Remove FutureMessage from sender TensorPipeAgent

[mrshenli]

Stack from ghstack:

• Apply clang-format to rpc cpp files #50236 Apply clang-format to rpc cpp files
• Completely remove FutureMessage type #50029 Completely remove FutureMessage type
• Completely Remove FutureMessage from RPC agents #50028 Completely Remove FutureMessage from RPC agents
• Completely Remove FutureMessage from RPC cpp tests #50027 Completely Remove FutureMessage from RPC cpp tests
• Remove FutureMessage from RPC request callback logic #50026 Remove FutureMessage from RPC request callback logic
• Completely remove FutureMessage from FaultyProcessGroupAgent #50025 Completely remove FutureMessage from FaultyProcessGroupAgent
• Remove FutureMessage from sender TensorPipeAgent #50024 Remove FutureMessage from sender TensorPipeAgent
• Remove FutureMessage from sender ProcessGroupAgent #50023 Remove FutureMessage from sender ProcessGroupAgent
• Completely remove FutureMessage from distributed autograd #50020 Completely remove FutureMessage from distributed autograd
• Completely remove FutureMessage from RPC TorchScript implementations #50005 Completely remove FutureMessage from RPC TorchScript implementations
• Completely remove FutureMessage from RRef Implementations #50004 Completely remove FutureMessage from RRef Implementations
• Replace FutureMessage with ivalue::Future in RpcAgent retry logic #49995 Replace FutureMessage with ivalue::Future in RpcAgent retry logic
• Replace FutureMessage with ivalue::Future in RRefContext #49960 Replace FutureMessage with ivalue::Future in RRefContext
• Replace FutureMessage with ivalue::Future in distributed/autograd/utils.* #49927 Replace FutureMessage with ivalue::Future in distributed/autograd/utils.*
• Let RpcAgent::send() return JitFuture #49906 Let RpcAgent::send() return JitFuture

Differential Revision: D25753386

[facebook-github-bot]

💊 CI failures summary and remediations

As of commit d485bb7 (more details on the Dr. CI page):

---

• 1/1 failures possibly* introduced in this PR
  • 1/1 non-CircleCI failure(s)

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions to the (internal) Dr. CI Users group.

This comment has been revised 15 times.

[lw]

This change (replacing a value with a shared_ptr to a value) is undoing a slight optimization I had added: instead of having two nested shared_ptrs (hence two allocations, and two separate control blocks, with two reference counts) the code was using a single allocation (for the AtomicFutureMessage) with a single control block, and was relying on the shared_ptr "aliasing" constructor to extract a shared_ptr to the FutureMessage field that however still used the former control block.

The merits of this optimization are arguable, and it also has downsides (it keeps the atomic_flag alive even once it's not needed anymore), thus I'm fine to remove it, I just wanted to make sure this was intentional.

[mrshenli]

This wasn't intentional. Let me add that to my todo list, and see if I can bring it back in a followup PR.

[codecov]

Codecov Report

Merging #50024 (d485bb7) into gh/mrshenli/276/base (392450d) will increase coverage by 10.21%.
The diff coverage is 100.00%.

@@                    Coverage Diff                    @@
##           gh/mrshenli/276/base   #50024       +/-   ##
=========================================================
+ Coverage                 70.46%   80.68%   +10.21%     
=========================================================
  Files                      1900     1900               
  Lines                    206321   206322        +1     
=========================================================
+ Hits                     145384   166463    +21079     
+ Misses                    60937    39859    -21078     

[facebook-github-bot]

@mrshenli merged this pull request in 0684d07.