mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-04-20 14:59:17 +00:00
Fix universal gemm profiler for pk_i4_t (#1790)
* Fix universal gemm profiler for pk_i4_t * fix
This commit is contained in:
@@ -1,5 +1,5 @@
|
||||
// SPDX-License-Identifier: MIT
|
||||
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
|
||||
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
|
||||
|
||||
#pragma once
|
||||
|
||||
@@ -44,10 +44,19 @@ std::ostream& LogRangeAsType(std::ostream& os, Range&& range, std::string delim)
|
||||
else
|
||||
os << delim;
|
||||
|
||||
if constexpr(std::is_same_v<T, ck::f8_t> || std::is_same_v<T, ck::bf8_t>)
|
||||
using RangeType = ck::remove_cvref_t<decltype(v)>;
|
||||
if constexpr(std::is_same_v<RangeType, ck::f8_t> || std::is_same_v<RangeType, ck::bf8_t> ||
|
||||
std::is_same_v<RangeType, ck::bhalf_t>)
|
||||
{
|
||||
os << ck::type_convert<float>(v);
|
||||
}
|
||||
else if constexpr(std::is_same_v<RangeType, ck::pk_i4_t>)
|
||||
{
|
||||
const auto packed_floats = ck::type_convert<ck::float2_t>(v);
|
||||
const ck::vector_type<float, 2> vector_of_floats{packed_floats};
|
||||
os << vector_of_floats.template AsType<float>()[ck::Number<0>{}] << delim
|
||||
<< vector_of_floats.template AsType<float>()[ck::Number<1>{}];
|
||||
}
|
||||
else
|
||||
{
|
||||
os << static_cast<T>(v);
|
||||
|
||||
Reference in New Issue
Block a user