Re: kernel_thread() & thread starting

[Philipp Rumpf]

On Mon, 19 Feb 2001, Russell King wrote:
> Philipp Rumpf writes:
> > That still won't catch keventd oopsing though - which I think might happen
> > quite easily in real life.
> 
> Maybe we should panic in that case?  For example, what happens if kswapd
> oopses?  kreclaimd?  bdflush?  kupdate?  All these have the same problem,

No.  If kswapd oopses it's a bug in kswapd (or related code).  If keventd
oopses most likely the broken code is actually the task queue you
scheduled, which belongs to your driver.