Skip to content

[NFI] Move vBlocks capping into getBlockIOTileSize<true> #6805

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
anmyachev merged 1 commit into from
May 1, 2026
+23 −23
Merged

[NFI] Move vBlocks capping into getBlockIOTileSize<true> #6805

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 0 additions & 18 deletions third_party/intel/lib/TritonIntelGPUToLLVM/LoadStoreOpToLLVM.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1980,16 +1980,6 @@ struct LoadOpToBlockIOConversion
if (totalBytesPerRowPerMatrix > MAX_WIDTH)
return failure();

// Load multiple dot operands by enlarging the vBlocks.
vBlocks = std::min(vBlocks,
static_cast<int>(MAX_WIDTH / totalBytesPerRowPerMatrix));
// vBlocks has HW limitation of 4.
vBlocks = std::min(vBlocks, 4);
// Limit vBlocks to 1 if block size is smaller than GRF size.
const unsigned GRF_SIZE = 64;
if (tileHeight * tileWidth * packedElemSizeInBits / 8 < GRF_SIZE)
vBlocks = 1;

Location loc = op.getLoc();
auto b = TritonLLVMOpBuilder(loc, rewriter);
MLIRContext *ctx = rewriter.getContext();
Expand Down Expand Up @@ -2361,14 +2351,6 @@ struct DescriptorLoadOpToBlockIOConversion
if (totalBytesPerRowPerMatrix > MAX_WIDTH)
return failure();

// Load multiple dot operands by enlarging the vBlocks.
vBlocks = std::min(vBlocks,
static_cast<int>(MAX_WIDTH / totalBytesPerRowPerMatrix));
vBlocks = std::min(vBlocks, 4);
const unsigned GRF_SIZE = 64;
if (tileHeight * tileWidth * packedElemSizeInBits / 8 < GRF_SIZE)
vBlocks = 1;

Location loc = op.getLoc();
auto b = TritonLLVMOpBuilder(loc, rewriter);
MLIRContext *ctx = rewriter.getContext();
Expand Down
28 changes: 23 additions & 5 deletions third_party/intel/lib/TritonIntelGPUTransforms/BlockIOUtils.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -330,11 +330,29 @@ getBlockIOTileSize(const LinearLayout &ll, unsigned memContiguousDim,
// insert the remaining register base.
regPackBases.insert(1 << regBaseIter);
}
return BlockIOTileSizeInfo(tileShape[transpose ? fastChangeDim : rowDim],
tileShape[transpose ? rowDim : fastChangeDim] /
numElemPerPackedVal,
numElemPerPackedVal, vBlocks, rowDim,
fastChangeDim, transpose, std::move(regPackBases));
int tileHeight = tileShape[transpose ? fastChangeDim : rowDim];
int tileWidth =
tileShape[transpose ? rowDim : fastChangeDim] / numElemPerPackedVal;

// Cap vBlocks for loads based on HW constraints.
if constexpr (isLoad) {
constexpr int MAX_WIDTH_BYTES = 64;
unsigned packedElemSizeInBits = elemSizeInBits * numElemPerPackedVal;
unsigned totalBytesPerRowPerMatrix = tileWidth * packedElemSizeInBits / 8;
if (totalBytesPerRowPerMatrix > 0) {
vBlocks =
std::min(vBlocks, static_cast<unsigned>(MAX_WIDTH_BYTES /
totalBytesPerRowPerMatrix));
}
vBlocks = std::min(vBlocks, 4u);
constexpr unsigned GRF_SIZE = 64;
if (tileHeight * tileWidth * packedElemSizeInBits / 8 < GRF_SIZE)
vBlocks = 1;
}

return BlockIOTileSizeInfo(tileHeight, tileWidth, numElemPerPackedVal,
vBlocks, rowDim, fastChangeDim, transpose,
std::move(regPackBases));
}

// Explicit instantiations.
Expand Down