diff --git a/embassy-embedded-hal/Cargo.toml b/embassy-embedded-hal/Cargo.toml
index 19d512585..ad2f14568 100644
--- a/embassy-embedded-hal/Cargo.toml
+++ b/embassy-embedded-hal/Cargo.toml
@@ -14,11 +14,14 @@ target = "x86_64-unknown-linux-gnu"
 [features]
 std = []
 # Enable nightly-only features
-nightly = ["embedded-hal-async", "embedded-storage-async"]
+nightly = ["embassy-futures", "embedded-hal-async", "embedded-storage-async"]
 
 [dependencies]
+embassy-futures = { version = "0.1.0", path = "../embassy-futures", optional = true }
 embassy-sync = { version = "0.2.0", path = "../embassy-sync" }
-embedded-hal-02 = { package = "embedded-hal", version = "0.2.6", features = ["unproven"] }
+embedded-hal-02 = { package = "embedded-hal", version = "0.2.6", features = [
+    "unproven",
+] }
 embedded-hal-1 = { package = "embedded-hal", version = "=1.0.0-alpha.10" }
 embedded-hal-async = { version = "=0.2.0-alpha.1", optional = true }
 embedded-storage = "0.3.0"
@@ -26,3 +29,6 @@ embedded-storage-async = { version = "0.4.0", optional = true }
 nb = "1.0.0"
 
 defmt = { version = "0.3", optional = true }
+
+[dev-dependencies]
+futures-test = "0.3.17"
diff --git a/embassy-embedded-hal/src/adapter.rs b/embassy-embedded-hal/src/adapter/blocking_async.rs
similarity index 99%
rename from embassy-embedded-hal/src/adapter.rs
rename to embassy-embedded-hal/src/adapter/blocking_async.rs
index 171ff6c9f..b996d6a75 100644
--- a/embassy-embedded-hal/src/adapter.rs
+++ b/embassy-embedded-hal/src/adapter/blocking_async.rs
@@ -1,5 +1,3 @@
-//! Adapters between embedded-hal traits.
-
 use embedded_hal_02::{blocking, serial};
 
 /// Wrapper that implements async traits using blocking implementations.
diff --git a/embassy-embedded-hal/src/adapter/mod.rs b/embassy-embedded-hal/src/adapter/mod.rs
new file mode 100644
index 000000000..28da56137
--- /dev/null
+++ b/embassy-embedded-hal/src/adapter/mod.rs
@@ -0,0 +1,7 @@
+//! Adapters between embedded-hal traits.
+
+mod blocking_async;
+mod yielding_async;
+
+pub use blocking_async::BlockingAsync;
+pub use yielding_async::YieldingAsync;
diff --git a/embassy-embedded-hal/src/adapter/yielding_async.rs b/embassy-embedded-hal/src/adapter/yielding_async.rs
new file mode 100644
index 000000000..96d5cca8e
--- /dev/null
+++ b/embassy-embedded-hal/src/adapter/yielding_async.rs
@@ -0,0 +1,232 @@
+use embassy_futures::yield_now;
+
+/// Wrapper that yields for each operation to the wrapped instance
+///
+/// This can be used in combination with BlockingAsync<T> to enforce yields
+/// between long running blocking operations.
+pub struct YieldingAsync<T> {
+    wrapped: T,
+}
+
+impl<T> YieldingAsync<T> {
+    /// Create a new instance of a wrapper that yields after each operation.
+    pub fn new(wrapped: T) -> Self {
+        Self { wrapped }
+    }
+}
+
+//
+// I2C implementations
+//
+impl<T> embedded_hal_1::i2c::ErrorType for YieldingAsync<T>
+where
+    T: embedded_hal_1::i2c::ErrorType,
+{
+    type Error = T::Error;
+}
+
+impl<T> embedded_hal_async::i2c::I2c for YieldingAsync<T>
+where
+    T: embedded_hal_async::i2c::I2c,
+{
+    async fn read(&mut self, address: u8, read: &mut [u8]) -> Result<(), Self::Error> {
+        self.wrapped.read(address, read).await?;
+        yield_now().await;
+        Ok(())
+    }
+
+    async fn write(&mut self, address: u8, write: &[u8]) -> Result<(), Self::Error> {
+        self.wrapped.write(address, write).await?;
+        yield_now().await;
+        Ok(())
+    }
+
+    async fn write_read(&mut self, address: u8, write: &[u8], read: &mut [u8]) -> Result<(), Self::Error> {
+        self.wrapped.write_read(address, write, read).await?;
+        yield_now().await;
+        Ok(())
+    }
+
+    async fn transaction(
+        &mut self,
+        address: u8,
+        operations: &mut [embedded_hal_1::i2c::Operation<'_>],
+    ) -> Result<(), Self::Error> {
+        self.wrapped.transaction(address, operations).await?;
+        yield_now().await;
+        Ok(())
+    }
+}
+
+//
+// SPI implementations
+//
+
+impl<T> embedded_hal_async::spi::ErrorType for YieldingAsync<T>
+where
+    T: embedded_hal_async::spi::ErrorType,
+{
+    type Error = T::Error;
+}
+
+impl<T> embedded_hal_async::spi::SpiBus<u8> for YieldingAsync<T>
+where
+    T: embedded_hal_async::spi::SpiBus,
+{
+    async fn transfer<'a>(&'a mut self, read: &'a mut [u8], write: &'a [u8]) -> Result<(), Self::Error> {
+        self.wrapped.transfer(read, write).await?;
+        yield_now().await;
+        Ok(())
+    }
+
+    async fn transfer_in_place<'a>(&'a mut self, words: &'a mut [u8]) -> Result<(), Self::Error> {
+        self.wrapped.transfer_in_place(words).await?;
+        yield_now().await;
+        Ok(())
+    }
+}
+
+impl<T> embedded_hal_async::spi::SpiBusFlush for YieldingAsync<T>
+where
+    T: embedded_hal_async::spi::SpiBusFlush,
+{
+    async fn flush(&mut self) -> Result<(), Self::Error> {
+        self.wrapped.flush().await?;
+        yield_now().await;
+        Ok(())
+    }
+}
+
+impl<T> embedded_hal_async::spi::SpiBusWrite<u8> for YieldingAsync<T>
+where
+    T: embedded_hal_async::spi::SpiBusWrite<u8>,
+{
+    async fn write(&mut self, data: &[u8]) -> Result<(), Self::Error> {
+        self.wrapped.write(data).await?;
+        yield_now().await;
+        Ok(())
+    }
+}
+
+impl<T> embedded_hal_async::spi::SpiBusRead<u8> for YieldingAsync<T>
+where
+    T: embedded_hal_async::spi::SpiBusRead<u8>,
+{
+    async fn read(&mut self, data: &mut [u8]) -> Result<(), Self::Error> {
+        self.wrapped.read(data).await?;
+        yield_now().await;
+        Ok(())
+    }
+}
+
+///
+/// NOR flash implementations
+///
+impl<T: embedded_storage::nor_flash::ErrorType> embedded_storage::nor_flash::ErrorType for YieldingAsync<T> {
+    type Error = T::Error;
+}
+
+impl<T: embedded_storage_async::nor_flash::ReadNorFlash> embedded_storage_async::nor_flash::ReadNorFlash
+    for YieldingAsync<T>
+{
+    const READ_SIZE: usize = T::READ_SIZE;
+
+    async fn read(&mut self, offset: u32, bytes: &mut [u8]) -> Result<(), Self::Error> {
+        self.wrapped.read(offset, bytes).await?;
+        Ok(())
+    }
+
+    fn capacity(&self) -> usize {
+        self.wrapped.capacity()
+    }
+}
+
+impl<T: embedded_storage_async::nor_flash::NorFlash> embedded_storage_async::nor_flash::NorFlash for YieldingAsync<T> {
+    const WRITE_SIZE: usize = T::WRITE_SIZE;
+    const ERASE_SIZE: usize = T::ERASE_SIZE;
+
+    async fn write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Self::Error> {
+        self.wrapped.write(offset, bytes).await?;
+        yield_now().await;
+        Ok(())
+    }
+
+    async fn erase(&mut self, from: u32, to: u32) -> Result<(), Self::Error> {
+        // Yield between each actual erase
+        for from in (from..to).step_by(T::ERASE_SIZE) {
+            let to = core::cmp::min(from + T::ERASE_SIZE as u32, to);
+            self.wrapped.erase(from, to).await?;
+            yield_now().await;
+        }
+        Ok(())
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use embedded_storage_async::nor_flash::NorFlash;
+
+    use super::*;
+
+    extern crate std;
+
+    #[derive(Default)]
+    struct FakeFlash(Vec<(u32, u32)>);
+
+    impl embedded_storage::nor_flash::ErrorType for FakeFlash {
+        type Error = std::convert::Infallible;
+    }
+
+    impl embedded_storage_async::nor_flash::ReadNorFlash for FakeFlash {
+        const READ_SIZE: usize = 1;
+
+        async fn read(&mut self, _offset: u32, _bytes: &mut [u8]) -> Result<(), Self::Error> {
+            unimplemented!()
+        }
+
+        fn capacity(&self) -> usize {
+            unimplemented!()
+        }
+    }
+
+    impl embedded_storage_async::nor_flash::NorFlash for FakeFlash {
+        const WRITE_SIZE: usize = 4;
+        const ERASE_SIZE: usize = 128;
+
+        async fn write(&mut self, _offset: u32, _bytes: &[u8]) -> Result<(), Self::Error> {
+            unimplemented!()
+        }
+
+        async fn erase(&mut self, from: u32, to: u32) -> Result<(), Self::Error> {
+            self.0.push((from, to));
+            Ok(())
+        }
+    }
+
+    #[futures_test::test]
+    async fn can_erase() {
+        let fake = FakeFlash::default();
+        let mut yielding = YieldingAsync::new(fake);
+
+        yielding.erase(0, 256).await.unwrap();
+
+        let fake = yielding.wrapped;
+        assert_eq!(2, fake.0.len());
+        assert_eq!((0, 128), fake.0[0]);
+        assert_eq!((128, 256), fake.0[1]);
+    }
+
+    #[futures_test::test]
+    async fn can_erase_wrong_erase_size() {
+        let fake = FakeFlash::default();
+        let mut yielding = YieldingAsync::new(fake);
+
+        yielding.erase(0, 257).await.unwrap();
+
+        let fake = yielding.wrapped;
+        assert_eq!(3, fake.0.len());
+        assert_eq!((0, 128), fake.0[0]);
+        assert_eq!((128, 256), fake.0[1]);
+        assert_eq!((256, 257), fake.0[2]);
+    }
+}
diff --git a/embassy-embedded-hal/src/flash.rs b/embassy-embedded-hal/src/flash.rs
new file mode 100644
index 000000000..9a6e4bd92
--- /dev/null
+++ b/embassy-embedded-hal/src/flash.rs
@@ -0,0 +1,286 @@
+//! Utilities related to flash.
+
+use embedded_storage::nor_flash::{ErrorType, NorFlash, NorFlashError, ReadNorFlash};
+#[cfg(feature = "nightly")]
+use embedded_storage_async::nor_flash::{NorFlash as AsyncNorFlash, ReadNorFlash as AsyncReadNorFlash};
+
+/// Convenience helper for concatenating two consecutive flashes into one.
+/// This is especially useful if used with "flash regions", where one may
+/// want to concatenate multiple regions into one larger region.
+pub struct ConcatFlash<First, Second>(First, Second);
+
+impl<First, Second> ConcatFlash<First, Second> {
+    /// Create a new flash that concatenates two consecutive flashes.
+    pub fn new(first: First, second: Second) -> Self {
+        Self(first, second)
+    }
+}
+
+const fn get_read_size(first_read_size: usize, second_read_size: usize) -> usize {
+    if first_read_size != second_read_size {
+        panic!("The read size for the concatenated flashes must be the same");
+    }
+    first_read_size
+}
+
+const fn get_write_size(first_write_size: usize, second_write_size: usize) -> usize {
+    if first_write_size != second_write_size {
+        panic!("The write size for the concatenated flashes must be the same");
+    }
+    first_write_size
+}
+
+const fn get_max_erase_size(first_erase_size: usize, second_erase_size: usize) -> usize {
+    let max_erase_size = if first_erase_size > second_erase_size {
+        first_erase_size
+    } else {
+        second_erase_size
+    };
+    if max_erase_size % first_erase_size != 0 || max_erase_size % second_erase_size != 0 {
+        panic!("The erase sizes for the concatenated flashes must have have a gcd equal to the max erase size");
+    }
+    max_erase_size
+}
+
+impl<First, Second, E> ErrorType for ConcatFlash<First, Second>
+where
+    First: ErrorType<Error = E>,
+    Second: ErrorType<Error = E>,
+    E: NorFlashError,
+{
+    type Error = E;
+}
+
+impl<First, Second, E> ReadNorFlash for ConcatFlash<First, Second>
+where
+    First: ReadNorFlash<Error = E>,
+    Second: ReadNorFlash<Error = E>,
+    E: NorFlashError,
+{
+    const READ_SIZE: usize = get_read_size(First::READ_SIZE, Second::READ_SIZE);
+
+    fn read(&mut self, mut offset: u32, mut bytes: &mut [u8]) -> Result<(), E> {
+        if offset < self.0.capacity() as u32 {
+            let len = core::cmp::min(self.0.capacity() - offset as usize, bytes.len());
+            self.0.read(offset, &mut bytes[..len])?;
+            offset += len as u32;
+            bytes = &mut bytes[len..];
+        }
+
+        if !bytes.is_empty() {
+            self.1.read(offset - self.0.capacity() as u32, bytes)?;
+        }
+
+        Ok(())
+    }
+
+    fn capacity(&self) -> usize {
+        self.0.capacity() + self.1.capacity()
+    }
+}
+
+impl<First, Second, E> NorFlash for ConcatFlash<First, Second>
+where
+    First: NorFlash<Error = E>,
+    Second: NorFlash<Error = E>,
+    E: NorFlashError,
+{
+    const WRITE_SIZE: usize = get_write_size(First::WRITE_SIZE, Second::WRITE_SIZE);
+    const ERASE_SIZE: usize = get_max_erase_size(First::ERASE_SIZE, Second::ERASE_SIZE);
+
+    fn write(&mut self, mut offset: u32, mut bytes: &[u8]) -> Result<(), E> {
+        if offset < self.0.capacity() as u32 {
+            let len = core::cmp::min(self.0.capacity() - offset as usize, bytes.len());
+            self.0.write(offset, &bytes[..len])?;
+            offset += len as u32;
+            bytes = &bytes[len..];
+        }
+
+        if !bytes.is_empty() {
+            self.1.write(offset - self.0.capacity() as u32, bytes)?;
+        }
+
+        Ok(())
+    }
+
+    fn erase(&mut self, mut from: u32, to: u32) -> Result<(), E> {
+        if from < self.0.capacity() as u32 {
+            let to = core::cmp::min(self.0.capacity() as u32, to);
+            self.0.erase(from, to)?;
+            from = self.0.capacity() as u32;
+        }
+
+        if from < to {
+            self.1
+                .erase(from - self.0.capacity() as u32, to - self.0.capacity() as u32)?;
+        }
+
+        Ok(())
+    }
+}
+
+#[cfg(feature = "nightly")]
+impl<First, Second, E> AsyncReadNorFlash for ConcatFlash<First, Second>
+where
+    First: AsyncReadNorFlash<Error = E>,
+    Second: AsyncReadNorFlash<Error = E>,
+    E: NorFlashError,
+{
+    const READ_SIZE: usize = get_read_size(First::READ_SIZE, Second::READ_SIZE);
+
+    async fn read(&mut self, mut offset: u32, mut bytes: &mut [u8]) -> Result<(), E> {
+        if offset < self.0.capacity() as u32 {
+            let len = core::cmp::min(self.0.capacity() - offset as usize, bytes.len());
+            self.0.read(offset, &mut bytes[..len]).await?;
+            offset += len as u32;
+            bytes = &mut bytes[len..];
+        }
+
+        if !bytes.is_empty() {
+            self.1.read(offset - self.0.capacity() as u32, bytes).await?;
+        }
+
+        Ok(())
+    }
+
+    fn capacity(&self) -> usize {
+        self.0.capacity() + self.1.capacity()
+    }
+}
+
+#[cfg(feature = "nightly")]
+impl<First, Second, E> AsyncNorFlash for ConcatFlash<First, Second>
+where
+    First: AsyncNorFlash<Error = E>,
+    Second: AsyncNorFlash<Error = E>,
+    E: NorFlashError,
+{
+    const WRITE_SIZE: usize = get_write_size(First::WRITE_SIZE, Second::WRITE_SIZE);
+    const ERASE_SIZE: usize = get_max_erase_size(First::ERASE_SIZE, Second::ERASE_SIZE);
+
+    async fn write(&mut self, mut offset: u32, mut bytes: &[u8]) -> Result<(), E> {
+        if offset < self.0.capacity() as u32 {
+            let len = core::cmp::min(self.0.capacity() - offset as usize, bytes.len());
+            self.0.write(offset, &bytes[..len]).await?;
+            offset += len as u32;
+            bytes = &bytes[len..];
+        }
+
+        if !bytes.is_empty() {
+            self.1.write(offset - self.0.capacity() as u32, bytes).await?;
+        }
+
+        Ok(())
+    }
+
+    async fn erase(&mut self, mut from: u32, to: u32) -> Result<(), E> {
+        if from < self.0.capacity() as u32 {
+            let to = core::cmp::min(self.0.capacity() as u32, to);
+            self.0.erase(from, to).await?;
+            from = self.0.capacity() as u32;
+        }
+
+        if from < to {
+            self.1
+                .erase(from - self.0.capacity() as u32, to - self.0.capacity() as u32)
+                .await?;
+        }
+
+        Ok(())
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn can_write_and_read_across_flashes() {
+        let first = MemFlash::<64, 16, 4>::new();
+        let second = MemFlash::<64, 64, 4>::new();
+        let mut f = ConcatFlash::new(first, second);
+
+        f.write(60, &[0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77, 0x88]).unwrap();
+
+        assert_eq!(&[0x11, 0x22, 0x33, 0x44], &f.0 .0[60..]);
+        assert_eq!(&[0x55, 0x66, 0x77, 0x88], &f.1 .0[0..4]);
+
+        let mut read_buf = [0; 8];
+        f.read(60, &mut read_buf).unwrap();
+
+        assert_eq!(&[0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77, 0x88], &read_buf);
+    }
+
+    #[test]
+    fn can_erase_across_flashes() {
+        let mut first = MemFlash::<128, 16, 4>::new();
+        let mut second = MemFlash::<128, 64, 4>::new();
+        first.0.fill(0x00);
+        second.0.fill(0x00);
+
+        let mut f = ConcatFlash::new(first, second);
+
+        f.erase(64, 192).unwrap();
+
+        assert_eq!(&[0x00; 64], &f.0 .0[0..64]);
+        assert_eq!(&[0xff; 64], &f.0 .0[64..128]);
+        assert_eq!(&[0xff; 64], &f.1 .0[0..64]);
+        assert_eq!(&[0x00; 64], &f.1 .0[64..128]);
+    }
+
+    pub struct MemFlash<const SIZE: usize, const ERASE_SIZE: usize, const WRITE_SIZE: usize>([u8; SIZE]);
+
+    impl<const SIZE: usize, const ERASE_SIZE: usize, const WRITE_SIZE: usize> MemFlash<SIZE, ERASE_SIZE, WRITE_SIZE> {
+        pub const fn new() -> Self {
+            Self([0xff; SIZE])
+        }
+    }
+
+    impl<const SIZE: usize, const ERASE_SIZE: usize, const WRITE_SIZE: usize> ErrorType
+        for MemFlash<SIZE, ERASE_SIZE, WRITE_SIZE>
+    {
+        type Error = core::convert::Infallible;
+    }
+
+    impl<const SIZE: usize, const ERASE_SIZE: usize, const WRITE_SIZE: usize> ReadNorFlash
+        for MemFlash<SIZE, ERASE_SIZE, WRITE_SIZE>
+    {
+        const READ_SIZE: usize = 1;
+
+        fn read(&mut self, offset: u32, bytes: &mut [u8]) -> Result<(), Self::Error> {
+            let len = bytes.len();
+            bytes.copy_from_slice(&self.0[offset as usize..offset as usize + len]);
+            Ok(())
+        }
+
+        fn capacity(&self) -> usize {
+            SIZE
+        }
+    }
+
+    impl<const SIZE: usize, const ERASE_SIZE: usize, const WRITE_SIZE: usize> NorFlash
+        for MemFlash<SIZE, ERASE_SIZE, WRITE_SIZE>
+    {
+        const WRITE_SIZE: usize = WRITE_SIZE;
+        const ERASE_SIZE: usize = ERASE_SIZE;
+
+        fn erase(&mut self, from: u32, to: u32) -> Result<(), Self::Error> {
+            let from = from as usize;
+            let to = to as usize;
+            assert_eq!(0, from % ERASE_SIZE);
+            assert_eq!(0, to % ERASE_SIZE);
+            self.0[from..to].fill(0xff);
+            Ok(())
+        }
+
+        fn write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Self::Error> {
+            let offset = offset as usize;
+            assert_eq!(0, bytes.len() % WRITE_SIZE);
+            assert_eq!(0, offset % WRITE_SIZE);
+            assert!(offset + bytes.len() <= SIZE);
+
+            self.0[offset..offset + bytes.len()].copy_from_slice(bytes);
+            Ok(())
+        }
+    }
+}
diff --git a/embassy-embedded-hal/src/lib.rs b/embassy-embedded-hal/src/lib.rs
index 73c81b465..3aad838bd 100644
--- a/embassy-embedded-hal/src/lib.rs
+++ b/embassy-embedded-hal/src/lib.rs
@@ -7,6 +7,8 @@
 #[cfg(feature = "nightly")]
 pub mod adapter;
 
+pub mod flash;
+
 pub mod shared_bus;
 
 /// Set the configuration of a peripheral driver.
diff --git a/embassy-stm32/build.rs b/embassy-stm32/build.rs
index dcee535b5..540981727 100644
--- a/embassy-stm32/build.rs
+++ b/embassy-stm32/build.rs
@@ -206,6 +206,7 @@ fn main() {
                 erase_size: #erase_size,
                 write_size: #write_size,
                 erase_value: #erase_value,
+                _ensure_internal: (),
             };
         });
 
diff --git a/embassy-stm32/src/flash/common.rs b/embassy-stm32/src/flash/common.rs
index 6d7f55974..432c8a43b 100644
--- a/embassy-stm32/src/flash/common.rs
+++ b/embassy-stm32/src/flash/common.rs
@@ -17,6 +17,7 @@ impl<'d> Flash<'d> {
     }
 
     pub fn into_regions(self) -> FlashLayout<'d> {
+        family::set_default_layout();
         FlashLayout::new(self.release())
     }
 
@@ -25,11 +26,11 @@ impl<'d> Flash<'d> {
     }
 
     pub fn blocking_write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Error> {
-        unsafe { blocking_write(FLASH_BASE as u32, FLASH_SIZE as u32, offset, bytes) }
+        unsafe { blocking_write_chunked(FLASH_BASE as u32, FLASH_SIZE as u32, offset, bytes) }
     }
 
     pub fn blocking_erase(&mut self, from: u32, to: u32) -> Result<(), Error> {
-        unsafe { blocking_erase(FLASH_BASE as u32, from, to) }
+        unsafe { blocking_erase_sectored(FLASH_BASE as u32, from, to) }
     }
 
     pub(crate) fn release(self) -> PeripheralRef<'d, crate::peripherals::FLASH> {
@@ -37,7 +38,7 @@ impl<'d> Flash<'d> {
     }
 }
 
-fn blocking_read(base: u32, size: u32, offset: u32, bytes: &mut [u8]) -> Result<(), Error> {
+pub(super) fn blocking_read(base: u32, size: u32, offset: u32, bytes: &mut [u8]) -> Result<(), Error> {
     if offset + bytes.len() as u32 > size {
         return Err(Error::Size);
     }
@@ -48,7 +49,7 @@ fn blocking_read(base: u32, size: u32, offset: u32, bytes: &mut [u8]) -> Result<
     Ok(())
 }
 
-unsafe fn blocking_write(base: u32, size: u32, offset: u32, bytes: &[u8]) -> Result<(), Error> {
+pub(super) unsafe fn blocking_write_chunked(base: u32, size: u32, offset: u32, bytes: &[u8]) -> Result<(), Error> {
     if offset + bytes.len() as u32 > size {
         return Err(Error::Size);
     }
@@ -81,7 +82,7 @@ unsafe fn blocking_write(base: u32, size: u32, offset: u32, bytes: &[u8]) -> Res
     Ok(())
 }
 
-unsafe fn blocking_erase(base: u32, from: u32, to: u32) -> Result<(), Error> {
+pub(super) unsafe fn blocking_erase_sectored(base: u32, from: u32, to: u32) -> Result<(), Error> {
     let start_address = base + from;
     let end_address = base + to;
     let regions = family::get_flash_regions();
@@ -154,11 +155,11 @@ impl FlashRegion {
     }
 
     pub fn blocking_write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Error> {
-        unsafe { blocking_write(self.base, self.size, offset, bytes) }
+        unsafe { blocking_write_chunked(self.base, self.size, offset, bytes) }
     }
 
     pub fn blocking_erase(&mut self, from: u32, to: u32) -> Result<(), Error> {
-        unsafe { blocking_erase(self.base, from, to) }
+        unsafe { blocking_erase_sectored(self.base, from, to) }
     }
 }
 
@@ -199,11 +200,11 @@ foreach_flash_region! {
             }
 
             pub fn blocking_write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Error> {
-                unsafe { blocking_write(self.0.base, self.0.size, offset, bytes) }
+                unsafe { blocking_write_chunked(self.0.base, self.0.size, offset, bytes) }
             }
 
             pub fn blocking_erase(&mut self, from: u32, to: u32) -> Result<(), Error> {
-                unsafe { blocking_erase(self.0.base, from, to) }
+                unsafe { blocking_erase_sectored(self.0.base, from, to) }
             }
         }
 
diff --git a/embassy-stm32/src/flash/f0.rs b/embassy-stm32/src/flash/f0.rs
index 033afecd5..c6441ef16 100644
--- a/embassy-stm32/src/flash/f0.rs
+++ b/embassy-stm32/src/flash/f0.rs
@@ -7,6 +7,8 @@ use super::{FlashRegion, FlashSector, FLASH_REGIONS, WRITE_SIZE};
 use crate::flash::Error;
 use crate::pac;
 
+pub const fn set_default_layout() {}
+
 pub const fn get_flash_regions() -> &'static [&'static FlashRegion] {
     &FLASH_REGIONS
 }
diff --git a/embassy-stm32/src/flash/f3.rs b/embassy-stm32/src/flash/f3.rs
index 10a09c42c..4c8172203 100644
--- a/embassy-stm32/src/flash/f3.rs
+++ b/embassy-stm32/src/flash/f3.rs
@@ -7,6 +7,8 @@ use super::{FlashRegion, FlashSector, FLASH_REGIONS, WRITE_SIZE};
 use crate::flash::Error;
 use crate::pac;
 
+pub const fn set_default_layout() {}
+
 pub const fn get_flash_regions() -> &'static [&'static FlashRegion] {
     &FLASH_REGIONS
 }
diff --git a/embassy-stm32/src/flash/f4.rs b/embassy-stm32/src/flash/f4.rs
index 60ac62c17..7f1c5f671 100644
--- a/embassy-stm32/src/flash/f4.rs
+++ b/embassy-stm32/src/flash/f4.rs
@@ -11,8 +11,11 @@ mod alt_regions {
     use embassy_hal_common::PeripheralRef;
     use stm32_metapac::FLASH_SIZE;
 
-    use crate::_generated::flash_regions::{BANK1_REGION1, BANK1_REGION2, BANK1_REGION3};
-    use crate::flash::{Bank1Region1, Bank1Region2, Flash, FlashBank, FlashRegion};
+    use crate::_generated::flash_regions::{OTPRegion, BANK1_REGION1, BANK1_REGION2, BANK1_REGION3, OTP_REGION};
+    use crate::flash::{
+        blocking_erase_sectored, blocking_read, blocking_write_chunked, Bank1Region1, Bank1Region2, Error, Flash,
+        FlashBank, FlashRegion,
+    };
     use crate::peripherals::FLASH;
 
     pub const ALT_BANK1_REGION3: FlashRegion = FlashRegion {
@@ -45,20 +48,19 @@ mod alt_regions {
         &ALT_BANK2_REGION3,
     ];
 
-    pub type AltBank1Region1<'d> = Bank1Region1<'d>;
-    pub type AltBank1Region2<'d> = Bank1Region2<'d>;
     pub struct AltBank1Region3<'d>(pub &'static FlashRegion, PeripheralRef<'d, FLASH>);
     pub struct AltBank2Region1<'d>(pub &'static FlashRegion, PeripheralRef<'d, FLASH>);
     pub struct AltBank2Region2<'d>(pub &'static FlashRegion, PeripheralRef<'d, FLASH>);
     pub struct AltBank2Region3<'d>(pub &'static FlashRegion, PeripheralRef<'d, FLASH>);
 
     pub struct AltFlashLayout<'d> {
-        pub bank1_region1: AltBank1Region1<'d>,
-        pub bank1_region2: AltBank1Region2<'d>,
+        pub bank1_region1: Bank1Region1<'d>,
+        pub bank1_region2: Bank1Region2<'d>,
         pub bank1_region3: AltBank1Region3<'d>,
         pub bank2_region1: AltBank2Region1<'d>,
         pub bank2_region2: AltBank2Region2<'d>,
         pub bank2_region3: AltBank2Region3<'d>,
+        pub otp_region: OTPRegion<'d>,
     }
 
     impl<'d> Flash<'d> {
@@ -66,7 +68,7 @@ mod alt_regions {
             unsafe { crate::pac::FLASH.optcr().modify(|r| r.set_db1m(true)) };
 
             // SAFETY: We never expose the cloned peripheral references, and their instance is not public.
-            // Also, all flash region operations are protected with a cs.
+            // Also, all blocking flash region operations are protected with a cs.
             let p = self.release();
             AltFlashLayout {
                 bank1_region1: Bank1Region1(&BANK1_REGION1, unsafe { p.clone_unchecked() }),
@@ -75,22 +77,74 @@ mod alt_regions {
                 bank2_region1: AltBank2Region1(&ALT_BANK2_REGION1, unsafe { p.clone_unchecked() }),
                 bank2_region2: AltBank2Region2(&ALT_BANK2_REGION2, unsafe { p.clone_unchecked() }),
                 bank2_region3: AltBank2Region3(&ALT_BANK2_REGION3, unsafe { p.clone_unchecked() }),
+                otp_region: OTPRegion(&OTP_REGION, unsafe { p.clone_unchecked() }),
             }
         }
     }
 
-    impl Drop for AltFlashLayout<'_> {
-        fn drop(&mut self) {
-            unsafe {
-                super::lock();
-                crate::pac::FLASH.optcr().modify(|r| r.set_db1m(false))
-            };
-        }
+    macro_rules! foreach_altflash_region {
+        ($type_name:ident, $region:ident) => {
+            impl $type_name<'_> {
+                pub fn blocking_read(&mut self, offset: u32, bytes: &mut [u8]) -> Result<(), Error> {
+                    blocking_read(self.0.base, self.0.size, offset, bytes)
+                }
+
+                pub fn blocking_write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Error> {
+                    unsafe { blocking_write_chunked(self.0.base, self.0.size, offset, bytes) }
+                }
+
+                pub fn blocking_erase(&mut self, from: u32, to: u32) -> Result<(), Error> {
+                    unsafe { blocking_erase_sectored(self.0.base, from, to) }
+                }
+            }
+
+            impl embedded_storage::nor_flash::ErrorType for $type_name<'_> {
+                type Error = Error;
+            }
+
+            impl embedded_storage::nor_flash::ReadNorFlash for $type_name<'_> {
+                const READ_SIZE: usize = 1;
+
+                fn read(&mut self, offset: u32, bytes: &mut [u8]) -> Result<(), Self::Error> {
+                    self.blocking_read(offset, bytes)
+                }
+
+                fn capacity(&self) -> usize {
+                    self.0.size as usize
+                }
+            }
+
+            impl embedded_storage::nor_flash::NorFlash for $type_name<'_> {
+                const WRITE_SIZE: usize = $region.write_size as usize;
+                const ERASE_SIZE: usize = $region.erase_size as usize;
+
+                fn write(&mut self, offset: u32, bytes: &[u8]) -> Result<(), Self::Error> {
+                    self.blocking_write(offset, bytes)
+                }
+
+                fn erase(&mut self, from: u32, to: u32) -> Result<(), Self::Error> {
+                    self.blocking_erase(from, to)
+                }
+            }
+        };
     }
+
+    foreach_altflash_region!(AltBank1Region3, ALT_BANK1_REGION3);
+    foreach_altflash_region!(AltBank2Region1, ALT_BANK2_REGION1);
+    foreach_altflash_region!(AltBank2Region2, ALT_BANK2_REGION2);
+    foreach_altflash_region!(AltBank2Region3, ALT_BANK2_REGION3);
 }
 
 #[cfg(any(stm32f427, stm32f429, stm32f437, stm32f439, stm32f469, stm32f479))]
-pub use alt_regions::{AltFlashLayout, ALT_FLASH_REGIONS};
+pub use alt_regions::*;
+
+#[cfg(any(stm32f427, stm32f429, stm32f437, stm32f439, stm32f469, stm32f479))]
+pub fn set_default_layout() {
+    unsafe { crate::pac::FLASH.optcr().modify(|r| r.set_db1m(false)) };
+}
+
+#[cfg(not(any(stm32f427, stm32f429, stm32f437, stm32f439, stm32f469, stm32f479)))]
+pub const fn set_default_layout() {}
 
 #[cfg(any(stm32f427, stm32f429, stm32f437, stm32f439, stm32f469, stm32f479))]
 pub fn get_flash_regions() -> &'static [&'static FlashRegion] {
diff --git a/embassy-stm32/src/flash/f7.rs b/embassy-stm32/src/flash/f7.rs
index 6427d5a09..ac2834a84 100644
--- a/embassy-stm32/src/flash/f7.rs
+++ b/embassy-stm32/src/flash/f7.rs
@@ -6,6 +6,8 @@ use super::{FlashRegion, FlashSector, FLASH_REGIONS, WRITE_SIZE};
 use crate::flash::Error;
 use crate::pac;
 
+pub const fn set_default_layout() {}
+
 pub const fn get_flash_regions() -> &'static [&'static FlashRegion] {
     &FLASH_REGIONS
 }
diff --git a/embassy-stm32/src/flash/h7.rs b/embassy-stm32/src/flash/h7.rs
index 4f38d50c0..1b1631068 100644
--- a/embassy-stm32/src/flash/h7.rs
+++ b/embassy-stm32/src/flash/h7.rs
@@ -7,6 +7,8 @@ use super::{FlashRegion, FlashSector, BANK1_REGION, FLASH_REGIONS, WRITE_SIZE};
 use crate::flash::Error;
 use crate::pac;
 
+pub const fn set_default_layout() {}
+
 const fn is_dual_bank() -> bool {
     FLASH_REGIONS.len() == 2
 }
diff --git a/embassy-stm32/src/flash/l.rs b/embassy-stm32/src/flash/l.rs
index 7d9cc6ea3..c94f61900 100644
--- a/embassy-stm32/src/flash/l.rs
+++ b/embassy-stm32/src/flash/l.rs
@@ -6,6 +6,8 @@ use super::{FlashRegion, FlashSector, FLASH_REGIONS, WRITE_SIZE};
 use crate::flash::Error;
 use crate::pac;
 
+pub const fn set_default_layout() {}
+
 pub const fn get_flash_regions() -> &'static [&'static FlashRegion] {
     &FLASH_REGIONS
 }
diff --git a/embassy-stm32/src/flash/mod.rs b/embassy-stm32/src/flash/mod.rs
index f6efa7753..b93270ae1 100644
--- a/embassy-stm32/src/flash/mod.rs
+++ b/embassy-stm32/src/flash/mod.rs
@@ -19,6 +19,7 @@ pub struct FlashRegion {
     pub erase_size: u32,
     pub write_size: u32,
     pub erase_value: u8,
+    pub(crate) _ensure_internal: (),
 }
 
 #[derive(Debug, PartialEq)]
diff --git a/embassy-stm32/src/flash/other.rs b/embassy-stm32/src/flash/other.rs
index c151cb828..556034654 100644
--- a/embassy-stm32/src/flash/other.rs
+++ b/embassy-stm32/src/flash/other.rs
@@ -2,6 +2,8 @@
 
 use super::{Error, FlashRegion, FlashSector, FLASH_REGIONS, WRITE_SIZE};
 
+pub const fn set_default_layout() {}
+
 pub const fn get_flash_regions() -> &'static [&'static FlashRegion] {
     &FLASH_REGIONS
 }