Don't use new-llvm-pass-manager=no in test
This commit is contained in:
parent
e2661bac6d
commit
57717eb8ad
@ -1,10 +1,12 @@
|
||||
// compile-flags: -Z panic-in-drop=abort -O -Z new-llvm-pass-manager=no
|
||||
// compile-flags: -Z panic-in-drop=abort -O
|
||||
// ignore-msvc
|
||||
|
||||
// Ensure that unwinding code paths are eliminated from the output after
|
||||
// optimization.
|
||||
|
||||
// This test uses -Z new-llvm-pass-manager=no, because the expected optimization does not happen
|
||||
// on targets using SEH exceptions (i.e. MSVC) anymore. The core issue is that Rust promises that
|
||||
// This test uses ignore-msvc, because the expected optimization does not happen on targets using
|
||||
// SEH exceptions with the new LLVM pass manager anymore, see
|
||||
// https://github.com/llvm/llvm-project/issues/51311. The core issue is that Rust promises that
|
||||
// the drop_in_place() function can't unwind, but implements it in a way that *can*, because we
|
||||
// currently go out of our way to allow longjmps, which also use the unwinding mechanism on MSVC
|
||||
// targets. We should either forbid longjmps, or not assume nounwind, making this optimization
|
||||
|
Loading…
x
Reference in New Issue
Block a user