lkml.org 
[lkml]   [2016]   [Mar]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v6 1/4] lib/percpu-list: Per-cpu list with associated per-cpu locks
On Fri 18-03-16 15:44:01, Waiman Long wrote:
> +static __always_inline bool
> +__pcpu_list_next_cpu(struct pcpu_list_head *head, struct pcpu_list_state *state)
> +{
> + if (state->lock)
> + spin_unlock(state->lock);
> +next_cpu:
> + /*
> + * for_each_possible_cpu(cpu)
> + */
> + state->cpu = cpumask_next(state->cpu, cpu_possible_mask);
> + if (state->cpu >= nr_cpu_ids)
> + return false; /* All the per-cpu lists iterated */
> +
> + state->head = &per_cpu_ptr(head, state->cpu)->list;
> + if (list_empty(state->head))
> + goto next_cpu;
> +
> + state->lock = &per_cpu_ptr(head, state->cpu)->lock;
> + spin_lock(state->lock);
> + state->curr = list_entry(state->head->next,
> + struct pcpu_list_node, list);
> + return true;

Waiman, I repeat it for the third time as you keep ignoring it: This is
*racy*. The list for state->cpu can be empty by the time you acquire
state->lock and thus state->curr will point somewhere around the head of
the list but definitely not to a list member where it should.

Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR

\
 
 \ /
  Last update: 2016-03-21 11:21    [W:0.102 / U:1.064 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site