Merge pull request #1208 from embassy-rs/nrf-uarte-lockfree

nrf/buffered_uarte: make it work without rts/cts, and lock-free.

16 files changed, 1096 insertions, 365 deletions

diff --git a/ci.sh b/ci.sh
index 417937d07..bbcb26bdb 100755
--- a/ci.sh
+++ b/ci.sh
@@ -133,6 +133,7 @@ cargo batch  \
     --- build --release --manifest-path tests/stm32/Cargo.toml --target thumbv7em-none-eabi --features stm32wb55rg --out-dir out/tests/nucleo-stm32wb55rg \
     --- build --release --manifest-path tests/stm32/Cargo.toml --target thumbv7em-none-eabi --features stm32u585ai --out-dir out/tests/iot-stm32u585ai \
     --- build --release --manifest-path tests/rp/Cargo.toml --target thumbv6m-none-eabi --out-dir out/tests/rpi-pico \
+    --- build --release --manifest-path tests/nrf/Cargo.toml --target thumbv7em-none-eabi --out-dir out/tests/nrf52840-dk \
     $BUILD_EXTRA
 
 
diff --git a/embassy-nrf/src/buffered_uarte.rs b/embassy-nrf/src/buffered_uarte.rs
index 112f084c1..ab639aeea 100644
--- a/embassy-nrf/src/buffered_uarte.rs
+++ b/embassy-nrf/src/buffered_uarte.rs
@@ -1,10 +1,5 @@
 //! Async buffered UART driver.
 //!
-//! WARNING!!! The functionality provided here is intended to be used only
-//! in situations where hardware flow control are available i.e. CTS and RTS.
-//! This is a problem that should be addressed at a later stage and can be
-//! fully explained at <https://github.com/embassy-rs/embassy/issues/536>.
-//!
 //! Note that discarding a future from a read or write operation may lead to losing
 //! data. For example, when using `futures_util::future::select` and completion occurs
 //! on the "other" future, you should capture the incomplete future and continue to use
@@ -13,82 +8,128 @@
 //!
 //! Please also see [crate::uarte] to understand when [BufferedUarte] should be used.
 
-use core::cell::RefCell;
 use core::cmp::min;
 use core::future::poll_fn;
-use core::sync::atomic::{compiler_fence, Ordering};
+use core::slice;
+use core::sync::atomic::{compiler_fence, AtomicU8, AtomicUsize, Ordering};
 use core::task::Poll;
 
-use embassy_cortex_m::peripheral::{PeripheralMutex, PeripheralState, StateStorage};
-use embassy_hal_common::ring_buffer::RingBuffer;
+use embassy_cortex_m::interrupt::Interrupt;
+use embassy_hal_common::atomic_ring_buffer::RingBuffer;
 use embassy_hal_common::{into_ref, PeripheralRef};
-use embassy_sync::waitqueue::WakerRegistration;
+use embassy_sync::waitqueue::AtomicWaker;
 // Re-export SVD variants to allow user to directly set values
 pub use pac::uarte0::{baudrate::BAUDRATE_A as Baudrate, config::PARITY_A as Parity};
 
-use crate::gpio::{self, Pin as GpioPin};
+use crate::gpio::sealed::Pin;
+use crate::gpio::{self, AnyPin, Pin as GpioPin, PselBits};
 use crate::interrupt::InterruptExt;
-use crate::ppi::{AnyConfigurableChannel, ConfigurableChannel, Event, Ppi, Task};
-use crate::timer::{Frequency, Instance as TimerInstance, Timer};
+use crate::ppi::{
+    self, AnyConfigurableChannel, AnyGroup, Channel, ConfigurableChannel, Event, Group, Ppi, PpiGroup, Task,
+};
+use crate::timer::{Instance as TimerInstance, Timer};
 use crate::uarte::{apply_workaround_for_enable_anomaly, Config, Instance as UarteInstance};
 use crate::{pac, Peripheral};
 
-#[derive(Copy, Clone, Debug, PartialEq)]
-enum RxState {
-    Idle,
-    Receiving,
-}
+mod sealed {
+    use super::*;
 
-#[derive(Copy, Clone, Debug, PartialEq)]
-enum TxState {
-    Idle,
-    Transmitting(usize),
-}
+    pub struct State {
+        pub tx_waker: AtomicWaker,
+        pub tx_buf: RingBuffer,
+        pub tx_count: AtomicUsize,
 
-/// A type for storing the state of the UARTE peripheral that can be stored in a static.
-pub struct State<'d, U: UarteInstance, T: TimerInstance>(StateStorage<StateInner<'d, U, T>>);
-impl<'d, U: UarteInstance, T: TimerInstance> State<'d, U, T> {
-    /// Create an instance for storing UARTE peripheral state.
-    pub fn new() -> Self {
-        Self(StateStorage::new())
+        pub rx_waker: AtomicWaker,
+        pub rx_buf: RingBuffer,
+        pub rx_bufs: AtomicU8,
+        pub rx_ppi_ch: AtomicU8,
     }
 }
 
-struct StateInner<'d, U: UarteInstance, T: TimerInstance> {
-    _peri: PeripheralRef<'d, U>,
-    timer: Timer<'d, T>,
-    _ppi_ch1: Ppi<'d, AnyConfigurableChannel, 1, 2>,
-    _ppi_ch2: Ppi<'d, AnyConfigurableChannel, 1, 1>,
+/// UART error.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+#[cfg_attr(feature = "defmt", derive(defmt::Format))]
+#[non_exhaustive]
+pub enum Error {
+    // No errors for now
+}
 
-    rx: RingBuffer<'d>,
-    rx_state: RxState,
-    rx_waker: WakerRegistration,
+pub(crate) use sealed::State;
 
-    tx: RingBuffer<'d>,
-    tx_state: TxState,
-    tx_waker: WakerRegistration,
+impl State {
+    pub(crate) const fn new() -> Self {
+        Self {
+            tx_waker: AtomicWaker::new(),
+            tx_buf: RingBuffer::new(),
+            tx_count: AtomicUsize::new(0),
+
+            rx_waker: AtomicWaker::new(),
+            rx_buf: RingBuffer::new(),
+            rx_bufs: AtomicU8::new(0),
+            rx_ppi_ch: AtomicU8::new(0),
+        }
+    }
 }
 
 /// Buffered UARTE driver.
 pub struct BufferedUarte<'d, U: UarteInstance, T: TimerInstance> {
-    inner: RefCell<PeripheralMutex<'d, StateInner<'d, U, T>>>,
+    _peri: PeripheralRef<'d, U>,
+    timer: Timer<'d, T>,
+    _ppi_ch1: Ppi<'d, AnyConfigurableChannel, 1, 1>,
+    _ppi_ch2: Ppi<'d, AnyConfigurableChannel, 1, 2>,
+    _ppi_group: PpiGroup<'d, AnyGroup>,
 }
 
 impl<'d, U: UarteInstance, T: TimerInstance> Unpin for BufferedUarte<'d, U, T> {}
 
 impl<'d, U: UarteInstance, T: TimerInstance> BufferedUarte<'d, U, T> {
-    /// Create a new instance of a BufferedUarte.
+    /// Create a new BufferedUarte without hardware flow control.
     ///
-    /// See the [module documentation](crate::buffered_uarte) for more details about the intended use.
+    /// # Panics
     ///
-    /// The BufferedUarte uses the provided state to store the buffers and peripheral state. The timer and ppi channels are used to 'emulate' idle line detection so that read operations
-    /// can return early if there is no data to receive.
+    /// Panics if `rx_buffer.len()` is odd.
     pub fn new(
-        state: &'d mut State<'d, U, T>,
-        peri: impl Peripheral<P = U> + 'd,
+        uarte: impl Peripheral<P = U> + 'd,
+        timer: impl Peripheral<P = T> + 'd,
+        ppi_ch1: impl Peripheral<P = impl ConfigurableChannel> + 'd,
+        ppi_ch2: impl Peripheral<P = impl ConfigurableChannel> + 'd,
+        ppi_group: impl Peripheral<P = impl Group> + 'd,
+        irq: impl Peripheral<P = U::Interrupt> + 'd,
+        rxd: impl Peripheral<P = impl GpioPin> + 'd,
+        txd: impl Peripheral<P = impl GpioPin> + 'd,
+        config: Config,
+        rx_buffer: &'d mut [u8],
+        tx_buffer: &'d mut [u8],
+    ) -> Self {
+        into_ref!(rxd, txd, ppi_ch1, ppi_ch2, ppi_group);
+        Self::new_inner(
+            uarte,
+            timer,
+            ppi_ch1.map_into(),
+            ppi_ch2.map_into(),
+            ppi_group.map_into(),
+            irq,
+            rxd.map_into(),
+            txd.map_into(),
+            None,
+            None,
+            config,
+            rx_buffer,
+            tx_buffer,
+        )
+    }
+
+    /// Create a new BufferedUarte with hardware flow control (RTS/CTS)
+    ///
+    /// # Panics
+    ///
+    /// Panics if `rx_buffer.len()` is odd.
+    pub fn new_with_rtscts(
+        uarte: impl Peripheral<P = U> + 'd,
         timer: impl Peripheral<P = T> + 'd,
-        ppi_ch1: impl Peripheral<P = impl ConfigurableChannel + 'd> + 'd,
-        ppi_ch2: impl Peripheral<P = impl ConfigurableChannel + 'd> + 'd,
+        ppi_ch1: impl Peripheral<P = impl ConfigurableChannel> + 'd,
+        ppi_ch2: impl Peripheral<P = impl ConfigurableChannel> + 'd,
+        ppi_group: impl Peripheral<P = impl Group> + 'd,
         irq: impl Peripheral<P = U::Interrupt> + 'd,
         rxd: impl Peripheral<P = impl GpioPin> + 'd,
         txd: impl Peripheral<P = impl GpioPin> + 'd,
@@ -98,11 +139,44 @@ impl<'d, U: UarteInstance, T: TimerInstance> BufferedUarte<'d, U, T> {
         rx_buffer: &'d mut [u8],
         tx_buffer: &'d mut [u8],
     ) -> Self {
-        into_ref!(peri, ppi_ch1, ppi_ch2, irq, rxd, txd, cts, rts);
+        into_ref!(rxd, txd, cts, rts, ppi_ch1, ppi_ch2, ppi_group);
+        Self::new_inner(
+            uarte,
+            timer,
+            ppi_ch1.map_into(),
+            ppi_ch2.map_into(),
+            ppi_group.map_into(),
+            irq,
+            rxd.map_into(),
+            txd.map_into(),
+            Some(cts.map_into()),
+            Some(rts.map_into()),
+            config,
+            rx_buffer,
+            tx_buffer,
+        )
+    }
 
-        let r = U::regs();
+    fn new_inner(
+        peri: impl Peripheral<P = U> + 'd,
+        timer: impl Peripheral<P = T> + 'd,
+        ppi_ch1: PeripheralRef<'d, AnyConfigurableChannel>,
+        ppi_ch2: PeripheralRef<'d, AnyConfigurableChannel>,
+        ppi_group: PeripheralRef<'d, AnyGroup>,
+        irq: impl Peripheral<P = U::Interrupt> + 'd,
+        rxd: PeripheralRef<'d, AnyPin>,
+        txd: PeripheralRef<'d, AnyPin>,
+        cts: Option<PeripheralRef<'d, AnyPin>>,
+        rts: Option<PeripheralRef<'d, AnyPin>>,
+        config: Config,
+        rx_buffer: &'d mut [u8],
+        tx_buffer: &'d mut [u8],
+    ) -> Self {
+        into_ref!(peri, timer, irq);
+
+        assert!(rx_buffer.len() % 2 == 0);
 
-        let mut timer = Timer::new(timer);
+        let r = U::regs();
 
         rxd.conf().write(|w| w.input().connect().drive().h0h1());
         r.psel.rxd.write(|w| unsafe { w.bits(rxd.psel_bits()) });
@@ -111,92 +185,200 @@ impl<'d, U: UarteInstance, T: TimerInstance> BufferedUarte<'d, U, T> {
         txd.conf().write(|w| w.dir().output().drive().h0h1());
         r.psel.txd.write(|w| unsafe { w.bits(txd.psel_bits()) });
 
-        cts.conf().write(|w| w.input().connect().drive().h0h1());
+        if let Some(pin) = &cts {
+            pin.conf().write(|w| w.input().connect().drive().h0h1());
+        }
         r.psel.cts.write(|w| unsafe { w.bits(cts.psel_bits()) });
 
-        rts.set_high();
-        rts.conf().write(|w| w.dir().output().drive().h0h1());
+        if let Some(pin) = &rts {
+            pin.set_high();
+            pin.conf().write(|w| w.dir().output().drive().h0h1());
+        }
         r.psel.rts.write(|w| unsafe { w.bits(rts.psel_bits()) });
 
-        r.baudrate.write(|w| w.baudrate().variant(config.baudrate));
-        r.config.write(|w| w.parity().variant(config.parity));
+        // Initialize state
+        let s = U::buffered_state();
+        s.tx_count.store(0, Ordering::Relaxed);
+        s.rx_bufs.store(0, Ordering::Relaxed);
+        let len = tx_buffer.len();
+        unsafe { s.tx_buf.init(tx_buffer.as_mut_ptr(), len) };
+        let len = rx_buffer.len();
+        unsafe { s.rx_buf.init(rx_buffer.as_mut_ptr(), len) };
 
         // Configure
         r.config.write(|w| {
-            w.hwfc().bit(true);
+            w.hwfc().bit(false);
             w.parity().variant(config.parity);
             w
         });
         r.baudrate.write(|w| w.baudrate().variant(config.baudrate));
 
-        // Enable interrupts
-        r.intenset.write(|w| w.endrx().set().endtx().set());
+        // clear errors
+        let errors = r.errorsrc.read().bits();
+        r.errorsrc.write(|w| unsafe { w.bits(errors) });
 
-        // Disable the irq, let the Registration enable it when everything is set up.
-        irq.disable();
-        irq.pend();
+        r.events_rxstarted.reset();
+        r.events_txstarted.reset();
+        r.events_error.reset();
+        r.events_endrx.reset();
+        r.events_endtx.reset();
+
+        // Enable interrupts
+        r.intenclr.write(|w| unsafe { w.bits(!0) });
+        r.intenset.write(|w| {
+            w.endtx().set();
+            w.rxstarted().set();
+            w.error().set();
+            w
+        });
 
         // Enable UARTE instance
         apply_workaround_for_enable_anomaly(&r);
         r.enable.write(|w| w.enable().enabled());
 
-        // BAUDRATE register values are `baudrate * 2^32 / 16000000`
-        // source: https://devzone.nordicsemi.com/f/nordic-q-a/391/uart-baudrate-register-values
-        //
-        // We want to stop RX if line is idle for 2 bytes worth of time
-        // That is 20 bits (each byte is 1 start bit + 8 data bits + 1 stop bit)
-        // This gives us the amount of 16M ticks for 20 bits.
-        let timeout = 0x8000_0000 / (config.baudrate as u32 / 40);
+        // Configure byte counter.
+        let mut timer = Timer::new_counter(timer);
+        timer.cc(1).write(rx_buffer.len() as u32 * 2);
+        timer.cc(1).short_compare_clear();
+        timer.clear();
+        timer.start();
 
-        timer.set_frequency(Frequency::F16MHz);
-        timer.cc(0).write(timeout);
-        timer.cc(0).short_compare_clear();
-        timer.cc(0).short_compare_stop();
-
-        let mut ppi_ch1 = Ppi::new_one_to_two(
-            ppi_ch1.map_into(),
-            Event::from_reg(&r.events_rxdrdy),
-            timer.task_clear(),
-            timer.task_start(),
-        );
+        let mut ppi_ch1 = Ppi::new_one_to_one(ppi_ch1, Event::from_reg(&r.events_rxdrdy), timer.task_count());
         ppi_ch1.enable();
 
-        let mut ppi_ch2 = Ppi::new_one_to_one(
-            ppi_ch2.map_into(),
-            timer.cc(0).event_compare(),
-            Task::from_reg(&r.tasks_stoprx),
+        s.rx_ppi_ch.store(ppi_ch2.number() as u8, Ordering::Relaxed);
+        let mut ppi_group = PpiGroup::new(ppi_group);
+        let mut ppi_ch2 = Ppi::new_one_to_two(
+            ppi_ch2,
+            Event::from_reg(&r.events_endrx),
+            Task::from_reg(&r.tasks_startrx),
+            ppi_group.task_disable_all(),
         );
-        ppi_ch2.enable();
+        ppi_ch2.disable();
+        ppi_group.add_channel(&ppi_ch2);
+
+        irq.disable();
+        irq.set_handler(Self::on_interrupt);
+        irq.pend();
+        irq.enable();
 
         Self {
-            inner: RefCell::new(PeripheralMutex::new(irq, &mut state.0, move || StateInner {
-                _peri: peri,
-                timer,
-                _ppi_ch1: ppi_ch1,
-                _ppi_ch2: ppi_ch2,
-
-                rx: RingBuffer::new(rx_buffer),
-                rx_state: RxState::Idle,
-                rx_waker: WakerRegistration::new(),
-
-                tx: RingBuffer::new(tx_buffer),
-                tx_state: TxState::Idle,
-                tx_waker: WakerRegistration::new(),
-            })),
+            _peri: peri,
+            timer,
+            _ppi_ch1: ppi_ch1,
+            _ppi_ch2: ppi_ch2,
+            _ppi_group: ppi_group,
         }
     }
 
-    /// Adjust the baud rate to the provided value.
-    pub fn set_baudrate(&mut self, baudrate: Baudrate) {
-        self.inner.borrow_mut().with(|state| {
-            let r = U::regs();
+    fn pend_irq() {
+        unsafe { <U::Interrupt as Interrupt>::steal() }.pend()
+    }
+
+    fn on_interrupt(_: *mut ()) {
+        //trace!("irq: start");
+        let r = U::regs();
+        let s = U::buffered_state();
 
-            let timeout = 0x8000_0000 / (baudrate as u32 / 40);
-            state.timer.cc(0).write(timeout);
-            state.timer.clear();
+        let buf_len = s.rx_buf.len();
+        let half_len = buf_len / 2;
+        let mut tx = unsafe { s.tx_buf.reader() };
+        let mut rx = unsafe { s.rx_buf.writer() };
 
-            r.baudrate.write(|w| w.baudrate().variant(baudrate));
-        });
+        if r.events_error.read().bits() != 0 {
+            r.events_error.reset();
+            let errs = r.errorsrc.read();
+            r.errorsrc.write(|w| unsafe { w.bits(errs.bits()) });
+
+            if errs.overrun().bit() {
+                panic!("BufferedUarte overrun");
+            }
+        }
+
+        // Received some bytes, wake task.
+        if r.inten.read().rxdrdy().bit_is_set() && r.events_rxdrdy.read().bits() != 0 {
+            r.intenclr.write(|w| w.rxdrdy().clear());
+            r.events_rxdrdy.reset();
+            s.rx_waker.wake();
+        }
+
+        // If not RXing, start.
+        if s.rx_bufs.load(Ordering::Relaxed) == 0 {
+            let (ptr, len) = rx.push_buf();
+            if len >= half_len {
+                //trace!("  irq_rx: starting {:?}", half_len);
+                s.rx_bufs.store(1, Ordering::Relaxed);
+
+                // Set up the DMA read
+                r.rxd.ptr.write(|w| unsafe { w.ptr().bits(ptr as u32) });
+                r.rxd.maxcnt.write(|w| unsafe { w.maxcnt().bits(half_len as _) });
+
+                // Start UARTE Receive transaction
+                r.tasks_startrx.write(|w| unsafe { w.bits(1) });
+                rx.push_done(half_len);
+                r.intenset.write(|w| w.rxstarted().set());
+            }
+        }
+
+        if r.events_rxstarted.read().bits() != 0 {
+            //trace!("  irq_rx: rxstarted");
+            let (ptr, len) = rx.push_buf();
+            if len >= half_len {
+                //trace!("  irq_rx: starting second {:?}", half_len);
+
+                // Set up the DMA read
+                r.rxd.ptr.write(|w| unsafe { w.ptr().bits(ptr as u32) });
+                r.rxd.maxcnt.write(|w| unsafe { w.maxcnt().bits(half_len as _) });
+
+                let chn = s.rx_ppi_ch.load(Ordering::Relaxed);
+
+                ppi::regs().chenset.write(|w| unsafe { w.bits(1 << chn) });
+
+                rx.push_done(half_len);
+
+                r.events_rxstarted.reset();
+            } else {
+                //trace!("  irq_rx: rxstarted no buf");
+                r.intenclr.write(|w| w.rxstarted().clear());
+            }
+        }
+
+        // =============================
+
+        // TX end
+        if r.events_endtx.read().bits() != 0 {
+            r.events_endtx.reset();
+
+            let n = s.tx_count.load(Ordering::Relaxed);
+            //trace!("  irq_tx: endtx {:?}", n);
+            tx.pop_done(n);
+            s.tx_waker.wake();
+            s.tx_count.store(0, Ordering::Relaxed);
+        }
+
+        // If not TXing, start.
+        if s.tx_count.load(Ordering::Relaxed) == 0 {
+            let (ptr, len) = tx.pop_buf();
+            if len != 0 {
+                //trace!("  irq_tx: starting {:?}", len);
+                s.tx_count.store(len, Ordering::Relaxed);
+
+                // Set up the DMA write
+                r.txd.ptr.write(|w| unsafe { w.ptr().bits(ptr as u32) });
+                r.txd.maxcnt.write(|w| unsafe { w.maxcnt().bits(len as _) });
+
+                // Start UARTE Transmit transaction
+                r.tasks_starttx.write(|w| unsafe { w.bits(1) });
+            }
+        }
+
+        //trace!("irq: end");
+    }
+
+    /// Adjust the baud rate to the provided value.
+    pub fn set_baudrate(&mut self, baudrate: Baudrate) {
+        let r = U::regs();
+        r.baudrate.write(|w| w.baudrate().variant(baudrate));
     }
 
     /// Split the UART in reader and writer parts.
@@ -206,120 +388,142 @@ impl<'d, U: UarteInstance, T: TimerInstance> BufferedUarte<'d, U, T> {
         (BufferedUarteRx { inner: self }, BufferedUarteTx { inner: self })
     }
 
-    async fn inner_read<'a>(&'a self, buf: &'a mut [u8]) -> Result<usize, core::convert::Infallible> {
+    async fn inner_read(&self, buf: &mut [u8]) -> Result<usize, Error> {
+        let data = self.inner_fill_buf().await?;
+        let n = data.len().min(buf.len());
+        buf[..n].copy_from_slice(&data[..n]);
+        self.inner_consume(n);
+        Ok(n)
+    }
+
+    async fn inner_write<'a>(&'a self, buf: &'a [u8]) -> Result<usize, Error> {
         poll_fn(move |cx| {
-            let mut do_pend = false;
-            let mut inner = self.inner.borrow_mut();
-            let res = inner.with(|state| {
-                compiler_fence(Ordering::SeqCst);
-                trace!("poll_read");
-
-                // We have data ready in buffer? Return it.
-                let data = state.rx.pop_buf();
-                if !data.is_empty() {
-                    trace!("  got {:?} {:?}", data.as_ptr() as u32, data.len());
-                    let len = data.len().min(buf.len());
-                    buf[..len].copy_from_slice(&data[..len]);
-                    state.rx.pop(len);
-                    do_pend = true;
-                    return Poll::Ready(Ok(len));
-                }
-
-                trace!("  empty");
-                state.rx_waker.register(cx.waker());
-                Poll::Pending
-            });
-            if do_pend {
-                inner.pend();
+            //trace!("poll_write: {:?}", buf.len());
+            let s = U::buffered_state();
+            let mut tx = unsafe { s.tx_buf.writer() };
+
+            let tx_buf = tx.push_slice();
+            if tx_buf.is_empty() {
+                //trace!("poll_write: pending");
+                s.tx_waker.register(cx.waker());
+                return Poll::Pending;
             }
 
-            res
+            let n = min(tx_buf.len(), buf.len());
+            tx_buf[..n].copy_from_slice(&buf[..n]);
+            tx.push_done(n);
+
+            //trace!("poll_write: queued {:?}", n);
+
+            compiler_fence(Ordering::SeqCst);
+            Self::pend_irq();
+
+            Poll::Ready(Ok(n))
         })
         .await
     }
 
-    async fn inner_write<'a>(&'a self, buf: &'a [u8]) -> Result<usize, core::convert::Infallible> {
+    async fn inner_flush<'a>(&'a self) -> Result<(), Error> {
         poll_fn(move |cx| {
-            let mut inner = self.inner.borrow_mut();
-            let res = inner.with(|state| {
-                trace!("poll_write: {:?}", buf.len());
+            //trace!("poll_flush");
+            let s = U::buffered_state();
+            if !s.tx_buf.is_empty() {
+                //trace!("poll_flush: pending");
+                s.tx_waker.register(cx.waker());
+                return Poll::Pending;
+            }
 
-                let tx_buf = state.tx.push_buf();
-                if tx_buf.is_empty() {
-                    trace!("poll_write: pending");
-                    state.tx_waker.register(cx.waker());
-                    return Poll::Pending;
-                }
+            Poll::Ready(Ok(()))
+        })
+        .await
+    }
 
-                let n = min(tx_buf.len(), buf.len());
-                tx_buf[..n].copy_from_slice(&buf[..n]);
-                state.tx.push(n);
+    async fn inner_fill_buf<'a>(&'a self) -> Result<&'a [u8], Error> {
+        poll_fn(move |cx| {
+            compiler_fence(Ordering::SeqCst);
+            //trace!("poll_read");
 
-                trace!("poll_write: queued {:?}", n);
+            let r = U::regs();
+            let s = U::buffered_state();
+
+            // Read the RXDRDY counter.
+            T::regs().tasks_capture[0].write(|w| unsafe { w.bits(1) });
+            let mut end = T::regs().cc[0].read().bits() as usize;
+            //trace!("  rxdrdy count = {:?}", end);
+
+            // We've set a compare channel that resets the counter to 0 when it reaches `len*2`.
+            // However, it's unclear if that's instant, or there's a small window where you can
+            // still read `len()*2`.
+            // This could happen if in one clock cycle the counter is updated, and in the next the
+            // clear takes effect. The docs are very sparse, they just say "Task delays: After TIMER
+            // is started, the CLEAR, COUNT, and STOP tasks are guaranteed to take effect within one
+            // clock cycle of the PCLK16M." :shrug:
+            // So, we wrap the counter ourselves, just in case.
+            if end > s.rx_buf.len() * 2 {
+                end = 0
+            }
 
-                compiler_fence(Ordering::SeqCst);
+            // This logic mirrors `atomic_ring_buffer::Reader::pop_buf()`
+            let mut start = s.rx_buf.start.load(Ordering::Relaxed);
+            let len = s.rx_buf.len();
+            if start == end {
+                //trace!("  empty");
+                s.rx_waker.register(cx.waker());
+                r.intenset.write(|w| w.rxdrdy().set_bit());
+                return Poll::Pending;
+            }
 
-                Poll::Ready(Ok(n))
-            });
+            if start >= len {
+                start -= len
+            }
+            if end >= len {
+                end -= len
+            }
 
-            inner.pend();
+            let n = if end > start { end - start } else { len - start };
+            assert!(n != 0);
+            //trace!("  uarte ringbuf: pop_buf {:?}..{:?}", start, start + n);
 
-            res
+            let buf = s.rx_buf.buf.load(Ordering::Relaxed);
+            Poll::Ready(Ok(unsafe { slice::from_raw_parts(buf.add(start), n) }))
         })
         .await
     }
 
-    async fn inner_flush<'a>(&'a self) -> Result<(), core::convert::Infallible> {
-        poll_fn(move |cx| {
-            self.inner.borrow_mut().with(|state| {
-                trace!("poll_flush");
+    fn inner_consume(&self, amt: usize) {
+        if amt == 0 {
+            return;
+        }
 
-                if !state.tx.is_empty() {
-                    trace!("poll_flush: pending");
-                    state.tx_waker.register(cx.waker());
-                    return Poll::Pending;
-                }
+        let s = U::buffered_state();
+        let mut rx = unsafe { s.rx_buf.reader() };
+        rx.pop_done(amt);
+        U::regs().intenset.write(|w| w.rxstarted().set());
+    }
 
-                Poll::Ready(Ok(()))
-            })
-        })
-        .await
+    /// Pull some bytes from this source into the specified buffer, returning how many bytes were read.
+    pub async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Error> {
+        self.inner_read(buf).await
     }
 
-    async fn inner_fill_buf<'a>(&'a self) -> Result<&'a [u8], core::convert::Infallible> {
-        poll_fn(move |cx| {
-            self.inner.borrow_mut().with(|state| {
-                compiler_fence(Ordering::SeqCst);
-                trace!("fill_buf");
-
-                // We have data ready in buffer? Return it.
-                let buf = state.rx.pop_buf();
-                if !buf.is_empty() {
-                    trace!("  got {:?} {:?}", buf.as_ptr() as u32, buf.len());
-                    let buf: &[u8] = buf;
-                    // Safety: buffer lives as long as uart
-                    let buf: &[u8] = unsafe { core::mem::transmute(buf) };
-                    return Poll::Ready(Ok(buf));
-                }
-
-                trace!("  empty");
-                state.rx_waker.register(cx.waker());
-                Poll::<Result<&[u8], core::convert::Infallible>>::Pending
-            })
-        })
-        .await
+    /// Return the contents of the internal buffer, filling it with more data from the inner reader if it is empty.
+    pub async fn fill_buf(&mut self) -> Result<&[u8], Error> {
+        self.inner_fill_buf().await
     }
 
-    fn inner_consume(&self, amt: usize) {
-        let mut inner = self.inner.borrow_mut();
-        let signal = inner.with(|state| {
-            let full = state.rx.is_full();
-            state.rx.pop(amt);
-            full
-        });
-        if signal {
-            inner.pend();
-        }
+    /// Tell this buffer that `amt` bytes have been consumed from the buffer, so they should no longer be returned in calls to `fill_buf`.
+    pub fn consume(&mut self, amt: usize) {
+        self.inner_consume(amt)
+    }
+
+    /// Write a buffer into this writer, returning how many bytes were written.
+    pub async fn write(&mut self, buf: &[u8]) -> Result<usize, Error> {
+        self.inner_write(buf).await
+    }
+
+    /// Flush this output stream, ensuring that all intermediately buffered contents reach their destination.
+    pub async fn flush(&mut self) -> Result<(), Error> {
+        self.inner_flush().await
     }
 }
 
@@ -328,76 +532,116 @@ pub struct BufferedUarteTx<'u, 'd, U: UarteInstance, T: TimerInstance> {
     inner: &'u BufferedUarte<'d, U, T>,
 }
 
+impl<'u, 'd, U: UarteInstance, T: TimerInstance> BufferedUarteTx<'u, 'd, U, T> {
+    /// Write a buffer into this writer, returning how many bytes were written.
+    pub async fn write(&mut self, buf: &[u8]) -> Result<usize, Error> {
+        self.inner.inner_write(buf).await
+    }
+
+    /// Flush this output stream, ensuring that all intermediately buffered contents reach their destination.
+    pub async fn flush(&mut self) -> Result<(), Error> {
+        self.inner.inner_flush().await
+    }
+}
+
 /// Writer part of the buffered UARTE driver.
 pub struct BufferedUarteRx<'u, 'd, U: UarteInstance, T: TimerInstance> {
     inner: &'u BufferedUarte<'d, U, T>,
 }
 
-impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarte<'d, U, T> {
-    type Error = core::convert::Infallible;
-}
+impl<'u, 'd, U: UarteInstance, T: TimerInstance> BufferedUarteRx<'u, 'd, U, T> {
+    /// Pull some bytes from this source into the specified buffer, returning how many bytes were read.
+    pub async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Error> {
+        self.inner.inner_read(buf).await
+    }
 
-impl<'u, 'd, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarteRx<'u, 'd, U, T> {
-    type Error = core::convert::Infallible;
-}
+    /// Return the contents of the internal buffer, filling it with more data from the inner reader if it is empty.
+    pub async fn fill_buf(&mut self) -> Result<&[u8], Error> {
+        self.inner.inner_fill_buf().await
+    }
 
-impl<'u, 'd, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarteTx<'u, 'd, U, T> {
-    type Error = core::convert::Infallible;
+    /// Tell this buffer that `amt` bytes have been consumed from the buffer, so they should no longer be returned in calls to `fill_buf`.
+    pub fn consume(&mut self, amt: usize) {
+        self.inner.inner_consume(amt)
+    }
 }
 
-impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Read for BufferedUarte<'d, U, T> {
-    async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Self::Error> {
-        self.inner_read(buf).await
+#[cfg(feature = "nightly")]
+mod _embedded_io {
+    use super::*;
+
+    impl embedded_io::Error for Error {
+        fn kind(&self) -> embedded_io::ErrorKind {
+            match *self {}
+        }
     }
-}
 
-impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Read for BufferedUarteRx<'u, 'd, U, T> {
-    async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Self::Error> {
-        self.inner.inner_read(buf).await
+    impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarte<'d, U, T> {
+        type Error = Error;
     }
-}
 
-impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::BufRead for BufferedUarte<'d, U, T> {
-    async fn fill_buf(&mut self) -> Result<&[u8], Self::Error> {
-        self.inner_fill_buf().await
+    impl<'u, 'd, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarteRx<'u, 'd, U, T> {
+        type Error = Error;
     }
 
-    fn consume(&mut self, amt: usize) {
-        self.inner_consume(amt)
+    impl<'u, 'd, U: UarteInstance, T: TimerInstance> embedded_io::Io for BufferedUarteTx<'u, 'd, U, T> {
+        type Error = Error;
     }
-}
 
-impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::BufRead for BufferedUarteRx<'u, 'd, U, T> {
-    async fn fill_buf(&mut self) -> Result<&[u8], Self::Error> {
-        self.inner.inner_fill_buf().await
+    impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Read for BufferedUarte<'d, U, T> {
+        async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Self::Error> {
+            self.inner_read(buf).await
+        }
     }
 
-    fn consume(&mut self, amt: usize) {
-        self.inner.inner_consume(amt)
+    impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Read for BufferedUarteRx<'u, 'd, U, T> {
+        async fn read(&mut self, buf: &mut [u8]) -> Result<usize, Self::Error> {
+            self.inner.inner_read(buf).await
+        }
     }
-}
 
-impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Write for BufferedUarte<'d, U, T> {
-    async fn write(&mut self, buf: &[u8]) -> Result<usize, Self::Error> {
-        self.inner_write(buf).await
+    impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::BufRead for BufferedUarte<'d, U, T> {
+        async fn fill_buf(&mut self) -> Result<&[u8], Self::Error> {
+            self.inner_fill_buf().await
+        }
+
+        fn consume(&mut self, amt: usize) {
+            self.inner_consume(amt)
+        }
     }
 
-    async fn flush(&mut self) -> Result<(), Self::Error> {
-        self.inner_flush().await
+    impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::BufRead for BufferedUarteRx<'u, 'd, U, T> {
+        async fn fill_buf(&mut self) -> Result<&[u8], Self::Error> {
+            self.inner.inner_fill_buf().await
+        }
+
+        fn consume(&mut self, amt: usize) {
+            self.inner.inner_consume(amt)
+        }
     }
-}
 
-impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Write for BufferedUarteTx<'u, 'd, U, T> {
-    async fn write(&mut self, buf: &[u8]) -> Result<usize, Self::Error> {
-        self.inner.inner_write(buf).await
+    impl<'d, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Write for BufferedUarte<'d, U, T> {
+        async fn write(&mut self, buf: &[u8]) -> Result<usize, Self::Error> {
+            self.inner_write(buf).await
+        }
+
+        async fn flush(&mut self) -> Result<(), Self::Error> {
+            self.inner_flush().await
+        }
     }
 
-    async fn flush(&mut self) -> Result<(), Self::Error> {
-        self.inner.inner_flush().await
+    impl<'u, 'd: 'u, U: UarteInstance, T: TimerInstance> embedded_io::asynch::Write for BufferedUarteTx<'u, 'd, U, T> {
+        async fn write(&mut self, buf: &[u8]) -> Result<usize, Self::Error> {
+            self.inner.inner_write(buf).await
+        }
+
+        async fn flush(&mut self) -> Result<(), Self::Error> {
+            self.inner.inner_flush().await
+        }
     }
 }
 
-impl<'a, U: UarteInstance, T: TimerInstance> Drop for StateInner<'a, U, T> {
+impl<'a, U: UarteInstance, T: TimerInstance> Drop for BufferedUarte<'a, U, T> {
     fn drop(&mut self) {
         let r = U::regs();
 
@@ -418,108 +662,11 @@ impl<'a, U: UarteInstance, T: TimerInstance> Drop for StateInner<'a, U, T> {
         gpio::deconfigure_pin(r.psel.txd.read().bits());
         gpio::deconfigure_pin(r.psel.rts.read().bits());
         gpio::deconfigure_pin(r.psel.cts.read().bits());
-    }
-}
 
-impl<'a, U: UarteInstance, T: TimerInstance> PeripheralState for StateInner<'a, U, T> {
-    type Interrupt = U::Interrupt;
-    fn on_interrupt(&mut self) {
-        trace!("irq: start");
-        let r = U::regs();
-
-        loop {
-            match self.rx_state {
-                RxState::Idle => {
-                    trace!("  irq_rx: in state idle");
-
-                    let buf = self.rx.push_buf();
-                    if !buf.is_empty() {
-                        trace!("  irq_rx: starting {:?}", buf.len());
-                        self.rx_state = RxState::Receiving;
-
-                        // Set up the DMA read
-                        r.rxd.ptr.write(|w|
-                            // The PTR field is a full 32 bits wide and accepts the full range
-                            // of values.
-                            unsafe { w.ptr().bits(buf.as_ptr() as u32) });
-                        r.rxd.maxcnt.write(|w|
-                            // We're giving it the length of the buffer, so no danger of
-                            // accessing invalid memory. We have verified that the length of the
-                            // buffer fits in an `u8`, so the cast to `u8` is also fine.
-                            //
-                            // The MAXCNT field is at least 8 bits wide and accepts the full
-                            // range of values.
-                            unsafe { w.maxcnt().bits(buf.len() as _) });
-                        trace!("  irq_rx: buf {:?} {:?}", buf.as_ptr() as u32, buf.len());
-
-                        // Start UARTE Receive transaction
-                        r.tasks_startrx.write(|w| unsafe { w.bits(1) });
-                    }
-                    break;
-                }
-                RxState::Receiving => {
-                    trace!("  irq_rx: in state receiving");
-                    if r.events_endrx.read().bits() != 0 {
-                        self.timer.stop();
-
-                        let n: usize = r.rxd.amount.read().amount().bits() as usize;
-                        trace!("  irq_rx: endrx {:?}", n);
-                        self.rx.push(n);
-
-                        r.events_endrx.reset();
-
-                        self.rx_waker.wake();
-                        self.rx_state = RxState::Idle;
-                    } else {
-                        break;
-                    }
-                }
-            }
-        }
-
-        loop {
-            match self.tx_state {
-                TxState::Idle => {
-                    trace!("  irq_tx: in state Idle");
-                    let buf = self.tx.pop_buf();
-                    if !buf.is_empty() {
-                        trace!("  irq_tx: starting {:?}", buf.len());
-                        self.tx_state = TxState::Transmitting(buf.len());
-
-                        // Set up the DMA write
-                        r.txd.ptr.write(|w|
-                            // The PTR field is a full 32 bits wide and accepts the full range
-                            // of values.
-                            unsafe { w.ptr().bits(buf.as_ptr() as u32) });
-                        r.txd.maxcnt.write(|w|
-                            // We're giving it the length of the buffer, so no danger of
-                            // accessing invalid memory. We have verified that the length of the
-                            // buffer fits in an `u8`, so the cast to `u8` is also fine.
-                            //
-                            // The MAXCNT field is 8 bits wide and accepts the full range of
-                            // values.
-                            unsafe { w.maxcnt().bits(buf.len() as _) });
-
-                        // Start UARTE Transmit transaction
-                        r.tasks_starttx.write(|w| unsafe { w.bits(1) });
-                    }
-                    break;
-                }
-                TxState::Transmitting(n) => {
-                    trace!("  irq_tx: in state Transmitting");
-                    if r.events_endtx.read().bits() != 0 {
-                        r.events_endtx.reset();
-
-                        trace!("  irq_tx: endtx {:?}", n);
-                        self.tx.pop(n);
-                        self.tx_waker.wake();
-                        self.tx_state = TxState::Idle;
-                    } else {
-                        break;
-                    }
-                }
-            }
+        let s = U::buffered_state();
+        unsafe {
+            s.rx_buf.deinit();
+            s.tx_buf.deinit();
         }
-        trace!("irq: end");
     }
 }
diff --git a/embassy-nrf/src/lib.rs b/embassy-nrf/src/lib.rs
index a9683df44..6b7dc7791 100644
--- a/embassy-nrf/src/lib.rs
+++ b/embassy-nrf/src/lib.rs
@@ -37,7 +37,6 @@ pub(crate) mod util;
 #[cfg(feature = "_time-driver")]
 mod time_driver;
 
-#[cfg(feature = "nightly")]
 pub mod buffered_uarte;
 pub mod gpio;
 #[cfg(feature = "gpiote")]
diff --git a/embassy-nrf/src/ppi/dppi.rs b/embassy-nrf/src/ppi/dppi.rs
index 0908cd7be..3a1e7f170 100644
--- a/embassy-nrf/src/ppi/dppi.rs
+++ b/embassy-nrf/src/ppi/dppi.rs
@@ -6,7 +6,7 @@ use crate::{pac, Peripheral};
 const DPPI_ENABLE_BIT: u32 = 0x8000_0000;
 const DPPI_CHANNEL_MASK: u32 = 0x0000_00FF;
 
-fn regs() -> &'static pac::dppic::RegisterBlock {
+pub(crate) fn regs() -> &'static pac::dppic::RegisterBlock {
     unsafe { &*pac::DPPIC::ptr() }
 }
 
diff --git a/embassy-nrf/src/ppi/mod.rs b/embassy-nrf/src/ppi/mod.rs
index b76eccf0b..7c18da6ee 100644
--- a/embassy-nrf/src/ppi/mod.rs
+++ b/embassy-nrf/src/ppi/mod.rs
@@ -17,16 +17,16 @@
 
 use core::ptr::NonNull;
 
-use embassy_hal_common::{impl_peripheral, PeripheralRef};
+use embassy_hal_common::{impl_peripheral, into_ref, PeripheralRef};
 
 use crate::{peripherals, Peripheral};
 
-#[cfg(feature = "_dppi")]
-mod dppi;
-#[cfg(feature = "_ppi")]
-mod ppi;
+#[cfg_attr(feature = "_dppi", path = "dppi.rs")]
+#[cfg_attr(feature = "_ppi", path = "ppi.rs")]
+mod _version;
+pub(crate) use _version::*;
 
-/// An instance of the Programmable peripheral interconnect on nRF devices.
+/// PPI channel driver.
 pub struct Ppi<'d, C: Channel, const EVENT_COUNT: usize, const TASK_COUNT: usize> {
     ch: PeripheralRef<'d, C>,
     #[cfg(feature = "_dppi")]
@@ -35,6 +35,88 @@ pub struct Ppi<'d, C: Channel, const EVENT_COUNT: usize, const TASK_COUNT: usize
     tasks: [Task; TASK_COUNT],
 }
 
+/// PPI channel group driver.
+pub struct PpiGroup<'d, G: Group> {
+    g: PeripheralRef<'d, G>,
+}
+
+impl<'d, G: Group> PpiGroup<'d, G> {
+    /// Create a new PPI group driver.
+    ///
+    /// The group is initialized as containing no channels.
+    pub fn new(g: impl Peripheral<P = G> + 'd) -> Self {
+        into_ref!(g);
+
+        let r = regs();
+        let n = g.number();
+        r.chg[n].write(|w| unsafe { w.bits(0) });
+
+        Self { g }
+    }
+
+    /// Add a PPI channel to this group.
+    ///
+    /// If the channel is already in the group, this is a no-op.
+    pub fn add_channel<C: Channel, const EVENT_COUNT: usize, const TASK_COUNT: usize>(
+        &mut self,
+        ch: &Ppi<'_, C, EVENT_COUNT, TASK_COUNT>,
+    ) {
+        let r = regs();
+        let ng = self.g.number();
+        let nc = ch.ch.number();
+        r.chg[ng].modify(|r, w| unsafe { w.bits(r.bits() | 1 << nc) });
+    }
+
+    /// Remove a PPI channel from this group.
+    ///
+    /// If the channel is already not in the group, this is a no-op.
+    pub fn remove_channel<C: Channel, const EVENT_COUNT: usize, const TASK_COUNT: usize>(
+        &mut self,
+        ch: &Ppi<'_, C, EVENT_COUNT, TASK_COUNT>,
+    ) {
+        let r = regs();
+        let ng = self.g.number();
+        let nc = ch.ch.number();
+        r.chg[ng].modify(|r, w| unsafe { w.bits(r.bits() & !(1 << nc)) });
+    }
+
+    /// Enable all the channels in this group.
+    pub fn enable_all(&mut self) {
+        let n = self.g.number();
+        regs().tasks_chg[n].en.write(|w| unsafe { w.bits(1) });
+    }
+
+    /// Disable all the channels in this group.
+    pub fn disable_all(&mut self) {
+        let n = self.g.number();
+        regs().tasks_chg[n].dis.write(|w| unsafe { w.bits(1) });
+    }
+
+    /// Get a reference to the "enable all" task.
+    ///
+    /// When triggered, it will enable all the channels in this group.
+    pub fn task_enable_all(&self) -> Task {
+        let n = self.g.number();
+        Task::from_reg(&regs().tasks_chg[n].en)
+    }
+
+    /// Get a reference to the "disable all" task.
+    ///
+    /// When triggered, it will disable all the channels in this group.
+    pub fn task_disable_all(&self) -> Task {
+        let n = self.g.number();
+        Task::from_reg(&regs().tasks_chg[n].dis)
+    }
+}
+
+impl<'d, G: Group> Drop for PpiGroup<'d, G> {
+    fn drop(&mut self) {
+        let r = regs();
+        let n = self.g.number();
+        r.chg[n].write(|w| unsafe { w.bits(0) });
+    }
+}
+
 #[cfg(feature = "_dppi")]
 const REGISTER_DPPI_CONFIG_OFFSET: usize = 0x80 / core::mem::size_of::<u32>();
 
@@ -112,7 +194,7 @@ pub(crate) mod sealed {
 }
 
 /// Interface for PPI channels.
-pub trait Channel: sealed::Channel + Peripheral<P = Self> + Sized {
+pub trait Channel: sealed::Channel + Peripheral<P = Self> + Sized + 'static {
     /// Returns the number of the channel
     fn number(&self) -> usize;
 }
@@ -130,7 +212,7 @@ pub trait StaticChannel: Channel + Into<AnyStaticChannel> {
 }
 
 /// Interface for a group of PPI channels.
-pub trait Group: sealed::Group + Sized {
+pub trait Group: sealed::Group + Peripheral<P = Self> + Into<AnyGroup> + Sized + 'static {
     /// Returns the number of the group.
     fn number(&self) -> usize;
     /// Convert into a type erased group.
@@ -248,6 +330,12 @@ macro_rules! impl_group {
                 $number
             }
         }
+
+        impl From<peripherals::$type> for crate::ppi::AnyGroup {
+            fn from(val: peripherals::$type) -> Self {
+                crate::ppi::Group::degrade(val)
+            }
+        }
     };
 }
 
diff --git a/embassy-nrf/src/ppi/ppi.rs b/embassy-nrf/src/ppi/ppi.rs
index a96ab50b7..f1eeaee1e 100644
--- a/embassy-nrf/src/ppi/ppi.rs
+++ b/embassy-nrf/src/ppi/ppi.rs
@@ -14,7 +14,7 @@ impl Event {
     }
 }
 
-fn regs() -> &'static pac::ppi::RegisterBlock {
+pub(crate) fn regs() -> &'static pac::ppi::RegisterBlock {
     unsafe { &*pac::PPI::ptr() }
 }
 
diff --git a/embassy-nrf/src/timer.rs b/embassy-nrf/src/timer.rs
index d1ae57237..3b0d2f1ca 100644
--- a/embassy-nrf/src/timer.rs
+++ b/embassy-nrf/src/timer.rs
@@ -132,7 +132,21 @@ impl<'d, T: Instance> Timer<'d, T, Awaitable> {
         irq.unpend();
         irq.enable();
 
-        Self::new_inner(timer)
+        Self::new_inner(timer, false)
+    }
+
+    /// Create a new async-capable timer driver in counter mode.
+    pub fn new_awaitable_counter(
+        timer: impl Peripheral<P = T> + 'd,
+        irq: impl Peripheral<P = T::Interrupt> + 'd,
+    ) -> Self {
+        into_ref!(irq);
+
+        irq.set_handler(Self::on_interrupt);
+        irq.unpend();
+        irq.enable();
+
+        Self::new_inner(timer, true)
     }
 }
 
@@ -142,7 +156,15 @@ impl<'d, T: Instance> Timer<'d, T, NotAwaitable> {
     /// This can be useful for triggering tasks via PPI
     /// `Uarte` uses this internally.
     pub fn new(timer: impl Peripheral<P = T> + 'd) -> Self {
-        Self::new_inner(timer)
+        Self::new_inner(timer, false)
+    }
+
+    /// Create a `Timer` driver in counter mode without an interrupt, meaning `Cc::wait` won't work.
+    ///
+    /// This can be useful for triggering tasks via PPI
+    /// `Uarte` uses this internally.
+    pub fn new_counter(timer: impl Peripheral<P = T> + 'd) -> Self {
+        Self::new_inner(timer, true)
     }
 }
 
@@ -150,7 +172,7 @@ impl<'d, T: Instance, I: TimerType> Timer<'d, T, I> {
     /// Create a `Timer` without an interrupt, meaning `Cc::wait` won't work.
     ///
     /// This is used by the public constructors.
-    fn new_inner(timer: impl Peripheral<P = T> + 'd) -> Self {
+    fn new_inner(timer: impl Peripheral<P = T> + 'd, is_counter: bool) -> Self {
         into_ref!(timer);
 
         let regs = T::regs();
@@ -164,8 +186,11 @@ impl<'d, T: Instance, I: TimerType> Timer<'d, T, I> {
         // since changing BITMODE while running can cause 'unpredictable behaviour' according to the specification.
         this.stop();
 
-        // Set the instance to timer mode.
-        regs.mode.write(|w| w.mode().timer());
+        if is_counter {
+            regs.mode.write(|w| w.mode().counter());
+        } else {
+            regs.mode.write(|w| w.mode().timer());
+        }
 
         // Make the counter's max value as high as possible.
         // TODO: is there a reason someone would want to set this lower?
@@ -225,6 +250,14 @@ impl<'d, T: Instance, I: TimerType> Timer<'d, T, I> {
         Task::from_reg(&T::regs().tasks_clear)
     }
 
+    /// Returns the COUNT task, for use with PPI.
+    ///
+    /// When triggered, this task increments the timer's counter by 1.
+    /// Only works in counter mode.
+    pub fn task_count(&self) -> Task {
+        Task::from_reg(&T::regs().tasks_count)
+    }
+
     /// Change the timer's frequency.
     ///
     /// This will stop the timer if it isn't already stopped,
diff --git a/embassy-nrf/src/uarte.rs b/embassy-nrf/src/uarte.rs
index 48457744b..00afbd059 100644
--- a/embassy-nrf/src/uarte.rs
+++ b/embassy-nrf/src/uarte.rs
@@ -883,6 +883,7 @@ pub(crate) mod sealed {
     pub trait Instance {
         fn regs() -> &'static pac::uarte0::RegisterBlock;
         fn state() -> &'static State;
+        fn buffered_state() -> &'static crate::buffered_uarte::State;
     }
 }
 
@@ -902,6 +903,10 @@ macro_rules! impl_uarte {
                 static STATE: crate::uarte::sealed::State = crate::uarte::sealed::State::new();
                 &STATE
             }
+            fn buffered_state() -> &'static crate::buffered_uarte::State {
+                static STATE: crate::buffered_uarte::State = crate::buffered_uarte::State::new();
+                &STATE
+            }
         }
         impl crate::uarte::Instance for peripherals::$type {
             type Interrupt = crate::interrupt::$irq;
diff --git a/examples/nrf52840/src/bin/buffered_uart.rs b/examples/nrf52840/src/bin/buffered_uart.rs
index ea566f4b2..5b934b7d6 100644
--- a/examples/nrf52840/src/bin/buffered_uart.rs
+++ b/examples/nrf52840/src/bin/buffered_uart.rs
@@ -4,10 +4,9 @@
 
 use defmt::*;
 use embassy_executor::Spawner;
-use embassy_nrf::buffered_uarte::{BufferedUarte, State};
+use embassy_nrf::buffered_uarte::BufferedUarte;
 use embassy_nrf::{interrupt, uarte};
-use embedded_io::asynch::{BufRead, Write};
-use futures::pin_mut;
+use embedded_io::asynch::Write;
 use {defmt_rtt as _, panic_probe as _};
 
 #[embassy_executor::main]
@@ -21,24 +20,19 @@ async fn main(_spawner: Spawner) {
     let mut rx_buffer = [0u8; 4096];
 
     let irq = interrupt::take!(UARTE0_UART0);
-    let mut state = State::new();
-    // Please note - important to have hardware flow control (https://github.com/embassy-rs/embassy/issues/536)
-    let u = BufferedUarte::new(
-        &mut state,
+    let mut u = BufferedUarte::new(
         p.UARTE0,
         p.TIMER0,
         p.PPI_CH0,
         p.PPI_CH1,
+        p.PPI_GROUP0,
         irq,
         p.P0_08,
         p.P0_06,
-        p.P0_07,
-        p.P0_05,
         config,
         &mut rx_buffer,
         &mut tx_buffer,
     );
-    pin_mut!(u);
 
     info!("uarte initialized!");
 
diff --git a/tests/nrf/.cargo/config.toml b/tests/nrf/.cargo/config.toml
new file mode 100644
index 000000000..4eec189d4
--- /dev/null
+++ b/tests/nrf/.cargo/config.toml
@@ -0,0 +1,9 @@
+[target.'cfg(all(target_arch = "arm", target_os = "none"))']
+#runner = "teleprobe local run --chip nRF52840_xxAA --elf"
+runner = "teleprobe client run --target nrf52840-dk --elf"
+
+[build]
+target = "thumbv7em-none-eabi"
+
+[env]
+DEFMT_LOG = "trace"
diff --git a/tests/nrf/Cargo.toml b/tests/nrf/Cargo.toml
new file mode 100644
index 000000000..2a4e8cf41
--- /dev/null
+++ b/tests/nrf/Cargo.toml
@@ -0,0 +1,20 @@
+[package]
+edition = "2021"
+name = "embassy-nrf-examples"
+version = "0.1.0"
+license = "MIT OR Apache-2.0"
+
+[dependencies]
+embassy-futures = { version = "0.1.0", path = "../../embassy-futures" }
+embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt", "nightly"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "nightly", "integrated-timers"] }
+embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "nightly", "defmt-timestamp-uptime"] }
+embassy-nrf = { version = "0.1.0", path = "../../embassy-nrf", features = ["defmt", "nightly", "unstable-traits", "nrf52840", "time-driver-rtc1", "gpiote", "unstable-pac"] }
+embedded-io = { version = "0.4.0", features = ["async"] }
+
+defmt = "0.3"
+defmt-rtt = "0.4"
+
+cortex-m = { version = "0.7.6", features = ["critical-section-single-core"] }
+cortex-m-rt = "0.7.0"
+panic-probe = { version = "0.3", features = ["print-defmt"] }
\ No newline at end of file
diff --git a/tests/nrf/build.rs b/tests/nrf/build.rs
new file mode 100644
index 000000000..6f4872249
--- /dev/null
+++ b/tests/nrf/build.rs
@@ -0,0 +1,16 @@
+use std::error::Error;
+use std::path::PathBuf;
+use std::{env, fs};
+
+fn main() -> Result<(), Box<dyn Error>> {
+    let out = PathBuf::from(env::var("OUT_DIR").unwrap());
+    fs::write(out.join("link_ram.x"), include_bytes!("link_ram.x")).unwrap();
+    println!("cargo:rustc-link-search={}", out.display());
+    println!("cargo:rerun-if-changed=link_ram.x");
+
+    println!("cargo:rustc-link-arg-bins=--nmagic");
+    println!("cargo:rustc-link-arg-bins=-Tlink_ram.x");
+    println!("cargo:rustc-link-arg-bins=-Tdefmt.x");
+
+    Ok(())
+}
diff --git a/tests/nrf/link_ram.x b/tests/nrf/link_ram.x
new file mode 100644
index 000000000..26da86baa
--- /dev/null
+++ b/tests/nrf/link_ram.x
@@ -0,0 +1,254 @@
+/* ##### EMBASSY NOTE
+    Originally from https://github.com/rust-embedded/cortex-m-rt/blob/master/link.x.in
+    Adjusted to put everything in RAM
+*/
+
+/* # Developer notes
+
+- Symbols that start with a double underscore (__) are considered "private"
+
+- Symbols that start with a single underscore (_) are considered "semi-public"; they can be
+  overridden in a user linker script, but should not be referred from user code (e.g. `extern "C" {
+  static mut __sbss }`).
+
+- `EXTERN` forces the linker to keep a symbol in the final binary. We use this to make sure a
+  symbol if not dropped if it appears in or near the front of the linker arguments and "it's not
+  needed" by any of the preceding objects (linker arguments)
+
+- `PROVIDE` is used to provide default values that can be overridden by a user linker script
+
+- On alignment: it's important for correctness that the VMA boundaries of both .bss and .data *and*
+  the LMA of .data are all 4-byte aligned. These alignments are assumed by the RAM initialization
+  routine. There's also a second benefit: 4-byte aligned boundaries means that you won't see
+  "Address (..) is out of bounds" in the disassembly produced by `objdump`.
+*/
+
+/* Provides information about the memory layout of the device */
+/* This will be provided by the user (see `memory.x`) or by a Board Support Crate */
+INCLUDE memory.x
+
+/* # Entry point = reset vector */
+EXTERN(__RESET_VECTOR);
+EXTERN(Reset);
+ENTRY(Reset);
+
+/* # Exception vectors */
+/* This is effectively weak aliasing at the linker level */
+/* The user can override any of these aliases by defining the corresponding symbol themselves (cf.
+   the `exception!` macro) */
+EXTERN(__EXCEPTIONS); /* depends on all the these PROVIDED symbols */
+
+EXTERN(DefaultHandler);
+
+PROVIDE(NonMaskableInt = DefaultHandler);
+EXTERN(HardFaultTrampoline);
+PROVIDE(MemoryManagement = DefaultHandler);
+PROVIDE(BusFault = DefaultHandler);
+PROVIDE(UsageFault = DefaultHandler);
+PROVIDE(SecureFault = DefaultHandler);
+PROVIDE(SVCall = DefaultHandler);
+PROVIDE(DebugMonitor = DefaultHandler);
+PROVIDE(PendSV = DefaultHandler);
+PROVIDE(SysTick = DefaultHandler);
+
+PROVIDE(DefaultHandler = DefaultHandler_);
+PROVIDE(HardFault = HardFault_);
+
+/* # Interrupt vectors */
+EXTERN(__INTERRUPTS); /* `static` variable similar to `__EXCEPTIONS` */
+
+/* # Pre-initialization function */
+/* If the user overrides this using the `pre_init!` macro or by creating a `__pre_init` function,
+   then the function this points to will be called before the RAM is initialized. */
+PROVIDE(__pre_init = DefaultPreInit);
+
+/* # Sections */
+SECTIONS
+{
+  PROVIDE(_stack_start = ORIGIN(RAM) + LENGTH(RAM));
+
+  /* ## Sections in RAM */
+  /* ### Vector table */
+  .vector_table ORIGIN(RAM) :
+  {
+    /* Initial Stack Pointer (SP) value */
+    LONG(_stack_start);
+
+    /* Reset vector */
+    KEEP(*(.vector_table.reset_vector)); /* this is the `__RESET_VECTOR` symbol */
+    __reset_vector = .;
+
+    /* Exceptions */
+    KEEP(*(.vector_table.exceptions)); /* this is the `__EXCEPTIONS` symbol */
+    __eexceptions = .;
+
+    /* Device specific interrupts */
+    KEEP(*(.vector_table.interrupts)); /* this is the `__INTERRUPTS` symbol */
+  } > RAM
+
+  PROVIDE(_stext = ADDR(.vector_table) + SIZEOF(.vector_table));
+
+  /* ### .text */
+  .text _stext :
+  {
+    __stext = .;
+    *(.Reset);
+
+    *(.text .text.*);
+
+    /* The HardFaultTrampoline uses the `b` instruction to enter `HardFault`,
+       so must be placed close to it. */
+    *(.HardFaultTrampoline);
+    *(.HardFault.*);
+
+    . = ALIGN(4); /* Pad .text to the alignment to workaround overlapping load section bug in old lld */
+    __etext = .;
+  } > RAM
+
+  /* ### .rodata */
+  .rodata : ALIGN(4)
+  {
+    . = ALIGN(4);
+    __srodata = .;
+    *(.rodata .rodata.*);
+
+    /* 4-byte align the end (VMA) of this section.
+       This is required by LLD to ensure the LMA of the following .data
+       section will have the correct alignment. */
+    . = ALIGN(4);
+    __erodata = .;
+  } > RAM
+
+  /* ## Sections in RAM */
+  /* ### .data */
+  .data : ALIGN(4)
+  {
+    . = ALIGN(4);
+    __sdata = .;
+    __edata = .;
+    *(.data .data.*);
+    . = ALIGN(4); /* 4-byte align the end (VMA) of this section */
+  } > RAM
+  /* Allow sections from user `memory.x` injected using `INSERT AFTER .data` to
+   * use the .data loading mechanism by pushing __edata. Note: do not change
+   * output region or load region in those user sections! */
+  . = ALIGN(4);
+
+  /* LMA of .data */
+  __sidata = LOADADDR(.data);
+
+  /* ### .gnu.sgstubs
+     This section contains the TrustZone-M veneers put there by the Arm GNU linker. */
+  /* Security Attribution Unit blocks must be 32 bytes aligned. */
+  /* Note that this pads the RAM usage to 32 byte alignment. */
+  .gnu.sgstubs : ALIGN(32)
+  {
+    . = ALIGN(32);
+    __veneer_base = .;
+    *(.gnu.sgstubs*)
+    . = ALIGN(32);
+    __veneer_limit = .;
+  } > RAM
+
+  /* ### .bss */
+  .bss (NOLOAD) : ALIGN(4)
+  {
+    . = ALIGN(4);
+    __sbss = .;
+    *(.bss .bss.*);
+    *(COMMON); /* Uninitialized C statics */
+    . = ALIGN(4); /* 4-byte align the end (VMA) of this section */
+  } > RAM
+  /* Allow sections from user `memory.x` injected using `INSERT AFTER .bss` to
+   * use the .bss zeroing mechanism by pushing __ebss. Note: do not change
+   * output region or load region in those user sections! */
+  . = ALIGN(4);
+  __ebss = .;
+
+  /* ### .uninit */
+  .uninit (NOLOAD) : ALIGN(4)
+  {
+    . = ALIGN(4);
+    __suninit = .;
+    *(.uninit .uninit.*);
+    . = ALIGN(4);
+    __euninit = .;
+  } > RAM
+
+  /* Place the heap right after `.uninit` in RAM */
+  PROVIDE(__sheap = __euninit);
+
+  /* ## .got */
+  /* Dynamic relocations are unsupported. This section is only used to detect relocatable code in
+     the input files and raise an error if relocatable code is found */
+  .got (NOLOAD) :
+  {
+    KEEP(*(.got .got.*));
+  }
+
+  /* ## Discarded sections */
+  /DISCARD/ :
+  {
+    /* Unused exception related info that only wastes space */
+    *(.ARM.exidx);
+    *(.ARM.exidx.*);
+    *(.ARM.extab.*);
+  }
+}
+
+/* Do not exceed this mark in the error messages below                                    | */
+/* # Alignment checks */
+ASSERT(ORIGIN(RAM) % 4 == 0, "
+ERROR(cortex-m-rt): the start of the RAM region must be 4-byte aligned");
+
+ASSERT(__sdata % 4 == 0 && __edata % 4 == 0, "
+BUG(cortex-m-rt): .data is not 4-byte aligned");
+
+ASSERT(__sidata % 4 == 0, "
+BUG(cortex-m-rt): the LMA of .data is not 4-byte aligned");
+
+ASSERT(__sbss % 4 == 0 && __ebss % 4 == 0, "
+BUG(cortex-m-rt): .bss is not 4-byte aligned");
+
+ASSERT(__sheap % 4 == 0, "
+BUG(cortex-m-rt): start of .heap is not 4-byte aligned");
+
+/* # Position checks */
+
+/* ## .vector_table */
+ASSERT(__reset_vector == ADDR(.vector_table) + 0x8, "
+BUG(cortex-m-rt): the reset vector is missing");
+
+ASSERT(__eexceptions == ADDR(.vector_table) + 0x40, "
+BUG(cortex-m-rt): the exception vectors are missing");
+
+ASSERT(SIZEOF(.vector_table) > 0x40, "
+ERROR(cortex-m-rt): The interrupt vectors are missing.
+Possible solutions, from most likely to less likely:
+- Link to a svd2rust generated device crate
+- Check that you actually use the device/hal/bsp crate in your code
+- Disable the 'device' feature of cortex-m-rt to build a generic application (a dependency
+may be enabling it)
+- Supply the interrupt handlers yourself. Check the documentation for details.");
+
+/* ## .text */
+ASSERT(ADDR(.vector_table) + SIZEOF(.vector_table) <= _stext, "
+ERROR(cortex-m-rt): The .text section can't be placed inside the .vector_table section
+Set _stext to an address greater than the end of .vector_table (See output of `nm`)");
+
+ASSERT(_stext + SIZEOF(.text) < ORIGIN(RAM) + LENGTH(RAM), "
+ERROR(cortex-m-rt): The .text section must be placed inside the RAM memory.
+Set _stext to an address smaller than 'ORIGIN(RAM) + LENGTH(RAM)'");
+
+/* # Other checks */
+ASSERT(SIZEOF(.got) == 0, "
+ERROR(cortex-m-rt): .got section detected in the input object files
+Dynamic relocations are not supported. If you are linking to C code compiled using
+the 'cc' crate then modify your build script to compile the C code _without_
+the -fPIC flag. See the documentation of the `cc::Build.pic` method for details.");
+/* Do not exceed this mark in the error messages above                                    | */
+
+
+/* Provides weak aliases (cf. PROVIDED) for device specific interrupt handlers */
+/* This will usually be provided by a device crate generated using svd2rust (see `device.x`) */
+INCLUDE device.x
\ No newline at end of file
diff --git a/tests/nrf/memory.x b/tests/nrf/memory.x
new file mode 100644
index 000000000..58900a7bd
--- /dev/null
+++ b/tests/nrf/memory.x
@@ -0,0 +1,5 @@
+MEMORY
+{
+  FLASH : ORIGIN = 0x00000000, LENGTH = 1024K
+  RAM : ORIGIN = 0x20000000, LENGTH = 256K
+}
diff --git a/tests/nrf/src/bin/buffered_uart.rs b/tests/nrf/src/bin/buffered_uart.rs
new file mode 100644
index 000000000..0550b0bb7
--- /dev/null
+++ b/tests/nrf/src/bin/buffered_uart.rs
@@ -0,0 +1,74 @@
+#![no_std]
+#![no_main]
+#![feature(type_alias_impl_trait)]
+
+use defmt::{assert_eq, *};
+use embassy_executor::Spawner;
+use embassy_futures::join::join;
+use embassy_nrf::buffered_uarte::BufferedUarte;
+use embassy_nrf::{interrupt, uarte};
+use {defmt_rtt as _, panic_probe as _};
+
+#[embassy_executor::main]
+async fn main(_spawner: Spawner) {
+    let p = embassy_nrf::init(Default::default());
+    let mut config = uarte::Config::default();
+    config.parity = uarte::Parity::EXCLUDED;
+    config.baudrate = uarte::Baudrate::BAUD1M;
+
+    let mut tx_buffer = [0u8; 1024];
+    let mut rx_buffer = [0u8; 1024];
+
+    let mut u = BufferedUarte::new(
+        p.UARTE0,
+        p.TIMER0,
+        p.PPI_CH0,
+        p.PPI_CH1,
+        p.PPI_GROUP0,
+        interrupt::take!(UARTE0_UART0),
+        p.P1_03,
+        p.P1_02,
+        config.clone(),
+        &mut rx_buffer,
+        &mut tx_buffer,
+    );
+
+    info!("uarte initialized!");
+
+    let (mut rx, mut tx) = u.split();
+
+    const COUNT: usize = 40_000;
+
+    let tx_fut = async {
+        let mut tx_buf = [0; 215];
+        let mut i = 0;
+        while i < COUNT {
+            let n = tx_buf.len().min(COUNT - i);
+            let tx_buf = &mut tx_buf[..n];
+            for (j, b) in tx_buf.iter_mut().enumerate() {
+                *b = (i + j) as u8;
+            }
+            let n = unwrap!(tx.write(tx_buf).await);
+            i += n;
+        }
+    };
+    let rx_fut = async {
+        let mut i = 0;
+        while i < COUNT {
+            let buf = unwrap!(rx.fill_buf().await);
+
+            for &b in buf {
+                assert_eq!(b, i as u8);
+                i = i + 1;
+            }
+
+            let n = buf.len();
+            rx.consume(n);
+        }
+    };
+
+    join(rx_fut, tx_fut).await;
+
+    info!("Test OK");
+    cortex_m::asm::bkpt();
+}
diff --git a/tests/nrf/src/bin/buffered_uart_spam.rs b/tests/nrf/src/bin/buffered_uart_spam.rs
new file mode 100644
index 000000000..57aaeca45
--- /dev/null
+++ b/tests/nrf/src/bin/buffered_uart_spam.rs
@@ -0,0 +1,86 @@
+#![no_std]
+#![no_main]
+#![feature(type_alias_impl_trait)]
+
+use core::mem;
+use core::ptr::NonNull;
+
+use defmt::{assert_eq, *};
+use embassy_executor::Spawner;
+use embassy_nrf::buffered_uarte::BufferedUarte;
+use embassy_nrf::gpio::{Level, Output, OutputDrive};
+use embassy_nrf::ppi::{Event, Ppi, Task};
+use embassy_nrf::uarte::Uarte;
+use embassy_nrf::{interrupt, pac, uarte};
+use embassy_time::{Duration, Timer};
+use {defmt_rtt as _, panic_probe as _};
+
+#[embassy_executor::main]
+async fn main(_spawner: Spawner) {
+    let mut p = embassy_nrf::init(Default::default());
+    let mut config = uarte::Config::default();
+    config.parity = uarte::Parity::EXCLUDED;
+    config.baudrate = uarte::Baudrate::BAUD1M;
+
+    let mut tx_buffer = [0u8; 1024];
+    let mut rx_buffer = [0u8; 1024];
+
+    mem::forget(Output::new(&mut p.P1_02, Level::High, OutputDrive::Standard));
+
+    let mut u = BufferedUarte::new(
+        p.UARTE0,
+        p.TIMER0,
+        p.PPI_CH0,
+        p.PPI_CH1,
+        p.PPI_GROUP0,
+        interrupt::take!(UARTE0_UART0),
+        p.P1_03,
+        p.P1_04,
+        config.clone(),
+        &mut rx_buffer,
+        &mut tx_buffer,
+    );
+
+    info!("uarte initialized!");
+
+    // uarte needs some quiet time to start rxing properly.
+    Timer::after(Duration::from_millis(10)).await;
+
+    // Tx spam in a loop.
+    const NSPAM: usize = 17;
+    static mut TX_BUF: [u8; NSPAM] = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16];
+    let _spam = Uarte::new(p.UARTE1, interrupt::take!(UARTE1), p.P1_01, p.P1_02, config.clone());
+    let spam_peri: pac::UARTE1 = unsafe { mem::transmute(()) };
+    let event = unsafe { Event::new_unchecked(NonNull::new_unchecked(&spam_peri.events_endtx as *const _ as _)) };
+    let task = unsafe { Task::new_unchecked(NonNull::new_unchecked(&spam_peri.tasks_starttx as *const _ as _)) };
+    let mut spam_ppi = Ppi::new_one_to_one(p.PPI_CH2, event, task);
+    spam_ppi.enable();
+    let p = unsafe { TX_BUF.as_mut_ptr() };
+    spam_peri.txd.ptr.write(|w| unsafe { w.ptr().bits(p as u32) });
+    spam_peri.txd.maxcnt.write(|w| unsafe { w.maxcnt().bits(NSPAM as _) });
+    spam_peri.tasks_starttx.write(|w| unsafe { w.bits(1) });
+
+    let mut i = 0;
+    let mut total = 0;
+    while total < 256 * 1024 {
+        let buf = unwrap!(u.fill_buf().await);
+        //info!("rx {}", buf);
+
+        for &b in buf {
+            assert_eq!(b, unsafe { TX_BUF[i] });
+
+            i = i + 1;
+            if i == NSPAM {
+                i = 0;
+            }
+        }
+
+        // Read bytes have to be explicitly consumed, otherwise fill_buf() will return them again
+        let n = buf.len();
+        u.consume(n);
+        total += n;
+    }
+
+    info!("Test OK");
+    cortex_m::asm::bkpt();
+}