Disable the SLP vectorizer to prevent unnecessary wait (#1008)

* Disable the SLP vectorizer to prevent unnecessary wait * Add comment to the reason of adding flag * Fix wording

Disable the SLP vectorizer to prevent unnecessary wait (#1008)
* Disable the SLP vectorizer to prevent unnecessary wait * Add comment to the reason of adding flag * Fix wording
ebcfdb3b · Po Yen Chen · Jun Liu · dcb013fc · ebcfdb3b
Commit ebcfdb3b authored Nov 01, 2023 by Po Yen Chen Committed by Jun Liu Nov 07, 2023
Hide whitespace changes
Inline Side-by-side

Showing with 4 additions and 2 deletions

library/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt ...ary/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt +4 -2

No files found.
--- a/library/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt
@@ -108,13 +108,15 @@ if (ENABLE_PIPELINE_V2_OPT)
       CK_EXPERIMENTAL_PIPELINE_V2_IGLP_OPT=1
    )

+    # TODO: The "-vectorize-slp=false" LLVM option is a workaround to prevent inefficient instruction scheduling
+    #       caused by the SLP Vectorizer. Remove this option after fix the SLP Vectorizer issue.
    # layout=NT
    set_source_files_properties(device_gemm_xdl_f16_f16_f16/km_kn_mn_default_pipeline_v2_opt_instance.cpp PROPERTIES
-       COMPILE_OPTIONS ";;"
+       COMPILE_OPTIONS ";-mllvm;-vectorize-slp=false"
       COMPILE_DEFINITIONS "${WAVES_PER_EU_DEFS};${IGLP_OPT_DEFS}")
    # layout=NN
    set_source_files_properties(device_gemm_xdl_f16_f16_f16/km_nk_mn_default_pipeline_v2_opt_instance.cpp PROPERTIES
-       COMPILE_OPTIONS ";;"
+       COMPILE_OPTIONS ";-mllvm;-vectorize-slp=false"
       COMPILE_DEFINITIONS "${WAVES_PER_EU_DEFS};${IGLP_OPT_DEFS}")
    # layout=TT
    set_source_files_properties(device_gemm_xdl_f16_f16_f16/mk_kn_mn_default_pipeline_v2_opt_instance.cpp PROPERTIES