[PATCH] D28390: [DWARF] LICM should null out the debug loc of hoisted loop invariant instructions

[Wolfgang Pieb via Phabricator via llvm-commits]

wolfgangp added a comment.



> @danielcdh @wolfgangp -- Would switching to line 0 locations for hoisted instructions have an adverse effect on Sample PGO?

It does indeed seem preferrable, since the hoisted instruction is currently attributed to the same line as the previous instruction, which is not what we want. I think our usage and understanding of 0 linenumbers has evolved since this change was made and we should examine all the places where we're removing the line number and see if setting it to 0 instead is better.


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D28390/new/

https://reviews.llvm.org/D28390