atomr-accel-tensorrt 0.10.0

//! # atomr-accel-tensorrt
//!
//! TensorRT engine builder + runtime as supervised atomr actors.
//! Wraps NVIDIA's libnvinfer (and optionally libnvonnxparser) at
//! runtime — the library itself is **not** vendored because it is
//! proprietary; users opt in via the `tensorrt-link` feature and
//! either install TensorRT system-wide or set `LIBNVINFER_PATH`.
//!
//! ## Features
//!
//! - `tensorrt-link` — actually link libnvinfer at build time.
//!   Off-by-default so the crate compiles on hosts without
//!   TensorRT (used by CI + unit tests).
//! - `tensorrt-onnx` — pull in `nvonnxparser` for ONNX import.
//! - `tensorrt-plugin` — `IPluginV3` Rust trampolines.
//! - `tensorrt-int8` — INT8 calibration helpers (entropy / minmax).
//! - `tensorrt-fp8` — FP8 PTQ helpers (Hopper-class GPUs).
//!
//! ## Public surface
//!
//! - [`actor::TrtActor`] / [`actor::TrtMsg`] — sibling actor to
//!   `atomr_accel_cuda::DeviceActor`. Shares `Arc<CudaStream>` with
//!   the device actor so inference rides the same execution
//!   timeline.
//! - [`builder::IBuilderConfig`] — pure-Rust mirror of the TensorRT
//!   builder config, with knobs for precision, DLA, structured
//!   sparsity, tactic sources, timing cache, and engine refit.
//! - [`engine::TrtEngine`] — owned, immutable engine handle that's
//!   `Send + Sync` via newtype.
//! - [`runtime::TrtRuntime`] / [`runtime::ExecutionContext`] — load
//!   serialised plans + drive `enqueueV3` on a shared CUDA stream.
//! - [`onnx::OnnxParser`] — gated on `tensorrt-onnx`.
//! - [`calibration`] — gated on `tensorrt-int8` / `tensorrt-fp8`.
//! - [`plugin`] — gated on `tensorrt-plugin`.

#![allow(
    clippy::type_complexity,
    clippy::too_many_arguments,
    clippy::arc_with_non_send_sync
)]

// The `tensorrt-link` feature compiles `csrc/nvinfer_shim.cpp` and
// links the resulting static lib against system libnvinfer (and
// libnvonnxparser / libnvinfer_plugin when their sub-features are
// also on). See `build.rs` for the probe order and the env-var
// contract (`LIBNVINFER_PATH`, `TENSORRT_INCLUDE_PATH`, `CUDA_PATH`).

pub mod actor;
pub mod builder;
pub mod engine;
pub mod error;
pub mod runtime;
pub mod sys;

#[cfg(feature = "tensorrt-onnx")]
pub mod onnx;

#[cfg(feature = "tensorrt-int8")]
pub mod calibration;

#[cfg(feature = "tensorrt-plugin")]
pub mod plugin;

pub use actor::{
    BuildFromOnnxReply, BuildReply, CreateContextReply, DeserializeReply, EnqueueReply,
    ExecuteReply, NetworkSource, RefitReply, RefitWeights, TrtActor, TrtMsg,
};
pub use builder::{
    BuilderFlags, DeviceType, IBuilderConfig, Precision, RefitPolicy, TacticSources,
};
pub use engine::{EnginePlan, TrtEngine, TrtRefitter};
pub use error::TrtError;
pub use runtime::{EnqueueRequest, ExecutionBindings, ExecutionContext, TensorShape, TrtRuntime};

/// Install the Rust→`tracing` logger bridge into the C++ shim's
/// `RustBridgeLogger`. Called from `TrtActor::new`, `TrtRuntime::new`,
/// and `IBuilderConfig::default` so any entry point sets it up before
/// the first TRT call. `Once` makes it idempotent across the process
/// lifetime.
#[cfg(feature = "tensorrt-link")]
pub fn init_logger() {
    use std::sync::Once;
    static INIT: Once = Once::new();
    INIT.call_once(|| unsafe {
        sys::atomr_trt_install_logger(rust_log_trampoline, std::ptr::null_mut());
    });
}

/// No-op when the `tensorrt-link` feature is off so callers don't
/// need to gate their `init_logger()` calls.
#[cfg(not(feature = "tensorrt-link"))]
pub fn init_logger() {}

#[cfg(feature = "tensorrt-link")]
unsafe extern "C" fn rust_log_trampoline(
    sev: std::os::raw::c_int,
    msg: *const std::os::raw::c_char,
    len: usize,
    _user: *mut std::os::raw::c_void,
) {
    if msg.is_null() || len == 0 {
        return;
    }
    let bytes = std::slice::from_raw_parts(msg as *const u8, len);
    let text = String::from_utf8_lossy(bytes);
    match sev {
        0 | 1 => tracing::error!(target: "tensorrt", "{text}"),
        2 => tracing::warn!(target: "tensorrt", "{text}"),
        3 => tracing::info!(target: "tensorrt", "{text}"),
        _ => tracing::debug!(target: "tensorrt", "{text}"),
    }
}