PR feedback + make VirtualLock actually work

Author: comex

CMakeLists.txt

@@ -152,6 +152,10 @@ if (LLAMA_ALL_WARNINGS)
 
 endif()
 
+if (MSVC)
+    add_compile_definitions(_CRT_SECURE_NO_WARNINGS)
+endif()
+
 if (LLAMA_LTO)
     include(CheckIPOSupported)
     check_ipo_supported(RESULT result OUTPUT output)

llama.cpp

@@ -1,8 +1,6 @@
 #ifndef LLAMA_UTIL_H
 #define LLAMA_UTIL_H
 
-#define _CRT_SECURE_NO_WARNINGS
-
 #include <cstdio>
 #include <cstdint>
 #include <cerrno>
@@ -26,7 +24,7 @@
 #if defined(_WIN32)
     #define WIN32_LEAN_AND_MEAN
     #define NOMINMAX
-    #include <Windows.h>
+    #include <windows.h>
     #include <io.h>
     #include <stdio.h> // for _fseeki64
 #endif
@@ -238,7 +236,7 @@ struct llama_mmap {
 struct llama_mlock {
     void * addr = NULL;
     size_t size = 0;
-        bool failed_already = false;
+    bool failed_already = false;
 
     llama_mlock() {}
     llama_mlock(const llama_mlock &) = delete;
@@ -318,7 +316,7 @@ struct llama_mlock {
                 return true;
             }
             if (tries == 2) {
-                fprintf(stderr, "warning: failed to VirtualLock %zu-byte buffer (after previously locking %zu bytes): %s\n", 
+                fprintf(stderr, "warning: failed to VirtualLock %zu-byte buffer (after previously locking %zu bytes): %s\n",
                         size, this->size, llama_format_win_err(GetLastError()).c_str());
                 return false;
             }
@@ -331,8 +329,13 @@ struct llama_mlock {
                         llama_format_win_err(GetLastError()).c_str());
                 return false;
             }
-            // No way to make this atomic, so hopefully nobody is doing the
-            // same on other threads.
+            // Per MSDN: "The maximum number of pages that a process can lock
+            // is equal to the number of pages in its minimum working set minus
+            // a small overhead."
+            // Hopefully a megabyte is enough overhead:
+            size_t increment = size + 1048576;
+            // The minimum must be <= the maximum, so we need to increase both:
+            min_ws_size += size;
             max_ws_size += size;
             if (!SetProcessWorkingSetSize(GetCurrentProcess(), min_ws_size, max_ws_size)) {
                 fprintf(stderr, "warning: SetProcessWorkingSetSize failed: %s\n",