added tests to check the for the correct number of clocks during valid

hil-test/Cargo.toml

@@ -99,6 +99,14 @@ harness = false
 name    = "systimer"
 harness = false
 
+[[test]]
+name    = "parl_io_tx"
+harness = false
+
+[[test]]
+name    = "parl_io_tx_async"
+harness = false
+
 [[test]]
 name    = "pcnt"
 harness = false

hil-test/tests/parl_io_tx.rs

@@ -0,0 +1,181 @@
+//! PARL_IO TX test
+
+//% CHIPS: esp32c6 esp32h2
+
+#![no_std]
+#![no_main]
+
+use esp_hal::{
+    dma::{ChannelCreator, Dma, DmaPriority},
+    gpio::{interconnect::InputSignal, AnyPin, Io, NoPin},
+    parl_io::{
+        BitPackOrder,
+        ClkOutPin,
+        ParlIoTxOnly,
+        SampleEdge,
+        TxEightBits,
+        TxPinConfigIncludingValidPin,
+        TxPinConfigWithValidPin,
+        TxSixteenBits,
+    },
+    pcnt::{
+        channel::{CtrlMode, EdgeMode},
+        unit::Unit,
+        Pcnt,
+    },
+    peripherals::PARL_IO,
+    prelude::*,
+};
+use hil_test as _;
+
+struct Context {
+    parl_io: PARL_IO,
+    dma_channel: ChannelCreator<0>,
+    clock: AnyPin,
+    valid: AnyPin,
+    clock_loopback: InputSignal,
+    valid_loopback: InputSignal,
+    pcnt_unit: Unit<'static, 0>,
+}
+
+#[cfg(test)]
+#[embedded_test::tests]
+mod tests {
+    // defmt::* is load-bearing, it ensures that the assert in dma_buffers! is not
+    // using defmt's non-const assert. Doing so would result in a compile error.
+    #[allow(unused_imports)]
+    use defmt::{assert_eq, *};
+
+    use super::*;
+
+    #[init]
+    fn init() -> Context {
+        let peripherals = esp_hal::init(esp_hal::Config::default());
+
+        let io = Io::new(peripherals.GPIO, peripherals.IO_MUX);
+        let (clock, _) = hil_test::common_test_pins!(io);
+        let valid = io.pins.gpio0.degrade();
+        let clock_loopback = clock.peripheral_input();
+        let valid_loopback = valid.peripheral_input();
+        let clock = clock.degrade();
+        let pcnt = Pcnt::new(peripherals.PCNT);
+        let pcnt_unit = pcnt.unit0;
+        let dma = Dma::new(peripherals.DMA);
+        let dma_channel = dma.channel0;
+
+        let parl_io = peripherals.PARL_IO;
+
+        Context {
+            parl_io,
+            dma_channel,
+            clock,
+            valid,
+            clock_loopback,
+            valid_loopback,
+            pcnt_unit,
+        }
+    }
+
+    #[test]
+    #[timeout(3)]
+    fn test_parl_io_tx_16bit_valid_clock_count(ctx: Context) {
+        const BUFFER_SIZE: usize = 64;
+        let tx_buffer = [0u16; BUFFER_SIZE];
+        let (_, tx_descriptors) = esp_hal::dma_descriptors!(0, 2 * BUFFER_SIZE);
+
+        let pins = TxSixteenBits::new(
+            NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin,
+            NoPin, NoPin, NoPin, ctx.valid,
+        );
+        let mut pins = TxPinConfigIncludingValidPin::new(pins);
+        let mut clock_pin = ClkOutPin::new(ctx.clock);
+
+        let pio = ParlIoTxOnly::new(
+            ctx.parl_io,
+            ctx.dma_channel.configure(false, DmaPriority::Priority0),
+            tx_descriptors,
+            20.MHz(),
+        )
+        .unwrap();
+
+        let mut pio = pio
+            .tx
+            .with_config(
+                &mut pins,
+                &mut clock_pin,
+                0,
+                SampleEdge::Invert,
+                BitPackOrder::Msb,
+            )
+            .unwrap(); // TODO: handle error
+
+        // use a PCNT unit to count the negitive clock edges only when valid is high
+        let clock_unit = ctx.pcnt_unit;
+        clock_unit.channel0.set_edge_signal(ctx.clock_loopback);
+        clock_unit.channel0.set_ctrl_signal(ctx.valid_loopback);
+        clock_unit
+            .channel0
+            .set_input_mode(EdgeMode::Increment, EdgeMode::Hold);
+        clock_unit
+            .channel0
+            .set_ctrl_mode(CtrlMode::Disable, CtrlMode::Keep);
+
+        for _ in 0..100 {
+            clock_unit.clear();
+            let xfer = pio.write_dma(&tx_buffer).unwrap();
+            xfer.wait().unwrap();
+            info!("clock count: {}", clock_unit.get_value());
+            assert_eq!(clock_unit.get_value(), BUFFER_SIZE as _);
+        }
+    }
+
+    #[test]
+    #[timeout(3)]
+    fn test_parl_io_tx_8bit_valid_clock_count(ctx: Context) {
+        const BUFFER_SIZE: usize = 64;
+        let tx_buffer = [0u8; BUFFER_SIZE];
+        let (_, tx_descriptors) = esp_hal::dma_descriptors!(0, 2 * BUFFER_SIZE);
+
+        let pins = TxEightBits::new(NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin);
+        let mut pins = TxPinConfigWithValidPin::new(pins, ctx.valid);
+        let mut clock_pin = ClkOutPin::new(ctx.clock);
+
+        let pio = ParlIoTxOnly::new(
+            ctx.parl_io,
+            ctx.dma_channel.configure(false, DmaPriority::Priority0),
+            tx_descriptors,
+            20.MHz(),
+        )
+        .unwrap();
+
+        let mut pio = pio
+            .tx
+            .with_config(
+                &mut pins,
+                &mut clock_pin,
+                0,
+                SampleEdge::Invert,
+                BitPackOrder::Msb,
+            )
+            .unwrap(); // TODO: handle error
+
+        // use a PCNT unit to count the negitive clock edges only when valid is high
+        let clock_unit = ctx.pcnt_unit;
+        clock_unit.channel0.set_edge_signal(ctx.clock_loopback);
+        clock_unit.channel0.set_ctrl_signal(ctx.valid_loopback);
+        clock_unit
+            .channel0
+            .set_input_mode(EdgeMode::Increment, EdgeMode::Hold);
+        clock_unit
+            .channel0
+            .set_ctrl_mode(CtrlMode::Disable, CtrlMode::Keep);
+
+        for _ in 0..100 {
+            clock_unit.clear();
+            let xfer = pio.write_dma(&tx_buffer).unwrap();
+            xfer.wait().unwrap();
+            info!("clock count: {}", clock_unit.get_value());
+            assert_eq!(clock_unit.get_value(), BUFFER_SIZE as _);
+        }
+    }
+}

hil-test/tests/parl_io_tx_async.rs

@@ -0,0 +1,182 @@
+//! PARL_IO TX async test
+
+//% CHIPS: esp32c6 esp32h2
+//% FEATURES: generic-queue
+
+#![no_std]
+#![no_main]
+
+use esp_hal::{
+    dma::{ChannelCreator, Dma, DmaPriority},
+    gpio::{interconnect::InputSignal, AnyPin, Io, NoPin},
+    parl_io::{
+        BitPackOrder,
+        ClkOutPin,
+        ParlIoTxOnly,
+        SampleEdge,
+        TxEightBits,
+        TxPinConfigIncludingValidPin,
+        TxPinConfigWithValidPin,
+        TxSixteenBits,
+    },
+    pcnt::{
+        channel::{CtrlMode, EdgeMode},
+        unit::Unit,
+        Pcnt,
+    },
+    peripherals::PARL_IO,
+    prelude::*,
+};
+use hil_test as _;
+
+struct Context {
+    parl_io: PARL_IO,
+    dma_channel: ChannelCreator<0>,
+    clock: AnyPin,
+    valid: AnyPin,
+    clock_loopback: InputSignal,
+    valid_loopback: InputSignal,
+    pcnt_unit: Unit<'static, 0>,
+}
+
+#[cfg(test)]
+#[embedded_test::tests(executor = esp_hal_embassy::Executor::new())]
+mod tests {
+    // defmt::* is load-bearing, it ensures that the assert in dma_buffers! is not
+    // using defmt's non-const assert. Doing so would result in a compile error.
+    #[allow(unused_imports)]
+    use defmt::{assert_eq, *};
+
+    use super::*;
+
+    #[init]
+    async fn init() -> Context {
+        let peripherals = esp_hal::init(esp_hal::Config::default());
+
+        let io = Io::new(peripherals.GPIO, peripherals.IO_MUX);
+        let (clock, _) = hil_test::common_test_pins!(io);
+        let valid = io.pins.gpio0.degrade();
+        let clock_loopback = clock.peripheral_input();
+        let valid_loopback = valid.peripheral_input();
+        let clock = clock.degrade();
+        let pcnt = Pcnt::new(peripherals.PCNT);
+        let pcnt_unit = pcnt.unit0;
+        let dma = Dma::new(peripherals.DMA);
+        let dma_channel = dma.channel0;
+
+        let parl_io = peripherals.PARL_IO;
+
+        Context {
+            parl_io,
+            dma_channel,
+            clock,
+            valid,
+            clock_loopback,
+            valid_loopback,
+            pcnt_unit,
+        }
+    }
+
+    #[test]
+    #[timeout(3)]
+    async fn test_parl_io_tx_async_16bit_valid_clock_count(ctx: Context) {
+        const BUFFER_SIZE: usize = 64;
+        let tx_buffer = [0u16; BUFFER_SIZE];
+        let (_, tx_descriptors) = esp_hal::dma_descriptors!(0, 2 * BUFFER_SIZE);
+
+        let pins = TxSixteenBits::new(
+            NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin,
+            NoPin, NoPin, NoPin, ctx.valid,
+        );
+        let mut pins = TxPinConfigIncludingValidPin::new(pins);
+        let mut clock_pin = ClkOutPin::new(ctx.clock);
+
+        let pio = ParlIoTxOnly::new(
+            ctx.parl_io,
+            ctx.dma_channel
+                .configure_for_async(false, DmaPriority::Priority0),
+            tx_descriptors,
+            20.MHz(),
+        )
+        .unwrap();
+
+        let mut pio = pio
+            .tx
+            .with_config(
+                &mut pins,
+                &mut clock_pin,
+                0,
+                SampleEdge::Invert,
+                BitPackOrder::Msb,
+            )
+            .unwrap();
+
+        // use a PCNT unit to count the negitive clock edges only when valid is high
+        let clock_unit = ctx.pcnt_unit;
+        clock_unit.channel0.set_edge_signal(ctx.clock_loopback);
+        clock_unit.channel0.set_ctrl_signal(ctx.valid_loopback);
+        clock_unit
+            .channel0
+            .set_input_mode(EdgeMode::Increment, EdgeMode::Hold);
+        clock_unit
+            .channel0
+            .set_ctrl_mode(CtrlMode::Disable, CtrlMode::Keep);
+
+        for _ in 0..100 {
+            clock_unit.clear();
+            pio.write_dma_async(&tx_buffer).await.unwrap();
+            info!("clock count: {}", clock_unit.get_value());
+            assert_eq!(clock_unit.get_value(), BUFFER_SIZE as _);
+        }
+    }
+
+    #[test]
+    #[timeout(3)]
+    async fn test_parl_io_tx_async_8bit_valid_clock_count(ctx: Context) {
+        const BUFFER_SIZE: usize = 64;
+        let tx_buffer = [0u8; BUFFER_SIZE];
+        let (_, tx_descriptors) = esp_hal::dma_descriptors!(0, 2 * BUFFER_SIZE);
+
+        let pins = TxEightBits::new(NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin, NoPin);
+        let mut pins = TxPinConfigWithValidPin::new(pins, ctx.valid);
+        let mut clock_pin = ClkOutPin::new(ctx.clock);
+
+        let pio = ParlIoTxOnly::new(
+            ctx.parl_io,
+            ctx.dma_channel
+                .configure_for_async(false, DmaPriority::Priority0),
+            tx_descriptors,
+            20.MHz(),
+        )
+        .unwrap();
+
+        let mut pio = pio
+            .tx
+            .with_config(
+                &mut pins,
+                &mut clock_pin,
+                0,
+                SampleEdge::Invert,
+                BitPackOrder::Msb,
+            )
+            .unwrap();
+
+        // use a PCNT unit to count the negitive clock edges only when valid is high
+        let clock_unit = ctx.pcnt_unit;
+        clock_unit.channel0.set_edge_signal(ctx.clock_loopback);
+        clock_unit.channel0.set_ctrl_signal(ctx.valid_loopback);
+        clock_unit
+            .channel0
+            .set_input_mode(EdgeMode::Increment, EdgeMode::Hold);
+        clock_unit
+            .channel0
+            .set_ctrl_mode(CtrlMode::Disable, CtrlMode::Keep);
+
+        for _ in 0..100 {
+            clock_unit.clear();
+            pio.write_dma_async(&tx_buffer).await.unwrap();
+            info!("clock count: {}", clock_unit.get_value());
+            assert_eq!(clock_unit.get_value(), BUFFER_SIZE as _);
+        }
+    }
+}