September 2022 - Wine-gitlab

[PATCH v2 0/6] MR832: uiautomationcore: Implement basic HWND based HUIANODE behavior.
by Connor McAdams (＠cmcadams) 14 Sep '22

14 Sep '22

This series has been split up more than the prior two, keeping the non-tests diffs below 200. Hopefully it's easier to review now. If not, please let me know and I can _try_ to split it further, but I'm not sure it can get much more split up than this. -- v2: uiautomationcore: Add tests for UiaNodeFromHandle. uiautomationcore: Create UI Automation client thread. uiautomationcore: Implement UiaNodeFromHandle. uiautomationcore: Shutdown provider thread when all returned nodes are released. uiautomationcore: Increment module reference count when starting provider thread. uiautomationcore: Implement UiaReturnRawElementProvider. https://gitlab.winehq.org/wine/wine/-/merge_requests/832

2 9

[PATCH 0/2] MR845: tests: urlmon and wininet fixes
by Jacek Caban (＠jacek) 14 Sep '22

14 Sep '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/845

3 3

[PATCH 0/7] MR832: uiautomationcore: Implement basic HWND based HUIANODE behavior.
by Connor McAdams (＠cmcadams) 14 Sep '22

14 Sep '22

This series has been split up more than the prior two, keeping the non-tests diffs below 200. Hopefully it's easier to review now. If not, please let me know and I can _try_ to split it further, but I'm not sure it can get much more split up than this. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/832

3 23

Re: [PATCH v4 0/8] MR814: win32u: Avoid copies in display adapter settings logic.
by Henri Verbeet (＠hverbeet) 14 Sep '22

14 Sep '22

This [patch](/uploads/c1e4bf81d3bd864581162cfd244d6b3d/patch.diff) helps for me; I'll send it in, along with some other fixes. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/814#note_8417

1 0

Re: [PATCH v2 0/2] MR820: Implement application hives in the registry
by Huw Davies (＠huw) 14 Sep '22

14 Sep '22

On Wed Sep 14 11:53:01 2022 +0000, Santino Mazza wrote: > Because it simplifies the implementation for saving the registry into > the file when NtClose get's called. If I want to keep the object handle > implementation the server should take control for closing the file > handle, and I don't like that idea so much, I think the caller should be > responsible for closing the objects it opens. Ok, let me look into that more. There are a couple more changes you could make in the meantime: - Split the `REG_LOAD_APP_KEY` logic and the return key logic into two commits. - Move the `NtClose()` outside of the `SERVER_START_REQ` block. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/820#note_8415

1 0

Re: [PATCH v2 0/2] MR820: Implement application hives in the registry
by Santino Mazza (＠tati1454) 14 Sep '22

14 Sep '22

On Wed Sep 14 10:35:23 2022 +0000, Huw Davies wrote: > Why are you now sending the filename, rather than the file object > handle, to the server? Because it simplifies the implementation for saving the registry into the file when NtClose get's called. If I want to keep the object handle implementation the server should take control for closing the file handle, and I don't like that idea so much, I think the caller should be responsible for closing the objects it opens. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/820#note_8405

1 0

Re: [PATCH v4 0/8] MR814: win32u: Avoid copies in display adapter settings logic.
by Bernhard Kölbl 14 Sep '22

14 Sep '22

On Wed Sep 14 10:36:44 2022 +0000, Alexandre Julliard wrote: > > I bisected the failure and it looks like it's caused instead by > 44d6f2adbcca119faebc095f3cfbeeafc50d04f3, from > https://gitlab.winehq.org/wine/wine/-/merge_requests/657. Reverting it > alone fixes it: > Ah sorry about that, reverting your MR fixed it for me but it's possible > that I screwed up the test. - Just a test message from E-Mail. - Am Mi., 14. Sept. 2022 um 11:05 Uhr schrieb Alexandre Julliard (@julliard) <gitlab(a)winehq.org>: -- https://gitlab.winehq.org/wine/wine/-/merge_requests/814#note_8404

1 0

Re: [PATCH v2 0/2] MR820: Implement application hives in the registry
by Huw Davies (＠huw) 14 Sep '22

14 Sep '22

Huw Davies (@huw) commented about server/protocol.def: > > /* Load a registry branch from a file */ > @REQ(load_registry) > - obj_handle_t file; /* file to load from */ > + unsigned int flags; /* flags */ > + unsigned int access; /* wanted access rights */ > VARARG(objattr,object_attributes); /* object attributes */ > + VARARG(filename,string); /* file to load name */ Why are you now sending the filename, rather than the file object handle, to the server? -- https://gitlab.winehq.org/wine/wine/-/merge_requests/820#note_8403

1 0

Re: [PATCH v2 0/2] MR820: Implement application hives in the registry
by Huw Davies (＠huw) 14 Sep '22

14 Sep '22

Huw Davies (@huw) commented about dlls/advapi32/tests/registry.c: > /* Test simple key load */ > /* Check if the changes are saved */ > ret = RegLoadAppKeyA(hivefilepath, &appkey, KEY_READ | KEY_WRITE, 0, 0); > - ok(ret == ERROR_SUCCESS, "expected ERROR_SUCCESS, got %ld\n", ret); > - ok(appkey != NULL, "got a null key\n"); > + todo_wine ok(ret == ERROR_SUCCESS, "expected ERROR_SUCCESS, got %ld\n", ret); > + todo_wine ok(appkey != NULL, "got a null key\n"); I'd suggest swapping the order of these commits so that you don't end up having to flip-flop these `todo_wine`s. Also note that there's now no need to add the Signed-off-by tags. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/820#note_8402

1 0

Re: [PATCH v4 0/8] MR814: win32u: Avoid copies in display adapter settings logic.
by Alexandre Julliard (＠julliard) 14 Sep '22

14 Sep '22

> I bisected the failure and it looks like it's caused instead by 44d6f2adbcca119faebc095f3cfbeeafc50d04f3, from https://gitlab.winehq.org/wine/wine/-/merge_requests/657. Reverting it alone fixes it: Ah sorry about that, reverting your MR fixed it for me but it's possible that I screwed up the test. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/814#note_8398

1 0