Disable freelist for macos

It's faster on my mac m1 without the freelist
Fix arm detection on macos
2024-11-19 14:28:15 -08:00 · 2024-11-19 13:24:13 -08:00
2 changed files with 9 additions and 1 deletions
@@ -69,7 +69,8 @@ if(HAS_FULL_RELRO)
 endif()
 cmake_pop_check_state()

-if(CMAKE_SYSTEM_PROCESSOR STREQUAL aarch64)
+if(CMAKE_SYSTEM_PROCESSOR STREQUAL aarch64 OR CMAKE_SYSTEM_PROCESSOR STREQUAL
+                                              arm64)
  add_compile_options(-mbranch-protection=standard)
 else()
  add_compile_options(-fcf-protection)
@@ -694,7 +694,12 @@ constexpr int getMaxCapacity(Node *self) {
                        self->partialKeyLen);
 }

+#ifdef __APPLE__
+// Disabling the free list altogether is faster on my mac m1
+constexpr int64_t kMaxFreeListBytes = 0;
+#else
 constexpr int64_t kMaxFreeListBytes = 1 << 20;
+#endif

 // Maintains a free list up to kMaxFreeListBytes. If the top element of the list
 // doesn't meet the capacity constraints, it's freed and a new node is allocated
@@ -729,6 +734,8 @@ template <class T> struct NodeAllocator {
  }

  void release(T *p) {
+    assume(p->partialKeyCapacity >= 0);
+    assume(freeListSize >= 0);
    if (freeListSize + sizeof(T) + p->partialKeyCapacity > kMaxFreeListBytes) {
      removeNode(p);
      return safe_free(p, sizeof(T) + p->partialKeyCapacity);
Author	SHA1	Message	Date
andrew	ceecc62a63	Disable freelist for macos Tests / 64 bit versions total: 8220, passed: 8220 Details Tests / Debug total: 8218, passed: 8218 Details Tests / SIMD fallback total: 8220, passed: 8220 Details Tests / Release [clang] total: 8220, passed: 8220 Details Clang \|Total\|New\|Outstanding\|Fixed\|Trend \|:-:\|:-:\|:-:\|:-:\|:-: \|0\|0\|0\|0\|:clap: Details Tests / gcc total: 8220, passed: 8220 Details GNU C Compiler (gcc) \|Total\|New\|Outstanding\|Fixed\|Trend \|:-:\|:-:\|:-:\|:-:\|:-: \|0\|0\|0\|0\|:clap: Details Tests / Release [clang,aarch64] total: 5446, passed: 5446 Details Tests / Coverage total: 5497, passed: 5497 Details Code Coverage #### Project Overview No changes detected, that affect the code coverage. * Line Coverage: 97.69% (3168/3243) * Branch Coverage: 42.25% (19291/45654) * Complexity Density: 0.00 * Lines of Code: 3243 #### Quality Gates Summary Output truncated. Details weaselab/conflict-set/pipeline/head This commit looks good Details It's faster on my mac m1 without the freelist	2024-11-19 14:28:15 -08:00
andrew	80f0697e79	Fix arm detection on macos	2024-11-19 13:24:13 -08:00