Enum AxisType 

pub enum AxisType {
    Outer,
    Global,
    Warp,
    Local,
    Loop,
    GroupReduce,
    Reduce,
    Upcast,
    Unroll,
    Thread,
    Placeholder,
}

Axis type for loop ranges and reductions.

Variants

Outer

Outer kernel-level scheduling dimension (doesn’t go inside kernels).

Used to mark ranges that exist at the scheduling/orchestration level but don’t become part of kernel execution. These ranges are used during kernel splitting to identify boundaries.

Global

GPU grid dimension.

Warp

Warp/wavefront dimension.

Local

GPU block/workgroup dimension (local memory scope).

Loop

Regular loop.

GroupReduce

Grouped reduction.

Reduce

Reduction axis.

Upcast

Vectorization axis (upcast).

Unroll

Unrolled loop.

Thread

Thread dimension.

Placeholder

Temporary canonicalized range for RESHAPE caching (Tinygrad: AxisType.PLACEHOLDER). Substituted in before _apply_reshape and substituted back after.

Implementations

impl AxisType

pub const fn is_kernel_boundary(&self) -> bool

Returns true if this axis type represents a kernel boundary.

Kernel boundary ranges (Outer) exist at the scheduling level and don’t go inside individual kernels. During kernel splitting, operations with outer ranges are skipped from being packaged into KERNEL ops.

pub const fn priority(self) -> i32

Returns the priority for sorting ranges.

Lower values are outer loops, higher values are inner loops. Matches Tinygrad’s axis_to_pos ordering for kernel optimization.

Priority Order:

• Outer: -2 (kernel-level boundary)
• Loop: -1 (not yet parallelized)
• Global/Thread: 0 (outer parallelism)
• Warp: 1 (sub-group parallelism)
• Local/GroupReduce: 2 (workgroup parallelism + synchronization)
• Upcast: 3 (vectorization)
• Reduce: 4 (reduction loops)
• Unroll: 5 (unrolled loops, innermost)

pub const fn letter(self) -> char

Returns the single-letter code for this axis type.

Used in kernel name generation and debug output.

Letter Codes:

• O: Outer
• L: Loop
• g: Global
• t: Thread
• w: Warp
• l: Local
• G: GroupReduce
• u: Upcast
• R: Reduce
• r: Unroll

pub const fn is_parallel(self) -> bool

Returns true if this is a parallelizable axis type.

Parallel axes represent GPU/thread dispatch dimensions that don’t contribute to accumulator placement in reduce_to_acc.

pub const fn is_reduce(self) -> bool

Returns true if this is a reduction axis type.

Trait Implementations

impl Clone for AxisType

fn clone(&self) -> AxisType

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for AxisType

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for AxisType

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Display for AxisType

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Hash for AxisType

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl Ord for AxisType

fn cmp(&self, other: &Self) -> Ordering

This method returns an Ordering between self and other.

fn max(self, other: Self) -> Self

where Self: Sized,

Compares and returns the maximum of two values.

fn min(self, other: Self) -> Self

where Self: Sized,

Compares and returns the minimum of two values.

fn clamp(self, min: Self, max: Self) -> Self

where Self: Sized,

Restrict a value to a certain interval.

impl PartialEq for AxisType

fn eq(&self, other: &AxisType) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl PartialOrd for AxisType

fn partial_cmp(&self, other: &Self) -> Option<Ordering>

This method returns an ordering between self and other values if one exists.

fn lt(&self, other: &Rhs) -> bool

Tests less than (for self and other) and is used by the < operator.

fn le(&self, other: &Rhs) -> bool

Tests less than or equal to (for self and other) and is used by the <= operator.

fn gt(&self, other: &Rhs) -> bool

Tests greater than (for self and other) and is used by the > operator.

fn ge(&self, other: &Rhs) -> bool

Tests greater than or equal to (for self and other) and is used by the >= operator.

impl Serialize for AxisType

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for AxisType

impl Eq for AxisType

impl StructuralPartialEq for AxisType