On Tue, Jun 30, 2015 at 11:55:59AM +0100, Peter Maydell wrote: > On 30 June 2015 at 11:50, Zhu Guihua <zhugh.f...@cn.fujitsu.com> wrote: > > Originally, there are some devices (such as hpet, rtc) reset before apic > > reset. > > When these devices reset, they would send irq to apic. > > In our current (arguably broken) reset design, this is a bug in > those devices -- they shouldn't be raising interrupts from their > reset functions.
Do we know exactly which devices are currently broken and need to be fixed? Should we add a warning to the interrupt code to detect those cases? It looks like we won't need a solution to ensure reset ordering if we fix the devices (or at least the solution would be just temporary until we fix them). -- Eduardo
