On Thu, 2021-11-11 at 10:01 +0100, David Hildenbrand wrote:
> On 10.11.21 21:45, Eric Farman wrote:
> > With the USER_SIGP capability, the kernel will pass most (but not
> > all)
> > SIGP orders to userspace for processing. But that means that the
> > kernel
> > is unable to determine if/when the o
On 10.11.21 21:45, Eric Farman wrote:
> With the USER_SIGP capability, the kernel will pass most (but not all)
> SIGP orders to userspace for processing. But that means that the kernel
> is unable to determine if/when the order has been completed by userspace,
> and could potentially return an inco
With the USER_SIGP capability, the kernel will pass most (but not all)
SIGP orders to userspace for processing. But that means that the kernel
is unable to determine if/when the order has been completed by userspace,
and could potentially return an incorrect answer (CC1 with status bits
versus CC2