From cc25cff8642b1662d7110ffd80eec5051d661c26 Mon Sep 17 00:00:00 2001 From: Shubham Date: Thu, 6 Apr 2023 02:33:08 +0530 Subject: [PATCH] Added AVX512 flag for d24xk pack kernel for windows - on windows 24xk kernel is compiled without avx512 flag which causes out of bounds writes for DTRSM. - to fix this avx512 flag has been added to the CMakeLists.txt file for 24xk kernel. AMD-Internal: [CPUPL-3186] Change-Id: I0314dea88302fc4964a303853a4b9b719ecd8064 --- CMakeLists.txt | 1 + kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c | 1 - 2 files changed, 1 insertion(+), 1 deletion(-) diff --git a/CMakeLists.txt b/CMakeLists.txt index dc7aaaa28..bc9465cb3 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -368,6 +368,7 @@ if(${TARGET_ARCH} STREQUAL zen4 OR set_source_files_properties(${CMAKE_CURRENT_SOURCE_DIR}/kernels/zen4/1m/bli_packm_zen4_asm_z4xk.c PROPERTIES COMPILE_FLAGS /arch:AVX512) set_source_files_properties(${CMAKE_CURRENT_SOURCE_DIR}/kernels/zen4/1m/bli_packm_zen4_asm_z12xk.c PROPERTIES COMPILE_FLAGS /arch:AVX512) set_source_files_properties(${CMAKE_CURRENT_SOURCE_DIR}/kernels/zen4/1m/bli_packm_zen4_asm_d8xk.c PROPERTIES COMPILE_FLAGS /arch:AVX512) + set_source_files_properties(${CMAKE_CURRENT_SOURCE_DIR}/kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c PROPERTIES COMPILE_FLAGS /arch:AVX512) endif() set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} /W0 ") diff --git a/kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c b/kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c index 89755885c..34ccb63d5 100644 --- a/kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c +++ b/kernels/zen4/1m/bli_packm_zen4_asm_d24xk.c @@ -32,7 +32,6 @@ */ -#include #include "blis.h" #define BLIS_ASM_SYNTAX_ATT