Preemption patch
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line · Peter Chubb
Peter> The main changes are in entry.S to check the RESCHEDULE flag Peter> for kernel threads if preempt_count() is zero, and call Peter> schedule() if so. As my IA64 assembly language skills are Peter> shaky, I'd appreciate someone reviewing that bit of code (as Peter> it's also performance critical).
I looked over your changes to entry.S and they look fine to me. The code probably could be scheduled better, but I assume you're aware of that and you are focusing on correctness for now.
--david