02:04 <Wizzup> freemangordon: btw, no need to look at this per se, but - https://phoenix.maemo.org/view/Images/job/leste-image-droid4-daedalus/2/console

11:04 <Wizzup> freemangordon: ping

11:05 <Wizzup> I am seeing segfaults on armhf (only, not arm64 and amd64) in rtcom el core code, it segfaults in sqlite3_vmprintf / sqlite3_str_vappendf

11:07 <Wizzup> I am not sure how to debug, I have debug symbols and glib attached, but don't se eanything super obvious wrong, and the va_args kind of masks what is going on

11:13 <Wizzup> gdb isn't very helpful is printing va_args

11:14 <Wizzup> sqlite seems to be reading/getting an address of 0xa, which is clearly off

11:17 <Wizzup> I do see:

11:17 <Wizzup> Uninitialised value was created by a stack allocation at 0x48B5104: _add_event_core (eventlogger.c:785)

11:17 <Wizzup> hm..

14:53 <freemangordon> Wizzup: valgrind/asan?

14:55 <Wizzup> valgrind mostly just confused me

14:56 <Wizzup> It is very likely some type/size issue because the problem doesn't occur on 64 bit

18:44 <maemodaedalus> Hi

18:44 <Wizzup> :)

18:56 <freemangordon> heh

19:03 <Wizzup> let's hope this one completes: https://phoenix.maemo.org/view/Images/job/leste-image-droid4-daedalus/4/

19:03 <Wizzup> I might try dist-upgrade over the weekend

20:38 <Wizzup> the oom killer intervened

20:55 <freemangordon> :(

20:55 <Wizzup> running again with more ram :)

20:55 <Wizzup> the whole thing is now tmpfs backed so it no longer takes 5-6 hours

20:56 <Wizzup> I am going to try dist-upgrade on my n900, but it looks like there two more things to sort out qrt qt and this qt base abi thing

20:56 <Wizzup> ooh, right, this is the priority issue

20:56 <Wizzup> freemangordon: do you have notes handy how you solved that?

21:05 <arno11> Wizzup: btw if you dist-upgrade your n900 with so much packages, you should disable boost mode imo and renice xorg,systemui,mce and hildon-desktop instead. this way the device should stay quite responsive during the process and it should avoid potential errors during long updates (apt really doesn't like overclock)

21:15 <freemangordon> Wizzup: sorry, solve what?

21:40 <Wizzup> freemangordon: hi

21:41 <Wizzup> freemangordon: so when you dist upgraded your vm to daedalus, you set some apt repo priority to cause apt to pick the daedalus pkgs over chimaera, because of the extension of the pkgs being +m7 vs +4m7

21:41 <Wizzup> arno11: understood

21:58 <Wizzup> # cat /etc/apt/preferences.d/pin-daedalus

21:58 <Wizzup> Package: *

21:58 <Wizzup> Pin: release o=Maemo Leste,n=daedalus

21:58 <Wizzup> Pin-Priority: 1001

21:59 <Wizzup> this does it

22:01 <Wizzup> 1048 upgraded, 177 newly installed, 1388 downgraded, 6 to remove and 0 not upgraded.

22:01 <Wizzup> lol :D

22:01 <Wizzup> more like dist-downgrade :)

22:03 <arno11> lol indeed

22:37 <Wizzup> would need the above preferences for o=Maemo Leste Extras too I think

22:48 <Wizzup> lol, forgot to change the mkdir :)

23:46 <Wizzup> btw

23:46 <Wizzup> it looks like this is the gconf issue

23:46 <Wizzup> Unpacking libgconf-2-4:armhf (3.2.6-8) over (3.2.6-7) ...

23:46 <Wizzup> dpkg: cycle found while processing triggers:

23:46 <Wizzup> chain of packages whose triggers are or may be responsible:

23:46 <Wizzup> gconf2 -> gconf2

23:46 <Wizzup> it's a cycle from libgconf

23:46 <Wizzup> a combination of these fixes it manually, but I think we might be able to fix this

23:46 <Wizzup> dpkg --configure --force-all dbus-x11

23:46 <Wizzup> (manual)

23:46 <Wizzup> dpkg --configure --force-all gconf2

23:46 <Wizzup> dpkg-reconfigure --force dbus-x11

23:46 <Wizzup> dpkg-reconfigure --force gconf2