- wine-gitlab - winehq.org

Re: [PATCH v5 0/2] MR3301: Improve trace output
by Gabriel Ivăncescu (＠insn) 13 Jul '23

13 Jul '23

On Thu Jul 13 16:37:28 2023 +0000, Jeffrey Smith wrote: > Compilers can (and at least some do) treat an enumeration as a _signed_ > integer. For the comparison to work as desired you'll want to cast > `InformationClass` to an unsigned type. I think `size_t` would be appropriate. There's no need for a cast though. `ARRAY_SIZE` is already `size_t` (because it's implemented with `sizeof`), and a comparison between a signed integer and an unsigned integer with at least as many bits is always promoted to unsigned. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3301#note_38928

1 0

[PATCH 0/1] MR3325: shell32: Don't fall back to $HOME for missing user special folders.
by Tim Clem (＠tclem) 13 Jul '23

13 Jul '23

Fall through in create_link and allow SHGetFolderPathAndSubDirW to create an empty folder in place if the alternatives are not available. Creating a link back to $HOME makes a recursive directory structure, which results in crashes in programs that naively search %USERPROFILE% (e.g. Quicken). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3325

2 1

Re: [PATCH v5 0/2] MR3301: Improve trace output
by Jeffrey Smith (＠whydoubt) 13 Jul '23

13 Jul '23

Jeffrey Smith (@whydoubt) commented about dlls/secur32/secur32.c: > + "(unknown)", > + "(unknown)", > + "NameUniqueId", > + "NameCanonical", > + "NameUserPrincipal", > + "NameCanonicalEx", > + "NameServicePrincipal", > + "(unknown)", > + "NameDnsDomain", > + "NameGivenName", > + "NameSurname" > +}; > + > +static LPCSTR debugstr_NameFormat(IN EXTENDED_NAME_FORMAT NameFormat) > +{ > + if (NameFormat >= ARRAY_SIZE(string_NameFormat)) Same issue with comparison of an enumeration. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3301#note_38927

1 0

Re: [PATCH v5 0/2] MR3301: Improve trace output
by Jeffrey Smith (＠whydoubt) 13 Jul '23

13 Jul '23

Jeffrey Smith (@whydoubt) commented about dlls/advapi32/lsa.c: > + "PolicyReplicaSourceInformation", > + "PolicyDefaultQuotaInformation", > + "PolicyModificationInformation", > + "PolicyAuditFullSetInformation", > + "PolicyAuditFullQueryInformation", > + "PolicyDnsDomainInformation", > + "PolicyDnsDomainInformationInt", > + "PolicyLocalAccountDomainInformation", > + "PolicyMachineAccountInformation", > + "PolicyMachineAccountInformation2", > + "PolicyLastEntry" > +}; > + > +static LPCSTR debugstr_InformationClass(IN POLICY_INFORMATION_CLASS InformationClass) > +{ > + if (InformationClass >= ARRAY_SIZE(string_InformationClass)) Compilers can (and at least some do) treat an enumeration as a _signed_ integer. For the comparison to work as desired you'll want to cast `InformationClass` to an unsigned type. I think `size_t` would be appropriate. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3301#note_38926

1 0

Re: [PATCH v2 0/1] MR3315: ntdll: Fix NtContinue from within exception context on ARM64
by Jinoh Kang (＠iamahuman) 13 Jul '23

13 Jul '23

This leads to a regression: NtContinue() can no longer properly restore X16 and X17, since they are clobbered by `__wine_syscall_dispatcher_return`. The proper fix is at https://www.winehq.org/mailman3/hyperkitty/list/wine-devel@winehq.org/threa…, which is a superset of this patchset. I don't have an ARM64 build machine at the moment; can you rebase it and see if it works? Many thanks for your work! -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3315#note_38925

1 0

[PATCH 0/1] MR3324: ntdll: Fail loudly if RtlActivateActivationContextEx fails to allocate memory.
by Jinoh Kang (＠iamahuman) 13 Jul '23

13 Jul '23

Many built-in callers of ActivateActCtx() just assume that it will always succeed. If it ever fails, then DeactivateActCtx() will notice that the cookie is invalid and raise an exception anyway. Based on comment https://gitlab.winehq.org/wine/wine/-/merge_requests/2068#note_37340 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3324

3 2

[PATCH 0/1] MR3323: win32u: Don't clip captured mouse pointer only if it's done by the system.
by Gabriel Ivăncescu (＠insn) 13 Jul '23

13 Jul '23

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/3323

3 2

[PATCH v2 0/3] MR265: vkd3d-shader/hlsl: Add some more constant folding operations.
by Nikolay Sivov (＠nsivov) 13 Jul '23

13 Jul '23

-- v2: vkd3d-shader/hlsl: Use type width in fold_rcp(). vkd3d-shader/hlsl: Add constant folding for 'log2'. vkd3d-shader/hlsl: Add constant folding for 'sqrt'. https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/265

3 9

[PATCH 0/7] MR3262: dxgi: Introduce a worker thread for each D3D12 swapchain (second part).
by Giovanni Mascellani (＠giomasce) 13 Jul '23

13 Jul '23

This MR completes https://gitlab.winehq.org/wine/wine/-/merge_requests/3165. The architecture was changed slightly from what is explained there after some discussion: instead of adding special support for `vkQueueSubmit()` and `vkQueuePresentKHR()`, vkd3d is not changed. Instead, the behavior of `vkd3d_acquire_vk_queue()` is specified, with a recommendation on how to ensure that frames are presented in order (see https://gitlab.winehq.org/wine/vkd3d/-/commit/b2a1f6b5e4f59fbc7f91ada7e5656…). The recommended behavior is implemented in this MR, and it is known to fix presentation bugs on The Medium. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3262

4 11

[PATCH v5 0/1] MR3269: wininet: Handle http headers correctly when length is 0
by Hans Lehnert (＠hanslehnert) 13 Jul '23

13 Jul '23

In HttpSendRequestW and HttpSendRequestExW, if the header pointer is not null but the length parameter is 0, the header length should be derived from the string length instead. In HttpSendRequestA and HttpSendRequestExA, on the same scenario, the function should fail instead. -- v5: wininet: Handle http headers correctly when length is 0 https://gitlab.winehq.org/wine/wine/-/merge_requests/3269

3 3

2025

2024

2023

2022

wine-gitlab