Optimizing Compilers, EDA230, 2016

You should present your opinion on the paper in terms of:

1. What performance problem does it try to solve?
   
   
2. How much time do you think you would need to implement it? Take one of:
   • days,
   • weeks,
   • months,
   • years.
   
   
3. Assuming the compiler you use does not perform this optimisation, is it possible to do by hand? If yes, is it a good idea? (always, sometimes, or never?)

Reserved articles

• Hampus Londögård: Operator Strength Reduction
• Mikael Johnsson: The priority-based coloring approach to register allocation
• Linus Röman: Rematerialization
• Johan Ju: Minimizing register usage penalty at procedure calls
• Gustav Ek: A data locality optimizing algorithm
• Jonatan Westholm: Spill Code Minimization via Interference Region Spilling
• Süleyman Savaş:Whole-Function Vectorization
• David Åkerman: An SSA-based algorithm for optimal speculative code motion under an execution profile
• Axel Larsson: Some new approaches to partial inlining
• Henrik Lehtonen: Eliminating branches using a superoptimizer and the GNU C compiler
• Niklas Jonsson: Dynamic memory disambiguation for array references
• Joakim Ericson: Division by invariant integers using multiplication
• Essayas Gebrewahid: Exploiting mixed SIMD parallelism by reducing data reorganization overhead
• Adam Nilsson: Efficient Instruction Scheduling for a Pipelined Architecture
• Erik Hedblom: Using Node Merging to Enhance Graph Coloring
• Alexander Andersson: Global register allocation at link time
• Alexander Andersson: Global register allocation at link time
• Mattias Eklund: Register allocation using lazy saves, eager restores, and greedy shuffling