linux-kernel - list splicing and list POISONing

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <alpine.LFD.1.00.0803250307110.29947@localhost.localdomain>
Date:	Tue, 25 Mar 2008 03:15:40 -0400 (EDT)
From:	"Robert P. J. Day" <rpjday@...shcourse.ca>
To:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: list splicing and list POISONing

  based on a conversation we've been having on the newbies list, i'm
curious about the rationale behind list "POISON"ing.

  from <linux/list.h>, when you splice one list into another, the
head element from the original list is left hanging in space with its
original prev and next pointer values unchanged:

===
static inline void __list_splice(struct list_head *list,
                                 struct list_head *head)
{
        struct list_head *first = list->next;
        struct list_head *last = list->prev;
        struct list_head *at = head->next;

        first->prev = head;
        head->next = first;

        last->next = at;
        at->prev = last;
}
===

  notice how neither list->prev nor list->next is set to, say, NULL to
represent that that list head now represents an empty list.  instead,
those pointers are left (rather dangerously) pointing into the
newly-spliced list, which means that, if you erroneously try to
traverse the list represented by the "list" list head, you'll end up
immediately over in the new list and loop there forever.  it would
seem to make more sense to set those pointers to either NULL or
LIST_POISON{1,2}, would it not?  just for safety's sake?

  it seems like that's what's done when an element is *deleted* (that
is, unlinked) from a list:

===
static inline void list_del(struct list_head *entry)
{
        __list_del(entry->prev, entry->next);
        entry->next = LIST_POISON1;
        entry->prev = LIST_POISON2;
}
===

  so shouldn't the same reasoning apply in the splicing example?

rday
--

========================================================================
Robert P. J. Day
Linux Consulting, Training and Annoying Kernel Pedantry:
    Have classroom, will lecture.

http://crashcourse.ca                          Waterloo, Ontario, CANADA
========================================================================
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/