Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Pedro Alves <palves@redhat.com>
To: gdb-patches@sourceware.org
Subject: [PATCH v2 23/25] Handle UI terminal closed
Date: Mon, 21 Mar 2016 15:27:00 -0000	[thread overview]
Message-ID: <1458573675-15478-24-git-send-email-palves@redhat.com> (raw)
In-Reply-To: <1458573675-15478-1-git-send-email-palves@redhat.com>

Without this, GDB exits if a secondary UIs terminal/input stream is
closed:

 $ ./gdb -ex "new-ui mi /dev/pts/6"
 New UI allocated
	 <<< close /dev/pts/6
 (gdb) Error detected on fd 9
 $

We want that for the main UI, but not secondary UIs.
---
 gdb/event-top.c | 25 ++++++++++++++++++-------
 gdb/top.c       | 31 +++++++++++++++++++++++++++++++
 gdb/top.h       |  1 +
 3 files changed, 50 insertions(+), 7 deletions(-)

diff --git a/gdb/event-top.c b/gdb/event-top.c
index 69f6be7..39b9633 100644
--- a/gdb/event-top.c
+++ b/gdb/event-top.c
@@ -420,19 +420,30 @@ stdin_event_handler (int error, gdb_client_data client_data)
 {
   struct ui *ui = (struct ui *) client_data;
 
-  /* Switch to the UI whose input descriptor woke up the event
-     loop.  */
-  current_ui = ui;
-
   if (error)
     {
-      printf_unfiltered (_("error detected on stdin\n"));
+      /* Switch to the main UI, so diagnostics always go there.  */
+      current_ui = main_ui;
+
       delete_file_handler (ui->input_fd);
-      /* If stdin died, we may as well kill gdb.  */
-      quit_command ((char *) 0, stdin == ui->instream);
+      if (main_ui == ui)
+	{
+	  /* If stdin died, we may as well kill gdb.  */
+	  printf_unfiltered (_("error detected on stdin\n"));
+	  quit_command ((char *) 0, stdin == ui->instream);
+	}
+      else
+	{
+	  /* Simply delete the UI.  */
+	  delete_ui (ui);
+	}
     }
   else
     {
+      /* Switch to the UI whose input descriptor woke up the event
+	 loop.  */
+      current_ui = ui;
+
       do
 	{
 	  call_stdin_event_handler_again_p = 0;
diff --git a/gdb/top.c b/gdb/top.c
index 12064bc..7d6d7cb 100644
--- a/gdb/top.c
+++ b/gdb/top.c
@@ -282,6 +282,37 @@ new_ui (FILE *instream, FILE *outstream, FILE *errstream)
   return ui;
 }
 
+static void
+free_ui (struct ui *ui)
+{
+  ui_file_delete (ui->m_gdb_stdin);
+  ui_file_delete (ui->m_gdb_stdout);
+  ui_file_delete (ui->m_gdb_stderr);
+
+  xfree (ui);
+}
+
+void
+delete_ui (struct ui *todel)
+{
+  struct ui *ui, *uiprev;
+
+  uiprev = NULL;
+
+  for (ui = ui_list; ui != NULL; uiprev = ui, ui = ui->next)
+    if (ui == todel)
+      break;
+
+  gdb_assert (ui != NULL);
+
+  if (uiprev != NULL)
+    uiprev->next = ui->next;
+  else
+    ui_list = ui->next;
+
+  free_ui (ui);
+}
+
 /* Handler for SIGHUP.  */
 
 #ifdef SIGHUP
diff --git a/gdb/top.h b/gdb/top.h
index df76204..bac5a33 100644
--- a/gdb/top.h
+++ b/gdb/top.h
@@ -161,6 +161,7 @@ extern void switch_thru_all_uis_next (struct switch_thru_all_uis *state);
   for (UI = ui_list; UI; UI = UI->next)		\
 
 extern struct ui *new_ui (FILE *instream, FILE *outstream, FILE *errstream);
+extern void delete_ui (struct ui *todel);
 
 extern void restore_ui_cleanup (void *data);
 
-- 
2.5.0


  parent reply	other threads:[~2016-03-21 15:27 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1458573675-15478-1-git-send-email-palves@redhat.com>
2016-03-21 15:21 ` [PATCH v2 03/25] Make the interpreters be per UI Pedro Alves
2016-03-21 15:21 ` [PATCH v2 22/25] Make main_ui be heap allocated Pedro Alves
2016-03-22 10:14   ` Yao Qi
2016-05-06 11:50     ` Pedro Alves
2016-03-21 15:21 ` [PATCH v2 01/25] Introduce "struct ui" Pedro Alves
2016-03-21 15:21 ` [PATCH v2 02/25] Make gdb_stdout&co be per UI Pedro Alves
2016-03-21 15:21 ` [PATCH v2 12/25] Make command line editing (use of readline) " Pedro Alves
2016-03-21 15:22 ` [PATCH v2 18/25] Replace the sync_execution global with a new enum prompt_state tristate Pedro Alves
2016-03-21 15:22 ` [PATCH v2 13/25] Always process target events in the main UI Pedro Alves
2016-03-22 10:26   ` Yao Qi
2016-05-06 11:53     ` Pedro Alves
2016-03-21 15:22 ` [PATCH v2 24/25] Add new command to create extra console/mi UI channels Pedro Alves
2016-03-21 16:31   ` Eli Zaretskii
2016-03-21 16:51     ` Pedro Alves
2016-03-21 17:12       ` Eli Zaretskii
2016-03-21 17:57         ` Pedro Alves
2016-05-26 11:43           ` Pedro Alves
2016-05-26 15:46             ` Eli Zaretskii
2016-05-26 16:03               ` Pedro Alves
2016-05-26 16:36                 ` Eli Zaretskii
2016-05-26 16:41                   ` Pedro Alves
2016-03-21 15:26 ` [PATCH v2 08/25] Make input_fd be per UI Pedro Alves
2016-03-22  9:46   ` Yao Qi
2016-05-06 11:53     ` Pedro Alves
2016-03-21 15:26 ` [PATCH v2 10/25] Delete def_uiout Pedro Alves
2016-03-21 15:27 ` [PATCH v2 21/25] Only send sync execution command output to the UI that ran the command Pedro Alves
2016-03-21 15:27 ` [PATCH v2 15/25] Introduce display_mi_prompt Pedro Alves
2016-03-21 15:27 ` Pedro Alves [this message]
2016-03-21 15:27 ` [PATCH v2 09/25] Make outstream be per UI Pedro Alves
2016-03-21 15:29 ` [PATCH v2 04/25] Introduce interpreter factories Pedro Alves
2016-03-22  8:55   ` Yao Qi
2016-05-06 11:49     ` Pedro Alves
2016-03-21 15:29 ` [PATCH v2 05/25] Make the intepreters output to all UIs Pedro Alves
2016-03-22  9:33   ` Yao Qi
2016-05-06 12:19     ` Pedro Alves
2016-03-21 15:29 ` [PATCH v2 11/25] Make current_ui_out be per UI Pedro Alves
2016-03-21 15:29 ` [PATCH v2 17/25] Make gdb_in_secondary_prompt_p() " Pedro Alves
2016-03-21 15:29 ` [PATCH v2 16/25] Simplify starting the command event loop Pedro Alves
2016-03-21 15:30 ` [PATCH v2 06/25] Always run async signal handlers in the main UI Pedro Alves
2016-03-21 15:30 ` [PATCH v2 19/25] New function should_print_stop_to_console Pedro Alves
2016-03-21 15:30 ` [PATCH v2 07/25] Make instream and serial_stdin be per UI Pedro Alves
2016-03-21 15:30 ` [PATCH v2 20/25] Push thread->control.command_interp to the struct thread_fsm Pedro Alves
2016-03-21 15:30 ` [PATCH v2 25/25] Add command to list UIs Pedro Alves
2016-03-22 10:36   ` Yao Qi
2016-05-06 11:49     ` Pedro Alves
2016-03-21 15:39 ` [PATCH v2 14/25] Make target_terminal_inferior/ours almost nops on non-main UIs Pedro Alves
2016-03-21 16:34 ` [PATCH v2 00/25] Towards great frontend GDB consoles Eli Zaretskii
2016-03-21 17:02   ` Pedro Alves
2016-03-21 17:17     ` Eli Zaretskii
2016-03-21 17:43     ` Marc Khouzam
2016-03-21 18:35 ` Marc Khouzam
2016-03-21 18:51   ` Pedro Alves
2016-03-21 19:06     ` Marc Khouzam
2016-05-06 12:58     ` Pedro Alves
2016-03-22 10:41 ` Yao Qi
2016-05-06 11:58   ` Pedro Alves

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1458573675-15478-24-git-send-email-palves@redhat.com \
    --to=palves@redhat.com \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox