merge with the develop branch

f6ceef78 · ThomasNing · 536c5458 · 25935b57 · f6ceef78 · f6ceef78
Commit f6ceef78 authored Aug 26, 2024 by ThomasNing
20 changed files
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_dl_v1r3.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_dl_v1r3.hpp
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3.hpp
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d.hpp
--- a/include/ck/tensor_operation/operator_transform/transform_conv_fwd_to_gemm.hpp
+++ b/include/ck/tensor_operation/operator_transform/transform_conv_fwd_to_gemm.hpp
--- a/include/ck/utility/amd_buffer_addressing.hpp
+++ b/include/ck/utility/amd_buffer_addressing.hpp
--- a/include/ck/utility/dynamic_buffer.hpp
+++ b/include/ck/utility/dynamic_buffer.hpp
--- a/include/ck/utility/f8_utils.hpp
+++ b/include/ck/utility/f8_utils.hpp
--- a/include/ck/utility/type_convert.hpp
+++ b/include/ck/utility/type_convert.hpp
--- a/include/ck_tile/core/algorithm/coordinate_transform.hpp
+++ b/include/ck_tile/core/algorithm/coordinate_transform.hpp
--- a/include/ck_tile/core/numeric/vector_type.hpp
+++ b/include/ck_tile/core/numeric/vector_type.hpp
--- a/include/ck_tile/core/tensor/tile_distribution.hpp
+++ b/include/ck_tile/core/tensor/tile_distribution.hpp
@@ -746,8 +746,9 @@ CK_TILE_HOST_DEVICE constexpr auto slice_distribution_from_x(
    return make_tuple(
        make_static_tile_distribution(
            tile_distribution_encoding<typename Encoding::RsLengths,
-                                       decltype(sliced_h_lengths), // only need to change the
-                                                                   // h_lengths type
+                                       remove_cvref_t<decltype(sliced_h_lengths)>, // only need to
+                                                                                   // change the
+                                                                                   // h_lengths type
                                       typename Encoding::Ps2RHssMajor,
                                       typename Encoding::Ps2RHssMinor,
                                       typename Encoding::Ys2RHsMajor,

--- a/include/ck_tile/core/tensor/tile_window.hpp
+++ b/include/ck_tile/core/tensor/tile_window.hpp
--- a/include/ck_tile/core/utility/philox_rand.hpp
+++ b/include/ck_tile/core/utility/philox_rand.hpp
--- a/include/ck_tile/ops/fmha.hpp
+++ b/include/ck_tile/ops/fmha.hpp
--- a/include/ck_tile/ops/fmha/block/block_dropout.hpp
+++ b/include/ck_tile/ops/fmha/block/block_dropout.hpp
--- a/include/ck_tile/ops/fmha/kernel/fmha_bwd_kernel.hpp
+++ b/include/ck_tile/ops/fmha/kernel/fmha_bwd_kernel.hpp
--- a/include/ck_tile/ops/fmha/kernel/fmha_bwd_tile_partitioner.hpp
+++ b/include/ck_tile/ops/fmha/kernel/fmha_bwd_tile_partitioner.hpp
--- a/include/ck_tile/ops/fmha/kernel/fmha_fwd_kernel.hpp
+++ b/include/ck_tile/ops/fmha/kernel/fmha_fwd_kernel.hpp
--- a/include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_combine_kernel.hpp
+++ b/include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_combine_kernel.hpp
--- a/include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_kernel.hpp
+++ b/include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_kernel.hpp