Hi, it seems, that using...
kaffe -vmdebug INIT,VMTHREAD,GCSYSALLOC,thread HelloWorld ... (and some additional debug messages I added), the last life signs come usually just before sigwait in KaffePThread_WaitForResume. Where should that signal come from? I'm using this configuration: ./configure --build=i686-linux --host=armv5b-softfloat-linux --enable-xscale --without-x --disable-alsatest --disable-esdtest --disable-sound --enable-pure-java-math --with-jikes=/opt/local/bin/jikes --without-classpath-gtk-awt --without-kaffe-qt-awt --without-kaffe-x-awt --with-threads=unix-pthreads --with-engine=jit --enable-debug --enable-xdebugging Cheers, Jari On Fri, 26 Aug 2005, Jari Korva wrote: > Hi Dalibor and Nils, > > I haven't done any thorough debugging but at least for me the behaviour > seems to be pretty much the same using jit and intrp. In some rare cases > my HelloWorld executes fine but usually it just hangs and I > have to "kill -9" it (often Ctrl-C and kill have no effect). Using > -vmdebug I have found that the hang happens always after > "createDaemon gc" but the final location seems to vary which makes the > debugging a bit difficult... > > Last time I pulled a working Kaffe out of cvs was in February, I think. > Haven't tried libffi lately. > > Cheers, > Jari > > PS. Most of the information I wrote in this mail comes out of memory > so there may be errors :/ > > > On Tue, 23 Aug 2005, Dalibor Topic wrote: > > Hi Nils, > > > > you could try --with-libffi and/or --with-engine=intrp , or, if you want > > to try debugging the problem, see FAQ/FAQ.debugging. > > > > cheers, > > dalibor topic > > > > > _______________________________________________ kaffe mailing list [email protected] http://kaffe.org/cgi-bin/mailman/listinfo/kaffe
