fix performance issue in simde package

CMakeLists.txt

@@ -205,12 +205,12 @@ if(EXISTS  "/proc/cpuinfo")
    message("AVX512 is ${AVX512}")
    message("AVX2 is ${AVX2}")
    if ("${AVX512}" STREQUAL "False")
-      set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -mno-avx512f")
+      set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -mno-avx512f -march=native -DSIMDE_X86_AVX2_NATIVE -DSIMDE_X86_AVX2_NATIVE -DSIMDE_X86_AVX512BW_NATIVE -DSIMDE_X86_AVX512F_NATIVE -DSIMDE_X86_AVX512VL_NATIVE -DSIMDE_X86_AVX_NATIVE -DSIMDE_X86_AVX_NATIVE -DSIMDE_X86_F16C_NATIVE -DSIMDE_X86_FMA_NATIVE -DSIMDE_X86_GFNI_NATIVE -DSIMDE_X86_MMX_NATIVE -DSIMDE_X86_PCLMUL_NATIVE -DSIMDE_X86_SSE2_NATIVE -DSIMDE_X86_SSE3_NATIVE -DSIMDE_X86_SSE4_1_NATIVE -DSIMDE_X86_SSE4_2_NATIVE -DSIMDE_X86_SSE_NATIVE -DSIMDE_X86_SSSE3_NATIVE -DSIMDE_X86_VPCLMULQDQ_NATIVE -DSIMDE_X86_XOP_HAVE_COM_ -DSIMDE_X86_XOP_NATIVE")
    else()
       set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -mavx512bw -march=skylake-avx512 -mtune=skylake-avx512")   
    endif()
    if (CPUINFO MATCHES "avx2" AND "${AVX2}" STREQUAL "True")
-      set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -DSIMDE_X86_AVX2_NATIVE")
+      set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -DSIMDE_X86_AVX2_NATIVE -DSIMDE_X86_VPCLMULQDQ_NATIVE")
    endif()
    if (CPUINFO MATCHES "sse4_1")
       set(C_FLAGS_PROCESSOR "${C_FLAGS_PROCESSOR} -DSIMDE_X86_SSE4_1_NATIVE")

cmake_targets/tools/build_helper

@@ -927,11 +927,33 @@ install_asn1c_from_source(){
 
 install_simde_from_source(){
     echo_info "\nInstalling SIMDE from source without test cases (header files only)"
+    cd /tmp
     $SUDO rm -rf /tmp/simde
     git clone https://github.com/simd-everywhere/simde-no-tests.git /tmp/simde
     cd /tmp/simde
     # brute force copy into /usr/include
     $SUDO \cp -rv ../simde /usr/include
+    $SUDO patch /usr/include/simde/x86/avx.h << FIN
+--- /usr/include/simde/x86/avx.h.old	2022-10-03 19:17:39.828223432 +0200
++++ /usr/include/simde/x86/avx.h	2022-10-05 16:19:55.086019445 +0200
+@@ -3636,12 +3636,16 @@
+ SIMDE_FUNCTION_ATTRIBUTES
+ simde__m256i simde_mm256_insertf128_si256(simde__m256i a, simde__m128i b, int imm8)
+     SIMDE_REQUIRE_CONSTANT_RANGE(imm8, 0, 1) {
++#if defined(SIMDE_X86_AVX_NATIVE) 
++   return _mm256_insertf128_si256(a, b, imm8);
++#else
+   simde__m256i_private a_ = simde__m256i_to_private(a);
+   simde__m128i_private b_ = simde__m128i_to_private(b);
+ 
+   a_.m128i_private[imm8] = b_;
+ 
+   return simde__m256i_from_private(a_);
++#endif
+ }
+ #if defined(SIMDE_X86_AVX_ENABLE_NATIVE_ALIASES)
+   #undef _mm256_insertf128_si256
+FIN
 }
 
 #################################################