[PATCH] D136526: [AAPointerInfo] refactor how offsets and Access objects are tracked

[Sameer Sahasrabuddhe via Phabricator via llvm-commits]

sameerds added inline comments.


================
Comment at: llvm/test/Transforms/Attributor/value-simplify-pointer-info.ll:784
 ; TUNIT-NEXT:    [[I3:%.*]] = getelementptr inbounds [[STRUCT_S]], %struct.S* [[AGG_RESULT]], i64 0, i32 2
-; TUNIT-NEXT:    store i32 4, i32* [[I3]], align 4, !tbaa [[TBAA14]]
+; TUNIT-NEXT:    store i32 [[ADD2]], i32* [[I3]], align 4, !tbaa [[TBAA14]]
 ; TUNIT-NEXT:    ret void
----------------
jdoerfert wrote:
> Why do we miss out on these propagations? There are no PHIs involved, right? Something is amiss.
When I don't distinguish between calls that reach the same remote inst, that remote inst is conservatively treated as unknown. I thought it would be okay to "improve" this later, but then realized that the solution is very relevant to the refactoring. The new change does it right, and this side-effect disappears.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D136526/new/

https://reviews.llvm.org/D136526