sammccall added inline comments.
================
Comment at: clang/lib/AST/ComputeDependence.cpp:74
for (const auto *I : D->specific_attrs<AlignedAttr>()) {
+ // FIXME: should we propagate the error bit?
if (I->isAlignmentDependent())
----------------
maybe just do this unless it causes regressions?
================
Comment at: clang/lib/Sema/SemaDecl.cpp:2434
for (auto *I : Old->specific_attrs<AlignedAttr>()) {
+ // FIXME: we should bail out if I is error dependent.
// FIXME: We have no way of representing inherited dependent alignments
----------------
I actually don't think we should add this FIXME.
Currently error dependence implies dependence. If we break this invariant,
there are going to be hundreds of places that check for dependence and need to
be updated. We haven't annotated them all with fixmes.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D78085/new/
https://reviews.llvm.org/D78085
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
