On 8/18/20 10:06 AM, Alexei Starovoitov wrote:
On Tue, Aug 18, 2020 at 9:24 AM Yonghong Song <yhs@xxxxxx> wrote:index f21b5e1e4540..885b14cab2c0 100644 --- a/kernel/bpf/task_iter.c +++ b/kernel/bpf/task_iter.c @@ -27,6 +27,8 @@ static struct task_struct *task_seq_get_next(struct pid_namespace *ns, struct task_struct *task = NULL; struct pid *pid; + cond_resched(); + rcu_read_lock(); retry: pid = idr_get_next(&ns->idr, tid); @@ -137,6 +139,8 @@ task_file_seq_get_next(struct bpf_iter_seq_task_file_info *info, struct task_struct *curr_task; int curr_fd = info->fd; + cond_resched(); +Instead of adding it to every *seq_get_next() it probably should be in bpf_seq_read().
Yes, we can add cond_resched() to bpf_seq_read(). This should cover both cases. Will make the change.
If cond_resched() is needed in task_file_seq_get_next() it should probably be after 'again:'.
We probably do not need here unless all tasks have zero files or each file just closed with f->f_count == 0 but the file pointer still there.
