Initial support for CLIC

[bjoernQ]

This adds support for interrupts on ESP32-P4

You can test it with this (you need to add the critical-section dependency)

#![no_std]
#![no_main]

use core::cell::RefCell;

use critical_section::Mutex;
use esp32p4_hal::{
    gpio::{Gpio35, Input, PullDown, IO},
    interrupt,
    peripherals,
    peripherals::Peripherals,
    prelude::*,
};
use esp_backtrace as _;
use esp_println::println;

static BUTTON: Mutex<RefCell<Option<Gpio35<Input<PullDown>>>>> = Mutex::new(RefCell::new(None));

#[entry]
fn main() -> ! {
    esp_println::logger::init_logger_from_env();
    let peripherals = Peripherals::take();

    let io = IO::new(peripherals.GPIO, peripherals.IO_MUX);
    let mut led = io.pins.gpio36.into_push_pull_output();
    let mut button = io.pins.gpio35.into_pull_down_input();

    button.listen(esp32p4_hal::gpio::Event::FallingEdge);

    critical_section::with(|cs| BUTTON.borrow_ref_mut(cs).replace(button));
    interrupt::enable(
        peripherals::Interrupt::GPIO_INT0,
        interrupt::Priority::Priority3,
    )
    .unwrap();

    led.set_high().unwrap();
    loop {
        led.toggle().unwrap();

        #[cfg(debug_assertions)]
        let count = 150000;

        #[cfg(not(debug_assertions))]
        let count = 3500000;

        for _ in 0..count {
            unsafe {
                core::arch::asm!("nop");
            }
        }
        println!("blink!");
    }
}

#[interrupt]
fn GPIO_INT0() {
    critical_section::with(|cs| {
        println!("GPIO interrupt");
        BUTTON
            .borrow_ref_mut(cs)
            .as_mut()
            .unwrap()
            .clear_interrupt();
    });
}

This blinks an LED on GPIO36 and handles an interrupt when using the boot button.

Only external external peripheral interrupts are supported.

Doesn't support "interrupt-preemption" and "direct-vectoring" features - there is apparently support for pre-emption by CLIC and a different way to do direct vectored interrupts. We can explore that later.

Probably the naming of "very_large_intr_status" and "large_intr_status" could be improved. I don't expect to have "extremely_large_intr_status" in future but probably we can come up with better naming in future. (Out of scope of this PR)

Not tested with the second core and anything but the GPIO interrupt for obvious reasons.

[MabezDev]

Awesome! Mostly looks good, I'd just like to see a comment or two explaining some of the operations.

[MabezDev]

Thanks!