[scheduler][restart] Reduce live-locking by using data produced on failure

Summary: When a worker fails because it can't a get the lock of a `Procname` it will include it in the exception that it throws so the `RestartScheduler` can record it as a dependency. Then when scheduling a new work item from `RestartScheduler.next` it will check if this dependency is already met, if it isn't it will not schedule the `Procname` yet.

Reviewed By: ngorogiannis

Differential Revision: D19820331

fbshipit-source-id: b48cacc9a

infer/src/absint/AbstractInterpreter.ml

@@ -163,7 +163,7 @@ module AbstractInterpreterCommon (TransferFunctions : TransferFunctions.SIL) = s
             Ok post
           with exn ->
             IExn.reraise_if exn ~f:(fun () ->
-                match exn with ProcessPool.ProcnameAlreadyLocked -> true | _ -> false ) ;
+                match exn with RestartScheduler.ProcnameAlreadyLocked _ -> true | _ -> false ) ;
             (* delay reraising to get a chance to write the debug HTML *)
             let backtrace = Caml.Printexc.get_raw_backtrace () in
             Error (exn, backtrace, instr)

infer/src/backend/FileScheduler.mli

@@ -6,4 +6,4 @@
  *)
 open! IStd
 
-val make : SourceFile.t list -> SchedulerTypes.target ProcessPool.TaskGenerator.t
+val make : SourceFile.t list -> (SchedulerTypes.target, Procname.t) ProcessPool.TaskGenerator.t

infer/src/backend/InferAnalyze.ml

@@ -20,15 +20,18 @@ let clear_caches () =
   BufferOverrunUtils.clear_cache ()
 
 
-let analyze_target : SchedulerTypes.target Tasks.doer =
+let analyze_target : (SchedulerTypes.target, Procname.t) Tasks.doer =
   let analyze_source_file exe_env source_file =
     if Topl.is_active () then DB.Results_dir.init (Topl.sourcefile ()) ;
     DB.Results_dir.init source_file ;
     L.task_progress SourceFile.pp source_file ~f:(fun () ->
-        Ondemand.analyze_file exe_env source_file ;
+        try
-        if Topl.is_active () && Config.debug_mode then
+          Ondemand.analyze_file exe_env source_file ;
-          DotCfg.emit_frontend_cfg (Topl.sourcefile ()) (Topl.cfg ()) ;
+          if Topl.is_active () && Config.debug_mode then
-        if Config.write_html then Printer.write_all_html_files source_file )
+            DotCfg.emit_frontend_cfg (Topl.sourcefile ()) (Topl.cfg ()) ;
+          if Config.write_html then Printer.write_all_html_files source_file ;
+          None
+        with RestartScheduler.ProcnameAlreadyLocked pname -> Some pname )
   in
   (* In call-graph scheduling, log progress every [per_procedure_logging_granularity] procedures.
      The default roughly reflects the average number of procedures in a C++ file. *)
@@ -41,7 +44,10 @@ let analyze_target : SchedulerTypes.target Tasks.doer =
       L.log_task "Analysing block of %d procs, starting with %a@." per_procedure_logging_granularity
         Procname.pp proc_name ;
       procs_left := per_procedure_logging_granularity ) ;
-    Ondemand.analyze_proc_name_toplevel exe_env proc_name
+    try
+      Ondemand.analyze_proc_name_toplevel exe_env proc_name ;
+      None
+    with RestartScheduler.ProcnameAlreadyLocked pname -> Some pname
   in
   fun target ->
     let exe_env = Exe_env.mk () in

infer/src/backend/ProcLocker.ml

@@ -6,6 +6,7 @@
  *)
 
 open! IStd
+module L = Logging
 
 exception UnlockNotLocked of Procname.t
 
@@ -45,3 +46,10 @@ let try_lock pname =
         Unix.symlink ~target:locks_target ~link_name:(filename_from pname) ;
         true
       with Unix.Unix_error (Unix.EEXIST, _, _) -> false )
+
+
+let is_locked pname =
+  try
+    ignore (Unix.lstat (filename_from pname)) ;
+    true
+  with Unix.Unix_error (Unix.ENOENT, _, _) -> false

infer/src/backend/ProcLocker.mli

@@ -22,3 +22,5 @@ val unlock : Procname.t -> unit
 val clean : unit -> unit
 (** This should be called when locks will no longer be used to remove any files or state that's not
     necessary. *)
+
+val is_locked : Procname.t -> bool

infer/src/backend/RestartScheduler.ml

@@ -7,13 +7,30 @@
 open! IStd
 module L = Logging
 
-let of_list (lst : 'a list) : 'a ProcessPool.TaskGenerator.t =
+exception ProcnameAlreadyLocked of Procname.t
+
+type work_with_dependency = {work: SchedulerTypes.target; need_to_finish: Procname.t option}
+
+let of_list (lst : work_with_dependency list) : ('a, Procname.t) ProcessPool.TaskGenerator.t =
   let content = Queue.of_list lst in
   let remaining = ref (Queue.length content) in
   let remaining_tasks () = !remaining in
   let is_empty () = Int.equal !remaining 0 in
-  let finished ~completed work = if completed then decr remaining else Queue.enqueue content work in
+  let finished ~result work =
-  let next () = Queue.dequeue content in
+    match result with
+    | None ->
+        decr remaining
+    | Some _ as need_to_finish ->
+        Queue.enqueue content {work; need_to_finish}
+  in
+  let work_if_dependency_allows w =
+    match w.need_to_finish with
+    | Some pname when ProcLocker.is_locked pname ->
+        Queue.enqueue content w ; None
+    | None | Some _ ->
+        Some w.work
+  in
+  let next () = Option.bind (Queue.dequeue content) ~f:(fun w -> work_if_dependency_allows w) in
   {remaining_tasks; is_empty; finished; next}
 
 
@@ -21,9 +38,12 @@ let make sources =
   let pnames =
     List.map sources ~f:SourceFiles.proc_names_of_source
     |> List.concat
-    |> List.rev_map ~f:(fun procname -> SchedulerTypes.Procname procname)
+    |> List.rev_map ~f:(fun procname ->
+           {work= SchedulerTypes.Procname procname; need_to_finish= None} )
+  in
+  let files =
+    List.map sources ~f:(fun file -> {work= SchedulerTypes.File file; need_to_finish= None})
   in
-  let files = List.map sources ~f:(fun file -> SchedulerTypes.File file) in
   let permute = List.permute ~random_state:(Random.State.make (Array.create ~len:1 0)) in
   permute pnames @ permute files |> of_list
 
@@ -44,7 +64,7 @@ let lock_exn pname =
       if ProcLocker.try_lock pname then record_locked_proc pname
       else (
         unlock_all () ;
-        raise ProcessPool.ProcnameAlreadyLocked ) )
+        raise (ProcnameAlreadyLocked pname) ) )
 
 
 let unlock pname =

infer/src/backend/RestartScheduler.mli

@@ -6,6 +6,8 @@
  *)
 open! IStd
 
+exception ProcnameAlreadyLocked of Procname.t
+
 val setup : unit -> unit
 
 val clean : unit -> unit
@@ -14,4 +16,4 @@ val lock_exn : Procname.t -> unit
 
 val unlock : Procname.t -> unit
 
-val make : SourceFile.t list -> SchedulerTypes.target ProcessPool.TaskGenerator.t
+val make : SourceFile.t list -> (SchedulerTypes.target, Procname.t) ProcessPool.TaskGenerator.t

infer/src/backend/SyntacticCallGraph.ml

@@ -46,7 +46,7 @@ let build_from_sources sources =
   g
 
 
-let bottom_up sources : SchedulerTypes.target ProcessPool.TaskGenerator.t =
+let bottom_up sources : (SchedulerTypes.target, Procname.t) ProcessPool.TaskGenerator.t =
   let open SchedulerTypes in
   let syntactic_call_graph = build_from_sources sources in
   let remaining = ref (CallGraph.n_procs syntactic_call_graph) in
@@ -83,7 +83,7 @@ let bottom_up sources : SchedulerTypes.target ProcessPool.TaskGenerator.t =
         CallGraph.flag syntactic_call_graph n.pname ;
         Some (Procname n.pname)
   in
-  let finished ~completed:_ = function
+  let finished ~result:_ = function
     | Procname pname ->
         decr remaining ;
         decr scheduled ;

infer/src/backend/SyntacticCallGraph.mli

@@ -6,7 +6,7 @@
  *)
 open! IStd
 
-val make : SourceFile.t list -> SchedulerTypes.target ProcessPool.TaskGenerator.t
+val make : SourceFile.t list -> (SchedulerTypes.target, Procname.t) ProcessPool.TaskGenerator.t
 (** task generator that works by
 
     - loading the syntactic call graph from the capture DB

infer/src/backend/Tasks.ml

@@ -7,7 +7,7 @@
 
 open! IStd
 
-type 'a doer = 'a -> unit
+type ('a, 'b) doer = 'a -> 'b option
 
 let fork_protect ~f x = BackendStats.reset () ; ForkUtils.protect ~f x
 
@@ -17,7 +17,7 @@ let with_new_db_connection ~f () =
 
 
 module Runner = struct
-  type ('work, 'final) t = ('work, 'final) ProcessPool.t
+  type ('work, 'final, 'result) t = ('work, 'final, 'result) ProcessPool.t
 
   let create ~jobs ~f ~child_epilogue ~tasks =
     PerfEvent.(
@@ -41,7 +41,7 @@ module Runner = struct
     ProcessPool.run runner
 end
 
-let run_sequentially ~(f : 'a doer) (tasks : 'a list) : unit =
+let run_sequentially ~(f : ('a, 'b) doer) (tasks : 'a list) : unit =
   let task_generator = ProcessPool.TaskGenerator.of_list tasks in
   let task_bar = TaskBar.create ~jobs:1 in
   (ProcessPoolState.update_status :=
@@ -53,8 +53,8 @@ let run_sequentially ~(f : 'a doer) (tasks : 'a list) : unit =
   let rec run_tasks () =
     if not (task_generator.is_empty ()) then (
       Option.iter (task_generator.next ()) ~f:(fun t ->
-          f t ;
+          let result = f t in
-          task_generator.finished ~completed:true t ) ;
+          task_generator.finished ~result t ) ;
       TaskBar.set_remaining_tasks task_bar (task_generator.remaining_tasks ()) ;
       TaskBar.refresh task_bar ;
       run_tasks () )

infer/src/backend/Tasks.mli

@@ -7,9 +7,9 @@
 
 open! IStd
 
-type 'a doer = 'a -> unit
+type ('a, 'b) doer = 'a -> 'b option
 
-val run_sequentially : f:'a doer -> 'a list -> unit
+val run_sequentially : f:('a, 'b) doer -> 'a list -> unit
 (** Run the tasks sequentially *)
 
 val fork_protect : f:('a -> 'b) -> 'a -> 'b
@@ -17,16 +17,16 @@ val fork_protect : f:('a -> 'b) -> 'a -> 'b
 
 (** A runner accepts new tasks repeatedly for parallel execution *)
 module Runner : sig
-  type ('work, 'final) t
+  type ('work, 'final, 'result) t
 
   val create :
        jobs:int
-    -> f:'work doer
+    -> f:('work, 'result) doer
     -> child_epilogue:(unit -> 'final)
-    -> tasks:(unit -> 'work ProcessPool.TaskGenerator.t)
+    -> tasks:(unit -> ('work, 'result) ProcessPool.TaskGenerator.t)
-    -> ('work, 'final) t
+    -> ('work, 'final, 'result) t
   (** Create a runner running [jobs] jobs in parallel *)
 
-  val run : (_, 'final) t -> 'final option Array.t
+  val run : (_, 'final, _) t -> 'final option Array.t
   (** Start the given tasks with the runner and wait until completion *)
 end

infer/src/backend/ondemand.ml

@@ -219,7 +219,7 @@ let run_proc_analysis ~caller_pdesc callee_pdesc =
     let backtrace = Printexc.get_backtrace () in
     IExn.reraise_if exn ~f:(fun () ->
         match exn with
-        | ProcessPool.ProcnameAlreadyLocked ->
+        | RestartScheduler.ProcnameAlreadyLocked _ ->
             clear_actives () ; true
         | _ ->
             if not !logged_error then (

infer/src/base/Logging.ml

@@ -213,7 +213,11 @@ let log_task fmt =
   log ~to_console progress_file_fmts fmt
 
 
-let task_progress ~f pp x = log_task "%a starting@." pp x ; f () ; log_task "%a DONE@." pp x
+let task_progress ~f pp x =
+  log_task "%a starting@." pp x ;
+  let result = f () in
+  log_task "%a DONE@." pp x ; result
+
 
 let user_warning fmt = log ~to_console:(not Config.quiet) user_warning_file_fmts fmt
 

infer/src/base/Logging.mli

@@ -26,7 +26,7 @@ val progress : ('a, F.formatter, unit) format -> 'a
 val log_task : ('a, F.formatter, unit) format -> 'a
 (** log progress in the log file and on the console unless there is an active task bar *)
 
-val task_progress : f:(unit -> unit) -> (F.formatter -> 'a -> unit) -> 'a -> unit
+val task_progress : f:(unit -> 'a) -> (F.formatter -> 'b -> unit) -> 'b -> 'a
 (** [task_progress ~f pp x] executes [f] and log progress [pp x] in the log file and also on the
     console unless there is an active task bar *)
 

infer/src/base/ProcessPool.ml

@@ -9,16 +9,14 @@ open! IStd
 module F = Format
 module L = Logging
 
-exception ProcnameAlreadyLocked
-
 module TaskGenerator = struct
-  type 'a t =
+  type ('a, 'b) t =
     { remaining_tasks: unit -> int
     ; is_empty: unit -> bool
-    ; finished: completed:bool -> 'a -> unit
+    ; finished: result:'b option -> 'a -> unit
     ; next: unit -> 'a option }
 
-  let chain (gen1 : 'a t) (gen2 : 'a t) : 'a t =
+  let chain (gen1 : ('a, 'b) t) (gen2 : ('a, 'b) t) : ('a, 'b) t =
     let remaining_tasks () = gen1.remaining_tasks () + gen2.remaining_tasks () in
     let gen1_returned_empty = ref false in
     let gen1_is_empty () =
@@ -26,20 +24,19 @@ module TaskGenerator = struct
       !gen1_returned_empty
     in
     let is_empty () = gen1_is_empty () && gen2.is_empty () in
-    let finished ~completed work_item =
+    let finished ~result work_item =
-      if gen1_is_empty () then gen2.finished ~completed work_item
+      if gen1_is_empty () then gen2.finished ~result work_item else gen1.finished ~result work_item
-      else gen1.finished ~completed work_item
     in
     let next x = if gen1_is_empty () then gen2.next x else gen1.next x in
     {remaining_tasks; is_empty; finished; next}
 
 
-  let of_list (lst : 'a list) : 'a t =
+  let of_list (lst : 'a list) : ('a, _) t =
     let content = ref lst in
     let length = ref (List.length lst) in
     let remaining_tasks () = !length in
     let is_empty () = List.is_empty !content in
-    let finished ~completed:_ _work_item = decr length in
+    let finished ~result:_ _work_item = decr length in
     let next () =
       match !content with
       | [] ->
@@ -65,7 +62,7 @@ type child_info = {pid: Pid.t; down_pipe: Out_channel.t}
 type 'a child_state = Initializing | Idle | Processing of 'a
 
 (** the state of the pool *)
-type ('work, 'final) t =
+type ('work, 'final, 'result) t =
   { jobs: int
         (** number of jobs running in parallel, i.e. number of children we are responsible for *)
   ; slots: child_info Array.t
@@ -75,7 +72,7 @@ type ('work, 'final) t =
   ; children_updates: Unix.File_descr.t
         (** all the children send updates up the same pipe to the pool *)
   ; task_bar: TaskBar.t
-  ; tasks: 'work TaskGenerator.t  (** generator for work remaining to be done *)
+  ; tasks: ('work, 'result) TaskGenerator.t  (** generator for work remaining to be done *)
   ; file_lock: Utils.file_lock  (** file lock for sending worker messages *) }
 
 (** {2 Constants} *)
@@ -99,11 +96,11 @@ let buffer_size = 65_535
     the pipe will crash in the parent process. This is a limitation of the way we read from the pipe
     for now. To lift it, it should be possible to extend the buffer to the required length if we
     notice that we are trying to read more than [buffer_size] for example. *)
-type worker_message =
+type 'result worker_message =
   | UpdateStatus of int * Mtime.t * string
       (** [(i, t, status)]: starting a task from slot [i], at start time [t], with description
           [status]. Watch out that [status] must not be too close in length to [buffer_size]. *)
-  | Ready of {worker: int; completed: bool}
+  | Ready of {worker: int; result: 'result}
       (** Sent after finishing initializing or after finishing a given task. When received by
           master, this moves the worker state from [Initializing] or [Processing _] to [Idle]. *)
   | Crash of int  (** there was an error and the child is no longer receiving messages *)
@@ -265,12 +262,12 @@ let process_updates pool buffer =
            (* clean crash, give the child process a chance to cleanup *)
            Unix.wait (`Pid pid) |> ignore ;
            killall pool ~slot "see backtrace above"
-       | Ready {worker= slot; completed} ->
+       | Ready {worker= slot; result} ->
            ( match pool.children_states.(slot) with
            | Initializing ->
                ()
            | Processing work ->
-               pool.tasks.finished ~completed work
+               pool.tasks.finished ~result work
            | Idle ->
                L.die InternalError "Received a Ready message from an idle worker@." ) ;
            TaskBar.set_remaining_tasks pool.task_bar (pool.tasks.remaining_tasks ()) ;
@@ -284,7 +281,7 @@ let process_updates pool buffer =
 
 type 'a final_worker_message = Finished of int * 'a option | FinalCrash of int
 
-let collect_results (pool : (_, 'final) t) =
+let collect_results (pool : (_, 'final, _) t) =
   let failed = ref false in
   let updates_in = Unix.in_channel_of_descr pool.children_updates in
   (* use [Array.init] just to collect n messages, the order in the array will not be the same as the
@@ -332,9 +329,8 @@ let wait_all pool =
 
 
 (** worker loop: wait for tasks and run [f] on them until we are told to go home *)
-let rec child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilogue ~prev_completed
+let rec child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilogue ~prev_result =
-    =
+  send_to_parent (Ready {worker= slot; result= prev_result}) ;
-  send_to_parent (Ready {worker= slot; completed= prev_completed}) ;
   match receive_from_parent () with
   | GoHome -> (
     match epilogue () with
@@ -352,23 +348,21 @@ let rec child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilo
               true ) ) )
   | Do stuff ->
       let result =
-        try f stuff ; true with
+        try f stuff
-        | ProcnameAlreadyLocked ->
+        with e ->
-            false
+          IExn.reraise_if e ~f:(fun () ->
-        | e ->
+              if Config.keep_going then (
-            IExn.reraise_if e ~f:(fun () ->
+                L.internal_error "Error in subprocess %d: %a@." slot Exn.pp e ;
-                if Config.keep_going then (
+                (* do not raise and continue accepting jobs *)
-                  L.internal_error "Error in subprocess %d: %a@." slot Exn.pp e ;
+                false )
-                  (* do not raise and continue accepting jobs *)
+              else (
-                  false )
+                (* crash hard, but first let the master know that we have crashed *)
-                else (
+                send_to_parent (Crash slot) ;
-                  (* crash hard, but first let the master know that we have crashed *)
+                true ) ) ;
-                  send_to_parent (Crash slot) ;
+          None
-                  true ) ) ;
-            true
       in
       child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilogue
-        ~prev_completed:result
+        ~prev_result:result
 
 
 (** Fork a new child and start it so that it is ready for work.
@@ -387,7 +381,7 @@ let fork_child ~file_lock ~child_prelude ~slot (updates_r, updates_w) ~f ~epilog
       ProcessPoolState.reset_pid () ;
       child_prelude () ;
       let updates_oc = Unix.out_channel_of_descr updates_w in
-      let send_to_parent (message : worker_message) =
+      let send_to_parent (message : 'b worker_message) =
         marshal_to_pipe ~file_lock updates_oc message
       in
       let send_final (final_message : 'a final_worker_message) =
@@ -414,8 +408,7 @@ let fork_child ~file_lock ~child_prelude ~slot (updates_r, updates_w) ~f ~epilog
         PerfEvent.(log (fun logger -> log_end_event logger ())) ;
         x
       in
-      child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilogue
+      child_loop ~slot send_to_parent send_final receive_from_parent ~f ~epilogue ~prev_result:None ;
-        ~prev_completed:true ;
       Out_channel.close updates_oc ;
       In_channel.close orders_ic ;
       Epilogues.run () ;
@@ -428,10 +421,10 @@ let fork_child ~file_lock ~child_prelude ~slot (updates_r, updates_w) ~f ~epilog
 let create :
        jobs:int
     -> child_prelude:(unit -> unit)
-    -> f:('work -> unit)
+    -> f:('work -> 'result option)
     -> child_epilogue:(unit -> 'final)
-    -> tasks:(unit -> 'work TaskGenerator.t)
+    -> tasks:(unit -> ('work, 'result) TaskGenerator.t)
-    -> ('work, 'final) t =
+    -> ('work, 'final, 'result) t =
  fun ~jobs ~child_prelude ~f ~child_epilogue ~tasks ->
   let file_lock = Utils.create_file_lock () in
   let task_bar = TaskBar.create ~jobs in

infer/src/base/ProcessPool.mli

@@ -7,29 +7,29 @@
 
 open! IStd
 
-exception ProcnameAlreadyLocked
-
 module TaskGenerator : sig
   (** abstraction for generating jobs *)
-  type 'a t =
+  type ('a, 'b) t =
     { remaining_tasks: unit -> int
           (** number of tasks remaining to complete -- only used for reporting, so imprecision is
               not a bug *)
     ; is_empty: unit -> bool
           (** when should the main loop of the task manager stop expecting new tasks *)
-    ; finished: completed:bool -> 'a -> unit
+    ; finished: result:'b option -> 'a -> unit
-          (** Process pool calls [finished x] when a worker finishes item [x]. This is only called
+          (** Process pool calls [finished result:r x] when a worker finishes item [x]. result is
-              if [next ()] has previously returned [Some x] and [x] was sent to a worker. *)
+              None when the item was completed successfully and Some pname when it failed because it
+              could not lock pname. This is only called if [next ()] has previously returned
+              [Some x] and [x] was sent to a worker. *)
     ; next: unit -> 'a option
           (** [next ()] generates the next work item. If [is_empty ()] is true then [next ()] must
               return [None]. However, it is OK to for [next ()] to return [None] when [is_empty] is
               false. This corresponds to the case where there is more work to be done, but it is not
               schedulable until some already scheduled work is finished. *) }
 
-  val chain : 'a t -> 'a t -> 'a t
+  val chain : ('a, 'b) t -> ('a, 'b) t -> ('a, 'b) t
   (** chain two generators in order *)
 
-  val of_list : 'a list -> 'a t
+  val of_list : 'a list -> ('a, 'b) t
   (** schedule tasks out of a concrete list *)
 end
 
@@ -51,17 +51,17 @@ end
 
 (** A [('work, 'final) t] process pool accepts tasks of type ['work] and produces an array of
     results of type ['final]. ['work] and ['final] will be marshalled over a Unix pipe.*)
-type (_, _) t
+type (_, _, _) t
 
 val create :
      jobs:int
   -> child_prelude:(unit -> unit)
-  -> f:('work -> unit)
+  -> f:('work -> 'result option)
   -> child_epilogue:(unit -> 'final)
-  -> tasks:(unit -> 'work TaskGenerator.t)
+  -> tasks:(unit -> ('work, 'result) TaskGenerator.t)
-  -> ('work, 'final) t
+  -> ('work, 'final, 'result) t
 (** Create a new pool of processes running [jobs] jobs in parallel *)
 
-val run : (_, 'final) t -> 'final option Array.t
+val run : (_, 'final, 'result) t -> 'final option Array.t
 (** use the processes in the given process pool to run all the given tasks in parallel and return
     the results of the epilogues *)

infer/src/integration/CaptureCompilationDatabase.ml

@@ -26,13 +26,14 @@ let create_cmd (source_file, (compilation_data : CompilationDatabase.compilation
 
 let invoke_cmd (source_file, (cmd : CompilationDatabase.compilation_data)) =
   let argv = cmd.executable :: cmd.escaped_arguments in
-  ( match Spawn.spawn ~cwd:(Path cmd.directory) ~prog:cmd.executable ~argv () with
+  ( ( match Spawn.spawn ~cwd:(Path cmd.directory) ~prog:cmd.executable ~argv () with
-  | pid ->
+    | pid ->
-      !ProcessPoolState.update_status (Mtime_clock.now ()) (SourceFile.to_string source_file) ;
+        !ProcessPoolState.update_status (Mtime_clock.now ()) (SourceFile.to_string source_file) ;
-      Unix.waitpid (Pid.of_int pid)
+        Unix.waitpid (Pid.of_int pid)
-      |> Result.map_error ~f:(fun unix_error -> Unix.Exit_or_signal.to_string_hum (Error unix_error))
+        |> Result.map_error ~f:(fun unix_error ->
-  | exception Unix.Unix_error (err, f, arg) ->
+               Unix.Exit_or_signal.to_string_hum (Error unix_error) )
-      Error (F.asprintf "%s(%s): %s@." f arg (Unix.Error.message err)) )
+    | exception Unix.Unix_error (err, f, arg) ->
+        Error (F.asprintf "%s(%s): %s@." f arg (Unix.Error.message err)) )
   |> function
   | Ok () ->
       ()
@@ -42,7 +43,8 @@ let invoke_cmd (source_file, (cmd : CompilationDatabase.compilation_data)) =
         else L.die ExternalError fmt
       in
       log_or_die "Error running compilation for '%a': %a:@\n%s@." SourceFile.pp source_file
-        Pp.cli_args argv error
+        Pp.cli_args argv error ) ;
+  None
 
 
 let run_compilation_database compilation_database should_capture_file =