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
158
159
160
161
162

//! GPGPU framework for Rust based on [CUDA Driver API]
//!
//! [CUDA Driver API]: https://docs.nvidia.com/cuda/cuda-driver-api/
//!
//! Basic Examples
//! --------------
//!
//! ### Vector Add
//!
//! ```
//! use accel::*;
//! use accel_derive::kernel;
//!
//! #[kernel]
//! unsafe fn add(a: *const f64, b: *const f64, c: *mut f64, n: usize) {
//!     let i = accel_core::index();
//!     if (i as usize) < n {
//!         *c.offset(i) = *a.offset(i) + *b.offset(i);
//!     }
//! }
//!
//! fn main() -> error::Result<()> {
//!     let device = Device::nth(0)?;
//!     let ctx = device.create_context();
//!
//!     // Allocate memories on GPU
//!     let n = 32;
//!     let mut a = DeviceMemory::<f64>::new(&ctx, n);
//!     let mut b = DeviceMemory::<f64>::new(&ctx, n);
//!     let mut c = DeviceMemory::<f64>::new(&ctx, n);
//!
//!     // Accessible from CPU as usual Rust slice (though this will be slow)
//!     for i in 0..n {
//!         a[i] = i as f64;
//!         b[i] = 2.0 * i as f64;
//!     }
//!     println!("a = {:?}", a.as_slice());
//!     println!("b = {:?}", b.as_slice());
//!
//!     // Launch kernel synchronously
//!     add(&ctx,
//!         1 /* grid */,
//!         n /* block */,
//!         &(&a.as_ptr(), &b.as_ptr(), &c.as_mut_ptr(), &n)
//!     ).expect("Kernel call failed");
//!
//!     println!("c = {:?}", c.as_slice());
//!     Ok(())
//! }
//! ```
//!
//! ### Assertion on GPU
//!
//! ```
//! use accel::*;
//! use accel_derive::kernel;
//!
//! #[kernel]
//! fn assert() {
//!     accel_core::assert_eq!(1 + 2, 4);  // will fail
//! }
//!
//! fn main() -> error::Result<()> {
//!     let device = Device::nth(0)?;
//!     let ctx = device.create_context();
//!     let result = assert(&ctx, 1 /* grid */, 4 /* block */, &());
//!     assert!(result.is_err()); // assertion failed
//!     Ok(())
//! }
//! ```
//!
//! ### Print from GPU
//!
//! ```
//! use accel::*;
//! use accel_derive::kernel;
//!
//! #[kernel]
//! pub fn print() {
//!     let i = accel_core::index();
//!     accel_core::println!("Hello from {}", i);
//! }
//!
//! fn main() -> error::Result<()> {
//!     let device = Device::nth(0)?;
//!     let ctx = device.create_context();
//!     print(&ctx, 1, 4, &())?;
//!     Ok(())
//! }
//! ```
//!
//! Advanced Examples
//! -----------------
//!
//! ### Get compiled PTX as `String`
//!
//! The proc-macro `#[kernel]` creates a submodule `add::` in addition to a function `add`.
//! Kernel Rust code is compiled into PTX string using rustc's `nvptx64-nvidia-cuda` toolchain.
//! Generated PTX string is embedded into proc-macro output as `{kernel_name}::PTX_STR`.
//!
//! ```
//! use accel_derive::kernel;
//!
//! #[kernel]
//! unsafe fn add(a: *const f64, b: *const f64, c: *mut f64, n: usize) {
//!     let i = accel_core::index();
//!     if (i as usize) < n {
//!         *c.offset(i) = *a.offset(i) + *b.offset(i);
//!     }
//! }
//!
//! fn main() {
//!     // PTX assembler code is embedded as `add::PTX_STR`
//!     println!("{}", add::PTX_STR);
//! }
//! ```
//!
//! ### Asynchronous launch
//!
//! `#[kernel]` creates `assert::Module` type definition which implements [Launchable] trait.
//! This struct will read `PTX_STR` using [Module].
//!
//! [Module]:     ./module/struct.Module.html
//! [Launchable]: ./module/trait.Launchable.html
//!
//! ```
//! use accel::*;
//! use accel_derive::kernel;
//!
//! #[kernel]
//! fn assert() {
//!     accel_core::assert_eq!(1 + 2, 4);
//! }
//!
//! fn main() -> error::Result<()> {
//!     let device = Device::nth(0)?;
//!     let ctx = device.create_context();
//!     let stream = Stream::new(&ctx);
//!
//!     let module = assert::Module::new(&ctx)?;
//!     module.stream_launch(&stream, 1, 4, &())?; // lanch will succeed
//!     assert!(stream.sync().is_err()); // assertion failed is detected in next sync
//!     Ok(())
//! }
//! ```

extern crate cuda_driver_sys as cuda;

pub mod array;
pub mod device;
pub mod error;
pub mod linker;
pub mod memory;
pub mod module;
pub mod stream;

pub use array::*;
pub use device::*;
pub use linker::*;
pub use memory::*;
pub use module::*;
pub use stream::*;