TFILLPAD

Instruction Diagram

Introduction

Copies a tile and fills the elements outside the valid region with a compile-time fill value.

Copies the source tile into the destination tile and fills the remaining (padding) elements with a compile-time fill value selected by TileDataDst::PadVal (such as PadValue::Min/PadValue::Max).

This operation is commonly used to deterministically materialize a specific value outside the runtime valid region, allowing subsequent operations to compute on the complete static tile shape.

Mathematical Semantics

Assume VR = src.GetValidRow() and VC = src.GetValidCol(). For each destination element (i, j):

pad is determined by TileDataDst::PadVal and the element type (for example, +inf/-inf when supported by the floating-point type, otherwise std::numeric_limits<T>::max()/min()).

Assembly Syntax

Synchronous form (conceptual):

%dst = tfillpad %src : !pto.tile<...> -> !pto.tile<...>

AS Level 1 (SSA)

%dst = pto.tfillpad %src : !pto.tile<...> -> !pto.tile<...>

AS Level 2 (DPS)

pto.tfillpad ins(%src : !pto.tile_buf<...>) outs(%dst : !pto.tile_buf<...>)

C++ Built-in APIs

Implemented in the backend header file introduced by include/pto/common/pto_instr_impl.hpp:

template <typename TileData, PadValue PadVal = PadValue::Zero, typename... WaitEvents>
PTO_INST RecordEvent TFILLPAD(TileData &dst, TileData &src, WaitEvents &... events);

template <typename DstTileData, typename SrcTileData, typename... WaitEvents>
PTO_INST RecordEvent TFILLPAD(DstTileData &dst, SrcTileData &src, WaitEvents &... events);

Constraints

• TileDataDst::PadVal != PadValue::Null (Vec type overload).
• sizeof(TileDataDst::DType) == sizeof(TileDataSrc::DType) and the element size must be 1, 2, or 4 bytes.
• TFILLPAD: TileDataDst::Rows/Cols must match TileDataSrc::Rows/Cols.
• TFILLPAD_EXPAND: TileDataDst::Rows >= TileDataSrc::Rows and TileDataDst::Cols >= TileDataSrc::Cols.
• TFILLPAD(TileData &dst, TileData &src) (Mat type overload): when TileData::TileType is Mat, the layout must satisfy !TileData::isRowMajor && TileData::SLayout::RowMajor, and PadVal must be PadValue::Zero or PadValue::Null. This Mat overload and the first Vec overload (PadVal != PadValue::Null) belong to different SFINAE overloads and do not conflict.

Examples

#include <pto/pto-inst.hpp>

using namespace pto;

void example1() {
  using SrcT = Tile<TileType::Vec, float, 16, 16>;
  using DstT = Tile<TileType::Vec, float, 16, 16, BLayout::RowMajor, 16, 16, SLayout::NoneBox, TileConfig::fractalABSize, PadValue::Min>;

  SrcT src;
  DstT dst;
  TFILLPAD(dst, src);
}

void example2() {
  using TileMatData = Tile<TileType::Mat, float, 16, 256, BLayout::ColMajor, 1, 224, SLayout::RowMajor, 512>;

  TileMatData matTile;
  TFILLPAD(matTile, matTile);
}

ASM Examples

Automatic Mode

# Automatic mode: the compiler/runtime is responsible for resource placement and scheduling.
%dst = pto.tfillpad %src : !pto.tile<...> -> !pto.tile<...>

Manual Mode

# Manual mode: explicitly bind resources first, then issue the instruction.
# Optional (when the instruction contains tile operands):
# pto.tassign %arg0, @tile(0x1000)
# pto.tassign %arg1, @tile(0x2000)
%dst = pto.tfillpad %src : !pto.tile<...> -> !pto.tile<...>

PTO Assembly Form

%dst = pto.tfillpad %src : !pto.tile<...> -> !pto.tile<...>
# AS Level 2 (DPS)
pto.tfillpad ins(%src : !pto.tile_buf<...>) outs(%dst : !pto.tile_buf<...>)