemangordon_d0t: anyway, GF will kill me if I hang here for one more minute :D00:23
*** nox- has joined #maemo-ssu00:24
_d0tfreemangordon: ok, cya00:24
freemangordonbut that kernel boots for sure00:24
freemangordonnight00:24
*** dafox has quit IRC00:43
*** lrtz_ has quit IRC00:44
*** lrtz has joined #maemo-ssu00:46
*** dos1 has quit IRC01:16
*** MrPingu has quit IRC01:32
*** dos1 has joined #maemo-ssu01:53
DocScrutinizer05_d0t: you might want to attach serial (aka RS232) to console testpoints under battery, to actually debug kernel startup02:01
_d0tDocScrutinizer05: how do I do that?02:15
_d0tah.. testpoints, ok02:15
_d0tstill I dont think the kernel is even started. JTAG would be nice.02:15
*** freemangordon_ has joined #maemo-ssu02:19
*** freemangordon has quit IRC02:19
DocScrutinizer05why JTAG?02:25
_d0tbecause it allows direct access to memory02:38
_d0tbtw, that script won't help since I dont have bme, and i2cget/set cannot access those lines02:39
_d0tbut I got the idea that I have to discharge the battery and charge it til full02:40
*** M4rtinK has quit IRC02:43
*** Martix has quit IRC03:21
*** _d0t has quit IRC04:02
*** kolp has quit IRC04:13
*** kolp_ has joined #maemo-ssu04:13
*** kolp_ is now known as kolp04:17
*** LauRoman has joined #maemo-ssu04:29
*** LauRoman has quit IRC05:12
*** amiconn_ has joined #maemo-ssu05:33
*** amiconn has quit IRC05:33
*** amiconn_ is now known as amiconn05:33
*** okias has quit IRC05:39
*** nox- has quit IRC05:46
*** jonwil has joined #maemo-ssu06:36
*** dafox has joined #maemo-ssu08:00
*** trx has quit IRC08:16
*** trx has joined #maemo-ssu08:20
*** dafox has quit IRC08:36
*** gggs has left #maemo-ssu08:39
*** Pali has joined #maemo-ssu08:44
*** Pali has quit IRC09:15
*** MadeAllUp has quit IRC10:07
*** MadeAllUp has joined #maemo-ssu10:12
*** sunny_s has joined #maemo-ssu10:24
*** luf has joined #maemo-ssu10:36
*** Martix has joined #maemo-ssu10:57
*** LaoLang_cool has joined #maemo-ssu12:00
*** DrCode has quit IRC12:05
*** DrCode has joined #maemo-ssu12:06
*** jonwil has quit IRC12:06
*** dos1 has quit IRC12:07
*** LaoLang_cool has quit IRC12:08
*** sunny_s has quit IRC12:18
*** Martix has quit IRC12:38
*** Martix has joined #maemo-ssu12:38
*** Martix has quit IRC13:00
*** Martix has joined #maemo-ssu13:02
*** Martix has joined #maemo-ssu13:04
*** lizardo has joined #maemo-ssu13:28
*** jonwil has joined #maemo-ssu14:28
*** Martix has quit IRC14:50
*** dos1 has joined #maemo-ssu14:51
*** Pali has joined #maemo-ssu15:03
*** LauRoman has joined #maemo-ssu15:14
*** trx has quit IRC15:21
*** thedead1440 has joined #maemo-ssu15:22
*** trx has joined #maemo-ssu15:26
*** LauRoman has quit IRC15:37
*** DrCode has quit IRC15:45
*** DrCode has joined #maemo-ssu15:47
*** LauRoman has joined #maemo-ssu16:00
*** l4m3rx has quit IRC16:09
*** l4m3rx has joined #maemo-ssu16:10
*** RST38h has quit IRC16:30
*** okias has joined #maemo-ssu16:32
*** RST38h has joined #maemo-ssu16:33
*** Martix has joined #maemo-ssu17:16
*** _rd_ has joined #maemo-ssu17:19
*** _rd_ has quit IRC17:28
*** _rd_ has joined #maemo-ssu17:29
*** _rd_ has quit IRC17:34
*** Martix has quit IRC17:46
*** Martix has joined #maemo-ssu18:15
*** dos11 has joined #maemo-ssu18:18
*** dos1 has quit IRC18:20
*** NIN101 has joined #maemo-ssu18:31
*** Martix has quit IRC18:36
*** MadeAllUp has quit IRC18:49
*** sunny_s has joined #maemo-ssu18:50
*** okias has quit IRC18:53
Paliagain: https://gitorious.org/community-ssu/calendar-backend/merge_requests/118:53
Paliany comments about this merge request?18:53
* freemangordon_ is going to check18:53
jonwilok, so who plans to work on cssu MCE?18:56
Palimerlin1991: again this is for you! https://gitorious.org/community-ssu/qt-x11-maemo/merge_requests/5#comment_9987318:56
Palijonwil: remove /proc/bootreason component_version usage18:56
Palijonwil: I will do it18:56
Pali(something I already done in diablo mce...)18:56
PaliI will port this change18:57
jonwilok :)18:57
freemangordon_Pali: please merge rc5 first18:57
Paliwhat is rc5?18:57
freemangordon_3.1218:57
Paliok18:57
Palilinux18:57
*** freemangordon_ is now known as freemangordon18:57
Paliand there is: https://gitorious.org/community-ssu/status-area-orientationlock-applet/merge_requests/618:58
freemangordonPali: hmm, if you remove bootreason, what will happen with ACT_DEAD mode?18:58
Palifreemangordon: nothing, kernel has already hardcoded pwr btn18:58
*** M4rtinK has joined #maemo-ssu18:58
Palibootreason and component version does not working since 3.5 (or 3.6?)18:58
freemangordonPali: exactly18:58
freemangordonwe'll break user experiens18:59
freemangordon*experience18:59
Palithis is already with any 3.x kernel18:59
freemangordonwe should fix the kernel then18:59
Palinot easy...18:59
Paliatag omap table was totally removed from kernel18:59
freemangordonPali: how it is supposed to pass data from bootloader to the kernel now atags are removed?19:00
Palirather ask on LKML how to propagate bootreason to userspace19:00
Palifreemangordon: atag omap table was removed19:00
Palinot all atags19:00
freemangordonis that make any difference?19:00
freemangordons/is/does/19:01
infobotfreemangordon meant: does that make any difference?19:01
Palineed to ask on LKML for proper way how to tell kernel about bootreason19:01
freemangordonok19:01
Palior maybe I can look at some code, if we can read bootreason from kernel (and ignore data from NOLO)19:01
freemangordonPali: I don;t think we should remove any functionality from mce19:02
*** Martix has joined #maemo-ssu19:02
Palinot remove, but make sure mce not crash/fail19:02
Palisame as in other parts which are already patched19:02
jonwilObviously MCE shouldn't be changed in ways that make it incompatible with the stock kernel19:02
freemangordonI guess only GPIO etc names should be fixed19:02
Palijonwil: nobody is going to do that19:02
jonwilok19:02
freemangordonit is not about compatibility, but about functionality19:03
Palilook at other bootreason changes in cssu19:03
Paliwith 3.x kernel there is no way to tell bootreason (yet)19:03
freemangordonPali: those are fine while we're fixing the kernel as they allow us to boot. but for a "production" I don;t think those hacks should remain19:04
Paliand kernel provide fake bootreason file for apps which crashing if that file does not exist19:04
freemangordonI know19:04
PaliI rather use some better way as nokia special proc file...19:04
freemangordonPali: can't we read atags from userspace?19:04
Paliso need to ask on lkml for proper way to tell booreson from kernel to userspace19:04
Palionly as root19:05
Paliand need to parsing them19:05
freemangordonwhat is the problem? read them from linuxrc, store data in /tmp or in /run or wherever19:05
Palibut this is I think irrelevant...19:05
Palithere are other bugs with higher priority19:05
freemangordonsure19:06
Palinot working act dead bootreason is maybe last bug...19:06
freemangordonPali: seems I am not going to get much help for omapfb :(19:06
freemangordonmaybe I should address the issue with MM guys19:06
freemangordontry to19:07
*** Martix has quit IRC19:07
jonwilbtw, the things that are using bootreason are 10hildon_weclome script, fb-progress.sh script & getbootstate binary19:11
freemangordonI guess getbootstate can be fixed to parse atags and write some tmp files19:12
freemangordonand use those files if they already exist19:13
freemangordonor we can just symlink /dev/bootreason and /dev/comp_version to those files in /tmp19:16
freemangordonsre is doing great job https://sre.ring0.de/linux/log/?h=n900-dt :)19:18
jonwilfor component_version, its getbootstate, hald, ham-notifier-status-menu-item.so, ham-updates-status-menu-item.so, http, https, libaccelerometer.so, libalarm.so, libaudiorouting.so, libbattery.so, libcallstate.so, libcamera.so, libdisplay.so, libfilter-brightness-als.so, libfilter-brightness-simple.so, libhomekey.so, libinactivity.so, libkeypad.so, libled.so, libossoproductinfo.so, libsysinfo.so, l19:18
jonwilibvitrator.so, libxul.so, mce, preinit script, rcS script, sgx script, softupd.rcS script & testserver19:18
jonwilof those only libossoproductinfo and libsysinfo are closed19:19
jonwilwell and testserver but we can ignore that19:19
jonwilI think someone should just go ahead and clone libossoproductinfo and libsysinfo (both of which have x86 binaries)19:19
jonwilI would have a go but I have other things I am working on (non-N900)19:20
freemangordonls -al /usr/lib/libossoproductinfo.so.0.7.319:20
freemangordon-rw-r--r--    1 root     root        11168 Aug 28  2009 /usr/lib/libossoproductinfo.so.0.7.319:20
freemangordonNokia-N900:~# ls -al /usr/lib/libsysinfo.so.019:20
freemangordon-rw-r--r--    1 root     root        16516 Mar  3  2010 /usr/lib/libsysinfo.so.019:20
jonwilplus I just finished the largest piece of reverse engineering work in the history of Maemo so I think I deserve a break :)19:20
freemangordonI bet those have x86 binaries19:20
jonwilyes they do19:20
jonwilthey definatly do19:20
freemangordonI may have a look19:20
jonwiland I posted header files for both of them the other day19:21
Palijonwil: when you are writing list what is using them, please remove that which are already patches in cssu, so you will not confuse me (and others)19:23
Palifreemangordon: by default uboot not adding omap atag table to 3.x kernels19:23
*** Martix has joined #maemo-ssu19:23
freemangordonPali: aren;t those available in the memory?19:23
jonwilI dont know which packages those bins come from or which ones are in cssu19:24
Palifreemangordon: where in memory?19:24
freemangordon"/dev/memory" :)19:24
freemangordonno idea19:24
Palino idea too :-)19:25
Paliyou need to tell uboot to generate non standard omap atag table19:25
freemangordonPali: where u-boot gets them from?19:25
Paliall omap atag table is hardcoded in uboot19:26
freemangordonhmm, iirc there was some data on 0x402?????19:26
Palibut it trying to read bootreason from nolo...19:26
freemangordonwill check in the trm19:26
*** Martix has quit IRC19:33
*** dos11 is now known as dos119:35
*** jonwil has quit IRC19:36
*** M4rtinK has quit IRC19:39
Palifreemangordon: in cssu git osso-systemui-dev has version 0.4, but in cssu-devel and testing repo only 0.319:39
Palimaybe you should update -devel repo?19:40
Palipackages which are in git, but missing in -devel: bluez calendar-backend getbootstate hildon-im-vkbrenderer3 hildon-input-method-plugins libconnui-dev libgtkadi mce osso-systemui-devlock-dev osso-systemui-modechange-dev osso-systemui-splashscreen-dev19:42
Paliluf: what is state of bluez?19:42
Palifor calendar-backend there is pending patch19:43
freemangordonPali: I didn;t pu -dev version in -devel repo19:43
Palifreemangordon: that -dev package has older version in testing and devel19:44
Paliso remove that package from -testing or update it19:44
freemangordonPali: I was expecting merlin1991 to do the update ASAP and to put it in -testing19:44
Paliok19:44
freemangordonPali: will upload 0.4 in -devel repo19:44
Palipackage getbootstate: it is not finished yet (missing testing /sbin/preinit)19:44
Palihildon-im-vkbrenderer3 hildon-input-method-plugins: this is _nicolai_ work?19:45
freemangordonpartially :)19:45
Paliso what is state of those packages?19:45
Paliincomplete? or not working? or needs testing?19:45
freemangordonI guess needs testing, but for sure it is nicolai to answer that question, as he was working on it19:46
freemangordonimplementing portrait support19:46
Pali~seen _nicolai_19:46
freemangordonI don;t know at what state he left it19:46
infobot_nicolai_ <~nicolai@pop8-126.catv.wtnet.de> was last seen on IRC in channel #maemo-ssu, 27d 21h 46m 4s ago, saying: 'hi'.19:46
Palisomebody please ping nicolai on TMO to join #maemo-ssu19:47
freemangordonping on TMO?19:47
freemangordonPM?19:47
Palihe is more online in TMO...19:47
Palithen on IRC19:47
freemangordonosso-system-ui-dev 0.4 in -devel19:49
Paliok19:50
Palilook here: http://repository.maemo.org/community-testing/dists/fremantle/free/source/Sources.gz there is missing battery plugin19:50
Palimerlin1991: why source code of battery plugin is missing in CSSU-T?19:50
Palimerlin1991: please delete package location-supl-fix from cssu devel repository19:51
freemangordonmy daughter is hungry, going to cook something for dinner, bbl19:55
*** DrCode has quit IRC19:59
*** dos1 has quit IRC20:07
*** dos1 has joined #maemo-ssu20:07
*** xes has joined #maemo-ssu20:15
Palion cssu gitorious is some empty repo telepathy-gabble, why?20:17
*** xes has quit IRC20:21
*** xes has joined #maemo-ssu20:27
freemangordonPali: there was a problem with FB chat accounts, but FB fixed it berfore us :)20:33
Palifreemangordon: I remember that there was problem with google talk video calls...20:34
Palibut do not know if it was fixed or not...20:34
Paliyear or two ago20:34
freemangordongoogle fixed it iirc20:34
Palifreemangordon: is commit "OMAP: RX-51: rx51_defconfig - increase VRAM size to 9 MB" needed?20:45
freemangordonPali: no20:45
Paliok20:46
Paliand others yes?20:46
freemangordonPali: I have some uncommited changes (enable CMA, omapfb resolution and bpp), but those can wait until omapfb is functional20:47
freemangordonPali: lemme check, but iirc all others are needed20:47
Paliok20:47
freemangordonPali: all others are needed20:48
Paliok20:48
*** arcean has joined #maemo-ssu20:51
Palifreemangordon: 143 patches are in 3.12-rc1-n90020:54
freemangordonnot good :(20:54
freemangordonPali: are there any big ones, besides PVR and SSI20:55
freemangordon?20:55
Palibig wl1251 series20:56
freemangordonwut? those are not upstreamed?20:56
Paliand more rx51_defconfig commits20:56
Palifreemangordon: yes20:56
Palinot in upstream20:56
freemangordonhmm, why?20:56
freemangordonmaemo specific?20:56
Palisome (2 or 3) are maemo specific20:57
Palibut others are hw support20:57
Palipacket injection support20:57
Palimonitor support20:57
Palihw filtering...20:57
freemangordonok20:58
freemangordoncan some of the patches be merged? for sure at least 5-6 of mine can20:59
freemangordonor maybe this is not a good idea20:59
freemangordonwe'll lose history20:59
freemangordonooh, libossoproductinfo uses libcal :)21:00
freemangordonPali: does it make sense to RE it?21:01
Paliwhat?21:01
freemangordonlibossoproductinfo21:02
Palido not know what is using it and if we need it...21:02
freemangordonthe one that uses /dev/component_version21:02
freemangordonPali: see jonwil's post ^^^21:03
freemangordon<jonwil> of those only libossoproductinfo and libsysinfo are closed21:03
freemangordon19:19 <jonwil> well and testserver but we can ignore that21:03
freemangordon<jonwil> I think someone should just go ahead and clone libossoproductinfo and libsysinfo (both of which have x86 binaries)21:03
Paliif it using /proc/component_version then if you can RE it, it could be nice21:04
Paliso we can get rid of component_version21:04
freemangordonit is 11k x86 binary21:04
freemangordonsysinfo is a bit bigger, but not that much, some 17k21:04
Paliif you need to know in C code if binary running on n900 device, look into /proc/cpuinfo21:04
freemangordonyep, I can21:04
Palithere is something like rx51 board21:04
freemangordonPali: I don;t think it is that simple, there are HW revisions21:05
PaliI will fix that in MCE (add support for device detection also from (/proc/cpuinfo)21:05
Paliyou can read also hw revision from /proc/cpuinfo21:05
Paliit is serial number field21:05
Palilook into 0xFFFF source code21:05
freemangordonPali: wait, isn't it better to just provide that data in /dev/comp_whatever?21:06
freemangordonbetter/easier21:06
Paliyes, but if that proc file does not exists, fallback to cpuinfo21:06
freemangordonwe can symlink /dev/xxx to some ordinary file21:06
Paliit is not dev!21:06
Paliit it proc"21:06
freemangordonoh21:06
Paliso no symlink21:06
Palionly RE it21:07
PaliI will add cpuinfo parsing support....21:07
PaliI already created that code21:07
freemangordonPali: ok21:07
Palifreemangordon: in commit "Support both new and old DSP API" old is harmattan and new is 3.x?21:13
Palior old is fremantle and new is 3.x?21:13
freemangordonold is maemo and new is harmattan :)21:13
Paliso harmattan is compatible with 3.x?21:14
freemangordonyep21:14
Paliok21:14
Paliand 3.12 kernel has enabled which support?21:14
Paliold or new?21:14
freemangordonnew21:14
freemangordonoh, wait21:14
freemangordonboth old and new21:14
Paliand kp?21:14
freemangordonsame21:14
Paliok21:14
freemangordonyou can disable old support21:15
freemangordonwith CONFIG_DSP_API_NEW or whatever I called that21:15
Paliok21:15
*** amiconn has quit IRC21:22
*** M4rtinK has joined #maemo-ssu21:22
*** amiconn has joined #maemo-ssu21:24
*** Martix has joined #maemo-ssu21:38
*** Martix has quit IRC21:39
*** Martix has joined #maemo-ssu21:41
Palifreemangordon: n900 kernel patches rebased21:48
Palidoing compile check now21:48
freemangordonPali: great21:49
freemangordonPali: ever heard for xstat function?21:49
Palixstat? no. only about stat and lstat21:50
Palifreemangordon: do you know that CONFIG_TIDSPBRIDGE_DVFS is disabled?21:59
freemangordonI enabled it here21:59
Palino21:59
freemangordonthis is part of uncommited changes21:59
Paliok21:59
Palidue to and it is not possible to enable CONFIG_TIDSPBRIDGE_DVFS due21:59
freemangordonthere are more22:00
PaliOMAP_PM_SRF22:00
freemangordonPali: well, just enable it22:00
Palienabled22:00
Paliand compile error22:00
freemangordonhmm, weird22:00
freemangordonwhat is broken?22:00
Paliyou need to remove OMAP_PM_SRF from "depends on TIDSPBRIDGE && OMAP_PM_SRF && CPU_FREQ"22:01
Paliin drivers/staging/tidspbridge/Kconfig22:01
Palibecause OMAP_PM_SRF is invalid config name22:01
Paliso Kconfig will always disable TIDSPBRIDGE_DVFS22:01
freemangordonso OMAP_PM_SRF was removed?22:02
Palilooks like it does not exists22:02
Palierror log: http://pastebin.com/QPaYpbqM22:04
freemangordonweird22:04
Palidiff: http://pastebin.com/db7FRvMU22:04
freemangordonPali: hmm, that is weird http://www.spinics.net/lists/linux-driver-devel/msg35438.html22:06
freemangordonit's been like that since 3 years22:06
freemangordonor something22:06
freemangordonPali: seems OPP definitions have moved to somne other header file22:07
Palifreemangordon: branch v3.12-rc5-n900 pushed to gitorious22:08
Palihappy hacking :-)22:08
freemangordonPali: unbuildable?22:08
Palibranch has disabled dvfs22:09
freemangordonok22:09
Paliso buidling working22:09
freemangordonwill check it when I finish REing22:09
*** jon_y has quit IRC22:14
*** jon_y has joined #maemo-ssu22:15
*** M4rtinK has quit IRC22:18
*** okias has joined #maemo-ssu22:25
*** Martix has quit IRC22:42
*** Martix has joined #maemo-ssu22:43
*** dos1 has quit IRC22:46
*** sunny_s has quit IRC22:59
*** Pali has quit IRC23:35
*** arcean has quit IRC23:40
*** M4rtinK has joined #maemo-ssu23:41
*** _rd_ has joined #maemo-ssu23:53
*** lizardo has quit IRC23:55

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!