[PATCH] gdb/tui: Don't cast between window types.

[PATCH] gdb/tui: Don't cast between window types.

[Andrew Burgess]

I'll push the following patch as obvious in a couple of days if no one
objects.

Though the existing code is currently fine, I think the new version is
clearer and more robust.

OK to apply?

Thanks,
Andrew

---

Instead of casting between structure types to get the 'tui_gen_win_info'
info from a 'tui_win_info' access the generic member variable.  This is
inline with what is done throughout the rest of the tui code.

gdb/ChangeLog:

	* tui/tui-win.c (tui_set_focus): Use structure member 'generic'
	instead of casting the structure type.
---
 gdb/ChangeLog     | 5 +++++
 gdb/tui/tui-win.c | 3 +--
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index 4636653..13d6da9 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,5 +1,10 @@
 2015-07-06  Andrew Burgess  <andrew.burgess@embecosm.com>
 
+	* tui/tui-win.c (tui_set_focus): Use structure member 'generic'
+	instead of casting the structure type.
+
+2015-07-06  Andrew Burgess  <andrew.burgess@embecosm.com>
+
 	* tui/tui-data.c (tui_partial_win_by_name): Window name is const.
 	(tui_win_name): Make parameter and result const.
 	* tui/tui-data.h (tui_win_name): Make parameter and result const.
diff --git a/gdb/tui/tui-win.c b/gdb/tui/tui-win.c
index feb360b..629d54d 100644
--- a/gdb/tui/tui-win.c
+++ b/gdb/tui/tui-win.c
@@ -1059,8 +1059,7 @@ The window name specified must be valid and visible.\n"));
 	tui_refresh_data_win ();
       xfree (buf_ptr);
       printf_filtered (_("Focus set to %s window.\n"),
-		       tui_win_name ((struct tui_gen_win_info *)
-				     tui_win_with_focus ()));
+		       tui_win_name (&tui_win_with_focus ()->generic));
     }
   else
     warning (_("Incorrect Number of Arguments.\n%s"), FOCUS_USAGE);
-- 
2.4.0

Re: [PATCH] gdb/tui: Don't cast between window types.

[Pedro Alves]

On 07/06/2015 05:17 PM, Andrew Burgess wrote:
> I'll push the following patch as obvious in a couple of days if no one
> objects.
> 
> Though the existing code is currently fine, I think the new version is
> clearer and more robust.
> 
> OK to apply?

Yes, please go ahead.

Thanks,
Pedro Alves