- wine-gitlab - winehq.org

Re: [PATCH v5 0/3] MR1031: mshtml: Document element props improvements. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1031

1 0

Re: [PATCH v2 0/2] MR1050: winegstreamer: Register AAC / WMV decoder MFT stubs. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1050

1 0

Re: [PATCH v2 0/2] MR1047: Link vcruntime140_1 and msvcp140 to vcruntime140. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1047

1 0

Re: [PATCH v2 0/1] MR1028: nsi: Use the correct members for the error and discard counts. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1028

1 0

Re: [PATCH v6 0/1] MR745: ntdll: Improve stub of NtQueryEaFile. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/745

1 0

Re: [PATCH v5 0/3] MR1033: winhttp: Add a few websocket options. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1033

1 0

Re: [PATCH v2 0/2] MR1049: winegstreamer: More factoring for MF transforms. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1049

1 0

Re: [PATCH v3 0/3] MR1027: windows.devices.enumeration: Create stub DeviceAccessInformation class. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1027

1 0

Re: [PATCH v2 0/1] MR1044: windowscodecs/tests: Accpet new error code from win11. - approved
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1044

1 0

[PATCH v3 0/1] MR1011: user32: Correctly detect and assign an int resource
by Alistair Leslie-Hughes (＠alesliehughes) 13 Oct '22

13 Oct '22

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=53646 Signed-off-by: Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> -- v3: user.exe16: Correctly detect and assign an int resource https://gitlab.winehq.org/wine/wine/-/merge_requests/1011

4 4

[PATCH v4 0/3] MR1031: mshtml: Document element props improvements.
by Gabriel Ivăncescu (＠insn) 13 Oct '22

13 Oct '22

-- v4: mshtml: Don't create dynamic prop before checking if elem prop even exists. mshtml: Allow accessing some document elements as props via id. mshtml: Expose props via element name only for specific element types. https://gitlab.winehq.org/wine/wine/-/merge_requests/1031

4 6

Re: [PATCH v5 0/5] MR944: win32u: Make desktop resize logic asynchronous.
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

On Thu Oct 13 19:11:33 2022 +0000, Alexandre Julliard wrote: > This is causing random deadlocks when running the d3d tests here; > basically the entire Wine session freezes. > Applying just the first patch is enough to trigger it. Sorry, I'll have a look. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/944#note_10591

1 0

[PATCH 0/2] MR1050: winegstreamer: Register AAC / WMV decoder MFT stubs.
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1050

4 4

[PATCH 0/2] MR1049: winegstreamer: More factoring for MF transforms.
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1049

4 4

Re: [PATCH v5 0/5] MR944: win32u: Make desktop resize logic asynchronous.
by Alexandre Julliard (＠julliard) 13 Oct '22

13 Oct '22

This is causing random deadlocks when running the d3d tests here; basically the entire Wine session freezes. Applying just the first patch is enough to trigger it. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/944#note_10586

1 0

Re: [PATCH v2 0/10] MR969: combase: CoMsgWaitForMultipleHandles message pump.
by Huw Davies (＠huw) 13 Oct '22

13 Oct '22

I'd suggest moving some of the existing code into helper functions first. It's already hard to follow and these changes to it aren't making it any simpler. Also, ten commits per MR is rather too many (especially when they're non trivial). Try sending the first five or so. If you want to show where it's going, then include a link to a branch which contains the finished set. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/969#note_10582

1 0

[PATCH v8 0/4] MR23: shell32: Sanitize Program Manager icon and group names
by Alex Henrie (＠alexhenrie) 13 Oct '22

13 Oct '22

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52506 Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> -- v8: shell32: Sanitize Program Manager icon and group names. shell32: Introduce combine_path helper for DDE. shell32: Use standard C functions for memory allocation in dde.c. shell32: Move strndupW to dde.c. https://gitlab.winehq.org/wine/wine/-/merge_requests/23

4 8

[PATCH 0/1] MR1053: include: Add LV_COLUMN[AW].
by Alex Henrie (＠alexhenrie) 13 Oct '22

13 Oct '22

Needed to compile Tera Term. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1053

3 2

[PATCH v2 0/1] MR16: vkd3d-shader/hlsl: Lower complex broadcasts.
by Francisco Casas (＠fcasas) 13 Oct '22

13 Oct '22

A rebase of the "Lower complex broadcasts" patch. -- v2: vkd3d-shader/hlsl: Lower broadcasts to complex types. https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/16

2 3

[PATCH v3 0/3] MR1031: mshtml: Document element props improvements.
by Gabriel Ivăncescu (＠insn) 13 Oct '22

13 Oct '22

-- v3: mshtml: Don't create dynamic prop before checking if elem prop even exists. mshtml: Allow accessing some document elements as props via id. mshtml: Expose props via element name only for specific element types. https://gitlab.winehq.org/wine/wine/-/merge_requests/1031

4 5

[PATCH v9 0/7] MR551: win32u: Complete support for display modes in nulldrv.
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

With this series it's now possible to run and pass `user32:monitor` and `user32:sysparams` tests with nulldrv, and so most `user32` tests (except for a few desktop cursor position tests). This still requires some prefix configuration to enable the nulldrv driver, or a change like https://gitlab.winehq.org/rbernon/wine/-/commit/753368ad0ec52f03f8d6e78ca79… to enable it when `DISPLAY` environment variable is unset. This then shows that some of the user32 tests are failing with winex11 but passing with nulldrv, as in https://gitlab.winehq.org/rbernon/wine/-/commit/6d5f4109a514a0dc266899fcacf…. -- v9: win32u: Read mode from the registry if GetCurrentDisplaySettings fails. win32u: Write display settings to the registry in apply_display_settings. win32u: Lock display devices while applying display settings. win32u: Use an internal WINE_DM_PRIMARY_DEVICE flag on primary adapter modes. win32u: Remove the device name parameter from GetCurrentDisplaySettings. win32u: Force update display cache after NtUserChangeDisplaySettingsEx. win32u: Add a BOOL force parameter to update_display_cache. https://gitlab.winehq.org/wine/wine/-/merge_requests/551

3 9

[PATCH 0/2] MR1047: Link vcruntime140_1 and msvcp140 to vcruntime140.
by Paul Gofman (＠gofman) 13 Oct '22

13 Oct '22

I used the following test program (compiled with MSVC with /MD): ``` #include <windows.h> #include <stdio.h> typedef void** (__cdecl *pcurrent_exception)(void); int main() { HMODULE hvcr = LoadLibraryA("vcruntime140.dll"); HMODULE ucrtb = LoadLibraryA("ucrtbase.dll"); pcurrent_exception p__current_exception1, p__current_exception2; p__current_exception1 = (pcurrent_exception)GetProcAddress(hvcr, "__current_exception"); p__current_exception2 = (pcurrent_exception)GetProcAddress(ucrtb, "__current_exception"); try { throw("ex1"); } catch (const char *s) { EXCEPTION_RECORD* rec1 = (EXCEPTION_RECORD *)*p__current_exception1(); EXCEPTION_RECORD* rec2 = (EXCEPTION_RECORD *)*p__current_exception2(); printf("Exception %s, rec1 %p (code %#lx), rec2 %p.\n", s, rec1, rec1->ExceptionCode, rec2); } } ``` This outputs the following on Windows: ``` Exception ex1, rec1 00000089CAF9F990 (code 0xe06d7363), rec2 0000000000000000. ``` So I think that confirms that: 1. vcruntime140 and ucrtbase have distinct local data (or at least current exception data) on Windows; 2. vcruntime140_1 links to vcruntime140. 3. msvcp140 should link to vcruntime140's exception data too as it is apparently able to get the correct current exception in ?__ExceptionPtrCurrentException@@YAXPEAX@Z. These patches allow the current exception to be correctly located when there are mixed builtin and native VC runtime 140. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1047

4 5

[PATCH v4 0/3] MR1033: winhttp: Add a few websocket options.
by Paul Gofman (＠gofman) 13 Oct '22

13 Oct '22

-- v4: winhttp: Support WINHTTP_OPTION_WEB_SOCKET_SEND_BUFFER_SIZE. winhttp: Stub WINHTTP_OPTION_WEB_SOCKET_RECEIVE_BUFFER_SIZE. https://gitlab.winehq.org/wine/wine/-/merge_requests/1033

4 6

[PATCH 0/4] MR1051: opengl32: Cleanup make_opengl in preparation for https://gitlab.winehq.org/wine/wine/-/merge_requests/1010
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1051

3 5

[PATCH v8 0/6] MR551: win32u: Complete support for display modes in nulldrv.
by Rémi Bernon (＠rbernon) 13 Oct '22

13 Oct '22

With this series it's now possible to run and pass `user32:monitor` and `user32:sysparams` tests with nulldrv, and so most `user32` tests (except for a few desktop cursor position tests). This still requires some prefix configuration to enable the nulldrv driver, or a change like https://gitlab.winehq.org/rbernon/wine/-/commit/753368ad0ec52f03f8d6e78ca79… to enable it when `DISPLAY` environment variable is unset. This then shows that some of the user32 tests are failing with winex11 but passing with nulldrv, as in https://gitlab.winehq.org/rbernon/wine/-/commit/6d5f4109a514a0dc266899fcacf…. -- v8: win32u: Read mode from the registry if GetCurrentDisplaySettings fails. win32u: Write display settings to the registry in apply_display_settings. win32u: Lock display devices while applying display settings. win32u: Use an internal WINE_DM_PRIMARY_DEVICE flag on primary adapter modes. win32u: Force update display cache after NtUserChangeDisplaySettingsEx. win32u: Add a BOOL force parameter to update_display_cache. https://gitlab.winehq.org/wine/wine/-/merge_requests/551

3 7

Re: [PATCH v3 0/1] MR1040: imm32/tests: Don't expect IME window if the parent is message-only window descendant. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1040

1 0

Re: [PATCH v3 0/7] MR1034: msvcp140: Do not import from msvcp120. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1034

1 0

Re: [PATCH v2 0/1] MR992: imm32/tests: Mark tests failing randomly on Windows as flaky. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/992

1 0

Re: [PATCH v2 0/2] MR1038: Sapi datakey - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1038

1 0

Re: [PATCH v2 0/2] MR1039: Schannel test - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1039

1 0

Re: [PATCH v5 0/5] MR944: win32u: Make desktop resize logic asynchronous. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/944

1 0

Re: [PATCH v3 0/1] MR962: wined3d: Use a Wine debug channel for vkd3d debug output. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/962

1 0

Re: [PATCH v2 0/3] MR1036: qcap: Various video capture patches. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1036

1 0

Re: [PATCH v2 0/6] MR971: winegstreamer: Complete transition to using IWMSyncReader2 from the WM async reader. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/971

1 0

Re: [PATCH v3 0/5] MR937: winegstreamer: Small refactor of wg_format video formats and major types. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/937

1 0

Re: [PATCH v3 0/2] MR920: kernel32: Remove shared heap functionality. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/920

1 0

Re: [PATCH v8 0/6] MR908: uiautomationcore: Add more nested node features. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/908

1 0

Re: [PATCH v4 0/5] MR25: vkd3d-shader/hlsl: Implement asuint() and completely move init_node() to hlsl.c. - approved
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/25

1 0

[PATCH v2 0/6] MR1001: winebuild: Reimplement dlltool for import lib and delay import lib generation.
by Rémi Bernon (＠rbernon) 12 Oct '22

12 Oct '22

This is slightly different from dlltool but I think it should be compatible. The transition is done by first replacing dlltool with its bugs, and fixing them in separate changes. I based the ARM / AARCH64 implementation on the existing code around, but I have no idea if it is correct, and dlltool also doesn't include any delay load implementation for ARM. -- v2: winebuild: Put the delay import descriptor in data section. winebuild: Fix import hint value for symbols imported by name. winebuild: Implement import lib generation to replace dlltool. winebuild: Split output_static_lib into output_(import|static)_lib. winebuild: Use asm_name to add underscore prefix in asm_globl. winebuild: Pass the desired symbol name to get_link_name. https://gitlab.winehq.org/wine/wine/-/merge_requests/1001

5 11

[PATCH v2 0/3] MR1027: windows.devices.enumeration: Create stub DeviceAccessInformation class.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

-- v2: windows.devices.enumeration: Add stub IDeviceAccessInformation interface. windows.devices.enumeration: Create stub DeviceAccessInformation class. include/windows.devices.enumeration: Define DeviceAccessInformation class. https://gitlab.winehq.org/wine/wine/-/merge_requests/1027

4 6

[PATCH v3 0/3] MR1033: winhttp: Add a few websocket options.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

-- v3: winhttp: Support WINHTTP_OPTION_WEB_SOCKET_SEND_BUFFER_SIZE. winhttp: Stub WINHTTP_OPTION_WEB_SOCKET_RECEIVE_BUFFER_SIZE. winhttp: Support WINHTTP_OPTION_WEB_SOCKET_KEEPALIVE_INTERVAL. https://gitlab.winehq.org/wine/wine/-/merge_requests/1033

3 5

[PATCH 0/1] MR1044: windowscodecs/tests: Accpet new error code from win11.
by Esme Povirk (＠madewokherd) 12 Oct '22

12 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1044

2 1

[PATCH v7 0/2] MR23: shell32: Sanitize Program Manager icon and group names
by Alex Henrie (＠alexhenrie) 12 Oct '22

12 Oct '22

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52506 Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> -- v7: shell32: Sanitize Program Manager icon and group names. shell32: Introduce combine_path helper for DDE. https://gitlab.winehq.org/wine/wine/-/merge_requests/23

4 5

[PATCH v4 0/5] MR874: mfmediaengine: Avoid vertically flipping video.
by Rémi Bernon (＠rbernon) 12 Oct '22

12 Oct '22

This shows that media session samples are correctly oriented, but that mfmediaengine renders them upside down. -- v4: mfmediaengine: Remove vertical flipping of video frames. mfmediaengine/tests: Check IMFMediaEngine_TransferVideoFrames output orientation. mfmediaengine/tests: Pass a device and output format to create_media_engine. mf/tests: Check sample grabber RGB / NV12 orientation. mf/tests: Factor test grabber callback implementations. https://gitlab.winehq.org/wine/wine/-/merge_requests/874

4 9

[PATCH 0/1] MR1041: shell32: Introduce combine_path helper for DDE.
by Alex Henrie (＠alexhenrie) 12 Oct '22

12 Oct '22

Split from merge request !23 per Huw's suggestion -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1041

4 5

[PATCH v2 0/7] MR1034: msvcp140: Do not import from msvcp120.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

The practical aspect of this (besides that msvcp140 just should not probably load msvcp120) is that for an app using VC runtime 140 mspcp140.__ExceptionPtrCurrentException gets the thread data from msvcp120.dll while vcruntime140_1 sets those thread data from ucrtbase.dll which are different data. So the application calling mspcp140.__ExceptionPtrCurrentException in catch block gets NULL exception while it was technically set by vcruntime140_1 before calling the app's catch block handler. After this patchset all the current exception data should end up in ucrtbase for VC runtime 140. There is a bit more to it (not covered by this patchset) when there is a mix of native and builtin VC runtime 140 dlls. From the app behaviour and some tests it looks like with native DLLs msvcp140 and vcruntime140_1 handle current exception data through vcruntime140.dll and vcruntime140.dll does not forward that to ucrtbase. I am not sure if the latter (not forwarding vcruntime140->ucrtbase) is practically important but to avoid the similar problem when mixing native and builtin VC runtime DLLs we should also probably link msvcp140 and vcruntime140_1 to vcruntime140. -- v2: msvcp140: Import __ExceptionPtrCompare implementation. msvcp140: Import __ExceptionPtrCopyException implementation. msvcp140: Import __ExceptionPtrToBool implementation. msvcp140: Import __ExceptionPtrCurrentException implementation. msvcp140: Import __ExceptionPtrRethrow implementation. msvcp140: Import __ExceptionPtrAssign implementation. msvcp140: Import __ExceptionPtrCopy implementation. https://gitlab.winehq.org/wine/wine/-/merge_requests/1034

4 10

[PATCH 0/1] MR1024: xmllite/writer: Correctly partially duplicate strings
by David Kahurani (＠kahurani) 12 Oct '22

12 Oct '22

Currently, the code copies one extra character than requested and does not terminate the string. Signed-off-by: David Kahurani <k.kahurani(a)gmail.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1024

3 3

[PATCH v2 0/3] MR1033: winhttp: Add a few websocket options.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

-- v2: winhttp: Support WINHTTP_OPTION_WEB_SOCKET_SEND_BUFFER_SIZE. https://gitlab.winehq.org/wine/wine/-/merge_requests/1033

5 12

[PATCH v2 0/2] MR1031: mshtml: Document element props improvements.
by Gabriel Ivăncescu (＠insn) 12 Oct '22

12 Oct '22

-- v2: mshtml: Don't create dynamic prop before checking if elem prop even exists. mshtml: Allow accessing document elements as props via id if they have name. https://gitlab.winehq.org/wine/wine/-/merge_requests/1031

4 7

Re: [PATCH v2 0/1] MR1040: imm32/tests: Don't expect IME window if the parent is message-only window descendant.
by Akihiro Sagawa (＠sgwaki) 12 Oct '22

12 Oct '22

I just mark the test as `todo_wine` for now because I'm not hundred percent sure that the fix for Wine is conducive. Do we need the fix for Wine, too? Here is a draft patch for Wine, https://gitlab.winehq.org/sgwaki/wine/-/commit/84b27c904d93ef2f45c230681172…. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1040#note_10394

1 0

[PATCH 0/2] MR1039: Schannel test
by Hans Leidekker (＠hans) 12 Oct '22

12 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1039

2 2

Re: [PATCH v2 0/2] MR1037: win32u: Add font link for MS UI Gothic.
by Huw Davies (＠huw) 12 Oct '22

12 Oct '22

Could you split the first commit further? Just make the code change with the commit msg something like "win32u: Add mechanism for font specific system links", but don't add the MS UI Gothic entries. A second commit would add the Gothic entries, and a third the additional Tahoma ones. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1037#note_10391

1 0

[PATCH v4 0/5] MR944: win32u: Make desktop resize logic asynchronous.
by Rémi Bernon (＠rbernon) 12 Oct '22

12 Oct '22

To avoid direct callbacks from display drivers to win32u while changing display settings. -- v4: winemac.drv: Resize desktop window on WM_DISPLAYCHANGE message. winex11.drv: Resize desktop window on WM_DISPLAYCHANGE message. win32u: Send WM_DISPLAYCHANGE message to the desktop window. winex11.drv: Send WM_X11DRV_CLIP_CURSOR_REQUEST message from the deskop. winex11.drv: Sync window positions in a WM_X11DRV_DESKTOP_RESIZED message. https://gitlab.winehq.org/wine/wine/-/merge_requests/944

3 9

[PATCH v2 0/1] MR962: wined3d: Use a Wine debug channel for vkd3d debug output.
by Alexandre Julliard (＠julliard) 12 Oct '22

12 Oct '22

-- v2: wined3d: Use a Wine debug channel for vkd3d debug output. https://gitlab.winehq.org/wine/wine/-/merge_requests/962

5 8

[PATCH v3 0/5] MR25: vkd3d-shader/hlsl: Implement asuint() and completely move init_node() to hlsl.c.
by Zebediah Figura (＠zfigura) 12 Oct '22

12 Oct '22

-- v3: vkd3d-shader/hlsl: Pass a location pointer to init_node(). vkd3d-shader/hlsl: Introduce a hlsl_new_expr() helper. vkd3d-shader/hlsl: Write SM4 reinterpret instructions. vkd3d-shader/hlsl: Parse the asuint() intrinsic. tests: Add a test for asuint(). https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/25

4 7

[PATCH 0/7] MR1034: msvcp140: Do not import from msvcp120.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

The practical aspect of this (besides that msvcp140 just should not probably load msvcp120) is that for an app using VC runtime 140 mspcp140.__ExceptionPtrCurrentException gets the thread data from msvcp120.dll while vcruntime140_1 sets those thread data from ucrtbase.dll which are different data. So the application calling mspcp140.__ExceptionPtrCurrentException in catch block gets NULL exception while it was technically set by vcruntime140_1 before calling the app's catch block handler. After this patchset all the current exception data should end up in ucrtbase for VC runtime 140. There is a bit more to it (not covered by this patchset) when there is a mix of native and builtin VC runtime 140 dlls. From the app behaviour and some tests it looks like with native DLLs msvcp140 and vcruntime140_1 handle current exception data through vcruntime140.dll and vcruntime140.dll does not forward that to ucrtbase. I am not sure if the latter (not forwarding vcruntime140->ucrtbase) is practically important but to avoid the similar problem when mixing native and builtin VC runtime DLLs we should also probably link msvcp140 and vcruntime140_1 to vcruntime140. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1034

4 9

[PATCH 0/2] MR1038: Sapi datakey
by Alistair Leslie-Hughes (＠alesliehughes) 12 Oct '22

12 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1038

3 3

[PATCH 0/1] MR1015: wpcap: use UNIX 98/BSD 4.4 sockaddr header
by Marc-Aurel Zent (＠marzent) 12 Oct '22

12 Oct '22

Interpreting the fist two bytes of the `sockaddr` struct as `sa_family` produces incorrect behaviour on macOS and (presumbly) BSD. This change is however still backwards compatible with Linux since `sa_family` is of effective type `uint8_t` and `sa_len` is currently disregarded. For reference https://www.ibm.com/docs/en/i/7.3?topic=family-af-inet-address and this is from the <sys/socket.h> on macOS ``` /* * [XSI] Structure used by kernel to store most addresses. */ struct sockaddr { __uint8_t sa_len; /* total length */ sa_family_t sa_family; /* [XSI] address family */ char sa_data[14]; /* [XSI] addr value (actually larger) */ }; ``` -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1015

5 12

Re: [PATCH 0/1] MR74: ntdll: Prevent double free (Coverity)
by Sven Baars (＠sbaars) 12 Oct '22

12 Oct '22

Sven Baars (@sbaars) commented about dlls/ntdll/unix/file.c: > if (size > 0x10000) size = 0x10000; > if ((tmpbuf = malloc( size ))) > { > + if (needs_close) close( fd ); > if (!server_get_unix_fd( handle, FILE_READ_DATA, &fd, &needs_close, NULL, NULL )) > { > int res = recv( fd, tmpbuf, size, MSG_PEEK ); > info->MessagesAvailable = (res > 0); > info->NextMessageSize = (res >= 0) ? res : MAILSLOT_NO_MESSAGE; > if (needs_close) close( fd ); I think you want to remove this one as well. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/74#note_10372

1 0

[PATCH 0/5] MR1001: winebuild: Reimplement dlltool for import lib and delay import lib generation.
by Rémi Bernon (＠rbernon) 12 Oct '22

12 Oct '22

This is slightly different from dlltool but I think it should be compatible. The transition is done by first replacing dlltool with its bugs, and fixing them in separate changes. I based the ARM / AARCH64 implementation on the existing code around, but I have no idea if it is correct, and dlltool also doesn't include any delay load implementation for ARM. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1001

3 7

[PATCH 0/2] MR1027: windows.devices.enumeration: Create stub DeviceAccessInformation class.
by Paul Gofman (＠gofman) 12 Oct '22

12 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1027

4 6

[PATCH 0/3] MR1036: qcap: Various video capture patches.
by Zebediah Figura (＠zfigura) 12 Oct '22

12 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1036

3 4

[PATCH 0/6] MR971: winegstreamer: Complete transition to using IWMSyncReader2 from the WM async reader.
by Rémi Bernon (＠rbernon) 11 Oct '22

11 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/971

4 9

[PATCH v2 0/5] MR937: winegstreamer: Small refactor of wg_format video formats and major types.
by Rémi Bernon (＠rbernon) 11 Oct '22

11 Oct '22

I'm going to need to add a few more compressed formats, WMV which is straightforward, but also AAC (so MPEG v2/v4 audio), and I think wg_format could be cleaned up a bit like this beforehand. Then, AAC could either be added as a separate `AUDIO_MPEG4` format, or, with an version field, to the (renamed) `AUDIO_MPEG` format, though it doesn't share any of the current `AUDIO_MPEG1` properties. FWIW GStreamer caps is `audio/mpeg` for both, with an `mpegversion` that can be 1, 2 or 4. Could be very well added as `AUDIO_AAC` too, I don't really mind. -- v2: winegstreamer: Reorder video format struct in wg_format. winegstreamer: Rename WG_MAJOR_TYPE_H264 to WG_MAJOR_TYPE_VIDEO_H264. winegstreamer: Rename WG_MAJOR_TYPE_WMA to WG_MAJOR_TYPE_AUDIO_WMA. winegstreamer: Rename WG_MAJOR_TYPE_MPEG1_AUDIO to WG_MAJOR_TYPE_AUDIO_MPEG1. winegstreamer: Move CINEPAK encoded format to a separate major type. https://gitlab.winehq.org/wine/wine/-/merge_requests/937

4 11

[PATCH 0/1] MR498: server: Allow creating named pipes using \Device\NamedPipe\ as RootDirectory. (#52105)
by Jinoh Kang (＠iamahuman) 11 Oct '22

11 Oct '22

On Windows, \Device\NamedPipe\ is the root directory of the named pipe file system (NPFS), and can be used as RootDirectory to skip its path when accessing the NPFS namespace. This introduces a regression that makes it possible to use \Device\NamedPipe (note the lack of trailing backslash) as RootDirectory as well, since Wine does not distinguish between \Device\NamedPipe (the device itself) and \Device\NamedPipe\ (the root directory of the filesystem). (Previous iteration of this patch did distinguish between the two, but the method in which it was implemented was deemed hacky.) -- https://gitlab.winehq.org/wine/wine/-/merge_requests/498

5 11

Re: [PATCH v3 0/2] MR897: oleaut32: dereference VT_RECORD|VT_BYREF in place.
by Kevin Puetz (＠puetzk) 11 Oct '22

11 Oct '22

Revereted in 02ce1008b7876d9f86dfc5844c4be17307b8e055 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/897#note_10350

1 0

[PATCH v2 0/2] MR920: kernel32: Map system heap anywhere if the standard address is unavailable.
by Brendan Shanks (＠bshanks) 11 Oct '22

11 Oct '22

`mscomctl.ocx` (part of the VB6 common controls) calls `HeapCreate(0x04000000,1,0)`, that flag is the old and undocumented `HEAP_SHARED`. `HEAP_CreateSystemHeap()` tries to map memory at `0x80000000`, this fails on Wow64 (and probably should for all 32-bit platforms for non-LAA EXEs). As a workaround, map the memory anywhere if it fails at `0x80000000`. I wasn't able to find much documentation about `HEAP_SHARED`, it dates back to Win9x and I guess Wine may have needed it for some native DLLs. It's not clear whether modern Windows really implements any shared heap functionality any more, every 'shared' heap created returns a different pointer (even in the same process). Maybe Wine should remove the functionality too? -- v2: kernel32: Remove shared heap functionality. ntdll: Remove shared heap functionality. https://gitlab.winehq.org/wine/wine/-/merge_requests/920

3 3

Re: [PATCH v5 0/1] MR1006: gdi32: Fix possible overflow - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1006

1 0

Re: [PATCH v2 0/1] MR994: gdi32/tests: Mark tests failing randomly on Windows as flaky. - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/994

1 0

Re: [PATCH v4 0/2] MR957: rpcrt4: MIDL_STUB_MESSAGE needs to keep a refcount on IRpcChannelBuffer. - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/957

1 0

Re: [PATCH v26 0/1] MR778: ntoskrnl: KeInsertQueueDpc stub - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/778

1 0

Re: [PATCH v2 0/1] MR1029: include: Add missing SP defines - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1029

1 0

Re: [PATCH v3 0/2] MR1026: shell32: Handle drive letters properly when parsing MyComputer. - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1026

1 0

Re: [PATCH v2 0/2] MR981: dxgi/tests: Mark tests failing randomly on Windows as flaky. - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/981

1 0

Re: [PATCH v2 0/3] MR956: Return focus window in swapchain desc from IDirect3DSwapChain9_GetPresentParameters() if hDeviceWindow is NULL. - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/956

1 0

Re: [PATCH v3 0/4] MR980: wintrust: Support WSS_GET_SECONDARY_SIG_COUNT in singature settings in WinVerifyTrust(). - approved
by Alexandre Julliard (＠julliard) 11 Oct '22

11 Oct '22

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/980

1 0

[PATCH v7 0/6] MR908: uiautomationcore: Add more nested node features.
by Connor McAdams (＠cmcadams) 11 Oct '22

11 Oct '22

Add the ability to disconnect nested node providers, and return element properties for nested nodes. -- v7: uiautomationcore: Add UIAutomationType_Element property support for nested node providers. uiautomationcore: Store provider thread nodes that have a runtime ID in an rbtree. uiautomationcore: Implement UiaDisconnectProvider. uiautomationcore: Track all HUIANODEs returned from the provider thread. uiautomationcore: Don't use nested node providers for same-thread HWNDs. uiautomationcore: Use a separate vtbl for nested node IWineUiaProviders. https://gitlab.winehq.org/wine/wine/-/merge_requests/908

4 12

[PATCH 0/3] MR1033: winhttp: Add a few websocket options.
by Paul Gofman (＠gofman) 11 Oct '22

11 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1033

3 4

[PATCH 0/2] MR1031: mshtml: Document element props improvements.
by Gabriel Ivăncescu (＠insn) 11 Oct '22

11 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1031

3 3

[PATCH v4 0/1] MR1006: gdi32: Fix possible overflow
by Mark Jansen (＠learn-more) 11 Oct '22

11 Oct '22

According to the documentation of ScriptShape function, the psva argument should have the number of elements indicated by cMaxGlyphs. -- v4: gdi32: Fix possible overflow https://gitlab.winehq.org/wine/wine/-/merge_requests/1006

4 5

[PATCH v3 0/1] MR1006: gdi32: Fix possible overflow
by Mark Jansen (＠learn-more) 11 Oct '22

11 Oct '22

According to the documentation of ScriptShape function, the psva argument should have the number of elements indicated by cMaxGlyphs. -- v3: gdi32: Fix possible overflow https://gitlab.winehq.org/wine/wine/-/merge_requests/1006

4 3

[PATCH v3 0/2] MR1016: ntoskrnl.exe: Implement IoCreateNotificationEvent
by Etaash Mathamsetty (＠etaash.mathamsetty) 11 Oct '22

11 Oct '22

xigncode3 seems to use it -- v3: ntoskrnl.exe: fix formatting in IoCreateSynchronizationEvent. ntoskrnl.exe: Implement IoCreateNotificationEvent. https://gitlab.winehq.org/wine/wine/-/merge_requests/1016

3 4

[PATCH v2 0/1] MR1006: gdi32: Fix possible overflow
by Mark Jansen (＠learn-more) 11 Oct '22

11 Oct '22

According to the documentation of ScriptShape function, the psva argument should have the number of elements indicated by cMaxGlyphs. -- v2: gdi32: Fix possible overflow https://gitlab.winehq.org/wine/wine/-/merge_requests/1006

3 2

[PATCH v3 0/2] MR957: rpcrt4: MIDL_STUB_MESSAGE needs to keep a refcount on IRpcChannelBuffer.
by Kevin Puetz (＠puetzk) 11 Oct '22

11 Oct '22

It's possible for a proxy to be released during the middle of an Invoke. A specific scenario where this happened was a single-shot event sink which, upon receiving the event it was waiting for, would immediately call DispEventUnadvise. This removed the proxy pointing to that sink from the connection point's list of subscribers and released it. For such cases all state used to complete the Invoke must be owned by the __proxy_frame kept on the stack; after calling NdrProxySendReceive (which pumps STA messages and permits reentrancy), anything inside *This could be use-after-free. MIDL_STUB_MESSAGE already had its own IRPCChannelBuffer *pRpcChannelBuffer, but lacked its own refcount. When this does crash, the exception is caught by RpcTryFinally, but still leads to leaks since NdrProxyFreeBuffer wasn't able to call IRPCChannelBuffer::FreeBuffer. To fix this, StdProxy_GetChannel() now calls AddRef() when setting __proxy_frame::_StubMsg::pRpcChannelBuffer and NdrProxyFreeBuffer() calls the corresponding Release(). -- v3: rpcrt4: Add a refcount owned by MIDL_STUB_MESSAGE. rpcrt4/tests: Add tests for lifetime of IRpcChannelBuffer. https://gitlab.winehq.org/wine/wine/-/merge_requests/957

4 5

[PATCH 0/1] MR1006: gdi32: Fix possible overflow
by Mark Jansen (＠learn-more) 11 Oct '22

11 Oct '22

According to the documentation of ScriptShape function, the psva argument should have the number of elements indicated by cMaxGlyphs. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1006

4 5

[PATCH 0/2] MR1030: Add stubs and tests for GetConsoleOriginalTitleA/W
by Hugh McMaster (＠hmc) 11 Oct '22

11 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1030

3 3

[PATCH 0/1] MR1029: include: Add missing SP defines
by Alistair Leslie-Hughes (＠alesliehughes) 11 Oct '22

11 Oct '22

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=49775 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1029

3 2

[PATCH v2 0/1] MR1011: user32: Correctly detect and assign an int resource
by Alistair Leslie-Hughes (＠alesliehughes) 11 Oct '22

11 Oct '22

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=53646 Signed-off-by: Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> -- v2: user32: Correctly detect and assign an int resource https://gitlab.winehq.org/wine/wine/-/merge_requests/1011

3 3

[PATCH v25 0/1] MR778: ntoskrnl: KeInsertQueueDpc stub
by Etaash Mathamsetty (＠etaash.mathamsetty) 11 Oct '22

11 Oct '22

this gets the YSDrv service to start successfully (with the MmAllocateContiguousMemorySpecifyCache semi-stub) this service is used by the nox android emulator -- v25: ntoskrnl.exe: KeInsertQueueDpc stub. https://gitlab.winehq.org/wine/wine/-/merge_requests/778

3 2

[PATCH v22 0/1] MR778: ntoskrnl: KeInsertQueueDpc stub
by Etaash Mathamsetty (＠etaash.mathamsetty) 11 Oct '22

11 Oct '22

this gets the YSDrv service to start successfully (with the MmAllocateContiguousMemorySpecifyCache semi-stub) this service is used by the nox android emulator -- v22: ntoskrnl: KeInsertQueueDpc stub https://gitlab.winehq.org/wine/wine/-/merge_requests/778

3 2

[PATCH 0/1] MR1028: nsi: Use the correct members for the error and discard counts.
by Huw Davies (＠huw) 11 Oct '22

11 Oct '22

This is instead of !984. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/1028

3 2

[PATCH v2 0/2] MR1026: shell32: Handle drive letters properly when parsing MyComputer.
by Gabriel Ivăncescu (＠insn) 11 Oct '22

11 Oct '22

-- v2: shell32: Handle drive letters properly when parsing MyComputer. shell32/tests: Test drive letters parsing for My Computer. https://gitlab.winehq.org/wine/wine/-/merge_requests/1026

3 3

[PATCH v3 0/6] MR1025: ntdll: Support memory address requirements in NtAllocateVirtualMemoryEx()
by Paul Gofman (＠gofman) 11 Oct '22

11 Oct '22

Various notes: - LowestStartingAddress is still unsupported; - Patch 5 ("wow64: Set HighestEndingAddress in wow64_NtAllocateVirtualMemoryEx() if it is absent") is fixing existing WOW issue: NtAllocateVirtualMemoryEx() called from wow64 currently does not constraing the allocation to 32 bit address space (wow64_NtAllocateVirtualMemory() passes zero_bits for that). - I initially thought of using a single inter process APC but added a different one due to zero_bits handling which is easier to convert in the target process. -- v3: ntdll: Support specified alignment in NtAllocateVirtualMemoryEx(). wow64: Set HighestEndingAddress in wow64_NtAllocateVirtualMemoryEx() if it is absent. wow64: Support MEM_ADDRESS_REQUIREMENTS in wow64_NtAllocateVirtualMemoryEx(). ntdll/tests: Add tests for memory address requiements. https://gitlab.winehq.org/wine/wine/-/merge_requests/1025

3 11

[PATCH 0/2] MR1026: shell32: Handle drive letters properly when parsing MyComputer.
by Gabriel Ivăncescu (＠insn) 11 Oct '22

11 Oct '22

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/1026

3 3

[PATCH v3 0/6] MR965: jscript/mshtml: Enumeration fixes.
by Gabriel Ivăncescu (＠insn) 11 Oct '22

11 Oct '22

-- v3: mshtml: Enumerate document elements with name. mshtml: Split the lookup for an element with name into a separate helper. mshtml: Do not enumerate internal dynamic props. mshtml: Fix enumerating first custom prop after builtins. jscript: Refill the props at end of enumeration in html mode and retry. jscript: Fill the builtin props for enumeration on prototypes as well. https://gitlab.winehq.org/wine/wine/-/merge_requests/965

4 13

[PATCH v2 0/4] MR980: wintrust: Support WSS_GET_SECONDARY_SIG_COUNT in singature settings in WinVerifyTrust().
by Paul Gofman (＠gofman) 11 Oct '22

11 Oct '22

Nvidia's Streamline library depends on that (https://github.com/NVIDIAGameWorks/Streamline/). -- v2: wintrust: Support WSS_GET_SECONDARY_SIG_COUNT in Softpub provider. wintrust: Load secondary signatures in SoftpubLoadSignature(). https://gitlab.winehq.org/wine/wine/-/merge_requests/980

3 5

Re: [PATCH v6 0/1] MR23: shell32: Sanitize Program Manager icon and group names
by Huw Davies (＠huw) 10 Oct '22

10 Oct '22

Huw Davies (@huw) commented about dlls/shell32/dde.c: > } > if (argc >= 2) > { > - len = lstrlenW(last_group) + 1 + lstrlenW(argv[1]) + 5; > - name = heap_alloc(len * sizeof(*name)); > - swprintf( name, len, L"%s/%s.lnk", last_group, argv[1] ); > + name = combine_path(last_group, argv[1], L".lnk", TRUE); > } > else > { > - const WCHAR *filename = PathFindFileNameW(argv[0]); > - len = PathFindExtensionW(filename) - filename; > - name = heap_alloc((lstrlenW(last_group) + 1 + len + 5) * sizeof(*name)); > - swprintf( name, lstrlenW(last_group) + 1 + len + 5, L"%s/%.*s.lnk", last_group, len, filename ); > + WCHAR *filename = PathFindFileNameW(argv[0]); > + *PathFindExtensionW(filename) = '\0'; You're modifying `argv[0]` here which is pretty ugly. More broadly, you might find things easier if you add the `combine_path()` helper without the `sanitize` param in a first commit and then turn on the sanitizing in a second commit. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/23#note_10235

1 0

Re: [PATCH v2 0/1] MR957: rpcrt4: MIDL_STUB_MESSAGE needs to keep a refcount on IRpcChannelBuffer.
by Kevin Puetz (＠puetzk) 10 Oct '22

10 Oct '22

On Mon Oct 10 21:33:57 2022 +0000, Huw Davies wrote: > This looks correct to me, however afaict, Windows also sets > `stubMessage.pRpcChannelBuffer` to `NULL` when it releases it, so it > would be good to have a test for that, as well as it added to the implementation. > It would be clearer to split this into two commits: the first commit > would contain just the tests with appropriate `todo_wine`s. A second > commit would then add the implementation changes and remove the > `todo_wine`s. That way it's clear what the new changes are fixing. Ok, easy to do -- https://gitlab.winehq.org/wine/wine/-/merge_requests/957#note_10232

1 0

[PATCH v5 0/1] MR745: ntdll: Improve stub of NtQueryEaFile.
by Joel Holdsworth (＠jhol) 10 Oct '22

10 Oct '22

Based on a patch by Qian Hong. -- v5: ntdll: Improve stub of NtQueryEaFile. https://gitlab.winehq.org/wine/wine/-/merge_requests/745

4 7

Re: [PATCH v2 0/1] MR957: rpcrt4: MIDL_STUB_MESSAGE needs to keep a refcount on IRpcChannelBuffer.
by Huw Davies (＠huw) 10 Oct '22

10 Oct '22

Huw Davies (@huw) commented about dlls/rpcrt4/tests/cstub.c: > + /* stubMessage doesn't add its own refcounts on proxy_if1 or proxy_buffer, > + * so it's possible these are freed out from under it. > + * E.g. an event sink might unadvise upon receiving the event it was waiting for; > + * this unadvise could be reentrant to Invoke because SendReceive pumps STA messages. > + * The source would then erase that conection point entry and Release the proxy. */ > + IRpcProxyBuffer_Disconnect(proxy_buffer); > + ok(test_chanbuf.RefCount == 1, "got %ld\n", test_chanbuf.RefCount); > + IRpcProxyBuffer_Release(proxy_buffer); > + refs = IUnknown_Release(proxy_if1); > + ok(refs == 0, "got %ld\n", refs); > + ok(test_chanbuf.RefCount == 1, "got %ld\n", test_chanbuf.RefCount); > + > + /* NdrProxyFreeBuffer must not dereference the the now-freed proxy_if1, > + * yet should still free the remaining reference on test_chanbuf */ > + NdrProxyFreeBuffer(proxy_if1, &stubMessage); > + ok(test_chanbuf.RefCount == 0, "got %ld\n", test_chanbuf.RefCount); This looks correct to me, however afaict, Windows also sets `stubMessage.pRpcChannelBuffer` to `NULL` when it releases it, so it would be good to have a test for that, as well as it added to the implementation. It would be clearer to split this into two commits: the first commit would contain just the tests with appropriate `todo_wine`s. A second commit would then add the implementation changes and remove the `todo_wine`s. That way it's clear what the new changes are fixing. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/957#note_10221

1 0

2025

2024

2023

2022

wine-gitlab