mem: 优化 C 内存统计, 缓解多线程竞争导致的 skynet_malloc 性能下降 (#2083)

* mem: 优化 C 内存统计, 缓解多线程竞争导致的 skynet_malloc 性能下降

* rm unused SPMC_STEP_LG macro

---------

Co-authored-by: efve.zff <efve.zff@alibaba-inc.com>

skynet-src/mem_info.h

@@ -0,0 +1,42 @@
+#ifndef _MEM_INFO_H_
+#define _MEM_INFO_H_
+
+#include <stdalign.h>
+#include <stddef.h>
+
+#include "atomic.h"
+
+#define CACHE_LINE_SIZE 64
+
+typedef struct {
+    size_t alloc;
+    size_t alloc_count;
+    size_t free;
+    size_t free_count;
+} MemInfo;
+
+typedef struct {
+    // alloc 与 free 可能不在一个线程
+    // 为了避免竞争，这里也拆成两个 CacheLine 大小
+    alignas(CACHE_LINE_SIZE)
+    ATOM_SIZET alloc;
+    ATOM_SIZET alloc_count;
+
+    alignas(CACHE_LINE_SIZE)
+    ATOM_SIZET free;
+    ATOM_SIZET free_count;
+} AtomicMemInfo;
+
+void meminfo_init(MemInfo *info);
+void atomic_meminfo_init(AtomicMemInfo *info);
+
+void meminfo_alloc(MemInfo *info, size_t size);
+void atomic_meminfo_alloc(AtomicMemInfo *info, size_t size);
+
+void meminfo_free(MemInfo *info, size_t size);
+void atomic_meminfo_free(AtomicMemInfo *info, size_t size);
+
+void meminfo_merge(MemInfo *dest, const MemInfo *src);
+void atomic_meminfo_merge(MemInfo *dest, const AtomicMemInfo *src);
+
+#endif // _MEM_INFO_H_