Merge branch 'dev' into dev-slice

.gitattributes

@@ -10,4 +10,3 @@
 *.dll binary
 *.lib binary
 *.exe binary
-bin export-ignore

.gitignore

@@ -7,3 +7,5 @@ ide/vs20??/VTune*
 out/
 docs/
 *.zip
+*.tar
+*.gz

docker/alpine-arm32v7/Dockerfile

@@ -0,0 +1,28 @@
+# install from an image
+# download first an appropiate tar.gz image into the current directory 
+# from: <https://github.com/alpinelinux/docker-alpine/tree/edge/armv7>
+FROM scratch
+
+# Substitute the image name that was downloaded
+ADD alpine-minirootfs-20240329-armv7.tar.gz /    
+
+# Install tools
+RUN apk add build-base make cmake
+RUN apk add git
+RUN apk add vim
+
+RUN mkdir -p  /home/dev
+WORKDIR /home/dev
+
+# Get mimalloc
+RUN git clone https://github.com/microsoft/mimalloc -b dev-slice
+RUN mkdir -p mimalloc/out/release
+RUN mkdir -p mimalloc/out/debug
+
+# Build mimalloc debug
+WORKDIR /home/dev/mimalloc/out/debug
+RUN cmake ../.. -DMI_DEBUG_FULL=ON
+RUN make -j
+RUN make test
+
+CMD ["/bin/sh"]

docker/alpine/Dockerfile

@@ -0,0 +1,23 @@
+# alpine image  
+FROM alpine
+
+# Install tools
+RUN apk add build-base make cmake
+RUN apk add git
+RUN apk add vim
+
+RUN mkdir -p  /home/dev
+WORKDIR /home/dev
+
+# Get mimalloc
+RUN git clone https://github.com/microsoft/mimalloc -b dev-slice
+RUN mkdir -p mimalloc/out/release
+RUN mkdir -p mimalloc/out/debug
+
+# Build mimalloc debug
+WORKDIR /home/dev/mimalloc/out/debug
+RUN cmake ../.. -DMI_DEBUG_FULL=ON
+RUN make -j
+RUN make test
+
+CMD ["/bin/sh"]

docker/manylinux-x64/Dockerfile

@@ -0,0 +1,23 @@
+FROM quay.io/pypa/manylinux2014_x86_64
+
+# Install tools
+RUN yum install -y openssl-devel
+RUN yum install -y gcc gcc-c++ kernel-devel make
+RUN yum install -y git cmake
+RUN yum install -y vim
+
+RUN mkdir -p  /home/dev
+WORKDIR /home/dev
+
+# Get mimalloc
+RUN git clone https://github.com/microsoft/mimalloc -b dev-slice
+RUN mkdir -p mimalloc/out/release
+RUN mkdir -p mimalloc/out/debug
+
+# Build mimalloc debug
+WORKDIR /home/dev/mimalloc/out/debug
+RUN cmake ../.. -DMI_DEBUG_FULL=ON
+RUN make -j
+RUN make test
+
+CMD ["/bin/sh"]

docker/readme.md

@@ -0,0 +1,10 @@
+Various example docker files used for testing.
+
+Usage:
+
+```
+> cd <host>
+> docker build -t <host>-mimalloc .
+> docker run -it <host>-mimalloc
+>> make test
+```

include/mimalloc/prim.h

@@ -198,7 +198,7 @@ static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexce
     tcb[slot] = value;
   #elif defined(__APPLE__) && defined(__POWERPC__) // ppc, issue #781
     MI_UNUSED(ofs);
-    pthread_setspecific(slot, value);    
+    pthread_setspecific(slot, value);
   #endif
 }
 
@@ -208,13 +208,18 @@ static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexce
 // but unfortunately, it seems we cannot test for this reliably at this time (see issue #883)
 // Nevertheless, it seems needed on older graviton platforms (see issue #851).
 // For now, we only enable this for specific platforms.
-#if defined(__GNUC__) && (__GNUC__ >= 7) && defined(__aarch64__) /* special case aarch64 for older gcc versions (issue #851) */ \
-    && !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>)*/ \
+#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(MI_LIBC_MUSL) \
     && (!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_USE_BUILTIN_THREAD_POINTER  1
+  #if    (defined(__GNUC__) && (__GNUC__ >= 7)  && defined(__aarch64__)) /* aarch64 for older gcc versions (issue #851) */ \
+      || (defined(__GNUC__) && (__GNUC__ >= 11) && defined(__x86_64__)) \
+      || (defined(__clang_major__) && (__clang_major__ >= 14) && (defined(__aarch64__) || defined(__x86_64__)))
+    #define MI_USE_BUILTIN_THREAD_POINTER  1
+  #endif
 #endif
 
 
+
 // defined in `init.c`; do not use these directly
 extern mi_decl_thread mi_heap_t* _mi_heap_default;  // default heap to allocate from
 extern bool _mi_process_is_initialized;             // has mi_process_init been called?
@@ -239,11 +244,11 @@ static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
   return (uintptr_t)NtCurrentTeb();
 }
 
-#elif MI_USE_BUILTIN_THREAD_POINTER 
+#elif MI_USE_BUILTIN_THREAD_POINTER
 
 static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
   // Works on most Unix based platforms with recent compilers
-  return (uintptr_t)__builtin_thread_pointer();  
+  return (uintptr_t)__builtin_thread_pointer();
 }
 
 #elif defined(MI_HAS_TLS_SLOT)

src/prim/osx/alloc-override-zone.c

@@ -422,6 +422,7 @@ __attribute__((constructor(0)))
 #else
 __attribute__((constructor))      // seems not supported by g++-11 on the M1
 #endif
+__attribute__((used))
 static void _mi_macos_override_malloc(void) {
   malloc_zone_t* purgeable_zone = NULL;