[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

[serge via Phabricator via llvm-commits]

serge-sans-paille marked 3 inline comments as done.
serge-sans-paille added inline comments.


================
Comment at: CMakeLists.txt:521
+endforeach()
+file(APPEND "${CMAKE_BINARY_DIR}/include/llvm/Support/Extension.def" "#undef HANDLE_EXTENSION\n")
 
----------------
mehdi_amini wrote:
> serge-sans-paille wrote:
> > mehdi_amini wrote:
> > > Is this file gonna be re-written for each re-execution of CMake?
> > > Will this trigger more recompilation than needed?
> > > If so what can be done about this? Can we generate a copy and replace the final one only if it differs in content? Can we make re-writing the file a rule based on actual dependencies (a target depending on the content of `LLVM_EXTENSION_LIBS` changing)?
> > Using as shA256 is not 100% bullet proof but should be ok, right?
> I'm not concerned about using a hash, but I'm wonder if there isn't a primitive "copy_if_changed" or similar in CMake that would just diff and copy?
> (your solution is fine but less boilerplate/complexity is always welcome)
There was indeed a way :-)


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61446/new/

https://reviews.llvm.org/D61446