Struct BatchNorm2dNchwBackward 

pub struct BatchNorm2dNchwBackward<D: Float> {
    pub name: &'static str,
    pub id: String,
    pub block_hw: i32,
    pub kernel_source: String,
    pub entry_point_source: String,
    pub source: String,
    /* private fields */
}

Computes gradients for NCHW BatchNorm2d inference.

Since running_mean and running_var are frozen constants, the backward of y = gamma * (x - mean) * rstd + beta (with respect to x) is simply:

dx[b,c,h,w]   = gamma[c] * rstd[c] * dy[b,c,h,w]
dweight[c]     = Σ_{b,h,w} dy * xhat
dbias[c]       = Σ_{b,h,w} dy

A single loop over (b, hw) computes all three simultaneously.

Grid: [C] — one CTA per channel.

Fields

name: &'static str

The kernel function’s name (e.g. "flash_attention2_forward").

id: String

Unique kernel identifier: fn_name + dtype(s) + const values joined by “__”.

block_hw: i32

Compile-time kernel constant, from the annotated fn’s const generics.

kernel_source: String

The original kernel function source.

entry_point_source: String

The Rust source of the generated C-ABI entry-point wrapper function.

source: String

Combined source (kernel_source + "\n\n" + entry_point_source); used by the Kernel trait.

Implementations

impl<D: Float> BatchNorm2dNchwBackward<D>

pub fn new(block_hw: i32) -> Self

Constructs a new kernel instance for these compile-time parameters.

Trait Implementations

impl<D: Float> Kernel for BatchNorm2dNchwBackward<D>

type Args<'__a> = (*mut D, *mut D, *mut D, *mut D, *mut D, *mut D, *mut D, *mut D, i32, i32, i32, f32)

This kernel’s launch-argument tuple type.

fn id(&self) -> String

A content hash of this kernel’s source, used as a cache key.

fn name(&self) -> &str

This kernel’s name.

fn source(&self) -> &str

This kernel’s full source (DSL + kernel body + entry point wrapper).

fn kernel_source(&self) -> &str

This kernel’s body source, without the entry-point wrapper.

fn entry_point_source(&self) -> &str

Returns the Rust source of the generated C-ABI entry-point wrapper function.

fn entry_point_name(&self) -> String

Returns the PTX symbol name for this kernel: "{name}_entry_point".