[PATCH] D76384: Move FPFeatures from BinaryOperator bitfields to Trailing storage - preliminary

[Melanie Blower via Phabricator via cfe-commits]

mibintc added a comment.

There's a bug in this patch that i haven't solved yet. I think it's just 1 bug. The bug is that the AST statement visitor is going to the "fallback" instead of the right destination when walking CompoundAssignmentOperator.  This patch collapses CompoundAssignmentOperator class into BinaryOperator class.  In a spot check, i think all the failing tests are because the Visitor walk isn't working correctly for CompoundAssign (for example += ). These are the test fails reported by check-clang in my sandbox,
Failing Tests (15):

  Clang :: Analysis/loopexit-cfg-output.cpp
  Clang :: CXX/drs/dr2xx.cpp
  Clang :: CXX/expr/expr.const/p2-0x.cpp
  Clang :: CXX/expr/expr.prim/expr.prim.lambda/expr.prim.lambda.capture/p17.cpp
  Clang :: CodeGen/ffp-contract-fast-option.cpp
  Clang :: CodeGenCXX/const-init-cxx1y.cpp
  Clang :: CodeGenCXX/const-init-cxx2a.cpp
  Clang :: CodeGenCXX/non-const-init-cxx2a.cpp
  Clang :: Sema/warn-unreachable.c
  Clang :: Sema/warn-unsequenced.c
  Clang :: SemaCXX/constant-expression-cxx1y.cpp
  Clang :: SemaCXX/constant-expression-cxx2a.cpp
  Clang :: SemaCXX/decomposed-condition.cpp
  Clang :: SemaCXX/integer-overflow.cpp
  Clang :: SemaCXX/warn-unsequenced.cpp


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D76384/new/

https://reviews.llvm.org/D76384