[all-commits] [llvm/llvm-project] 35e6cd: [Clang][Sema][NFCI] Simplify `resolveAllocationOve...

[Igor Kudrin via All-commits]

  Branch: refs/heads/users/ikudrin/clang-findallocationfunction-simplify
  Home:   https://github.com/llvm/llvm-project
  Commit: 35e6cd3ee18756869cf75d01236588b444c135fa
      https://github.com/llvm/llvm-project/commit/35e6cd3ee18756869cf75d01236588b444c135fa
  Author: Igor Kudrin <ikudrin at accesssoftek.com>
  Date:   2026-06-13 (Sat, 13 Jun 2026)

  Changed paths:
    M clang/lib/Sema/SemaExprCXX.cpp

  Log Message:
  -----------
  [Clang][Sema][NFCI] Simplify `resolveAllocationOverload()`

`resolveAllocationOverload()` performs multiple rounds of overload
resolution (typed and untyped, aligned and unaligned), each requiring a
slightly different argument list. Previously, the argument vector was
mutated in-place, which made the flow hard to follow.

This refactor prepares the list of arguments before calling
`resolveAllocationOverload()`. The preferred argument list is passed in
`PrefArgs`, while the fallback arguments are passed in `FallbackArgs`.
If the fallback resolution is not required, `FallbackArgs` is empty.
When making a nested call to perform the resolution with the fallback
arguments, the current set of candidates is passed in `PrefCandidates`
(formerly, `AlignedCandidates`). This argument also serves as a flag
used to distinguish the top-level call from nested fallback calls.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications