Rebase, add commit message

Rebase, clarify commit message

Use an option instead

In D34750#790841, @mhorne wrote:

I guess this commit gives a little more context. So because the device tree declares &intc to be the interrupt parent for several peripherals, we need this stub driver in order to properly route things back to the PLIC, is that right?

Correct pass level, use ofw_compat_data table

In D34749#787764, @mhorne wrote:

The driver changes look good to me although I am not the authority there. As mentioned please drop the diff to machdep.h.

Check out 'Edit Related Revisions' on the right hand side.

Update the commit message, and remove unused 'res'

Drop machdep.h

In D34748#787522, @jrtc27 wrote:

In D34748#787517, @julien.cassette_gmail.com wrote:

In D34748#787493, @jrtc27 wrote:

Why do we need this? Device trees should come from the firmware in this day and age...

Hello Jessica.
Since this is done for other devices, I guessed this was needed for some cases (e.g. maybe the firmware does not always provide the exact DTB we expect?), or even just for documentation purposes.
Regards, Julien.

This was added for the SiFive Unleashed because initially there was no proper firmware, U-Boot came later and instead you just had to direct boot a kernel. There is no need for it any more even there.

In D34750#787496, @jrtc27 wrote:

Don't see why the PLIC changes are needed, but if they are they should be separated out from the addition of the D1 interrupt controller.

In D34749#787499, @jrtc27 wrote:

I don't see any reason why this needs to be part of the revision stack, it isn't dependent on anything else

In D34748#787493, @jrtc27 wrote:

Why do we need this? Device trees should come from the firmware in this day and age...