[RFC] Implement 'detach pid'.

[RFC] Implement 'detach pid'.

[Vladimir Prus]

As Marc has noted, -target-detach does not accept a pid to detach from,
and CLI detach does not accept pid either. This patch fixes it. Does it
look OK?

- Volodya

	* infcmd.c (_initialize_infcmd): Make 'detach' accept
	parameter.
	* mi/mi-cmds.c (mi_cmds): Make '-target-detach' pass parameter
	to 'detach'.
	* remote.c (remote_detach_1): Interpret the passed
	parameter as a pid to detach from.
---
 gdb/infcmd.c     |    2 +-
 gdb/mi/mi-cmds.c |    2 +-
 gdb/remote.c     |   13 +++++++++++--
 3 files changed, 13 insertions(+), 4 deletions(-)

diff --git a/gdb/infcmd.c b/gdb/infcmd.c
index e650789..dcdaa7f 100644
--- a/gdb/infcmd.c
+++ b/gdb/infcmd.c
@@ -2250,7 +2250,7 @@ to specify the program, and to load its symbol table."));
 Detach a process or file previously attached.\n\
 If a process, it is no longer traced, and it continues its execution.  If\n\
 you were debugging a file, the file is closed and gdb no longer accesses it."),
-		  &detachlist, "detach ", 0, &cmdlist);
+		  &detachlist, "detach ", 1, &cmdlist);
 
   add_com ("disconnect", class_run, disconnect_command, _("\
 Disconnect from a target.\n\
diff --git a/gdb/mi/mi-cmds.c b/gdb/mi/mi-cmds.c
index 68c5353..b9c4f8e 100644
--- a/gdb/mi/mi-cmds.c
+++ b/gdb/mi/mi-cmds.c
@@ -120,7 +120,7 @@ struct mi_cmd mi_cmds[] =
   { "symbol-type", { NULL, 0 }, NULL },
   { "target-attach", { "attach", 1 }, NULL },
   { "target-compare-sections", { NULL, 0 }, NULL },
-  { "target-detach", { "detach", 0 }, 0 },
+  { "target-detach", { "detach", 1 }, 0 },
   { "target-disconnect", { "disconnect", 0 }, 0 },
   { "target-download", { NULL, 0 }, mi_cmd_target_download},
   { "target-exec-status", { NULL, 0 }, NULL },
diff --git a/gdb/remote.c b/gdb/remote.c
index 51b8ec5..d8a7500 100644
--- a/gdb/remote.c
+++ b/gdb/remote.c
@@ -2917,11 +2917,20 @@ remote_open_1 (char *name, int from_tty, struct target_ops *target, int extended
 static void
 remote_detach_1 (char *args, int from_tty, int extended)
 {
-  int pid = ptid_get_pid (inferior_ptid);
+  int pid;
   struct remote_state *rs = get_remote_state ();
 
   if (args)
-    error (_("Argument given to \"detach\" when remotely debugging."));
+    {      
+      char *end = args;
+      pid = strtol (args, &end, 10);
+      if (*end != '\0')
+	error (_("Cannot parse process id '%s'"), args);
+      if (!in_process_list (pid))
+	error (_("Invalid process id %d"), pid);
+    }
+  else
+    pid = ptid_get_pid (inferior_ptid);
 
   if (!target_has_execution)
     error (_("No process to detach from."));
-- 
1.5.3.5

Re: [RFC] Implement 'detach pid'.

[Vladimir Prus]

Vladimir Prus wrote:

> 
> As Marc has noted, -target-detach does not accept a pid to detach from,
> and CLI detach does not accept pid either. This patch fixes it. Does it
> look OK?

Doh! This patch was not meant for gdb-patches, as the pre-requisite patch
allowing GDB to attach to more that one process is not submitted yet.

Sorry for the noise,
Volodya

Re: [RFC] Implement 'detach pid'.

[Tom Tromey]

>>>>> "Vladimir" == Vladimir Prus <vladimir@codesourcery.com> writes:

Vladimir> As Marc has noted, -target-detach does not accept a pid to
Vladimir> detach from, and CLI detach does not accept pid either. This
Vladimir> patch fixes it. Does it look OK?

Channelling Eli here ... it needs a documentation update and perhaps a
NEWS entry :-)

Tom

Re: [RFC] Implement 'detach pid'.

[Eli Zaretskii]

> Cc: gdb-patches@sources.redhat.com
> From: Tom Tromey <tromey@redhat.com>
> Reply-To: tromey@redhat.com
> Date: Sat, 05 Jul 2008 11:21:03 -0600
> 
> >>>>> "Vladimir" == Vladimir Prus <vladimir@codesourcery.com> writes:
> 
> Vladimir> As Marc has noted, -target-detach does not accept a pid to
> Vladimir> detach from, and CLI detach does not accept pid either. This
> Vladimir> patch fixes it. Does it look OK?
> 
> Channelling Eli here

Thanks!