Re: Segfault on ctrl-C (curpos)

Hello Liam,

Thanks for reporting!

On Tue, Dec 20, 2016, at 09:21, Liam Gretton wrote:
> I build Nano for our SLES and CentOS systems. For quite a few versions
> including 2.7.2 my SLES builds segfault if ctrl-C (curpos) is used.
> 
> I've finally got round to rebuilding with debugging enabled and the
> following is reported when ctrl-C is pressed (new Nano session, nothing
> in the buffer):
> 
>  get_key_buffer(): the sequence of hex codes: 163
>  before parsing:  keycode = 355, escapes = 0, byte_digits = 0
>  whoops, returning null given func 4270736 in menu 1
>  Segmentation fault

Please try the attached patch.  It fixes it for me.

> The error therefore suggests Cancel is being called out of context
> instead of curpos.
> 
> Any ideas what might be wrong?

Your keyboard/input-system produces an actual KEY_CANCEL
code when you press ^C, instead of simply 0x03, what most
other systems seem to do.  Do you compile ncurses in a special
way on SLES to make it report ^C as KEY_CANCEL?

> Nanorc:
> 
> set backup
> set const
> set cut

Wow.  You actually use 'set cut'?  Not being used to it,
I find that behavior quite confusing.  (I've tried running
for a moment with your config. :) )

You may wish to update "const" to "constantshow" --
its new name.

Benno

-- 
http://www.fastmail.com - The professional email service

_______________________________________________
Help-nano mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/help-nano

diff --git a/src/winio.c b/src/winio.c
index 7237b8ac..8a5d70a3 100644
--- a/src/winio.c
+++ b/src/winio.c
@@ -667,7 +667,7 @@ int parse_kbinput(WINDOW *win)
 #endif
 	/* Slang doesn't support KEY_CANCEL. */
 	case KEY_CANCEL:
-	    return first_sc_for(currmenu, do_cancel)->keycode;
+	    return sc_seq_or(do_cancel, 0x03);
 #endif
 #ifdef KEY_SUSPEND
 #ifdef KEY_SSUSPEND