[prev in list] [next in list] [prev in thread] [next in thread] 

List:       binutils-bugs
Subject:    [Bug binutils/29285] New: gdb ignores stub socket disconnect when no program running
From:       "Manfred.Steiner at gmx dot at" <sourceware-bugzilla () sourceware ! org>
Date:       2022-06-25 7:13:52
Message-ID: bug-29285-70 () http ! sourceware ! org/bugzilla/
[Download RAW message or body]

https://sourceware.org/bugzilla/show_bug.cgi?id=29285

            Bug ID: 29285
           Summary: gdb ignores stub socket disconnect when no program
                    running
           Product: binutils
           Version: 2.39 (HEAD)
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: binutils
          Assignee: unassigned at sourceware dot org
          Reporter: Manfred.Steiner at gmx dot at
  Target Milestone: ---

I am working on git version commit 325600eb1175e6f989f201ae885c8fae82953264
(Thu Jun 23 00:00:24 2022 +0000).

When no (stub-)thread is active, gdb ignores the TCP FIN packet and does not
close the remote debugging connection.

Steps to reproduce:

1) Shell 1: start gdbserver
   $ gdbserver --multi localhost:2424
   Listening on port 2424
   Remote debugging from host 127.0.0.1, port 33498
   Remote side has terminated connection.  GDBserver will reopen the
connection.
   Listening on port 2424

2) Shell 2: start gdb
   $ gdb
   (gdb) target extended-remote localhost:2424
   Remote debugging using localhost:2424

3) close shell 1 (gdbServer) with CTRL-C
   -> TCP/FIN is send to gdb (but ignored there)

4) Shell 2
   (gdb) target extended-remote localhost:2424
   Already connected to a remote target.  Disconnect? (y or n) 

gdb thinks that connection to server is already open.

If you do the same with an established connection (thread) to a running program
you get on gdb-server CTRL-C ...

  (gdb) c
  Continuing.
  Remote connection closed

If the program is not running, TCP FIN packet is also ignored and gdb thinks
that connection is already established.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
[prev in list] [next in list] [prev in thread] [next in thread]