Interprocedural optimization of OpenMP kernels
When compiling OpenMP constructs, the bodies of OpenMP regions are outlined into separate functions, which are later called indirectly by libgomp built-ins. This outlining process disables interprocedural optimizations for the kernels. In this short talk, we outline a mechanism to partially restore interprocedural optimization capabilities for the kernels, starting with constant propagation, and we discuss its applications beyond OpenMP.
Josef Melcr
I’m a graduate student pursuing my Master’s with a passion for system programming, compilers, and open-source development. Though I’m new to contributing to GCC, I hope to make it a central part of my career one day.