https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68230
--- Comment #5 from Joshua T, Fisher <j.fisher at digipen dot edu> --- (In reply to Jonathan Wakely from comment #2) > It's rare that my reaction to a -Wunused-parameter warning is to change the > function to remove the parameter entirely. YMMV. Totally reasonable, I too rarely change the signature and use either attributes, removing the parameter name, or the void cast trick. However I have seen bugs pop up a couple times from important parameters being included, but not used, see: https://github.com/nvpro-pipeline/VkHLF/commit/b6646c4773e8aef49c40e8684eca1382bf2e9d50 and my blog where I first encountered this in a closed source codebase: http://www.nullterminatedstrings.com/c++/recursive-warning Perhaps this doesn't qualify as a bug in the current -Wunused-parameter warning implementation, but would it be reasonable to add it as a separate warning? Thanks for looking into this by the way folks!
