|
From: Erik de C. L. <eri...@me...> - 2005-12-25 23:56:54
|
HI all, I'm running valgrind 3.0.1 installed from Ubuntu Linux on an machine running Ubuntu. I running it on Csound5 compiled from CVS on sourceforge and I get: ==16631== ==16631== Conditional jump or move depends on uninitialised value(s) ==16631== at 0x1B8EC852: (within /lib/ld-2.3.5.so) ==16631== by 0x1BB86212: dl_open_worker (dl-open.c:343) ==16631== by 0x1B8EF105: (within /lib/ld-2.3.5.so) ==16631== by 0x1BB869B8: _dl_open (dl-open.c:577) ==16631== by 0x1BA27CD4: dlopen_doit (dlopen.c:59) ==16631== by 0x1B8EF105: (within /lib/ld-2.3.5.so) ==16631== by 0x1BA28284: _dlerror_run (dlerror.c:162) ==16631== by 0x1BA27D24: dlopen@@GLIBC_2.1 (dlopen.c:78) ==16631== by 0x1B997EA1: csoundOpenLibrary (Top/csmodule.c:560) ==16631== by 0x1B998095: csoundLoadExternal (Top/csmodule.c:226) ==16631== by 0x1B9988A3: csoundLoadModules (Top/csmodule.c:384) ==16631== by 0x1B99A839: csoundPreCompile (Top/csound.c:930) @@ don't know what type '_' is @@ parsing __common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; gave NULL type (_common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; remains) @@ expected ':' at struct method MANGLE-ARGS (remains="common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;;") @@ parsing (0,909)=s1__mt_alloc<char,__gnu_cxx::__common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; gave NULL type (s1__mt_alloc<char,__gnu_cxx::__common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; remains) --16631-- INTERNAL ERROR: Valgrind received a signal 11 (SIGSEGV) - exiting --16631-- si_code=1; Faulting address: 0x0; sp: 0xB0B29B00 valgrind: the 'impossible' happened: Killed by fatal signal Erik -- +-----------------------------------------------------------+ Erik de Castro Lopo +-----------------------------------------------------------+ "Any sufficiently complicated C or Fortran program contains an ad-hoc, informally-specified, bug-ridden, slow implementation of half of CommonLisp." -- Greenspuns Tenth Rule Of Programming |
|
From: Erik de C. L. <eri...@me...> - 2005-12-27 05:10:33
|
Andr=E9 W=F6bbeking wrote: > Do you compile with visibility support? I just used the standard compile options used by Csound5 from CVS. > Debian's gcc was "broken" so you could=20 >=20 > 1) disable visibility support > 2) set the enviroment variable GLIBCXX_FORCE_NEW=3D1 > 3) update the compiler Sorry, do I do 1, 2 *and* 3 or 1, 2 *or* 3? I tried setting GLIBCXX_FORCE_NEW=3D1 but valgrind still crashes. I'm also not sure how I should disable visibility support.=20 The compiler is currently g++-4.0.2. Would moving to 3.4 or 3.3 help? Erik --=20 +-----------------------------------------------------------+ Erik de Castro Lopo +-----------------------------------------------------------+ "The X-files is too optimistic. The truth is not out there." -- Anthony Ord |
|
From: <Woe...@on...> - 2005-12-27 22:55:49
|
On Tuesday 27 December 2005 06:10, Erik de Castro Lopo wrote: > Andr=E9 W=F6bbeking wrote: > > Do you compile with visibility support? > > I just used the standard compile options used by Csound5 from CVS. > > > Debian's gcc was "broken" so you could > > > > 1) disable visibility support > > 2) set the enviroment variable GLIBCXX_FORCE_NEW=3D1 > > 3) update the compiler > > Sorry, do I do 1, 2 *and* 3 or 1, 2 *or* 3? or > I tried setting GLIBCXX_FORCE_NEW=3D1 but valgrind still crashes. As already said in other mails it's probably also a valgrind bug. My=20 tips help to run a STL using C++ program compiled with the broken=20 compiler. > I'm also not sure how I should disable visibility support. RTFM, look after -fvisibility* > The compiler is currently g++-4.0.2. Would moving to 3.4 or 3.3 > help? Probably. Cheers, Andr=E9 |
|
From: Julian S. <js...@ac...> - 2005-12-27 13:49:27
|
> @@ parsing __common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; > gave NULL type (_common_pool_policy<__gnu_cxx::__pool, true> > >:(0,908),0,64;; remains) Looks like a bug in the .stabs reader. Try 3.1.0 and/or use dwarf instead. J |
|
From: Erik de C. L. <eri...@me...> - 2005-12-31 13:54:57
|
Julian Seward wrote: > > > @@ parsing __common_pool_policy<__gnu_cxx::__pool, true> >:(0,908),0,64;; > > gave NULL type (_common_pool_policy<__gnu_cxx::__pool, true> > > >:(0,908),0,64;; remains) > > Looks like a bug in the .stabs reader. Try 3.1.0 and/or use dwarf instead. Yep, worked fine wiyj -gdwarf-3 instead of -gstabs. Erik -- +-----------------------------------------------------------+ Erik de Castro Lopo +-----------------------------------------------------------+ "It's pretty simple. [Large companies are] allowed to own either a content creation empire - a studio or label - or a playback device empire (Walkmans, or PCs). But ... can't have both." -- Andrew Orlowski |