修复C++11 thread_local对象析构函数与实际内存释放动作顺序相反问题

3 files changed, 278 insertions, 9 deletions

diff --git a/components/libc/posix/pthreads/pthread.c b/components/libc/posix/pthreads/pthread.c
index d035515be4..5cc3bbd7a7 100644
--- a/components/libc/posix/pthreads/pthread.c
+++ b/components/libc/posix/pthreads/pthread.c
@@ -9,6 +9,8 @@
  *                             thread.
  * 2019-02-07     Bernard      Add _pthread_destroy to release pthread resource.
  * 2022-05-10     xiangxistu   Modify the recycle logic about resource of pthread.
+ * 2024-04-15     atwww        Modify the recycle logic of TLS in function _pthread_data_destroy,
+ *                                    make it safe for C++11's thread_local destructors.
  */
 
 #include <rthw.h>
@@ -85,9 +87,27 @@ pthread_t _pthread_data_create(void)
     return index;
 }
 
-void _pthread_data_destroy(_pthread_data_t *ptd)
+static inline void _destroy_item(int index, _pthread_data_t *ptd)
 {
     extern _pthread_key_data_t _thread_keys[PTHREAD_KEY_MAX];
+    void *data;
+
+    if (_thread_keys[index].is_used)
+    {
+        data = ptd->tls[index];
+        if (data && _thread_keys[index].destructor)
+        {
+            _thread_keys[index].destructor(data);
+        }
+    }
+}
+
+#ifdef RT_USING_CPLUSPLUS11
+    #define NOT_USE_CXX_TLS -1
+#endif
+
+void _pthread_data_destroy(_pthread_data_t *ptd)
+{
     pthread_t pth;
 
     if (ptd)
@@ -97,18 +117,37 @@ void _pthread_data_destroy(_pthread_data_t *ptd)
          */
         if (ptd->tls != RT_NULL)
         {
-            void *data;
-            rt_uint32_t index;
-            for (index = 0; index < PTHREAD_KEY_MAX; index ++)
+            int index;
+#ifdef RT_USING_CPLUSPLUS11
+            /* If C++11 is enabled and emutls is used,
+             * destructors of C++ object must be called safely.
+             */
+            extern pthread_key_t emutls_get_pthread_key(void);
+            pthread_key_t emutls_pthread_key = emutls_get_pthread_key();
+
+            if (emutls_pthread_key != NOT_USE_CXX_TLS)
             {
-                if (_thread_keys[index].is_used)
+                /* If execution reaches here, C++ 'thread_local' may be used.
+                 * Destructors of c++ class object must be called before emutls_key_destructor.
+                 */
+                int start = ((emutls_pthread_key - 1 + PTHREAD_KEY_MAX) % PTHREAD_KEY_MAX);
+                int i = 0;
+                for (index = start; i < PTHREAD_KEY_MAX; index = (index - 1 + PTHREAD_KEY_MAX) % PTHREAD_KEY_MAX, i ++)
                 {
-                    data = ptd->tls[index];
-                    if (data && _thread_keys[index].destructor)
-                        _thread_keys[index].destructor(data);
+                    _destroy_item(index, ptd);
+                }
+            }
+            else
+#endif
+            {
+                /* If only C TLS is used, that is, POSIX TLS or __Thread_local,
+                 * just iterate the _thread_keys from index 0.
+                 */
+                for (index = 0; index < PTHREAD_KEY_MAX; index ++)
+                {
+                    _destroy_item(index, ptd);
                 }
             }
-
             /* release tls area */
             rt_free(ptd->tls);
             ptd->tls = RT_NULL;
diff --git a/components/libc/posix/tls/SConscript b/components/libc/posix/tls/SConscript
new file mode 100644
index 0000000000..adcd07ee49
--- /dev/null
+++ b/components/libc/posix/tls/SConscript
@@ -0,0 +1,9 @@
+from building import *
+
+cwd        = GetCurrentDir()
+src        = Glob('*.c')
+CPPPATH    = [cwd]
+
+group = DefineGroup('POSIX', src, depend = ['RT_USING_PTHREADS'], CPPPATH = CPPPATH)
+
+Return('group')
diff --git a/components/libc/posix/tls/emutls.c b/components/libc/posix/tls/emutls.c
new file mode 100644
index 0000000000..0bec347cf9
--- /dev/null
+++ b/components/libc/posix/tls/emutls.c
@@ -0,0 +1,221 @@
+/*
+ * Copyright (c) 2006-2021, RT-Thread Development Team
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * Change Logs:
+ * Date           Author       Notes
+ * 2021-04-27     peterfan     Add copyright header.
+ * 2024-04-15     atwww        Add emutls_get_pthread_key to make c++11's thread_local destructe safely.
+ *                                 Use emutls_pthread_key to determine whether C++11 thread_local is used.
+ *                             Move this file from components\libc\cplusplus\cpp11 to components\libc\posix\tls,
+ *                                 because _Thread_local in C will also use emutls.
+ */
+
+/* ===---------- emutls.c - Implements __emutls_get_address ---------------===
+ *
+ *                     The LLVM Compiler Infrastructure
+ *
+ * This file is dual licensed under the MIT and the University of Illinois Open
+ * Source Licenses. See LICENSE.TXT for details.
+ *
+ * ===----------------------------------------------------------------------===
+ */
+
+#include <pthread.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+
+extern int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
+extern int pthread_key_delete(pthread_key_t key);
+extern void *pthread_getspecific(pthread_key_t key);
+extern int pthread_setspecific(pthread_key_t key, const void *value);
+
+/* Default is not to use posix_memalign, so systems like Android
+ * can use thread local data without heavier POSIX memory allocators.
+ */
+#ifndef EMUTLS_USE_POSIX_MEMALIGN
+#define EMUTLS_USE_POSIX_MEMALIGN 0
+#endif
+
+/* For every TLS variable xyz,
+ * there is one __emutls_control variable named __emutls_v.xyz.
+ * If xyz has non-zero initial value, __emutls_v.xyz's "value"
+ * will point to __emutls_t.xyz, which has the initial value.
+ */
+typedef struct __emutls_control
+{
+    size_t size;  /* size of the object in bytes */
+    size_t align; /* alignment of the object in bytes */
+    union
+    {
+        uintptr_t index; /* data[index-1] is the object address */
+        void *address;   /* object address, when in single thread env */
+    } object;
+    void *value; /* null or non-zero initial value for the object */
+} __emutls_control;
+
+static __inline void *emutls_memalign_alloc(size_t align, size_t size)
+{
+    void *base;
+#if EMUTLS_USE_POSIX_MEMALIGN
+    if (posix_memalign(&base, align, size) != 0)
+        abort();
+#else
+#define EXTRA_ALIGN_PTR_BYTES (align - 1 + sizeof(void *))
+    char *object;
+    if ((object = (char *)malloc(EXTRA_ALIGN_PTR_BYTES + size)) == NULL)
+        abort();
+    base = (void *)(((uintptr_t)(object + EXTRA_ALIGN_PTR_BYTES)) & ~(uintptr_t)(align - 1));
+
+    ((void **)base)[-1] = object;
+#endif
+    return base;
+}
+
+static __inline void emutls_memalign_free(void *base)
+{
+#if EMUTLS_USE_POSIX_MEMALIGN
+    free(base);
+#else
+    /* The mallocated address is in ((void**)base)[-1] */
+    free(((void **)base)[-1]);
+#endif
+}
+
+/* Emulated TLS objects are always allocated at run-time. */
+static __inline void *emutls_allocate_object(__emutls_control *control)
+{
+    size_t size = control->size;
+    size_t align = control->align;
+    if (align < sizeof(void *))
+        align = sizeof(void *);
+    /* Make sure that align is power of 2. */
+    if ((align & (align - 1)) != 0)
+        abort();
+
+    void *base = emutls_memalign_alloc(align, size);
+    if (control->value)
+        memcpy(base, control->value, size);
+    else
+        memset(base, 0, size);
+    return base;
+}
+
+static pthread_mutex_t emutls_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+static size_t emutls_num_object = 0; /* number of allocated TLS objects */
+
+typedef struct emutls_address_array
+{
+    uintptr_t size; /* number of elements in the 'data' array */
+    void *data[];
+} emutls_address_array;
+
+static pthread_key_t emutls_pthread_key = -1; /* -1 means that TLS in C or C++ is not used. */
+
+pthread_key_t emutls_get_pthread_key(void)
+{
+    /* If program uses C or C++ TLS keyword, _Thread_local、__thread or thread_local,
+     * the function emutls_get_index will ensure that emutls_pthread_key is initialized once
+     * when it is first used. Therefore, there is no need to use synchronization measures here.
+     */
+    return emutls_pthread_key;
+}
+
+static void emutls_key_destructor(void *ptr)
+{
+    emutls_address_array *array = (emutls_address_array *)ptr;
+    uintptr_t i;
+    for (i = 0; i < array->size; ++i)
+    {
+        if (array->data[i])
+            emutls_memalign_free(array->data[i]);
+    }
+    free(ptr);
+}
+
+static void emutls_init(void)
+{
+    if (pthread_key_create(&emutls_pthread_key, emutls_key_destructor) != 0)
+        abort();
+}
+
+/* Returns control->object.index; set index if not allocated yet. */
+static __inline uintptr_t emutls_get_index(__emutls_control *control)
+{
+    uintptr_t index = __atomic_load_n(&control->object.index, __ATOMIC_ACQUIRE);
+    if (!index)
+    {
+        static pthread_once_t once = PTHREAD_ONCE_INIT;
+        pthread_once(&once, emutls_init);
+        pthread_mutex_lock(&emutls_mutex);
+        index = control->object.index;
+        if (!index)
+        {
+            index = ++emutls_num_object;
+            __atomic_store_n(&control->object.index, index, __ATOMIC_RELEASE);
+        }
+        pthread_mutex_unlock(&emutls_mutex);
+    }
+    return index;
+}
+
+/* Updates newly allocated thread local emutls_address_array. */
+static __inline void emutls_check_array_set_size(emutls_address_array *array,
+                                                 uintptr_t size)
+{
+    if (array == NULL)
+        abort();
+    array->size = size;
+    pthread_setspecific(emutls_pthread_key, (void *)array);
+}
+
+/* Returns the new 'data' array size, number of elements,
+ * which must be no smaller than the given index.
+ */
+static __inline uintptr_t emutls_new_data_array_size(uintptr_t index)
+{
+    /* Need to allocate emutls_address_array with one extra slot
+     * to store the data array size.
+     * Round up the emutls_address_array size to multiple of 16.
+     */
+    return ((index + 1 + 15) & ~((uintptr_t)15)) - 1;
+}
+
+/* Returns the thread local emutls_address_array.
+ * Extends its size if necessary to hold address at index.
+ */
+static __inline emutls_address_array *
+emutls_get_address_array(uintptr_t index)
+{
+    emutls_address_array *array = pthread_getspecific(emutls_pthread_key);
+    if (array == NULL)
+    {
+        uintptr_t new_size = emutls_new_data_array_size(index);
+        array = (emutls_address_array *)calloc(new_size + 1, sizeof(void *));
+        emutls_check_array_set_size(array, new_size);
+    }
+    else if (index > array->size)
+    {
+        uintptr_t orig_size = array->size;
+        uintptr_t new_size = emutls_new_data_array_size(index);
+        array = (emutls_address_array *)realloc(array, (new_size + 1) * sizeof(void *));
+
+        if (array)
+            memset(array->data + orig_size, 0,
+                   (new_size - orig_size) * sizeof(void *));
+        emutls_check_array_set_size(array, new_size);
+    }
+    return array;
+}
+
+void *__emutls_get_address(void *control)
+{
+    uintptr_t index = emutls_get_index((__emutls_control *)control);
+    emutls_address_array *array = emutls_get_address_array(index);
+    if (array->data[index - 1] == NULL)
+        array->data[index - 1] = emutls_allocate_object((__emutls_control *)control);
+    return array->data[index - 1];
+}