Skip to content

Remove fused_rms_norm_ext from WIN32 banned list #76007

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
merged 1 commit into from
Oct 23, 2025
Merged

Remove fused_rms_norm_ext from WIN32 banned list #76007

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
7 changes: 6 additions & 1 deletion paddle/phi/kernels/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,7 +74,6 @@ if(((WITH_GPU) AND (CUDA_VERSION VERSION_LESS 12.0))
"legacy/gpu/moe_gate_dispatch_kernel.cu"
"legacy/gpu/moe_gate_dispatch_grad_kernel.cu"
"legacy/gpu/int_bincount.cu"
"legacy/gpu/layer_norm_cuda_kernel.cu"
"legacy/gpu/fp8_gemm_blockwise_kernel.cu"
"legacy/gpu/fp8_quant_blockwise_kernel.cu"
"fusion/gpu/fused_act_dequant_kernel.cu"
Expand All @@ -85,6 +84,12 @@ if(((WITH_GPU) AND (CUDA_VERSION VERSION_LESS 12.0))
"fusion/gpu/fused_weighted_swiglu_act_quant_kernel.cu")
endif()

if(((WITH_GPU) AND (CUDA_VERSION VERSION_LESS 12.0))
OR APPLE
OR WITH_ROCM)
list(REMOVE_ITEM kernel_cu "legacy/gpu/layer_norm_cuda_kernel.cu")
endif()

# Get flag for CUDA arch >= 80
set(has_arch_ge80 FALSE)
foreach(arch ${NVCC_ARCH_BIN})
Expand Down
10 changes: 9 additions & 1 deletion test/legacy_test/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -551,7 +551,6 @@ if(NOT WITH_GPU
test_incubate_cal_aux_loss
test_incubate_expand_modality_expert_id
test_incubate_fused_loss
test_incubate_fused_rmsnorm_ext
test_incubate_int_bincount
test_incubate_moe_combine
test_incubate_moe_combine_no_weight
Expand All @@ -571,6 +570,15 @@ if(NOT WITH_GPU
test_fused_weighted_swiglu_act_quant_op)
endif()

if(NOT WITH_GPU
OR APPLE
OR WITH_ROCM
OR (${CUDA_ARCH_NAME} STREQUAL "Volta") # Affects the accuracy of op tests
OR ((WITH_GPU) AND (CUDA_VERSION VERSION_LESS 12.0))
)# Restrict the use of older versions of CUB
list(REMOVE_ITEM TEST_OPS test_incubate_fused_rmsnorm_ext)
endif()

set(has_arch_ge80 FALSE)
foreach(arch ${NVCC_ARCH_BIN})
if(${arch} GREATER_EQUAL 80)
Expand Down
Loading