linux-kernel - Re: [PATCH v2 1/9] llist: Provide a safe version for llist_for

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>] [<thread-prev] [day] [month] [year] [list]

Date:   Tue, 14 Feb 2017 14:59:36 +0800
From:   "Huang\, Ying" <ying.huang@...el.com>
To:     Byungchul Park <byungchul.park@....com>
Cc:     "Huang\, Ying" <ying.huang@...el.com>, <peterz@...radead.org>,
        <mingo@...nel.org>, <neilb@...e.de>, <nab@...ux-iscsi.org>,
        <viro@...iv.linux.org.uk>, <oleg@...hat.com>, <shli@...nel.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 1/9] llist: Provide a safe version for llist_for_each

Byungchul Park <byungchul.park@....com> writes:

> On Mon, Feb 13, 2017 at 04:58:05PM +0900, Byungchul Park wrote:
>> On Mon, Feb 13, 2017 at 03:52:44PM +0800, Huang, Ying wrote:
>> > Byungchul Park <byungchul.park@....com> writes:
>> > 
>> > > On Mon, Feb 13, 2017 at 03:36:33PM +0800, Huang, Ying wrote:
>> > >> Byungchul Park <byungchul.park@....com> writes:
>> > >> 
>> > >> > Sometimes we have to dereference next field of llist node before entering
>> > >> > loop becasue the node might be deleted or the next field might be
>> > >> > modified within the loop. So this adds the safe version of llist_for_each,
>> > >> > that is, llist_for_each_safe.
>> > >> >
>> > >> > Signed-off-by: Byungchul Park <byungchul.park@....com>
>> > >> > ---
>> > >> >  include/linux/llist.h | 19 +++++++++++++++++++
>> > >> >  1 file changed, 19 insertions(+)
>> > >> >
>> > >> > diff --git a/include/linux/llist.h b/include/linux/llist.h
>> > >> > index fd4ca0b..4c508a5 100644
>> > >> > --- a/include/linux/llist.h
>> > >> > +++ b/include/linux/llist.h
>> > >> > @@ -105,6 +105,25 @@ static inline void init_llist_head(struct llist_head *list)
>> > >> >  	for ((pos) = (node); pos; (pos) = (pos)->next)
>> > >> >  
>> > >> >  /**
>> > >> > + * llist_for_each_safe - iterate over some deleted entries of a lock-less list
>> > >> > + *			 safe against removal of list entry
>> > >> > + * @pos:	the &struct llist_node to use as a loop cursor
>> > >> > + * @n:		another type * to use as temporary storage
>> > >> 
>> > >> s/type */&struct llist_node/
>> > >
>> > > Yes.
>> > >
>> > >> 
>> > >> > + * @node:	the first entry of deleted list entries
>> > >> > + *
>> > >> > + * In general, some entries of the lock-less list can be traversed
>> > >> > + * safely only after being deleted from list, so start with an entry
>> > >> > + * instead of list head.
>> > >> > + *
>> > >> > + * If being used on entries deleted from lock-less list directly, the
>> > >> > + * traverse order is from the newest to the oldest added entry.  If
>> > >> > + * you want to traverse from the oldest to the newest, you must
>> > >> > + * reverse the order by yourself before traversing.
>> > >> > + */
>> > >> > +#define llist_for_each_safe(pos, n, node)			\
>> > >> > +	for ((pos) = (node); (pos) && ((n) = (pos)->next, true); (pos) = (n))
>> > >> > +
>> > >> 
>> > >> Following the style of other xxx_for_each_safe,
>> > >> 
>> > >> #define llist_for_each_safe(pos, n, node)			\
>> > >> 	for (pos = (node), (pos && (n = pos->next)); pos; pos = n, n = pos->next)
>> > >
>> > > Do you think it should be modified? I think mine is simpler. No?
>> > 
>> > Personally I prefer the style of other xxx_for_each_safe().
>> 
>> Yes, I will modify it as you recommand.
>> 
>> Thank you very much.
>
> I wanted to modify it as you recommanded but it has a bug. It should be
> (to fix the bug):
>
>    for (pos = (node), (pos && (n = pos->next)); pos; pos = n, (pos && \
>    (n = pos->next)))
>
> Don't you think this is too messy? Or do I miss something? I still think
> the following is neater and simpler.
>
>    for (pos = node; pos && (n = pos->next, true); pos = n)

OK.  This looks better.

Best Regards,
Huang, Ying

> Or could you recommand another preference?