This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[PATCH] Prune program spaces in remove-inferiors
- From: Doug Evans <xdje42 at gmail dot com>
- To: gdb-patches at sourceware dot org
- Date: Sun, 02 Mar 2014 13:23:16 -0800
- Subject: [PATCH] Prune program spaces in remove-inferiors
- Authentication-results: sourceware.org; auth=none
Hi.
While writing a test for Guile program space support I found that
after I removed an inferior I would still see it show up in the
progspaces list. And then if I did "mt info program-spaces" it would
then go away. While info commands shouldn't in general exhibit side
effects like this, I'm not too worried here (*1).
But I think remove-inferiors should call prune_program_spaces.
Is there a reason for it to not to?
Regression tested on amd64-linux.
2014-03-02 Doug Evans <xdje42@gmail.com>
* inferior.c (prune_inferiors): Fix comment.
(remove_inferior_command): Call prune_program_spaces.
diff --git a/gdb/inferior.c b/gdb/inferior.c
index 90d9649..23da0c7 100644
--- a/gdb/inferior.c
+++ b/gdb/inferior.c
@@ -475,8 +475,8 @@ have_live_inferiors (void)
return inf != NULL;
}
-/* Prune away automatically added program spaces that aren't required
- anymore. */
+/* Prune away any unused inferiors, and then prune away no longer used
+ program spaces. */
void
prune_inferiors (void)
@@ -788,6 +788,8 @@ remove_inferior_command (char *args, int from_tty)
delete_inferior_1 (inf, 1);
}
+
+ prune_program_spaces ();
}
struct inferior *
---
(*1): For reference sake, "info threads" is a more serious
case of this when debugging remotely, ref: "Finding new threads in the
inferior" in https://sourceware.org/gdb/wiki/LocalRemoteFeatureParity