From 146f9d2333bc0aeed1593dc659b17e65b38a33e7 Mon Sep 17 00:00:00 2001
From: Daan <daanl@outlook.com>
Date: Wed, 24 Apr 2024 19:48:04 -0700
Subject: [PATCH] make TLS slot default instead of __builtin_thread_pointer.
 Potentially fixes build errors on various platforms (see issue #883)

---
 include/mimalloc/prim.h | 29 +++++++++++++++--------------
 1 file changed, 15 insertions(+), 14 deletions(-)

diff --git a/include/mimalloc/prim.h b/include/mimalloc/prim.h
index f8a40323..6063d44a 100644
--- a/include/mimalloc/prim.h
+++ b/include/mimalloc/prim.h
@@ -203,11 +203,14 @@ static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexce
 
 #endif
 
-// Do we have __builtin_thread_pointer? (do not make this a compound test as it fails on older gcc's, see issue #851)
-#if defined(__has_builtin)
-#if __has_builtin(__builtin_thread_pointer)
-#define MI_HAS_BUILTIN_THREAD_POINTER  1
-#endif
+// Do we have __builtin_thread_pointer? (do not make this a compound test as that fails on older gcc's, see issue #851)
+#ifdef __has_builtin
+  #if __has_builtin(__builtin_thread_pointer)
+    #if (!defined(__APPLE__)) && /* on apple (M1) the wrong register is read (tpidr_el0 instead of tpidrro_el0) so fall back to TLS slot assembly (<https://github.com/microsoft/mimalloc/issues/343#issuecomment-763272369>)*/ \
+        (!defined(__clang_major__) || __clang_major__ >= 14)  // older clang versions emit bad code; fall back to using the TLS slot (<https://lore.kernel.org/linux-arm-kernel/202110280952.352F66D8@keescook/T/>)
+      #define MI_HAS_BUILTIN_THREAD_POINTER  1
+    #endif
+  #endif
 #elif defined(__GNUC__) && (__GNUC__ >= 7) && defined(__aarch64__)  // special case aarch64 for older gcc versions (issue #851)
 #define MI_HAS_BUILTIN_THREAD_POINTER  1
 #endif
@@ -231,15 +234,6 @@ static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
   return (uintptr_t)NtCurrentTeb();
 }
 
-#elif MI_HAS_BUILTIN_THREAD_POINTER && \
-      (!defined(__APPLE__)) && /* on apple (M1) the wrong register is read (tpidr_el0 instead of tpidrro_el0) so fall back to TLS slot assembly (<https://github.com/microsoft/mimalloc/issues/343#issuecomment-763272369>)*/ \
-      (!defined(__clang_major__) || __clang_major__ >= 14)  // older clang versions emit bad code; fall back to using the TLS slot (<https://lore.kernel.org/linux-arm-kernel/202110280952.352F66D8@keescook/T/>)
-
-static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
-  // Works on most Unix based platforms
-  return (uintptr_t)__builtin_thread_pointer();  
-}
-
 #elif defined(MI_HAS_TLS_SLOT)
 
 static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
@@ -255,6 +249,13 @@ static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
   #endif
 }
 
+#elif MI_HAS_BUILTIN_THREAD_POINTER 
+
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
+  // Works on most Unix based platforms
+  return (uintptr_t)__builtin_thread_pointer();  
+}
+
 #else
 
 // otherwise use portable C, taking the address of a thread local variable (this is still very fast on most platforms).