Simulator segfaults
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines · Peter Chubb
"David" = David Mosberger <[email protected]> writes:
Peter> however, I'd expect that attempting to execute an IA32 Peter> instruction that wasn't implemented would generate an Peter> unimplemented instruction trap to the operating system, not Peter> cause a segfault.
David> I agree the Ski behavior isn't pretty for this case, but I David> suspect we may have to just live with it. AFAIK, nobody David> seriously cares about x86 emulation on Ski, because it's David> woefully incomplete at any rate (well, it may be useful for David> DOS-area programs, but certainly not for x86 Linux programs).
Peter> Yes, well, I was trying to use it to work out how to fix IA32 Peter> coredumps on the IA64 kernel --- curerntly an easy way to Peter> crash 2.6.0 is to run an IA32 program that segfaults.
Hmmh, I think Arun has patches to make ia32 core-dumps work. Arun, can you confirm?
--david
Quoted 2 lines · David Mosberger
Hmmh, I think Arun has patches to make ia32 core-dumps work. Arun, can you confirm?
I just posted a patch to linux-ia64.
-Arun