Merge branch 'dev3' into dev3-bin

CMakeLists.txt

@@ -586,10 +586,13 @@ if(MI_BUILD_SHARED)
   install(EXPORT mimalloc DESTINATION ${mi_install_cmakedir})
 
   if(WIN32)
-    # On windows, the import library name for the dll would clash with the static mimalloc.lib
+    # On windows, the import library name for the dll would clash with the static mimalloc.lib library
-    # so we postfix the dll import library with `.dll.lib`.
+    # so we postfix the dll import library with `.dll.lib` (and also the .pdb debug file)
     set_property(TARGET mimalloc PROPERTY ARCHIVE_OUTPUT_NAME "${mi_libname}.dll" )
     install(FILES "$<TARGET_FILE_DIR:mimalloc>/${mi_libname}.dll.lib" DESTINATION ${CMAKE_INSTALL_LIBDIR})
+    set_property(TARGET mimalloc PROPERTY PDB_NAME "${mi_libname}.dll")
+    # don't try to install the pdb since it may not be generated depending on the configuration
+    # install(FILES "$<TARGET_FILE_DIR:mimalloc>/${mi_libname}.dll.pdb" DESTINATION ${CMAKE_INSTALL_LIBDIR})
   endif()
   if(WIN32 AND MI_WIN_REDIRECT)
     # On windows, link and copy the mimalloc redirection dll too.

bin/readme.md

@@ -1,15 +1,14 @@
 # Windows Override
 
-<span id="override_on_windows">Dynamically overriding on mimalloc on Windows</span> 
+<span id="override_on_windows">We use a separate redirection DLL to override mimalloc on Windows</span> 
-is robust and has the particular advantage to be able to redirect all malloc/free calls 
+such that we redirect all malloc/free calls that go through the (dynamic) C runtime allocator, 
-that go through the (dynamic) C runtime allocator, including those from other DLL's or 
+including those from other DLL's or libraries. As it intercepts all allocation calls on a low level, 
-libraries. As it intercepts all allocation calls on a low level, it can be used reliably 
+it can be used reliably on large programs that include other 3rd party components.
-on large programs that include other 3rd party components.
 There are four requirements to make the overriding work well:
 
 1. Use the C-runtime library as a DLL (using the `/MD` or `/MDd` switch).
 
-2. Link your program explicitly with the `mimalloc.lib` export library for
+2. Link your program explicitly with the `mimalloc.dll.lib` export library for
    the `mimalloc.dll` -- which contains all mimalloc functionality.
    To ensure the `mimalloc.dll` is actually loaded at run-time it is easiest 
    to insert some call to the mimalloc API in the `main` function, like `mi_version()`

contrib/vcpkg/readme.md

@@ -9,7 +9,7 @@ to check out a specific commit, version, or branch of mimalloc, or set further o
 You can install such custom port as:
 
 ```sh
-$ vcpkg install mimalloc[override] --recurse --overlay-ports=./contrib/vcpkg
+$ vcpkg install "mimalloc[override]" --recurse --overlay-ports=./contrib/vcpkg
 ```
 
 This will also show the correct sha512 hash if you use a custom version.

contrib/vcpkg/vcpkg-cmake-wrapper.cmake

@@ -17,4 +17,5 @@ endif()
 if(TARGET mimalloc-static AND NOT TARGET mimalloc)
   add_library(mimalloc INTERFACE IMPORTED)
   set_target_properties(mimalloc PROPERTIES INTERFACE_LINK_LIBRARIES mimalloc-static)
+  set_target_properties(mimalloc PROPERTIES INTERFACE_INCLUDE_DIRECTORIES mimalloc-static)
 endif()

readme.md

@@ -165,8 +165,8 @@ mimalloc is used in various large scale low-latency services and programs, for e
 ## Windows
 
 Open `ide/vs2022/mimalloc.sln` in Visual Studio 2022 and build.
-The `mimalloc` project builds a static library (in `out/msvc-x64`), while the
+The `mimalloc-lib` project builds a static library (in `out/msvc-x64`), while the
-`mimalloc-override` project builds a DLL for overriding malloc
+`mimalloc-override-dll` project builds a DLL for overriding malloc
 in the entire program.
 
 ## Linux, macOS, BSD, etc.
@@ -475,16 +475,15 @@ the [shell](https://stackoverflow.com/questions/43941322/dyld-insert-libraries-i
 
 # Windows Override
 
-<span id="override_on_windows">Dynamically overriding on mimalloc on Windows</span> 
+<span id="override_on_windows">We use a separate redirection DLL to override mimalloc on Windows</span> 
-is robust and has the particular advantage to be able to redirect all malloc/free calls 
+such that we redirect all malloc/free calls that go through the (dynamic) C runtime allocator, 
-that go through the (dynamic) C runtime allocator, including those from other DLL's or 
+including those from other DLL's or libraries. As it intercepts all allocation calls on a low level, 
-libraries. As it intercepts all allocation calls on a low level, it can be used reliably 
+it can be used reliably on large programs that include other 3rd party components.
-on large programs that include other 3rd party components.
 There are four requirements to make the overriding work well:
 
 1. Use the C-runtime library as a DLL (using the `/MD` or `/MDd` switch).
 
-2. Link your program explicitly with the `mimalloc.lib` export library for the `mimalloc.dll`.
+2. Link your program explicitly with the `mimalloc.dll.lib` export library for the `mimalloc.dll`.
    (which must be compiled with `-DMI_OVERRIDE=ON`, which is the default though).
    To ensure the `mimalloc.dll` is actually loaded at run-time it is easiest 
    to insert some call to the mimalloc API in the `main` function, like `mi_version()`
@@ -501,9 +500,8 @@ There are four requirements to make the overriding work well:
    list of the final executable (so it can intercept all potential allocations).
    You can use `minject -l <exe>` to check this if needed.
 
-For best performance on Windows with C++, it
+For best performance on Windows with C++, it is also recommended to also override 
-is also recommended to also override the `new`/`delete` operations (by including
+the `new`/`delete` operations (by including [`mimalloc-new-delete.h`](include/mimalloc-new-delete.h)
-[`mimalloc-new-delete.h`](include/mimalloc-new-delete.h)
 a single(!) source file in your project).
 
 The environment variable `MIMALLOC_DISABLE_REDIRECT=1` can be used to disable dynamic

src/prim/unix/prim.c

@@ -201,7 +201,8 @@ static void* unix_mmap_prim(void* addr, size_t size, size_t try_alignment, int p
   void* p = NULL;
   #if defined(MAP_ALIGNED)  // BSD
   if (addr == NULL && try_alignment > 1 && (try_alignment % _mi_os_page_size()) == 0) {
-    size_t n = mi_bsr(try_alignment);
+    size_t idx;
+    size_t n = mi_bsr(try_alignment, &idx);
     if (((size_t)1 << n) == try_alignment && n >= 12 && n <= 30) {  // alignment is a power of 2 and 4096 <= alignment <= 1GiB
       p = mmap(addr, size, protect_flags, flags | MAP_ALIGNED(n), fd, 0);
       if (p==MAP_FAILED || !_mi_is_aligned(p,try_alignment)) {