Fix LIST_INSERT breakage
PATCH 1/1 Fix LIST_INSERT breakage
In 2.4.25, this patch was included:
which has a questionable change in LIST_INSERT due to the disappearing cmpfn. But it appears this macro is unused anywhere in the tree, so perhaps removing it is better than fixing it?
Phil Oester
linux/netfilter_ipv4/listhelp.h 2004-03-23 20:01:09.603032216 -0500 renamed from linux/netfilter_ipv4/listhelp.h-orig 2004-03-23 19:57:00.696775216 -0500 +0 -11
@@ -89,17 +89,6 @@ list_add(new, head); } -/* Insert according to ordering function; insert before first true. */ -#define LIST_INSERT(head, new, cmpfn) \ -do { \ - struct list_head *__i; \ - ASSERT_WRITE_LOCK(head); \ - list_for_each(__i, (head)) \ - if ((new), (typeof (new))__i) \ - break; \ - list_add((struct list_head *)(new), __i->prev); \ -} while(0) - /* If the field after the list_head is a nul-terminated string, you can use these functions. */ static inline int __list_cmp_name(const void *i, const char *name)
Quoted 6 lines · Phil Oester
In 2.4.25, this patch was included:
which has a questionable change in
LIST_INSERTdue to the disappearing cmpfn.
Could you be more specific about your objections against that changeset?
Quoted 2 lines
But it appears this macro is unused anywhere in the tree, so perhaps removing it is better than fixing it?
mh, we also don't use it in patch-o-matic-ng. So we could remove it, true.
Quoted 1 line
Phil Oester
Signature · 5 lines
- Harald Welte <[email protected]> http://www.netfilter.org/ ============================================================================ "Fragmentation is like classful addressing -- an interesting early architectural error that shows how much experimentation was going on while IP was being designed." -- Paul Vixie
LIST_INSERT takes a 'cmpfn' argument, and is used pre-change:
- for (__i = (head)->next; \
- !cmpfn((new), (typeof (new))__i) && __i != (head); \
- __i = __i->next); \
but not post-change:
+ list_for_each(__i, (head)) \
+ if ((new), (typeof (new))__i) \ <---
+ break; \
'if (cmpfn((new), ......' is probably what was intended?
But either way, it is unused and thus moot.
Phil
Quoted 9 lines · Harald Welte
Quoted 6 lines · Phil Oester
In 2.4.25, this patch was included:
which has a questionable change in
LIST_INSERTdue to the disappearing cmpfn.Could you be more specific about your objections against that changeset?