https://llvm.org/bugs/show_bug.cgi?id=24185
Sanjoy Das <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |INVALID --- Comment #2 from Sanjoy Das <[email protected]> --- (In reply to comment #1) > I think that all `mayThrow` ensures is that the specific instruction doesn't > immediately unwind. That `invoke` transfers control to instructions which > may throw isn't conceptually different from a `br` which does the same, no? I was under the impression that an invoke does not always unwind to its landingpad, but may "skip" frames if the catch clause does not match the type of the thrown exception. Given that that isn't the case, and invokes always unwind to their landingpad if an exception is thrown, this isn't a bug. -- You are receiving this mail because: You are on the CC list for the bug.
_______________________________________________ LLVMbugs mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
