On Mon, 2010-08-02 at 13:47 +0200, [email protected] wrote: > On 2010-08-02 14:20, Sven Neumann wrote: > > I've tried that now, unfortunately without much success. Here's the > > output from a simple test application with debug output enabled. As you > > can see, dwarf_step() fails and then the code crashes in the alternative > > approach using the frame chain: > > .... > > >_ULarm_step: dwarf_step() failed (ret=0), trying > > frame-chain > > caught signal 11 > > It's strange that dwarf_step() failed, less strange that frame-chain > unwinding could fail. What options did you use to compile your test > program? IIRC frame chain unwinding requires -fno-omit-frame-pointer and > -mapcs-frame options enabled.
I didn't add these extra options, and of course I would prefer if I wouldn't have to. So far I have tried to add -funwind-tables, but that did not seem to help with dwarf_step(). The compiler options in use at the moment are -mtune=iwmmxt -march=iwmmxt -mabi=aapcs-linux -msoft-float -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O2 -pipe -funwind-tables Sven _______________________________________________ Libunwind-devel mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/libunwind-devel
