[v5,4/5] gdb/infrun: extract out a code piece into 'mark_non_executing_threads' function

Message ID c13b23eec880df76c05e25a0f096bc5d44479a56.1586187408.git.tankut.baris.aktemur@intel.com
State Superseded
Headers show
Series
  • Handling already-exited threads in 'stop_all_threads'
Related show

Commit Message

Pedro Alves via Gdb-patches April 6, 2020, 3:45 p.m.
This is a refactoring.  The extracted function is placed deliberately
before 'stop_all_threads' because the function will be re-used there
in a subsequent patch for handling an exit status kind received from
a thread that GDB attempted to stop.

gdb/ChangeLog:
2019-11-04  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

	* infrun.c (handle_inferior_event): Extract out a piece of code
	into...
	(mark_non_executing_threads): ...this new function.

Change-Id: I2b088f4a724f4260cb37068264964525cf62a118
---
 gdb/infrun.c | 77 ++++++++++++++++++++++++++++------------------------
 1 file changed, 42 insertions(+), 35 deletions(-)

-- 
2.17.1

Comments

Pedro Alves via Gdb-patches April 16, 2020, 4:24 p.m. | #1
On 4/6/20 4:45 PM, Tankut Baris Aktemur via Gdb-patches wrote:
> This is a refactoring.  The extracted function is placed deliberately

> before 'stop_all_threads' because the function will be re-used there

> in a subsequent patch for handling an exit status kind received from

> a thread that GDB attempted to stop.

> 

> gdb/ChangeLog:

> 2019-11-04  Tankut Baris Aktemur  <tankut.baris.aktemur@intel.com>

> 

> 	* infrun.c (handle_inferior_event): Extract out a piece of code

> 	into...

> 	(mark_non_executing_threads): ...this new function.


OK.

Thanks,
Pedro Alves

Patch

diff --git a/gdb/infrun.c b/gdb/infrun.c
index f5d3b79e0d0..ce8b544ab8d 100644
--- a/gdb/infrun.c
+++ b/gdb/infrun.c
@@ -4704,6 +4704,47 @@  save_waitstatus (struct thread_info *tp, const target_waitstatus *ws)
     }
 }
 
+/* Mark the non-executing threads accordingly.  In all-stop, all
+   threads of all processes are stopped when we get any event
+   reported.  In non-stop mode, only the event thread stops.  */
+
+static void
+mark_non_executing_threads (process_stratum_target *target,
+			    ptid_t event_ptid,
+			    struct target_waitstatus ws)
+{
+  ptid_t mark_ptid;
+
+  if (!target_is_non_stop_p ())
+    mark_ptid = minus_one_ptid;
+  else if (ws.kind == TARGET_WAITKIND_SIGNALLED
+	   || ws.kind == TARGET_WAITKIND_EXITED)
+    {
+      /* If we're handling a process exit in non-stop mode, even
+	 though threads haven't been deleted yet, one would think
+	 that there is nothing to do, as threads of the dead process
+	 will be soon deleted, and threads of any other process were
+	 left running.  However, on some targets, threads survive a
+	 process exit event.  E.g., for the "checkpoint" command,
+	 when the current checkpoint/fork exits, linux-fork.c
+	 automatically switches to another fork from within
+	 target_mourn_inferior, by associating the same
+	 inferior/thread to another fork.  We haven't mourned yet at
+	 this point, but we must mark any threads left in the
+	 process as not-executing so that finish_thread_state marks
+	 them stopped (in the user's perspective) if/when we present
+	 the stop to the user.  */
+      mark_ptid = ptid_t (event_ptid.pid ());
+    }
+  else
+    mark_ptid = event_ptid;
+
+  set_executing (target, mark_ptid, false);
+
+  /* Likewise the resumed flag.  */
+  set_resumed (target, mark_ptid, false);
+}
+
 /* See infrun.h.  */
 
 void
@@ -5137,41 +5178,7 @@  handle_inferior_event (struct execution_control_state *ecs)
 	}
     }
 
-  /* Mark the non-executing threads accordingly.  In all-stop, all
-     threads of all processes are stopped when we get any event
-     reported.  In non-stop mode, only the event thread stops.  */
-  {
-    ptid_t mark_ptid;
-
-    if (!target_is_non_stop_p ())
-      mark_ptid = minus_one_ptid;
-    else if (ecs->ws.kind == TARGET_WAITKIND_SIGNALLED
-	     || ecs->ws.kind == TARGET_WAITKIND_EXITED)
-      {
-	/* If we're handling a process exit in non-stop mode, even
-	   though threads haven't been deleted yet, one would think
-	   that there is nothing to do, as threads of the dead process
-	   will be soon deleted, and threads of any other process were
-	   left running.  However, on some targets, threads survive a
-	   process exit event.  E.g., for the "checkpoint" command,
-	   when the current checkpoint/fork exits, linux-fork.c
-	   automatically switches to another fork from within
-	   target_mourn_inferior, by associating the same
-	   inferior/thread to another fork.  We haven't mourned yet at
-	   this point, but we must mark any threads left in the
-	   process as not-executing so that finish_thread_state marks
-	   them stopped (in the user's perspective) if/when we present
-	   the stop to the user.  */
-	mark_ptid = ptid_t (ecs->ptid.pid ());
-      }
-    else
-      mark_ptid = ecs->ptid;
-
-    set_executing (ecs->target, mark_ptid, false);
-
-    /* Likewise the resumed flag.  */
-    set_resumed (ecs->target, mark_ptid, false);
-  }
+  mark_non_executing_threads (ecs->target, ecs->ptid, ecs->ws);
 
   switch (ecs->ws.kind)
     {