`wcsncpy` and `GetLocaleInfoEx` take length in number of characters, but `size` and `ret` counts number of bytes.
Previous commit changed a call to `GetLocaleInfoW` which counts lenght in `TCHAR`s (aka bytes), to the current `GetLocaleInfoEx`, which is probably the source of this confusion.
--
v2: msvcrt: fix out-of-bound access in create_locinfo
https://gitlab.winehq.org/wine/wine/-/merge_requests/3358
It's a clean-room reimplementation that mimics Windows 10 program's output format.
It prints all the information that is available via KerbQueryTicketCacheMessage.
Also tested to work on Windows if dynamically linked + built with winegcc.
For further extension of the functonality, implementing
KerbQueryTicketCacheEx{,2,3}Message is required.
--
v11: klist: Add a program that lists Kerberos tickets.
https://gitlab.winehq.org/wine/wine/-/merge_requests/3347
It's a clean-room reimplementation that mimics Windows 10 program's output format.
It prints all the information that is available via KerbQueryTicketCacheMessage.
Also tested to work on Windows if dynamically linked + built with winegcc.
For further extension of the functonality, implementing
KerbQueryTicketCacheEx{,2,3}Message is required.
--
v10: klist: Add a program that lists Kerberos tickets.
https://gitlab.winehq.org/wine/wine/-/merge_requests/3347
Shader visibility is currently ignored, but we don't want to create Vulkan descriptor sets for CPU heaps.
--
v2: vkd3d: Do not create Vulkan descriptor sets for non-shader-visible heaps.
vkd3d: Return a null handle from GetGPUDescriptorHandleForHeapStart() for non-shader-visible heaps.
tests: Test GetGPUDescriptorHandleForHeapStart() on a non-shader-visible heap.
vkd3d: Enable Vulkan-backed heaps for each heap instead of per device.
https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/272
Sufficient for compiling a no-op pixel shader.
This should probably be rebased on top of !263 because it introduces vkd3d_spirv_get_type_id_for_data_type(), which 263 renders unnecessary.
!263 is not essential, but I think using two different type systems in the backend is not ideal.
--
v3: vkd3d-shader/dxil: Emit undefined constants.
vkd3d-shader/spirv: Introduce an undefined register type.
vkd3d-shader/dxil: Emit the shader instructions.
vkd3d-shader/spirv: Do not normalise Shader Model 6 shaders.
https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/278
Logically the change here is that the parent shell folder resolves
each child path in 'FORPARSING' mode, instead of resolving itself
and doing a bunch more string concatenation for assumed simple child
pidls.
This also removes hacks for change notifications in the delete path
by actually performing notifications for deletion and trash operations
within the shell op.
Background:
I noticed a comment in the ReactOS version which suggested the first
change, it makes sense to not query the children in this way here.
The notifications hack also seemed really off. I'm not sure why the shell
operation wasn't performing these notifications already.
Testing:
All tests pass, and loading up wine explorer was easy to test the deletion.
Items were deleted (or trashed) and the shell folder was notified.
Unfortunately, copy and paste seem very broken at the moment, as the
paste_pidls function only calls ISFHelper_CopyItems with one pidl at a time,
and it itself seems to fail to correctly resolve the from path.
I did some hacking so that it did work correctly if the 'from' folder was the
desktop, and it worked fine with these changes (but those hacks wouldn't pass
review).
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3360