On Fri, May 31, 2019 at 02:32:34PM +0200, Miroslav Benes wrote:
> On Fri, 31 May 2019, Petr Mladek wrote:
>
> > WARN_ON_ONCE() could not be called safely under rq lock because
> > of console deadlock issues.
> >
> > It can be simply removed. A better descriptive message is written
> > in klp_enable_patch() when klp_have_reliable_stack() fails.
> > The remaining debug message is good enough.
> >
> > Signed-off-by: Petr Mladek <pmladek@xxxxxxxx>
> > ---
> > kernel/livepatch/transition.c | 1 -
> > 1 file changed, 1 deletion(-)
> >
> > diff --git a/kernel/livepatch/transition.c b/kernel/livepatch/transition.c
> > index abb2a4a2cbb2..1bf362df76e1 100644
> > --- a/kernel/livepatch/transition.c
> > +++ b/kernel/livepatch/transition.c
> > @@ -247,7 +247,6 @@ static int klp_check_stack(struct task_struct *task, char *err_buf)
> > int ret, nr_entries;
> >
> > ret = stack_trace_save_tsk_reliable(task, entries, ARRAY_SIZE(entries));
> > - WARN_ON_ONCE(ret == -ENOSYS);
> > if (ret < 0) {
> > snprintf(err_buf, STACK_ERR_BUF_SIZE,
> > "%s: %s:%d has an unreliable stack\n",
>
> The current situation is not the best, but I think the patch improves it
> only slightly. I see two possible solutions.
>
> 1. we either revert commit 1d98a69e5cef ("livepatch: Remove reliable
> stacktrace check in klp_try_switch_task()"), so that klp_check_stack()
> returns right away.
>
> 2. or we test ret from stack_trace_save_tsk_reliable() for ENOSYS and
> return.
>
> In my opinion either of them is better than what we have now (and what we
> would have with the patch), because klp_check_stack() returns, but it
> prints out that a task has an unreliable stack. Yes, it is pr_debug() only
> in the end, but still.
>
> I don't have a preference and my understanding is that Petr does not want
> to do v4. I can prepare a patch, but it would be nice to choose now. Josh?
> Anyone else?
My vote would be #1 -- just revert 1d98a69e5cef.
--
Josh
