Constants

The following constants are defined for use with the memory interface. For conceptual usage, see Memory Transactions and Lifecycle.

Transaction Types

The following memory transaction types are defined:

static const uint32_t rogue::interfaces::memory::Read = 0x1
[header]

Memory read transaction type.

Exposed to Python as rogue.interfaces.memory.Read.

static const uint32_t rogue::interfaces::memory::Write = 0x2
[header]

Memory write transaction type.

Exposed to Python as rogue.interfaces.memory.Write.

static const uint32_t rogue::interfaces::memory::Post = 0x3
[header]

Memory posted write transaction type.

Exposed to Python as rogue.interfaces.memory.Post.

static const uint32_t rogue::interfaces::memory::Verify = 0x4
[header]

Memory verify readback transaction type.

Exposed to Python as rogue.interfaces.memory.Verify.

Block Process Types

The following block processing types are defined:

static const uint8_t rogue::interfaces::memory::PyFunc = 0x00
[header]

Block access type for Python callback functions.

Exposed to Python as rogue.interfaces.memory.PyFunc.

static const uint8_t rogue::interfaces::memory::Bytes = 0x01
[header]

Block access type for raw byte data.

Exposed to Python as rogue.interfaces.memory.Bytes.

static const uint8_t rogue::interfaces::memory::UInt = 0x02
[header]

Block access type for unsigned integer data.

Exposed to Python as rogue.interfaces.memory.UInt.

static const uint8_t rogue::interfaces::memory::Int = 0x03
[header]

Block access type for signed integer data.

Exposed to Python as rogue.interfaces.memory.Int.

static const uint8_t rogue::interfaces::memory::Bool = 0x04
[header]

Block access type for boolean data.

Exposed to Python as rogue.interfaces.memory.Bool.

static const uint8_t rogue::interfaces::memory::String = 0x05
[header]

Block access type for string data.

Exposed to Python as rogue.interfaces.memory.String.

static const uint8_t rogue::interfaces::memory::Float = 0x06
[header]

Block access type for floating-point (float) data.

Exposed to Python as rogue.interfaces.memory.Float.

static const uint8_t rogue::interfaces::memory::Double = 0x07
[header]

Block access type for double-precision (double) data.

Exposed to Python as rogue.interfaces.memory.Double.

static const uint8_t rogue::interfaces::memory::Float16 = 0x0A
[header]

Block access type for half-precision floating-point data.

Exposed to Python as rogue.interfaces.memory.Float16.

static const uint8_t rogue::interfaces::memory::Float8 = 0x0B
[header]

Block access type for 8-bit E4M3 floating-point data.

Exposed to Python as rogue.interfaces.memory.Float8.

static const uint8_t rogue::interfaces::memory::BFloat16 = 0x0C
[header]

Block access type for BFloat16 (Brain Float 16) data.

Exposed to Python as rogue.interfaces.memory.BFloat16.

static const uint8_t rogue::interfaces::memory::TensorFloat32 = 0x0D
[header]

Block access type for TensorFloat32 (NVIDIA TF32) data.

Exposed to Python as rogue.interfaces.memory.TensorFloat32.

static const uint8_t rogue::interfaces::memory::Float6 = 0x0E
[header]

Block access type for 6-bit E3M2 floating-point data.

Exposed to Python as rogue.interfaces.memory.Float6.

static const uint8_t rogue::interfaces::memory::Float4 = 0x0F
[header]

Block access type for 4-bit E2M1 floating-point data.

Exposed to Python as rogue.interfaces.memory.Float4.

static const uint8_t rogue::interfaces::memory::Fixed = 0x08
[header]

Block access type for fixed-point numeric data.

Exposed to Python as rogue.interfaces.memory.Fixed.

static const uint8_t rogue::interfaces::memory::Custom = 0x80
[header]

Block access type for custom handlers.

Exposed to Python as rogue.interfaces.memory.Custom.