[lldb] [llvm] [llvm][lldb] Handle oversized LEB128 operands in DWARF expressions (PR #219149)

[via llvm-commits]

MrEven132 wrote:

Updated the PR to address the review feedback:

- Centralized operand decoding failures through `FailOperandDecode`.
- Replaced the signedness boolean at the call sites with explicitly named signed and unsigned LEB128 helpers.
- Added `isError()` checks in `DWARFUnit::updateVariableDieMap` before reading both the initial address operation and the optional `DW_OP_plus_uconst`.
- Added regression tests for operand errors in both positions.


https://github.com/llvm/llvm-project/pull/219149