diff options
| author | Atmn Patel <atmnp@nvidia.com> | 2025-10-23 15:21:13 -0500 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-10-23 13:21:13 -0700 |
| commit | 1c30038e5af5256aeda45946ddc0b5f801749e15 (patch) | |
| tree | e270163cb5a8415d649745884dd321b2b570d2fb /lldb/packages/Python/lldbsuite/test/gdbclientutils.py | |
| parent | d87200e8ca2a2c593baea2e113b90bf43409b534 (diff) | |
| download | llvm-1c30038e5af5256aeda45946ddc0b5f801749e15.zip llvm-1c30038e5af5256aeda45946ddc0b5f801749e15.tar.gz llvm-1c30038e5af5256aeda45946ddc0b5f801749e15.tar.bz2 | |
[flang][mlir] add missing type conversion when lowering atomiccas (#164865)
When lowering `atomiccas`, flang does not convert the output of the
`llvm.extract_value` op to result type expected in the expression being
lowered. This results in invalid MLIR being generated such as when the
output of the atomiccas is being used for an equality check in a `do
while` loop condition, where the `arith.cmpi` would be comparing an `i64
0` with an `i1`. This change ensures that the appropriate cast is
inserted.
Reviewers: @clementval @vzakhari
Diffstat (limited to 'lldb/packages/Python/lldbsuite/test/gdbclientutils.py')
0 files changed, 0 insertions, 0 deletions