On Tue, 2022-07-26 at 15:19 +0200, Jiri Olsa wrote:
> On Tue, Jul 26, 2022 at 02:13:17PM +0200, Jiri Olsa wrote:
> > > -static struct task_struct *task_seq_get_next(struct
> > > pid_namespace *ns,
> > > +static struct task_struct *task_seq_get_next(struct
> > > bpf_iter_seq_task_common *common,
> > > u32 *tid,
> > > bool
> > > skip_if_dup_files)
> > > {
> > > struct task_struct *task = NULL;
> > > struct pid *pid;
> > >
> > > + if (common->type == BPF_TASK_ITER_TID) {
> > > + if (*tid)
> > > + return NULL;
> >
> > I tested and this condition breaks it for fd iterations, not sure
> > about
> > the task and vma, because they share this function
> >
> > if bpf_seq_read is called with small buffer there will be multiple
> > calls
> > to task_file_seq_get_next and second one will stop in here, even if
> > there
> > are more files to be displayed for the task in filter
>
> I mean there will be multiple calls of following sequence:
>
> bpf_seq_read
> task_file_seq_start
> task_seq_get_next
>
> and 2nd one will return NULL in task_seq_get_next,
> because info->tid is already set
Ok! I got your point. I will fix it ASAP.
