compnerd marked 2 inline comments as done.
================
Comment at: source/Interpreter/CommandObject.cpp:339
@@ -338,3 +338,3 @@
m_exe_ctx.Clear();
- m_api_locker.Unlock();
+ m_api_locker.unlock();
}
----------------
zturner wrote:
> ```
> if (m_api_locker)
> m_api_locker.unlock();
> ```
>
> Otherwise this is undefined behavior. Also make sure you do this check
> anywhere else in the patch where you do manual unlocking / locking like this.
Yeah, this was the only place where the lock was unlocked and there wasn't a
comment indicating that the assumption was that the lock will be held. A few
places where there were unlocks were early releases of the unique_lock. A
couple of places indicated that the lock must have been acquired prior to the
destructor running.
Repository:
rL LLVM
http://reviews.llvm.org/D20351
_______________________________________________
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
