On Thu, Oct 14, 2021 at 07:03:07PM +0100, Mark Rutland wrote:

On Fri, Oct 08, 2021 at 03:45:59PM +0200, Peter Zijlstra wrote:

quoted

On Fri, Oct 08, 2021 at 01:40:52PM +0100, Mark Rutland wrote:

quoted

[Adding Josh, since there might be a concern here from a livepatch pov]

quoted

quoted

+static unsigned long __get_wchan(struct task_struct *p)
+{
+	unsigned long entry = 0;
+
+	stack_trace_save_tsk(p, &entry, 1, 0);

This assumes stack_trace_save_tsk() will skip sched functions, but I
don't think that's ever been a requirement? It's certinaly not
documented anywhere that I could find, and arm64 doesn't do so today,
and this patch causes wchan to just log `__switch_to` for everything.

Confused, arm64 has arch_stack_walk() and should thus use
kernel/stacktrace.c's stack_trace_consume_entry_nosched.

Looking at this arm64's *current* get_wchan() unwinds once before
checking in_sched_functions(), so it skips __switch_to(). As of this
patch, we check in_sched_functions() first, which stops the unwind
immediately as __switch_to() isn't marked as __sched.

I think x86 gets away with this because switch_to() is asm, and that
tail-calls __switch_to() when returning.

Does switch_to() and below need to be marked __sched?

Yes, I would think so, for arches where they otherwise show up on the
stacktrace.

-- 
Josh