Skip to content

Fix fp4 pointer arithmetic in CUDA codegen #1524

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
LeiWang1999 merged 2 commits into from
Dec 24, 2025
Merged

Fix fp4 pointer arithmetic in CUDA codegen #1524

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
65654e5
Fix fp4 pointer arithmetic in CUDA codegen
LJC00118 Dec 24, 2025
7fcbfa8
Fix fp4 pointer arithmetic in CUDA codegen
LJC00118 Dec 24, 2025
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
17 changes: 13 additions & 4 deletions src/target/codegen_cuda.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1319,13 +1319,22 @@ std::string CodeGenTileLangCUDA::GetBufferRef(DataType t,
// int32. Therefore, we need to divide by the ratio of their
// sizes in that case.
int div_factor = (t.lanes() == 1) ? (32 / t.bits()) : t.lanes();
index_str =
PrintExpr(arith::Analyzer().Simplify(truncdiv(index, div_factor)));

os << "*("
<< "(" << ptr_cast(t) << vid << ")"
<< " + " << index_str << " / " << div_factor << ")";
os << "*((" << ptr_cast(t) << vid << ")" << " + " << index_str << ")";
} else if (t == buffer_element_dtype) {
os << buffer_str << "[" << index_str << "]";
} else {
// Fix fp4 pointer arithmetic: fp4 elements are 4-bit packed 2 per byte.
// fp4* + n incorrectly advances n bytes (skipping 2n elements).
int div_factor = 1;
if (buffer_element_dtype.is_float4() && buffer_element_dtype.lanes() == 1) {
div_factor = 2;
}
index_str =
PrintExpr(arith::Analyzer().Simplify(truncdiv(index, div_factor)));

os << "*" << ptr_cast(t) << "(" << buffer_str << " + " << index_str << ")";
}

Expand Down Expand Up @@ -2177,7 +2186,7 @@ void CodeGenTileLangCUDA::VisitExpr_(const CallNode *op, std::ostream &os) {
<< "[(i % 8) / 4 * " + smem_stride +
" * 16 + (threadIdx.x % 4) * 4 * " + smem_stride +
"+ (i % 4) * " + smem_stride +
" + threadIdx.x / 4 + (i / 8) * 8];\n";
" + threadIdx.x / 4 + (i / 8) * 8];\n";
os << "}\n";
} else {
std::string smem_elem_offset = this->PrintExpr(op->args[6]);
Expand Down
Loading