#2019: add move constructor/assign to TraceScopedEvent and TraceScopedNote

[nmm0]

Fixes #2019

[github-actions]

Pipelines results

PR tests (gcc-5, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-6, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-3.9, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-9, ubuntu, mpich, zoltan)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-10, ubuntu, openmpi, no LB)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-5.0, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-9, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-7, ubuntu, mpich, trace runtime, LB)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-11, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-12, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-13, alpine, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-13, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-14, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (intel icpx, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (nvidia cuda 10.1, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-11, ubuntu, mpich, json schema test)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-12, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (clang-10, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (nvidia cuda 11.0, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (intel icpc, ubuntu, mpich)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

PR tests (gcc-8, ubuntu, mpich, address sanitizer)

Build for 22d105f (2022-11-21 18:36:01 UTC)

Compilation - successful

Testing - passed

Build log

---

[codecov]

Codecov Report

Merging #2020 (22d105f) into develop (1259a37) will decrease coverage by 0.01%.
The diff coverage is n/a.

@@             Coverage Diff             @@
##           develop    #2020      +/-   ##
===========================================
- Coverage    84.50%   84.49%   -0.02%     
===========================================
  Files          731      731              
  Lines        25860    25860              
===========================================
- Hits         21854    21850       -4     
- Misses        4006     4010       +4     

Impacted Files	Coverage Δ
src/vt/trace/trace_user.h	71.42% <ø> (ø)
src/vt/phase/phase_manager.cc	89.86% <0.00%> (-2.71%)	⬇️

[PhilMiller]

I know calling swap is idiomatic, but I'd be more comfortable if the moved from objects had their event fields assigned the no event value explicitly, moreso in the move assignment case.

[nmm0]

I know calling swap is idiomatic, but I'd be more comfortable if the moved from objects had their event fields assigned the no event value explicitly, moreso in the move assignment case.

For the move constructor it is set explicitly then swapped (from the initializer on event_), but for move assignment it has to use a swap to be correct if not the old value (if it is not no_trace_event) will never be released

[PhilMiller]

I feel like if that case ever arises, it represents a logic bug, which we should maybe assert against.

What's the actual use case for allowing any sort of move construction or assignment of these types? I figured it was for something like moving into a closure, where there's no existing event to be dealt with.

[nmm0]

I feel like if that case ever arises, it represents a logic bug, which we should maybe assert against.

What's the actual use case for allowing any sort of move construction or assignment of these types? I figured it was for something like moving into a closure, where there's no existing event to be dealt with.

One I can think of off the top of my head would be an object that needs to extend the lifetime of the TraceScopedEvent/Note until it is passed a new event. When passed a new event (lets say at the end of an iteration or something), it would need to move the new event into its storage, releasing the previous scope.

[PhilMiller]

Ok, I'll buy it.

[cz4rs]

Looks good to me.