[PATCH 0/1] MR1456: imm32: Prevent deadlock on unload

[Fabian Maurer (＠DarkShadow44)]

We need to make sure all (important) cleanup is finished when we exit DllMain,
otherwise we might already unload krnl386 and deadlock
Since we can't have a synchronous DestroyWindow, use an extra message

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52511
Signed-off-by: Fabian Maurer

Code flow is like follows:

- DOSVM_Exit -> RtlExitUserThread -> LdrShutdownThread
- acquire loader lock
- send DLL_THREAD_DETACH to imm32
- - calls DestroyWindow on its window
- - WM_DESTROY is handled asynchonously!
- send DLL_THREAD_DETACH to krnl386.exe 
- - TASK_ExitTask
- -  try get win16 lock, is blocked by ime32

meanwhile, at ime32

- handles WM_DESTROY
- already have win16 lock from user32
- __wine_ime_wnd_proc (WM_DESTROY) -> imm_couninit_thread -> CoRevokeInitializeSpy
- LdrGetProcedureAddress (want CoRevokeInitializeSpy)
- try get loader_lock, is blocked

Solution is to wait in ime32 DllMain until the cleanup as happened, aka CoRevokeInitializeSpy has finished.

[PATCH 1/1] imm32: Prevent deadlock on unload

[Fabian Maurer]

From: Fabian Maurer

We need to make sure all (important) cleanup is finished when we exit DllMain,
otherwise we might already unload krnl386 and deadlock
Since we can't have a synchronous DestroyWindow, use an extra message

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52511
Signed-off-by: Fabian Maurer 
---
 dlls/imm32/imm.c | 21 ++++++++++++++-------
 1 file changed, 14 insertions(+), 7 deletions(-)

diff --git a/dlls/imm32/imm.c b/dlls/imm32/imm.c
index 780d544c0e9..3a334bc0de5 100644
--- a/dlls/imm32/imm.c
+++ b/dlls/imm32/imm.c
@@ -51,6 +51,7 @@ static UINT WM_MSIME_RECONVERTREQUEST;
 static UINT WM_MSIME_RECONVERT;
 static UINT WM_MSIME_QUERYPOSITION;
 static UINT WM_MSIME_DOCUMENTFEED;
+static UINT WM_WINE_IME_DESTROY;

typedef struct _tagImmHkl{
     struct list entry;
@@ -554,7 +555,10 @@ static void IMM_FreeAllImmHkl(void)
             FreeLibrary(ptr->hIME);
         }
         if (ptr->UIWnd)
+        {
+            SendMessageA(ptr->UIWnd, WM_WINE_IME_DESTROY, 0, 0);
             DestroyWindow(ptr->UIWnd);
+        }
         HeapFree(GetProcessHeap(),0,ptr);
     }
 }
@@ -3185,6 +3189,7 @@ static void init_messages(void)
     WM_MSIME_RECONVERT = RegisterWindowMessageW(L"MSIMEReconvert");
     WM_MSIME_QUERYPOSITION = RegisterWindowMessageW(L"MSIMEQueryPosition");
     WM_MSIME_DOCUMENTFEED = RegisterWindowMessageW(L"MSIMEDocumentFeed");
+    WM_WINE_IME_DESTROY = RegisterWindowMessageW(L"__WINE_IME_DESTROY");
     initialized = TRUE;
 }

@@ -3198,16 +3203,18 @@ LRESULT WINAPI __wine_ime_wnd_proc(HWND hwnd, UINT msg, WPARAM wparam, LPARAM lp
         init_messages();
         return TRUE;

-    case WM_DESTROY:
+    case WM_IME_INTERNAL:
+        return ime_internal_msg(wparam, lparam);
+    }
+
+    if (msg == WM_WINE_IME_DESTROY)
+    {
+        HWND default_hwnd = ImmGetDefaultIMEWnd(0);
+        if (!default_hwnd || hwnd == default_hwnd)
         {
-            HWND default_hwnd = ImmGetDefaultIMEWnd(0);
-            if (!default_hwnd || hwnd == default_hwnd)
-                imm_couninit_thread(TRUE);
+            imm_couninit_thread(TRUE);
         }
         return TRUE;
-
-    case WM_IME_INTERNAL:
-        return ime_internal_msg(wparam, lparam);
     }

if (is_ime_ui_msg(msg))

[PATCH 1/1] imm32: Prevent deadlock on unload

[Marvin]

Hi,

It looks like your patch introduced the new failures shown below.
Please investigate and fix them before resubmitting your patch.
If they are not new, fixing them anyway would help a lot. Otherwise
please ask for the known failures list to be updated.

The tests also ran into some preexisting test failures. If you know how
to fix them that would be helpful. See the TestBot job for the details:

The full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=126508

Your paranoid android.

=== debian11 (32 bit report) ===

d3d9:
stateblock: Timeout
visual: Timeout

d3dcompiler_43:
asm: Timeout
blob: Timeout
hlsl_d3d11: Timeout
hlsl_d3d9: Timeout
reflection: Timeout

d3dcompiler_46:
asm: Timeout
blob: Timeout
hlsl_d3d11: Timeout
hlsl_d3d9: Timeout
reflection: Timeout

d3dcompiler_47:
asm: Timeout
blob: Timeout
hlsl_d3d11: Timeout
hlsl_d3d9: Timeout
reflection: Timeout

d3drm:
d3drm: Timeout
vector: Timeout

d3dx10_34:
d3dx10: Timeout

d3dx10_35:
d3dx10: Timeout

Report validation errors:
d3dx10: Timeout

=== debian11 (build log) ===

WineRunWineTest.pl:error: The task timed out

=== debian11b (64 bit WoW report) ===

imm32:
imm32.c:2309: Test failed: CoGetApartmentType returned 0
imm32.c:2309: Test failed: type 3
imm32.c:2331: Test failed: CoGetApartmentType returned 0
imm32.c:2331: Test failed: type 3
imm32.c:2341: Test failed: CoGetApartmentType returned 0
imm32.c:2341: Test failed: type 1
imm32.c:2345: Test failed: CoGetApartmentType returned 0
imm32.c:2345: Test failed: type 1
imm32.c:2347: Test failed: type 1
imm32.c:2349: Test failed: CoGetApartmentType returned 0
imm32.c:2349: Test failed: type 1

[Fabian Maurer (＠DarkShadow44)]

Just noticed, we should probably also wait until the window is closed, so we don't run the code inside __wine_ime_wnd_proc after it's unloaded. An idea how to do that?