hi list,
this patch adds a check that
the address p is larger the current buffer adress text when accessing p-1
because at begining p==text is possible.
the first place have a dot>text (cursor pos) it could be that p=q is assumed
no idea if this is true
re,
wh
--- editors/vi_org.c 2008-06-23 17:42:13.000000000 +0200
+++ ../vi.c 2008-06-23 17:38:10.000000000 +0200
@@ -1659,12 +1659,12 @@
cmdcnt = 0;
end_cmd_q(); // stop adding to q
last_status_cksum = 0; // force status update
- if ((p > text) && (p[-1] != '\n') && (dot > text)) {
+ if ((p[-1] != '\n') && (dot > text)) {
p--;
}
} else if (c == erase_char || c == 8 || c == 127) { // Is this a BS
// 123456789
- if ((p > text) && (p[-1] != '\n') && (dot>text)) {
+ if ((p[-1] != '\n') && (dot>text)) {
p--;
p = text_hole_delete(p, p); // shrink buffer 1 char
}
@@ -3594,7 +3594,7 @@
case 'O': // O- open a empty line above
// 0i\n ESC -i
p = begin_line(dot);
- if ((p > text) && p[-1] == '\n') {
+ if (p[-1] == '\n') {
dot_prev();
case 'o': // o- open a empty line below; Yes, I
know it is in the midd
le of the "if (..."
dot_end();
_______________________________________________
busybox mailing list
[email protected]
http://busybox.net/cgi-bin/mailman/listinfo/busybox
