[PATCH v3 0/2] MR9246: winemac: Use event handle to detect macdrv ime event processing done.
Since !9128, macdrv ime processing has been broken. In my opinion, this seems to be an issue with the macdrv ime code rather than a problem with !9128. In macdrv_ImeProcessKey: ``` macdrv_send_keydown_to_input_source(flags, repeat, keyc, himc, &done); while (!done) NtUserMsgWaitForMultipleObjectsEx(0, NULL, INFINITE, QS_POSTMESSAGE | QS_SENDMESSAGE, 0); ``` Within the macdrv WINE_IME_POST_UPDATE code path, there are no message post/send operations, causing it to wait infinitely here. I haven't analyzed in detail why it worked before, but I guess it escaped the wait due to unrelated messages. This code likely represents a remnant of the old macdrv ime code. I recall the old macdrv ime code used SendMessage(). In ime processing, some key inputs reach macdrv_sent_text_input without calling macdrv_ime_set_text. Therefore, I modified the code to clearly detect when macdrv_sent_text_input is reached. I left a working fallback code in case NtCreateEvent() fails. But that might be superfluous. -- v3: win32u: Fix typo. winemac: Use event handle to detect macdrv ime processing done. https://gitlab.winehq.org/wine/wine/-/merge_requests/9246
From: Byeongsik Jeon <bsjeon(a)hanmail.net> --- dlls/winemac.drv/cocoa_window.m | 6 ++++-- dlls/winemac.drv/event.c | 4 ++++ dlls/winemac.drv/keyboard.c | 12 ++++++++++-- dlls/winemac.drv/macdrv.h | 1 + dlls/winemac.drv/macdrv_cocoa.h | 5 +++-- dlls/winemac.drv/macdrv_main.c | 4 ++++ 6 files changed, 26 insertions(+), 6 deletions(-) diff --git a/dlls/winemac.drv/cocoa_window.m b/dlls/winemac.drv/cocoa_window.m index 326b5925c5f..2dd7d180c24 100644 --- a/dlls/winemac.drv/cocoa_window.m +++ b/dlls/winemac.drv/cocoa_window.m @@ -4034,13 +4034,14 @@ uint32_t macdrv_window_background_color(void) } /*********************************************************************** - * macdrv_send_keydown_to_input_source + * macdrv_ime_process_key * * Sends a key down event to the active window's inputContext so that it can be * processed by input sources (AKA IMEs). This is only called when there is an * active non-keyboard input source. */ -void macdrv_send_keydown_to_input_source(unsigned int flags, int repeat, int keyc, void* himc, int* done) +void macdrv_ime_process_key(int keyc, unsigned int flags, int repeat, void *himc, + int *done, void *ime_done_event) { OnMainThreadAsync(^{ BOOL ret; @@ -4080,6 +4081,7 @@ void macdrv_send_keydown_to_input_source(unsigned int flags, int repeat, int key event = macdrv_create_event(SENT_TEXT_INPUT, window); event->sent_text_input.handled = ret; event->sent_text_input.done = done; + event->sent_text_input.ime_done_event = ime_done_event; [[window queue] postEvent:event]; macdrv_release_event(event); }); diff --git a/dlls/winemac.drv/event.c b/dlls/winemac.drv/event.c index b97b876cb70..6841fd3603e 100644 --- a/dlls/winemac.drv/event.c +++ b/dlls/winemac.drv/event.c @@ -190,8 +190,12 @@ static void macdrv_im_set_text(const macdrv_event *event) */ static void macdrv_sent_text_input(const macdrv_event *event) { + HANDLE ime_done_event = event->sent_text_input.ime_done_event; + TRACE_(imm)("handled: %s\n", event->sent_text_input.handled ? "TRUE" : "FALSE"); + *event->sent_text_input.done = event->sent_text_input.handled ? 1 : -1; + if (ime_done_event) NtSetEvent(ime_done_event, NULL); } diff --git a/dlls/winemac.drv/keyboard.c b/dlls/winemac.drv/keyboard.c index 645d082b166..6c615d10e3d 100644 --- a/dlls/winemac.drv/keyboard.c +++ b/dlls/winemac.drv/keyboard.c @@ -1139,8 +1139,16 @@ UINT macdrv_ImeProcessKey(HIMC himc, UINT wparam, UINT lparam, const BYTE *key_s TRACE("flags 0x%08x keyc 0x%04x\n", flags, keyc); - macdrv_send_keydown_to_input_source(flags, repeat, keyc, himc, &done); - while (!done) NtUserMsgWaitForMultipleObjectsEx(0, NULL, INFINITE, QS_POSTMESSAGE | QS_SENDMESSAGE, 0); + if (!thread_data->ime_done_event) + { + NTSTATUS status; + status = NtCreateEvent(&thread_data->ime_done_event, EVENT_ALL_ACCESS, NULL, + SynchronizationEvent, FALSE); + if (status != STATUS_SUCCESS) ERR("NtCreateEvent call failed.\n"); + } + + macdrv_ime_process_key(keyc, flags, repeat, himc, &done, thread_data->ime_done_event); + NtUserMsgWaitForMultipleObjectsEx(1, &thread_data->ime_done_event, INFINITE, 0, 0); return done > 0; } diff --git a/dlls/winemac.drv/macdrv.h b/dlls/winemac.drv/macdrv.h index 22e96fa1129..5b9515f086f 100644 --- a/dlls/winemac.drv/macdrv.h +++ b/dlls/winemac.drv/macdrv.h @@ -112,6 +112,7 @@ static inline RECT rect_from_cgrect(CGRect cgrect) HKL active_keyboard_layout; WORD keyc2vkey[128]; WORD keyc2scan[128]; + HANDLE ime_done_event; }; extern struct macdrv_thread_data *macdrv_init_thread_data(void); diff --git a/dlls/winemac.drv/macdrv_cocoa.h b/dlls/winemac.drv/macdrv_cocoa.h index f91e9f9f169..bd6af9ed25e 100644 --- a/dlls/winemac.drv/macdrv_cocoa.h +++ b/dlls/winemac.drv/macdrv_cocoa.h @@ -381,6 +381,7 @@ extern int macdrv_set_display_mode(const struct macdrv_display* display, struct { int handled; int *done; + void *ime_done_event; } sent_text_input; struct { macdrv_status_item item; @@ -545,8 +546,8 @@ extern void macdrv_order_cocoa_window(macdrv_window w, macdrv_window prev, extern int macdrv_get_view_backing_size(macdrv_view v, int backing_size[2]); extern void macdrv_set_view_backing_size(macdrv_view v, const int backing_size[2]); extern uint32_t macdrv_window_background_color(void); -extern void macdrv_send_keydown_to_input_source(unsigned int flags, int repeat, int keyc, - void* data, int* done); +extern void macdrv_ime_process_key(int keyc, unsigned int flags, int repeat, void *data, + int *done, void *ime_done_event); extern int macdrv_is_any_wine_window_visible(void); diff --git a/dlls/winemac.drv/macdrv_main.c b/dlls/winemac.drv/macdrv_main.c index 20295d1c6ee..f5e2b5e08d2 100644 --- a/dlls/winemac.drv/macdrv_main.c +++ b/dlls/winemac.drv/macdrv_main.c @@ -464,6 +464,8 @@ void macdrv_ThreadDetach(void) macdrv_destroy_event_queue(data->queue); if (data->keyboard_layout_uchr) CFRelease(data->keyboard_layout_uchr); + if (data->ime_done_event) + NtClose(data->ime_done_event); free(data); /* clear data in case we get re-entered from user32 before the thread is truly dead */ NtUserGetThreadInfo()->driver_data = 0; @@ -523,6 +525,8 @@ struct macdrv_thread_data *macdrv_init_thread_data(void) NtTerminateProcess(0, 1); } + data->ime_done_event = NULL; + macdrv_get_input_source_info(&data->keyboard_layout_uchr, &data->keyboard_type, &data->iso_keyboard, &input_source); data->active_keyboard_layout = macdrv_get_hkl_from_source(input_source); CFRelease(input_source); -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/9246
From: Byeongsik Jeon <bsjeon(a)hanmail.net> --- dlls/win32u/imm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/dlls/win32u/imm.c b/dlls/win32u/imm.c index 8b416a654d9..3337f66e291 100644 --- a/dlls/win32u/imm.c +++ b/dlls/win32u/imm.c @@ -683,7 +683,7 @@ LRESULT ime_driver_call( HWND hwnd, enum wine_ime_call call, WPARAM wparam, LPAR res = TRUE; } - TRACE( "processing scan %#x, vkey %#x -> %lu\n", LOWORD(wparam), HIWORD(lparam), res ); + TRACE( "processing vkey %#x, scan %#x -> %lu\n", LOWORD(wparam), HIWORD(lparam), res ); return res; } case WINE_IME_TO_ASCII_EX: -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/9246
This merge request was approved by Rémi Bernon. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9246
Within the macdrv WINE_IME_POST_UPDATE code path, there are no message post/send operations
We still need `NtUserMsgWaitForMultipleObjectsEx` to drain the driver events though, since that is the mechanism by which `macdrv_sent_text_input` eventually gets called (which still happens here and would break if replaced by `NtWaitForSingleObject`). Creating an NT event just for this purpose seems a bit overkill IMO, there is really no reason this call to `macdrv_ime_process_key` needs to be asynchronous with the current design in the first place. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9246#note_119300
On Wed Oct 22 11:20:09 2025 +0000, Marc-Aurel Zent wrote:
Within the macdrv WINE_IME_POST_UPDATE code path, there are no message post/send operations We still need `NtUserMsgWaitForMultipleObjectsEx` to drain the driver events though, since that is the mechanism by which `macdrv_sent_text_input` eventually gets called (which still happens here and would break if replaced by `NtWaitForSingleObject`). Creating an NT event just for this purpose seems a bit overkill IMO, there is really no reason this call to `macdrv_ime_process_key` needs to be asynchronous with the current design in the first place. NtUserMsgWaitForMultipleObjectsEx will not return anymore until the queue is signaled. With the given masks it will unlikely be.
-- https://gitlab.winehq.org/wine/wine/-/merge_requests/9246#note_119301
On Wed Oct 22 11:20:09 2025 +0000, Rémi Bernon wrote:
NtUserMsgWaitForMultipleObjectsEx will not return anymore until the queue is signaled. With the given masks it will unlikely be. Also, this doesn't replace the wait with `NtWaitForSingleObject`, it only adds the event so that the wait eventually returns even if the queue isn't signaled.
-- https://gitlab.winehq.org/wine/wine/-/merge_requests/9246#note_119302
NtUserMsgWaitForMultipleObjectsEx will not return anymore until the queue is signaled. With the given masks it will unlikely be.
Yeah, and if it did that would be quite problematic.
Also, this doesn't replace the wait with `NtWaitForSingleObject`, it only adds the event so that the wait eventually returns even if the queue isn't signaled.
The `NtWaitForSingleObject` was just a hypothetical, not something the current MR does; maybe my phrasing was a bit messy there. In its current form it is functionally correct, I am just wondering if there is a simpler solution to the problem here. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9246#note_119306
participants (4)
-
Byeongsik Jeon -
Byeongsik Jeon (＠bsjeon)
-
Marc-Aurel Zent (＠mzent)
-
Rémi Bernon