fix a case of potentially use of undefined variable when handling error in distributed message processing (#51019)

There is a use of `oldstate` on line 244 which has the possibility of
being undefined. This bug seems to have been introduced in
40c622b#diff-39fa44ff86c5b38bd6b9e7f60733b25724a9efd9221ca38f776ed3f3ab01dec2.
I don't have a repro for this but this PR reverts back to the situation
where `oldstate` is defined at the topmost level of the catch block as
it was before the offending commit.

Author: KristofferC

stdlib/Distributed/src/process_messages.jl

@@ -210,6 +210,9 @@ function message_handler_loop(r_stream::IO, w_stream::IO, incoming::Bool)
             handle_msg(msg, header, r_stream, w_stream, version)
         end
     catch e
+        werr = worker_from_id(wpid)
+        oldstate = werr.state
+
         # Check again as it may have been set in a message handler but not propagated to the calling block above
         if wpid < 1
             wpid = worker_id_from_socket(r_stream)
@@ -219,8 +222,6 @@ function message_handler_loop(r_stream::IO, w_stream::IO, incoming::Bool)
             println(stderr, e, CapturedException(e, catch_backtrace()))
             println(stderr, "Process($(myid())) - Unknown remote, closing connection.")
         elseif !(wpid in map_del_wrkr)
-            werr = worker_from_id(wpid)
-            oldstate = werr.state
             set_worker_state(werr, W_TERMINATED)
 
             # If unhandleable error occurred talking to pid 1, exit