Re: drivers for l4-hurd

[Peter 'p2' De Schrijver]

Hi,

> 
> We introduced that scheme for three main reasons. 
> First, exporting the IRQ controller to user level allows to disable
> timer interrupts which could halt the system. One could argue that
> device drivers are part of the trusted computing base, but that is not
> always true. For example a serial line device driver does not
> necessarily have access to DMA and can therefore completely
> encapsulated.
> Second, on many systems (not IA32) interrupt controllers cannot be
> accessed from user level. Hence, the kernel must take care of the IRQ

On which systems would this be ? I can think of the IBM PPC 405 which
uses the DCB for the interrupt controller using the mtdcr and mfdcr
which are supervisor mode only. On most other systems at least the
second and lower level interrupt controllers can be memory mapped to
user space.

All in all, this doesn't seem to have much impact on the idea of putting
all IRQ threads in 1 AS. I still think this is a good idea as some
devices really need low interrupt latency and it allows for interrupt
mitigation schemes by disabling the IRQ and going to a polling mode.
This can be useful for gige cards and other high bandwidth interfaces.

Cheers,

Peter.