1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
/***********************************************************************************************************************
 * Copyright (c) 2019 by the authors
 *
 * Author: André Borrmann
 * License: MIT / Apache License 2.0
 **********************************************************************************************************************/
#![doc(html_root_url = "https://docs.rs/ruspiro-singleton/0.4.3")]
#![no_std]
#![feature(const_fn_trait_bound)]

//! # Singleton pattern implementation
//!
//! Provide a cross core synchronisation safe singleton implementation pattern. The `Singleton` is intended to be used
//! to declare crate level static variables that require safe access accross cores. This is helpful where the data
//! structure used within the `Singleton` represents a peripheral where the crate shall only hand out a single instance
//! to safely represent to unique existance of the peripheral.
//!
//! # HINT
//! Safe lazy initialization is ensured using atomics. On the Raspberry Pi atmomic operations require the *MMU* to be
//! configured and active. Otherwise the executing CPU core will hang when trying to execute the atomic operation.
//!
//! # Example
//! ```no_run
//! # use ruspiro_singleton::*;
//! // define the static variable with an inizialization closure
//! static FOO:Singleton<u32> = Singleton::new(20);
//!
//! // define the static variable with an inizialization closure
//! static DEMO:Singleton<Box<Demo>> = Singleton::lazy(&|| {
//!     Box::new(
//!         Demo::new()
//!     )
//! });
//!
//! // define the type to be accessible as singleton
//! struct Demo {
//!     pub count: u32,
//! }
//!
//! // implement the type that should provided as singlton
//! impl Demo {
//!     pub const fn new() -> Self {
//!         Demo {
//!             count: 0,
//!         }
//!     }
//! }
//!
//! fn main() {
//!     // safely use the singleton inside the closure passed to [with_mut] to update it's contents
//!     DEMO.with_mut(|d| {
//!         d.count += 10;
//!     });
//!
//!     // safely use the singleton inside the closure passed to [with_ref] if read-only access is required
//!     DEMO.with_mut(|d| {
//!         println!("Value: {}", d.count);
//!     });
//!
//!     // you may also return a value from the singleton to work with it after the safe singleton access
//!     let val = DEMO.with_ref(|d| {
//!         if d.count != 0 {
//!             true
//!         } else {
//!             false
//!         }
//!     });
//! }
//! ```

mod lazy;

use lazy::LazyValue;
use ruspiro_lock::RWLock;

/// The Singleton wrapper stores any type
pub struct Singleton<T: 'static> {
  /// the inner value wrapping the contained data for safe read/write access
  inner: RWLock<LazyValue<T>>,
}

// The Singleton need to implement Send & Sync to ensure cross core compile check mechanics
// this is safe as the inner RWLock ensures cross core safety
// but we need to be conditional on the inner type to prevent interior mutable types beeing used
// inside a singleton
unsafe impl<T> Sync for Singleton<T> where T: Sync {}
unsafe impl<T> Send for Singleton<T> where T: Send {}

impl<T: 'static> Singleton<T> {
  /// Create a new [Singleton] instance to be used in a static variable. Only ``const fn`` constructors are allowed
  /// here.
  /// # Example
  /// ```no_run
  /// # use ruspiro_singleton::*;
  /// static FOO: Singleton<u32> = Singleton::new(20);
  /// # fn main() {}
  /// ```
  pub const fn new(value: T) -> Self {
    Singleton {
      inner: RWLock::new(LazyValue::with_value(value)),
    }
  }

  /// Create a new [Singleton] instance passing a closure that will be evaluated at first access to the contents of
  /// the singleton that will provide its value
  /// # Example
  /// ```no_run
  /// # use ruspiro_singleton::*;
  /// static FOO: Singleton<String> = Singleton::lazy(&|| String::from("foo"));
  /// # fn main() {}
  /// ```
  pub const fn lazy<F>(init: &'static F) -> Self
  where
    F: Fn() -> T,
  {
    Self {
      inner: RWLock::new(LazyValue::with_init(init)),
    }
  }

  /// Take the stored singleton for whatever operation and prevent usage by other cores
  /// Safe access to the singleton mutable instance is guarantied inside the given closure.
  ///
  pub fn with_mut<F, R>(&self, f: F) -> R
  where
    F: FnOnce(&mut T) -> R,
  {
    let inner = self.inner.lock();
    // use write lock to mutably access the inner value of the singleton. As long
    // as the write lock exists no other write or read lock is possible
    let r = f(inner.get_mut());

    // explicitly release the lock befor providing the result of the closure to the caller
    drop(inner);

    r
  }

  /// Immutable access to a singleton for a specific operation.
  /// This access does not enforce any lock nor guarantees safe atomic access to the instance. However, it is usefull
  /// in read-only access scenarios like inside interrupt handlers.
  ///
  pub fn with_ref<F, R>(&self, f: F) -> R
  where
    F: FnOnce(&T) -> R,
  {
    let inner = self.inner.read();
    // multiple read locks are possible when accessing the inner data of the singleton
    // all read locks are required to be released before the next write lock could happen
    let r = f(inner.get());

    // explicitly release the lock befor providing the result of the closure to the caller
    drop(inner);

    r
  }
}