50 files changed, 663 insertions, 518 deletions

diff --git a/docs/modules/ROOT/examples/basic/Cargo.toml b/docs/modules/ROOT/examples/basic/Cargo.toml
index d9f8a285a..e3e446e63 100644
--- a/docs/modules/ROOT/examples/basic/Cargo.toml
+++ b/docs/modules/ROOT/examples/basic/Cargo.toml
@@ -6,7 +6,7 @@ version = "0.1.0"
 license = "MIT OR Apache-2.0"
 
 [dependencies]
-embassy-executor = { version = "0.1.0", path = "../../../../../embassy-executor", features = ["defmt", "nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../../embassy-executor", features = ["defmt", "nightly", "integrated-timers", "arch-cortex-m", "executor-thread"] }
 embassy-time = { version = "0.1.0", path = "../../../../../embassy-time", features = ["defmt", "nightly"] }
 embassy-nrf = { version = "0.1.0", path = "../../../../../embassy-nrf", features = ["defmt", "nrf52840", "time-driver-rtc1", "gpiote", "nightly"] }
 
diff --git a/docs/modules/ROOT/examples/layer-by-layer/blinky-async/Cargo.toml b/docs/modules/ROOT/examples/layer-by-layer/blinky-async/Cargo.toml
index c9a963d4d..a11a7e0ba 100644
--- a/docs/modules/ROOT/examples/layer-by-layer/blinky-async/Cargo.toml
+++ b/docs/modules/ROOT/examples/layer-by-layer/blinky-async/Cargo.toml
@@ -8,7 +8,7 @@ license = "MIT OR Apache-2.0"
 cortex-m = "0.7"
 cortex-m-rt = "0.7"
 embassy-stm32 = { version = "0.1.0", features = ["stm32l475vg", "memory-x", "exti"], default-features = false  }
-embassy-executor = { version = "0.1.0", default-features = false, features = ["nightly"] }
+embassy-executor = { version = "0.1.0", default-features = false, features = ["nightly", "arch-cortex-m", "executor-thread"] }
 
 defmt = "0.3.0"
 defmt-rtt = "0.3.0"
diff --git a/embassy-cortex-m/src/executor.rs b/embassy-cortex-m/src/executor.rs
deleted file mode 100644
index 558539e73..000000000
--- a/embassy-cortex-m/src/executor.rs
+++ /dev/null
@@ -1,116 +0,0 @@
-//! Executor specific to cortex-m devices.
-
-use core::cell::UnsafeCell;
-use core::mem::MaybeUninit;
-
-use atomic_polyfill::{AtomicBool, Ordering};
-use cortex_m::interrupt::InterruptNumber;
-use cortex_m::peripheral::NVIC;
-pub use embassy_executor::*;
-
-#[derive(Clone, Copy)]
-struct N(u16);
-unsafe impl cortex_m::interrupt::InterruptNumber for N {
-    fn number(self) -> u16 {
-        self.0
-    }
-}
-
-fn pend_by_number(n: u16) {
-    cortex_m::peripheral::NVIC::pend(N(n))
-}
-
-/// Interrupt mode executor.
-///
-/// This executor runs tasks in interrupt mode. The interrupt handler is set up
-/// to poll tasks, and when a task is woken the interrupt is pended from software.
-///
-/// This allows running async tasks at a priority higher than thread mode. One
-/// use case is to leave thread mode free for non-async tasks. Another use case is
-/// to run multiple executors: one in thread mode for low priority tasks and another in
-/// interrupt mode for higher priority tasks. Higher priority tasks will preempt lower
-/// priority ones.
-///
-/// It is even possible to run multiple interrupt mode executors at different priorities,
-/// by assigning different priorities to the interrupts. For an example on how to do this,
-/// See the 'multiprio' example for 'embassy-nrf'.
-///
-/// To use it, you have to pick an interrupt that won't be used by the hardware.
-/// Some chips reserve some interrupts for this purpose, sometimes named "software interrupts" (SWI).
-/// If this is not the case, you may use an interrupt from any unused peripheral.
-///
-/// It is somewhat more complex to use, it's recommended to use the thread-mode
-/// [`Executor`] instead, if it works for your use case.
-pub struct InterruptExecutor {
-    started: AtomicBool,
-    executor: UnsafeCell<MaybeUninit<raw::Executor>>,
-}
-
-unsafe impl Send for InterruptExecutor {}
-unsafe impl Sync for InterruptExecutor {}
-
-impl InterruptExecutor {
-    /// Create a new, not started `InterruptExecutor`.
-    #[inline]
-    pub const fn new() -> Self {
-        Self {
-            started: AtomicBool::new(false),
-            executor: UnsafeCell::new(MaybeUninit::uninit()),
-        }
-    }
-
-    /// Executor interrupt callback.
-    ///
-    /// # Safety
-    ///
-    /// You MUST call this from the interrupt handler, and from nowhere else.
-    pub unsafe fn on_interrupt(&'static self) {
-        let executor = unsafe { (&*self.executor.get()).assume_init_ref() };
-        executor.poll();
-    }
-
-    /// Start the executor.
-    ///
-    /// This initializes the executor, enables the interrupt, and returns.
-    /// The executor keeps running in the background through the interrupt.
-    ///
-    /// This returns a [`SendSpawner`] you can use to spawn tasks on it. A [`SendSpawner`]
-    /// is returned instead of a [`Spawner`](embassy_executor::Spawner) because the executor effectively runs in a
-    /// different "thread" (the interrupt), so spawning tasks on it is effectively
-    /// sending them.
-    ///
-    /// To obtain a [`Spawner`](embassy_executor::Spawner) for this executor, use [`Spawner::for_current_executor()`](embassy_executor::Spawner::for_current_executor()) from
-    /// a task running in it.
-    ///
-    /// # Interrupt requirements
-    ///
-    /// You must write the interrupt handler yourself, and make it call [`on_interrupt()`](Self::on_interrupt).
-    ///
-    /// This method already enables (unmasks) the interrupt, you must NOT do it yourself.
-    ///
-    /// You must set the interrupt priority before calling this method. You MUST NOT
-    /// do it after.
-    ///
-    pub fn start(&'static self, irq: impl InterruptNumber) -> SendSpawner {
-        if self
-            .started
-            .compare_exchange(false, true, Ordering::Acquire, Ordering::Relaxed)
-            .is_err()
-        {
-            panic!("InterruptExecutor::start() called multiple times on the same executor.");
-        }
-
-        unsafe {
-            (&mut *self.executor.get()).as_mut_ptr().write(raw::Executor::new(
-                |ctx| pend_by_number(ctx as u16),
-                irq.number() as *mut (),
-            ))
-        }
-
-        let executor = unsafe { (&*self.executor.get()).assume_init_ref() };
-
-        unsafe { NVIC::unmask(irq) }
-
-        executor.spawner().make_send()
-    }
-}
diff --git a/embassy-cortex-m/src/lib.rs b/embassy-cortex-m/src/lib.rs
index fba23367b..e4b713a06 100644
--- a/embassy-cortex-m/src/lib.rs
+++ b/embassy-cortex-m/src/lib.rs
@@ -5,6 +5,6 @@
 // This mod MUST go first, so that the others see its macros.
 pub(crate) mod fmt;
 
-pub mod executor;
+pub use embassy_executor as executor;
 pub mod interrupt;
 pub mod peripheral;
diff --git a/embassy-executor/Cargo.toml b/embassy-executor/Cargo.toml
index 8ad3fd698..bb8a46c82 100644
--- a/embassy-executor/Cargo.toml
+++ b/embassy-executor/Cargo.toml
@@ -31,9 +31,22 @@ flavors = [
 features = ["std", "nightly", "defmt"]
 
 [features]
-default = []
-std = ["critical-section/std"]
-wasm = ["dep:wasm-bindgen", "dep:js-sys"]
+
+# Architecture
+_arch = [] # some arch was picked
+arch-std = ["_arch", "critical-section/std"]
+arch-cortex-m = ["_arch", "dep:cortex-m"]
+arch-xtensa = ["_arch"]
+arch-riscv32 = ["_arch"]
+arch-wasm = ["_arch", "dep:wasm-bindgen", "dep:js-sys"]
+
+# Enable creating a `Pender` from an arbitrary function pointer callback.
+pender-callback = []
+
+# Enable the thread-mode executor (using WFE/SEV in Cortex-M, WFI in other embedded archs)
+executor-thread = []
+# Enable the interrupt-mode executor (available in Cortex-M only)
+executor-interrupt = []
 
 # Enable nightly-only features
 nightly = []
@@ -55,9 +68,11 @@ embassy-macros  = { version = "0.1.0", path = "../embassy-macros" }
 embassy-time  = { version = "0.1.0", path = "../embassy-time", optional = true}
 atomic-polyfill = "1.0.1"
 critical-section = "1.1"
-cfg-if = "1.0.0"
 static_cell = "1.0"
 
-# WASM dependencies
+# arch-cortex-m dependencies
+cortex-m = { version = "0.7.6", optional = true }
+
+# arch-wasm dependencies
 wasm-bindgen = { version = "0.2.82", optional = true }
 js-sys = { version = "0.3", optional = true }
diff --git a/embassy-executor/src/arch/cortex_m.rs b/embassy-executor/src/arch/cortex_m.rs
index 4b27a264e..d6a55c4c7 100644
--- a/embassy-executor/src/arch/cortex_m.rs
+++ b/embassy-executor/src/arch/cortex_m.rs
@@ -1,59 +1,209 @@
-use core::arch::asm;
-use core::marker::PhantomData;
-use core::ptr;
-
-use super::{raw, Spawner};
-
-/// Thread mode executor, using WFE/SEV.
-///
-/// This is the simplest and most common kind of executor. It runs on
-/// thread mode (at the lowest priority level), and uses the `WFE` ARM instruction
-/// to sleep when it has no more work to do. When a task is woken, a `SEV` instruction
-/// is executed, to make the `WFE` exit from sleep and poll the task.
-///
-/// This executor allows for ultra low power consumption for chips where `WFE`
-/// triggers low-power sleep without extra steps. If your chip requires extra steps,
-/// you may use [`raw::Executor`] directly to program custom behavior.
-pub struct Executor {
-    inner: raw::Executor,
-    not_send: PhantomData<*mut ()>,
+#[cfg(feature = "executor-thread")]
+pub use thread::*;
+#[cfg(feature = "executor-thread")]
+mod thread {
+    use core::arch::asm;
+    use core::marker::PhantomData;
+
+    #[cfg(feature = "nightly")]
+    pub use embassy_macros::main_cortex_m as main;
+
+    use crate::raw::{Pender, PenderInner};
+    use crate::{raw, Spawner};
+
+    #[derive(Copy, Clone)]
+    pub(crate) struct ThreadPender;
+
+    impl ThreadPender {
+        pub(crate) fn pend(self) {
+            unsafe { core::arch::asm!("sev") }
+        }
+    }
+
+    /// Thread mode executor, using WFE/SEV.
+    ///
+    /// This is the simplest and most common kind of executor. It runs on
+    /// thread mode (at the lowest priority level), and uses the `WFE` ARM instruction
+    /// to sleep when it has no more work to do. When a task is woken, a `SEV` instruction
+    /// is executed, to make the `WFE` exit from sleep and poll the task.
+    ///
+    /// This executor allows for ultra low power consumption for chips where `WFE`
+    /// triggers low-power sleep without extra steps. If your chip requires extra steps,
+    /// you may use [`raw::Executor`] directly to program custom behavior.
+    pub struct Executor {
+        inner: raw::Executor,
+        not_send: PhantomData<*mut ()>,
+    }
+
+    impl Executor {
+        /// Create a new Executor.
+        pub fn new() -> Self {
+            Self {
+                inner: raw::Executor::new(Pender(PenderInner::Thread(ThreadPender))),
+                not_send: PhantomData,
+            }
+        }
+
+        /// Run the executor.
+        ///
+        /// The `init` closure is called with a [`Spawner`] that spawns tasks on
+        /// this executor. Use it to spawn the initial task(s). After `init` returns,
+        /// the executor starts running the tasks.
+        ///
+        /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
+        /// for example by passing it as an argument to the initial tasks.
+        ///
+        /// This function requires `&'static mut self`. This means you have to store the
+        /// Executor instance in a place where it'll live forever and grants you mutable
+        /// access. There's a few ways to do this:
+        ///
+        /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
+        /// - a `static mut` (unsafe)
+        /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+        ///
+        /// This function never returns.
+        pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
+            init(self.inner.spawner());
+
+            loop {
+                unsafe {
+                    self.inner.poll();
+                    asm!("wfe");
+                };
+            }
+        }
+    }
 }
 
-impl Executor {
-    /// Create a new Executor.
-    pub fn new() -> Self {
-        Self {
-            inner: raw::Executor::new(|_| unsafe { asm!("sev") }, ptr::null_mut()),
-            not_send: PhantomData,
+#[cfg(feature = "executor-interrupt")]
+pub use interrupt::*;
+#[cfg(feature = "executor-interrupt")]
+mod interrupt {
+    use core::cell::UnsafeCell;
+    use core::mem::MaybeUninit;
+
+    use atomic_polyfill::{AtomicBool, Ordering};
+    use cortex_m::interrupt::InterruptNumber;
+    use cortex_m::peripheral::NVIC;
+
+    use crate::raw::{self, Pender, PenderInner};
+
+    #[derive(Clone, Copy)]
+    pub(crate) struct InterruptPender(u16);
+
+    impl InterruptPender {
+        pub(crate) fn pend(self) {
+            // STIR is faster, but is only available in v7 and higher.
+            #[cfg(not(armv6m))]
+            {
+                let mut nvic: cortex_m::peripheral::NVIC = unsafe { core::mem::transmute(()) };
+                nvic.request(self);
+            }
+
+            #[cfg(armv6m)]
+            cortex_m::peripheral::NVIC::pend(self);
         }
     }
 
-    /// Run the executor.
+    unsafe impl cortex_m::interrupt::InterruptNumber for InterruptPender {
+        fn number(self) -> u16 {
+            self.0
+        }
+    }
+
+    /// Interrupt mode executor.
     ///
-    /// The `init` closure is called with a [`Spawner`] that spawns tasks on
-    /// this executor. Use it to spawn the initial task(s). After `init` returns,
-    /// the executor starts running the tasks.
+    /// This executor runs tasks in interrupt mode. The interrupt handler is set up
+    /// to poll tasks, and when a task is woken the interrupt is pended from software.
     ///
-    /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
-    /// for example by passing it as an argument to the initial tasks.
+    /// This allows running async tasks at a priority higher than thread mode. One
+    /// use case is to leave thread mode free for non-async tasks. Another use case is
+    /// to run multiple executors: one in thread mode for low priority tasks and another in
+    /// interrupt mode for higher priority tasks. Higher priority tasks will preempt lower
+    /// priority ones.
     ///
-    /// This function requires `&'static mut self`. This means you have to store the
-    /// Executor instance in a place where it'll live forever and grants you mutable
-    /// access. There's a few ways to do this:
+    /// It is even possible to run multiple interrupt mode executors at different priorities,
+    /// by assigning different priorities to the interrupts. For an example on how to do this,
+    /// See the 'multiprio' example for 'embassy-nrf'.
     ///
-    /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
-    /// - a `static mut` (unsafe)
-    /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+    /// To use it, you have to pick an interrupt that won't be used by the hardware.
+    /// Some chips reserve some interrupts for this purpose, sometimes named "software interrupts" (SWI).
+    /// If this is not the case, you may use an interrupt from any unused peripheral.
     ///
-    /// This function never returns.
-    pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
-        init(self.inner.spawner());
+    /// It is somewhat more complex to use, it's recommended to use the thread-mode
+    /// [`Executor`] instead, if it works for your use case.
+    pub struct InterruptExecutor {
+        started: AtomicBool,
+        executor: UnsafeCell<MaybeUninit<raw::Executor>>,
+    }
+
+    unsafe impl Send for InterruptExecutor {}
+    unsafe impl Sync for InterruptExecutor {}
+
+    impl InterruptExecutor {
+        /// Create a new, not started `InterruptExecutor`.
+        #[inline]
+        pub const fn new() -> Self {
+            Self {
+                started: AtomicBool::new(false),
+                executor: UnsafeCell::new(MaybeUninit::uninit()),
+            }
+        }
+
+        /// Executor interrupt callback.
+        ///
+        /// # Safety
+        ///
+        /// You MUST call this from the interrupt handler, and from nowhere else.
+        pub unsafe fn on_interrupt(&'static self) {
+            let executor = unsafe { (&*self.executor.get()).assume_init_ref() };
+            executor.poll();
+        }
+
+        /// Start the executor.
+        ///
+        /// This initializes the executor, enables the interrupt, and returns.
+        /// The executor keeps running in the background through the interrupt.
+        ///
+        /// This returns a [`SendSpawner`] you can use to spawn tasks on it. A [`SendSpawner`]
+        /// is returned instead of a [`Spawner`](embassy_executor::Spawner) because the executor effectively runs in a
+        /// different "thread" (the interrupt), so spawning tasks on it is effectively
+        /// sending them.
+        ///
+        /// To obtain a [`Spawner`](embassy_executor::Spawner) for this executor, use [`Spawner::for_current_executor()`](embassy_executor::Spawner::for_current_executor()) from
+        /// a task running in it.
+        ///
+        /// # Interrupt requirements
+        ///
+        /// You must write the interrupt handler yourself, and make it call [`on_interrupt()`](Self::on_interrupt).
+        ///
+        /// This method already enables (unmasks) the interrupt, you must NOT do it yourself.
+        ///
+        /// You must set the interrupt priority before calling this method. You MUST NOT
+        /// do it after.
+        ///
+        pub fn start(&'static self, irq: impl InterruptNumber) -> crate::SendSpawner {
+            if self
+                .started
+                .compare_exchange(false, true, Ordering::Acquire, Ordering::Relaxed)
+                .is_err()
+            {
+                panic!("InterruptExecutor::start() called multiple times on the same executor.");
+            }
 
-        loop {
             unsafe {
-                self.inner.poll();
-                asm!("wfe");
-            };
+                (&mut *self.executor.get())
+                    .as_mut_ptr()
+                    .write(raw::Executor::new(Pender(PenderInner::Interrupt(InterruptPender(
+                        irq.number(),
+                    )))))
+            }
+
+            let executor = unsafe { (&*self.executor.get()).assume_init_ref() };
+
+            unsafe { NVIC::unmask(irq) }
+
+            executor.spawner().make_send()
         }
     }
 }
diff --git a/embassy-executor/src/arch/riscv32.rs b/embassy-executor/src/arch/riscv32.rs
index e97a56cda..f66daeae4 100644
--- a/embassy-executor/src/arch/riscv32.rs
+++ b/embassy-executor/src/arch/riscv32.rs
@@ -1,72 +1,83 @@
-use core::marker::PhantomData;
-use core::ptr;
-use core::sync::atomic::{AtomicBool, Ordering};
+#[cfg(feature = "executor-interrupt")]
+compile_error!("`executor-interrupt` is not supported with `arch-riscv32`.");
 
-use super::{raw, Spawner};
+#[cfg(feature = "executor-thread")]
+pub use thread::*;
+#[cfg(feature = "executor-thread")]
+mod thread {
+    use core::marker::PhantomData;
+    use core::sync::atomic::{AtomicBool, Ordering};
 
-/// global atomic used to keep track of whether there is work to do since sev() is not available on RISCV
-///
-static SIGNAL_WORK_THREAD_MODE: AtomicBool = AtomicBool::new(false);
+    use crate::raw::{Pender, PenderInner};
+    use crate::{raw, Spawner};
 
-/// RISCV32 Executor
-pub struct Executor {
-    inner: raw::Executor,
-    not_send: PhantomData<*mut ()>,
-}
+    #[derive(Copy, Clone)]
+    pub(crate) struct ThreadPender;
 
-impl Executor {
-    /// Create a new Executor.
-    pub fn new() -> Self {
-        Self {
-            // use Signal_Work_Thread_Mode as substitute for local interrupt register
-            inner: raw::Executor::new(
-                |_| {
-                    SIGNAL_WORK_THREAD_MODE.store(true, Ordering::SeqCst);
-                },
-                ptr::null_mut(),
-            ),
-            not_send: PhantomData,
+    impl ThreadPender {
+        #[allow(unused)]
+        pub(crate) fn pend(self) {
+            SIGNAL_WORK_THREAD_MODE.store(true, core::sync::atomic::Ordering::SeqCst);
         }
     }
 
-    /// Run the executor.
-    ///
-    /// The `init` closure is called with a [`Spawner`] that spawns tasks on
-    /// this executor. Use it to spawn the initial task(s). After `init` returns,
-    /// the executor starts running the tasks.
-    ///
-    /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
-    /// for example by passing it as an argument to the initial tasks.
-    ///
-    /// This function requires `&'static mut self`. This means you have to store the
-    /// Executor instance in a place where it'll live forever and grants you mutable
-    /// access. There's a few ways to do this:
-    ///
-    /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
-    /// - a `static mut` (unsafe)
-    /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
-    ///
-    /// This function never returns.
-    pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
-        init(self.inner.spawner());
+    /// global atomic used to keep track of whether there is work to do since sev() is not available on RISCV
+    static SIGNAL_WORK_THREAD_MODE: AtomicBool = AtomicBool::new(false);
+
+    /// RISCV32 Executor
+    pub struct Executor {
+        inner: raw::Executor,
+        not_send: PhantomData<*mut ()>,
+    }
+
+    impl Executor {
+        /// Create a new Executor.
+        pub fn new() -> Self {
+            Self {
+                inner: raw::Executor::new(Pender(PenderInner::Thread(ThreadPender))),
+                not_send: PhantomData,
+            }
+        }
+
+        /// Run the executor.
+        ///
+        /// The `init` closure is called with a [`Spawner`] that spawns tasks on
+        /// this executor. Use it to spawn the initial task(s). After `init` returns,
+        /// the executor starts running the tasks.
+        ///
+        /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
+        /// for example by passing it as an argument to the initial tasks.
+        ///
+        /// This function requires `&'static mut self`. This means you have to store the
+        /// Executor instance in a place where it'll live forever and grants you mutable
+        /// access. There's a few ways to do this:
+        ///
+        /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
+        /// - a `static mut` (unsafe)
+        /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+        ///
+        /// This function never returns.
+        pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
+            init(self.inner.spawner());
 
-        loop {
-            unsafe {
-                self.inner.poll();
-                // we do not care about race conditions between the load and store operations, interrupts
-                //will only set this value to true.
-                critical_section::with(|_| {
-                    // if there is work to do, loop back to polling
-                    // TODO can we relax this?
-                    if SIGNAL_WORK_THREAD_MODE.load(Ordering::SeqCst) {
-                        SIGNAL_WORK_THREAD_MODE.store(false, Ordering::SeqCst);
-                    }
-                    // if not, wait for interrupt
-                    else {
-                        core::arch::asm!("wfi");
-                    }
-                });
-                // if an interrupt occurred while waiting, it will be serviced here
+            loop {
+                unsafe {
+                    self.inner.poll();
+                    // we do not care about race conditions between the load and store operations, interrupts
+                    //will only set this value to true.
+                    critical_section::with(|_| {
+                        // if there is work to do, loop back to polling
+                        // TODO can we relax this?
+                        if SIGNAL_WORK_THREAD_MODE.load(Ordering::SeqCst) {
+                            SIGNAL_WORK_THREAD_MODE.store(false, Ordering::SeqCst);
+                        }
+                        // if not, wait for interrupt
+                        else {
+                            core::arch::asm!("wfi");
+                        }
+                    });
+                    // if an interrupt occurred while waiting, it will be serviced here
+                }
             }
         }
     }
diff --git a/embassy-executor/src/arch/std.rs b/embassy-executor/src/arch/std.rs
index 701f0eb18..4e4a178f0 100644
--- a/embassy-executor/src/arch/std.rs
+++ b/embassy-executor/src/arch/std.rs
@@ -1,84 +1,100 @@
-use std::marker::PhantomData;
-use std::sync::{Condvar, Mutex};
+#[cfg(feature = "executor-interrupt")]
+compile_error!("`executor-interrupt` is not supported with `arch-std`.");
 
-use super::{raw, Spawner};
+#[cfg(feature = "executor-thread")]
+pub use thread::*;
+#[cfg(feature = "executor-thread")]
+mod thread {
+    use std::marker::PhantomData;
+    use std::sync::{Condvar, Mutex};
 
-/// Single-threaded std-based executor.
-pub struct Executor {
-    inner: raw::Executor,
-    not_send: PhantomData<*mut ()>,
-    signaler: &'static Signaler,
-}
+    #[cfg(feature = "nightly")]
+    pub use embassy_macros::main_std as main;
+
+    use crate::raw::{Pender, PenderInner};
+    use crate::{raw, Spawner};
 
-impl Executor {
-    /// Create a new Executor.
-    pub fn new() -> Self {
-        let signaler = &*Box::leak(Box::new(Signaler::new()));
-        Self {
-            inner: raw::Executor::new(
-                |p| unsafe {
-                    let s = &*(p as *const () as *const Signaler);
-                    s.signal()
-                },
-                signaler as *const _ as _,
-            ),
-            not_send: PhantomData,
-            signaler,
+    #[derive(Copy, Clone)]
+    pub(crate) struct ThreadPender(&'static Signaler);
+
+    impl ThreadPender {
+        #[allow(unused)]
+        pub(crate) fn pend(self) {
+            self.0.signal()
         }
     }
 
-    /// Run the executor.
-    ///
-    /// The `init` closure is called with a [`Spawner`] that spawns tasks on
-    /// this executor. Use it to spawn the initial task(s). After `init` returns,
-    /// the executor starts running the tasks.
-    ///
-    /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
-    /// for example by passing it as an argument to the initial tasks.
-    ///
-    /// This function requires `&'static mut self`. This means you have to store the
-    /// Executor instance in a place where it'll live forever and grants you mutable
-    /// access. There's a few ways to do this:
-    ///
-    /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
-    /// - a `static mut` (unsafe)
-    /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
-    ///
-    /// This function never returns.
-    pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
-        init(self.inner.spawner());
+    /// Single-threaded std-based executor.
+    pub struct Executor {
+        inner: raw::Executor,
+        not_send: PhantomData<*mut ()>,
+        signaler: &'static Signaler,
+    }
 
-        loop {
-            unsafe { self.inner.poll() };
-            self.signaler.wait()
+    impl Executor {
+        /// Create a new Executor.
+        pub fn new() -> Self {
+            let signaler = &*Box::leak(Box::new(Signaler::new()));
+            Self {
+                inner: raw::Executor::new(Pender(PenderInner::Thread(ThreadPender(signaler)))),
+                not_send: PhantomData,
+                signaler,
+            }
         }
-    }
-}
 
-struct Signaler {
-    mutex: Mutex<bool>,
-    condvar: Condvar,
-}
+        /// Run the executor.
+        ///
+        /// The `init` closure is called with a [`Spawner`] that spawns tasks on
+        /// this executor. Use it to spawn the initial task(s). After `init` returns,
+        /// the executor starts running the tasks.
+        ///
+        /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
+        /// for example by passing it as an argument to the initial tasks.
+        ///
+        /// This function requires `&'static mut self`. This means you have to store the
+        /// Executor instance in a place where it'll live forever and grants you mutable
+        /// access. There's a few ways to do this:
+        ///
+        /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
+        /// - a `static mut` (unsafe)
+        /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+        ///
+        /// This function never returns.
+        pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
+            init(self.inner.spawner());
 
-impl Signaler {
-    fn new() -> Self {
-        Self {
-            mutex: Mutex::new(false),
-            condvar: Condvar::new(),
+            loop {
+                unsafe { self.inner.poll() };
+                self.signaler.wait()
+            }
         }
     }
 
-    fn wait(&self) {
-        let mut signaled = self.mutex.lock().unwrap();
-        while !*signaled {
-            signaled = self.condvar.wait(signaled).unwrap();
-        }
-        *signaled = false;
+    struct Signaler {
+        mutex: Mutex<bool>,
+        condvar: Condvar,
     }
 
-    fn signal(&self) {
-        let mut signaled = self.mutex.lock().unwrap();
-        *signaled = true;
-        self.condvar.notify_one();
+    impl Signaler {
+        fn new() -> Self {
+            Self {
+                mutex: Mutex::new(false),
+                condvar: Condvar::new(),
+            }
+        }
+
+        fn wait(&self) {
+            let mut signaled = self.mutex.lock().unwrap();
+            while !*signaled {
+                signaled = self.condvar.wait(signaled).unwrap();
+            }
+            *signaled = false;
+        }
+
+        fn signal(&self) {
+            let mut signaled = self.mutex.lock().unwrap();
+            *signaled = true;
+            self.condvar.notify_one();
+        }
     }
 }
diff --git a/embassy-executor/src/arch/wasm.rs b/embassy-executor/src/arch/wasm.rs
index 98091cfbb..08ab16b99 100644
--- a/embassy-executor/src/arch/wasm.rs
+++ b/embassy-executor/src/arch/wasm.rs
@@ -1,74 +1,88 @@
-use core::marker::PhantomData;
+#[cfg(feature = "executor-interrupt")]
+compile_error!("`executor-interrupt` is not supported with `arch-wasm`.");
 
-use js_sys::Promise;
-use wasm_bindgen::prelude::*;
+#[cfg(feature = "executor-thread")]
+pub use thread::*;
+#[cfg(feature = "executor-thread")]
+mod thread {
 
-use super::raw::util::UninitCell;
-use super::raw::{self};
-use super::Spawner;
+    use core::marker::PhantomData;
 
-/// WASM executor, wasm_bindgen to schedule tasks on the JS event loop.
-pub struct Executor {
-    inner: raw::Executor,
-    ctx: &'static WasmContext,
-    not_send: PhantomData<*mut ()>,
-}
+    #[cfg(feature = "nightly")]
+    pub use embassy_macros::main_wasm as main;
+    use js_sys::Promise;
+    use wasm_bindgen::prelude::*;
 
-pub(crate) struct WasmContext {
-    promise: Promise,
-    closure: UninitCell<Closure<dyn FnMut(JsValue)>>,
-}
+    use crate::raw::util::UninitCell;
+    use crate::raw::{Pender, PenderInner};
+    use crate::{raw, Spawner};
+
+    /// WASM executor, wasm_bindgen to schedule tasks on the JS event loop.
+    pub struct Executor {
+        inner: raw::Executor,
+        ctx: &'static WasmContext,
+        not_send: PhantomData<*mut ()>,
+    }
+
+    pub(crate) struct WasmContext {
+        promise: Promise,
+        closure: UninitCell<Closure<dyn FnMut(JsValue)>>,
+    }
+
+    #[derive(Copy, Clone)]
+    pub(crate) struct ThreadPender(&'static WasmContext);
 
-impl WasmContext {
-    pub fn new() -> Self {
-        Self {
-            promise: Promise::resolve(&JsValue::undefined()),
-            closure: UninitCell::uninit(),
+    impl ThreadPender {
+        #[allow(unused)]
+        pub(crate) fn pend(self) {
+            let _ = self.0.promise.then(unsafe { self.0.closure.as_mut() });
         }
     }
-}
 
-impl Executor {
-    /// Create a new Executor.
-    pub fn new() -> Self {
-        let ctx = &*Box::leak(Box::new(WasmContext::new()));
-        let inner = raw::Executor::new(
-            |p| unsafe {
-                let ctx = &*(p as *const () as *const WasmContext);
-                let _ = ctx.promise.then(ctx.closure.as_mut());
-            },
-            ctx as *const _ as _,
-        );
-        Self {
-            inner,
-            not_send: PhantomData,
-            ctx,
+    impl WasmContext {
+        pub fn new() -> Self {
+            Self {
+                promise: Promise::resolve(&JsValue::undefined()),
+                closure: UninitCell::uninit(),
+            }
         }
     }
 
-    /// Run the executor.
-    ///
-    /// The `init` closure is called with a [`Spawner`] that spawns tasks on
-    /// this executor. Use it to spawn the initial task(s). After `init` returns,
-    /// the executor starts running the tasks.
-    ///
-    /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
-    /// for example by passing it as an argument to the initial tasks.
-    ///
-    /// This function requires `&'static mut self`. This means you have to store the
-    /// Executor instance in a place where it'll live forever and grants you mutable
-    /// access. There's a few ways to do this:
-    ///
-    /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
-    /// - a `static mut` (unsafe)
-    /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
-    pub fn start(&'static mut self, init: impl FnOnce(Spawner)) {
-        unsafe {
-            let executor = &self.inner;
-            self.ctx.closure.write(Closure::new(move |_| {
-                executor.poll();
-            }));
-            init(self.inner.spawner());
+    impl Executor {
+        /// Create a new Executor.
+        pub fn new() -> Self {
+            let ctx = &*Box::leak(Box::new(WasmContext::new()));
+            Self {
+                inner: raw::Executor::new(Pender(PenderInner::Thread(ThreadPender(ctx)))),
+                not_send: PhantomData,
+                ctx,
+            }
+        }
+
+        /// Run the executor.
+        ///
+        /// The `init` closure is called with a [`Spawner`] that spawns tasks on
+        /// this executor. Use it to spawn the initial task(s). After `init` returns,
+        /// the executor starts running the tasks.
+        ///
+        /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
+        /// for example by passing it as an argument to the initial tasks.
+        ///
+        /// This function requires `&'static mut self`. This means you have to store the
+        /// Executor instance in a place where it'll live forever and grants you mutable
+        /// access. There's a few ways to do this:
+        ///
+        /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
+        /// - a `static mut` (unsafe)
+        /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+        pub fn start(&'static mut self, init: impl FnOnce(Spawner)) {
+            unsafe {
+                let executor = &self.inner;
+                self.ctx.closure.write(Closure::new(move |_| {
+                    executor.poll();
+                }));
+                init(self.inner.spawner());
+            }
         }
     }
 }
diff --git a/embassy-executor/src/arch/xtensa.rs b/embassy-executor/src/arch/xtensa.rs
index 4ee0d9f78..61ea92c16 100644
--- a/embassy-executor/src/arch/xtensa.rs
+++ b/embassy-executor/src/arch/xtensa.rs
@@ -1,73 +1,84 @@
-use core::marker::PhantomData;
-use core::ptr;
-use core::sync::atomic::{AtomicBool, Ordering};
+#[cfg(feature = "executor-interrupt")]
+compile_error!("`executor-interrupt` is not supported with `arch-xtensa`.");
 
-use super::{raw, Spawner};
+#[cfg(feature = "executor-thread")]
+pub use thread::*;
+#[cfg(feature = "executor-thread")]
+mod thread {
+    use core::marker::PhantomData;
+    use core::sync::atomic::{AtomicBool, Ordering};
 
-/// global atomic used to keep track of whether there is work to do since sev() is not available on Xtensa
-///
-static SIGNAL_WORK_THREAD_MODE: AtomicBool = AtomicBool::new(false);
+    use crate::raw::{Pender, PenderInner};
+    use crate::{raw, Spawner};
 
-/// Xtensa Executor
-pub struct Executor {
-    inner: raw::Executor,
-    not_send: PhantomData<*mut ()>,
-}
+    #[derive(Copy, Clone)]
+    pub(crate) struct ThreadPender;
 
-impl Executor {
-    /// Create a new Executor.
-    pub fn new() -> Self {
-        Self {
-            // use Signal_Work_Thread_Mode as substitute for local interrupt register
-            inner: raw::Executor::new(
-                |_| {
-                    SIGNAL_WORK_THREAD_MODE.store(true, Ordering::SeqCst);
-                },
-                ptr::null_mut(),
-            ),
-            not_send: PhantomData,
+    impl ThreadPender {
+        #[allow(unused)]
+        pub(crate) fn pend(self) {
+            SIGNAL_WORK_THREAD_MODE.store(true, core::sync::atomic::Ordering::SeqCst);
         }
     }
 
-    /// Run the executor.
-    ///
-    /// The `init` closure is called with a [`Spawner`] that spawns tasks on
-    /// this executor. Use it to spawn the initial task(s). After `init` returns,
-    /// the executor starts running the tasks.
-    ///
-    /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
-    /// for example by passing it as an argument to the initial tasks.
-    ///
-    /// This function requires `&'static mut self`. This means you have to store the
-    /// Executor instance in a place where it'll live forever and grants you mutable
-    /// access. There's a few ways to do this:
-    ///
-    /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
-    /// - a `static mut` (unsafe)
-    /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
-    ///
-    /// This function never returns.
-    pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
-        init(self.inner.spawner());
+    /// global atomic used to keep track of whether there is work to do since sev() is not available on Xtensa
+    static SIGNAL_WORK_THREAD_MODE: AtomicBool = AtomicBool::new(false);
+
+    /// Xtensa Executor
+    pub struct Executor {
+        inner: raw::Executor,
+        not_send: PhantomData<*mut ()>,
+    }
+
+    impl Executor {
+        /// Create a new Executor.
+        pub fn new() -> Self {
+            Self {
+                inner: raw::Executor::new(Pender(PenderInner::Thread(ThreadPender))),
+                not_send: PhantomData,
+            }
+        }
+
+        /// Run the executor.
+        ///
+        /// The `init` closure is called with a [`Spawner`] that spawns tasks on
+        /// this executor. Use it to spawn the initial task(s). After `init` returns,
+        /// the executor starts running the tasks.
+        ///
+        /// To spawn more tasks later, you may keep copies of the [`Spawner`] (it is `Copy`),
+        /// for example by passing it as an argument to the initial tasks.
+        ///
+        /// This function requires `&'static mut self`. This means you have to store the
+        /// Executor instance in a place where it'll live forever and grants you mutable
+        /// access. There's a few ways to do this:
+        ///
+        /// - a [StaticCell](https://docs.rs/static_cell/latest/static_cell/) (safe)
+        /// - a `static mut` (unsafe)
+        /// - a local variable in a function you know never returns (like `fn main() -> !`), upgrading its lifetime with `transmute`. (unsafe)
+        ///
+        /// This function never returns.
+        pub fn run(&'static mut self, init: impl FnOnce(Spawner)) -> ! {
+            init(self.inner.spawner());
 
-        loop {
-            unsafe {
-                self.inner.poll();
-                // we do not care about race conditions between the load and store operations, interrupts
-                // will only set this value to true.
-                // if there is work to do, loop back to polling
-                // TODO can we relax this?
-                critical_section::with(|_| {
-                    if SIGNAL_WORK_THREAD_MODE.load(Ordering::SeqCst) {
-                        SIGNAL_WORK_THREAD_MODE.store(false, Ordering::SeqCst);
-                    } else {
-                        // waiti sets the PS.INTLEVEL when slipping into sleep
-                        // because critical sections in Xtensa are implemented via increasing
-                        // PS.INTLEVEL the critical section ends here
-                        // take care not add code after `waiti` if it needs to be inside the CS
-                        core::arch::asm!("waiti 0"); // critical section ends here
-                    }
-                });
+            loop {
+                unsafe {
+                    self.inner.poll();
+                    // we do not care about race conditions between the load and store operations, interrupts
+                    // will only set this value to true.
+                    // if there is work to do, loop back to polling
+                    // TODO can we relax this?
+                    critical_section::with(|_| {
+                        if SIGNAL_WORK_THREAD_MODE.load(Ordering::SeqCst) {
+                            SIGNAL_WORK_THREAD_MODE.store(false, Ordering::SeqCst);
+                        } else {
+                            // waiti sets the PS.INTLEVEL when slipping into sleep
+                            // because critical sections in Xtensa are implemented via increasing
+                            // PS.INTLEVEL the critical section ends here
+                            // take care not add code after `waiti` if it needs to be inside the CS
+                            core::arch::asm!("waiti 0"); // critical section ends here
+                        }
+                    });
+                }
             }
         }
     }
diff --git a/embassy-executor/src/lib.rs b/embassy-executor/src/lib.rs
index 8707995b4..3ce687eb6 100644
--- a/embassy-executor/src/lib.rs
+++ b/embassy-executor/src/lib.rs
@@ -1,5 +1,5 @@
-#![cfg_attr(not(any(feature = "std", feature = "wasm")), no_std)]
-#![cfg_attr(all(feature = "nightly", target_arch = "xtensa"), feature(asm_experimental_arch))]
+#![cfg_attr(not(any(feature = "arch-std", feature = "arch-wasm")), no_std)]
+#![cfg_attr(all(feature = "nightly", feature = "arch-xtensa"), feature(asm_experimental_arch))]
 #![allow(clippy::new_without_default)]
 #![doc = include_str!("../README.md")]
 #![warn(missing_docs)]
@@ -10,41 +10,35 @@ pub(crate) mod fmt;
 #[cfg(feature = "nightly")]
 pub use embassy_macros::task;
 
-cfg_if::cfg_if! {
-    if #[cfg(cortex_m)] {
-        #[path="arch/cortex_m.rs"]
-        mod arch;
-        pub use arch::*;
-        #[cfg(feature = "nightly")]
-        pub use embassy_macros::main_cortex_m as main;
-    }
-    else if #[cfg(target_arch="riscv32")] {
-        #[path="arch/riscv32.rs"]
-        mod arch;
-        pub use arch::*;
-        #[cfg(feature = "nightly")]
-        pub use embassy_macros::main_riscv as main;
-    }
-    else if #[cfg(all(target_arch="xtensa", feature = "nightly"))] {
-        #[path="arch/xtensa.rs"]
-        mod arch;
-        pub use arch::*;
-    }
-    else if #[cfg(feature="wasm")] {
-        #[path="arch/wasm.rs"]
-        mod arch;
-        pub use arch::*;
-        #[cfg(feature = "nightly")]
-        pub use embassy_macros::main_wasm as main;
-    }
-    else if #[cfg(feature="std")] {
-        #[path="arch/std.rs"]
-        mod arch;
-        pub use arch::*;
-        #[cfg(feature = "nightly")]
-        pub use embassy_macros::main_std as main;
-    }
+macro_rules! check_at_most_one {
+    (@amo [$($feats:literal)*] [] [$($res:tt)*]) => {
+        #[cfg(any($($res)*))]
+        compile_error!(concat!("At most one of these features can be enabled at the same time:", $(" `", $feats, "`",)*));
+    };
+    (@amo $feats:tt [$curr:literal $($rest:literal)*] [$($res:tt)*]) => {
+        check_at_most_one!(@amo $feats [$($rest)*] [$($res)* $(all(feature=$curr, feature=$rest),)*]);
+    };
+    ($($f:literal),*$(,)?) => {
+        check_at_most_one!(@amo [$($f)*] [$($f)*] []);
+    };
 }
+check_at_most_one!("arch-cortex-m", "arch-riscv32", "arch-xtensa", "arch-std", "arch-wasm",);
+
+#[cfg(feature = "_arch")]
+#[cfg_attr(feature = "arch-cortex-m", path = "arch/cortex_m.rs")]
+#[cfg_attr(feature = "arch-riscv32", path = "arch/riscv32.rs")]
+#[cfg_attr(feature = "arch-xtensa", path = "arch/xtensa.rs")]
+#[cfg_attr(feature = "arch-std", path = "arch/std.rs")]
+#[cfg_attr(feature = "arch-wasm", path = "arch/wasm.rs")]
+mod arch;
+
+#[cfg(feature = "_arch")]
+pub use arch::*;
+
+pub mod raw;
+
+mod spawner;
+pub use spawner::*;
 
 /// Implementation details for embassy macros.
 /// Do not use. Used for macros and HALs only. Not covered by semver guarantees.
@@ -72,8 +66,3 @@ pub mod _export {
         ($($tt:tt)*) => {};
     }
 }
-
-pub mod raw;
-
-mod spawner;
-pub use spawner::*;
diff --git a/embassy-executor/src/raw/mod.rs b/embassy-executor/src/raw/mod.rs
index 72c367c33..f6c66da5a 100644
--- a/embassy-executor/src/raw/mod.rs
+++ b/embassy-executor/src/raw/mod.rs
@@ -19,7 +19,6 @@ use core::marker::PhantomData;
 use core::mem;
 use core::pin::Pin;
 use core::ptr::NonNull;
-use core::sync::atomic::AtomicPtr;
 use core::task::{Context, Poll};
 
 use atomic_polyfill::{AtomicU32, Ordering};
@@ -290,10 +289,60 @@ impl<F: Future + 'static, const N: usize> TaskPool<F, N> {
     }
 }
 
+#[derive(Clone, Copy)]
+pub(crate) enum PenderInner {
+    #[cfg(feature = "executor-thread")]
+    Thread(crate::arch::ThreadPender),
+    #[cfg(feature = "executor-interrupt")]
+    Interrupt(crate::arch::InterruptPender),
+    #[cfg(feature = "pender-callback")]
+    Callback { func: fn(*mut ()), context: *mut () },
+}
+
+unsafe impl Send for PenderInner {}
+unsafe impl Sync for PenderInner {}
+
+/// Platform/architecture-specific action executed when an executor has pending work.
+///
+/// When a task within an executor is woken, the `Pender` is called. This does a
+/// platform/architecture-specific action to signal there is pending work in the executor.
+/// When this happens, you must arrange for [`Executor::poll`] to be called.
+///
+/// You can think of it as a waker, but for the whole executor.
+pub struct Pender(pub(crate) PenderInner);
+
+impl Pender {
+    /// Create a `Pender` that will call an arbitrary function pointer.
+    ///
+    /// # Arguments
+    ///
+    /// - `func`: The function pointer to call.
+    /// - `context`: Opaque context pointer, that will be passed to the function pointer.
+    #[cfg(feature = "pender-callback")]
+    pub fn new_from_callback(func: fn(*mut ()), context: *mut ()) -> Self {
+        Self(PenderInner::Callback {
+            func,
+            context: context.into(),
+        })
+    }
+}
+
+impl Pender {
+    pub(crate) fn pend(&self) {
+        match self.0 {
+            #[cfg(feature = "executor-thread")]
+            PenderInner::Thread(x) => x.pend(),
+            #[cfg(feature = "executor-interrupt")]
+            PenderInner::Interrupt(x) => x.pend(),
+            #[cfg(feature = "pender-callback")]
+            PenderInner::Callback { func, context } => func(context),
+        }
+    }
+}
+
 pub(crate) struct SyncExecutor {
     run_queue: RunQueue,
-    signal_fn: fn(*mut ()),
-    signal_ctx: AtomicPtr<()>,
+    pender: Pender,
 
     #[cfg(feature = "integrated-timers")]
     pub(crate) timer_queue: timer_queue::TimerQueue,
@@ -302,16 +351,13 @@ pub(crate) struct SyncExecutor {
 }
 
 impl SyncExecutor {
-    pub(crate) fn new(signal_fn: fn(*mut ()), signal_ctx: *mut ()) -> Self {
+    pub(crate) fn new(pender: Pender) -> Self {
         #[cfg(feature = "integrated-timers")]
         let alarm = unsafe { unwrap!(driver::allocate_alarm()) };
-        #[cfg(feature = "integrated-timers")]
-        driver::set_alarm_callback(alarm, signal_fn, signal_ctx);
 
         Self {
             run_queue: RunQueue::new(),
-            signal_fn,
-            signal_ctx: AtomicPtr::new(signal_ctx),
+            pender,
 
             #[cfg(feature = "integrated-timers")]
             timer_queue: timer_queue::TimerQueue::new(),
@@ -332,10 +378,16 @@ impl SyncExecutor {
         trace::task_ready_begin(task.as_ptr() as u32);
 
         if self.run_queue.enqueue(cs, task) {
-            (self.signal_fn)(self.signal_ctx.load(Ordering::Relaxed))
+            self.pender.pend();
         }
     }
 
+    #[cfg(feature = "integrated-timers")]
+    fn alarm_callback(ctx: *mut ()) {
+        let this: &Self = unsafe { &*(ctx as *const Self) };
+        this.pender.pend();
+    }
+
     pub(super) unsafe fn spawn(&'static self, task: TaskRef) {
         task.header().executor.set(Some(self));
 
@@ -351,6 +403,9 @@ impl SyncExecutor {
     ///
     /// Same as [`Executor::poll`], plus you must only call this on the thread this executor was created.
     pub(crate) unsafe fn poll(&'static self) {
+        #[cfg(feature = "integrated-timers")]
+        driver::set_alarm_callback(self.alarm, Self::alarm_callback, self as *const _ as *mut ());
+
         #[allow(clippy::never_loop)]
         loop {
             #[cfg(feature = "integrated-timers")]
@@ -417,14 +472,14 @@ impl SyncExecutor {
 ///
 /// - To get the executor to do work, call `poll()`. This will poll all queued tasks (all tasks
 ///   that "want to run").
-/// - You must supply a `signal_fn`. The executor will call it to notify you it has work
+/// - You must supply a [`Pender`]. The executor will call it to notify you it has work
 ///   to do. You must arrange for `poll()` to be called as soon as possible.
 ///
-/// `signal_fn` can be called from *any* context: any thread, any interrupt priority
+/// The [`Pender`] can be called from *any* context: any thread, any interrupt priority
 /// level, etc. It may be called synchronously from any `Executor` method call as well.
 /// You must deal with this correctly.
 ///
-/// In particular, you must NOT call `poll` directly from `signal_fn`, as this violates
+/// In particular, you must NOT call `poll` directly from the pender callback, as this violates
 /// the requirement for `poll` to not be called reentrantly.
 #[repr(transparent)]
 pub struct Executor {
@@ -437,15 +492,15 @@ impl Executor {
     pub(crate) unsafe fn wrap(inner: &SyncExecutor) -> &Self {
         mem::transmute(inner)
     }
+
     /// Create a new executor.
     ///
-    /// When the executor has work to do, it will call `signal_fn` with
-    /// `signal_ctx` as argument.
+    /// When the executor has work to do, it will call the [`Pender`].
     ///
-    /// See [`Executor`] docs for details on `signal_fn`.
-    pub fn new(signal_fn: fn(*mut ()), signal_ctx: *mut ()) -> Self {
+    /// See [`Executor`] docs for details on `Pender`.
+    pub fn new(pender: Pender) -> Self {
         Self {
-            inner: SyncExecutor::new(signal_fn, signal_ctx),
+            inner: SyncExecutor::new(pender),
             _not_sync: PhantomData,
         }
     }
@@ -468,16 +523,16 @@ impl Executor {
     /// This loops over all tasks that are queued to be polled (i.e. they're
     /// freshly spawned or they've been woken). Other tasks are not polled.
     ///
-    /// You must call `poll` after receiving a call to `signal_fn`. It is OK
-    /// to call `poll` even when not requested by `signal_fn`, but it wastes
+    /// You must call `poll` after receiving a call to the [`Pender`]. It is OK
+    /// to call `poll` even when not requested by the `Pender`, but it wastes
     /// energy.
     ///
     /// # Safety
     ///
     /// You must NOT call `poll` reentrantly on the same executor.
     ///
-    /// In particular, note that `poll` may call `signal_fn` synchronously. Therefore, you
-    /// must NOT directly call `poll()` from your `signal_fn`. Instead, `signal_fn` has to
+    /// In particular, note that `poll` may call the `Pender` synchronously. Therefore, you
+    /// must NOT directly call `poll()` from the `Pender` callback. Instead, the callback has to
     /// somehow schedule for `poll()` to be called later, at a time you know for sure there's
     /// no `poll()` already running.
     pub unsafe fn poll(&'static self) {
diff --git a/examples/boot/application/nrf/Cargo.toml b/examples/boot/application/nrf/Cargo.toml
index 888993255..e75c73cbd 100644
--- a/examples/boot/application/nrf/Cargo.toml
+++ b/examples/boot/application/nrf/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync" }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers", "arch-cortex-m", "executor-thread"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly"] }
 embassy-nrf = { version = "0.1.0", path = "../../../../embassy-nrf", features = ["time-driver-rtc1", "gpiote", "nightly"] }
 embassy-boot = { version = "0.1.0", path = "../../../../embassy-boot/boot" }
diff --git a/examples/boot/application/rp/Cargo.toml b/examples/boot/application/rp/Cargo.toml
index 8d826790b..8de2d2ebd 100644
--- a/examples/boot/application/rp/Cargo.toml
+++ b/examples/boot/application/rp/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync" }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers", "arch-cortex-m", "executor-thread"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly"] }
 embassy-rp = { version = "0.1.0", path = "../../../../embassy-rp", features = ["time-driver", "unstable-traits", "nightly"] }
 embassy-boot-rp = { version = "0.1.0", path = "../../../../embassy-boot/rp" }
diff --git a/examples/boot/application/stm32f3/Cargo.toml b/examples/boot/application/stm32f3/Cargo.toml
index aa279fb76..083607de5 100644
--- a/examples/boot/application/stm32f3/Cargo.toml
+++ b/examples/boot/application/stm32f3/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32f303re", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32f7/Cargo.toml b/examples/boot/application/stm32f7/Cargo.toml
index 1ec0643a6..74f508515 100644
--- a/examples/boot/application/stm32f7/Cargo.toml
+++ b/examples/boot/application/stm32f7/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32f767zi", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32h7/Cargo.toml b/examples/boot/application/stm32h7/Cargo.toml
index a4eefe2a5..898b9a47e 100644
--- a/examples/boot/application/stm32h7/Cargo.toml
+++ b/examples/boot/application/stm32h7/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync" }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32h743zi", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32l0/Cargo.toml b/examples/boot/application/stm32l0/Cargo.toml
index 36eada29b..e142c8481 100644
--- a/examples/boot/application/stm32l0/Cargo.toml
+++ b/examples/boot/application/stm32l0/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32l072cz", "time-driver-any", "exti", "memory-x"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32l1/Cargo.toml b/examples/boot/application/stm32l1/Cargo.toml
index 67efda748..f0e92e1ac 100644
--- a/examples/boot/application/stm32l1/Cargo.toml
+++ b/examples/boot/application/stm32l1/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32l151cb-a", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32l4/Cargo.toml b/examples/boot/application/stm32l4/Cargo.toml
index 4b2e02dd2..87689e9a9 100644
--- a/examples/boot/application/stm32l4/Cargo.toml
+++ b/examples/boot/application/stm32l4/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32l475vg", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/boot/application/stm32wl/Cargo.toml b/examples/boot/application/stm32wl/Cargo.toml
index fecbfc51d..a6708bf51 100644
--- a/examples/boot/application/stm32wl/Cargo.toml
+++ b/examples/boot/application/stm32wl/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../../../embassy-time", features = ["nightly", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../../../embassy-stm32", features = ["unstable-traits", "nightly", "stm32wl55jc-cm4", "time-driver-any", "exti"]  }
 embassy-boot-stm32 = { version = "0.1.0", path = "../../../../embassy-boot/stm32" }
diff --git a/examples/nrf-rtos-trace/Cargo.toml b/examples/nrf-rtos-trace/Cargo.toml
index d8c24dfad..7910b372a 100644
--- a/examples/nrf-rtos-trace/Cargo.toml
+++ b/examples/nrf-rtos-trace/Cargo.toml
@@ -17,7 +17,7 @@ log = [
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync" }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features=["rtos-trace", "rtos-trace-interrupt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "rtos-trace", "rtos-trace-interrupt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time" }
 embassy-nrf = { version = "0.1.0", path = "../../embassy-nrf", features = ["nrf52840", "time-driver-rtc1", "gpiote", "unstable-pac"] }
 
diff --git a/examples/nrf52840/Cargo.toml b/examples/nrf52840/Cargo.toml
index cc88d92c7..3ece24066 100644
--- a/examples/nrf52840/Cargo.toml
+++ b/examples/nrf52840/Cargo.toml
@@ -12,7 +12,7 @@ nightly = ["embassy-executor/nightly", "embassy-nrf/nightly", "embassy-net/night
 [dependencies]
 embassy-futures = { version = "0.1.0", path = "../../embassy-futures" }
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "executor-interrupt", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime"] }
 embassy-nrf = { version = "0.1.0", path = "../../embassy-nrf", features = ["defmt", "nrf52840", "time-driver-rtc1", "gpiote", "unstable-pac", "time"] }
 embassy-net = { version = "0.1.0", path = "../../embassy-net", features = ["defmt", "tcp", "dhcpv4", "medium-ethernet"], optional = true }
diff --git a/examples/nrf5340/Cargo.toml b/examples/nrf5340/Cargo.toml
index e88ddf2f7..4134db46f 100644
--- a/examples/nrf5340/Cargo.toml
+++ b/examples/nrf5340/Cargo.toml
@@ -9,7 +9,7 @@ embassy-futures = { version = "0.1.0", path = "../../embassy-futures" }
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = [
     "defmt",
 ] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = [
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", 
     "nightly",
     "defmt",
     "integrated-timers",
diff --git a/examples/rp/Cargo.toml b/examples/rp/Cargo.toml
index 1e8870ed7..aea61eec5 100644
--- a/examples/rp/Cargo.toml
+++ b/examples/rp/Cargo.toml
@@ -7,7 +7,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime"] }
 embassy-rp = { version = "0.1.0", path = "../../embassy-rp", features = ["defmt", "unstable-traits", "nightly", "unstable-pac", "time-driver", "pio", "critical-section-impl"] }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/std/Cargo.toml b/examples/std/Cargo.toml
index 8087df09a..ff08e378c 100644
--- a/examples/std/Cargo.toml
+++ b/examples/std/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["log"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["log", "std", "nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-std", "executor-thread", "log", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["log", "std", "nightly"] }
 embassy-net = { version = "0.1.0", path = "../../embassy-net", features=[ "std", "nightly", "log", "medium-ethernet", "tcp", "udp", "dns", "dhcpv4", "unstable-traits", "proto-ipv6"] }
 embassy-net-driver = { version = "0.1.0", path = "../../embassy-net-driver" }
diff --git a/examples/stm32c0/Cargo.toml b/examples/stm32c0/Cargo.toml
index 0095a680c..3b1d888f6 100644
--- a/examples/stm32c0/Cargo.toml
+++ b/examples/stm32c0/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "time-driver-any", "stm32c031c6", "memory-x", "unstable-pac", "exti"]  }
 
diff --git a/examples/stm32f0/Cargo.toml b/examples/stm32f0/Cargo.toml
index 89d99b6d3..5c82c5579 100644
--- a/examples/stm32f0/Cargo.toml
+++ b/examples/stm32f0/Cargo.toml
@@ -13,7 +13,7 @@ defmt = "0.3"
 defmt-rtt = "0.4"
 panic-probe = "0.3"
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "executor-interrupt", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "memory-x", "stm32f091rc", "time-driver-any", "exti", "unstable-pac"] }
 static_cell = "1.0"
diff --git a/examples/stm32f0/src/bin/multiprio.rs b/examples/stm32f0/src/bin/multiprio.rs
index e0dc8c989..430a805fc 100644
--- a/examples/stm32f0/src/bin/multiprio.rs
+++ b/examples/stm32f0/src/bin/multiprio.rs
@@ -62,7 +62,7 @@ use core::mem;
 use cortex_m::peripheral::NVIC;
 use cortex_m_rt::entry;
 use defmt::*;
-use embassy_stm32::executor::{Executor, InterruptExecutor};
+use embassy_executor::{Executor, InterruptExecutor};
 use embassy_stm32::interrupt;
 use embassy_stm32::pac::Interrupt;
 use embassy_time::{Duration, Instant, Timer};
diff --git a/examples/stm32f1/Cargo.toml b/examples/stm32f1/Cargo.toml
index 53f369b3a..387af783a 100644
--- a/examples/stm32f1/Cargo.toml
+++ b/examples/stm32f1/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32f103c8", "unstable-pac", "memory-x", "time-driver-any"]  }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/stm32f2/Cargo.toml b/examples/stm32f2/Cargo.toml
index afaf9a0c9..ffb232310 100644
--- a/examples/stm32f2/Cargo.toml
+++ b/examples/stm32f2/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32f207zg", "unstable-pac", "memory-x", "time-driver-any", "exti"]  }
 
diff --git a/examples/stm32f3/Cargo.toml b/examples/stm32f3/Cargo.toml
index 69ebef786..38f11201d 100644
--- a/examples/stm32f3/Cargo.toml
+++ b/examples/stm32f3/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "executor-interrupt", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32f303ze", "unstable-pac", "memory-x", "time-driver-any", "exti"]  }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/stm32f3/src/bin/multiprio.rs b/examples/stm32f3/src/bin/multiprio.rs
index 77df51ac7..5d010f799 100644
--- a/examples/stm32f3/src/bin/multiprio.rs
+++ b/examples/stm32f3/src/bin/multiprio.rs
@@ -62,7 +62,7 @@ use core::mem;
 use cortex_m::peripheral::NVIC;
 use cortex_m_rt::entry;
 use defmt::*;
-use embassy_stm32::executor::{Executor, InterruptExecutor};
+use embassy_executor::{Executor, InterruptExecutor};
 use embassy_stm32::interrupt;
 use embassy_stm32::pac::Interrupt;
 use embassy_time::{Duration, Instant, Timer};
diff --git a/examples/stm32f4/Cargo.toml b/examples/stm32f4/Cargo.toml
index 7a7bab5bb..d967d8501 100644
--- a/examples/stm32f4/Cargo.toml
+++ b/examples/stm32f4/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "executor-interrupt", "defmt", "integrated-timers", "arch-cortex-m", "executor-thread", "executor-interrupt"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "unstable-traits", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "unstable-traits", "defmt", "stm32f429zi", "unstable-pac", "memory-x", "time-driver-any", "exti"]  }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/stm32f4/src/bin/multiprio.rs b/examples/stm32f4/src/bin/multiprio.rs
index 77df51ac7..5d010f799 100644
--- a/examples/stm32f4/src/bin/multiprio.rs
+++ b/examples/stm32f4/src/bin/multiprio.rs
@@ -62,7 +62,7 @@ use core::mem;
 use cortex_m::peripheral::NVIC;
 use cortex_m_rt::entry;
 use defmt::*;
-use embassy_stm32::executor::{Executor, InterruptExecutor};
+use embassy_executor::{Executor, InterruptExecutor};
 use embassy_stm32::interrupt;
 use embassy_stm32::pac::Interrupt;
 use embassy_time::{Duration, Instant, Timer};
diff --git a/examples/stm32f7/Cargo.toml b/examples/stm32f7/Cargo.toml
index ea4cbd808..74e7bf53d 100644
--- a/examples/stm32f7/Cargo.toml
+++ b/examples/stm32f7/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32f767zi", "unstable-pac", "time-driver-any", "exti"]  }
 embassy-net = { path = "../../embassy-net", features = ["defmt", "nightly", "tcp", "dhcpv4", "medium-ethernet"] }
diff --git a/examples/stm32g0/Cargo.toml b/examples/stm32g0/Cargo.toml
index e7273c9fc..03bdbcea3 100644
--- a/examples/stm32g0/Cargo.toml
+++ b/examples/stm32g0/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "time-driver-any", "stm32g071rb", "memory-x", "unstable-pac", "exti"]  }
 
diff --git a/examples/stm32g4/Cargo.toml b/examples/stm32g4/Cargo.toml
index 8a57a8ef0..4e4150350 100644
--- a/examples/stm32g4/Cargo.toml
+++ b/examples/stm32g4/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "time-driver-any", "stm32g491re", "memory-x", "unstable-pac", "exti"]  }
 embassy-hal-common = {version = "0.1.0", path = "../../embassy-hal-common" }
diff --git a/examples/stm32h7/Cargo.toml b/examples/stm32h7/Cargo.toml
index a04134789..d0d6a9497 100644
--- a/examples/stm32h7/Cargo.toml
+++ b/examples/stm32h7/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "unstable-traits", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32h743bi", "time-driver-any", "exti", "unstable-pac", "unstable-traits"] }
 embassy-net = { path = "../../embassy-net", features = ["defmt", "nightly", "tcp", "dhcpv4", "medium-ethernet", "unstable-traits", "proto-ipv6"] }
diff --git a/examples/stm32l0/Cargo.toml b/examples/stm32l0/Cargo.toml
index 86933a629..413d5c18f 100644
--- a/examples/stm32l0/Cargo.toml
+++ b/examples/stm32l0/Cargo.toml
@@ -10,7 +10,7 @@ nightly = ["embassy-stm32/nightly", "embassy-lora", "lorawan-device", "lorawan",
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["defmt", "stm32l072cz", "time-driver-any", "exti", "unstable-traits", "memory-x"]  }
 embassy-lora = { version = "0.1.0", path = "../../embassy-lora", features = ["sx127x", "time", "defmt"], optional = true}
diff --git a/examples/stm32l1/Cargo.toml b/examples/stm32l1/Cargo.toml
index 6e3b2103c..cd9508d57 100644
--- a/examples/stm32l1/Cargo.toml
+++ b/examples/stm32l1/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32l151cb-a", "time-driver-any", "memory-x"]  }
 
diff --git a/examples/stm32l4/Cargo.toml b/examples/stm32l4/Cargo.toml
index 644c90b1a..7c254eba3 100644
--- a/examples/stm32l4/Cargo.toml
+++ b/examples/stm32l4/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-embedded-hal = { version = "0.1.0", path = "../../embassy-embedded-hal" }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "unstable-pac", "stm32l4s5vi", "time-driver-any", "exti", "unstable-traits"]  }
diff --git a/examples/stm32l5/Cargo.toml b/examples/stm32l5/Cargo.toml
index f880328dc..1c662b9da 100644
--- a/examples/stm32l5/Cargo.toml
+++ b/examples/stm32l5/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "unstable-pac", "stm32l552ze", "time-driver-any", "exti", "unstable-traits", "memory-x"]  }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/stm32u5/Cargo.toml b/examples/stm32u5/Cargo.toml
index 2b02eda92..ebef0a4f7 100644
--- a/examples/stm32u5/Cargo.toml
+++ b/examples/stm32u5/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "unstable-pac", "stm32u585ai", "time-driver-any", "memory-x" ]  }
 embassy-usb = { version = "0.1.0", path = "../../embassy-usb", features = ["defmt"] }
diff --git a/examples/stm32wb/Cargo.toml b/examples/stm32wb/Cargo.toml
index e27b4527c..ddf9729e6 100644
--- a/examples/stm32wb/Cargo.toml
+++ b/examples/stm32wb/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32wb55cc", "time-driver-any", "exti"]  }
 
diff --git a/examples/stm32wl/Cargo.toml b/examples/stm32wl/Cargo.toml
index 690481bbf..9fc7e0f4a 100644
--- a/examples/stm32wl/Cargo.toml
+++ b/examples/stm32wl/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "defmt-timestamp-uptime", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "stm32wl55jc-cm4", "time-driver-any", "memory-x", "subghz", "unstable-pac", "exti"]  }
 embassy-lora = { version = "0.1.0", path = "../../embassy-lora", features = ["stm32wl", "time", "defmt"] }
diff --git a/examples/wasm/Cargo.toml b/examples/wasm/Cargo.toml
index e0e799a34..430d0b4c7 100644
--- a/examples/wasm/Cargo.toml
+++ b/examples/wasm/Cargo.toml
@@ -9,7 +9,7 @@ crate-type = ["cdylib"]
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["log"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["log", "wasm", "nightly", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-wasm", "executor-thread", "log", "nightly", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["log", "wasm", "nightly"] }
 
 wasm-logger = "0.2.0"
diff --git a/tests/nrf/Cargo.toml b/tests/nrf/Cargo.toml
index 2a4e8cf41..912749e5d 100644
--- a/tests/nrf/Cargo.toml
+++ b/tests/nrf/Cargo.toml
@@ -7,7 +7,7 @@ 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-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "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"] }
diff --git a/tests/rp/Cargo.toml b/tests/rp/Cargo.toml
index 572a9ce88..eb447be35 100644
--- a/tests/rp/Cargo.toml
+++ b/tests/rp/Cargo.toml
@@ -6,7 +6,7 @@ license = "MIT OR Apache-2.0"
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt"] }
 embassy-rp = { version = "0.1.0", path = "../../embassy-rp", features = ["nightly", "defmt", "unstable-pac", "unstable-traits", "time-driver", "critical-section-impl"]  }
 embassy-futures = { version = "0.1.0", path = "../../embassy-futures" }
diff --git a/tests/stm32/Cargo.toml b/tests/stm32/Cargo.toml
index 08a775eae..17b640797 100644
--- a/tests/stm32/Cargo.toml
+++ b/tests/stm32/Cargo.toml
@@ -15,7 +15,7 @@ stm32u585ai = ["embassy-stm32/stm32u585ai"]     # IoT board
 
 [dependencies]
 embassy-sync = { version = "0.1.0", path = "../../embassy-sync", features = ["defmt"] }
-embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["defmt", "integrated-timers"] }
+embassy-executor = { version = "0.1.0", path = "../../embassy-executor", features = ["arch-cortex-m", "executor-thread", "defmt", "integrated-timers"] }
 embassy-time = { version = "0.1.0", path = "../../embassy-time", features = ["defmt", "tick-hz-32_768"] }
 embassy-stm32 = { version = "0.1.0", path = "../../embassy-stm32", features = ["nightly", "defmt", "unstable-pac", "memory-x", "time-driver-tim2"]  }