Hi Andre,
TBH, I would have expected this patch to be split in two:
- Introduction of the flag before patch #19
- Set flag in patch #19
This would have make easier to review the implementation of MAPTI.
On 12/04/17 01:44, Andre Przywara wrote:
When LPIs get unmapped by a guest,
Hi,
On 12/04/17 01:44, Andre Przywara wrote:
> When LPIs get unmapped by a guest, they might still be in some LR of
> some VCPU. Nevertheless we remove the corresponding pending_irq
> (possibly freeing it), and detect this case (irq_to_pending() returns
> NULL) when the LR gets cleaned up later.
>
When LPIs get unmapped by a guest, they might still be in some LR of
some VCPU. Nevertheless we remove the corresponding pending_irq
(possibly freeing it), and detect this case (irq_to_pending() returns
NULL) when the LR gets cleaned up later.
However a *new* LPI may get mapped with the same number
