round two :) On Tue, Jun 09, 2009 at 03:19:22PM +0200, Giacomo Boffi wrote: > > i can't reproduce this problem (and i'm also running 2.6.29/amd64). > > i'm running an AMD cpu, but debian's architecture is i386 ah, okay. i don't think it's relevant but thanks for clarifying. > > are you sure you don't have a compiz component installed from source, > > a non-standard library installed or perhaps something from experimental? > > i don't know let's try: sudo apt-get install apt-show-versions apt-show-versions -b | grep -v -E '/sid|/squeeze|/testing|/unstable' this should give a listing of all packages that aren't known in testing or unstable (there could possibly be a large number of "No available version in archive" lines output, which are probably okay but it would be good if i could go through them for the sake of being thorough. > ,----[ cat /etc/ld.so.conf /etc/ld.so.conf.d/* ] > | include /etc/ld.so.conf.d/*.conf > | > | /usr/lib/3dnow/atlas > | /usr/lib/3dnow > | /usr/lib/atlas i don't have these, but it doesn't look like anything strange in the ldd output below so i'm tentatively thinking they're not relevant. > ,----[ sudo updatedb && locate libcompizconfig.so | xargs ldd ] > | /usr/lib/libcompizconfig.so.0: <snip> how about this one? ldd `which compiz.real` > > dpkg -l '*compiz*' > ,---- > | Desired=Unknown/Install/Remove/Purge/Hold > | | Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend > | |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) > | ||/ Name Version Description > | +++-=================================-============================-============================================ these are the only differences to my system: > | pn compiz <none> (no description available) > | pn compiz-gnome <none> (no description available) > | pn compizconfig-backend-gconf <none> (no description available) what window manager are you using? it could be that the problem only asserts itself in non-gnome installations. > > i'm going to check to see if there's anything that has recently > > happened in testing/unstable that i've somehow missed but this > > information will be helpful in any case. in the meantime i'm > > downgrading the bug. i haven't found anything here... one last question. when you get the error: terminate called after throwing an instance of 'std::logic_error' what(): basic_string::_S_construct NULL not valid is it followed by an abort? from some googling (and a basic test) it seems this error stems from something somewhere allocating a std::string(0). if it does give an abort message, could you: * apt-get install gdb * ulimit -c unlimited * invoke compiz again on the cmdline to get the same error this should produce a core dump, which can then get us a backtrace to point at where the exception was thrown with echo bt | gdb /usr/bin/compiz.real core it may be the case that we will need debugging symbols available to read the backtrace. if so i'll provide some debuggable .debs for you to install. sean
Attachment:
signature.asc
Description: Digital signature