On Wed, Mar 26, 2008 at 11:18 AM, Balaji Rao <[EMAIL PROTECTED]> wrote:
> +
> +static s64 cpu_cgroup_read_stat(struct cpu_cgroup_stat *stat,
> + enum cpu_cgroup_stat_index idx)
> +{
> + int cpu;
> + s64 ret = 0;
> + for_each_possible_cpu(cpu)
> + ret += stat->cpustat[cpu].count[idx];
On a 32-bit architecture I think this could race with a non-atomic
update that crosses a 32-bit boundary and get a corrupted result.
Paul
_______________________________________________
Containers mailing list
[EMAIL PROTECTED]
https://lists.linux-foundation.org/mailman/listinfo/containers
_______________________________________________
Devel mailing list
[email protected]
https://openvz.org/mailman/listinfo/devel
