From: Pedro Alves <palves@redhat.com>
Date: Fri, 10 Jan 2020 20:05:45 +0000 (+0000)
Subject: Don't check target is running in remote_target::mourn_inferior
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=ddf5db90a175756b3a5c39ee87d549d9f9d09d28;p=binutils-gdb.git

Don't check target is running in remote_target::mourn_inferior

I believe the tail end of remote_target::mourn_inferior is broken, and
it's been broken for too long to even bother trying to fix.  Most
probably nobody needs it.  If the code is reached and we find the
target is running, we'd need to resync the thread list, at least,
since generic_mourn_inferior got rid of all the threads in the
inferior, otherwise, we'd hit an assertion on the next call to
inferior_thread(), for example.  A "correct" fix would probably
involve restarting the whole remote_target::start_remote requence,
exactly as if we had completely disconnected and reconnected from
scratch.

Note that regular stub debugging usually uses plain target remote, but
this code is only reachable in target extended-mode:

- The !remote_multi_process_p check means that it's only reacheable if
  the stub does not support multi-process.  I.e., there can only ever
  be one live process.

- remote_target::mourn_inferior has this at the top:

  /* In 'target remote' mode with one inferior, we close the connection.  */
  if (!rs->extended && number_of_live_inferiors () <= 1)
    {
      unpush_target (this);

      /* remote_close takes care of doing most of the clean up.  */
      generic_mourn_inferior ();
      return;
    }

  Which means that if we only had one live inferior (which for our
  case, must be true), we'll have closed the connection already,
  unless we're in extended-remote mode.

gdb/ChangeLog:
2020-01-10  Pedro Alves  <palves@redhat.com>

	* remote.c (remote_target::mourn_inferior): No longer check
	whether the target is running.
---

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index 6e97bd46a15..ff1bc14eca3 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,3 +1,8 @@
+2020-01-10  Pedro Alves  <palves@redhat.com>
+
+	* remote.c (remote_target::mourn_inferior): No longer check
+	whether the target is running.
+
 2020-01-10  Pedro Alves  <palves@redhat.com>
 
 	* corelow.c (core_target::has_execution): Change parameter type to
diff --git a/gdb/remote.c b/gdb/remote.c
index 539b27ebd29..51fa5da0c18 100644
--- a/gdb/remote.c
+++ b/gdb/remote.c
@@ -9843,26 +9843,6 @@ remote_target::mourn_inferior ()
 
   /* Call common code to mark the inferior as not running.  */
   generic_mourn_inferior ();
-
-  if (!have_inferiors ())
-    {
-      if (!remote_multi_process_p (rs))
-	{
-	  /* Check whether the target is running now - some remote stubs
-	     automatically restart after kill.	*/
-	  putpkt ("?");
-	  getpkt (&rs->buf, 0);
-
-	  if (rs->buf[0] == 'S' || rs->buf[0] == 'T')
-	    {
-	      /* Assume that the target has been restarted.  Set
-		 inferior_ptid so that bits of core GDB realizes
-		 there's something here, e.g., so that the user can
-		 say "kill" again.  */
-	      inferior_ptid = magic_null_ptid;
-	    }
-	}
-    }
 }
 
 bool