**** BEGIN LOGGING AT Thu Sep 25 03:00:01 2014
Sep 25 14:26:04 <Pali> ~poettering
Sep 25 14:26:05 <infobot> 'sth is poettering' means it acts invasive, possessive, destructive, and generally in an egocentric exacerbating negative way. ``this cancer is extremely poettering'', or you look here for Linus' notion on what's poettering: http://lkml.iu.edu/hypermail/linux/kernel/1404.0/01331.html, or http://lkml.iu.edu/hypermail/linux/kernel/1404.0/01488.html, or see ~systemd cabal
Sep 25 14:50:19 <CapsuL> hello
Sep 25 14:52:32 <CapsuL> anybody can be help me about maemo5 on mobile phone ?? (Nokia N900)
Sep 25 14:53:50 <jmlich> CapsuL, just ask the question. This is a right place.
Sep 25 14:53:56 <CapsuL> i looking for install QT4 library/driver , because it's impossible to install some application
Sep 25 14:54:08 <CapsuL> thx jmlich
Sep 25 14:54:43 <jmlich> have you tried maemo/extras/testing repositories?
Sep 25 14:54:48 <CapsuL> exemple of app : miniature (chess game)
Sep 25 14:57:43 <jmlich> In those ancient ages, I was trying to create apps using qt4 + qt components from meego. However, there was different (incompatible) version of Qt in extras/extras-testing/.. repos. 
Sep 25 14:58:16 <jmlich> I am not sure about current state of Qt in maemo. Currently, I am using Jolla.
Sep 25 14:59:25 <CapsuL> hmm
Sep 25 15:00:06 <jmlich> see http://wiki.maemo.org/Qt-Maemo
Sep 25 15:00:11 <CapsuL> how can i see wich OS on my phone ? Wwich command in terminal?
Sep 25 15:00:57 <jmlich> nokia n900 is fremantle, nokia n9 is harmattan
Sep 25 15:01:13 <CapsuL> yess thx u for link , why i didn't start by this lol
Sep 25 15:01:16 <jmlich> IIRC diablo is n810 
Sep 25 15:02:05 <CapsuL> i have n900 , with keyboard:)
Sep 25 15:03:10 <quentin__> jmlich , i am CapsuL (from my pc)
Sep 25 15:03:32 <jmlich> Here is something about http://wiki.maemo.org/Extras and http://wiki.maemo.org/Extras-devel ; Beware It may damage OS.
Sep 25 15:03:40 <quentin__> yes
Sep 25 15:03:49 <quentin__> i have put extras-devel
Sep 25 15:04:02 <quentin__> register , in application "market"
Sep 25 15:04:14 <stryngs> No bash update yet...
Sep 25 15:04:26 <quentin__> ôÔ?
Sep 25 15:04:51 <quentin__> i read , before , that will be dangerous , but it's ok for the moment
Sep 25 15:05:32 <quentin__> just same application doesn't want be install on phone .. from extras-devel
Sep 25 15:06:08 <quentin__> it told me : you need libqt4-*** ; libqt4-*** etc ...
Sep 25 15:07:08 <CapsuL> i am auentin__
Sep 25 15:07:17 <CapsuL> i am quentin__
Sep 25 15:08:06 <elias_a_> Are you also 1uentin__ ? :P
Sep 25 15:10:33 <quentin__> no ?
Sep 25 15:10:34 <quentin__> :)
Sep 25 15:10:40 <quentin__> look ip ;)
Sep 25 15:10:48 <quentin__> okok
Sep 25 15:11:22 <quentin__> so ... on step of my answer way is suspicious ...
Sep 25 15:11:54 <quentin__> if i put on my Nokia N900 's terminal that : apt-get update
Sep 25 15:12:18 <quentin__> i fear that broken my phone ... if it download too many files =/
Sep 25 15:18:01 <quentin__> maybe  sdk pr1.2 autobuilder ôÔ
Sep 25 15:21:36 <quentin__> this is maemo channel :( and nobody can be help me ...
Sep 25 15:21:49 <quentin__> i know n900 nokia is older 
Sep 25 15:22:26 <quentin__> but plz , i never can be install application without ur help 
Sep 25 15:24:01 <quentin__> i try to resume : i have extras-devel catalog , i looking for install somes applications (chess game) and it told me : impossible u missing/need libqt4-core , libqt4-script , libqt4...........
Sep 25 15:24:34 <quentin__> when i try to install from catalog QT4 python full .. same problem ...
Sep 25 15:25:24 <quentin__> i am blocked , google can be help me :(
Sep 25 15:26:01 <quentin__> i stay here , plz ask me , tell me when u will see precedent message 
Sep 25 15:34:20 <Sc0rpius> env x='() { :;}; echo vulnerable' bash -c "echo this is a test"
Sep 25 15:34:41 <Sc0rpius> quentin__, you have CSSU installed?
Sep 25 15:34:44 <Sc0rpius> if not, you should.
Sep 25 15:39:31 <quentin__> re
Sep 25 15:39:38 <quentin__> how can i saw that ?
Sep 25 15:39:47 <quentin__> can i see that ?
Sep 25 15:39:53 <quentin__> Sc0rpius: 
Sep 25 15:40:07 <Sc0rpius> go to Settings
Sep 25 15:40:17 <Sc0rpius> at the end you should see "About product"
Sep 25 15:40:27 <Sc0rpius> if you also see "About Community SSU" then you have CSSU installed
Sep 25 15:40:33 <Sc0rpius> if you don't, click on About product and tell us the version
Sep 25 15:40:40 <Sc0rpius> it should be 21.2011.38-1
Sep 25 15:40:57 <quentin__> okok
Sep 25 15:41:55 <quentin__> about terminal 
Sep 25 15:42:00 <quentin__> nokia n900
Sep 25 15:42:06 <quentin__> Maemo 5
Sep 25 15:42:31 <quentin__> version : 2.2009.51-1.205.1
Sep 25 15:43:03 <Sc0rpius> Jesus Christ.
Sep 25 15:43:09 <Sc0rpius> that's your problem
Sep 25 15:43:28 <Sc0rpius> you have never updated your N900 and the applications need a recent version of the OS and that's why they fail when you try to install it
Sep 25 15:44:27 <quentin__> yes i know -_-
Sep 25 15:44:50 <quentin__> but how can be upgrade it ?
Sep 25 15:45:01 <quentin__> but how can i upgrade it ?
Sep 25 15:45:04 <quentin__> ^
Sep 25 15:45:05 <quentin__> ^^
Sep 25 15:46:00 <Sc0rpius> you should have received the update automatically but if you don't, then I have no clue
Sep 25 15:46:07 <Sc0rpius> when I said you should have
Sep 25 15:46:10 <Sc0rpius> I meant back in 2010
Sep 25 15:46:12 <Sc0rpius> :)
Sep 25 15:46:47 <quentin__> ôÔ
Sep 25 15:47:08 <quentin__> i just receveid my phone
Sep 25 15:47:12 <quentin__> 2 week ago ;)
Sep 25 15:47:32 <Sc0rpius> why did you buy an extremely old phone like the N900
Sep 25 15:47:39 <quentin__> hehe
Sep 25 15:47:45 <quentin__> u know for what :)
Sep 25 15:48:02 <quentin__> keyboard
Sep 25 15:48:05 <quentin__> and linux ...
Sep 25 15:48:13 <quentin__> that sell me dream 
Sep 25 15:48:37 <quentin__> i install ubuntu on my pc since 4/5 month , and that is magic ...
Sep 25 15:49:10 <quentin__> after i switch on other linux os
Sep 25 15:50:06 <Sc0rpius> when you go to the application manager, and click on update, don't you receive the notification to upgrade the system?
Sep 25 15:50:27 <Sc0rpius> Android also runs in a Linux kernel
Sep 25 15:50:33 <Sc0rpius> and I believe iOS too.
Sep 25 15:50:48 <quentin__> i saw that after :)
Sep 25 15:50:51 <Sc0rpius> you can even run a whole Ubuntu in an Android phone
Sep 25 15:51:01 <quentin__> wooow
Sep 25 15:51:03 <quentin__> :D
Sep 25 15:51:08 <quentin__> didn't know that
Sep 25 15:51:16 <Sc0rpius> an a quad-core Android phone is just like $150
Sep 25 15:51:21 <Sc0rpius> (like the Moto G)
Sep 25 15:51:36 <quentin__> yes :)
Sep 25 15:51:46 <quentin__> but  i am nostalgic
Sep 25 15:51:58 <quentin__> i little option on N900
Sep 25 15:52:00 <quentin__> is nice 
Sep 25 15:52:25 <quentin__> like F.M UDP :D 
Sep 25 15:52:36 <quentin__> HDMI connection
Sep 25 15:52:40 <quentin__> keyboard
Sep 25 15:52:49 <quentin__> and geek mind
Sep 25 15:52:53 <Sc0rpius> I see
Sep 25 15:53:14 <Sc0rpius> well you have to figure out a way to install PR1.3.1 and if you don't receive the OTA update I don't have a clue how you can update it
Sep 25 15:53:28 <quentin__> :/
Sep 25 15:53:41 <freemangordon> nokia repos down?
Sep 25 15:54:08 <freemangordon> otherwise - just set your date so 2009 :)
Sep 25 15:54:10 <quentin__> ovi store is down 
Sep 25 15:54:11 <freemangordon> *to
Sep 25 15:54:44 <freemangordon> set the year to 2009, start HAM and refresh
Sep 25 15:54:56 <quentin__> what is HAM ?
Sep 25 15:55:08 <freemangordon> Hildon Application Manager
Sep 25 15:55:19 <quentin__> from application catalog ?
Sep 25 15:55:38 <freemangordon> what is "application catalog"?
Sep 25 15:55:53 <quentin__> like etras-devel
Sep 25 15:56:03 <quentin__> where i find application
Sep 25 15:56:05 <freemangordon> HAM is http://www.youtube.com/watch?v=XjAwC4nM5cE
Sep 25 15:56:35 <freemangordon> make sure you have extras-testing and extras-devel catalog disable
Sep 25 15:56:41 <freemangordon> *disabled
Sep 25 15:56:54 <freemangordon> and make sure you have all Nokia catalogs enabled
Sep 25 15:57:06 <quentin__> HAM is application catalog for me
Sep 25 15:57:19 <quentin__> we talk about same think :p
Sep 25 15:57:46 <freemangordon> well, you'd better use the correct names, otherwise you risk to be misunderstood :)
Sep 25 15:57:52 <quentin__> wait , i click on update , and it make verification ...
Sep 25 15:58:05 <quentin__> ^^
Sep 25 15:58:06 <quentin__> pardon
Sep 25 15:58:09 <quentin__> i am french 
Sep 25 15:58:10 <freemangordon> you need to click "Refresh" first
Sep 25 15:58:11 <quentin__> :D
Sep 25 15:58:20 <freemangordon> I am not, so what :P
Sep 25 15:58:35 <quentin__> because all french make battle with english lol
Sep 25 15:58:50 <freemangordon> in the menu there is a "Refresh" option
Sep 25 15:59:12 <freemangordon> you ned to use it, so HAM to update its repository info
Sep 25 15:59:16 <freemangordon> *need
Sep 25 15:59:20 <quentin__> i stay 6 month in london for learn more my studiant year lol and i am ever bad :p 
Sep 25 15:59:23 <quentin__> yes
Sep 25 15:59:35 <quentin__> i clik on refresh
Sep 25 15:59:36 <quentin__> i wait
Sep 25 16:00:09 <quentin__> it's ok
Sep 25 16:00:18 <freemangordon> it is ok as in?
Sep 25 16:00:19 <quentin__> but i just see application 
Sep 25 16:00:29 <freemangordon> give it some time
Sep 25 16:00:50 <quentin__> i download it before .. i think it's update for it
Sep 25 16:00:58 <quentin__> (application , no phone)
Sep 25 16:01:13 <freemangordon> you should see "Maemo Update" or similar
Sep 25 16:01:19 <quentin__> no
Sep 25 16:01:41 <freemangordon> and you're sure the date is 2009?
Sep 25 16:01:46 <freemangordon> year that is
Sep 25 16:01:57 <quentin__> 2 THEME , nmap (-__-) , wifiinfo , n900-fmrx-enabler
Sep 25 16:03:18 <quentin__> 2.2009.51-1.205.1
Sep 25 16:03:23 <freemangordon> hmm, are you sure you're not on the latest PR?
Sep 25 16:03:53 <quentin__> i don't think
Sep 25 16:03:59 <freemangordon> quentin__: in settings -> about product
Sep 25 16:04:04 <freemangordon> what is the version there?
Sep 25 16:04:11 <quentin__> 2.2009.51-1.205.1
Sep 25 16:04:17 <freemangordon> oh, 205
Sep 25 16:04:19 <quentin__> Maemo 5
Sep 25 16:04:26 <quentin__> nokia N900
Sep 25 16:04:30 <freemangordon> yeah, yeah
Sep 25 16:05:14 <freemangordon> this is a special firmware cooked for vodafone 
Sep 25 16:05:55 <freemangordon> you'd better reflash to global
Sep 25 16:06:02 <quentin__> arf ..
Sep 25 16:06:23 <quentin__> i begin i buy it for flashing in pentest version ...
Sep 25 16:06:30 <quentin__> but after try it 
Sep 25 16:06:42 <quentin__> i find n900 maemo perfect
Sep 25 16:06:52 <quentin__> and i forgot that 
Sep 25 16:06:56 <quentin__> :)
Sep 25 16:07:26 <quentin__> but ok .. thx u for ur diagnostic :)
Sep 25 16:07:46 <freemangordon> if you have nothing valuable on the phone, just flash rootfs and emmc and you'll be fine
Sep 25 16:08:26 <quentin__> yes , that take lot of time :/
Sep 25 16:08:30 <quentin__> i try later
Sep 25 16:08:59 <freemangordon> no, that takes 5 minutes
Sep 25 16:09:00 <quentin__> which is the last flash rom for n900 ?
Sep 25 16:09:03 <freemangordon> ~flashing
Sep 25 16:09:04 <infobot> hmm... maemo-flashing is http://wiki.maemo.org/Updating_the_tablet_firmware
Sep 25 16:09:05 <quentin__> :/
Sep 25 16:09:40 <freemangordon> 21.2011.38-1
Sep 25 16:09:45 <quentin__> for u 5 min
Sep 25 16:10:00 <Sc0rpius> maybe it's more than 5 minutes but it's fun
Sep 25 16:10:04 <quentin__> i need 15 for read the link :D 10 for understand
Sep 25 16:10:04 <Sc0rpius> don't take the fun out of it
Sep 25 16:10:06 <quentin__> ^^
Sep 25 16:10:47 <quentin__> and maybe 30 for try to install/flash it
Sep 25 16:10:48 <quentin__> ^^
Sep 25 16:12:11 <quentin__> and this freemangordon http://repository.maemo.org/pool/maemo5.0/free/q/qt4-x11/
Sep 25 16:12:39 <quentin__> and this, freemangordon,  http://repository.maemo.org/pool/maemo5.0/free/q/qt4-x11/
Sep 25 16:22:07 <Sc0rpius> quentin__, you can't install Qt if you don't upgrade
Sep 25 16:22:18 <Sc0rpius> I mean you can't upgrade your Qt if you don't upgrade the whole thing
Sep 25 16:23:01 <quentin__> okok
Sep 25 16:23:04 <quentin__> :)
Sep 25 16:23:09 <quentin__> i will flash it 
Sep 25 16:48:06 <sixwheeledbeast> freemangordon: hows nm-nav going :)
Sep 25 17:19:13 <freemangordon> sixwheeledbeast: just REed the last function :)
Sep 25 17:20:05 <sixwheeledbeast> freemangordon: cool, will it be going into cssu-devel for testing?
Sep 25 17:42:14 <stryngs> Any plans to patch bash guys/gals?
Sep 25 17:43:05 <APic> I use zsh ;-P
Sep 25 17:43:53 <Humpelstilzchen> I thought be have busybox
Sep 25 17:44:34 <ecc3g> some people have bash installed...
Sep 25 17:49:47 <sixwheeledbeast> stryngs: ask the maintainer it is not covered under CSSU, default shell is messybox
Sep 25 17:50:05 * stryngs doesnt use cssu =)
Sep 25 17:50:11 <sixwheeledbeast> bash is an extras pkg
Sep 25 17:50:35 <sixwheeledbeast> ~pkg
Sep 25 17:50:35 <infobot> i guess pkg is http://maemo.org/packages/
Sep 25 17:50:38 <stryngs> Right, i get that.  What i was getting at is wondering if there is any preplanned attack to work on the patch to bash for the maemo repo 
Sep 25 17:50:53 <sixwheeledbeast> :shrug:
Sep 25 17:51:04 <stryngs> Glad to see we're on point here in #maemo land
Sep 25 17:51:08 * stryngs goes back to reality
Sep 25 17:52:00 <sixwheeledbeast> marmistrz on TMO maybe work asking. He is the maintainer of bash-opt (optified bash4)
Sep 25 17:52:14 <stryngs> interesting
Sep 25 17:52:16 <stryngs> thanks
Sep 25 17:53:21 <sixwheeledbeast> http://maemo.org/profile/view/onion/ is the maintainer of bash4 package
Sep 25 17:54:08 <sixwheeledbeast> http://maemo.org/profile/view/zap/ is the maintainer of bash3
Sep 25 18:34:17 <freemangordon> sixwheeledbeast: sure, once I fix whatever I can catch as a bug
Sep 25 18:45:49 <stryngs> Anyone in here ncurses smart?
Sep 25 18:45:58 <stryngs> I'm trying to build ncurses5.9 for the n900
Sep 25 18:46:08 <stryngs> Well, hell, it's built
Sep 25 18:46:15 <stryngs> I want to know about ncursesw
Sep 25 19:19:57 <CapsuL> hey :) 
Sep 25 19:21:04 <stryngs> hi
Sep 25 19:31:59 <CapsuL> hi
Sep 25 19:32:36 <CapsuL> how can i know my nokia n900 phone password?
Sep 25 19:33:08 <CapsuL> before try to flashing it:)
Sep 25 19:37:18 <Tekk_> CapsuL: you could do sudo passwd
Sep 25 19:37:28 <Tekk_> and set one
Sep 25 19:37:31 <Tekk_> but that may ruin things, idk
Sep 25 19:37:33 <Tekk_> I haven't done it
Sep 25 19:39:13 <CapsuL> ok thx
Sep 25 19:39:43 <stryngs> CapsuL: WAIT
Sep 25 19:39:48 <stryngs> Which password
Sep 25 19:39:50 <stryngs> root password
Sep 25 19:39:53 <stryngs> or device lock password
Sep 25 19:40:07 <stryngs> Flashing DOES NOT RESET the device lock, at least from what i can tell
Sep 25 19:40:14 <stryngs> mine has stayed through massive reflashes
Sep 25 19:49:44 <Pali> freemangordon: what about RE device lock libraries and understand where and how is lock password stored?
Sep 25 19:49:45 <Pali> ^^^
Sep 25 19:50:03 <Tekk_> Pali: it's probably stored in the firmware
Sep 25 19:50:11 <Tekk_> so are your alarms actually
Sep 25 19:50:16 <Tekk_> your alarms will fire even if your phone is off
Sep 25 19:56:03 <freemangordon> Pali: why not, lemme finish nm-nav-provider first
Sep 25 19:56:19 <Pali> ok, this was just idea
Sep 25 19:56:57 <freemangordon> Pali: in that regard - http://pastebin.com/qZzG7AGE
Sep 25 19:57:11 <freemangordon> could you have a look at it
Sep 25 19:57:20 <Pali> ok, whats problem?
Sep 25 19:57:42 <freemangordon> G_LOCK is called (or could be called) 3 times ina a row
Sep 25 19:57:51 <freemangordon> GL_LOCK translates to g_mutex_lock
Sep 25 19:58:11 <freemangordon> whi has undefined behaviour if called with the mutex already held
Sep 25 19:58:15 <freemangordon> *which
Sep 25 19:58:16 <Pali> so G_(UN)LOCK is mutex?
Sep 25 19:58:20 <freemangordon> yes
Sep 25 19:58:37 <freemangordon> http://irtfweb.ifa.hawaii.edu/SoftwareDocs/gtk20/glib/glib-threads.html#G-LOCK-CAPS
Sep 25 19:59:42 <Pali> does not con_ic_connection_connect relase lock?
Sep 25 20:00:15 <freemangordon> it is called in the same thread I guess (this is multy-threaded program)
Sep 25 20:00:28 <freemangordon> oh, no, why should it
Sep 25 20:00:38 <freemangordon> the mutex is a static global variable
Sep 25 20:00:53 <Pali> are you sure that this code is correctly disassembled?
Sep 25 20:01:06 <freemangordon> I looked at the assembly code as well
Sep 25 20:01:19 <freemangordon> wanna habd you db?
Sep 25 20:01:19 <Pali> I doubt that somebody will write code lock(mutex); unlock(mutex)
Sep 25 20:01:24 <freemangordon> *hand
Sep 25 20:01:35 <freemangordon> me too
Sep 25 20:01:51 <Pali> this looks like gcc optimizer
Sep 25 20:01:59 <Pali> when it did something bad
Sep 25 20:02:07 <freemangordon> though, there might have been some code gcc optimized
Sep 25 20:02:31 <Pali> I saw code which gcc incorrectly compiled with -O3 (or -O2)
Sep 25 20:03:07 <freemangordon> besides that, my concern is that g_mutex_lock (or G_LOCK) is called more than once in the same thread with the mutex already held
Sep 25 20:03:22 <Pali> then it freeze
Sep 25 20:03:25 <freemangordon> yes
Sep 25 20:03:28 <freemangordon> exactly
Sep 25 20:03:40 <Pali> but if function con_ic_connection_connect release lock it is OK
Sep 25 20:03:59 <freemangordon> that function has no way of knowing about that mutex
Sep 25 20:04:08 <freemangordon> iiuc
Sep 25 20:04:22 <freemangordon> it is external library function
Sep 25 20:04:27 <freemangordon> from libconnic
Sep 25 20:04:34 <Pali> and there is g_signal_connect_data with callback con_ic_status_handler which release lock
Sep 25 20:04:39 <freemangordon> yes
Sep 25 20:04:45 <Pali> how signals in gtk working?
Sep 25 20:04:54 <Pali> are immediately callback executed?
Sep 25 20:05:07 <freemangordon> oh, so you think the callback will be called before we call the second G_LOCK?
Sep 25 20:05:28 <Pali> or executaion is postponed when main loop enters again?
Sep 25 20:05:47 <Pali> in Qt4 slots are executed immediately
Sep 25 20:05:54 <freemangordon> I doubt this is the main loop, this is executed in a thread
Sep 25 20:06:08 <freemangordon> hmm, going to attach a debugger
Sep 25 20:06:25 <Pali> but you can connect qt4 slots with some special argument, executation can be postponed until application returns back to mainloop
Sep 25 20:06:37 <freemangordon> yes
Sep 25 20:06:45 <Pali> freemangordon: with main loop I mean main glib loop
Sep 25 20:06:50 <freemangordon> but I dunno hows that in glib
Sep 25 20:06:54 <Pali> that loop which process executaion
Sep 25 20:06:55 <freemangordon> I know
Sep 25 20:07:08 <Pali> I think that glib signals are always postponed
Sep 25 20:07:17 <Pali> but not remeber this if it true
Sep 25 20:07:22 <freemangordon> but it makes no sense, you can't release mutex held by a thread A from the main loop :)
Sep 25 20:07:42 <freemangordon> so I conclude that callback is executed in the same thread
Sep 25 20:08:11 <Pali> if con_ic_status_handler is called from same thread and by function g_signal_connect_data then everything is OK
Sep 25 20:08:25 <Pali> in qt4 it is for sure
Sep 25 20:08:38 <Pali> in qt3 not
Sep 25 20:08:50 <freemangordon> g_signal_connect data calls nothin, it just connects the "signal" to a "slot" speaking in qt terms
Sep 25 20:09:12 <Pali> I know but there can be already sent signal which waiting
Sep 25 20:09:26 <Pali> and that connect can read it and process it
Sep 25 20:09:59 <freemangordon> then we have race condition
Sep 25 20:10:05 <Pali> but I think that code was incorrectly compiled or incorrectly decompiled
Sep 25 20:10:29 <freemangordon> yeah, going to attach a debugger, brb
Sep 25 20:11:00 <Pali> maybe branch instruction which is after con_ic_connection_connect and after g_signal_connect_data should point before G_UNLOCK (after G_LOCK)
Sep 25 20:12:48 <Pali> or I have another idea: that G_LOCK(conn_ic); G_UNLOCK(conn_ic); code could be barrier
Sep 25 20:13:01 <Pali> which waits until execution of con_ic_status_handler() finish
Sep 25 20:13:10 <freemangordon> yes, could be
Sep 25 20:13:33 <Pali> g_signal_connect_data could be used there for blocking call
Sep 25 20:13:34 <freemangordon> but what will stop it from executing in the thread A?
Sep 25 20:13:42 <freemangordon> no, it is not blocking
Sep 25 20:14:05 <freemangordon> only con_ic_connection_connect could be
Sep 25 20:14:08 <Pali> yes it is not blocking call and nokians used this hack to create blocking call :D
Sep 25 20:14:30 <Pali> and that thread stops execution
Sep 25 20:14:36 <freemangordon> Pali: it deadlocks
Sep 25 20:14:49 <freemangordon> if G_LOCK is called for the second time in the same thread
Sep 25 20:14:56 <Pali> if this is multi thread app then another thread can process con_ic_status_handler callback
Sep 25 20:15:11 <freemangordon> so what, thread A is already in deadlock
Sep 25 20:15:12 <Pali> and unblock mutex
Sep 25 20:15:23 <freemangordon> it can't, as it does not hold it
Sep 25 20:15:41 <freemangordon> it is held by thread A, which is in a deadlock
Sep 25 20:15:44 <Pali> is not conn_ic shared across all threads?
Sep 25 20:15:53 <freemangordon> it is mutex :)
Sep 25 20:16:03 <freemangordon> GStaticMutex to be precise
Sep 25 20:16:38 <freemangordon> so yes, all threads can get hold on it, but not simultaneously
Sep 25 20:16:51 <freemangordon> obviously :D
Sep 25 20:17:15 <Pali> is not conn_ic shared across all threads and any thread can lock or unlock it?
Sep 25 20:17:37 <freemangordon> sure
Sep 25 20:17:54 <Pali> so when A lock it, A try to lock again (will wait), and after some time B unlock it
Sep 25 20:18:08 <freemangordon> no, see http://irtfweb.ifa.hawaii.edu/SoftwareDocs/gtk20/glib/glib-threads.html#g-mutex-lock
Sep 25 20:18:31 <freemangordon> Pali: this is mutex, only the thread that lock it can unlock it
Sep 25 20:20:01 <Pali> so it is really not GStaticRecMutex?
Sep 25 20:20:43 <freemangordon> it is not
Sep 25 20:20:49 <freemangordon> aiui
Sep 25 20:20:54 <Pali> ok, no idea then...
Sep 25 20:21:05 <freemangordon> debugger attached, lets see
Sep 25 20:22:13 <freemangordon> Pali: in the meanwhile - that whole function is buggy as hell, there is a race condition in the first if() to start with
Sep 25 20:22:43 <Pali> ah... reason why it is closed? :D
Sep 25 20:23:02 <freemangordon> it is not protected by a mutex, so 2 threads could simultaneously do "if (!priv->con_ic_conn)" ;)
Sep 25 20:23:21 <freemangordon> anyway, got in the bp
Sep 25 20:35:45 <joga> hmm, is dhcp client in maemo vulnerable to this https://www.trustedsec.com/september-2014/shellshock-dhcp-rce-proof-concept/
Sep 25 20:37:04 <joga> apparently it's udhcpc, are the shell scripts involved?
Sep 25 20:37:07 <joga> *there
Sep 25 20:37:53 <joga> (bash I mean, but it's not there by default I guess?)
Sep 25 20:46:21 <Pali> joga: no maemo dhcp client is not vulnerable to that bash bug
Sep 25 20:46:43 <joga> k
Sep 25 20:46:43 <Pali> maemo (core) system does not use bash
Sep 25 20:46:54 <joga> yeah I guess it was busybox...
Sep 25 20:47:12 <joga> been using bash on it otherwise for so long forgot already ;)
Sep 25 20:50:11 <freemangordon> Pali: well, g_mutex_lock is simply called twice :(
Sep 25 20:50:33 <Pali> and deadlock
Sep 25 20:50:51 <freemangordon> no idea, still debugging
Sep 25 20:51:35 <freemangordon> no deadlock
Sep 25 20:52:08 <freemangordon> oh, it locked
Sep 25 20:52:45 <freemangordon> or rather the callback was called before pthread_mutex_lock exited
Sep 25 20:52:52 <freemangordon> so your theory seems right
Sep 25 20:53:43 <freemangordon> (gdb) info threads
Sep 25 20:53:45 <freemangordon>   Id   Target Id         Frame 
Sep 25 20:53:47 <freemangordon>   2    LWP 2938          0x4016f27c in pthread_mutex_lock ()
Sep 25 20:53:49 <freemangordon>    from /lib/libpthread.so.0
Sep 25 20:53:51 <freemangordon> * 1    LWP 2833          0x0000c014 in ?? ()
Sep 25 20:53:54 <freemangordon> Pali: callback is called from the main glib thread
Sep 25 20:54:01 * freemangordon is confused
Sep 25 20:56:36 <Pali> so callback called from main glib thread unlocked that mutex or not?
Sep 25 20:57:23 <freemangordon> will check now, I was missing a bp
Sep 25 21:01:37 <freemangordon> Pali: yes, it unlocked the motex
Sep 25 21:01:40 <freemangordon> *mutex
Sep 25 21:01:56 * freemangordon is even more confused
Sep 25 21:03:03 <Pali> so it is as I was expecting...
Sep 25 21:03:06 <freemangordon> seems like docs are not telling the whole truth :)
Sep 25 21:03:15 <freemangordon> yes, exactly as you were expecting
Sep 25 21:03:39 <freemangordon> but how is that possible?
Sep 25 21:03:59 <freemangordon> to release a mutex held by thread A from thread B?!?
Sep 25 21:06:23 <Pali> it is implemented by linux futexes?
Sep 25 21:06:32 <Pali> or something else?
Sep 25 21:06:41 <freemangordon> pthread_mutex_t
Sep 25 21:06:52 <freemangordon> and this is cute http://stackoverflow.com/questions/5454746/pthread-mutex-lock-unlock-by-different-threads
Sep 25 21:07:59 <freemangordon> "What you've done is simply not legal, and the behavior is undefined."
Sep 25 21:08:03 <freemangordon> :D
Sep 25 21:09:35 <Pali> so it using futexes
Sep 25 21:09:50 <freemangordon> ok, if you say so :)
Sep 25 21:09:57 <freemangordon> so?
Sep 25 21:10:04 <Pali> and I think you can unlock futex from another thread
Sep 25 21:10:24 <Pali> as it is same (mmapped) memory
Sep 25 21:10:36 <freemangordon> ok, my conclusion is that this is illegal and implementation specific
Sep 25 21:10:42 <Pali> yes
Sep 25 21:10:43 <freemangordon> we must rework it
Sep 25 21:11:15 <freemangordon> but at least now I understand what happens
Sep 25 21:11:19 <Pali> but on linux 2.6+ when futexes are used by default it is working
Sep 25 21:12:24 <freemangordon> I don;t believe this is the only way to make async function sync
Sep 25 21:12:54 <Pali> never! do not do that
Sep 25 21:13:26 <freemangordon> Pali: mind you, this is what we all have in our devices ;)
Sep 25 21:13:28 <Pali> 3x I thought that I have working code (in qt) which creating blocking function from async callls
Sep 25 21:14:09 <Pali> and after some time I found special case when it is not working
Sep 25 21:14:14 <freemangordon> I'd rather start another thread and wait() on it
Sep 25 21:14:35 <freemangordon> or something like that
Sep 25 21:14:52 <freemangordon> qnyway, I'll find the proper way once I am on it
Sep 25 21:16:16 <freemangordon> Pali: thanks a lot for the help
Sep 26 01:10:46 <Oksana> A question: is GPS firmware of N900 closed-source?
Sep 26 01:16:34 <jonwil> GPS on the N900 is handled by the cellular modem
Sep 26 01:16:38 <jonwil> and the firmware for that is closed source
Sep 26 01:16:55 <jonwil> On the userspace side its handled by a number of daemons and libraries, all of which are closed source
Sep 26 01:17:31 <jonwil> Although I have recently figured out a good whack of how the userspace bits work internally through reverse engineering
Sep 26 01:18:38 <Oksana> Thank you!
**** ENDING LOGGING AT Fri Sep 26 03:00:00 2014