April 2025 - Wine-gitlab - list.winehq.org

Re: [PATCH v25 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) 30 Apr '25

30 Apr '25

Ralf Habacker (@rhabacker) commented about server/sock.c: > + { > + sock->addr.un.sun_family = WS_AF_UNIX; > + memcpy(sock->addr.un.sun_path, params->addr.sa_data, sizeof(sock->addr.un.sun_path)); > + sock->addr_len = sizeof(sock->addr.un); > + } > + else > + sock->addr_len = sockaddr_from_unix( &bind_addr, &sock->addr.addr, sizeof(sock->addr) ); > } > > update_addr_usage( sock, &bind_addr, v6only ); > > if (get_reply_max_size() >= sock->addr_len) > set_reply_data( &sock->addr, sock->addr_len ); > + > + if (sock->family == WS_AF_UNIX && *params->addr.sa_data) > + fchdir(server_dir_fd); here also -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7650#note_102242

1 0

Re: [PATCH v25 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) 30 Apr '25

30 Apr '25

Ralf Habacker (@rhabacker) commented about server/sock.c: > + if (!(unix_path = mem_alloc( unix_path_len + 1 ))) > + return; > + > + memcpy( unix_path, (char *)(params + 1) + params->addr_len, unix_path_len ); > + unix_path[unix_path_len] = '\0'; > + > + base_name = strrchr(unix_path, '/'); > + if (base_name) > + { > + if (base_name != unix_path) > + (++base_name)[-1] = '\0'; > + } > + else > + base_name = unix_path; > + > + if (chdir( unix_path ) == -1) At https://gitlab.winehq.org/wine/wine/-/merge_requests/7519 it is pointed out that changing the working directory may not be a good idea, so it should be checked whether this can be avoided. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7650#note_102237

1 0

Re: [PATCH v25 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) 30 Apr '25

30 Apr '25

Ralf Habacker (@rhabacker) commented about server/sock.c: > return; > } > > + if (sock->family == WS_AF_UNIX && *addr->sa_data) > + fchdir(server_dir_fd); Here the working directory is also changed, which needs to be reviewed. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7650#note_102239

1 0

Re: [PATCH v25 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) 30 Apr '25

30 Apr '25

Ralf Habacker (@rhabacker) commented about server/sock.c: > + if (!(unix_path = mem_alloc( unix_path_len + 1 ))) > + return; > + > + memcpy( unix_path, (char *)(&params->addr) + sizeof(struct WS_sockaddr_un), unix_path_len ); > + unix_path[unix_path_len] = '\0'; > + > + base_name = strrchr(unix_path, '/'); > + if (base_name) > + { > + if (base_name != unix_path) > + (++base_name)[-1] = '\0'; > + } > + else > + base_name = unix_path; > + > + if (chdir( unix_path ) == -1) chdir() is also used here. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7650#note_102240

1 0

Re: [PATCH v25 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) 30 Apr '25

30 Apr '25

Ralf Habacker (@rhabacker) commented about server/sock.c: > if (errno == EADDRINUSE && sock->reuseaddr) > errno = EACCES; > > - set_error( sock_get_ntstatus( errno ) ); > - return; > + /* Windows' AF_UNIX implementation has an edge case allowing for a socket to bind to > + * an empty path. Linux doesn't, so it throws EINVAL. We check for this situation > + * here and avoid early-exiting if it's the case. */ > + if (!(errno == EINVAL && sock->family == WS_AF_UNIX && !*params->addr.sa_data)) > + { > + set_error( sock_get_ntstatus( errno ) ); > + if (sock->family == WS_AF_UNIX && *params->addr.sa_data) > + fchdir(server_dir_fd); This should be changed. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7650#note_102241

1 0

[PATCH 0/4] MR7939: ntdll: Implement ProcessBasePriority class in NtSetInformationProcess.
by Marc-Aurel Zent (＠mzent) 30 Apr '25

30 Apr '25

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/7939

2 4

[PATCH 0/1] MR7937: Draft: winewayland: Don't clip cursor coordinates to vscreen boundaries.
by Alexandros Frantzis (＠afrantzis) 30 Apr '25

30 Apr '25

Since the positions of win32 windows and their corresponding Wayland surfaces are not synchronized, there are cases where parts of a window are outside the vscreen boundaries, and thus inaccessible to input events, but still visible and accessible from a user (i.e., Wayland compositor) standpoint. Try to remedy this issue by instructing the Wine server to not clip the Wayland mouse event coordinates to vscreen boundaries. -- Relevant discussions: * https://gitlab.winehq.org/wine/wine/-/merge_requests/4014#note_47581 * https://gitlab.winehq.org/wine/wine/-/merge_requests/7919 @julliard @rbernon Do you think such an approach (or something along these lines) would have any future in upstream as (a possibly opt-in) workaround for the Wayland input issues? -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7937

3 2

[PATCH 0/1] MR7919: winewayland.drv: Move windows to the top-left of the screen
by Julian Orth (＠mahkoh) 30 Apr '25

30 Apr '25

When opening the Cheat Engine [1] settings window, the window is spawned at around 400x400, making the right and bottom sides of the window inaccessible due to clipping. This commit moves the window to 0x0 to ensure that all window contents on the monitor are accessible. [1]: https://github.com/cheat-engine/cheat-engine Signed-off-by: Julian Orth <ju.orth(a)gmail.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7919

3 6

[PATCH v2 0/2] MR7780: win32u: Avoid uneeded (and sometimes harmful) RedrawWindow() calls in expose_window_surface().
by Paul Gofman (＠gofman) 30 Apr '25

30 Apr '25

-- v2: win32u: Only inflate dirty rect when dpi conversion is performed in expose_window_surface(). win32u: Don't redraw window in expose_window_surface() if window has surface. https://gitlab.winehq.org/wine/wine/-/merge_requests/7780

3 4

[PATCH v4 0/5] MR7848: server: Use events as a sync for several objects.
by Rémi Bernon 30 Apr '25

30 Apr '25

Instead of https://gitlab.winehq.org/wine/wine/-/merge_requests/7815, for https://gitlab.winehq.org/wine/wine/-/merge_requests/7226, this only split the sync ops to a separate vtable and let objects delegate theirs to a separate object. This starts using a event-like interface for most objects, leaving the decision regarding if/how to split sync themselves / integrate inproc syncs for later. -- v4: server: Use an event sync for fd objects. server: Introduce a new event sync object. server: Redirect fd-based objects sync to the fd. server: Add an operation to retrieve an object sync. server: Move object grab/release out of (add|remove)_queue. https://gitlab.winehq.org/wine/wine/-/merge_requests/7848

1 6